diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,156821 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.44966375589681823,
+  "eval_steps": 500,
+  "global_step": 22400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.550151362374153,
+      "learning_rate": 0.0,
+      "loss": 0.7559,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.2983169633056395,
+      "learning_rate": 2e-05,
+      "loss": 0.6943,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.798603951571041,
+      "learning_rate": 2e-05,
+      "loss": 0.6973,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7491460463638027,
+      "learning_rate": 2e-05,
+      "loss": 0.5874,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6349292125919317,
+      "learning_rate": 2e-05,
+      "loss": 0.6094,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.754222799778281,
+      "learning_rate": 2e-05,
+      "loss": 0.7051,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.416359416902974,
+      "learning_rate": 2e-05,
+      "loss": 0.6611,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.198018996487873,
+      "learning_rate": 2e-05,
+      "loss": 0.6045,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.1040768427951475,
+      "learning_rate": 2e-05,
+      "loss": 0.501,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.29934121841981,
+      "learning_rate": 2e-05,
+      "loss": 0.5654,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.959602813199588,
+      "learning_rate": 2e-05,
+      "loss": 0.5088,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.36648275458843,
+      "learning_rate": 2e-05,
+      "loss": 0.4868,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.544409582857607,
+      "learning_rate": 2e-05,
+      "loss": 0.5278,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.136320238980888,
+      "learning_rate": 2e-05,
+      "loss": 0.4507,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.66853372819511,
+      "learning_rate": 2e-05,
+      "loss": 0.4521,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.741515562915936,
+      "learning_rate": 2e-05,
+      "loss": 0.3577,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.455827191063251,
+      "learning_rate": 2e-05,
+      "loss": 0.5386,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.5988573298945274,
+      "learning_rate": 2e-05,
+      "loss": 0.4824,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.207937684562415,
+      "learning_rate": 2e-05,
+      "loss": 0.374,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.921984668297045,
+      "learning_rate": 2e-05,
+      "loss": 0.3901,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.770817763568287,
+      "learning_rate": 2e-05,
+      "loss": 0.3005,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.652590753500778,
+      "learning_rate": 2e-05,
+      "loss": 0.3594,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.117244712822952,
+      "learning_rate": 2e-05,
+      "loss": 0.3682,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.159909840873931,
+      "learning_rate": 2e-05,
+      "loss": 0.3838,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.248089655865254,
+      "learning_rate": 2e-05,
+      "loss": 0.4111,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.745548124893142,
+      "learning_rate": 2e-05,
+      "loss": 0.3235,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.919532203730706,
+      "learning_rate": 2e-05,
+      "loss": 0.3203,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.23462885697629,
+      "learning_rate": 2e-05,
+      "loss": 0.4844,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.140704064797506,
+      "learning_rate": 2e-05,
+      "loss": 0.3218,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.383148227182253,
+      "learning_rate": 2e-05,
+      "loss": 0.2878,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.336910375369747,
+      "learning_rate": 2e-05,
+      "loss": 0.3667,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.451898559867859,
+      "learning_rate": 2e-05,
+      "loss": 0.4038,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.404377976401506,
+      "learning_rate": 2e-05,
+      "loss": 0.4058,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.887671410562186,
+      "learning_rate": 2e-05,
+      "loss": 0.3662,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.981691384872207,
+      "learning_rate": 2e-05,
+      "loss": 0.2756,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.36566181234587,
+      "learning_rate": 2e-05,
+      "loss": 0.354,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.061749353309699,
+      "learning_rate": 2e-05,
+      "loss": 0.2852,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.97932259281819,
+      "learning_rate": 2e-05,
+      "loss": 0.3281,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.478472657040973,
+      "learning_rate": 2e-05,
+      "loss": 0.3928,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.012578020891779,
+      "learning_rate": 2e-05,
+      "loss": 0.3826,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.986395928346488,
+      "learning_rate": 2e-05,
+      "loss": 0.3335,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.636805543608028,
+      "learning_rate": 2e-05,
+      "loss": 0.2766,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.926753902399091,
+      "learning_rate": 2e-05,
+      "loss": 0.3188,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.3751836346151896,
+      "learning_rate": 2e-05,
+      "loss": 0.3096,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9935341137990252,
+      "learning_rate": 2e-05,
+      "loss": 0.2505,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.210431990918068,
+      "learning_rate": 2e-05,
+      "loss": 0.3359,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.463184300584972,
+      "learning_rate": 2e-05,
+      "loss": 0.3613,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.724794364220003,
+      "learning_rate": 2e-05,
+      "loss": 0.2297,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.427114983269285,
+      "learning_rate": 2e-05,
+      "loss": 0.3311,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.8961430659140674,
+      "learning_rate": 2e-05,
+      "loss": 0.2715,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.753994918159333,
+      "learning_rate": 2e-05,
+      "loss": 0.281,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.280295902987638,
+      "learning_rate": 2e-05,
+      "loss": 0.3174,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.1385917004862955,
+      "learning_rate": 2e-05,
+      "loss": 0.2954,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.46514856823037,
+      "learning_rate": 2e-05,
+      "loss": 0.2532,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.615967096405438,
+      "learning_rate": 2e-05,
+      "loss": 0.3677,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.017464968203362,
+      "learning_rate": 2e-05,
+      "loss": 0.3225,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.075570009602108,
+      "learning_rate": 2e-05,
+      "loss": 0.3716,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.964061838995337,
+      "learning_rate": 2e-05,
+      "loss": 0.3577,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.535447894070176,
+      "learning_rate": 2e-05,
+      "loss": 0.3013,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.9703840726227595,
+      "learning_rate": 2e-05,
+      "loss": 0.3247,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.78787816122196,
+      "learning_rate": 2e-05,
+      "loss": 0.2905,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.05457203186313,
+      "learning_rate": 2e-05,
+      "loss": 0.3157,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.9751787169609445,
+      "learning_rate": 2e-05,
+      "loss": 0.3362,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.441273733646192,
+      "learning_rate": 2e-05,
+      "loss": 0.2905,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.035396499044031,
+      "learning_rate": 2e-05,
+      "loss": 0.2832,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.207866507515364,
+      "learning_rate": 2e-05,
+      "loss": 0.3157,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.572895856688707,
+      "learning_rate": 2e-05,
+      "loss": 0.3723,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.766601501104143,
+      "learning_rate": 2e-05,
+      "loss": 0.3809,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.4902553931138955,
+      "learning_rate": 2e-05,
+      "loss": 0.3579,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.338678284424429,
+      "learning_rate": 2e-05,
+      "loss": 0.323,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.282810138745868,
+      "learning_rate": 2e-05,
+      "loss": 0.3726,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.139185059357848,
+      "learning_rate": 2e-05,
+      "loss": 0.4023,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.302091583858356,
+      "learning_rate": 2e-05,
+      "loss": 0.4077,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.499015137277394,
+      "learning_rate": 2e-05,
+      "loss": 0.3315,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2041590068379993,
+      "learning_rate": 2e-05,
+      "loss": 0.3464,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.630925009297282,
+      "learning_rate": 2e-05,
+      "loss": 0.366,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.325690180223073,
+      "learning_rate": 2e-05,
+      "loss": 0.2402,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.660499125148761,
+      "learning_rate": 2e-05,
+      "loss": 0.3022,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.712611165240665,
+      "learning_rate": 2e-05,
+      "loss": 0.3483,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.512947685098363,
+      "learning_rate": 2e-05,
+      "loss": 0.3064,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.690261950654657,
+      "learning_rate": 2e-05,
+      "loss": 0.2524,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.365316104545883,
+      "learning_rate": 2e-05,
+      "loss": 0.3247,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.1105368989842175,
+      "learning_rate": 2e-05,
+      "loss": 0.2483,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.92673596507897,
+      "learning_rate": 2e-05,
+      "loss": 0.3191,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.457115159615562,
+      "learning_rate": 2e-05,
+      "loss": 0.3647,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.21145259243963,
+      "learning_rate": 2e-05,
+      "loss": 0.4531,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.309851062304625,
+      "learning_rate": 2e-05,
+      "loss": 0.3516,
+      "step": 87
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.1715279397079374,
+      "learning_rate": 2e-05,
+      "loss": 0.2708,
+      "step": 88
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.440788232754837,
+      "learning_rate": 2e-05,
+      "loss": 0.3335,
+      "step": 89
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.012051071769929,
+      "learning_rate": 2e-05,
+      "loss": 0.2693,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.855002274614604,
+      "learning_rate": 2e-05,
+      "loss": 0.2097,
+      "step": 91
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.067392793454087,
+      "learning_rate": 2e-05,
+      "loss": 0.272,
+      "step": 92
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.9031138510543855,
+      "learning_rate": 2e-05,
+      "loss": 0.2419,
+      "step": 93
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.37059201059938,
+      "learning_rate": 2e-05,
+      "loss": 0.2258,
+      "step": 94
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.242396231806007,
+      "learning_rate": 2e-05,
+      "loss": 0.3274,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.453422927444687,
+      "learning_rate": 2e-05,
+      "loss": 0.2979,
+      "step": 96
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.91218577897065,
+      "learning_rate": 2e-05,
+      "loss": 0.2424,
+      "step": 97
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.8396630549778745,
+      "learning_rate": 2e-05,
+      "loss": 0.2468,
+      "step": 98
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.373215665707451,
+      "learning_rate": 2e-05,
+      "loss": 0.1978,
+      "step": 99
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.680895819717507,
+      "learning_rate": 2e-05,
+      "loss": 0.2073,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.12381405929446,
+      "learning_rate": 2e-05,
+      "loss": 0.239,
+      "step": 101
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.630056736886395,
+      "learning_rate": 2e-05,
+      "loss": 0.1736,
+      "step": 102
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.82687176277245,
+      "learning_rate": 2e-05,
+      "loss": 0.2318,
+      "step": 103
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.238003239092093,
+      "learning_rate": 2e-05,
+      "loss": 0.1918,
+      "step": 104
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.778642687444174,
+      "learning_rate": 2e-05,
+      "loss": 0.2104,
+      "step": 105
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.7694636986619,
+      "learning_rate": 2e-05,
+      "loss": 0.2795,
+      "step": 106
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.890903463867327,
+      "learning_rate": 2e-05,
+      "loss": 0.2612,
+      "step": 107
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.995995372345286,
+      "learning_rate": 2e-05,
+      "loss": 0.1482,
+      "step": 108
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.900928286028787,
+      "learning_rate": 2e-05,
+      "loss": 0.1642,
+      "step": 109
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.297227048218883,
+      "learning_rate": 2e-05,
+      "loss": 0.2177,
+      "step": 110
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.617242835412766,
+      "learning_rate": 2e-05,
+      "loss": 0.1929,
+      "step": 111
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.80443187851541,
+      "learning_rate": 2e-05,
+      "loss": 0.1886,
+      "step": 112
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.665382387635718,
+      "learning_rate": 2e-05,
+      "loss": 0.2896,
+      "step": 113
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.0181550017038,
+      "learning_rate": 2e-05,
+      "loss": 0.1855,
+      "step": 114
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.3630691119621625,
+      "learning_rate": 2e-05,
+      "loss": 0.2031,
+      "step": 115
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.3331333262668235,
+      "learning_rate": 2e-05,
+      "loss": 0.1832,
+      "step": 116
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.717535237405416,
+      "learning_rate": 2e-05,
+      "loss": 0.2959,
+      "step": 117
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.647346614064125,
+      "learning_rate": 2e-05,
+      "loss": 0.2888,
+      "step": 118
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.724698224325484,
+      "learning_rate": 2e-05,
+      "loss": 0.1908,
+      "step": 119
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.33364863090189,
+      "learning_rate": 2e-05,
+      "loss": 0.2307,
+      "step": 120
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.72600378927761,
+      "learning_rate": 2e-05,
+      "loss": 0.2866,
+      "step": 121
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.603210972787514,
+      "learning_rate": 2e-05,
+      "loss": 0.2854,
+      "step": 122
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.08426871576725,
+      "learning_rate": 2e-05,
+      "loss": 0.2009,
+      "step": 123
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.712369065123873,
+      "learning_rate": 2e-05,
+      "loss": 0.2092,
+      "step": 124
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.014136570684657,
+      "learning_rate": 2e-05,
+      "loss": 0.323,
+      "step": 125
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.892060203664209,
+      "learning_rate": 2e-05,
+      "loss": 0.1862,
+      "step": 126
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.849826864711336,
+      "learning_rate": 2e-05,
+      "loss": 0.1664,
+      "step": 127
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.204220768692849,
+      "learning_rate": 2e-05,
+      "loss": 0.1544,
+      "step": 128
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.4503030115288515,
+      "learning_rate": 2e-05,
+      "loss": 0.2578,
+      "step": 129
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.782808093163351,
+      "learning_rate": 2e-05,
+      "loss": 0.1477,
+      "step": 130
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.46710633781568,
+      "learning_rate": 2e-05,
+      "loss": 0.207,
+      "step": 131
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.846979651929211,
+      "learning_rate": 2e-05,
+      "loss": 0.1582,
+      "step": 132
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.44668960358668,
+      "learning_rate": 2e-05,
+      "loss": 0.2224,
+      "step": 133
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.060021482969452,
+      "learning_rate": 2e-05,
+      "loss": 0.2153,
+      "step": 134
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.91510018764092,
+      "learning_rate": 2e-05,
+      "loss": 0.1384,
+      "step": 135
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.642288274495034,
+      "learning_rate": 2e-05,
+      "loss": 0.1682,
+      "step": 136
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.73974813774058,
+      "learning_rate": 2e-05,
+      "loss": 0.1548,
+      "step": 137
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.151971436080729,
+      "learning_rate": 2e-05,
+      "loss": 0.1147,
+      "step": 138
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.431987565909163,
+      "learning_rate": 2e-05,
+      "loss": 0.2001,
+      "step": 139
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.096740552761142,
+      "learning_rate": 2e-05,
+      "loss": 0.2463,
+      "step": 140
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 36.18594826120355,
+      "learning_rate": 2e-05,
+      "loss": 0.259,
+      "step": 141
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.493139459637296,
+      "learning_rate": 2e-05,
+      "loss": 0.2274,
+      "step": 142
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.05415443601996,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 143
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.992861565879151,
+      "learning_rate": 2e-05,
+      "loss": 0.2104,
+      "step": 144
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.925374137704765,
+      "learning_rate": 2e-05,
+      "loss": 0.1912,
+      "step": 145
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.77869222895987,
+      "learning_rate": 2e-05,
+      "loss": 0.2334,
+      "step": 146
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.797823880875427,
+      "learning_rate": 2e-05,
+      "loss": 0.2164,
+      "step": 147
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.934094613883614,
+      "learning_rate": 2e-05,
+      "loss": 0.192,
+      "step": 148
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.483392535629495,
+      "learning_rate": 2e-05,
+      "loss": 0.2349,
+      "step": 149
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.031968154357669,
+      "learning_rate": 2e-05,
+      "loss": 0.1758,
+      "step": 150
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.2213598554438025,
+      "learning_rate": 2e-05,
+      "loss": 0.1086,
+      "step": 151
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.80356447008165,
+      "learning_rate": 2e-05,
+      "loss": 0.1571,
+      "step": 152
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.975656414250057,
+      "learning_rate": 2e-05,
+      "loss": 0.193,
+      "step": 153
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.679106677050305,
+      "learning_rate": 2e-05,
+      "loss": 0.1628,
+      "step": 154
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.883938421989115,
+      "learning_rate": 2e-05,
+      "loss": 0.1298,
+      "step": 155
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.211123432308033,
+      "learning_rate": 2e-05,
+      "loss": 0.1196,
+      "step": 156
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.335537494729656,
+      "learning_rate": 2e-05,
+      "loss": 0.0874,
+      "step": 157
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.905580122586046,
+      "learning_rate": 2e-05,
+      "loss": 0.158,
+      "step": 158
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.012861684939938,
+      "learning_rate": 2e-05,
+      "loss": 0.1125,
+      "step": 159
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.381055986709665,
+      "learning_rate": 2e-05,
+      "loss": 0.1486,
+      "step": 160
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.022135980766441,
+      "learning_rate": 2e-05,
+      "loss": 0.1667,
+      "step": 161
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.818607368782739,
+      "learning_rate": 2e-05,
+      "loss": 0.1195,
+      "step": 162
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.714055703847305,
+      "learning_rate": 2e-05,
+      "loss": 0.1093,
+      "step": 163
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.302174146749227,
+      "learning_rate": 2e-05,
+      "loss": 0.148,
+      "step": 164
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.658546613775788,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 165
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.40597072085604,
+      "learning_rate": 2e-05,
+      "loss": 0.1482,
+      "step": 166
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.03144165996811,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 167
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.78121178661706,
+      "learning_rate": 2e-05,
+      "loss": 0.2369,
+      "step": 168
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.814941491269868,
+      "learning_rate": 2e-05,
+      "loss": 0.1445,
+      "step": 169
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.931179132307188,
+      "learning_rate": 2e-05,
+      "loss": 0.2134,
+      "step": 170
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.478616648219916,
+      "learning_rate": 2e-05,
+      "loss": 0.1245,
+      "step": 171
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.973042607232702,
+      "learning_rate": 2e-05,
+      "loss": 0.1925,
+      "step": 172
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.495565502166306,
+      "learning_rate": 2e-05,
+      "loss": 0.2483,
+      "step": 173
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.459087795308367,
+      "learning_rate": 2e-05,
+      "loss": 0.1912,
+      "step": 174
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.772806796466185,
+      "learning_rate": 2e-05,
+      "loss": 0.1392,
+      "step": 175
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.960761984776582,
+      "learning_rate": 2e-05,
+      "loss": 0.1118,
+      "step": 176
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.1182194761307365,
+      "learning_rate": 2e-05,
+      "loss": 0.1741,
+      "step": 177
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.7395434479205445,
+      "learning_rate": 2e-05,
+      "loss": 0.1892,
+      "step": 178
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.407225155972798,
+      "learning_rate": 2e-05,
+      "loss": 0.2319,
+      "step": 179
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.636663010735422,
+      "learning_rate": 2e-05,
+      "loss": 0.1814,
+      "step": 180
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.703626637621504,
+      "learning_rate": 2e-05,
+      "loss": 0.121,
+      "step": 181
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.17619801797331,
+      "learning_rate": 2e-05,
+      "loss": 0.1857,
+      "step": 182
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.984649235444594,
+      "learning_rate": 2e-05,
+      "loss": 0.1645,
+      "step": 183
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.331225444949785,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 184
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.975139615030733,
+      "learning_rate": 2e-05,
+      "loss": 0.1548,
+      "step": 185
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.494727123422111,
+      "learning_rate": 2e-05,
+      "loss": 0.1588,
+      "step": 186
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.78971481837293,
+      "learning_rate": 2e-05,
+      "loss": 0.1992,
+      "step": 187
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.102553522204003,
+      "learning_rate": 2e-05,
+      "loss": 0.2009,
+      "step": 188
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.385761214227544,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 189
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.63577478363743,
+      "learning_rate": 2e-05,
+      "loss": 0.1763,
+      "step": 190
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.381971006524078,
+      "learning_rate": 2e-05,
+      "loss": 0.249,
+      "step": 191
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.61174119947972,
+      "learning_rate": 2e-05,
+      "loss": 0.2454,
+      "step": 192
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.225062450948974,
+      "learning_rate": 2e-05,
+      "loss": 0.1475,
+      "step": 193
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.242602433604351,
+      "learning_rate": 2e-05,
+      "loss": 0.1398,
+      "step": 194
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.81258590101868,
+      "learning_rate": 2e-05,
+      "loss": 0.2532,
+      "step": 195
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.55491626680855,
+      "learning_rate": 2e-05,
+      "loss": 0.2185,
+      "step": 196
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.919082540681894,
+      "learning_rate": 2e-05,
+      "loss": 0.2603,
+      "step": 197
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.716486581944607,
+      "learning_rate": 2e-05,
+      "loss": 0.1353,
+      "step": 198
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.326442854374758,
+      "learning_rate": 2e-05,
+      "loss": 0.168,
+      "step": 199
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.242948064761851,
+      "learning_rate": 2e-05,
+      "loss": 0.1928,
+      "step": 200
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.891944457673156,
+      "learning_rate": 2e-05,
+      "loss": 0.2532,
+      "step": 201
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.425518203329034,
+      "learning_rate": 2e-05,
+      "loss": 0.1841,
+      "step": 202
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.443820850119709,
+      "learning_rate": 2e-05,
+      "loss": 0.0972,
+      "step": 203
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.1362198809601205,
+      "learning_rate": 2e-05,
+      "loss": 0.1226,
+      "step": 204
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.088538179543043,
+      "learning_rate": 2e-05,
+      "loss": 0.2417,
+      "step": 205
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.424160733854988,
+      "learning_rate": 2e-05,
+      "loss": 0.2054,
+      "step": 206
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7471232506196634,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 207
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.238516817352112,
+      "learning_rate": 2e-05,
+      "loss": 0.2444,
+      "step": 208
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.170416922691095,
+      "learning_rate": 2e-05,
+      "loss": 0.1149,
+      "step": 209
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.7413277949123245,
+      "learning_rate": 2e-05,
+      "loss": 0.229,
+      "step": 210
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.615047041712086,
+      "learning_rate": 2e-05,
+      "loss": 0.1026,
+      "step": 211
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.429344742581324,
+      "learning_rate": 2e-05,
+      "loss": 0.1466,
+      "step": 212
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.889080948166912,
+      "learning_rate": 2e-05,
+      "loss": 0.1593,
+      "step": 213
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.5042971565146175,
+      "learning_rate": 2e-05,
+      "loss": 0.1022,
+      "step": 214
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.809641397793209,
+      "learning_rate": 2e-05,
+      "loss": 0.1266,
+      "step": 215
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.1136055987830025,
+      "learning_rate": 2e-05,
+      "loss": 0.162,
+      "step": 216
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.429696992702592,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 217
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.093654279021035,
+      "learning_rate": 2e-05,
+      "loss": 0.1792,
+      "step": 218
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.536086559523004,
+      "learning_rate": 2e-05,
+      "loss": 0.1258,
+      "step": 219
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.8029471798648595,
+      "learning_rate": 2e-05,
+      "loss": 0.0881,
+      "step": 220
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.08496722035453,
+      "learning_rate": 2e-05,
+      "loss": 0.1069,
+      "step": 221
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.994550773581852,
+      "learning_rate": 2e-05,
+      "loss": 0.1267,
+      "step": 222
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.249310532454789,
+      "learning_rate": 2e-05,
+      "loss": 0.1042,
+      "step": 223
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.719132874392843,
+      "learning_rate": 2e-05,
+      "loss": 0.1809,
+      "step": 224
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.332887193775357,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 225
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.593037298048571,
+      "learning_rate": 2e-05,
+      "loss": 0.1883,
+      "step": 226
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.937594588139659,
+      "learning_rate": 2e-05,
+      "loss": 0.0985,
+      "step": 227
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.050408235057212,
+      "learning_rate": 2e-05,
+      "loss": 0.1624,
+      "step": 228
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.182418752038806,
+      "learning_rate": 2e-05,
+      "loss": 0.1733,
+      "step": 229
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.741371062724204,
+      "learning_rate": 2e-05,
+      "loss": 0.1572,
+      "step": 230
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.081503079307588,
+      "learning_rate": 2e-05,
+      "loss": 0.1538,
+      "step": 231
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.632253742384194,
+      "learning_rate": 2e-05,
+      "loss": 0.1739,
+      "step": 232
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.327156157156397,
+      "learning_rate": 2e-05,
+      "loss": 0.1985,
+      "step": 233
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.807736546573434,
+      "learning_rate": 2e-05,
+      "loss": 0.1324,
+      "step": 234
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.958821452316384,
+      "learning_rate": 2e-05,
+      "loss": 0.1718,
+      "step": 235
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.746289255730803,
+      "learning_rate": 2e-05,
+      "loss": 0.2681,
+      "step": 236
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.218713513002635,
+      "learning_rate": 2e-05,
+      "loss": 0.2296,
+      "step": 237
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.531466101762801,
+      "learning_rate": 2e-05,
+      "loss": 0.182,
+      "step": 238
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.820963365828266,
+      "learning_rate": 2e-05,
+      "loss": 0.1556,
+      "step": 239
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.305511006184339,
+      "learning_rate": 2e-05,
+      "loss": 0.1399,
+      "step": 240
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.230936042361105,
+      "learning_rate": 2e-05,
+      "loss": 0.1594,
+      "step": 241
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.551035442538672,
+      "learning_rate": 2e-05,
+      "loss": 0.1324,
+      "step": 242
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.275840292992903,
+      "learning_rate": 2e-05,
+      "loss": 0.165,
+      "step": 243
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.366628570360865,
+      "learning_rate": 2e-05,
+      "loss": 0.1851,
+      "step": 244
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.765971914498753,
+      "learning_rate": 2e-05,
+      "loss": 0.2258,
+      "step": 245
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.224771476747945,
+      "learning_rate": 2e-05,
+      "loss": 0.1039,
+      "step": 246
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.166502804340358,
+      "learning_rate": 2e-05,
+      "loss": 0.1638,
+      "step": 247
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.082943082586388,
+      "learning_rate": 2e-05,
+      "loss": 0.1627,
+      "step": 248
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.444507525489729,
+      "learning_rate": 2e-05,
+      "loss": 0.1548,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.332116206683529,
+      "learning_rate": 2e-05,
+      "loss": 0.1604,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.452903768833019,
+      "learning_rate": 2e-05,
+      "loss": 0.136,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.572777522715514,
+      "learning_rate": 2e-05,
+      "loss": 0.2498,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.671329507242742,
+      "learning_rate": 2e-05,
+      "loss": 0.3381,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.818895331601707,
+      "learning_rate": 2e-05,
+      "loss": 0.2065,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8112115402508184,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.164037307176371,
+      "learning_rate": 2e-05,
+      "loss": 0.11,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.888714791114978,
+      "learning_rate": 2e-05,
+      "loss": 0.1888,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.46832534717035,
+      "learning_rate": 2e-05,
+      "loss": 0.2515,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.764076252334333,
+      "learning_rate": 2e-05,
+      "loss": 0.1227,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.006304771323746,
+      "learning_rate": 2e-05,
+      "loss": 0.0897,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.874256492805002,
+      "learning_rate": 2e-05,
+      "loss": 0.1232,
+      "step": 261
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.29258041564613,
+      "learning_rate": 2e-05,
+      "loss": 0.1819,
+      "step": 262
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.18778028083787,
+      "learning_rate": 2e-05,
+      "loss": 0.1447,
+      "step": 263
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.424605018051281,
+      "learning_rate": 2e-05,
+      "loss": 0.189,
+      "step": 264
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.762602223050235,
+      "learning_rate": 2e-05,
+      "loss": 0.28,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.664811628490145,
+      "learning_rate": 2e-05,
+      "loss": 0.1151,
+      "step": 266
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.869087275871862,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 267
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.68634850401463,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 268
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.735412685371915,
+      "learning_rate": 2e-05,
+      "loss": 0.1084,
+      "step": 269
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.175204612646988,
+      "learning_rate": 2e-05,
+      "loss": 0.12,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.193143486559729,
+      "learning_rate": 2e-05,
+      "loss": 0.1396,
+      "step": 271
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.183440104992606,
+      "learning_rate": 2e-05,
+      "loss": 0.1537,
+      "step": 272
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.392397000753412,
+      "learning_rate": 2e-05,
+      "loss": 0.1116,
+      "step": 273
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.183153652662574,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 274
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.2099584035662785,
+      "learning_rate": 2e-05,
+      "loss": 0.1318,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.477609449200893,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 276
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.876016478019089,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 277
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.659634586475738,
+      "learning_rate": 2e-05,
+      "loss": 0.1061,
+      "step": 278
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.42397739548366,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 279
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.753143426321894,
+      "learning_rate": 2e-05,
+      "loss": 0.1356,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.3223028955117035,
+      "learning_rate": 2e-05,
+      "loss": 0.1327,
+      "step": 281
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.753846265101782,
+      "learning_rate": 2e-05,
+      "loss": 0.1693,
+      "step": 282
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.26561358985457,
+      "learning_rate": 2e-05,
+      "loss": 0.1111,
+      "step": 283
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.345333626624851,
+      "learning_rate": 2e-05,
+      "loss": 0.1156,
+      "step": 284
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.022980162737962,
+      "learning_rate": 2e-05,
+      "loss": 0.0984,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.844613077445039,
+      "learning_rate": 2e-05,
+      "loss": 0.1222,
+      "step": 286
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.920785934818603,
+      "learning_rate": 2e-05,
+      "loss": 0.1369,
+      "step": 287
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.549219110571748,
+      "learning_rate": 2e-05,
+      "loss": 0.1223,
+      "step": 288
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.822531777123029,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 289
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.704218850613249,
+      "learning_rate": 2e-05,
+      "loss": 0.1441,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9229013201056233,
+      "learning_rate": 2e-05,
+      "loss": 0.1058,
+      "step": 291
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.413423574819824,
+      "learning_rate": 2e-05,
+      "loss": 0.1167,
+      "step": 292
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.576141865298528,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 293
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.075299332747137,
+      "learning_rate": 2e-05,
+      "loss": 0.1204,
+      "step": 294
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8776889333998747,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 295
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.236163395871765,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 296
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1461345873622246,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 297
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.395646559858881,
+      "learning_rate": 2e-05,
+      "loss": 0.0918,
+      "step": 298
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.62110525458,
+      "learning_rate": 2e-05,
+      "loss": 0.2312,
+      "step": 299
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.028159193227331,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.503371127451238,
+      "learning_rate": 2e-05,
+      "loss": 0.0953,
+      "step": 301
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.008703083429108,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 302
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.029896245546198,
+      "learning_rate": 2e-05,
+      "loss": 0.1077,
+      "step": 303
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.862077810506189,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 304
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.054862106615762,
+      "learning_rate": 2e-05,
+      "loss": 0.1491,
+      "step": 305
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.371831607627782,
+      "learning_rate": 2e-05,
+      "loss": 0.0883,
+      "step": 306
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.6519570261191525,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 307
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.79559014173224,
+      "learning_rate": 2e-05,
+      "loss": 0.1373,
+      "step": 308
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.845724066925435,
+      "learning_rate": 2e-05,
+      "loss": 0.2122,
+      "step": 309
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.493238614298864,
+      "learning_rate": 2e-05,
+      "loss": 0.2733,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.90827743791831,
+      "learning_rate": 2e-05,
+      "loss": 0.1658,
+      "step": 311
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.916157883243532,
+      "learning_rate": 2e-05,
+      "loss": 0.1714,
+      "step": 312
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.338437106930199,
+      "learning_rate": 2e-05,
+      "loss": 0.1518,
+      "step": 313
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.702560773138494,
+      "learning_rate": 2e-05,
+      "loss": 0.0971,
+      "step": 314
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.945443494353111,
+      "learning_rate": 2e-05,
+      "loss": 0.1599,
+      "step": 315
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.74588528896168,
+      "learning_rate": 2e-05,
+      "loss": 0.1337,
+      "step": 316
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.563433876106007,
+      "learning_rate": 2e-05,
+      "loss": 0.1221,
+      "step": 317
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.694360640407959,
+      "learning_rate": 2e-05,
+      "loss": 0.1901,
+      "step": 318
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.632328255486146,
+      "learning_rate": 2e-05,
+      "loss": 0.1061,
+      "step": 319
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.595418565263723,
+      "learning_rate": 2e-05,
+      "loss": 0.1653,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.697589487049359,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 321
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.620399461201636,
+      "learning_rate": 2e-05,
+      "loss": 0.102,
+      "step": 322
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.271102375667576,
+      "learning_rate": 2e-05,
+      "loss": 0.1298,
+      "step": 323
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.18245802606201,
+      "learning_rate": 2e-05,
+      "loss": 0.2074,
+      "step": 324
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.69925747073358,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 325
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.801584803350611,
+      "learning_rate": 2e-05,
+      "loss": 0.1022,
+      "step": 326
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6747191165802993,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 327
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.84334949732434,
+      "learning_rate": 2e-05,
+      "loss": 0.1078,
+      "step": 328
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.806136643123357,
+      "learning_rate": 2e-05,
+      "loss": 0.1347,
+      "step": 329
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.493051995384627,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 330
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.518983848817435,
+      "learning_rate": 2e-05,
+      "loss": 0.1523,
+      "step": 331
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.294484027686836,
+      "learning_rate": 2e-05,
+      "loss": 0.109,
+      "step": 332
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.84159938994908,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 333
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.743479463867663,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 334
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.2082480498404,
+      "learning_rate": 2e-05,
+      "loss": 0.153,
+      "step": 335
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.987853152506044,
+      "learning_rate": 2e-05,
+      "loss": 0.0983,
+      "step": 336
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.702321095894744,
+      "learning_rate": 2e-05,
+      "loss": 0.1029,
+      "step": 337
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.4500731291798035,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 338
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.580924904850369,
+      "learning_rate": 2e-05,
+      "loss": 0.0903,
+      "step": 339
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.056894748976421,
+      "learning_rate": 2e-05,
+      "loss": 0.1465,
+      "step": 340
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.595292475926869,
+      "learning_rate": 2e-05,
+      "loss": 0.1384,
+      "step": 341
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7874653074985876,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 342
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.331675891261951,
+      "learning_rate": 2e-05,
+      "loss": 0.1343,
+      "step": 343
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.812083567806809,
+      "learning_rate": 2e-05,
+      "loss": 0.1173,
+      "step": 344
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.392701984119196,
+      "learning_rate": 2e-05,
+      "loss": 0.1361,
+      "step": 345
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.606131439389639,
+      "learning_rate": 2e-05,
+      "loss": 0.1578,
+      "step": 346
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.966028312262296,
+      "learning_rate": 2e-05,
+      "loss": 0.1661,
+      "step": 347
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.174661699429209,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 348
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.187219635154748,
+      "learning_rate": 2e-05,
+      "loss": 0.1505,
+      "step": 349
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.4786406096483,
+      "learning_rate": 2e-05,
+      "loss": 0.2081,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.427551878961249,
+      "learning_rate": 2e-05,
+      "loss": 0.1133,
+      "step": 351
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.872217100084024,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 352
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.049831404597698,
+      "learning_rate": 2e-05,
+      "loss": 0.1202,
+      "step": 353
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.34568030190636,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 354
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.293342993460795,
+      "learning_rate": 2e-05,
+      "loss": 0.1521,
+      "step": 355
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.972680048541319,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 356
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.9748768238416705,
+      "learning_rate": 2e-05,
+      "loss": 0.1414,
+      "step": 357
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.311868232758482,
+      "learning_rate": 2e-05,
+      "loss": 0.184,
+      "step": 358
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.425167598181471,
+      "learning_rate": 2e-05,
+      "loss": 0.1616,
+      "step": 359
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.40115787761275,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 360
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.902612282421385,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 361
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.080933537340042,
+      "learning_rate": 2e-05,
+      "loss": 0.1323,
+      "step": 362
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.311898769689489,
+      "learning_rate": 2e-05,
+      "loss": 0.1222,
+      "step": 363
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.575398020289315,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 364
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.438707269729504,
+      "learning_rate": 2e-05,
+      "loss": 0.1448,
+      "step": 365
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.862899591707664,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 366
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.440378679159139,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 367
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.922112546970948,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 368
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.308497458195857,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 369
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.3857011947386,
+      "learning_rate": 2e-05,
+      "loss": 0.1102,
+      "step": 370
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.422430092487145,
+      "learning_rate": 2e-05,
+      "loss": 0.1106,
+      "step": 371
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.177662675621718,
+      "learning_rate": 2e-05,
+      "loss": 0.1125,
+      "step": 372
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.107316935465908,
+      "learning_rate": 2e-05,
+      "loss": 0.1068,
+      "step": 373
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.602012627960211,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 374
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.264752756445372,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 375
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.199847607124106,
+      "learning_rate": 2e-05,
+      "loss": 0.1649,
+      "step": 376
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.056510562232158,
+      "learning_rate": 2e-05,
+      "loss": 0.1382,
+      "step": 377
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.798438231915735,
+      "learning_rate": 2e-05,
+      "loss": 0.1976,
+      "step": 378
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.281216293927755,
+      "learning_rate": 2e-05,
+      "loss": 0.1089,
+      "step": 379
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.146754421929023,
+      "learning_rate": 2e-05,
+      "loss": 0.1241,
+      "step": 380
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.296295215049767,
+      "learning_rate": 2e-05,
+      "loss": 0.1449,
+      "step": 381
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.694683861803803,
+      "learning_rate": 2e-05,
+      "loss": 0.1709,
+      "step": 382
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.457382682346653,
+      "learning_rate": 2e-05,
+      "loss": 0.1338,
+      "step": 383
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.771081851408722,
+      "learning_rate": 2e-05,
+      "loss": 0.1919,
+      "step": 384
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.292715493837806,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 385
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.588756527522539,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 386
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.518520807150232,
+      "learning_rate": 2e-05,
+      "loss": 0.1008,
+      "step": 387
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.343649176387723,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 388
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.234640359126598,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 389
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.578864919045448,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 390
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.955599290191971,
+      "learning_rate": 2e-05,
+      "loss": 0.0785,
+      "step": 391
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.38486719502349,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 392
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.171142225585812,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 393
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.301249138548437,
+      "learning_rate": 2e-05,
+      "loss": 0.1938,
+      "step": 394
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.39966675151086,
+      "learning_rate": 2e-05,
+      "loss": 0.1432,
+      "step": 395
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.811870186948617,
+      "learning_rate": 2e-05,
+      "loss": 0.0949,
+      "step": 396
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.14020690271146,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 397
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.634691184153787,
+      "learning_rate": 2e-05,
+      "loss": 0.2109,
+      "step": 398
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.341545135229145,
+      "learning_rate": 2e-05,
+      "loss": 0.1061,
+      "step": 399
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.144244760288503,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.323432075472503,
+      "learning_rate": 2e-05,
+      "loss": 0.1251,
+      "step": 401
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.322137871262119,
+      "learning_rate": 2e-05,
+      "loss": 0.0981,
+      "step": 402
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.54200704519529,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 403
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.677708507960513,
+      "learning_rate": 2e-05,
+      "loss": 0.1323,
+      "step": 404
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 40.609638020983546,
+      "learning_rate": 2e-05,
+      "loss": 0.1133,
+      "step": 405
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.1417511739657895,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 406
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.867946118648244,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 407
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8271450734786825,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 408
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.255433143955593,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 409
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.530488042559319,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 410
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.854511121286343,
+      "learning_rate": 2e-05,
+      "loss": 0.1497,
+      "step": 411
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.232969656278136,
+      "learning_rate": 2e-05,
+      "loss": 0.1428,
+      "step": 412
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.680086964937722,
+      "learning_rate": 2e-05,
+      "loss": 0.105,
+      "step": 413
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.778688852469462,
+      "learning_rate": 2e-05,
+      "loss": 0.166,
+      "step": 414
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7966974161949754,
+      "learning_rate": 2e-05,
+      "loss": 0.098,
+      "step": 415
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.176196954820499,
+      "learning_rate": 2e-05,
+      "loss": 0.1063,
+      "step": 416
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.552194664714369,
+      "learning_rate": 2e-05,
+      "loss": 0.1594,
+      "step": 417
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.986796474338048,
+      "learning_rate": 2e-05,
+      "loss": 0.0886,
+      "step": 418
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.456631500354453,
+      "learning_rate": 2e-05,
+      "loss": 0.1257,
+      "step": 419
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.341320117394785,
+      "learning_rate": 2e-05,
+      "loss": 0.1884,
+      "step": 420
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.486688635112455,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 421
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.09321597913861,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 422
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.528123317077332,
+      "learning_rate": 2e-05,
+      "loss": 0.1321,
+      "step": 423
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.688080873017572,
+      "learning_rate": 2e-05,
+      "loss": 0.1472,
+      "step": 424
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.393464067821246,
+      "learning_rate": 2e-05,
+      "loss": 0.1346,
+      "step": 425
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8658209341685956,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 426
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.345120001256281,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 427
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.003393002875934,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 428
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.627225392013146,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 429
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.095658772311838,
+      "learning_rate": 2e-05,
+      "loss": 0.1303,
+      "step": 430
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.7288761922620495,
+      "learning_rate": 2e-05,
+      "loss": 0.1676,
+      "step": 431
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.387644177506112,
+      "learning_rate": 2e-05,
+      "loss": 0.1912,
+      "step": 432
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8349910275430177,
+      "learning_rate": 2e-05,
+      "loss": 0.1224,
+      "step": 433
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.6644116554481805,
+      "learning_rate": 2e-05,
+      "loss": 0.0883,
+      "step": 434
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.51615285103676,
+      "learning_rate": 2e-05,
+      "loss": 0.1383,
+      "step": 435
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 20.5026801613552,
+      "learning_rate": 2e-05,
+      "loss": 0.0853,
+      "step": 436
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.9488564554409455,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 437
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.208090382714412,
+      "learning_rate": 2e-05,
+      "loss": 0.1253,
+      "step": 438
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.5416960175758625,
+      "learning_rate": 2e-05,
+      "loss": 0.0938,
+      "step": 439
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.724561063106443,
+      "learning_rate": 2e-05,
+      "loss": 0.1247,
+      "step": 440
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.752747355828101,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 441
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3586192145827507,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 442
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.835480174434558,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 443
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.400083081588111,
+      "learning_rate": 2e-05,
+      "loss": 0.1493,
+      "step": 444
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.509946857713052,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 445
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.151580219776086,
+      "learning_rate": 2e-05,
+      "loss": 0.1495,
+      "step": 446
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.068281080469617,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 447
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.755883119182059,
+      "learning_rate": 2e-05,
+      "loss": 0.1486,
+      "step": 448
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.846186559266716,
+      "learning_rate": 2e-05,
+      "loss": 0.155,
+      "step": 449
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9135821080024193,
+      "learning_rate": 2e-05,
+      "loss": 0.1772,
+      "step": 450
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.260217613787034,
+      "learning_rate": 2e-05,
+      "loss": 0.1134,
+      "step": 451
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.715503308323053,
+      "learning_rate": 2e-05,
+      "loss": 0.1156,
+      "step": 452
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.27731937521079,
+      "learning_rate": 2e-05,
+      "loss": 0.1152,
+      "step": 453
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.454473028280124,
+      "learning_rate": 2e-05,
+      "loss": 0.0893,
+      "step": 454
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.274292085998222,
+      "learning_rate": 2e-05,
+      "loss": 0.1638,
+      "step": 455
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.982860082713999,
+      "learning_rate": 2e-05,
+      "loss": 0.096,
+      "step": 456
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.657635734408066,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 457
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.491395982816017,
+      "learning_rate": 2e-05,
+      "loss": 0.1429,
+      "step": 458
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.787089703208902,
+      "learning_rate": 2e-05,
+      "loss": 0.1076,
+      "step": 459
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.7175521846082376,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 460
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.961539356638783,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 461
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.5346699780724835,
+      "learning_rate": 2e-05,
+      "loss": 0.1517,
+      "step": 462
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.969235394224134,
+      "learning_rate": 2e-05,
+      "loss": 0.1716,
+      "step": 463
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.452063226057761,
+      "learning_rate": 2e-05,
+      "loss": 0.1121,
+      "step": 464
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.438609918823179,
+      "learning_rate": 2e-05,
+      "loss": 0.1461,
+      "step": 465
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.606616443298145,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 466
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.870108470340794,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 467
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.065055960278915,
+      "learning_rate": 2e-05,
+      "loss": 0.095,
+      "step": 468
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9141862739286846,
+      "learning_rate": 2e-05,
+      "loss": 0.1037,
+      "step": 469
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.2188096783973394,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 470
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.155618709875296,
+      "learning_rate": 2e-05,
+      "loss": 0.0889,
+      "step": 471
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7058462453734027,
+      "learning_rate": 2e-05,
+      "loss": 0.0882,
+      "step": 472
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.418549938974643,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 473
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.275908606101614,
+      "learning_rate": 2e-05,
+      "loss": 0.0903,
+      "step": 474
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.497709140664478,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 475
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.60646077038567,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 476
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8570062572773662,
+      "learning_rate": 2e-05,
+      "loss": 0.1243,
+      "step": 477
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.093537870331291,
+      "learning_rate": 2e-05,
+      "loss": 0.0878,
+      "step": 478
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.851026872948842,
+      "learning_rate": 2e-05,
+      "loss": 0.151,
+      "step": 479
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8668637174527833,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 480
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.47208882693,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 481
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2010074221791025,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 482
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.956468097751495,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 483
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.566862137409503,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 484
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.405586824623759,
+      "learning_rate": 2e-05,
+      "loss": 0.1219,
+      "step": 485
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.87931730372845,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 486
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.063667673452408,
+      "learning_rate": 2e-05,
+      "loss": 0.1328,
+      "step": 487
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.228440747027567,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 488
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.85177457152962,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 489
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.963353778483507,
+      "learning_rate": 2e-05,
+      "loss": 0.1633,
+      "step": 490
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.061689752332874,
+      "learning_rate": 2e-05,
+      "loss": 0.0903,
+      "step": 491
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.925702712482187,
+      "learning_rate": 2e-05,
+      "loss": 0.0864,
+      "step": 492
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.685832019317043,
+      "learning_rate": 2e-05,
+      "loss": 0.1149,
+      "step": 493
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3671648418611206,
+      "learning_rate": 2e-05,
+      "loss": 0.1084,
+      "step": 494
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.283176936631651,
+      "learning_rate": 2e-05,
+      "loss": 0.1254,
+      "step": 495
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.83586140401152,
+      "learning_rate": 2e-05,
+      "loss": 0.1702,
+      "step": 496
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.616711842713303,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 497
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.444561973433569,
+      "learning_rate": 2e-05,
+      "loss": 0.1167,
+      "step": 498
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.234817988460159,
+      "learning_rate": 2e-05,
+      "loss": 0.0931,
+      "step": 499
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3541650870815056,
+      "learning_rate": 2e-05,
+      "loss": 0.1055,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.551046339180922,
+      "learning_rate": 2e-05,
+      "loss": 0.1013,
+      "step": 501
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.43438666651188,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 502
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.682776302351858,
+      "learning_rate": 2e-05,
+      "loss": 0.1749,
+      "step": 503
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.1667437482697585,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 504
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.74240695798556,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 505
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.504453502143285,
+      "learning_rate": 2e-05,
+      "loss": 0.2559,
+      "step": 506
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.990646855186797,
+      "learning_rate": 2e-05,
+      "loss": 0.1763,
+      "step": 507
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.802634295165071,
+      "learning_rate": 2e-05,
+      "loss": 0.1064,
+      "step": 508
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.675179929934726,
+      "learning_rate": 2e-05,
+      "loss": 0.0741,
+      "step": 509
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.407092507620817,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 510
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.858938921258785,
+      "learning_rate": 2e-05,
+      "loss": 0.1376,
+      "step": 511
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4161820106705814,
+      "learning_rate": 2e-05,
+      "loss": 0.1196,
+      "step": 512
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.502527402674207,
+      "learning_rate": 2e-05,
+      "loss": 0.1956,
+      "step": 513
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.91095017696423,
+      "learning_rate": 2e-05,
+      "loss": 0.1535,
+      "step": 514
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1100737586888294,
+      "learning_rate": 2e-05,
+      "loss": 0.0962,
+      "step": 515
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9214632950455686,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 516
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.040327919148894,
+      "learning_rate": 2e-05,
+      "loss": 0.1113,
+      "step": 517
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.653614937448653,
+      "learning_rate": 2e-05,
+      "loss": 0.1328,
+      "step": 518
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.289561647369917,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 519
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.204256643244901,
+      "learning_rate": 2e-05,
+      "loss": 0.0741,
+      "step": 520
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.869913392785887,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 521
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.821729401180142,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 522
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.1034131334624435,
+      "learning_rate": 2e-05,
+      "loss": 0.0918,
+      "step": 523
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1794715451849265,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 524
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.798079455502007,
+      "learning_rate": 2e-05,
+      "loss": 0.1038,
+      "step": 525
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.172459365938814,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 526
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.228371346826954,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 527
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.275105384509072,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 528
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.228254394428586,
+      "learning_rate": 2e-05,
+      "loss": 0.2139,
+      "step": 529
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.423273527303701,
+      "learning_rate": 2e-05,
+      "loss": 0.1208,
+      "step": 530
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6992139806182958,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 531
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.750384442289999,
+      "learning_rate": 2e-05,
+      "loss": 0.0957,
+      "step": 532
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.423876614499829,
+      "learning_rate": 2e-05,
+      "loss": 0.0989,
+      "step": 533
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.9565896993117855,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 534
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.16654274082209,
+      "learning_rate": 2e-05,
+      "loss": 0.2798,
+      "step": 535
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.446716229716054,
+      "learning_rate": 2e-05,
+      "loss": 0.103,
+      "step": 536
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.15640646957194,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 537
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0404408206747338,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 538
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.655534224770787,
+      "learning_rate": 2e-05,
+      "loss": 0.1714,
+      "step": 539
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.378981601516965,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 540
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9086261083806346,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 541
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.173585672554124,
+      "learning_rate": 2e-05,
+      "loss": 0.1268,
+      "step": 542
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.117560874614041,
+      "learning_rate": 2e-05,
+      "loss": 0.1054,
+      "step": 543
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.129117124833203,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 544
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.727081867655675,
+      "learning_rate": 2e-05,
+      "loss": 0.1418,
+      "step": 545
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.003067029608816,
+      "learning_rate": 2e-05,
+      "loss": 0.1034,
+      "step": 546
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.4260362197705305,
+      "learning_rate": 2e-05,
+      "loss": 0.1364,
+      "step": 547
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.058150795050156,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 548
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.42676880542017,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 549
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.275177862678631,
+      "learning_rate": 2e-05,
+      "loss": 0.1855,
+      "step": 550
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.132350409456987,
+      "learning_rate": 2e-05,
+      "loss": 0.2366,
+      "step": 551
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.852729915543056,
+      "learning_rate": 2e-05,
+      "loss": 0.2644,
+      "step": 552
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.76382086498044,
+      "learning_rate": 2e-05,
+      "loss": 0.1085,
+      "step": 553
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8219856701576767,
+      "learning_rate": 2e-05,
+      "loss": 0.1279,
+      "step": 554
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.088801297571868,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 555
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.473155120428891,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 556
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.119751671092438,
+      "learning_rate": 2e-05,
+      "loss": 0.2419,
+      "step": 557
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.353546905994156,
+      "learning_rate": 2e-05,
+      "loss": 0.1753,
+      "step": 558
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.305813411121432,
+      "learning_rate": 2e-05,
+      "loss": 0.1341,
+      "step": 559
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8113141326162645,
+      "learning_rate": 2e-05,
+      "loss": 0.1077,
+      "step": 560
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.201941003203099,
+      "learning_rate": 2e-05,
+      "loss": 0.1042,
+      "step": 561
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.885572977801339,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 562
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.352389461271701,
+      "learning_rate": 2e-05,
+      "loss": 0.1696,
+      "step": 563
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.820065558333783,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 564
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9537530463320425,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 565
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2118475522339947,
+      "learning_rate": 2e-05,
+      "loss": 0.0925,
+      "step": 566
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.768866649475484,
+      "learning_rate": 2e-05,
+      "loss": 0.1431,
+      "step": 567
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.795229173442776,
+      "learning_rate": 2e-05,
+      "loss": 0.1274,
+      "step": 568
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.439818974415326,
+      "learning_rate": 2e-05,
+      "loss": 0.1885,
+      "step": 569
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.710649016547984,
+      "learning_rate": 2e-05,
+      "loss": 0.1147,
+      "step": 570
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.205701136318017,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 571
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.138344688325617,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 572
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.266206694684117,
+      "learning_rate": 2e-05,
+      "loss": 0.132,
+      "step": 573
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.954711986449029,
+      "learning_rate": 2e-05,
+      "loss": 0.1685,
+      "step": 574
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.342272280437671,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 575
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3599300901769467,
+      "learning_rate": 2e-05,
+      "loss": 0.114,
+      "step": 576
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4172675061900413,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 577
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1801471840754645,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 578
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.108683124985644,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 579
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6453524172445446,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 580
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.947308997601526,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 581
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.952957552244679,
+      "learning_rate": 2e-05,
+      "loss": 0.089,
+      "step": 582
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5448341716079215,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 583
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.546210696008807,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 584
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.055813214794615,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 585
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.346194773244498,
+      "learning_rate": 2e-05,
+      "loss": 0.1761,
+      "step": 586
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.741394130535655,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 587
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5981061273384456,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 588
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.2144165243027665,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 589
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.203273516319159,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 590
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.324093598915189,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 591
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.679976464292047,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 592
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.010117974634028,
+      "learning_rate": 2e-05,
+      "loss": 0.1488,
+      "step": 593
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.049335462155443,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 594
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.52060370665168,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 595
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.948234981659124,
+      "learning_rate": 2e-05,
+      "loss": 0.1522,
+      "step": 596
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.366887008892891,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 597
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.070717188086954,
+      "learning_rate": 2e-05,
+      "loss": 0.1678,
+      "step": 598
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.65086182480852,
+      "learning_rate": 2e-05,
+      "loss": 0.1324,
+      "step": 599
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.185116952516145,
+      "learning_rate": 2e-05,
+      "loss": 0.1119,
+      "step": 600
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.255044046150843,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 601
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.384602215234505,
+      "learning_rate": 2e-05,
+      "loss": 0.1023,
+      "step": 602
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.704712988417331,
+      "learning_rate": 2e-05,
+      "loss": 0.1144,
+      "step": 603
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.181190455890759,
+      "learning_rate": 2e-05,
+      "loss": 0.1222,
+      "step": 604
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.469399375508118,
+      "learning_rate": 2e-05,
+      "loss": 0.1268,
+      "step": 605
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7279047757867936,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 606
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.638868678588467,
+      "learning_rate": 2e-05,
+      "loss": 0.1335,
+      "step": 607
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.173299528085265,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 608
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0753007509115475,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 609
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.397772711381978,
+      "learning_rate": 2e-05,
+      "loss": 0.1442,
+      "step": 610
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.000572813029824,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 611
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.408494957309942,
+      "learning_rate": 2e-05,
+      "loss": 0.1091,
+      "step": 612
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.653567512018954,
+      "learning_rate": 2e-05,
+      "loss": 0.1034,
+      "step": 613
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4878149457802805,
+      "learning_rate": 2e-05,
+      "loss": 0.1472,
+      "step": 614
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3499989808493815,
+      "learning_rate": 2e-05,
+      "loss": 0.1199,
+      "step": 615
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.9761247906507435,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 616
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7530942231122566,
+      "learning_rate": 2e-05,
+      "loss": 0.113,
+      "step": 617
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.890226396183004,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 618
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4556117705024474,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 619
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.033479494042576,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 620
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4185325985813675,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 621
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.148224848525077,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 622
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8655340185537357,
+      "learning_rate": 2e-05,
+      "loss": 0.0884,
+      "step": 623
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.475892552471772,
+      "learning_rate": 2e-05,
+      "loss": 0.0897,
+      "step": 624
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.349684634546908,
+      "learning_rate": 2e-05,
+      "loss": 0.1453,
+      "step": 625
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.602323192582036,
+      "learning_rate": 2e-05,
+      "loss": 0.1194,
+      "step": 626
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6603730258645175,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 627
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.642557674277149,
+      "learning_rate": 2e-05,
+      "loss": 0.1041,
+      "step": 628
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.163375826502464,
+      "learning_rate": 2e-05,
+      "loss": 0.1268,
+      "step": 629
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.634915831294415,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 630
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3100598542486352,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 631
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2365896410539343,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 632
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.494542521279105,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 633
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.530859568645461,
+      "learning_rate": 2e-05,
+      "loss": 0.0874,
+      "step": 634
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.161062163521167,
+      "learning_rate": 2e-05,
+      "loss": 0.1332,
+      "step": 635
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.465747375952657,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 636
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.737477565940457,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 637
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.883995136407778,
+      "learning_rate": 2e-05,
+      "loss": 0.0871,
+      "step": 638
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.362204113602225,
+      "learning_rate": 2e-05,
+      "loss": 0.1174,
+      "step": 639
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.577224863392546,
+      "learning_rate": 2e-05,
+      "loss": 0.1178,
+      "step": 640
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6614887494939676,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 641
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.672236647956691,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 642
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.579960437876027,
+      "learning_rate": 2e-05,
+      "loss": 0.1058,
+      "step": 643
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6076956140641623,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 644
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.305007497942171,
+      "learning_rate": 2e-05,
+      "loss": 0.0834,
+      "step": 645
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.052862065380506,
+      "learning_rate": 2e-05,
+      "loss": 0.1663,
+      "step": 646
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.818240078795707,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 647
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.984577065983143,
+      "learning_rate": 2e-05,
+      "loss": 0.1318,
+      "step": 648
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.663401599990321,
+      "learning_rate": 2e-05,
+      "loss": 0.166,
+      "step": 649
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.308241054332763,
+      "learning_rate": 2e-05,
+      "loss": 0.1765,
+      "step": 650
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.385993603711221,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 651
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.448427116837055,
+      "learning_rate": 2e-05,
+      "loss": 0.1674,
+      "step": 652
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.381164564280774,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 653
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.015623414671057,
+      "learning_rate": 2e-05,
+      "loss": 0.1343,
+      "step": 654
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8218565396902515,
+      "learning_rate": 2e-05,
+      "loss": 0.1396,
+      "step": 655
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.940727241515042,
+      "learning_rate": 2e-05,
+      "loss": 0.1086,
+      "step": 656
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2993152138023825,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 657
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1174925078888585,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 658
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1489250600892054,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 659
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.767835560888079,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 660
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.257613151153243,
+      "learning_rate": 2e-05,
+      "loss": 0.0925,
+      "step": 661
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.833735007442867,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 662
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.984862533361727,
+      "learning_rate": 2e-05,
+      "loss": 0.1754,
+      "step": 663
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.896847517588928,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 664
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8883993991210524,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 665
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.170440751551808,
+      "learning_rate": 2e-05,
+      "loss": 0.1127,
+      "step": 666
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.537565257632992,
+      "learning_rate": 2e-05,
+      "loss": 0.1232,
+      "step": 667
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.913032247347165,
+      "learning_rate": 2e-05,
+      "loss": 0.0872,
+      "step": 668
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.394687714228489,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 669
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.586941181748148,
+      "learning_rate": 2e-05,
+      "loss": 0.0844,
+      "step": 670
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.001889110150191,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 671
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.3127899090818556,
+      "learning_rate": 2e-05,
+      "loss": 0.0978,
+      "step": 672
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1631860343262055,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 673
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5647694842171744,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 674
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.719539669004123,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 675
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.626352885559911,
+      "learning_rate": 2e-05,
+      "loss": 0.1696,
+      "step": 676
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.496949650451123,
+      "learning_rate": 2e-05,
+      "loss": 0.0969,
+      "step": 677
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.4776415231268825,
+      "learning_rate": 2e-05,
+      "loss": 0.1243,
+      "step": 678
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.516696911463873,
+      "learning_rate": 2e-05,
+      "loss": 0.1616,
+      "step": 679
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1126595023470314,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 680
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.3632350894197405,
+      "learning_rate": 2e-05,
+      "loss": 0.1259,
+      "step": 681
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.367543651273146,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 682
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.942231241346801,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 683
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.409893071960807,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 684
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.459943732785673,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 685
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3502092096848557,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 686
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.9392462784402,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 687
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.9299120150595535,
+      "learning_rate": 2e-05,
+      "loss": 0.1018,
+      "step": 688
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.65602104258879,
+      "learning_rate": 2e-05,
+      "loss": 0.1085,
+      "step": 689
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.5185598812011145,
+      "learning_rate": 2e-05,
+      "loss": 0.0881,
+      "step": 690
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.29166687113955,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 691
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.520064867279998,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 692
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8463589877660347,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 693
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.120000866651462,
+      "learning_rate": 2e-05,
+      "loss": 0.0886,
+      "step": 694
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8815723941367342,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 695
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.717289756657653,
+      "learning_rate": 2e-05,
+      "loss": 0.1141,
+      "step": 696
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.384152892530438,
+      "learning_rate": 2e-05,
+      "loss": 0.1059,
+      "step": 697
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.380989402658176,
+      "learning_rate": 2e-05,
+      "loss": 0.1296,
+      "step": 698
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.207647341096547,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 699
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3730957026226633,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 700
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.137170669787695,
+      "learning_rate": 2e-05,
+      "loss": 0.1592,
+      "step": 701
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7707137882908343,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 702
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.605297809352136,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 703
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.517011840677278,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 704
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9186138374539605,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 705
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.883662553748874,
+      "learning_rate": 2e-05,
+      "loss": 0.103,
+      "step": 706
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2717199075661187,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 707
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.264567820575617,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 708
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.015697419082676,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 709
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3773748670955346,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 710
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.090612342430555,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 711
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.800279734240208,
+      "learning_rate": 2e-05,
+      "loss": 0.1089,
+      "step": 712
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.142169013232114,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 713
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.382239323676448,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 714
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.991201320555789,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 715
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.278804331107728,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 716
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.274872199038848,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 717
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1255317235620366,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 718
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.371516172117182,
+      "learning_rate": 2e-05,
+      "loss": 0.1855,
+      "step": 719
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.367760419226165,
+      "learning_rate": 2e-05,
+      "loss": 0.1722,
+      "step": 720
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.355625097439717,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 721
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.3536315053695,
+      "learning_rate": 2e-05,
+      "loss": 0.1651,
+      "step": 722
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.992740128902262,
+      "learning_rate": 2e-05,
+      "loss": 0.1306,
+      "step": 723
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.038900460295017,
+      "learning_rate": 2e-05,
+      "loss": 0.106,
+      "step": 724
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5280216279610364,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 725
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2929935115005415,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 726
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.888127380846559,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 727
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8409456951210723,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 728
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.801088352794384,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 729
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.374577627204149,
+      "learning_rate": 2e-05,
+      "loss": 0.1533,
+      "step": 730
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.382986414208074,
+      "learning_rate": 2e-05,
+      "loss": 0.1315,
+      "step": 731
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.429020009208915,
+      "learning_rate": 2e-05,
+      "loss": 0.0997,
+      "step": 732
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9903007997877769,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 733
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.228330660274927,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 734
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.650674337469789,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 735
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.234671606957783,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 736
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1561293667935484,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 737
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.946178335108793,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 738
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9634141821586968,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 739
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.910217889942112,
+      "learning_rate": 2e-05,
+      "loss": 0.1457,
+      "step": 740
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.4076933064257915,
+      "learning_rate": 2e-05,
+      "loss": 0.1049,
+      "step": 741
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.824766262649214,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 742
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.615267230032759,
+      "learning_rate": 2e-05,
+      "loss": 0.0998,
+      "step": 743
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.629318269458288,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 744
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.247429899826091,
+      "learning_rate": 2e-05,
+      "loss": 0.152,
+      "step": 745
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.331286146040718,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 746
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.41662026002865,
+      "learning_rate": 2e-05,
+      "loss": 0.1381,
+      "step": 747
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.4292878048058,
+      "learning_rate": 2e-05,
+      "loss": 0.1425,
+      "step": 748
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2207737458619614,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 749
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.044927761336503,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 750
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.803680463253357,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 751
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.124834941904246,
+      "learning_rate": 2e-05,
+      "loss": 0.1386,
+      "step": 752
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6479546760453925,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 753
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.408903580952163,
+      "learning_rate": 2e-05,
+      "loss": 0.0878,
+      "step": 754
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.222854563204488,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 755
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.875481861503755,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 756
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.805851261892463,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 757
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.475672342097705,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 758
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.918247853360705,
+      "learning_rate": 2e-05,
+      "loss": 0.1456,
+      "step": 759
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.525807868119332,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 760
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7620095797240727,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 761
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.705009007055112,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 762
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.884264241878677,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 763
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.755162248905089,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 764
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.015590756801888,
+      "learning_rate": 2e-05,
+      "loss": 0.1732,
+      "step": 765
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.75488099469078,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 766
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.448620422445657,
+      "learning_rate": 2e-05,
+      "loss": 0.121,
+      "step": 767
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.625258530664636,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 768
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3678535811612456,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 769
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9803842755066836,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 770
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3990102000385085,
+      "learning_rate": 2e-05,
+      "loss": 0.0883,
+      "step": 771
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.679849102815396,
+      "learning_rate": 2e-05,
+      "loss": 0.0967,
+      "step": 772
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.1810731744883025,
+      "learning_rate": 2e-05,
+      "loss": 0.1414,
+      "step": 773
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.024756644052097,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 774
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.002364310429577,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 775
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.363423875006519,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 776
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.59143714486553,
+      "learning_rate": 2e-05,
+      "loss": 0.0989,
+      "step": 777
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.517702836489974,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 778
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9339957008202573,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 779
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5483422290504745,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 780
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5968260392073383,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 781
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.033687593859792,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 782
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5426344878899854,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 783
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0422842947006323,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 784
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.694587064781687,
+      "learning_rate": 2e-05,
+      "loss": 0.1078,
+      "step": 785
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.519125456137024,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 786
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.784018777822329,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 787
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0974746142409577,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 788
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.827510110850616,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 789
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.379017593225094,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 790
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.90285437610839,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 791
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.3310989626813665,
+      "learning_rate": 2e-05,
+      "loss": 0.1102,
+      "step": 792
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.068617180399228,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 793
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.770483531147535,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 794
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.658604442540896,
+      "learning_rate": 2e-05,
+      "loss": 0.1124,
+      "step": 795
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.20685525488024,
+      "learning_rate": 2e-05,
+      "loss": 0.0909,
+      "step": 796
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2018595358962676,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 797
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4364254745912115,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 798
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.634714741212739,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 799
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4262959388344774,
+      "learning_rate": 2e-05,
+      "loss": 0.0884,
+      "step": 800
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4140474157633016,
+      "learning_rate": 2e-05,
+      "loss": 0.0913,
+      "step": 801
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.2371247998611485,
+      "learning_rate": 2e-05,
+      "loss": 0.1517,
+      "step": 802
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.449402869804795,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 803
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.849010379337261,
+      "learning_rate": 2e-05,
+      "loss": 0.1131,
+      "step": 804
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.190101820089842,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 805
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.535740818800659,
+      "learning_rate": 2e-05,
+      "loss": 0.0924,
+      "step": 806
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.630181786282633,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 807
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.306635753256882,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 808
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.328842158756173,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 809
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.714690628662608,
+      "learning_rate": 2e-05,
+      "loss": 0.108,
+      "step": 810
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.959990451557444,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 811
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.399471738363215,
+      "learning_rate": 2e-05,
+      "loss": 0.1069,
+      "step": 812
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4066769393596803,
+      "learning_rate": 2e-05,
+      "loss": 0.0996,
+      "step": 813
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.203918835859134,
+      "learning_rate": 2e-05,
+      "loss": 0.1204,
+      "step": 814
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.53501144041916,
+      "learning_rate": 2e-05,
+      "loss": 0.1631,
+      "step": 815
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.256239406219608,
+      "learning_rate": 2e-05,
+      "loss": 0.108,
+      "step": 816
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7460913315050197,
+      "learning_rate": 2e-05,
+      "loss": 0.1007,
+      "step": 817
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2812636602208216,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 818
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3530664159771146,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 819
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.340128813400652,
+      "learning_rate": 2e-05,
+      "loss": 0.141,
+      "step": 820
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0836680118963478,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 821
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1223285700680012,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 822
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2499317748904932,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 823
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2162551823021315,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 824
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0003652350304124,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 825
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.351413536762805,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 826
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.488665161166083,
+      "learning_rate": 2e-05,
+      "loss": 0.1891,
+      "step": 827
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.845398960125805,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 828
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.670992053747057,
+      "learning_rate": 2e-05,
+      "loss": 0.1333,
+      "step": 829
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.995213505813081,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 830
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.099669404628012,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 831
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7670741161875996,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 832
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.500516014183362,
+      "learning_rate": 2e-05,
+      "loss": 0.1006,
+      "step": 833
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.005846519207797,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 834
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4412374812299253,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 835
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.202293795962273,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 836
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7262982472632094,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 837
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.012341533633388,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 838
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.953225302679115,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 839
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.627859108220248,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 840
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.639190391170272,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 841
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.037338648027074,
+      "learning_rate": 2e-05,
+      "loss": 0.0996,
+      "step": 842
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5596458646817384,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 843
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5499193216164566,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 844
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.341554603452542,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 845
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.973315999059394,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 846
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.309835356020606,
+      "learning_rate": 2e-05,
+      "loss": 0.1254,
+      "step": 847
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.011320011515245,
+      "learning_rate": 2e-05,
+      "loss": 0.1052,
+      "step": 848
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.3254703684907385,
+      "learning_rate": 2e-05,
+      "loss": 0.1318,
+      "step": 849
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.2830385801869815,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 850
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.768691568307349,
+      "learning_rate": 2e-05,
+      "loss": 0.1371,
+      "step": 851
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.844611109822947,
+      "learning_rate": 2e-05,
+      "loss": 0.1116,
+      "step": 852
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.113288896874057,
+      "learning_rate": 2e-05,
+      "loss": 0.0886,
+      "step": 853
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0314102917006167,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 854
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.119518049380801,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 855
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.016498190308258,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 856
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.116210486848129,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 857
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8275332174196595,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 858
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.89202229642819,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 859
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.9332894959653375,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 860
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.846141397136493,
+      "learning_rate": 2e-05,
+      "loss": 0.1634,
+      "step": 861
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.953929366494126,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 862
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.060067616791205,
+      "learning_rate": 2e-05,
+      "loss": 0.1221,
+      "step": 863
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.884676081578124,
+      "learning_rate": 2e-05,
+      "loss": 0.1635,
+      "step": 864
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.314614423396104,
+      "learning_rate": 2e-05,
+      "loss": 0.1122,
+      "step": 865
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.16358671615641,
+      "learning_rate": 2e-05,
+      "loss": 0.2212,
+      "step": 866
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.50774437705783,
+      "learning_rate": 2e-05,
+      "loss": 0.0882,
+      "step": 867
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8352186378342432,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 868
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.399667802177315,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 869
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.0970762906850355,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 870
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.970303623148489,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 871
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.989329242967566,
+      "learning_rate": 2e-05,
+      "loss": 0.154,
+      "step": 872
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.99682533199859,
+      "learning_rate": 2e-05,
+      "loss": 0.1254,
+      "step": 873
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.901071043914997,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 874
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3169412170206467,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 875
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.185716610505678,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 876
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.313281624624881,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 877
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.689954164365925,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 878
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7626035606351613,
+      "learning_rate": 2e-05,
+      "loss": 0.1111,
+      "step": 879
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.619052465133859,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 880
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.942936065170736,
+      "learning_rate": 2e-05,
+      "loss": 0.1956,
+      "step": 881
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.250918353458517,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 882
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.799643047164142,
+      "learning_rate": 2e-05,
+      "loss": 0.1146,
+      "step": 883
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8448660866145943,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 884
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.14557104143432,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 885
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.273395991428895,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 886
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.2036585262193285,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 887
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.1315098636476,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 888
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4176776569024967,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 889
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.201641061884635,
+      "learning_rate": 2e-05,
+      "loss": 0.0853,
+      "step": 890
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.277000654810123,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 891
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.962522531920671,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 892
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.241124234336188,
+      "learning_rate": 2e-05,
+      "loss": 0.1451,
+      "step": 893
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.091083803440958,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 894
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.806619831026454,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 895
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.412422684927829,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 896
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.86366021687103,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 897
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.479538012486461,
+      "learning_rate": 2e-05,
+      "loss": 0.2397,
+      "step": 898
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.6295017648713985,
+      "learning_rate": 2e-05,
+      "loss": 0.1409,
+      "step": 899
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.916779022769603,
+      "learning_rate": 2e-05,
+      "loss": 0.1086,
+      "step": 900
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.500666039238296,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 901
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.019693311160588,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 902
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.875371484485227,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 903
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5496697567288367,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 904
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.380316300861359,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 905
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.597762658403835,
+      "learning_rate": 2e-05,
+      "loss": 0.0874,
+      "step": 906
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1648967801127332,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 907
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.851584492107664,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 908
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.991286314333738,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 909
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.403559302847119,
+      "learning_rate": 2e-05,
+      "loss": 0.0755,
+      "step": 910
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.607943353128737,
+      "learning_rate": 2e-05,
+      "loss": 0.1361,
+      "step": 911
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.683373020497203,
+      "learning_rate": 2e-05,
+      "loss": 0.0949,
+      "step": 912
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5833527468369843,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 913
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3796429675638677,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 914
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.821561010369308,
+      "learning_rate": 2e-05,
+      "loss": 0.1971,
+      "step": 915
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.6124629819874,
+      "learning_rate": 2e-05,
+      "loss": 0.1573,
+      "step": 916
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.093741933675204,
+      "learning_rate": 2e-05,
+      "loss": 0.1583,
+      "step": 917
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.427390530727549,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 918
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.440259478435899,
+      "learning_rate": 2e-05,
+      "loss": 0.1289,
+      "step": 919
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0244357323671363,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 920
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.087992134018978,
+      "learning_rate": 2e-05,
+      "loss": 0.1761,
+      "step": 921
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.794745460343455,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 922
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5478889927539963,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 923
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6575136527838543,
+      "learning_rate": 2e-05,
+      "loss": 0.1324,
+      "step": 924
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3529507978397737,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 925
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5469482036905884,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 926
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.294214819301271,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 927
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9575839817380003,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 928
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.958878577640912,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 929
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.093689909481544,
+      "learning_rate": 2e-05,
+      "loss": 0.1873,
+      "step": 930
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.723491585359077,
+      "learning_rate": 2e-05,
+      "loss": 0.132,
+      "step": 931
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.462464144414348,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 932
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.531195665892849,
+      "learning_rate": 2e-05,
+      "loss": 0.1017,
+      "step": 933
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2158005913929917,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 934
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1508664589504063,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 935
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6310005509384777,
+      "learning_rate": 2e-05,
+      "loss": 0.1513,
+      "step": 936
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.0812791813839295,
+      "learning_rate": 2e-05,
+      "loss": 0.0878,
+      "step": 937
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3086619184332995,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 938
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.028118600435059,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 939
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.380815012692987,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 940
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.158143037029324,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 941
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.196727615119639,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 942
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.419189782756155,
+      "learning_rate": 2e-05,
+      "loss": 0.1252,
+      "step": 943
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.603373239903783,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 944
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0818715797518283,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 945
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1942887821062533,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 946
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.688468096157862,
+      "learning_rate": 2e-05,
+      "loss": 0.0949,
+      "step": 947
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.7003170259887606,
+      "learning_rate": 2e-05,
+      "loss": 0.1221,
+      "step": 948
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8416806262193506,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 949
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.587948827589523,
+      "learning_rate": 2e-05,
+      "loss": 0.1531,
+      "step": 950
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.990369797760956,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 951
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.37970545993352,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 952
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.252398991896139,
+      "learning_rate": 2e-05,
+      "loss": 0.0815,
+      "step": 953
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.597245492224619,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 954
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.365809900989289,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 955
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.306009688405903,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 956
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9248176934254277,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 957
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8169939377692077,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 958
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.634491943906069,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 959
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7714586014223828,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 960
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6198346315858734,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 961
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.619209004809698,
+      "learning_rate": 2e-05,
+      "loss": 0.098,
+      "step": 962
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.301769724294326,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 963
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.925251524454477,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 964
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.325379858661548,
+      "learning_rate": 2e-05,
+      "loss": 0.1092,
+      "step": 965
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.006703317578463,
+      "learning_rate": 2e-05,
+      "loss": 0.0942,
+      "step": 966
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7627458764526893,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 967
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4755704342956424,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 968
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.763963732721681,
+      "learning_rate": 2e-05,
+      "loss": 0.0773,
+      "step": 969
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.530920030726932,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 970
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.337695165992396,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 971
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.731774799548901,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 972
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.207644649099578,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 973
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.0409339686348735,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 974
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4682776327088964,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 975
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.44228373697824,
+      "learning_rate": 2e-05,
+      "loss": 0.109,
+      "step": 976
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7298315013227175,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 977
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.567923470813087,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 978
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.48803328093066,
+      "learning_rate": 2e-05,
+      "loss": 0.1539,
+      "step": 979
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9531958049720646,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 980
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.154916190760783,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 981
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.22887513493358,
+      "learning_rate": 2e-05,
+      "loss": 0.1037,
+      "step": 982
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.048543701212246,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 983
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.545532954491291,
+      "learning_rate": 2e-05,
+      "loss": 0.0821,
+      "step": 984
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.820852791429156,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 985
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.684185024018289,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 986
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7677651322215755,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 987
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.628403863615669,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 988
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2932815131719306,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 989
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8384818441170845,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 990
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.748219512414478,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 991
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.047335158013231,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 992
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.332146812326316,
+      "learning_rate": 2e-05,
+      "loss": 0.0675,
+      "step": 993
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.955756139145125,
+      "learning_rate": 2e-05,
+      "loss": 0.106,
+      "step": 994
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.13143721319133,
+      "learning_rate": 2e-05,
+      "loss": 0.1391,
+      "step": 995
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.91502060178005,
+      "learning_rate": 2e-05,
+      "loss": 0.0976,
+      "step": 996
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.5523395015237815,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 997
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9663381316044775,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 998
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.011316114705778,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 999
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7928767294945533,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.710759137796904,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 1001
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.915593703777639,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 1002
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5699143492679366,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 1003
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.014889167785569,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 1004
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.005768475264203,
+      "learning_rate": 2e-05,
+      "loss": 0.0775,
+      "step": 1005
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3575254094801985,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 1006
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9016634938417893,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 1007
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.249956970066718,
+      "learning_rate": 2e-05,
+      "loss": 0.1216,
+      "step": 1008
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.23103467345446,
+      "learning_rate": 2e-05,
+      "loss": 0.1174,
+      "step": 1009
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.937730228512926,
+      "learning_rate": 2e-05,
+      "loss": 0.1323,
+      "step": 1010
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.207740215383125,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 1011
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7330443714111734,
+      "learning_rate": 2e-05,
+      "loss": 0.1532,
+      "step": 1012
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.606660884491,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 1013
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.256481613601627,
+      "learning_rate": 2e-05,
+      "loss": 0.083,
+      "step": 1014
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.575572060428705,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 1015
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6770882414736947,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 1016
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.617475989084958,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 1017
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.300657135130641,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 1018
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7987133960021016,
+      "learning_rate": 2e-05,
+      "loss": 0.1365,
+      "step": 1019
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9572837963479459,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 1020
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3840663925181125,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 1021
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.110752231316926,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 1022
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.195683240978615,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 1023
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.010940598384897,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 1024
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.853591879384847,
+      "learning_rate": 2e-05,
+      "loss": 0.0969,
+      "step": 1025
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.32041733520159,
+      "learning_rate": 2e-05,
+      "loss": 0.1268,
+      "step": 1026
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.589160936577256,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 1027
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.138295565640356,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 1028
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1417703390512624,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 1029
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.705770457027227,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 1030
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.0778736366524795,
+      "learning_rate": 2e-05,
+      "loss": 0.1216,
+      "step": 1031
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7372898872028606,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 1032
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.64448833525958,
+      "learning_rate": 2e-05,
+      "loss": 0.1014,
+      "step": 1033
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.678193584830107,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 1034
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.044399523079017,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 1035
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.136947544164007,
+      "learning_rate": 2e-05,
+      "loss": 0.1242,
+      "step": 1036
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.509051256557905,
+      "learning_rate": 2e-05,
+      "loss": 0.1059,
+      "step": 1037
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.428524985150598,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 1038
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.881966962199021,
+      "learning_rate": 2e-05,
+      "loss": 0.1681,
+      "step": 1039
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.823115018447292,
+      "learning_rate": 2e-05,
+      "loss": 0.1132,
+      "step": 1040
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.566175887380716,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 1041
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.0644940115368895,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 1042
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.529266145143921,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 1043
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3839351914244373,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 1044
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5594548996597566,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 1045
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.626345258390159,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 1046
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.00255574594404,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 1047
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.478228977565096,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 1048
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8744941028940763,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 1049
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.6858487081685345,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 1050
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.879998213777426,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 1051
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0891306906753337,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 1052
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.630811274941975,
+      "learning_rate": 2e-05,
+      "loss": 0.2344,
+      "step": 1053
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.788851632461852,
+      "learning_rate": 2e-05,
+      "loss": 0.1671,
+      "step": 1054
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.877218279736243,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 1055
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9729385500730126,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 1056
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.586812220030754,
+      "learning_rate": 2e-05,
+      "loss": 0.0804,
+      "step": 1057
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7361472757599403,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 1058
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.057510366009636,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 1059
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.46521194193211,
+      "learning_rate": 2e-05,
+      "loss": 0.1636,
+      "step": 1060
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.700738733221035,
+      "learning_rate": 2e-05,
+      "loss": 0.1359,
+      "step": 1061
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9174318715922674,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 1062
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.40521660989865,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 1063
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.351955444355068,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 1064
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.818868778053099,
+      "learning_rate": 2e-05,
+      "loss": 0.0942,
+      "step": 1065
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1286159194948024,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 1066
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.819232936358814,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 1067
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7115371380252844,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 1068
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.106890246291856,
+      "learning_rate": 2e-05,
+      "loss": 0.1508,
+      "step": 1069
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5398495437335393,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 1070
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.052751412763275,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 1071
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7160592482312156,
+      "learning_rate": 2e-05,
+      "loss": 0.0913,
+      "step": 1072
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0642013392006446,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 1073
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.1967105217812914,
+      "learning_rate": 2e-05,
+      "loss": 0.1466,
+      "step": 1074
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7892122164387017,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 1075
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.306508643102349,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 1076
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3447494410392045,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 1077
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.911900104087797,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 1078
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9948177146145287,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 1079
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.014892986386894,
+      "learning_rate": 2e-05,
+      "loss": 0.082,
+      "step": 1080
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.122007266938209,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 1081
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.753764040057788,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 1082
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.739663494494193,
+      "learning_rate": 2e-05,
+      "loss": 0.1003,
+      "step": 1083
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.775319950108655,
+      "learning_rate": 2e-05,
+      "loss": 0.09,
+      "step": 1084
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1230946020466215,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 1085
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.201167107955803,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 1086
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.1253509372203245,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 1087
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.034509947911146,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 1088
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2990333640526184,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 1089
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.549806299432234,
+      "learning_rate": 2e-05,
+      "loss": 0.0889,
+      "step": 1090
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.857956380962382,
+      "learning_rate": 2e-05,
+      "loss": 0.1434,
+      "step": 1091
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.582242770174201,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 1092
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.952746281629768,
+      "learning_rate": 2e-05,
+      "loss": 0.1223,
+      "step": 1093
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5931755851101229,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 1094
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.396785834976398,
+      "learning_rate": 2e-05,
+      "loss": 0.1265,
+      "step": 1095
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5582554717335766,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 1096
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.658556475722925,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 1097
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5912500061608754,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 1098
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4874691760160563,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 1099
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.586587351168306,
+      "learning_rate": 2e-05,
+      "loss": 0.1019,
+      "step": 1100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.46238914428407,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 1101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6443507528759946,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 1102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.7187086426953995,
+      "learning_rate": 2e-05,
+      "loss": 0.214,
+      "step": 1103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5997233181382764,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 1104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.253686351829659,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 1105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.174875124603227,
+      "learning_rate": 2e-05,
+      "loss": 0.1482,
+      "step": 1106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.887403510266942,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 1107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.809016465168354,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 1108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5624657344843333,
+      "learning_rate": 2e-05,
+      "loss": 0.1002,
+      "step": 1109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.944084837614276,
+      "learning_rate": 2e-05,
+      "loss": 0.0858,
+      "step": 1110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.058499116761194,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 1111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.337175841957254,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 1112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.082077741484845,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 1113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7985985416185826,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 1114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.120798286321913,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 1115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4147879982531735,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 1116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.00080484714147,
+      "learning_rate": 2e-05,
+      "loss": 0.1459,
+      "step": 1117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0215178317071403,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 1118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.527982195455917,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 1119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.794767195002879,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 1120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.4889520508524114,
+      "learning_rate": 2e-05,
+      "loss": 0.2085,
+      "step": 1121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.353622488433342,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 1122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.21084111349879,
+      "learning_rate": 2e-05,
+      "loss": 0.1297,
+      "step": 1123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.142907653288072,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 1124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2514543947051093,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 1125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.877357499755367,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 1126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.541017515120575,
+      "learning_rate": 2e-05,
+      "loss": 0.13,
+      "step": 1127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.788748623845152,
+      "learning_rate": 2e-05,
+      "loss": 0.1367,
+      "step": 1128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.35897285360966,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 1129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.6068420692170475,
+      "learning_rate": 2e-05,
+      "loss": 0.0898,
+      "step": 1130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.591043664015072,
+      "learning_rate": 2e-05,
+      "loss": 0.0963,
+      "step": 1131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6891761461757566,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 1132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2599671322618406,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 1133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.092681398227513,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 1134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.917534617841388,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 1135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2771197481341436,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 1136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.240983936296482,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 1137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.657221340785235,
+      "learning_rate": 2e-05,
+      "loss": 0.0956,
+      "step": 1138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9901416411152764,
+      "learning_rate": 2e-05,
+      "loss": 0.1026,
+      "step": 1139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.062428869945203,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 1140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4277855676376663,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 1141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.013039079146481,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 1142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.6058382663484885,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 1143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5013809204916018,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 1144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.434503810087971,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 1145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.390666397599195,
+      "learning_rate": 2e-05,
+      "loss": 0.0754,
+      "step": 1146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7819326184557256,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 1147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1634979121088405,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 1148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5023331358844714,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 1149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.161371265630366,
+      "learning_rate": 2e-05,
+      "loss": 0.1497,
+      "step": 1150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.86633245787755,
+      "learning_rate": 2e-05,
+      "loss": 0.1285,
+      "step": 1151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.397121702227851,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 1152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.6822408317811846,
+      "learning_rate": 2e-05,
+      "loss": 0.1011,
+      "step": 1153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.265579721499504,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 1154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.669515564245755,
+      "learning_rate": 2e-05,
+      "loss": 0.1122,
+      "step": 1155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.146502517398726,
+      "learning_rate": 2e-05,
+      "loss": 0.0995,
+      "step": 1156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.062130538593904,
+      "learning_rate": 2e-05,
+      "loss": 0.0813,
+      "step": 1157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.787607952556375,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 1158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7569748302474126,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 1159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5173206902344942,
+      "learning_rate": 2e-05,
+      "loss": 0.0813,
+      "step": 1160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.220598020309923,
+      "learning_rate": 2e-05,
+      "loss": 0.1442,
+      "step": 1161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7774029732363477,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 1162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9917270714515642,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 1163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8686330202141725,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 1164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.175421912827444,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 1165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5482618616367536,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 1166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.211702356038566,
+      "learning_rate": 2e-05,
+      "loss": 0.1118,
+      "step": 1167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.998256422551109,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 1168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7199244848400617,
+      "learning_rate": 2e-05,
+      "loss": 0.0936,
+      "step": 1169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5334868731733637,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 1170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.679762653674413,
+      "learning_rate": 2e-05,
+      "loss": 0.1852,
+      "step": 1171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7510907870877013,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 1172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.610544657607702,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 1173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0447624512266,
+      "learning_rate": 2e-05,
+      "loss": 0.1054,
+      "step": 1174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.166488604210933,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 1175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.761703182072061,
+      "learning_rate": 2e-05,
+      "loss": 0.1456,
+      "step": 1176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.301300560159076,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 1177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.612153804254578,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 1178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.460821627900769,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 1179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.769152623269262,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 1180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.733392701343641,
+      "learning_rate": 2e-05,
+      "loss": 0.1497,
+      "step": 1181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.849906896897026,
+      "learning_rate": 2e-05,
+      "loss": 0.0776,
+      "step": 1182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.71957765938621,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 1183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5073018252255834,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 1184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.274030470450185,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 1185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.969205833827956,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 1186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.255989494543241,
+      "learning_rate": 2e-05,
+      "loss": 0.1347,
+      "step": 1187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.3758984100978795,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 1188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9453261041740046,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 1189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2276592565049174,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 1190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.786780397490736,
+      "learning_rate": 2e-05,
+      "loss": 0.1637,
+      "step": 1191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.462176372250706,
+      "learning_rate": 2e-05,
+      "loss": 0.0671,
+      "step": 1192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.172730051035683,
+      "learning_rate": 2e-05,
+      "loss": 0.1013,
+      "step": 1193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.289932246173337,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 1194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.955271985989623,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 1195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.361572996122968,
+      "learning_rate": 2e-05,
+      "loss": 0.1462,
+      "step": 1196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.834660364432965,
+      "learning_rate": 2e-05,
+      "loss": 0.1044,
+      "step": 1197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.511503750387355,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 1198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.466247414048263,
+      "learning_rate": 2e-05,
+      "loss": 0.1429,
+      "step": 1199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3373260963742815,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 1200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9142597323918746,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 1201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.629924514615521,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 1202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5971689326439975,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 1203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.038322768609687,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 1204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.285106620396854,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 1205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.840358394880314,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 1206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3026152732199712,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 1207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5403690117723703,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 1208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.6270617627160355,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 1209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.193431922914006,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 1210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.150576512381362,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 1211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.69832713609936,
+      "learning_rate": 2e-05,
+      "loss": 0.0872,
+      "step": 1212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.1843703861229,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 1213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.764826519782825,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 1214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.510292799388085,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 1215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.22883832407566,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 1216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.175195346475986,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 1217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.810176890233722,
+      "learning_rate": 2e-05,
+      "loss": 0.1215,
+      "step": 1218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.034207032384022,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 1219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5569858373017365,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 1220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.184868783218808,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 1221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.109402624399711,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 1222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.250625115590767,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 1223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.179858727019953,
+      "learning_rate": 2e-05,
+      "loss": 0.0703,
+      "step": 1224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.753079866237973,
+      "learning_rate": 2e-05,
+      "loss": 0.1137,
+      "step": 1225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.739953653419595,
+      "learning_rate": 2e-05,
+      "loss": 0.1039,
+      "step": 1226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.557632287819754,
+      "learning_rate": 2e-05,
+      "loss": 0.1611,
+      "step": 1227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.40336007910607,
+      "learning_rate": 2e-05,
+      "loss": 0.1194,
+      "step": 1228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5159435890802113,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 1229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.524396022735483,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 1230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0784992435882517,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 1231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2833744207633362,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 1232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.518197040785388,
+      "learning_rate": 2e-05,
+      "loss": 0.1285,
+      "step": 1233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.231362078760859,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 1234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4316461157066525,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 1235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9847550753156593,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 1236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5831074386206978,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 1237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6091327007923537,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 1238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.496756111000344,
+      "learning_rate": 2e-05,
+      "loss": 0.0754,
+      "step": 1239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.021820632341834,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 1240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.894873490615957,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 1241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.61419488026979,
+      "learning_rate": 2e-05,
+      "loss": 0.0996,
+      "step": 1242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.926568000653867,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 1243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.522942445114287,
+      "learning_rate": 2e-05,
+      "loss": 0.0995,
+      "step": 1244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.66151570496556,
+      "learning_rate": 2e-05,
+      "loss": 0.0868,
+      "step": 1245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0520329606381593,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 1246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.17962657848417,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 1247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5422805216986577,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 1248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.295565984912501,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 1249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8361406136557687,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.023896123245706,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 1251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.55859870784678,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 1252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.112880125710789,
+      "learning_rate": 2e-05,
+      "loss": 0.0909,
+      "step": 1253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.675619098652977,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 1254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.887495260450412,
+      "learning_rate": 2e-05,
+      "loss": 0.1332,
+      "step": 1255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.367966596758137,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 1256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4493990167361264,
+      "learning_rate": 2e-05,
+      "loss": 0.1328,
+      "step": 1257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5398383339332309,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 1258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.013883576761769,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 1259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.338779009787439,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 1260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0409681136401727,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 1261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.749608897301088,
+      "learning_rate": 2e-05,
+      "loss": 0.0957,
+      "step": 1262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6734692111959881,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 1263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6589386449842067,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 1264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6927108154862442,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 1265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.583751136611955,
+      "learning_rate": 2e-05,
+      "loss": 0.0861,
+      "step": 1266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0901016969222979,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 1267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.861034833521558,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 1268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.939384390585428,
+      "learning_rate": 2e-05,
+      "loss": 0.1108,
+      "step": 1269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.38258897443675,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 1270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0596512583388042,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 1271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.265173360648629,
+      "learning_rate": 2e-05,
+      "loss": 0.098,
+      "step": 1272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.723620543402931,
+      "learning_rate": 2e-05,
+      "loss": 0.1509,
+      "step": 1273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.705160702607607,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 1274
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7889015300898228,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 1275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.561141310585228,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 1276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.254116070463212,
+      "learning_rate": 2e-05,
+      "loss": 0.1492,
+      "step": 1277
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5903705595770468,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 1278
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9844341172359745,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 1279
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3998329829319156,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 1280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.750567399065757,
+      "learning_rate": 2e-05,
+      "loss": 0.2557,
+      "step": 1281
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.216870459246815,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 1282
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.356729092917377,
+      "learning_rate": 2e-05,
+      "loss": 0.1258,
+      "step": 1283
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.204793492529608,
+      "learning_rate": 2e-05,
+      "loss": 0.0775,
+      "step": 1284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2590606301907186,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 1285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2820234567848887,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 1286
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.749315635755356,
+      "learning_rate": 2e-05,
+      "loss": 0.0903,
+      "step": 1287
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.176793246563383,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 1288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2107460828757426,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 1289
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.847775466276831,
+      "learning_rate": 2e-05,
+      "loss": 0.1876,
+      "step": 1290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.009002591745319,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 1291
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.702484166105216,
+      "learning_rate": 2e-05,
+      "loss": 0.0889,
+      "step": 1292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8125185647987547,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 1293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0240307602907364,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 1294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6360950325664354,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 1295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.742567286129532,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 1296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.817396677713692,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 1297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.340038769066629,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 1298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.9074586307789945,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 1299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.599404827915185,
+      "learning_rate": 2e-05,
+      "loss": 0.0755,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4671268273881797,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 1301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.469378012427144,
+      "learning_rate": 2e-05,
+      "loss": 0.1393,
+      "step": 1302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.803281303112631,
+      "learning_rate": 2e-05,
+      "loss": 0.1326,
+      "step": 1303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.923656516619771,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 1304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.540134708256648,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 1305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.086951762864312,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 1306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6843054659167587,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 1307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7029140147519635,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 1308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7976262452903535,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 1309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2910447110176153,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 1310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5144629201899429,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 1311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.04931071647882,
+      "learning_rate": 2e-05,
+      "loss": 0.0909,
+      "step": 1312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3212283868384311,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 1313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.236887041518092,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 1314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9915128313231296,
+      "learning_rate": 2e-05,
+      "loss": 0.1909,
+      "step": 1315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.57677317926802,
+      "learning_rate": 2e-05,
+      "loss": 0.0862,
+      "step": 1316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.216617084267947,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 1317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.341323390807333,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 1318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.035914499428361,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 1319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.684264630524644,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 1320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1221384298168084,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 1321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.44044562624927,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 1322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4490103260288327,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 1323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1656452657865664,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 1324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.930984738838918,
+      "learning_rate": 2e-05,
+      "loss": 0.0992,
+      "step": 1325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0393020664539643,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 1326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.109591222558221,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 1327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.661730318924078,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 1328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.8595047225073555,
+      "learning_rate": 2e-05,
+      "loss": 0.1045,
+      "step": 1329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6102171076532352,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 1330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.077707310620054,
+      "learning_rate": 2e-05,
+      "loss": 0.1655,
+      "step": 1331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.664649504714267,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 1332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.303831598354499,
+      "learning_rate": 2e-05,
+      "loss": 0.1144,
+      "step": 1333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3525845763186393,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 1334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1543092701365985,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 1335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.820132259912767,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 1336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.410367948791301,
+      "learning_rate": 2e-05,
+      "loss": 0.1313,
+      "step": 1337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0445322273141233,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 1338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.136794230476834,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 1339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.879900905934891,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 1340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.092110864307418,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 1341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.536560957188678,
+      "learning_rate": 2e-05,
+      "loss": 0.1165,
+      "step": 1342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.9971319337383315,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 1343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8718891577658074,
+      "learning_rate": 2e-05,
+      "loss": 0.1218,
+      "step": 1344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3994066281608597,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 1345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8378376968335217,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 1346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2801354180660494,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 1347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1384142026309068,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 1348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.123136587261447,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 1349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.735144109189618,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 1350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.992159771106014,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 1351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.093610703137529,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 1352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.269737742954255,
+      "learning_rate": 2e-05,
+      "loss": 0.1373,
+      "step": 1353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.778007970086204,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 1354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9473622118554643,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 1355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4789862580466693,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 1356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.563416479950657,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 1357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.974110141321566,
+      "learning_rate": 2e-05,
+      "loss": 0.1396,
+      "step": 1358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.873478129159585,
+      "learning_rate": 2e-05,
+      "loss": 0.1241,
+      "step": 1359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3842282757430633,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 1360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.230431554432719,
+      "learning_rate": 2e-05,
+      "loss": 0.1312,
+      "step": 1361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0597540553368647,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 1362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.519706175377241,
+      "learning_rate": 2e-05,
+      "loss": 0.1285,
+      "step": 1363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.67046532070103,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 1364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0688483811990745,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 1365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.620302174592758,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 1366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7313265947566534,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 1367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.29106807497151,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 1368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.158698945163488,
+      "learning_rate": 2e-05,
+      "loss": 0.1163,
+      "step": 1369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.166904417248535,
+      "learning_rate": 2e-05,
+      "loss": 0.0809,
+      "step": 1370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.02388617226782,
+      "learning_rate": 2e-05,
+      "loss": 0.0834,
+      "step": 1371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0334513428198204,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 1372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1077451837987105,
+      "learning_rate": 2e-05,
+      "loss": 0.0741,
+      "step": 1373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.290445842740082,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 1374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2830358096123877,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 1375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.155150303882107,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 1376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.210886987901733,
+      "learning_rate": 2e-05,
+      "loss": 0.0962,
+      "step": 1377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.790028109309297,
+      "learning_rate": 2e-05,
+      "loss": 0.1298,
+      "step": 1378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.8080883860659895,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 1379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.774899088540884,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 1380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.259830997320483,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 1381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.238772877952328,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 1382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.864809480341293,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 1383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.884420938275968,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 1384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.463716996805594,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 1385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0655989246251045,
+      "learning_rate": 2e-05,
+      "loss": 0.0811,
+      "step": 1386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2080134608017246,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 1387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5033371678463456,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 1388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.299441298752258,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 1389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5094380562010605,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 1390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.325191081292106,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 1391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.529309788969204,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 1392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.482883000457342,
+      "learning_rate": 2e-05,
+      "loss": 0.083,
+      "step": 1393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6452898884148786,
+      "learning_rate": 2e-05,
+      "loss": 0.1076,
+      "step": 1394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.605014306763929,
+      "learning_rate": 2e-05,
+      "loss": 0.1284,
+      "step": 1395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.276552787851118,
+      "learning_rate": 2e-05,
+      "loss": 0.0886,
+      "step": 1396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3700427630510092,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 1397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.857952626837425,
+      "learning_rate": 2e-05,
+      "loss": 0.0923,
+      "step": 1398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7345256654919834,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 1399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.220730735059954,
+      "learning_rate": 2e-05,
+      "loss": 0.1355,
+      "step": 1400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3426350248264594,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 1401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4830355272039144,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 1402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8437159609591407,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 1403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9007275131091002,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 1404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6309573937843889,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 1405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.709381527091318,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 1406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9897924649504553,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 1407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.744011611586277,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 1408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8549815612452363,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 1409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1450637776299135,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 1410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3025915835761035,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 1411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0143202098507467,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 1412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9985064921509976,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 1413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7974264113407568,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 1414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8254785959394666,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 1415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.562127965644986,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 1416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.850193484510105,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 1417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.368137203258959,
+      "learning_rate": 2e-05,
+      "loss": 0.1353,
+      "step": 1418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2035949362305303,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 1419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0765227974441145,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 1420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.857675031249848,
+      "learning_rate": 2e-05,
+      "loss": 0.1603,
+      "step": 1421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5652530117063703,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 1422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.5096565058258165,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 1423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7056812386416178,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 1424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.5473790759075055,
+      "learning_rate": 2e-05,
+      "loss": 0.1368,
+      "step": 1425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.587949765818035,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 1426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.025447901885533,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 1427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4181281386399958,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 1428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.7849864598070955,
+      "learning_rate": 2e-05,
+      "loss": 0.1089,
+      "step": 1429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.446624713571051,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 1430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9545290417465138,
+      "learning_rate": 2e-05,
+      "loss": 0.0897,
+      "step": 1431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.862926370919542,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 1432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.758468094895728,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 1433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.269612337517475,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 1434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3348793516499895,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 1435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5957291973323269,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 1436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.545178112510889,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 1437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3652526243024303,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 1438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3684280216939912,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 1439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.682674422623866,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 1440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1369519428995605,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 1441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.927396728101391,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 1442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.624863848107146,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 1443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.120605902700608,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 1444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9881455475888092,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 1445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.065491932824566,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 1446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7659898075514513,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 1447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.227642044771868,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 1448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.877283180225912,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 1449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.414256090175558,
+      "learning_rate": 2e-05,
+      "loss": 0.0912,
+      "step": 1450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6152739404330667,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 1451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6297769224179697,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 1452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.168600174496072,
+      "learning_rate": 2e-05,
+      "loss": 0.1119,
+      "step": 1453
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.787922908739164,
+      "learning_rate": 2e-05,
+      "loss": 0.0964,
+      "step": 1454
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.069909610473815,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 1455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.847607657706196,
+      "learning_rate": 2e-05,
+      "loss": 0.105,
+      "step": 1456
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8489263152088564,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 1457
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.747970352246132,
+      "learning_rate": 2e-05,
+      "loss": 0.1234,
+      "step": 1458
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0367711079857953,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 1459
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.188750334656736,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 1460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.317999831481908,
+      "learning_rate": 2e-05,
+      "loss": 0.0963,
+      "step": 1461
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6971023450970453,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 1462
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.431507184400059,
+      "learning_rate": 2e-05,
+      "loss": 0.1023,
+      "step": 1463
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6638955877823367,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 1464
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7440265477242225,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 1465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.533157778523265,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 1466
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8597421818941946,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 1467
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.713946796801611,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 1468
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.413961661495005,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 1469
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8916439746834437,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 1470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.844872916713251,
+      "learning_rate": 2e-05,
+      "loss": 0.0809,
+      "step": 1471
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3583660651064084,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 1472
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.532651913333399,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 1473
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.088608282219995,
+      "learning_rate": 2e-05,
+      "loss": 0.1071,
+      "step": 1474
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.196989258395833,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 1475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.611132684271621,
+      "learning_rate": 2e-05,
+      "loss": 0.0809,
+      "step": 1476
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.014727619217078,
+      "learning_rate": 2e-05,
+      "loss": 0.1329,
+      "step": 1477
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.856299896180283,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 1478
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1979898126356894,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 1479
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.599712483087506,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 1480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.002922920372888,
+      "learning_rate": 2e-05,
+      "loss": 0.0736,
+      "step": 1481
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.899897544688812,
+      "learning_rate": 2e-05,
+      "loss": 0.1162,
+      "step": 1482
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.97585429065785,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 1483
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.744267869909957,
+      "learning_rate": 2e-05,
+      "loss": 0.116,
+      "step": 1484
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.38646572976237,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 1485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1382311147721613,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 1486
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.015494228580067,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 1487
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.1565155126952,
+      "learning_rate": 2e-05,
+      "loss": 0.0804,
+      "step": 1488
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0467975075352067,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 1489
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.76157069045412,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 1490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7950681143245504,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 1491
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3273486454200603,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 1492
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.5529593432995075,
+      "learning_rate": 2e-05,
+      "loss": 0.1021,
+      "step": 1493
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.174717538381316,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 1494
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.053557898256972,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 1495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.152551769796764,
+      "learning_rate": 2e-05,
+      "loss": 0.1634,
+      "step": 1496
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1244726087564523,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 1497
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5689323775219939,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 1498
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.033873940936074,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 1499
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3267050838339414,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 1500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.805821861452859,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 1501
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0083895199251356,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 1502
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9757975428683636,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 1503
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4568181376314424,
+      "learning_rate": 2e-05,
+      "loss": 0.1122,
+      "step": 1504
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.348640033744212,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 1505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.341440694535258,
+      "learning_rate": 2e-05,
+      "loss": 0.1205,
+      "step": 1506
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.619756911990802,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 1507
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.803161189613147,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 1508
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2855244136183193,
+      "learning_rate": 2e-05,
+      "loss": 0.0878,
+      "step": 1509
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6469688200602002,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 1510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.179740659003987,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 1511
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6806665450437432,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 1512
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9051978563500738,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 1513
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4828260596906855,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 1514
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4035930945647923,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 1515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.242686317861888,
+      "learning_rate": 2e-05,
+      "loss": 0.1202,
+      "step": 1516
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6077118101617707,
+      "learning_rate": 2e-05,
+      "loss": 0.0956,
+      "step": 1517
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8406774669204062,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 1518
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8153140930516796,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 1519
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.786245889968283,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 1520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8991918655780258,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 1521
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7135366874044387,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 1522
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.989212271198129,
+      "learning_rate": 2e-05,
+      "loss": 0.0721,
+      "step": 1523
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.179558247468216,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 1524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5314080459798385,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 1525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1628703567705383,
+      "learning_rate": 2e-05,
+      "loss": 0.1255,
+      "step": 1526
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.000062942009671,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 1527
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9929851057277954,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 1528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0773711020170675,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 1529
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4624182846397424,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 1530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1198137798639474,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 1531
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.333261443347636,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 1532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.576036937034164,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 1533
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.42116639630518,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 1534
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.370315994825383,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 1535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6089238673397412,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 1536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.759793019380719,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 1537
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.570078366639614,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 1538
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.322866519956499,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 1539
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5509932089870005,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 1540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.581968377882455,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 1541
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.745390920569999,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 1542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.679595839557509,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 1543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.621959196609765,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 1544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.253657682150309,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 1545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.749790588581351,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 1546
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2494727954209974,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 1547
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.693587179190375,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 1548
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.805417330251901,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 1549
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.308264158617111,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 1550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.521630079815422,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 1551
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7917478503803834,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 1552
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7485912219922257,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 1553
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.123967128100953,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 1554
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.71605136276691,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 1555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.946208149520321,
+      "learning_rate": 2e-05,
+      "loss": 0.1071,
+      "step": 1556
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.686804351050341,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 1557
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.55003286910231,
+      "learning_rate": 2e-05,
+      "loss": 0.1477,
+      "step": 1558
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.851993723158687,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 1559
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.142530024831937,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 1560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.001351846653382,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 1561
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3532800833573377,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 1562
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.116414933223547,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 1563
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.397915363759801,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 1564
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.7700386995868955,
+      "learning_rate": 2e-05,
+      "loss": 0.1616,
+      "step": 1565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.112523674445118,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 1566
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.491936134271612,
+      "learning_rate": 2e-05,
+      "loss": 0.1086,
+      "step": 1567
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.455819300051282,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 1568
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.697842334841983,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 1569
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.4335832099957315,
+      "learning_rate": 2e-05,
+      "loss": 0.1069,
+      "step": 1570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7390552640560448,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 1571
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.261194461543239,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 1572
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.454005478216273,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 1573
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.896685384357434,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 1574
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6392105544782347,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 1575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8063347634882785,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 1576
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.659482572412886,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 1577
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.027225111537598,
+      "learning_rate": 2e-05,
+      "loss": 0.0776,
+      "step": 1578
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.480482039335249,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 1579
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.379572849086324,
+      "learning_rate": 2e-05,
+      "loss": 0.0808,
+      "step": 1580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3641772376617367,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 1581
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.201806815987163,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 1582
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1095915292462073,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 1583
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6603387709868545,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 1584
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.940806136176923,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 1585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2368585615468035,
+      "learning_rate": 2e-05,
+      "loss": 0.1018,
+      "step": 1586
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1331029002589776,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 1587
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0738461570390205,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 1588
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9765376781595765,
+      "learning_rate": 2e-05,
+      "loss": 0.1261,
+      "step": 1589
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9801389979261788,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 1590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3448250726825317,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 1591
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3555123750559552,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 1592
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.147077317123626,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 1593
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0191733893601236,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 1594
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2047836877507345,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 1595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.436590182024626,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 1596
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6520042701970807,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 1597
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7102560896420194,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 1598
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0459079870702337,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 1599
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.371864557824569,
+      "learning_rate": 2e-05,
+      "loss": 0.0994,
+      "step": 1600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.057265342759375,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 1601
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.185599864002112,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 1602
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0844814293769685,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 1603
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.145210232068091,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 1604
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7922803920061394,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 1605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.03829103227264,
+      "learning_rate": 2e-05,
+      "loss": 0.122,
+      "step": 1606
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.948508239281492,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 1607
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2378356383498765,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 1608
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.95644468223719,
+      "learning_rate": 2e-05,
+      "loss": 0.238,
+      "step": 1609
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.232286056903911,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 1610
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5196781086075446,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 1611
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7052879720331435,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 1612
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4534892584850585,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 1613
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.547030422558458,
+      "learning_rate": 2e-05,
+      "loss": 0.0862,
+      "step": 1614
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0497488565824438,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 1615
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.254044500438009,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 1616
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9108578779381213,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 1617
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.212059536780295,
+      "learning_rate": 2e-05,
+      "loss": 0.0959,
+      "step": 1618
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9304992365033105,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 1619
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.164000923032896,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 1620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.741243211337208,
+      "learning_rate": 2e-05,
+      "loss": 0.0773,
+      "step": 1621
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.332157610048546,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 1622
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.436989768339083,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 1623
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9808166318752904,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 1624
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.547373374137045,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 1625
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7744577170096,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 1626
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.651946564634384,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 1627
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8777134652319634,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 1628
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.970499558393034,
+      "learning_rate": 2e-05,
+      "loss": 0.1365,
+      "step": 1629
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1392438138375436,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 1630
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.922030394992347,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 1631
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.147612581294415,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 1632
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0911900827200407,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 1633
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.104436418320993,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 1634
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.163606436556988,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 1635
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7334356044203463,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 1636
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.764379608088747,
+      "learning_rate": 2e-05,
+      "loss": 0.1481,
+      "step": 1637
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.579582483890009,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 1638
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.585176008896984,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 1639
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.101533910847594,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 1640
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1031756821487093,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 1641
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.125526138694129,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 1642
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5763247958171434,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 1643
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.246959714222683,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 1644
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7982282721143585,
+      "learning_rate": 2e-05,
+      "loss": 0.1237,
+      "step": 1645
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9761193452510819,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 1646
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3171420544973533,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 1647
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.518138343133798,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 1648
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.598095972494009,
+      "learning_rate": 2e-05,
+      "loss": 0.0675,
+      "step": 1649
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.849184142458927,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 1650
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7022566280188456,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 1651
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.260406897003872,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 1652
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.303652161930676,
+      "learning_rate": 2e-05,
+      "loss": 0.0858,
+      "step": 1653
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.637782807577912,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 1654
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.882667562855492,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 1655
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9097129135675615,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 1656
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8122420298629325,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 1657
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.336095858594673,
+      "learning_rate": 2e-05,
+      "loss": 0.0992,
+      "step": 1658
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9315924757592535,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 1659
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3885256689010443,
+      "learning_rate": 2e-05,
+      "loss": 0.1091,
+      "step": 1660
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.987188902213581,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 1661
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.233983182891434,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 1662
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6183629278597746,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 1663
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.169801632572511,
+      "learning_rate": 2e-05,
+      "loss": 0.1073,
+      "step": 1664
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8478551513072703,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 1665
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.154868399509732,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 1666
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.944906114199098,
+      "learning_rate": 2e-05,
+      "loss": 0.106,
+      "step": 1667
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.124082224069024,
+      "learning_rate": 2e-05,
+      "loss": 0.191,
+      "step": 1668
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1259135627535577,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 1669
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.326180046461263,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 1670
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8481689774018937,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 1671
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.446437101776876,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 1672
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3252015194505062,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 1673
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7159741141050606,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 1674
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9255003811368563,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 1675
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.3867367941943005,
+      "learning_rate": 2e-05,
+      "loss": 0.1539,
+      "step": 1676
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.435596424502334,
+      "learning_rate": 2e-05,
+      "loss": 0.0936,
+      "step": 1677
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.658001400863244,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 1678
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.42034292424214,
+      "learning_rate": 2e-05,
+      "loss": 0.1107,
+      "step": 1679
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.348154154300483,
+      "learning_rate": 2e-05,
+      "loss": 0.1467,
+      "step": 1680
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.817987855242746,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 1681
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.38963965760034,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 1682
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2048599229811088,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 1683
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.403361787513976,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 1684
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2898580354712403,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 1685
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8396076739086995,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 1686
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.4154668413847435,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 1687
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.892963382022188,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 1688
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9902321177801104,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 1689
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.262903420948287,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 1690
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9870907080170792,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 1691
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.634641546814171,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 1692
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.571247552019985,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 1693
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.885887284719099,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 1694
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.593170574456341,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 1695
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.876118654926681,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 1696
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3341858886174833,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 1697
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2249913076166674,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 1698
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1599643116154956,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 1699
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.531149825573733,
+      "learning_rate": 2e-05,
+      "loss": 0.1177,
+      "step": 1700
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.272992984134759,
+      "learning_rate": 2e-05,
+      "loss": 0.095,
+      "step": 1701
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.122485088609054,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 1702
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.961690600600286,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 1703
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.92388062278221,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 1704
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.687515258764227,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 1705
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.830375270441821,
+      "learning_rate": 2e-05,
+      "loss": 0.1173,
+      "step": 1706
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.102482690675845,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 1707
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5019471215348057,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 1708
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1487202406282098,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 1709
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.810884610869245,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 1710
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.633123846880343,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 1711
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4642059530478027,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 1712
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.719854270172261,
+      "learning_rate": 2e-05,
+      "loss": 0.1335,
+      "step": 1713
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.926131669360549,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 1714
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.783917455631816,
+      "learning_rate": 2e-05,
+      "loss": 0.1206,
+      "step": 1715
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.86092997125303,
+      "learning_rate": 2e-05,
+      "loss": 0.1256,
+      "step": 1716
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8797428270036822,
+      "learning_rate": 2e-05,
+      "loss": 0.1121,
+      "step": 1717
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.119730965431974,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 1718
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6885649872740807,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 1719
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.611529072210612,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 1720
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.308568074209535,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 1721
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.271442180054677,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 1722
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.363879217996425,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 1723
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5109183782651592,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 1724
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6876491254850469,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 1725
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5509074661817988,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 1726
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6126389790487647,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 1727
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.826870708896163,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 1728
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.543519554958365,
+      "learning_rate": 2e-05,
+      "loss": 0.1398,
+      "step": 1729
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.125837834063224,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 1730
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.022751240396833,
+      "learning_rate": 2e-05,
+      "loss": 0.1421,
+      "step": 1731
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.512552220954801,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 1732
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.18214280466554,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 1733
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.244379478365721,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 1734
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9176023526449213,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 1735
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5866072326235763,
+      "learning_rate": 2e-05,
+      "loss": 0.1135,
+      "step": 1736
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.045531286453736,
+      "learning_rate": 2e-05,
+      "loss": 0.0981,
+      "step": 1737
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1513614247539703,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 1738
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.083514748001231,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 1739
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.920313254869156,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 1740
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3311219006008947,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 1741
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.320791123896071,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 1742
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.551512522247206,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 1743
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.528293402194195,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 1744
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7626188900936832,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 1745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3506566358088996,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 1746
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5099439268193997,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 1747
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.7678235998950855,
+      "learning_rate": 2e-05,
+      "loss": 0.1217,
+      "step": 1748
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5240354855208893,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 1749
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.1056531616909115,
+      "learning_rate": 2e-05,
+      "loss": 0.1455,
+      "step": 1750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2481484640983114,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 1751
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8124692279403827,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 1752
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.7314300053952,
+      "learning_rate": 2e-05,
+      "loss": 0.1244,
+      "step": 1753
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7561480668593057,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 1754
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3351262624374092,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 1755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.176855527287913,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 1756
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.434593717766305,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 1757
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.940699049808253,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 1758
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.678049812423488,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 1759
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.624339849921753,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 1760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.130066083996558,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 1761
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2608976324228003,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 1762
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.555264514081317,
+      "learning_rate": 2e-05,
+      "loss": 0.1085,
+      "step": 1763
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5273148831306425,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 1764
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.504452745427062,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 1765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.6635476088541035,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 1766
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8341817333610217,
+      "learning_rate": 2e-05,
+      "loss": 0.0671,
+      "step": 1767
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7397687577346446,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 1768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7757019451391858,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 1769
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.574054675504421,
+      "learning_rate": 2e-05,
+      "loss": 0.0837,
+      "step": 1770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4265462416926764,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 1771
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.26650405141526,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 1772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.206492383301836,
+      "learning_rate": 2e-05,
+      "loss": 0.0864,
+      "step": 1773
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.991262113295084,
+      "learning_rate": 2e-05,
+      "loss": 0.1158,
+      "step": 1774
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1840500660150572,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 1775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.7833877471770725,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 1776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.487256088950284,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 1777
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.008815813653217,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 1778
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.448147163112687,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 1779
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7517299476505044,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 1780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.435367629637381,
+      "learning_rate": 2e-05,
+      "loss": 0.0996,
+      "step": 1781
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.0464071938682675,
+      "learning_rate": 2e-05,
+      "loss": 0.1055,
+      "step": 1782
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.784001733260241,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 1783
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.650485244837564,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 1784
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.626242187295505,
+      "learning_rate": 2e-05,
+      "loss": 0.0968,
+      "step": 1785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.33929437218426,
+      "learning_rate": 2e-05,
+      "loss": 0.1076,
+      "step": 1786
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.188692356068611,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 1787
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0607245140169692,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 1788
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.194935928817186,
+      "learning_rate": 2e-05,
+      "loss": 0.0979,
+      "step": 1789
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0223212384982037,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 1790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8471319571727576,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 1791
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8190667526336517,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 1792
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.739290574465088,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 1793
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6944367568392324,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 1794
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.224387112125393,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 1795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.855785681048055,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 1796
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6017167329176143,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 1797
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.289978469014459,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 1798
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.5502267131286045,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 1799
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.561607258391709,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 1800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1779980965948416,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 1801
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7584285750373825,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 1802
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0935093949901122,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 1803
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6056469449466175,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 1804
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1751506446320588,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 1805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1604439680041456,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 1806
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.855185516682951,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 1807
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1870997908953151,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 1808
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9919317151095575,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 1809
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.287265310731829,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 1810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.673505313627439,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 1811
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.196615744825442,
+      "learning_rate": 2e-05,
+      "loss": 0.1002,
+      "step": 1812
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.271463834541377,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 1813
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3776214060098328,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 1814
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.373373546725596,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 1815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.752120498362498,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 1816
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.375147795308622,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 1817
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0204162429407386,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 1818
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.921151627535737,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 1819
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.69659668895999,
+      "learning_rate": 2e-05,
+      "loss": 0.0921,
+      "step": 1820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8755571518563223,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 1821
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.2720884882302625,
+      "learning_rate": 2e-05,
+      "loss": 0.1037,
+      "step": 1822
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3868839687469583,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 1823
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.090003233911668,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 1824
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.455809779354052,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 1825
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.212600634360473,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 1826
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8136963765073193,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 1827
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8502617364458933,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 1828
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.527433749753564,
+      "learning_rate": 2e-05,
+      "loss": 0.1169,
+      "step": 1829
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.674791581867741,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 1830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.225883048508661,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 1831
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5462452541022527,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 1832
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.812341773828804,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 1833
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4775502253887267,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 1834
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6769841640682372,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 1835
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3454475867132576,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 1836
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.145051181055329,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 1837
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.235547118172431,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 1838
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.630584378502355,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 1839
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.093344236313111,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 1840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.16380036585745,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 1841
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9659605094692214,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 1842
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.566243658535487,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 1843
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5243577725832496,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 1844
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5804262715166475,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 1845
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.3671008051966265,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 1846
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.691009428471426,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 1847
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.248320079217596,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 1848
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 21.254377564958897,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 1849
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7439879279025063,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 1850
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3156989144185967,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 1851
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8122353383738634,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 1852
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4475664650510836,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 1853
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.757104885537276,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 1854
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4335595868074327,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 1855
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.942198169507086,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 1856
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8470950574244167,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 1857
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.444984837113741,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 1858
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.59412055564605,
+      "learning_rate": 2e-05,
+      "loss": 0.1393,
+      "step": 1859
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2719214668362753,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 1860
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8707846623173499,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 1861
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.959329986769279,
+      "learning_rate": 2e-05,
+      "loss": 0.0741,
+      "step": 1862
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5418431555960765,
+      "learning_rate": 2e-05,
+      "loss": 0.117,
+      "step": 1863
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.102229066000158,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 1864
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5599231323980454,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 1865
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.158555277406654,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 1866
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5315014379720635,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 1867
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.236995414227836,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 1868
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.460763758997898,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 1869
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9835454776571675,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 1870
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1719467079813817,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 1871
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5045587090408954,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 1872
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.057732588747674,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 1873
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.7046275810924225,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 1874
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.868362393923748,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 1875
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.330376664816342,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 1876
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8478678976560694,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 1877
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.100402784334048,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 1878
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.871210152331964,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 1879
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.788729245976335,
+      "learning_rate": 2e-05,
+      "loss": 0.1137,
+      "step": 1880
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.544766805812358,
+      "learning_rate": 2e-05,
+      "loss": 0.1191,
+      "step": 1881
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.704943739324622,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 1882
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3553428299934933,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 1883
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9584304542896818,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 1884
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.005247999878028,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 1885
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1636552713441013,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 1886
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.517311867040545,
+      "learning_rate": 2e-05,
+      "loss": 0.1216,
+      "step": 1887
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.798745409002112,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 1888
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.285344080603115,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 1889
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.329736728714853,
+      "learning_rate": 2e-05,
+      "loss": 0.0944,
+      "step": 1890
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.740480919339662,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 1891
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4922075519912117,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 1892
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.786530771660123,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 1893
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6303758464057339,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 1894
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8436170105296408,
+      "learning_rate": 2e-05,
+      "loss": 0.108,
+      "step": 1895
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.713836702695146,
+      "learning_rate": 2e-05,
+      "loss": 0.1043,
+      "step": 1896
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.02316135088428,
+      "learning_rate": 2e-05,
+      "loss": 0.1179,
+      "step": 1897
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8511645747946583,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 1898
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.327714000446257,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 1899
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.267549084832476,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 1900
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.570485064517286,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 1901
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.10506151516189,
+      "learning_rate": 2e-05,
+      "loss": 0.1107,
+      "step": 1902
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3265805668158843,
+      "learning_rate": 2e-05,
+      "loss": 0.1251,
+      "step": 1903
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4109928813357904,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 1904
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.489672746445465,
+      "learning_rate": 2e-05,
+      "loss": 0.0883,
+      "step": 1905
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.112057243016824,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 1906
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.812307742215236,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 1907
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.254936723694694,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 1908
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8036269998150034,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 1909
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.880834309651063,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 1910
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8662881441204946,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 1911
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.835182734984116,
+      "learning_rate": 2e-05,
+      "loss": 0.1531,
+      "step": 1912
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.26141305975719,
+      "learning_rate": 2e-05,
+      "loss": 0.1672,
+      "step": 1913
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.097124952166893,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 1914
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.408875879384194,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 1915
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9894208343329067,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 1916
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8009202545560123,
+      "learning_rate": 2e-05,
+      "loss": 0.0918,
+      "step": 1917
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4552851345913735,
+      "learning_rate": 2e-05,
+      "loss": 0.1291,
+      "step": 1918
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7803640036942534,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 1919
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4201674810101297,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 1920
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4605195778772262,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 1921
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1918855304891642,
+      "learning_rate": 2e-05,
+      "loss": 0.0883,
+      "step": 1922
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3739982770772956,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 1923
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.491931038874908,
+      "learning_rate": 2e-05,
+      "loss": 0.1169,
+      "step": 1924
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7618574784761787,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 1925
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.620990768149325,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 1926
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8618606494095695,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 1927
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.833678299342146,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 1928
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7743759072085945,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 1929
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.754806739273347,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 1930
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2803921986074425,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 1931
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.046222949559479,
+      "learning_rate": 2e-05,
+      "loss": 0.0941,
+      "step": 1932
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.790323012384038,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 1933
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.819838296434015,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 1934
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.073764869309583,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 1935
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.05022895316716,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 1936
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.477726226701829,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 1937
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7674791403159196,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 1938
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.022411741588878,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 1939
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6307716426727209,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 1940
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.592543832469398,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 1941
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9676170116563148,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 1942
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.63539645872819,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 1943
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.910017463395721,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 1944
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.01183907630903,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 1945
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.925889472039105,
+      "learning_rate": 2e-05,
+      "loss": 0.1008,
+      "step": 1946
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7224410690376546,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 1947
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7720408186831516,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 1948
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9789936900197875,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 1949
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.669174956210465,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 1950
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.75259838273768,
+      "learning_rate": 2e-05,
+      "loss": 0.1056,
+      "step": 1951
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.116561482604429,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 1952
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5055043659587053,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 1953
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.045988125627623,
+      "learning_rate": 2e-05,
+      "loss": 0.0853,
+      "step": 1954
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5767128546618463,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 1955
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.389600219621466,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 1956
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9198029055363879,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 1957
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.634497710730011,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 1958
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.130866824258177,
+      "learning_rate": 2e-05,
+      "loss": 0.1315,
+      "step": 1959
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.7836447309798675,
+      "learning_rate": 2e-05,
+      "loss": 0.1102,
+      "step": 1960
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.408325263078333,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 1961
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7221536400514987,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 1962
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2813043815422958,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 1963
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.945564161820581,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 1964
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.938462518184109,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 1965
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1725463927202404,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 1966
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.903168700821523,
+      "learning_rate": 2e-05,
+      "loss": 0.0986,
+      "step": 1967
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.274677999995188,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 1968
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.387202772775992,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 1969
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5711534046356856,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 1970
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.982904516210962,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 1971
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.108775105221047,
+      "learning_rate": 2e-05,
+      "loss": 0.1576,
+      "step": 1972
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.560610332334608,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 1973
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.669584612781227,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 1974
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.848804027534679,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 1975
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8990535516889206,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 1976
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.069892835206682,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 1977
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.737278416265281,
+      "learning_rate": 2e-05,
+      "loss": 0.1391,
+      "step": 1978
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.329584089219208,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 1979
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.817308321399541,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 1980
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.691495896117371,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 1981
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.50916341893559,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 1982
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.513261750826084,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 1983
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.318428275109293,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 1984
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.833152745698902,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 1985
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.841632969532784,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 1986
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.639783789431453,
+      "learning_rate": 2e-05,
+      "loss": 0.1338,
+      "step": 1987
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.245575993623847,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 1988
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1380054567800886,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 1989
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.088002712896811,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 1990
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7665600387424676,
+      "learning_rate": 2e-05,
+      "loss": 0.0991,
+      "step": 1991
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.106393299870172,
+      "learning_rate": 2e-05,
+      "loss": 0.1055,
+      "step": 1992
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.215148243905037,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 1993
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8895681911255742,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 1994
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4866090463675325,
+      "learning_rate": 2e-05,
+      "loss": 0.1252,
+      "step": 1995
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6619683641803706,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 1996
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.405643120254953,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 1997
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6361160154557193,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 1998
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.757780015452109,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 1999
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.684517122719381,
+      "learning_rate": 2e-05,
+      "loss": 0.1191,
+      "step": 2000
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.045075370293625,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 2001
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.8187609189915515,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 2002
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.304791402514765,
+      "learning_rate": 2e-05,
+      "loss": 0.0976,
+      "step": 2003
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0057170932604182,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 2004
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.162172101034959,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 2005
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.197349011430296,
+      "learning_rate": 2e-05,
+      "loss": 0.1503,
+      "step": 2006
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.6860581278498765,
+      "learning_rate": 2e-05,
+      "loss": 0.1115,
+      "step": 2007
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.50516041070427,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 2008
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7326065532730806,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 2009
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.088005605767735,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 2010
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.5747032527885585,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 2011
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8593859961966954,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 2012
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.7193001455360495,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 2013
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1933660057015203,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 2014
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.224711860898585,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 2015
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.042780530364194,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 2016
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5786175118191155,
+      "learning_rate": 2e-05,
+      "loss": 0.0706,
+      "step": 2017
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.471718271256532,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 2018
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8292389163789373,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 2019
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1147410767697647,
+      "learning_rate": 2e-05,
+      "loss": 0.0998,
+      "step": 2020
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1878853359302406,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 2021
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7018286294605323,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 2022
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9013027753991247,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 2023
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.459549402303111,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 2024
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.456157474762463,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 2025
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9206103038974476,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 2026
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7147065196816964,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 2027
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8061368332625105,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 2028
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0449419198345993,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 2029
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.079683794782105,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 2030
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.727664040960573,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 2031
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.365156598859697,
+      "learning_rate": 2e-05,
+      "loss": 0.1057,
+      "step": 2032
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.715434140943254,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 2033
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.876040106615106,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 2034
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5293572074126405,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 2035
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.414648978586498,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 2036
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.977588933289739,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 2037
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.544359029008443,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 2038
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.367408714566471,
+      "learning_rate": 2e-05,
+      "loss": 0.0873,
+      "step": 2039
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4515891133254204,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 2040
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4871809366606668,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 2041
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.820812788155416,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 2042
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8253554933193534,
+      "learning_rate": 2e-05,
+      "loss": 0.1147,
+      "step": 2043
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8237507319824757,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 2044
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.455797913857587,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 2045
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.553911674698256,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 2046
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.575694095482069,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 2047
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.107596831671178,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 2048
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.59590455858069,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 2049
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1635990509834366,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 2050
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6561552056589754,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 2051
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.697792384117823,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 2052
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.387563228025701,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 2053
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.452782212990419,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 2054
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.920739615137683,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 2055
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.373840384369875,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 2056
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.06887232828917,
+      "learning_rate": 2e-05,
+      "loss": 0.1587,
+      "step": 2057
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.26178273758022,
+      "learning_rate": 2e-05,
+      "loss": 0.1467,
+      "step": 2058
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.157416846750694,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 2059
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1978106962815187,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 2060
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0666108858868046,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 2061
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.29653284547632,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 2062
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.466575630759097,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 2063
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5339843314811714,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 2064
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.557246708264322,
+      "learning_rate": 2e-05,
+      "loss": 0.107,
+      "step": 2065
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8152718342628047,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 2066
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5575998143868515,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 2067
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9151139664900523,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 2068
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.791856759186304,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 2069
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.68684795239737,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 2070
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.315044813221793,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 2071
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6908927439213604,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 2072
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5142257305520364,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 2073
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.492872472880661,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 2074
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.878892940787806,
+      "learning_rate": 2e-05,
+      "loss": 0.1188,
+      "step": 2075
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.38305055290033,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 2076
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.185243137178653,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 2077
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5613656313172817,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 2078
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.018959412524775,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 2079
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.08075195372864,
+      "learning_rate": 2e-05,
+      "loss": 0.1111,
+      "step": 2080
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6051437902955654,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 2081
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.353320405706644,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 2082
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.328598920005784,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 2083
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.207221531617475,
+      "learning_rate": 2e-05,
+      "loss": 0.1021,
+      "step": 2084
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6097044691179523,
+      "learning_rate": 2e-05,
+      "loss": 0.0955,
+      "step": 2085
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2021179104693864,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 2086
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7205838910120401,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 2087
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7014866219128317,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 2088
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.528658639918095,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 2089
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.958733593255132,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 2090
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6084306115808658,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 2091
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.090724524166134,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 2092
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.121076638720303,
+      "learning_rate": 2e-05,
+      "loss": 0.0919,
+      "step": 2093
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0988085456786445,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 2094
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1265864340775766,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 2095
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.699055425226767,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 2096
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.962429989015662,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 2097
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.268161341012094,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 2098
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.544236390197283,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 2099
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.227281984372117,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 2100
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.636314615324543,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 2101
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.427012986764931,
+      "learning_rate": 2e-05,
+      "loss": 0.1337,
+      "step": 2102
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.536180431497902,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 2103
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9349042054547556,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 2104
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3484241924302247,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 2105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0404948411548665,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 2106
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.075644942098523,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 2107
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.656787593519218,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 2108
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9630497958046595,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 2109
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6435253845612685,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 2110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.98267096158183,
+      "learning_rate": 2e-05,
+      "loss": 0.0904,
+      "step": 2111
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3482010334558152,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 2112
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4134749145413867,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 2113
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2533588581242263,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 2114
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.365642282480783,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 2115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.902381373042233,
+      "learning_rate": 2e-05,
+      "loss": 0.1157,
+      "step": 2116
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.025168967371621,
+      "learning_rate": 2e-05,
+      "loss": 0.1182,
+      "step": 2117
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8544606893635613,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 2118
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.396193040456067,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 2119
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6657376557426993,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 2120
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.701194660599059,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 2121
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.1913390917098985,
+      "learning_rate": 2e-05,
+      "loss": 0.0868,
+      "step": 2122
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.045740147598265,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 2123
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.273190460070054,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 2124
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1150696713905432,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 2125
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.969003085868746,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 2126
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.143099760206156,
+      "learning_rate": 2e-05,
+      "loss": 0.1076,
+      "step": 2127
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.919865661275722,
+      "learning_rate": 2e-05,
+      "loss": 0.1167,
+      "step": 2128
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1051507543371635,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 2129
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.03229699410983,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 2130
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3606000219193324,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 2131
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.111929438681599,
+      "learning_rate": 2e-05,
+      "loss": 0.1078,
+      "step": 2132
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.782512223925525,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 2133
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.293315477723045,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 2134
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0194699773456595,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 2135
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.054820622664142,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 2136
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.968759586666057,
+      "learning_rate": 2e-05,
+      "loss": 0.1316,
+      "step": 2137
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0739985569779062,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 2138
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0476054563105746,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 2139
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.505532524152795,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 2140
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7954536455545047,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 2141
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.93763034612527,
+      "learning_rate": 2e-05,
+      "loss": 0.1019,
+      "step": 2142
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.313793439246099,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 2143
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5589742781590394,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 2144
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.998063500800094,
+      "learning_rate": 2e-05,
+      "loss": 0.1224,
+      "step": 2145
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6782526997167597,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 2146
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1758672673260087,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 2147
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.482840105867168,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 2148
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.647678001408553,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 2149
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5251910373445905,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 2150
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8849601716070583,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 2151
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6904586710172356,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 2152
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5279658169608252,
+      "learning_rate": 2e-05,
+      "loss": 0.1292,
+      "step": 2153
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.909485031653207,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 2154
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.440392963297116,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 2155
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1218897791484976,
+      "learning_rate": 2e-05,
+      "loss": 0.0763,
+      "step": 2156
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7304526300303782,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 2157
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7198937301351975,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 2158
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.222115537622948,
+      "learning_rate": 2e-05,
+      "loss": 0.121,
+      "step": 2159
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.441170276254609,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 2160
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0426670850452044,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 2161
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.806499267925994,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 2162
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4792603566044282,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 2163
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.489203535095797,
+      "learning_rate": 2e-05,
+      "loss": 0.1251,
+      "step": 2164
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5014915013126404,
+      "learning_rate": 2e-05,
+      "loss": 0.1151,
+      "step": 2165
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.951319223953866,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 2166
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.51057144468781,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 2167
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3923592629530075,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 2168
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4254577553577032,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 2169
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.760549486411099,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 2170
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8830632086389167,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 2171
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.4423455660053754,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 2172
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5175006579310173,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 2173
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.88509246680327,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 2174
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.345515489021704,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 2175
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7020809687641597,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 2176
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6057777507511677,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 2177
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.714153751187931,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 2178
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.789865391426505,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 2179
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9754768134900061,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 2180
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.501391798274001,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 2181
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.446290307762243,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 2182
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9016287807383745,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 2183
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.129024203359565,
+      "learning_rate": 2e-05,
+      "loss": 0.1524,
+      "step": 2184
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.103569707285228,
+      "learning_rate": 2e-05,
+      "loss": 0.0906,
+      "step": 2185
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1209821302856295,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 2186
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2045615347699212,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 2187
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0136807355670134,
+      "learning_rate": 2e-05,
+      "loss": 0.1058,
+      "step": 2188
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.755983864874039,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 2189
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3617534648780607,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 2190
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7566460702449933,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 2191
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.396785811421571,
+      "learning_rate": 2e-05,
+      "loss": 0.0902,
+      "step": 2192
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.825624909952555,
+      "learning_rate": 2e-05,
+      "loss": 0.1084,
+      "step": 2193
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.434266537609405,
+      "learning_rate": 2e-05,
+      "loss": 0.1181,
+      "step": 2194
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.0855665688125615,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 2195
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.894192002587548,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 2196
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.139755763730483,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 2197
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8028075311663296,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 2198
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3971184418124956,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 2199
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6182503890561115,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 2200
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.068912939932766,
+      "learning_rate": 2e-05,
+      "loss": 0.1309,
+      "step": 2201
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.714077024302674,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 2202
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.585099928822538,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 2203
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.496217454820859,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 2204
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7186753712200304,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 2205
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.364080014802961,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 2206
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.622373535068029,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 2207
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.563845118595266,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 2208
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5678595216183738,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 2209
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.838087031502585,
+      "learning_rate": 2e-05,
+      "loss": 0.1294,
+      "step": 2210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.925464534671742,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 2211
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.405386951685587,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 2212
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.053693859421627,
+      "learning_rate": 2e-05,
+      "loss": 0.1508,
+      "step": 2213
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2609904274482666,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 2214
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.87151342721824,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 2215
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.289540057876183,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 2216
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.95645422199845,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 2217
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.757559409719131,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 2218
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1674891035780637,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 2219
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.577766633723919,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 2220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.85202123697352,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 2221
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3178247342571894,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 2222
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7678680806449263,
+      "learning_rate": 2e-05,
+      "loss": 0.0775,
+      "step": 2223
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.398379825193581,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 2224
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.715020520390563,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 2225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.534525802674614,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 2226
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.536540974135444,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 2227
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.3010100993003,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 2228
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.465351398699313,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 2229
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8797009164038183,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 2230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3078658627432704,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 2231
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.692690605139648,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 2232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.864898188978519,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 2233
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.928062568138732,
+      "learning_rate": 2e-05,
+      "loss": 0.1252,
+      "step": 2234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3300592237366815,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 2235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.837914148431412,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 2236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.444129438284742,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 2237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7730873213955025,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 2238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.618442974437687,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 2239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.680354161220651,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 2240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.271795715206177,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 2241
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.733526099461232,
+      "learning_rate": 2e-05,
+      "loss": 0.1396,
+      "step": 2242
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.923139818572793,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 2243
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1511497439618807,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 2244
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.636061206350161,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 2245
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1636955142595187,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 2246
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.429924434185241,
+      "learning_rate": 2e-05,
+      "loss": 0.2222,
+      "step": 2247
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.312037567555365,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 2248
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0535004399588046,
+      "learning_rate": 2e-05,
+      "loss": 0.1267,
+      "step": 2249
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.330634646168923,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 2250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6727646929292734,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 2251
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8166291337832554,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 2252
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0284348991868857,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 2253
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.879091177531649,
+      "learning_rate": 2e-05,
+      "loss": 0.1536,
+      "step": 2254
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1143126269086343,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 2255
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.233843893266357,
+      "learning_rate": 2e-05,
+      "loss": 0.093,
+      "step": 2256
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.866481440016783,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 2257
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1550018457462348,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 2258
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.542927517851819,
+      "learning_rate": 2e-05,
+      "loss": 0.1258,
+      "step": 2259
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9414496821117293,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 2260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7165956402096367,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 2261
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.87982037229289,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 2262
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.8085403997218155,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 2263
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.654610421100598,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 2264
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5518517110127656,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 2265
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9821986706342707,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 2266
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5849816898309887,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 2267
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.183202281930583,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 2268
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.215405912556739,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 2269
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.651781136186665,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 2270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5672073049576145,
+      "learning_rate": 2e-05,
+      "loss": 0.1038,
+      "step": 2271
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2512989382817103,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 2272
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0997754976069087,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 2273
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0499636809690647,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 2274
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6784074828799787,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 2275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8226666122740443,
+      "learning_rate": 2e-05,
+      "loss": 0.1011,
+      "step": 2276
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.616500872913834,
+      "learning_rate": 2e-05,
+      "loss": 0.0993,
+      "step": 2277
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.029952244978835,
+      "learning_rate": 2e-05,
+      "loss": 0.1318,
+      "step": 2278
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.087061749722855,
+      "learning_rate": 2e-05,
+      "loss": 0.1826,
+      "step": 2279
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.843654677583141,
+      "learning_rate": 2e-05,
+      "loss": 0.1272,
+      "step": 2280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9218391477140535,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 2281
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.106114810941476,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 2282
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.206599618062039,
+      "learning_rate": 2e-05,
+      "loss": 0.1091,
+      "step": 2283
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.865032173407752,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 2284
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.156066223274013,
+      "learning_rate": 2e-05,
+      "loss": 0.082,
+      "step": 2285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.479578090964246,
+      "learning_rate": 2e-05,
+      "loss": 0.0776,
+      "step": 2286
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.256347628242586,
+      "learning_rate": 2e-05,
+      "loss": 0.1032,
+      "step": 2287
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2842466973256554,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 2288
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0093051922110132,
+      "learning_rate": 2e-05,
+      "loss": 0.0929,
+      "step": 2289
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4240881495355915,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 2290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.172224805362835,
+      "learning_rate": 2e-05,
+      "loss": 0.0763,
+      "step": 2291
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1090245648453814,
+      "learning_rate": 2e-05,
+      "loss": 0.0736,
+      "step": 2292
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0132731830167376,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 2293
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5975455765160036,
+      "learning_rate": 2e-05,
+      "loss": 0.1058,
+      "step": 2294
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.273700824553933,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 2295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.263164545356163,
+      "learning_rate": 2e-05,
+      "loss": 0.0882,
+      "step": 2296
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.177595127324097,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 2297
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8272297289088342,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 2298
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9127772161491743,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 2299
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.980677004968315,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 2300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.274893571128943,
+      "learning_rate": 2e-05,
+      "loss": 0.0732,
+      "step": 2301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.526575031467272,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 2302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8392705004058891,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 2303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.645075349884382,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 2304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.051248551050366,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 2305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.13783840155388,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 2306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.954231983174747,
+      "learning_rate": 2e-05,
+      "loss": 0.1037,
+      "step": 2307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5393042071330747,
+      "learning_rate": 2e-05,
+      "loss": 0.1257,
+      "step": 2308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6274234846971725,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 2309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.24595067714051,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 2310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.164394333004497,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 2311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.00297865706277,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 2312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3446357480828186,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 2313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8604262424579617,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 2314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.139215046784203,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 2315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1043127910923012,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 2316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6477942869760644,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 2317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.976128654201196,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 2318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.078020742751754,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 2319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.365937137572344,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 2320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3724243576106394,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 2321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.8536001417673145,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 2322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.651578881764946,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 2323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2092974481619505,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 2324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.582962540755362,
+      "learning_rate": 2e-05,
+      "loss": 0.1789,
+      "step": 2325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.014268935594659,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 2326
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.875135940594932,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 2327
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6717705957192326,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 2328
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.199424733839909,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 2329
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.422628762219186,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 2330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6453530684197717,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 2331
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4373546222809783,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 2332
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.788857426997454,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 2333
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8514887155245099,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 2334
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0947777959862603,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 2335
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.088936037013275,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 2336
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7874438235515479,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 2337
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8358881926923754,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 2338
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.537362552082275,
+      "learning_rate": 2e-05,
+      "loss": 0.0932,
+      "step": 2339
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.973728931583545,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 2340
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.41074127449524,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 2341
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.287356289159476,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 2342
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.370209672434294,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 2343
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.407306686832102,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 2344
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.028129528258078,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 2345
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.920357459955733,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 2346
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5154757997854222,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 2347
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8457389061098506,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 2348
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.284405163191803,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 2349
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.067656088590873,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 2350
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9855623928866146,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 2351
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7262914134801917,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 2352
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.597548954404156,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 2353
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.940811380823466,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 2354
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.2493664229645365,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 2355
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.24513916960612,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 2356
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.012346277532803,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 2357
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.071567735022309,
+      "learning_rate": 2e-05,
+      "loss": 0.1186,
+      "step": 2358
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2757482998749006,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 2359
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5895936134566027,
+      "learning_rate": 2e-05,
+      "loss": 0.091,
+      "step": 2360
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.869449897441707,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 2361
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.790048875921333,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 2362
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.3003816479604735,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 2363
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.958331633450327,
+      "learning_rate": 2e-05,
+      "loss": 0.1167,
+      "step": 2364
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.0703106191011535,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 2365
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.047502508214787,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 2366
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.434722030856816,
+      "learning_rate": 2e-05,
+      "loss": 0.1017,
+      "step": 2367
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.465926276436222,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 2368
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4585944169078215,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 2369
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3302695618589953,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 2370
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.564266124285031,
+      "learning_rate": 2e-05,
+      "loss": 0.1083,
+      "step": 2371
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.6212501789616125,
+      "learning_rate": 2e-05,
+      "loss": 0.1073,
+      "step": 2372
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.048215327898184,
+      "learning_rate": 2e-05,
+      "loss": 0.0941,
+      "step": 2373
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.5274406468815425,
+      "learning_rate": 2e-05,
+      "loss": 0.1047,
+      "step": 2374
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2675990968447617,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 2375
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3736133542939073,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 2376
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.193039830847221,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 2377
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.116359514683096,
+      "learning_rate": 2e-05,
+      "loss": 0.0873,
+      "step": 2378
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.007817832241701,
+      "learning_rate": 2e-05,
+      "loss": 0.1218,
+      "step": 2379
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1071314430389503,
+      "learning_rate": 2e-05,
+      "loss": 0.0938,
+      "step": 2380
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.228305388822024,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 2381
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3001784998850465,
+      "learning_rate": 2e-05,
+      "loss": 0.1264,
+      "step": 2382
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.300795118314061,
+      "learning_rate": 2e-05,
+      "loss": 0.0992,
+      "step": 2383
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1450733826718187,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 2384
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.639905381727937,
+      "learning_rate": 2e-05,
+      "loss": 0.1405,
+      "step": 2385
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.788236602319174,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 2386
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4533558147720305,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 2387
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5004458029948577,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 2388
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7867257723707777,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 2389
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8349112155676834,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 2390
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9732358175531897,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 2391
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8777527629094017,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 2392
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.906698540970676,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 2393
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8054947820277376,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 2394
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.871694816479357,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 2395
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7643344297951025,
+      "learning_rate": 2e-05,
+      "loss": 0.0811,
+      "step": 2396
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.980019663318136,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 2397
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1383454374157598,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 2398
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.04705810132653,
+      "learning_rate": 2e-05,
+      "loss": 0.0838,
+      "step": 2399
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.206861533332127,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 2400
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.899958801051922,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 2401
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3702508231781545,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 2402
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7638070528899181,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 2403
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8680592820115798,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 2404
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.89990047914731,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 2405
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7946712839359943,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 2406
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.3127941086587525,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 2407
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.86684459378444,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 2408
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0860214028843123,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 2409
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.840120283722713,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 2410
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.146325674806852,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 2411
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1086986323684784,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 2412
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2492511350758804,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 2413
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.339322955446209,
+      "learning_rate": 2e-05,
+      "loss": 0.121,
+      "step": 2414
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34665948513429634,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 2415
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6339340611607494,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 2416
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.740919308409858,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 2417
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2851758450571222,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 2418
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.062434738808745,
+      "learning_rate": 2e-05,
+      "loss": 0.128,
+      "step": 2419
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6723227369973857,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 2420
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.879036185709216,
+      "learning_rate": 2e-05,
+      "loss": 0.1136,
+      "step": 2421
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2960078106852455,
+      "learning_rate": 2e-05,
+      "loss": 0.0703,
+      "step": 2422
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0410088825089834,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 2423
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.364876287396066,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 2424
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.92211060944007,
+      "learning_rate": 2e-05,
+      "loss": 0.0776,
+      "step": 2425
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8591052588981398,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 2426
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7514400179808365,
+      "learning_rate": 2e-05,
+      "loss": 0.0893,
+      "step": 2427
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.659631717192697,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 2428
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5593011009344213,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 2429
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0241665219975893,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 2430
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.032615589979936,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 2431
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4739549067958726,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 2432
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9666748008990849,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 2433
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.147903131994053,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 2434
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9972320714975045,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 2435
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5344808234194645,
+      "learning_rate": 2e-05,
+      "loss": 0.1124,
+      "step": 2436
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9195219818918646,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 2437
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2170384212428726,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 2438
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0701429765500197,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 2439
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7306501254582347,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 2440
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2432630132656737,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 2441
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7458975922557576,
+      "learning_rate": 2e-05,
+      "loss": 0.2185,
+      "step": 2442
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5528621472350794,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 2443
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.91919472997175,
+      "learning_rate": 2e-05,
+      "loss": 0.0966,
+      "step": 2444
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.9729529976506,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 2445
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.120697939984624,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 2446
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5962083593327685,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 2447
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8690607737482363,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 2448
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3598859418696534,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 2449
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.218512294017334,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 2450
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.003135488307114,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 2451
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0353300931929534,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 2452
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5934669714891085,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 2453
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5399090912231044,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 2454
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7666232103749833,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 2455
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9942574312708365,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 2456
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.923004584826627,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 2457
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2214049713314292,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 2458
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6256034170886418,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 2459
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.052805542530816,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 2460
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.948766207173827,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 2461
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.74359129705232,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 2462
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1794760306082868,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 2463
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.47276142935135,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 2464
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.79613951188247,
+      "learning_rate": 2e-05,
+      "loss": 0.1662,
+      "step": 2465
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.098308249281914,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 2466
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.258027615050606,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 2467
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.991468378011144,
+      "learning_rate": 2e-05,
+      "loss": 0.1164,
+      "step": 2468
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.843678508295958,
+      "learning_rate": 2e-05,
+      "loss": 0.1133,
+      "step": 2469
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0631658461860036,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 2470
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.372548098555197,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 2471
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.986990217184541,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 2472
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.005390175504997,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 2473
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6396506052294098,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 2474
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3188891441375676,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 2475
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0851853625038967,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 2476
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9775259952906517,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 2477
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2733974333468694,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 2478
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2109940053236574,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 2479
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.603233221948723,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 2480
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8627584609172534,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 2481
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1351642986750283,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 2482
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.874679074484436,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 2483
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0236169727295685,
+      "learning_rate": 2e-05,
+      "loss": 0.1055,
+      "step": 2484
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.481885478186072,
+      "learning_rate": 2e-05,
+      "loss": 0.1021,
+      "step": 2485
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.580078125,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 2486
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.377139023753128,
+      "learning_rate": 2e-05,
+      "loss": 0.1056,
+      "step": 2487
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.388019812752531,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 2488
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.221343767530123,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 2489
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.499457586573795,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 2490
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1398136801137166,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 2491
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9093847288958576,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 2492
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8710057630139258,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 2493
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1990764023023033,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 2494
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.5886297368331475,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 2495
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9062907768809751,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 2496
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3220060874204387,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 2497
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7399559295832097,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 2498
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.5073055729373435,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 2499
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.8203358456434255,
+      "learning_rate": 2e-05,
+      "loss": 0.1032,
+      "step": 2500
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9786677551527876,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 2501
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3468091999214282,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 2502
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.760974659693958,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 2503
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6170261421218632,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 2504
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0882934847993955,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 2505
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9990323825444103,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 2506
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.703616702715005,
+      "learning_rate": 2e-05,
+      "loss": 0.083,
+      "step": 2507
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.399209965958088,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 2508
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.112089572794176,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 2509
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.286962109515078,
+      "learning_rate": 2e-05,
+      "loss": 0.2024,
+      "step": 2510
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.413642900068575,
+      "learning_rate": 2e-05,
+      "loss": 0.0884,
+      "step": 2511
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1166884618923616,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 2512
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.3146253130958225,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 2513
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7603595506179361,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 2514
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.095325446269294,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 2515
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.6658927538599055,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 2516
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.234631523564959,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 2517
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.383409314387182,
+      "learning_rate": 2e-05,
+      "loss": 0.1769,
+      "step": 2518
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.9502574191787865,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 2519
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5408328954855237,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 2520
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9915822624400419,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 2521
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2599723114990904,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 2522
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3784777772044694,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 2523
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.124002803669792,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 2524
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4646672256918947,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 2525
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.376287439211962,
+      "learning_rate": 2e-05,
+      "loss": 0.101,
+      "step": 2526
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.822108794976615,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 2527
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.304531596454871,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 2528
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.844628235969297,
+      "learning_rate": 2e-05,
+      "loss": 0.1008,
+      "step": 2529
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1902008581753796,
+      "learning_rate": 2e-05,
+      "loss": 0.1013,
+      "step": 2530
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.773066080131078,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 2531
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6977723729267824,
+      "learning_rate": 2e-05,
+      "loss": 0.0732,
+      "step": 2532
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.125844226760897,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 2533
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7426080581019896,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 2534
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.988107310183127,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 2535
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.415022307025667,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 2536
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3927061388582884,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 2537
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9028363539813293,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 2538
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5663767016420374,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 2539
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7009620612600185,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 2540
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4260250353441284,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 2541
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3345993181047535,
+      "learning_rate": 2e-05,
+      "loss": 0.0815,
+      "step": 2542
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.50276289998935,
+      "learning_rate": 2e-05,
+      "loss": 0.0919,
+      "step": 2543
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.05426599446882,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 2544
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.527174034493164,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 2545
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6563193828517053,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 2546
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4968900487608354,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 2547
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8245474215005877,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 2548
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4717872544879387,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 2549
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.517902835713482,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 2550
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.414767220831024,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 2551
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7005667515489287,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 2552
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7546479524790259,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 2553
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.31954865426869,
+      "learning_rate": 2e-05,
+      "loss": 0.0947,
+      "step": 2554
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.644720749707444,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 2555
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2952324839668297,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 2556
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.398295143194176,
+      "learning_rate": 2e-05,
+      "loss": 0.0736,
+      "step": 2557
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.332675992669895,
+      "learning_rate": 2e-05,
+      "loss": 0.0868,
+      "step": 2558
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.525069755010293,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 2559
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.5236859515262875,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 2560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5220390748360433,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 2561
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2056005110858026,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 2562
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.823168878528647,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 2563
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.756271777864018,
+      "learning_rate": 2e-05,
+      "loss": 0.1065,
+      "step": 2564
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.081897680469408,
+      "learning_rate": 2e-05,
+      "loss": 0.1622,
+      "step": 2565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0622067098449754,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 2566
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.830934546709182,
+      "learning_rate": 2e-05,
+      "loss": 0.1371,
+      "step": 2567
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6055996016968517,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 2568
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.344431637715241,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 2569
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.747140046077214,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 2570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.287546247672173,
+      "learning_rate": 2e-05,
+      "loss": 0.1078,
+      "step": 2571
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4708973879053473,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 2572
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.077304413578037,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 2573
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.636462672938682,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 2574
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.062343430337616,
+      "learning_rate": 2e-05,
+      "loss": 0.1495,
+      "step": 2575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.593213103543908,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 2576
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6663092547881901,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 2577
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7734794506070264,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 2578
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.481439149056606,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 2579
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.080399503783132,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 2580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3896314763011675,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 2581
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.074853044914195,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 2582
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3594696834335966,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 2583
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1761257371477765,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 2584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.396651705967982,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 2585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.711813393255127,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 2586
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.207212187515408,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 2587
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9775657688269379,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 2588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.802494307281674,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 2589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2875011661658013,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 2590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6121418611637917,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 2591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.721674792625888,
+      "learning_rate": 2e-05,
+      "loss": 0.0754,
+      "step": 2592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.889582430494003,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 2593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1523234787001955,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 2594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1806381266692796,
+      "learning_rate": 2e-05,
+      "loss": 0.057,
+      "step": 2595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2086477909966185,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 2596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6791091108014355,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 2597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8379958068300524,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 2598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8449304325665323,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 2599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.107010598961733,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 2600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.989559289139515,
+      "learning_rate": 2e-05,
+      "loss": 0.0986,
+      "step": 2601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4573476614938845,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 2602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.433329378536594,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 2603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.989543174624567,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 2604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5299462806569353,
+      "learning_rate": 2e-05,
+      "loss": 0.0858,
+      "step": 2605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.352600521743281,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 2606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.774913301798624,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 2607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.687089933116267,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 2608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1135884714356081,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 2609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2057213657923636,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 2610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9468546809671519,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 2611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2778630654202874,
+      "learning_rate": 2e-05,
+      "loss": 0.0792,
+      "step": 2612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4411550576930752,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 2613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6912713646648184,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 2614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5716738453156682,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 2615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.942319413459897,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 2616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.228789865529693,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 2617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.207775157264669,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 2618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.500307069659629,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 2619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7732169909863633,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 2620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3836095258652694,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 2621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9602703160119954,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 2622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.874963329510837,
+      "learning_rate": 2e-05,
+      "loss": 0.0987,
+      "step": 2623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.069833721724912,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 2624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4084899440944019,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 2625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.376103897539642,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 2626
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3136437139515285,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 2627
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.5032004525443785,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 2628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5205161365370103,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 2629
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6564436025647788,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 2630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.531715166709761,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 2631
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4893874815475083,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 2632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4508420176559276,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 2633
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.330898676904065,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 2634
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6842944463884602,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 2635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.919068326428618,
+      "learning_rate": 2e-05,
+      "loss": 0.1274,
+      "step": 2636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.646313606834632,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 2637
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7607100278061723,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 2638
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0656026530229918,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 2639
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.301535399470604,
+      "learning_rate": 2e-05,
+      "loss": 0.104,
+      "step": 2640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8882235573859458,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 2641
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.868675238606446,
+      "learning_rate": 2e-05,
+      "loss": 0.1089,
+      "step": 2642
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2410026131642655,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 2643
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5952662566802944,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 2644
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6392760845571798,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 2645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.758992336929821,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 2646
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7877407639527902,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 2647
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7613174600161847,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 2648
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.380318605142885,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 2649
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3529548749510716,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 2650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.543941948753226,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 2651
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.146154487295865,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 2652
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.948049188378171,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 2653
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.284003270423132,
+      "learning_rate": 2e-05,
+      "loss": 0.098,
+      "step": 2654
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6310293107589784,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 2655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1981589980718965,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 2656
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.609467080002744,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 2657
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8668900056845269,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 2658
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.789370221030293,
+      "learning_rate": 2e-05,
+      "loss": 0.1011,
+      "step": 2659
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.870639887542838,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 2660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.904180873369883,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 2661
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.319690888503754,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 2662
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.875044283318516,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 2663
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3024829810672571,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 2664
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.59807850553615,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 2665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9584896064280457,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 2666
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.984936005235255,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 2667
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.372271534150257,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 2668
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1534453735248196,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 2669
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.38960412473736,
+      "learning_rate": 2e-05,
+      "loss": 0.0953,
+      "step": 2670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.606477968047822,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 2671
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8939068266640566,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 2672
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 17.138758123892668,
+      "learning_rate": 2e-05,
+      "loss": 0.2024,
+      "step": 2673
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1179612213417465,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 2674
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.213036631466792,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 2675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.37558997921038,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 2676
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9141624781253315,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 2677
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6787071028261176,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 2678
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.977975723039397,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 2679
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.187000757319836,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 2680
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.505404935159629,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 2681
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2156652389966363,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 2682
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5182802472852415,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 2683
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.732666931915857,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 2684
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.462447695652651,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 2685
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.86338325263584,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 2686
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.833560167844064,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 2687
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.959294731558102,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 2688
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.311598524736289,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 2689
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.475734929371826,
+      "learning_rate": 2e-05,
+      "loss": 0.1481,
+      "step": 2690
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8196730800269667,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 2691
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7982975511759192,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 2692
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7171620055537162,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 2693
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6120846322326625,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 2694
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.986030216361456,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 2695
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.521180639096152,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 2696
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.992265412883467,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 2697
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.714939957615875,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 2698
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7700090247533005,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 2699
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5712902763746088,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 2700
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.379359009601078,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 2701
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.278595337450109,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 2702
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.022389851365701,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 2703
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7253926125566066,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 2704
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3511279956743125,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 2705
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.915818736027188,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 2706
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.618528839308722,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 2707
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5261188834395154,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 2708
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.015730397550961,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 2709
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.730298300294997,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 2710
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.864435210044486,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 2711
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.474298507562578,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 2712
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.713841446750677,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 2713
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8470288233591423,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 2714
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6672659041731204,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 2715
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.040174509886896,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 2716
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.21709539258457,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 2717
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.097971981559741,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 2718
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.012294709370047,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 2719
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.21646571426884295,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 2720
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.395909564014886,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 2721
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3900797342854847,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 2722
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.960318691325982,
+      "learning_rate": 2e-05,
+      "loss": 0.1119,
+      "step": 2723
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.957577139434659,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 2724
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2772945196248149,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 2725
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3717908790425644,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 2726
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.602610980751362,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 2727
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.90773403354728,
+      "learning_rate": 2e-05,
+      "loss": 0.0979,
+      "step": 2728
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3810769520754504,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 2729
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.136297487788006,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 2730
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.007324654786155,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 2731
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2858225014346334,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 2732
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.118402158301064,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 2733
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.956131707663296,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 2734
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.2914267661192635,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 2735
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.965297243069758,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 2736
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.157048174423146,
+      "learning_rate": 2e-05,
+      "loss": 0.0729,
+      "step": 2737
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1563974750393589,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 2738
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.692845442069802,
+      "learning_rate": 2e-05,
+      "loss": 0.0813,
+      "step": 2739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6479523250921067,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 2740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4926898091537284,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 2741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7923558345946495,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 2742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9069617863357428,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 2743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.588814509318825,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 2744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8340145970774584,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 2745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.30506189175553,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 2746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.680482417942925,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 2747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.082473703784265,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 2748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6328249844160507,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 2749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1572088918949515,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 2750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.535198098655639,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 2751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.193061437803101,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 2752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.786250315553155,
+      "learning_rate": 2e-05,
+      "loss": 0.0834,
+      "step": 2753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1025854460806261,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 2754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.342472025638972,
+      "learning_rate": 2e-05,
+      "loss": 0.0757,
+      "step": 2755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.413575157182323,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 2756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9028535195131633,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 2757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5532526650718275,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 2758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6108188860173196,
+      "learning_rate": 2e-05,
+      "loss": 0.1215,
+      "step": 2759
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.510712654951993,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 2760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.443706416735882,
+      "learning_rate": 2e-05,
+      "loss": 0.0963,
+      "step": 2761
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7179358201212478,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 2762
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.352020219125579,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 2763
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0347255956037924,
+      "learning_rate": 2e-05,
+      "loss": 0.1206,
+      "step": 2764
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.712359414672417,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 2765
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.762675981396667,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 2766
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.096423212584589,
+      "learning_rate": 2e-05,
+      "loss": 0.0776,
+      "step": 2767
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6655013597979034,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 2768
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.994942375482122,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 2769
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9517085317763057,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 2770
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4252738990566527,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 2771
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.5384102909223625,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 2772
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.269034195011416,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 2773
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7541525439786647,
+      "learning_rate": 2e-05,
+      "loss": 0.128,
+      "step": 2774
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.815475930908814,
+      "learning_rate": 2e-05,
+      "loss": 0.0838,
+      "step": 2775
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.377970694750369,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 2776
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0671715431361863,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 2777
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4830436738160446,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 2778
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1506833379785757,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 2779
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.216760705815178,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 2780
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1429865820709844,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 2781
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.288436126937777,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 2782
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.53437431112876,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 2783
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.665835023260379,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 2784
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0341172156078606,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 2785
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7787035832417807,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 2786
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5498268515040534,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 2787
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6033506112709297,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 2788
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7697997400214693,
+      "learning_rate": 2e-05,
+      "loss": 0.0872,
+      "step": 2789
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.803951416990088,
+      "learning_rate": 2e-05,
+      "loss": 0.1088,
+      "step": 2790
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.195939043182103,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 2791
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4299111113231482,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 2792
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.13493243799655,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 2793
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.448110123630424,
+      "learning_rate": 2e-05,
+      "loss": 0.1011,
+      "step": 2794
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.921737790055664,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 2795
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.681219954960847,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 2796
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2155900400951984,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 2797
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7803714521876195,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 2798
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0462171850675597,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 2799
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.357452314612304,
+      "learning_rate": 2e-05,
+      "loss": 0.0872,
+      "step": 2800
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1665068698620265,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 2801
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0496871099853893,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 2802
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.364677453064167,
+      "learning_rate": 2e-05,
+      "loss": 0.0889,
+      "step": 2803
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.743179621890749,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 2804
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.144972336404148,
+      "learning_rate": 2e-05,
+      "loss": 0.0947,
+      "step": 2805
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.062331221639488,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 2806
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.8906191717063345,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 2807
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.473609198311831,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 2808
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.578789184408614,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 2809
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6303376924537223,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 2810
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.758986450308907,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 2811
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.504799757975446,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 2812
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.889732791332251,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 2813
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.319316305814471,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 2814
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.014959732936869,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 2815
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.822242728495289,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 2816
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4281523794716944,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 2817
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.821141065368154,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 2818
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.009720874975574,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 2819
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.120509927355796,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 2820
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.177986130763922,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 2821
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9851236253246998,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 2822
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1884577191250236,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 2823
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7740134917125707,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 2824
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8862183975595816,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 2825
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7242631836359696,
+      "learning_rate": 2e-05,
+      "loss": 0.1023,
+      "step": 2826
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8771006332050444,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 2827
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2500080108385885,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 2828
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.110385784968308,
+      "learning_rate": 2e-05,
+      "loss": 0.1211,
+      "step": 2829
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.007395529640905,
+      "learning_rate": 2e-05,
+      "loss": 0.0985,
+      "step": 2830
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5419592662882122,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 2831
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.392318670430495,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 2832
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.246195717353165,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 2833
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.689065900557956,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 2834
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5043757291620925,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 2835
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7273140216755403,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 2836
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.295145953987188,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 2837
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.96774795978468,
+      "learning_rate": 2e-05,
+      "loss": 0.0906,
+      "step": 2838
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.06727221395386,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 2839
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.09623479720479,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 2840
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3323638590246596,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 2841
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4214467516608646,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 2842
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3580045446169273,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 2843
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1755087115061413,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 2844
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4412839003949784,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 2845
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.865325914622079,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 2846
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8319305197021292,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 2847
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5499460096302404,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 2848
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.96206274081598,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 2849
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1959097379050823,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 2850
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6957654919125467,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 2851
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9009193307961106,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 2852
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.690182002723832,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 2853
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2772536088171758,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 2854
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0315158373484916,
+      "learning_rate": 2e-05,
+      "loss": 0.0867,
+      "step": 2855
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.589889575885757,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 2856
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.253439320173252,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 2857
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.191105898488339,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 2858
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6332819036298365,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 2859
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2343679500330222,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 2860
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0701905341163012,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 2861
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8835555387186607,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 2862
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5913818517872307,
+      "learning_rate": 2e-05,
+      "loss": 0.1167,
+      "step": 2863
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1081590326606117,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 2864
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2836772841819934,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 2865
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.077386890468172,
+      "learning_rate": 2e-05,
+      "loss": 0.0978,
+      "step": 2866
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.763348081823926,
+      "learning_rate": 2e-05,
+      "loss": 0.0896,
+      "step": 2867
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0901668165453984,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 2868
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.127231716858993,
+      "learning_rate": 2e-05,
+      "loss": 0.1472,
+      "step": 2869
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5171397019082504,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 2870
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.643809354373929,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 2871
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8633348738370517,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 2872
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0048470175734834,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 2873
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4494164838360721,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 2874
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.1586326710529535,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 2875
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6005523473300018,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 2876
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.505851332215405,
+      "learning_rate": 2e-05,
+      "loss": 0.1197,
+      "step": 2877
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.007150521782971,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 2878
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.144607372428774,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 2879
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.54751009272949,
+      "learning_rate": 2e-05,
+      "loss": 0.1552,
+      "step": 2880
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.121171032681068,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 2881
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6074817530645733,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 2882
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5601477637324797,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 2883
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.09958263260175,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 2884
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5795890135961883,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 2885
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6380598814595249,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 2886
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2553960943725575,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 2887
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.136563363907659,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 2888
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.947103890035433,
+      "learning_rate": 2e-05,
+      "loss": 0.0675,
+      "step": 2889
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9261167537998117,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 2890
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7894166132048106,
+      "learning_rate": 2e-05,
+      "loss": 0.1329,
+      "step": 2891
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4694163739144854,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 2892
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5606972021618855,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 2893
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.263502206316444,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 2894
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1259408678843044,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 2895
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.920587755909043,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 2896
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.042846205424324,
+      "learning_rate": 2e-05,
+      "loss": 0.1337,
+      "step": 2897
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6635678187222607,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 2898
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.288523163336388,
+      "learning_rate": 2e-05,
+      "loss": 0.1112,
+      "step": 2899
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8730039069965887,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 2900
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.4376072400526585,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 2901
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.039325838860648,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 2902
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2190742792983422,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 2903
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6017771222821455,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 2904
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6655330350482824,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 2905
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5313634263448135,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 2906
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8765960170165362,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 2907
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.1519937230066,
+      "learning_rate": 2e-05,
+      "loss": 0.0707,
+      "step": 2908
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.097686272903002,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 2909
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0552274099108672,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 2910
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.190177391405571,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 2911
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.762671984673493,
+      "learning_rate": 2e-05,
+      "loss": 0.0998,
+      "step": 2912
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8258985423477858,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 2913
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.754996231992173,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 2914
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6518729329727644,
+      "learning_rate": 2e-05,
+      "loss": 0.0792,
+      "step": 2915
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.938744363493232,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 2916
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4255160986594404,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 2917
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.393970716375419,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 2918
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.57089624894457,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 2919
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0017585368679587,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 2920
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1910796168533304,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 2921
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.084854595159861,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 2922
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.357343696113323,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 2923
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3713066237195415,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 2924
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4207833075374596,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 2925
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3883945806316276,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 2926
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2576670592198154,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 2927
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.243418101974084,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 2928
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.410623514394028,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 2929
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0864364655888674,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 2930
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7407023852694181,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 2931
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4076849080480953,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 2932
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.209405340752291,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 2933
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.085448475638623,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 2934
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2963222954530416,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 2935
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2942102931067687,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 2936
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.150774854343146,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 2937
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5673634180312934,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 2938
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8668870835125605,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 2939
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6306005946446147,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 2940
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.228037056181974,
+      "learning_rate": 2e-05,
+      "loss": 0.0886,
+      "step": 2941
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6908204647882425,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 2942
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.480023431936692,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 2943
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5710010490661634,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 2944
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.475777869674126,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 2945
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.09197143208655,
+      "learning_rate": 2e-05,
+      "loss": 0.0857,
+      "step": 2946
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.319356623666786,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 2947
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4199573034932764,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 2948
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5675017690977486,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 2949
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.439580510229714,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 2950
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.358435639587534,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 2951
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.039437095171394,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 2952
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.65623708869415,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 2953
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.301731094843469,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 2954
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.003804272373209,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 2955
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.628054658322052,
+      "learning_rate": 2e-05,
+      "loss": 0.1599,
+      "step": 2956
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7703480881214024,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 2957
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.266364055033463,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 2958
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7659209011892922,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 2959
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.218186404411071,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 2960
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.291278774832985,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 2961
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.678678232940348,
+      "learning_rate": 2e-05,
+      "loss": 0.1283,
+      "step": 2962
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4373951424761655,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 2963
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3176639539943316,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 2964
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1298287449077913,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 2965
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.03645851709509,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 2966
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8362440137460458,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 2967
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6746997736076414,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 2968
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6407446681384585,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 2969
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0141030063151826,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 2970
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2360450745167624,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 2971
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.359456345122188,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 2972
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.951310180412398,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 2973
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0830767155592884,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 2974
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.488030197917033,
+      "learning_rate": 2e-05,
+      "loss": 0.0951,
+      "step": 2975
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.466049380241778,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 2976
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.774972662916689,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 2977
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8412144679602553,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 2978
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5071229551190664,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 2979
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.205077202844596,
+      "learning_rate": 2e-05,
+      "loss": 0.0966,
+      "step": 2980
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.084551149727961,
+      "learning_rate": 2e-05,
+      "loss": 0.0707,
+      "step": 2981
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7234579629798608,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 2982
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.128053622808133,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 2983
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2020459846563734,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 2984
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.168557356927509,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 2985
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.3872700427184075,
+      "learning_rate": 2e-05,
+      "loss": 0.0936,
+      "step": 2986
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.926271764059118,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 2987
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3583587267711714,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 2988
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9366501821034139,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 2989
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8780688616882726,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 2990
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1315224347545723,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 2991
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.15688325732808,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 2992
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3073217786180726,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 2993
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.056658709411231,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 2994
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.086660874967586,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 2995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0354077555351715,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 2996
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0439626927190466,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 2997
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.127963748149846,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 2998
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.875995795245539,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 2999
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.564495519213556,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 3000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47703105530420925,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 3001
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4500706604090654,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 3002
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0613830316202173,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 3003
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9138171603168284,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 3004
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.771615732963305,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 3005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.3725580390362,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 3006
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2376335688476316,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 3007
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38199266890286687,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 3008
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4582972335320057,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 3009
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5273160373617216,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 3010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.916848171708948,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 3011
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.157580922395815,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 3012
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5323214635526874,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 3013
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7097777793117204,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 3014
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8104220006876716,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 3015
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5659842743260621,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 3016
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2157783493978647,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 3017
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.328066388225656,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 3018
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.966991377249895,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 3019
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.38657339673136,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 3020
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.267940566762443,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 3021
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.494018469909521,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 3022
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2943217300386913,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 3023
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2118974350626166,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 3024
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.116229787218492,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 3025
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4489935933952784,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 3026
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.497521795808157,
+      "learning_rate": 2e-05,
+      "loss": 0.1049,
+      "step": 3027
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.766098240595141,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 3028
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.027987750894847,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 3029
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0423591243949404,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 3030
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6262829687998686,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 3031
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7148771851438346,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 3032
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9742719413962837,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 3033
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.74691571939569,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 3034
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.0384844938867035,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 3035
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.575837155062837,
+      "learning_rate": 2e-05,
+      "loss": 0.0976,
+      "step": 3036
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.420958514992442,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 3037
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.734728370721158,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 3038
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4952037579771402,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 3039
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2728424085984762,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 3040
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.681487877694938,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 3041
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.8478722535729935,
+      "learning_rate": 2e-05,
+      "loss": 0.106,
+      "step": 3042
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.011529408994016,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 3043
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7275995694836963,
+      "learning_rate": 2e-05,
+      "loss": 0.0985,
+      "step": 3044
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.385337724822925,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 3045
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4753574983317814,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 3046
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.324070722209691,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 3047
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2629709417523087,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 3048
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.636129853218036,
+      "learning_rate": 2e-05,
+      "loss": 0.0964,
+      "step": 3049
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4638132768696268,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 3050
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6129537993623884,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 3051
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.205520409827967,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 3052
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4355816687694605,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 3053
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.278539271948874,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 3054
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3486634402212136,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 3055
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.887998727252499,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 3056
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.82123878040711,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 3057
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.131068765009286,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 3058
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.215975817840781,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 3059
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.772557753487743,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 3060
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5136675323069326,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 3061
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.371517221102746,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 3062
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.752308334828175,
+      "learning_rate": 2e-05,
+      "loss": 0.1063,
+      "step": 3063
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1059710975409287,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 3064
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.023971628722055,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 3065
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5407478137895796,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 3066
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.340047891589901,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 3067
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.035620934173946,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 3068
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9980087500372012,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 3069
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8567989038714814,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 3070
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.296575740592707,
+      "learning_rate": 2e-05,
+      "loss": 0.0918,
+      "step": 3071
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.901082356485477,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 3072
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7855094760393613,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 3073
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5138381965074994,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 3074
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.177330992753921,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 3075
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.673727988018294,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 3076
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4857255858432468,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 3077
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.288598429826412,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 3078
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.106518542096077,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 3079
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1290183501521125,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 3080
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.515800499063136,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 3081
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3242554786112217,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 3082
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.544252341860094,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 3083
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6785962966491166,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 3084
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2390651688420729,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 3085
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0751833936704562,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 3086
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.678610092973945,
+      "learning_rate": 2e-05,
+      "loss": 0.1017,
+      "step": 3087
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2400407782316973,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 3088
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.965771628172119,
+      "learning_rate": 2e-05,
+      "loss": 0.082,
+      "step": 3089
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5352994942491205,
+      "learning_rate": 2e-05,
+      "loss": 0.1004,
+      "step": 3090
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3179850095275705,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 3091
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.642075089692103,
+      "learning_rate": 2e-05,
+      "loss": 0.1445,
+      "step": 3092
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.31134025981663,
+      "learning_rate": 2e-05,
+      "loss": 0.0994,
+      "step": 3093
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.3500252821615115,
+      "learning_rate": 2e-05,
+      "loss": 0.0819,
+      "step": 3094
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.510442122346068,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 3095
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4149351055200823,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 3096
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5357821357599044,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 3097
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.412625926658219,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 3098
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.587262600682926,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 3099
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6733084844931415,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 3100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.253404373012209,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 3101
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.796956512056832,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 3102
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2789928818675644,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 3103
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9346513703051885,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 3104
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9855833153932076,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 3105
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.0252775201030175,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 3106
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.70020332708256,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 3107
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6245906147298435,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 3108
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3479318530461928,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 3109
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.742393026103567,
+      "learning_rate": 2e-05,
+      "loss": 0.1057,
+      "step": 3110
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7617029159747828,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 3111
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.499296579731152,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 3112
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.205568356945205,
+      "learning_rate": 2e-05,
+      "loss": 0.083,
+      "step": 3113
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.99352934282961,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 3114
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7219636712155055,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 3115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3863973697738654,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 3116
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7078308593706995,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 3117
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6431069532181066,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 3118
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1408027614921221,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 3119
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.046482315189953,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 3120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.253190582016415,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 3121
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.665826436000388,
+      "learning_rate": 2e-05,
+      "loss": 0.1388,
+      "step": 3122
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.467503940813219,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 3123
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.119648891402275,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 3124
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9429419316875762,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 3125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2375608224936414,
+      "learning_rate": 2e-05,
+      "loss": 0.1275,
+      "step": 3126
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.416863159434788,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 3127
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.320679695918187,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 3128
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5059466209297963,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 3129
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0845199226259137,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 3130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.37851743721227,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 3131
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7338807372599938,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 3132
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3253245055456904,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 3133
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.880870903035105,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 3134
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.100931638611576,
+      "learning_rate": 2e-05,
+      "loss": 0.0953,
+      "step": 3135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.602980402013546,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 3136
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.530184297747183,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 3137
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.6407161293170995,
+      "learning_rate": 2e-05,
+      "loss": 0.0909,
+      "step": 3138
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1946014881692038,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 3139
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.704949617604153,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 3140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7124721847160065,
+      "learning_rate": 2e-05,
+      "loss": 0.0946,
+      "step": 3141
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.423443745944978,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 3142
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6941887431560048,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 3143
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9234735885255321,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 3144
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.682307438973779,
+      "learning_rate": 2e-05,
+      "loss": 0.0947,
+      "step": 3145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7986099573411187,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 3146
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2757756640087106,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 3147
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6318103012709341,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 3148
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5978834368430717,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 3149
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8466361099521245,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 3150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2175521841005197,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 3151
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.289656598006731,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 3152
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.5339217103341625,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 3153
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8184511548736009,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 3154
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3833031419681654,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 3155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.944242003593359,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 3156
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.4594997888439485,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 3157
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.435639380416725,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 3158
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0449202427315596,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 3159
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.275189865090012,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 3160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.277760070122013,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 3161
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4039256713677357,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 3162
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.696843163223918,
+      "learning_rate": 2e-05,
+      "loss": 0.1001,
+      "step": 3163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.514253979799228,
+      "learning_rate": 2e-05,
+      "loss": 0.091,
+      "step": 3164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.641225085921327,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 3165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.467640071858111,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 3166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9635147950145574,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 3167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3701725133246994,
+      "learning_rate": 2e-05,
+      "loss": 0.0985,
+      "step": 3168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.822694359991371,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 3169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.977864014100392,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 3170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.390597857529249,
+      "learning_rate": 2e-05,
+      "loss": 0.0873,
+      "step": 3171
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.134733103541801,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 3172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5492026800956364,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 3173
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6193835716824547,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 3174
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7130332200169003,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 3175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7051017457226845,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 3176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.769706804073619,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 3177
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0455195615394164,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 3178
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.076800833966571,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 3179
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5616419216128175,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 3180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.368317085906567,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 3181
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.844996925565612,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 3182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9233485209803627,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 3183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8069265108475325,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 3184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6878669481088941,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 3185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4846977284600422,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 3186
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3340131050020918,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 3187
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.534885477215272,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 3188
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.471996540788606,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 3189
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.040937980691357,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 3190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2514300570161496,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 3191
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0352976451526725,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 3192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.463969200903085,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 3193
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1891608336619877,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 3194
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.902266202278268,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 3195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.077118594049834,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 3196
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.198920058569823,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 3197
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7351785808639448,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 3198
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.349109661137087,
+      "learning_rate": 2e-05,
+      "loss": 0.1189,
+      "step": 3199
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5365734652775385,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 3200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.034903300974291,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 3201
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.111859137483143,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 3202
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9524780016083422,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 3203
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.543121380188157,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 3204
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8204619135342062,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 3205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6535022182628516,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 3206
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2685024536267524,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 3207
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.200467550825827,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 3208
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1386217399348397,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 3209
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0746975505968663,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 3210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.728425461971323,
+      "learning_rate": 2e-05,
+      "loss": 0.1019,
+      "step": 3211
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4308809395929507,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 3212
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0711488581879953,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 3213
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2157717775515857,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 3214
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6734098158808381,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 3215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.548569866686421,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 3216
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7117630912292333,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 3217
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.021048004797352,
+      "learning_rate": 2e-05,
+      "loss": 0.1075,
+      "step": 3218
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.226918938198225,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 3219
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0885632165826022,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 3220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.336439322755796,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 3221
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9498422118684584,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 3222
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3996696255558785,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 3223
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.032649064867073,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 3224
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5095988547930799,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 3225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7420574190988591,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 3226
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.085333347339256,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 3227
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.835375853690204,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 3228
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.223953767610176,
+      "learning_rate": 2e-05,
+      "loss": 0.0864,
+      "step": 3229
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3843009395034844,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 3230
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.7051128946646745,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 3231
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2351598538099455,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 3232
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.125702435701891,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 3233
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.644714381109248,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 3234
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.080102433807432,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 3235
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5733808459904235,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 3236
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.969762734903483,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 3237
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5929279173612496,
+      "learning_rate": 2e-05,
+      "loss": 0.0703,
+      "step": 3238
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5854947829917072,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 3239
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.934415305297048,
+      "learning_rate": 2e-05,
+      "loss": 0.0729,
+      "step": 3240
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.974972111468846,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 3241
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.949948333029157,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 3242
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7787952519969426,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 3243
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.586024533561813,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 3244
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5699184571883359,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 3245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.533761540742535,
+      "learning_rate": 2e-05,
+      "loss": 0.0912,
+      "step": 3246
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0620236858261847,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 3247
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2637969962454587,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 3248
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.331234244655697,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 3249
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.157956112743334,
+      "learning_rate": 2e-05,
+      "loss": 0.1134,
+      "step": 3250
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0982460599308026,
+      "learning_rate": 2e-05,
+      "loss": 0.0903,
+      "step": 3251
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.601977341027672,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 3252
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.277368052513723,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 3253
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.512608965128778,
+      "learning_rate": 2e-05,
+      "loss": 0.105,
+      "step": 3254
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.14526934757563,
+      "learning_rate": 2e-05,
+      "loss": 0.0962,
+      "step": 3255
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.551742582863326,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 3256
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.826975861616456,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 3257
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3450765288987054,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 3258
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.872155482363057,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 3259
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.975872770881531,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 3260
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9948419576203444,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 3261
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.704029085069385,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 3262
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.53242371273842,
+      "learning_rate": 2e-05,
+      "loss": 0.1105,
+      "step": 3263
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4022907045155448,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 3264
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.836201468969193,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 3265
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0701115348732264,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 3266
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.881950970789941,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 3267
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.93305015438558,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 3268
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.028617326554931,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 3269
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.625268318932567,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 3270
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.37032853451932,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 3271
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7615723308074425,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 3272
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.04470395563877,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 3273
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4572737290613174,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 3274
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.642255209299119,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 3275
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.505287682009574,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 3276
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2478902178186746,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 3277
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.174680475902547,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 3278
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1105194099134947,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 3279
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.641390622096113,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 3280
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0352109498740514,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 3281
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.935637883448129,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 3282
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.029051544662786,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 3283
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.492707284225062,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 3284
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1885298019100365,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 3285
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7814855582962754,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 3286
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1712005446209153,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 3287
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0861862775858402,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 3288
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.835459266882596,
+      "learning_rate": 2e-05,
+      "loss": 0.0939,
+      "step": 3289
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1554009797799747,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 3290
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.192768806134957,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 3291
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2280506461518064,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 3292
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7015421207469035,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 3293
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.478550639404958,
+      "learning_rate": 2e-05,
+      "loss": 0.1035,
+      "step": 3294
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6055179783311386,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 3295
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.337231862294442,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 3296
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.515469887258408,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 3297
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.699186278981746,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 3298
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.858178980223201,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 3299
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4905560454009468,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 3300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8548033474845698,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 3301
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.022775655718972,
+      "learning_rate": 2e-05,
+      "loss": 0.1246,
+      "step": 3302
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6792838387827411,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 3303
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9464193358895914,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 3304
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.449559914462895,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 3305
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.78036356311928,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 3306
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2785389292895235,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 3307
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8726325022472355,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 3308
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.750368204401905,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 3309
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0621361029716754,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 3310
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7069916316046774,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 3311
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.048436161984775,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 3312
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.743145464145388,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 3313
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1181665678704191,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 3314
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5617657051308615,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 3315
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1514945180842164,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 3316
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8376561338562334,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 3317
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7231073648872393,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 3318
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.9359226846065045,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 3319
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.388543812523028,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 3320
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.983796062379408,
+      "learning_rate": 2e-05,
+      "loss": 0.1219,
+      "step": 3321
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3729264829007861,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 3322
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7749547336719114,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 3323
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.22470221975392,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 3324
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.329490253494346,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 3325
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.449553009051413,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 3326
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.069060911145027,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 3327
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4811882837725214,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 3328
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7930168860130213,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 3329
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4884964594595476,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 3330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.69206954249054,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 3331
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5533929280363976,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 3332
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9106800401630566,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 3333
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0352806595101507,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 3334
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.156094156993977,
+      "learning_rate": 2e-05,
+      "loss": 0.1285,
+      "step": 3335
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2590508245840044,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 3336
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.229467511402865,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 3337
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.305315972228698,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 3338
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.816620330504656,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 3339
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.022574368075478,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 3340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8054635824952725,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 3341
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5053315532198206,
+      "learning_rate": 2e-05,
+      "loss": 0.1094,
+      "step": 3342
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.868529152887046,
+      "learning_rate": 2e-05,
+      "loss": 0.1003,
+      "step": 3343
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.457423227176423,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 3344
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2556018967814766,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 3345
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8529648040048503,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 3346
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6563158683417968,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 3347
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.82873816737116,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 3348
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8232790394563477,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 3349
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.026887884526449,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 3350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6233500968126533,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 3351
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3690235835488003,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 3352
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8725463319844415,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 3353
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.342022014144213,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 3354
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0019528868140104,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 3355
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6173002089537407,
+      "learning_rate": 2e-05,
+      "loss": 0.1007,
+      "step": 3356
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4699648439005155,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 3357
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6013935022765247,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 3358
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.994044380916579,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 3359
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.38509794815265,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 3360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9487938176902175,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 3361
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1545362434658553,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 3362
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.395266823037857,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 3363
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7901341589408422,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 3364
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.290656025813026,
+      "learning_rate": 2e-05,
+      "loss": 0.1398,
+      "step": 3365
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7110065463691657,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 3366
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.995008456630631,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 3367
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.815993787146835,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 3368
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.4831805979937265,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 3369
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.382243119527673,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 3370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.57127613865175,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 3371
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.6893884227568785,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 3372
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7618881786945597,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 3373
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.836275620574096,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 3374
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1593137822607589,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 3375
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.539120154093025,
+      "learning_rate": 2e-05,
+      "loss": 0.1286,
+      "step": 3376
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0166034593511357,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 3377
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.338370800502108,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 3378
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7872118648296462,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 3379
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.620565398516044,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 3380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.519144017872851,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 3381
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4445359788716408,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 3382
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6938270354259732,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 3383
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.258128639908237,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 3384
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8812837411885077,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 3385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.954182513998989,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 3386
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.497869946087466,
+      "learning_rate": 2e-05,
+      "loss": 0.0732,
+      "step": 3387
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.678624532473559,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 3388
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.364384770203303,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 3389
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1942304313514598,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 3390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.845394868057659,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 3391
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7759125365208237,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 3392
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8153537777466149,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 3393
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7960162663407715,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 3394
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.339657987885122,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 3395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.984574364767807,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 3396
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1345789773581094,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 3397
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.011078075128232,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 3398
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9223020516899063,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 3399
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8631415164782317,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 3400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.063070969589152,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 3401
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7691032821498616,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 3402
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.875479001213698,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 3403
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.230493976296708,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 3404
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.885331808327729,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 3405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2943056691609474,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 3406
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4602429575463174,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 3407
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0929167161816202,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 3408
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.714730459360557,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 3409
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8979040469516751,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 3410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.31785626804145,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 3411
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5564059955728733,
+      "learning_rate": 2e-05,
+      "loss": 0.1247,
+      "step": 3412
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.485491255926643,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 3413
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6622715298260509,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 3414
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.332811388039755,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 3415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.262153078879013,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 3416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.403647662844363,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 3417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.626278611271976,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 3418
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1013588505315495,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 3419
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.983101555267296,
+      "learning_rate": 2e-05,
+      "loss": 0.0945,
+      "step": 3420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8551384275040456,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 3421
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.6017174267060375,
+      "learning_rate": 2e-05,
+      "loss": 0.0841,
+      "step": 3422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0246056449377217,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 3423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0889740176924745,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 3424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8110112806627843,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 3425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.507741415167426,
+      "learning_rate": 2e-05,
+      "loss": 0.1252,
+      "step": 3426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.255638073847168,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 3427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6645054661276326,
+      "learning_rate": 2e-05,
+      "loss": 0.1111,
+      "step": 3428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.606384755703558,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 3429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5283849257129577,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 3430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4643504214859666,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 3431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.433627353393351,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 3432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5500911839643905,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 3433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3006269471218053,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 3434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.183967124882748,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 3435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.5379366241642956,
+      "learning_rate": 2e-05,
+      "loss": 0.1359,
+      "step": 3436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.452465194705348,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 3437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2527570034470945,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 3438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.186962065769555,
+      "learning_rate": 2e-05,
+      "loss": 0.1128,
+      "step": 3439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3399963700965754,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 3440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.076134517382288,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 3441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6687560177812404,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 3442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0367231139972906,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 3443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.279598707162431,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 3444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9058451613490208,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 3445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3050343577204897,
+      "learning_rate": 2e-05,
+      "loss": 0.1186,
+      "step": 3446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4313756429504065,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 3447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4732140646468723,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 3448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.554871031657236,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 3449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4352124559685941,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 3450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0741446072896963,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 3451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.994875008892008,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 3452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5919626722991644,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 3453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.662021604729731,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 3454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1104755849979218,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 3455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9115285210972255,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 3456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1315336177482676,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 3457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.161241807011993,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 3458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.343946219813538,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 3459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1053818307950456,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 3460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.437661190082969,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 3461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4392387820613892,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 3462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1258829014006655,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 3463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5570668171956108,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 3464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9915657498330168,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 3465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9793291042120287,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 3466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4721510448084016,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 3467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2668757702174323,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 3468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0518903096237087,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 3469
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.054110506197378,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 3470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4117780152229558,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 3471
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.796386698183862,
+      "learning_rate": 2e-05,
+      "loss": 0.1236,
+      "step": 3472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.050748289704381,
+      "learning_rate": 2e-05,
+      "loss": 0.1485,
+      "step": 3473
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8319268479741493,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 3474
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.911816957458769,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 3475
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.826155506178816,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 3476
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3549393381285524,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 3477
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.588445916250239,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 3478
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.015732895124525,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 3479
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.56268672727989,
+      "learning_rate": 2e-05,
+      "loss": 0.0884,
+      "step": 3480
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.34565594453576565,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 3481
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3939787039833567,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 3482
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 21.56761926888859,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 3483
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6947921563241616,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 3484
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.4030661476810105,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 3485
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8703968765232708,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 3486
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5792982819933288,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 3487
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0881935854106812,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 3488
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9921451938101362,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 3489
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2530441410299586,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 3490
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7164914678476885,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 3491
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.112604199239448,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 3492
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5779252408406417,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 3493
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6204417278487626,
+      "learning_rate": 2e-05,
+      "loss": 0.0775,
+      "step": 3494
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9336584831730548,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 3495
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1742661301346895,
+      "learning_rate": 2e-05,
+      "loss": 0.0962,
+      "step": 3496
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4082064688083349,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 3497
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3570679566860409,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 3498
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5266709530306706,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 3499
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.313644772787979,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 3500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.647799515744783,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 3501
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1536814021819362,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 3502
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8623896694105139,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 3503
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5108256696088316,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 3504
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8443045030328529,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 3505
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9148128722572495,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 3506
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.303909369893587,
+      "learning_rate": 2e-05,
+      "loss": 0.0928,
+      "step": 3507
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.493721831263077,
+      "learning_rate": 2e-05,
+      "loss": 0.0962,
+      "step": 3508
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2752776515996866,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 3509
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3904664452848015,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 3510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.771635259777994,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 3511
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.664848324706362,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 3512
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3108801569885355,
+      "learning_rate": 2e-05,
+      "loss": 0.0931,
+      "step": 3513
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2450828033838937,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 3514
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.767592345161102,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 3515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.615845770905493,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 3516
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7414352521013248,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 3517
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2914328773739618,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 3518
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7370850899177467,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 3519
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.501436892243493,
+      "learning_rate": 2e-05,
+      "loss": 0.104,
+      "step": 3520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.71652552646024,
+      "learning_rate": 2e-05,
+      "loss": 0.1504,
+      "step": 3521
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5928351984419219,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 3522
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.804934765811698,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 3523
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.191393847098925,
+      "learning_rate": 2e-05,
+      "loss": 0.0943,
+      "step": 3524
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8836243682864606,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 3525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.784099105369945,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 3526
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6785063962762115,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 3527
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6674168408431407,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 3528
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.192443763160435,
+      "learning_rate": 2e-05,
+      "loss": 0.0844,
+      "step": 3529
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.33057497140382,
+      "learning_rate": 2e-05,
+      "loss": 0.1174,
+      "step": 3530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0160569588218027,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 3531
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5279560084499018,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 3532
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1566476099338017,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 3533
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1231867964327553,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 3534
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.020145013371781,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 3535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8473661075591947,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 3536
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2459511885803742,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 3537
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1790280455438387,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 3538
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1130636106666048,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 3539
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.0153394005768455,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 3540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8033158675644816,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 3541
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8314711865389333,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 3542
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3373594245834421,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 3543
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.336821191319213,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 3544
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6075457204246526,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 3545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.779961839652611,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 3546
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6689143709698022,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 3547
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.398002642144693,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 3548
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.196429292963408,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 3549
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8390087804923905,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 3550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.786825313378027,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 3551
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.953754799426485,
+      "learning_rate": 2e-05,
+      "loss": 0.1833,
+      "step": 3552
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.822800564049176,
+      "learning_rate": 2e-05,
+      "loss": 0.126,
+      "step": 3553
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9222371086092767,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 3554
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3318741333778314,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 3555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5232465400459563,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 3556
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6407658834367194,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 3557
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6570868627126796,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 3558
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4035214125410131,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 3559
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.737355977731405,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 3560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2550643197310154,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 3561
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3963177288037474,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 3562
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0889876089520816,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 3563
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2987049105413835,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 3564
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.771477492944239,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 3565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.605938870454449,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 3566
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.6700386852810425,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 3567
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.14433910848499,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 3568
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2242476539434684,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 3569
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3382210650609183,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 3570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.532624778838529,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 3571
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.951243839486417,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 3572
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.106876164755419,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 3573
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.704331213632956,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 3574
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0224747300426,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 3575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.16289082008002512,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 3576
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6967278825974477,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 3577
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.395572142351683,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 3578
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.846680927106934,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 3579
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.034646088328702,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 3580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.416858424328887,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 3581
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.884193978461835,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 3582
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7940626443108343,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 3583
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.280425061822321,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 3584
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.246060284409864,
+      "learning_rate": 2e-05,
+      "loss": 0.0811,
+      "step": 3585
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4834540137842933,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 3586
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2686651215191413,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 3587
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5840055965091344,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 3588
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.619092044610827,
+      "learning_rate": 2e-05,
+      "loss": 0.095,
+      "step": 3589
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.581225029444259,
+      "learning_rate": 2e-05,
+      "loss": 0.1346,
+      "step": 3590
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4931072756252246,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 3591
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.990565246176412,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 3592
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4935653476626483,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 3593
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.337539693653384,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 3594
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5161576743700422,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 3595
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.99033680790173,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 3596
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.300868937213597,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 3597
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9970788310823817,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 3598
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8912753745041186,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 3599
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7919469284653653,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 3600
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5260514930979676,
+      "learning_rate": 2e-05,
+      "loss": 0.1121,
+      "step": 3601
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0526695640488177,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 3602
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9982368923307627,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 3603
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9703820847087187,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 3604
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.757014265576717,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 3605
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4570713395645123,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 3606
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.692603524006441,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 3607
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2587807239159208,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 3608
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3548458987933771,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 3609
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.258025619926153,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 3610
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.221010695121259,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 3611
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0368214582952702,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 3612
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2643850156060843,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 3613
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31403761712945066,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 3614
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.775834114565511,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 3615
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.088642552349927,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 3616
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0692933075039095,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 3617
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9550392840602022,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 3618
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.431853576895206,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 3619
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7823202238589437,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 3620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2132462181442685,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 3621
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.057738964530895,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 3622
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8065746363510458,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 3623
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7853479092047513,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 3624
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0137517773857554,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 3625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3986212942671394,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 3626
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.69278813883136,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 3627
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6930852225902184,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 3628
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8283550582577113,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 3629
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7106009374395856,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 3630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.490781990917099,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 3631
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.190616567380513,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 3632
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6351565095379108,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 3633
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.26431709775071,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 3634
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6133784895218046,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 3635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2211961827925943,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 3636
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.205491091324991,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 3637
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.672856290100176,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 3638
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1880266839015112,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 3639
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.1476611187301495,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 3640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.182636740509453,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 3641
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.323065982600893,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 3642
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5196013684828777,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 3643
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.746270565493485,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 3644
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8181598113610664,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 3645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5955691105653393,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 3646
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6002099257083269,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 3647
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.612389108595006,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 3648
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3603390533839888,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 3649
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.70770193711816,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 3650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.047900684827091,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 3651
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.675084061300354,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 3652
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.4066915087978344,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 3653
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.299033665055412,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 3654
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.014621178532759,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 3655
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3857664715237172,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 3656
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6995320349219165,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 3657
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9815757411483708,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 3658
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8512629939564618,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 3659
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.793851282005087,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 3660
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2663947225489633,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 3661
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.108457523763384,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 3662
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.212630384084123,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 3663
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8108140404708208,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 3664
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.584426392271425,
+      "learning_rate": 2e-05,
+      "loss": 0.0821,
+      "step": 3665
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.064300474956376,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 3666
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.817625250211366,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 3667
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5913026007193838,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 3668
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.385707392368336,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 3669
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9183998060949925,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 3670
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5699391087186243,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 3671
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.25758511895732,
+      "learning_rate": 2e-05,
+      "loss": 0.1575,
+      "step": 3672
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6003067980629048,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 3673
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9913011331346067,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 3674
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.927019135376686,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 3675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0078923984763617,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 3676
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.571777734875624,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 3677
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.111966204567455,
+      "learning_rate": 2e-05,
+      "loss": 0.0909,
+      "step": 3678
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8045894850071806,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 3679
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4782831284598608,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 3680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.610228577171375,
+      "learning_rate": 2e-05,
+      "loss": 0.0819,
+      "step": 3681
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.29053975919315,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 3682
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3702813591892997,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 3683
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.555777112362744,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 3684
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.19677305676885,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 3685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1939774581148717,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 3686
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.642915984643853,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 3687
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.866072672472553,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 3688
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7597765991004242,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 3689
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.228060698319829,
+      "learning_rate": 2e-05,
+      "loss": 0.096,
+      "step": 3690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3163354357852013,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 3691
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.2310423305040645,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 3692
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0730311555411998,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 3693
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1765646255718725,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 3694
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.2123809142197075,
+      "learning_rate": 2e-05,
+      "loss": 0.1435,
+      "step": 3695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.33456492845056,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 3696
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2987616271011615,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 3697
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4651875410372444,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 3698
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3643001913604396,
+      "learning_rate": 2e-05,
+      "loss": 0.1365,
+      "step": 3699
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3147909515351512,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 3700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.917977078608126,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 3701
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1154072054743271,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 3702
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.902033124289408,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 3703
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6143792874600154,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 3704
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.389005640679448,
+      "learning_rate": 2e-05,
+      "loss": 0.0741,
+      "step": 3705
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6560573732429984,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 3706
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.547894026099534,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 3707
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8133022309307747,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 3708
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7106698339271226,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 3709
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3559899710294414,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 3710
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.730338956054064,
+      "learning_rate": 2e-05,
+      "loss": 0.0787,
+      "step": 3711
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.60176414993489,
+      "learning_rate": 2e-05,
+      "loss": 0.126,
+      "step": 3712
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6195882769719308,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 3713
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6099787030867523,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 3714
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6801830181840656,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 3715
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4589761225659765,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 3716
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.746937327626151,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 3717
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.713804347335399,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 3718
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.395362768553638,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 3719
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.610205500856839,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 3720
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2729921556846897,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 3721
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9364045645759491,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 3722
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9126107879650625,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 3723
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15028883226240886,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 3724
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5151853837551053,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 3725
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.819667769559548,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 3726
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.2022504499368125,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 3727
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8106882313809842,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 3728
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.892466850123773,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 3729
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3600018407523526,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 3730
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6611764438797,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 3731
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.207544533786882,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 3732
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.818002438366286,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 3733
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8841600463134498,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 3734
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6936073106116485,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 3735
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.314031866825168,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 3736
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9806536895849423,
+      "learning_rate": 2e-05,
+      "loss": 0.1055,
+      "step": 3737
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.767584579723281,
+      "learning_rate": 2e-05,
+      "loss": 0.1618,
+      "step": 3738
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5596305721517643,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 3739
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.005746934646204,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 3740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0007967553958568,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 3741
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.176441216767973,
+      "learning_rate": 2e-05,
+      "loss": 0.0878,
+      "step": 3742
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.196161954679323,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 3743
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7290988263344635,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 3744
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.510853097632967,
+      "learning_rate": 2e-05,
+      "loss": 0.0924,
+      "step": 3745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8717578832196153,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 3746
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5356779440681416,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 3747
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5026538868983055,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 3748
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5412842637601645,
+      "learning_rate": 2e-05,
+      "loss": 0.0947,
+      "step": 3749
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8745458052791022,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 3750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5609936892121024,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 3751
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9512270807620596,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 3752
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.253283134264383,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 3753
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5515219765205788,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 3754
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4214530231752303,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 3755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1184143790831733,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 3756
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8088503958277158,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 3757
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8208774459465245,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 3758
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4457950786174534,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 3759
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.952242896056098,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 3760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2061601354784286,
+      "learning_rate": 2e-05,
+      "loss": 0.1151,
+      "step": 3761
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.393603993238341,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 3762
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6025634169175422,
+      "learning_rate": 2e-05,
+      "loss": 0.0955,
+      "step": 3763
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.131674035500248,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 3764
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5305596956003593,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 3765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.82660375808738,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 3766
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.05574742099953,
+      "learning_rate": 2e-05,
+      "loss": 0.1321,
+      "step": 3767
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9196998344421238,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 3768
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8023508088303741,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 3769
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.424283706429141,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 3770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1725660702405152,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 3771
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2186952725373938,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 3772
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5915651426229385,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 3773
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0309735988170656,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 3774
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8936362588168962,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 3775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.734086937528723,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 3776
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6398636836422935,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 3777
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4076561781313757,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 3778
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6968839192024312,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 3779
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1985089046705553,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 3780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8732326538036315,
+      "learning_rate": 2e-05,
+      "loss": 0.1141,
+      "step": 3781
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7276362825447236,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 3782
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.236635181341253,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 3783
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.709197152339702,
+      "learning_rate": 2e-05,
+      "loss": 0.1634,
+      "step": 3784
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.76928374095121,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 3785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2021551500036276,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 3786
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.635762901832506,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 3787
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0703249178823295,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 3788
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9999606605475502,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 3789
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.6379909039630665,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 3790
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.269371694228788,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 3791
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.675674874755942,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 3792
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.810090580823422,
+      "learning_rate": 2e-05,
+      "loss": 0.0884,
+      "step": 3793
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.170314311158983,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 3794
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8656046395013481,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 3795
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.502228030156061,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 3796
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.631308780356523,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 3797
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.318994530158165,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 3798
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6090239354896614,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 3799
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.483773500574185,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 3800
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2394192184524364,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 3801
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6515315519408869,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 3802
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.469974494375457,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 3803
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.631401329367108,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 3804
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4615399577654526,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 3805
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8725876016431195,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 3806
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3753873106188765,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 3807
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.546206427507061,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 3808
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.170237520816335,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 3809
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4241914561361315,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 3810
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7094815125136558,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 3811
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.891163873704965,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 3812
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2583658647858353,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 3813
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9028916088993946,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 3814
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.800161636319261,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 3815
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1932134944947976,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 3816
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.757520449468297,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 3817
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.778475022014834,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 3818
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.696139450371263,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 3819
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3888507317543963,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 3820
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2309222453480575,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 3821
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.330755922131999,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 3822
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1781082174565856,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 3823
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2505557245355057,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 3824
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.830648402957271,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 3825
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0903512190254951,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 3826
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2387298349072995,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 3827
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5700075844085046,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 3828
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.873993573048201,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 3829
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1011736768922953,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 3830
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.246002158133164,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 3831
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7487376973808058,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 3832
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.593682311375021,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 3833
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.226983210824203,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 3834
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7990678757929632,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 3835
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.479600649336096,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 3836
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.108378822637095,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 3837
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4922447078440864,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 3838
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0420286156183987,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 3839
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.71198128880472,
+      "learning_rate": 2e-05,
+      "loss": 0.1163,
+      "step": 3840
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4525248400577511,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 3841
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4464506050059809,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 3842
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5607511553833895,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 3843
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8551860204912898,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 3844
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.821989397320858,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 3845
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.637674077119226,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 3846
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8337740575671773,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 3847
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.153378509212375,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 3848
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5022137316758477,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 3849
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.962689113616334,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 3850
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.955766707126509,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 3851
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2879356936828095,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 3852
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.34592474675209,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 3853
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.01903630077628,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 3854
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7186909112276257,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 3855
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.642631026112782,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 3856
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.254156212166479,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 3857
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.009994547639389,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 3858
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3964816433539717,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 3859
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.889760096307229,
+      "learning_rate": 2e-05,
+      "loss": 0.1201,
+      "step": 3860
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0077286349408037,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 3861
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.060379385115759,
+      "learning_rate": 2e-05,
+      "loss": 0.238,
+      "step": 3862
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.531896239802713,
+      "learning_rate": 2e-05,
+      "loss": 0.1201,
+      "step": 3863
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.604246212230491,
+      "learning_rate": 2e-05,
+      "loss": 0.0864,
+      "step": 3864
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3240916567398116,
+      "learning_rate": 2e-05,
+      "loss": 0.0858,
+      "step": 3865
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6137099206698435,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 3866
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.331216298804229,
+      "learning_rate": 2e-05,
+      "loss": 0.1165,
+      "step": 3867
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.310484041793978,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 3868
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0479840729865515,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 3869
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.926063519858395,
+      "learning_rate": 2e-05,
+      "loss": 0.1099,
+      "step": 3870
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8255369874460214,
+      "learning_rate": 2e-05,
+      "loss": 0.0919,
+      "step": 3871
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9456560681757895,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 3872
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9167198357958384,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 3873
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.469957852949949,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 3874
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7070073818899516,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 3875
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.962931555136709,
+      "learning_rate": 2e-05,
+      "loss": 0.1505,
+      "step": 3876
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6795153326217647,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 3877
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9989378956377357,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 3878
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.36998925847416,
+      "learning_rate": 2e-05,
+      "loss": 0.1223,
+      "step": 3879
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.370430006897708,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 3880
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.930393632349609,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 3881
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2212953608168755,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 3882
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.197732480785227,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 3883
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2264270406642845,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 3884
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.17450458536254,
+      "learning_rate": 2e-05,
+      "loss": 0.1725,
+      "step": 3885
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.354910865117341,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 3886
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1981662980023,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 3887
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.856099192905545,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 3888
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6169498110097917,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 3889
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9414790732773497,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 3890
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9379123741118218,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 3891
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9726852657785379,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 3892
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3980357257453307,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 3893
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.338311663377898,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 3894
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3989539568904346,
+      "learning_rate": 2e-05,
+      "loss": 0.0755,
+      "step": 3895
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.277927424001557,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 3896
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.975502966563576,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 3897
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8646308663788345,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 3898
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7955456003864165,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 3899
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9479462056570405,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 3900
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.745899958711525,
+      "learning_rate": 2e-05,
+      "loss": 0.1191,
+      "step": 3901
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6139197021420113,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 3902
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0770550031142445,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 3903
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4493663126630447,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 3904
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6899406935248774,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 3905
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7797441156206513,
+      "learning_rate": 2e-05,
+      "loss": 0.0981,
+      "step": 3906
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4788644002657314,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 3907
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.3740281504933165,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 3908
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5181710763494967,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 3909
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.162730447567761,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 3910
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.046510958320196,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 3911
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2085284549449464,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 3912
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2602352949076576,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 3913
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.027269659728176,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 3914
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7260751856432937,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 3915
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.677422775446691,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 3916
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1572946764710235,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 3917
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8362646643195815,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 3918
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4780311344981534,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 3919
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7799988083353233,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 3920
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5822117078550237,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 3921
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1914231336093333,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 3922
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4969215512859764,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 3923
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3901009577412835,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 3924
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5975940209299067,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 3925
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2607145299027445,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 3926
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.288156549215079,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 3927
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5873734558416372,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 3928
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9870501302338835,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 3929
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.460665927010777,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 3930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.316175041631289,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 3931
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.52389760195753,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 3932
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4848151626770307,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 3933
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4971384728449666,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 3934
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2472454237862578,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 3935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.080174403581682,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 3936
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1378346632066882,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 3937
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9668893650283115,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 3938
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6373026335574834,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 3939
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.139646766066497,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 3940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0731666324771982,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 3941
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.628813743380785,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 3942
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.740625362029731,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 3943
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.752467051371808,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 3944
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6845086215488457,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 3945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9904444230845908,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 3946
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.055748727101513,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 3947
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6260792671204016,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 3948
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8200427747246373,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 3949
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5564073414143378,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 3950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.317056653486732,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 3951
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.409486265665864,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 3952
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9070343271806642,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 3953
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1960768268466477,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 3954
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8021480897466604,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 3955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.574964466751514,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 3956
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.500263885359342,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 3957
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1233877910215955,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 3958
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.24318161476737,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 3959
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3122713981776233,
+      "learning_rate": 2e-05,
+      "loss": 0.1002,
+      "step": 3960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.081085734715318,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 3961
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6502802524113969,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 3962
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.183897921918929,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 3963
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.543014513016496,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 3964
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.929042998410913,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 3965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.499927111275422,
+      "learning_rate": 2e-05,
+      "loss": 0.0707,
+      "step": 3966
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1026844522512596,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 3967
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.6732367883507955,
+      "learning_rate": 2e-05,
+      "loss": 0.1185,
+      "step": 3968
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5720778832980207,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 3969
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3758311360644133,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 3970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.827951516915002,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 3971
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.04177854219424,
+      "learning_rate": 2e-05,
+      "loss": 0.0909,
+      "step": 3972
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0362191088202555,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 3973
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.056660911988915,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 3974
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.571317436573754,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 3975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.432998325690325,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 3976
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.034092622262265,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 3977
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7312514248043428,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 3978
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5940293608547178,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 3979
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.428243618968056,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 3980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0805202910403744,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 3981
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.5786313763122894,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 3982
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4515452965755959,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 3983
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1775267055747032,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 3984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4419846970987056,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 3985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.7100256765587805,
+      "learning_rate": 2e-05,
+      "loss": 0.1355,
+      "step": 3986
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.272991172412312,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 3987
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.456670838884891,
+      "learning_rate": 2e-05,
+      "loss": 0.1149,
+      "step": 3988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.284431556200381,
+      "learning_rate": 2e-05,
+      "loss": 0.0813,
+      "step": 3989
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7564415601990175,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 3990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.032736249192934,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 3991
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6120935103278232,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 3992
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.277920186759434,
+      "learning_rate": 2e-05,
+      "loss": 0.0868,
+      "step": 3993
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.194854330404578,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 3994
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2908997104957507,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 3995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.088788935803345,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 3996
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.800360943567282,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 3997
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6779506381772453,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 3998
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.784175662730562,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 3999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6206118827216787,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 4000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.253882934394322,
+      "learning_rate": 2e-05,
+      "loss": 0.1069,
+      "step": 4001
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.414958775105337,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 4002
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.595352788074587,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 4003
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6305135211729038,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 4004
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.044237139373703,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 4005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.267504867398315,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 4006
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8834906658632968,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 4007
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4913900751162974,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 4008
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.601107586353605,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 4009
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8368933083159105,
+      "learning_rate": 2e-05,
+      "loss": 0.0775,
+      "step": 4010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7375835261405252,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 4011
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4830693978080054,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 4012
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3352491710625842,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 4013
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.178629331229737,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 4014
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9582494420148546,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 4015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5030521494129805,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 4016
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5193698669874136,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 4017
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6372671051613383,
+      "learning_rate": 2e-05,
+      "loss": 0.1187,
+      "step": 4018
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.017478529081607,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 4019
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2824715164620613,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 4020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3986800851811585,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 4021
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.675217172238571,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 4022
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6605036168438594,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 4023
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0423646110197127,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 4024
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.105573379096711,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 4025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9184535268324292,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 4026
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6111314353549421,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 4027
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1389804286551857,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 4028
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0562628389740683,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 4029
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8444337127734767,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 4030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.289880703979631,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 4031
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1296071989687237,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 4032
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8496122239255692,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 4033
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.183788529718319,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 4034
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8093051990837736,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 4035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.749009737763034,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 4036
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8304102381538283,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 4037
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5484107514658787,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 4038
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5955569925245676,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 4039
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.161341419658246,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 4040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.611443898272701,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 4041
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.369077218606252,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 4042
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.42412129961212,
+      "learning_rate": 2e-05,
+      "loss": 0.1281,
+      "step": 4043
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.510788907091578,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 4044
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.345847245293077,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 4045
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5502395405176697,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 4046
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1884897228863167,
+      "learning_rate": 2e-05,
+      "loss": 0.0844,
+      "step": 4047
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.42861914988522,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 4048
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6371608674150773,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 4049
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5971805078707573,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 4050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6159644158477557,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 4051
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.380282398851711,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 4052
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8137790738420816,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 4053
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4247347400796555,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 4054
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5491200348783654,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 4055
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9227926679770615,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 4056
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.864380670950143,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 4057
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8164813406203395,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 4058
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.138843541817867,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 4059
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8073754141824363,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 4060
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.078477866154469,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 4061
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7074993067179345,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 4062
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.49726715387317594,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 4063
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9265245469462808,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 4064
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9413440471792816,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 4065
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6550823450379082,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 4066
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.399119376056826,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 4067
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8332903989642997,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 4068
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1660621416535364,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 4069
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.545073953852347,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 4070
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6173987459930754,
+      "learning_rate": 2e-05,
+      "loss": 0.1257,
+      "step": 4071
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6217224372743198,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 4072
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7194161251340088,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 4073
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9166172223772315,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 4074
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.670590179969307,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 4075
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8571481805593097,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 4076
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6777896666713548,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 4077
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.037158032493275,
+      "learning_rate": 2e-05,
+      "loss": 0.0808,
+      "step": 4078
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1374855130245662,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 4079
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.400710827148063,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 4080
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.422795658990595,
+      "learning_rate": 2e-05,
+      "loss": 0.1068,
+      "step": 4081
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9736956284845255,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 4082
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.834860821858956,
+      "learning_rate": 2e-05,
+      "loss": 0.1365,
+      "step": 4083
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.133356648106962,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 4084
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.404420470464131,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 4085
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.117375498072437,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 4086
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8874152431186353,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 4087
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.27660628821912,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 4088
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.72421105636905,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 4089
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3456683105926135,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 4090
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.34702204867109,
+      "learning_rate": 2e-05,
+      "loss": 0.0819,
+      "step": 4091
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.568223617207736,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 4092
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1596250506038477,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 4093
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.175747906058407,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 4094
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6436141621554878,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 4095
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.367216530517619,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 4096
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8558893309311146,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 4097
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.974266788945425,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 4098
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4140136121225453,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 4099
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7794884850077871,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 4100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0883572519621287,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 4101
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0806138873465785,
+      "learning_rate": 2e-05,
+      "loss": 0.1269,
+      "step": 4102
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6207087061002724,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 4103
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.244203827556377,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 4104
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.563951964846368,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 4105
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.175691613040817,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 4106
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1424412576045264,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 4107
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5748744006258346,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 4108
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9435310946507394,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 4109
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.544444322609193,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 4110
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.1526112056838205,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 4111
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.098104247944248,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 4112
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1708671813814568,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 4113
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.484969426244005,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 4114
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0920976419007553,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 4115
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7294065776295322,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 4116
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5315243222269614,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 4117
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5157953540195768,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 4118
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4619916081318554,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 4119
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1317583111794258,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 4120
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.212382030035991,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 4121
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.415064925331931,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 4122
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3938208728783577,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 4123
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.526807032377577,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 4124
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.099435994924448,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 4125
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3924145173201077,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 4126
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6912791708765713,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 4127
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8420514918754286,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 4128
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8329945155991814,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 4129
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5702111510408876,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 4130
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.605112300758696,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 4131
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.843328940169922,
+      "learning_rate": 2e-05,
+      "loss": 0.0955,
+      "step": 4132
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.06853722806371,
+      "learning_rate": 2e-05,
+      "loss": 0.124,
+      "step": 4133
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4576512876220267,
+      "learning_rate": 2e-05,
+      "loss": 0.1124,
+      "step": 4134
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.014199507997311,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 4135
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9304692967291546,
+      "learning_rate": 2e-05,
+      "loss": 0.0955,
+      "step": 4136
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7993442993797184,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 4137
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1371616831956417,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 4138
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.856150572769172,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 4139
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3410537194281327,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 4140
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.218225517166179,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 4141
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2300295956961342,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 4142
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2458017898104554,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 4143
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.762109361759372,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 4144
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.564050716951902,
+      "learning_rate": 2e-05,
+      "loss": 0.0959,
+      "step": 4145
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9346780838426831,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 4146
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.880393235593083,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 4147
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9083954041671345,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 4148
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.908938278974936,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 4149
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.18193654597001807,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 4150
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9614882227395056,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 4151
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7810301913852133,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 4152
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.892700800947914,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 4153
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.072578196015503,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 4154
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1859257074001177,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 4155
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.434316855288341,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 4156
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.423356519830417,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 4157
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9360881091040225,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 4158
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4792301361364624,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 4159
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.696109023779088,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 4160
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3434734764912504,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 4161
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1994186171255965,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 4162
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.919917610533086,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 4163
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.36068245927848,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 4164
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7268132747542464,
+      "learning_rate": 2e-05,
+      "loss": 0.1083,
+      "step": 4165
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6416032647404846,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 4166
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.197441598427749,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 4167
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3090236128592685,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 4168
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.18854160033463,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 4169
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.085480486350269,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 4170
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9639939801613826,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 4171
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3043812375177708,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 4172
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.49999186389979206,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 4173
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4833445781308097,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 4174
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5431918794918102,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 4175
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7777404892396365,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 4176
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9706231546196735,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 4177
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0639835546843155,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 4178
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7883958532325825,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 4179
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0529522548796475,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 4180
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.141099434788857,
+      "learning_rate": 2e-05,
+      "loss": 0.1178,
+      "step": 4181
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3139104923289513,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 4182
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.21939330670998,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 4183
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5677900838757317,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 4184
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4975990312338818,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 4185
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2647327056263344,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 4186
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.41510149521606,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 4187
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1138747081818425,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 4188
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.692409526490264,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 4189
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.213092758174254,
+      "learning_rate": 2e-05,
+      "loss": 0.1986,
+      "step": 4190
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4684777411495986,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 4191
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2147936162365256,
+      "learning_rate": 2e-05,
+      "loss": 0.1344,
+      "step": 4192
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8308480316234292,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 4193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.769996114285404,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 4194
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46727236866768557,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 4195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.579010861459094,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 4196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1171331081934452,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 4197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0633789241866791,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 4198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.785850072875124,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 4199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.597690572579825,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 4200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1405632315854966,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 4201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.928095234002791,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 4202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3970534740121257,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 4203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6279541770229351,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 4204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.897599919612301,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 4205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9707487104079715,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 4206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1417174510154533,
+      "learning_rate": 2e-05,
+      "loss": 0.1223,
+      "step": 4207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.660942261937391,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 4208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.944890950536683,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 4209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3478065738596343,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 4210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3922732863507114,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 4211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6167109198140843,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 4212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.938775903761155,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 4213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2325744534457335,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 4214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.96455408467607,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 4215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3177493796407722,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 4216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.347592762360535,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 4217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1873104546506177,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 4218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2825965783264355,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 4219
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.952531220776272,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 4220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7854552462941613,
+      "learning_rate": 2e-05,
+      "loss": 0.0861,
+      "step": 4221
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1763574656952467,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 4222
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2887582593296136,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 4223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2652794936242677,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 4224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.308662374809383,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 4225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9282057530031755,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 4226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7329543885459773,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 4227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.409780947891957,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 4228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.975481162389378,
+      "learning_rate": 2e-05,
+      "loss": 0.1285,
+      "step": 4229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.143686235294509,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 4230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.657896429237968,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 4231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8716410113326016,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 4232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5453575474852743,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 4233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2250485275634344,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 4234
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.5665607932039345,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 4235
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5137338934997073,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 4236
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4219770840695536,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 4237
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8797556014182355,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 4238
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.35599450643133,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 4239
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5771433720616606,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 4240
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2593678400863175,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 4241
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.610754961610525,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 4242
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.04864887258471,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 4243
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8958027456422046,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 4244
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7236071533368076,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 4245
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.829446743479405,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 4246
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.787139913796645,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 4247
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4806804143077272,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 4248
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.138022078710554,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 4249
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5336918726121544,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 4250
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3467631387339115,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 4251
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8443783080786585,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 4252
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0878507719170636,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 4253
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6508257065341203,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 4254
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.769206060645438,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 4255
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4143812839427268,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 4256
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.206996681145202,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 4257
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6295568689967324,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 4258
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3699116658999553,
+      "learning_rate": 2e-05,
+      "loss": 0.1429,
+      "step": 4259
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1709930246036433,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 4260
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.150992670019807,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 4261
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.610844364023228,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 4262
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.417070892116593,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 4263
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5873607896675791,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 4264
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.0525331745990165,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 4265
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.296247956733562,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 4266
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7021521493298957,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 4267
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6325646901421025,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 4268
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5289094124723777,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 4269
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.309503592383763,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 4270
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.188518571098207,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 4271
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.914445052551175,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 4272
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.556207377642464,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 4273
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0452928396640417,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 4274
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0343718822396446,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 4275
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47394766813023076,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 4276
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3177281305784258,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 4277
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5988696025208697,
+      "learning_rate": 2e-05,
+      "loss": 0.0757,
+      "step": 4278
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.196538479649822,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 4279
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9662355716963673,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 4280
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.211017097173393,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 4281
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9682407271071422,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 4282
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.05727390714716,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 4283
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0659978856078682,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 4284
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.921809063532435,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 4285
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.396638817533879,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 4286
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9373586866050655,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 4287
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.647323329027009,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 4288
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3901432788883015,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 4289
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3474419570588436,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 4290
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5087163563560324,
+      "learning_rate": 2e-05,
+      "loss": 0.1028,
+      "step": 4291
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.623067146477649,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 4292
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.433684737158331,
+      "learning_rate": 2e-05,
+      "loss": 0.1663,
+      "step": 4293
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9350414058779617,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 4294
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0977468346726567,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 4295
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.994953882645217,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 4296
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.909274140218413,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 4297
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6071391320248622,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 4298
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.50678058439813,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 4299
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.010446801890366,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 4300
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5436809122660837,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 4301
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.973318340119976,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 4302
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1257877120008652,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 4303
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0824412316278322,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 4304
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.16439352813157232,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 4305
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.993979150515055,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 4306
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9311905579509384,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 4307
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9625573022336904,
+      "learning_rate": 2e-05,
+      "loss": 0.0942,
+      "step": 4308
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.126661186816265,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 4309
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8243529735026227,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 4310
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6152045637829,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 4311
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8142373546984842,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 4312
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2782568957499563,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 4313
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.16725067201566,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 4314
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2373675034617027,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 4315
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8166293877235757,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 4316
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.444564783268537,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 4317
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0675471230566047,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 4318
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.481426192401003,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 4319
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4188507951734155,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 4320
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5916906777746107,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 4321
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.020957512592764,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 4322
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.218362295914699,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 4323
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8873414427384225,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 4324
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.540013053640553,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 4325
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9122770117119785,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 4326
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1552173700430974,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 4327
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6498658559195947,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 4328
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.719092601734116,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 4329
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4633130022749588,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 4330
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.352229123787862,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 4331
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.735039940774287,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 4332
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.468872550143842,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 4333
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1191041723535284,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 4334
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.442667252466945,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 4335
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8004097516847684,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 4336
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.677094129760615,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 4337
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7461900898512797,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 4338
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.145129707985084,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 4339
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6972336402472825,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 4340
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4501049136061503,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 4341
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7131280798033246,
+      "learning_rate": 2e-05,
+      "loss": 0.107,
+      "step": 4342
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1592278360415116,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 4343
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3649912298898816,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 4344
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6101091062408268,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 4345
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.936544877724731,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 4346
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8904997098408467,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 4347
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6064702458171458,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 4348
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.006492106945158,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 4349
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.417715327427938,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 4350
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0366641020870888,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 4351
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46402883825009994,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 4352
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2635424238427913,
+      "learning_rate": 2e-05,
+      "loss": 0.1305,
+      "step": 4353
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.231774584226598,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 4354
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.071058785602002,
+      "learning_rate": 2e-05,
+      "loss": 0.0962,
+      "step": 4355
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5005294342306497,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 4356
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1608812657208762,
+      "learning_rate": 2e-05,
+      "loss": 0.0939,
+      "step": 4357
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8878014027029866,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 4358
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3144518888446064,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 4359
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8974633248746562,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 4360
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1093230436811472,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 4361
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5916554768486872,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 4362
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6903588813561554,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 4363
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1869988140945456,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 4364
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4900694070664087,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 4365
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6689542409659142,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 4366
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2019132481686654,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 4367
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4070291620064075,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 4368
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.032783047157045,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 4369
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.956394663083993,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 4370
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4208299904918955,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 4371
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.509190537453314,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 4372
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6448003045939004,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 4373
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.584006843788138,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 4374
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3105771648243787,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 4375
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5492194548435503,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 4376
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8780057914429916,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 4377
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5767351583372342,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 4378
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.497188131203188,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 4379
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.341163852410113,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 4380
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.41414546211057,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 4381
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.989564838571218,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 4382
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4982510543200829,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 4383
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6093158829439647,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 4384
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.898662552997943,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 4385
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6203618982890835,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 4386
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1531706865559184,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 4387
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1856866845235745,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 4388
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9157376386082557,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 4389
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.36453561832723,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 4390
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0574560863120444,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 4391
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5764980427747701,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 4392
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2247037957643552,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 4393
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.37973877532808,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 4394
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9808228306003082,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 4395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.692430206858287,
+      "learning_rate": 2e-05,
+      "loss": 0.1217,
+      "step": 4396
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6355799534752817,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 4397
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.551120986629531,
+      "learning_rate": 2e-05,
+      "loss": 0.0955,
+      "step": 4398
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.511531847851408,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 4399
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.82613087527491,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 4400
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.689846021998166,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 4401
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3190550816793447,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 4402
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4562919315964635,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 4403
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6491283397387742,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 4404
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.187163837661236,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 4405
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1420144801983854,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 4406
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8349583675947293,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 4407
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1918425057547823,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 4408
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7319835637350078,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 4409
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3435121813734026,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 4410
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1036287050670244,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 4411
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9865935317902542,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 4412
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2910766587135432,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 4413
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1332015915763693,
+      "learning_rate": 2e-05,
+      "loss": 0.1116,
+      "step": 4414
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0160453513378926,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 4415
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.6109585499546695,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 4416
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.275620922827794,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 4417
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.54821925267036,
+      "learning_rate": 2e-05,
+      "loss": 0.0868,
+      "step": 4418
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2672996744160734,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 4419
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5977013168880596,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 4420
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2638198755060817,
+      "learning_rate": 2e-05,
+      "loss": 0.1164,
+      "step": 4421
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6246583285962388,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 4422
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9217904262218541,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 4423
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9963361557354413,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 4424
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8372347468276145,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 4425
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8941723632512535,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 4426
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8684297204174483,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 4427
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3496894867854368,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 4428
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9714696355653816,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 4429
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.383453818154943,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 4430
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0402066466440947,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 4431
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8836367373978273,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 4432
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4312655607769926,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 4433
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2388601303801687,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 4434
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4915492351392334,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 4435
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9243074750191693,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 4436
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8363973281544355,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 4437
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.900107671120049,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 4438
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4533171365901243,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 4439
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.698183861232125,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 4440
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2927877985249845,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 4441
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.227498471055977,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 4442
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.079694774654803,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 4443
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4551964143579147,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 4444
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9469830796809937,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 4445
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.101439962827294,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 4446
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4515207931390577,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 4447
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3561497717180775,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 4448
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6020605289868413,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 4449
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.183879462088776,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 4450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.201505968256613,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 4451
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.226800631454485,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 4452
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4793958371024614,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 4453
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.067010628881934,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 4454
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.413816470352534,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 4455
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.587936194259834,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 4456
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3553141582654976,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 4457
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0049548557337027,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 4458
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7862599524761114,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 4459
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9853422555925793,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 4460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.201256033668719,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 4461
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.095819276291086,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 4462
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2442958369750468,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 4463
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3716293549513283,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 4464
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1943287664685895,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 4465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.320210074509181,
+      "learning_rate": 2e-05,
+      "loss": 0.1437,
+      "step": 4466
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8454247560548068,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 4467
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.122896387404546,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 4468
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9533759604393277,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 4469
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.917762677577389,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 4470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.880401181783117,
+      "learning_rate": 2e-05,
+      "loss": 0.0949,
+      "step": 4471
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3229982890779102,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 4472
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0982165220777875,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 4473
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5967970162694916,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 4474
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.650879131098604,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 4475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5502343758595933,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 4476
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.313313222423696,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 4477
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6775782732422951,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 4478
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4870266970794936,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 4479
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3994304740967674,
+      "learning_rate": 2e-05,
+      "loss": 0.096,
+      "step": 4480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8260866196258898,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 4481
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.994230225128277,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 4482
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5657406296501786,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 4483
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0891443558392273,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 4484
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9246475202972677,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 4485
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.690779437197478,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 4486
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.020962116655775,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 4487
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.871799993519724,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 4488
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9065932577465239,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 4489
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.22374367552743,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 4490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6735263438902095,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 4491
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.237993230799788,
+      "learning_rate": 2e-05,
+      "loss": 0.1108,
+      "step": 4492
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9573453738007519,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 4493
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2971202204861518,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 4494
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.008205662413415,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 4495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0732286176987613,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 4496
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4146494722784526,
+      "learning_rate": 2e-05,
+      "loss": 0.0967,
+      "step": 4497
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.358727170412237,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 4498
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8859582404075447,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 4499
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6706060424596767,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 4500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.068655583187542,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 4501
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.669012715083404,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 4502
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.224299870553972,
+      "learning_rate": 2e-05,
+      "loss": 0.0956,
+      "step": 4503
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.962867771933125,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 4504
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.494591339546471,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 4505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5901754175528477,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 4506
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6120082460244596,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 4507
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5022964420664215,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 4508
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.028690340391062,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 4509
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.092148353261705,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 4510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.776737010555463,
+      "learning_rate": 2e-05,
+      "loss": 0.0828,
+      "step": 4511
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4309485038805752,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 4512
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.8367737769569095,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 4513
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.140025067320463,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 4514
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.650619013482052,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 4515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0240118323647445,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 4516
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3632403698232505,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 4517
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.803674456479614,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 4518
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.902244674953909,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 4519
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.29981672232322,
+      "learning_rate": 2e-05,
+      "loss": 0.1169,
+      "step": 4520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7222891622682749,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 4521
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6806888068367636,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 4522
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.100280484133019,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 4523
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7452807247122846,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 4524
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.699737462472573,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 4525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7599516104050084,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 4526
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.594929661312916,
+      "learning_rate": 2e-05,
+      "loss": 0.0898,
+      "step": 4527
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.670325233647095,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 4528
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.924959361093359,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 4529
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.763589832179979,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 4530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4784774356482115,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 4531
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4645445478374493,
+      "learning_rate": 2e-05,
+      "loss": 0.1099,
+      "step": 4532
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.277641836431158,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 4533
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.12956118527442,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 4534
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0640032176169882,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 4535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.457512203064227,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 4536
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6449233650050394,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 4537
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7539558658506225,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 4538
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1944628194381837,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 4539
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9899364843727472,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 4540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3242592687841546,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 4541
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2460303215697113,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 4542
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.314562029783088,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 4543
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.564920910067666,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 4544
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.056159132775782,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 4545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.183579095428693,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 4546
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.822926257101532,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 4547
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7073044896492795,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 4548
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4668397859069446,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 4549
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1140772065506779,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 4550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.620746163119879,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 4551
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6726077604147318,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 4552
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1546969422404714,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 4553
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3478971058610965,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 4554
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5759710705515685,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 4555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7017107349907874,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 4556
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.897152964986451,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 4557
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.507320378028055,
+      "learning_rate": 2e-05,
+      "loss": 0.1857,
+      "step": 4558
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.270281961866197,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 4559
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8985666415910862,
+      "learning_rate": 2e-05,
+      "loss": 0.0987,
+      "step": 4560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.858004345304666,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 4561
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2676823224941491,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 4562
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.127089314790162,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 4563
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7690515358297425,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 4564
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5972293945741174,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 4565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8538812338749888,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 4566
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.550044515632573,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 4567
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9658056373256882,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 4568
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9712744981867791,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 4569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8820158353295415,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 4570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8671430638344764,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 4571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.858163839858764,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 4572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0160568129617777,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 4573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4969503873504695,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 4574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.887749473726946,
+      "learning_rate": 2e-05,
+      "loss": 0.1015,
+      "step": 4575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1114163164180986,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 4576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6430329966908106,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 4577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.49997123993175435,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 4578
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8805664721757012,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 4579
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7836795099521165,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 4580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.149868533338633,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 4581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7567754102101227,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 4582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6621488214601592,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 4583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.79844614945492,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 4584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.044731017476886,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 4585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.492316263633328,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 4586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.944125629203958,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 4587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9615862447529993,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 4588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9511961258258319,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 4589
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9884079928369287,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 4590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7414491055171044,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 4591
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9160281507859784,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 4592
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.081731778305712,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 4593
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.907346457103257,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 4594
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.264108189555537,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 4595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6480457837678917,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 4596
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.866576718235085,
+      "learning_rate": 2e-05,
+      "loss": 0.082,
+      "step": 4597
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0028274338918846,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 4598
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8603714909324676,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 4599
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6877161523348796,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 4600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8899587292681266,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 4601
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3168685477501114,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 4602
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.94500682266559,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 4603
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2397010437456584,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 4604
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.566126327693435,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 4605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3945637089332448,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 4606
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6237442100864476,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 4607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.592820676884688,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 4608
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6971216133203677,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 4609
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7682702296821886,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 4610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5464806633605779,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 4611
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.2717895645608515,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 4612
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5209419351062023,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 4613
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4811160803668018,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 4614
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9278374223198157,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 4615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0908799253343595,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 4616
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6859984528003817,
+      "learning_rate": 2e-05,
+      "loss": 0.1124,
+      "step": 4617
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.290482964126548,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 4618
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6616652867047783,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 4619
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9479243868558852,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 4620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.205109372183665,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 4621
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2754853403609907,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 4622
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.495823136586379,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 4623
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2329902080148076,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 4624
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9741697982811717,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 4625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.070826887168234,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 4626
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.840171658606924,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 4627
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8323693197370376,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 4628
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.0160942545616924,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 4629
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5488769943906499,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 4630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.642660884137658,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 4631
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3257720299547964,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 4632
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.758748438868635,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 4633
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.937583759818138,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 4634
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7481195701102907,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 4635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.015746593504162,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 4636
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.904020349507015,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 4637
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5688157421182054,
+      "learning_rate": 2e-05,
+      "loss": 0.0813,
+      "step": 4638
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.0914187098853025,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 4639
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9441177818745303,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 4640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7190668680967343,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 4641
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.372461565278001,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 4642
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8066966704694615,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 4643
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9617453799988315,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 4644
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.670852611022218,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 4645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0694806425042267,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 4646
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8866534817070115,
+      "learning_rate": 2e-05,
+      "loss": 0.0921,
+      "step": 4647
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7362668713310474,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 4648
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.777377593629277,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 4649
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.432927645322222,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 4650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6819323377339567,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 4651
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2216588055894206,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 4652
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.543829749073318,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 4653
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7025494993456847,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 4654
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.536335957190616,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 4655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3576017631057369,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 4656
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.318115337900452,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 4657
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.611250072091562,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 4658
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1421347258473777,
+      "learning_rate": 2e-05,
+      "loss": 0.0844,
+      "step": 4659
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7209243848753366,
+      "learning_rate": 2e-05,
+      "loss": 0.0986,
+      "step": 4660
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.054452490398789,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 4661
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7660815963292618,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 4662
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8322149318858216,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 4663
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2584369152380286,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 4664
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.528913352595055,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 4665
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9716296312483601,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 4666
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2454336684185232,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 4667
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.932378309617437,
+      "learning_rate": 2e-05,
+      "loss": 0.0997,
+      "step": 4668
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.936409366420164,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 4669
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2436699477003295,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 4670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6901090510503414,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 4671
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9880260011404094,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 4672
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7169060959862343,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 4673
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.769641527524869,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 4674
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6385824998249285,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 4675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6237832429109793,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 4676
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.3869224494251675,
+      "learning_rate": 2e-05,
+      "loss": 0.0931,
+      "step": 4677
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9314347360888247,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 4678
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5886795763037127,
+      "learning_rate": 2e-05,
+      "loss": 0.0938,
+      "step": 4679
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.348531163721878,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 4680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.10432244925961,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 4681
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.240182864354397,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 4682
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.012456464650737,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 4683
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8705904772555684,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 4684
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.078910621773224,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 4685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8558780258934096,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 4686
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.163140056508697,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 4687
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.980318038149705,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 4688
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5622794725674833,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 4689
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.672349675717353,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 4690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.100687972409986,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 4691
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1547426457943994,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 4692
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4908884995163496,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 4693
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.67164759257888,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 4694
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5325987189453696,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 4695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9872692792299126,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 4696
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.293925772852073,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 4697
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0511734650982936,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 4698
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8302974892338573,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 4699
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7856082803052242,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 4700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5433550773527942,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 4701
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4960084895846761,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 4702
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.480929340791654,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 4703
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5281928548589165,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 4704
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1170086661960132,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 4705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.13397263494534,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 4706
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.451298513598659,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 4707
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2922819681934719,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 4708
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.480676645292521,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 4709
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.70229112133034,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 4710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7486597105744246,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 4711
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3978626180394607,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 4712
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1182194991716237,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 4713
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3058161671225483,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 4714
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.131867922701533,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 4715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.414261913014942,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 4716
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.196509689359186,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 4717
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7634509720288545,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 4718
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.109525536008217,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 4719
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2119081114346988,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 4720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.992163620773777,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 4721
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.463202508937627,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 4722
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5086472330436103,
+      "learning_rate": 2e-05,
+      "loss": 0.0838,
+      "step": 4723
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7899522859191832,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 4724
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9848411606747485,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 4725
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7584234235683835,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 4726
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4678600130924913,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 4727
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.323862303525899,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 4728
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8467674693852897,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 4729
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4684079090145623,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 4730
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.075427491607139,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 4731
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3893222625204884,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 4732
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.200449492606881,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 4733
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5227431059677994,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 4734
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.060436488917459,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 4735
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2810791646064539,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 4736
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0872032177329314,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 4737
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.101676895726136,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 4738
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0714645922869135,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 4739
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0373907141835144,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 4740
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6949005329575755,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 4741
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9552758447105513,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 4742
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5001722020974824,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 4743
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.208522803357092,
+      "learning_rate": 2e-05,
+      "loss": 0.0838,
+      "step": 4744
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7797181271971985,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 4745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4578652130115055,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 4746
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.313610544252487,
+      "learning_rate": 2e-05,
+      "loss": 0.1512,
+      "step": 4747
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.467125370144472,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 4748
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8753340968852358,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 4749
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.155543030981123,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 4750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4875727691802398,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 4751
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6193633254236595,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 4752
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9422157890665764,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 4753
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0581626368283508,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 4754
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4933258028919087,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 4755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.497957667115494,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 4756
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.902598558538333,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 4757
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.709712421832355,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 4758
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.199296912638816,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 4759
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0767838589016043,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 4760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5196730936029597,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 4761
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5155789388858607,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 4762
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5133511798219583,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 4763
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4654746362787778,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 4764
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8860566921997206,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 4765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3721400341632688,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 4766
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6517906005432703,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 4767
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.3547645170575455,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 4768
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.002991566324354,
+      "learning_rate": 2e-05,
+      "loss": 0.0884,
+      "step": 4769
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0500790695197173,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 4770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9425774186781046,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 4771
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8963097296248298,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 4772
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.91701982137005,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 4773
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.1904540326040065,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 4774
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8080020867774533,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 4775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3119456683576067,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 4776
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8415940821341277,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 4777
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4743415956207635,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 4778
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9351404331279647,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 4779
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9823125133436457,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 4780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.675324472018042,
+      "learning_rate": 2e-05,
+      "loss": 0.1042,
+      "step": 4781
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3533084509113156,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 4782
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3949047197597213,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 4783
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5174638270634073,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 4784
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0825017463899054,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 4785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5021212995848177,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 4786
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9018402676625643,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 4787
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5193493681322754,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 4788
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5844695715734611,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 4789
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.240375051980994,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 4790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.168376929237217,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 4791
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2170790441573085,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 4792
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.623666788047478,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 4793
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4325133365988922,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 4794
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.278165276771805,
+      "learning_rate": 2e-05,
+      "loss": 0.057,
+      "step": 4795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.107091000626448,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 4796
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.884893648093618,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 4797
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.875294976241178,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 4798
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7507361498585412,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 4799
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9754952425463759,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 4800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7436653335416454,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 4801
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7590969251582225,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 4802
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8274740143582585,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 4803
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.496993548769865,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 4804
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.147269276644926,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 4805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.860478589489055,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 4806
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6715466940995216,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 4807
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.960970674595652,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 4808
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.714180454921542,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 4809
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.589177813938161,
+      "learning_rate": 2e-05,
+      "loss": 0.095,
+      "step": 4810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5080341942264623,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 4811
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.259754746089752,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 4812
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6926458421403231,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 4813
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6265179806643886,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 4814
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1948864246009399,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 4815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6246933530898247,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 4816
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.361734386764993,
+      "learning_rate": 2e-05,
+      "loss": 0.093,
+      "step": 4817
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.177670959486067,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 4818
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3514684302608175,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 4819
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.397567462265397,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 4820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.466574612897786,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 4821
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.941924549817484,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 4822
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7557355351211381,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 4823
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8133876426796429,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 4824
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.403845962365256,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 4825
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9104210791637299,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 4826
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3352377324829066,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 4827
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.651411788874397,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 4828
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3150847140604527,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 4829
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5147841962702613,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 4830
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.858034835594875,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 4831
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4402260150809551,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 4832
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0559811309067335,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 4833
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5186101087656003,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 4834
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6304542965278102,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 4835
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4434967446365135,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 4836
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2088264956341167,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 4837
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8453135122248163,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 4838
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0579161134904345,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 4839
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1710706224879038,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 4840
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.553031454625439,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 4841
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1280010778780447,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 4842
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0224521348751034,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 4843
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8339571172664,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 4844
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0224717803642482,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 4845
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.714065549999186,
+      "learning_rate": 2e-05,
+      "loss": 0.0938,
+      "step": 4846
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.515964745265113,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 4847
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8287133835135405,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 4848
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32809987426108383,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 4849
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.573663446157662,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 4850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8541319287083518,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 4851
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.692543000114316,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 4852
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3246667766711413,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 4853
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8851928859143434,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 4854
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.0457052653845595,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 4855
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.905726231784755,
+      "learning_rate": 2e-05,
+      "loss": 0.0757,
+      "step": 4856
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9714160392308315,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 4857
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.101986181114008,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 4858
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6545482778804388,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 4859
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.076374311295399,
+      "learning_rate": 2e-05,
+      "loss": 0.1118,
+      "step": 4860
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.428986083222086,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 4861
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6437593635322536,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 4862
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.24374367545679,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 4863
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5191040577291364,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 4864
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7136105076219261,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 4865
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7396882609932223,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 4866
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8639284357570829,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 4867
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0992816551878577,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 4868
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5483134289677881,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 4869
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.795659316012859,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 4870
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6858739338851705,
+      "learning_rate": 2e-05,
+      "loss": 0.0869,
+      "step": 4871
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6355129707636105,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 4872
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3646588292483075,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 4873
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2126481309836943,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 4874
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.004452993304698,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 4875
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1478450051224716,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 4876
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.747207730018717,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 4877
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.340324972616673,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 4878
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6812173908091346,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 4879
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6108901241533915,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 4880
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0371104669654974,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 4881
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2043469156614488,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 4882
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.927732086953271,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 4883
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6244477287235506,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 4884
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9948975682882937,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 4885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.97782242368334,
+      "learning_rate": 2e-05,
+      "loss": 0.1392,
+      "step": 4886
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6804322514342522,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 4887
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.553144920917317,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 4888
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.315633558085042,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 4889
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9879704767055526,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 4890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3799309207696546,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 4891
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7974039377320714,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 4892
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4552565497254424,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 4893
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9175299552358196,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 4894
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9782111023369756,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 4895
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3545848672604532,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 4896
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.313355674177718,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 4897
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.847906653204748,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 4898
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.584912791450477,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 4899
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.110973854231912,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 4900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.376322760256727,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 4901
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4093458399016154,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 4902
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3937954982325892,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 4903
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2354023677948365,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 4904
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7879680634277006,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 4905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.185016635143408,
+      "learning_rate": 2e-05,
+      "loss": 0.1047,
+      "step": 4906
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.432097033640011,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 4907
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1571123914092056,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 4908
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4628863563231587,
+      "learning_rate": 2e-05,
+      "loss": 0.0857,
+      "step": 4909
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9901323157795443,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 4910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.177557904596004,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 4911
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7929204797168599,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 4912
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3179430596614767,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 4913
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.261441265797841,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 4914
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.4204545721993265,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 4915
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.234835810593539,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 4916
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9352839026650087,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 4917
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9229139295608806,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 4918
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.003645904264436,
+      "learning_rate": 2e-05,
+      "loss": 0.0956,
+      "step": 4919
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2398858036155604,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 4920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9594430890337151,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 4921
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3301885408893781,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 4922
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.572778973517921,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 4923
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1486937463573064,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 4924
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9230695517487131,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 4925
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.773378033776971,
+      "learning_rate": 2e-05,
+      "loss": 0.1211,
+      "step": 4926
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.951301118567081,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 4927
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.384339949118916,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 4928
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7345907943587378,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 4929
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.562940327098892,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 4930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3224141475595723,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 4931
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.399712825315794,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 4932
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6228904421835844,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 4933
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.440853354008085,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 4934
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1055347665830157,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 4935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7312907117938425,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 4936
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.510602213438261,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 4937
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.948176360177255,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 4938
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3415968096032023,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 4939
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4416616044149664,
+      "learning_rate": 2e-05,
+      "loss": 0.102,
+      "step": 4940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5788477684062333,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 4941
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8143308259543085,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 4942
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7737169087651996,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 4943
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8577802947344553,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 4944
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4842126635808718,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 4945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.18619866582190697,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 4946
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7650437199956532,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 4947
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.787546124953393,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 4948
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7193899168895523,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 4949
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.50238122456153,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 4950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1761005379647713,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 4951
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.878882233352588,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 4952
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.688836986808801,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 4953
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9867145833801667,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 4954
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.947880458749625,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 4955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6259075489772546,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 4956
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0782794429548557,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 4957
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1378901476923113,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 4958
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6801945954460114,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 4959
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.249248332390724,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 4960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.557046145814347,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 4961
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5053745780753545,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 4962
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5287192399200773,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 4963
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.324025583654936,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 4964
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8353819079561124,
+      "learning_rate": 2e-05,
+      "loss": 0.0706,
+      "step": 4965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.552284195554439,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 4966
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9079562928502726,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 4967
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7106552239543507,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 4968
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8326688848426322,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 4969
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8076209883502422,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 4970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4462406237407435,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 4971
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6207816317565287,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 4972
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.637373688913952,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 4973
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.540277997886881,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 4974
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.117203715920372,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 4975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.019395607296208,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 4976
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.733107603151694,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 4977
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2830126497751175,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 4978
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3493461429724465,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 4979
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.622689159560479,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 4980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0846255173329977,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 4981
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.7452402862389915,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 4982
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.201113220271414,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 4983
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.130580017888491,
+      "learning_rate": 2e-05,
+      "loss": 0.1362,
+      "step": 4984
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2527663854987146,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 4985
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3283934865478186,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 4986
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2502764396167816,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 4987
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.471442749694181,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 4988
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8470838115659107,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 4989
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7687036652487595,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 4990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.438127500197331,
+      "learning_rate": 2e-05,
+      "loss": 0.1014,
+      "step": 4991
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.905303229990761,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 4992
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3021074470830296,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 4993
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.675484878945007,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 4994
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8552419674179776,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 4995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7559626361156446,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 4996
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7528841230585808,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 4997
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4391837540177774,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 4998
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.22840162458601,
+      "learning_rate": 2e-05,
+      "loss": 0.1527,
+      "step": 4999
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6138034179132745,
+      "learning_rate": 2e-05,
+      "loss": 0.0808,
+      "step": 5000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3885999787037366,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 5001
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2707870082038073,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 5002
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.314709529270506,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 5003
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.2077765269535465,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 5004
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.928161882885498,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 5005
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.469292073962287,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 5006
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.005284582236273,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 5007
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2841754079895247,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 5008
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6285431744920484,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 5009
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.695753597709592,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 5010
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8095064126108027,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 5011
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.466244742247231,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 5012
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.9813684466170685,
+      "learning_rate": 2e-05,
+      "loss": 0.0941,
+      "step": 5013
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.553465851498086,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 5014
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5604287265581367,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 5015
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.039951291531263,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 5016
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5330117658995956,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 5017
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4304477358846972,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 5018
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1284277701978094,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 5019
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3658510488059867,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 5020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1280312142046434,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 5021
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.410817159243972,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 5022
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.594501467795308,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 5023
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.72072319247916,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 5024
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2509136199005988,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 5025
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0792980766560194,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 5026
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0786913045847553,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 5027
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2611830196988205,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 5028
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8149254454125674,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 5029
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4124921646576012,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 5030
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9944172412279135,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 5031
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.113211413796697,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 5032
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8008754675420744,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 5033
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.007402192579917,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 5034
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3307297235981936,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 5035
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2970434332246563,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 5036
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3280294341931467,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 5037
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.394043468542585,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 5038
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.249897397071658,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 5039
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7804410561048064,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 5040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4245975081711273,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 5041
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.214728648835626,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 5042
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7630669860473507,
+      "learning_rate": 2e-05,
+      "loss": 0.0937,
+      "step": 5043
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4268219661100177,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 5044
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8988264817442911,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 5045
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.786015946370845,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 5046
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5359237344305414,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 5047
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2153172304079502,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 5048
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7753109446597999,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 5049
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6811366144367232,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 5050
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6263845853950398,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 5051
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9972073820179004,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 5052
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.608208149778455,
+      "learning_rate": 2e-05,
+      "loss": 0.0919,
+      "step": 5053
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8043031607961673,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 5054
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2001249841386965,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 5055
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0525057853217605,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 5056
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.060349398604228,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 5057
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.429783859059403,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 5058
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0424113051021298,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 5059
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0412868125059218,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 5060
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8272550420280533,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 5061
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.689872582382717,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 5062
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.368644243448425,
+      "learning_rate": 2e-05,
+      "loss": 0.0898,
+      "step": 5063
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.212526107892671,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 5064
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6369779468045942,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 5065
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.591859953401898,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 5066
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.323664238441405,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 5067
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.199524581879131,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 5068
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9154613132458467,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 5069
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.35462182473302,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 5070
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9350836130485211,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 5071
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.386174312688953,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 5072
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.143644556060321,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 5073
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.042843386377853,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 5074
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.751871599190521,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 5075
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.514063854229019,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 5076
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.495072563354627,
+      "learning_rate": 2e-05,
+      "loss": 0.0707,
+      "step": 5077
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4292338153158246,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 5078
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8718316329492284,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 5079
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5826420864541415,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 5080
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.185892113629577,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 5081
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7698347829229695,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 5082
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3350633403516356,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 5083
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.292602808338565,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 5084
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.17195535415091,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 5085
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8380890369024597,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 5086
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0103731375425447,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 5087
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.039443988667974,
+      "learning_rate": 2e-05,
+      "loss": 0.1393,
+      "step": 5088
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7763333473238117,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 5089
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.064167273588968,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 5090
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.101036517613423,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 5091
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2599513077343234,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 5092
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.801327202567093,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 5093
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.796481615092616,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 5094
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6126291245962051,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 5095
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8567547969116687,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 5096
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.053459119391225,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 5097
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.929348287653904,
+      "learning_rate": 2e-05,
+      "loss": 0.1035,
+      "step": 5098
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5803377770203006,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 5099
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5012910982059561,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 5100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4716296241781435,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 5101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5130566731167847,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 5102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2816145096198466,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 5103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.40114761386836,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 5104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5303892567642641,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 5105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.462932831344344,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 5106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0207847401997079,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 5107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3886728731217253,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 5108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4695301725046175,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 5109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.487624906538617,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 5110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6975367146315194,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 5111
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.036680972001719,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 5112
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.179023997181249,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 5113
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0211546803701013,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 5114
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.520039970761351,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 5115
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.195033115624539,
+      "learning_rate": 2e-05,
+      "loss": 0.1419,
+      "step": 5116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4677174723993573,
+      "learning_rate": 2e-05,
+      "loss": 0.1033,
+      "step": 5117
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9341374413176378,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 5118
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.010632587508036,
+      "learning_rate": 2e-05,
+      "loss": 0.0828,
+      "step": 5119
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3759606018660175,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 5120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4272278854038656,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 5121
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.001319926539547,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 5122
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6535499444829407,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 5123
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3419260909539947,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 5124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7176264471735492,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 5125
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8157012555757932,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 5126
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5754958502638915,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 5127
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6140316788021868,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 5128
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6855765029694774,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 5129
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5681105020764359,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 5130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5598612056024113,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 5131
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3153044732328345,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 5132
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.031209729113939,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 5133
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.399324859493824,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 5134
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1129862076486563,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 5135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7192576142779923,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 5136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.813928107278578,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 5137
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.139079851925832,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 5138
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.070496655652924,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 5139
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0173188033294216,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 5140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7154818329174146,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 5141
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6170882932165216,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 5142
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3903074365263732,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 5143
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.203395853910954,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 5144
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8886910232208494,
+      "learning_rate": 2e-05,
+      "loss": 0.0984,
+      "step": 5145
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5357908797640976,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 5146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3330574557264727,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 5147
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.280882436747809,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 5148
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0419159908732343,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 5149
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9134847657343272,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 5150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.577074823993877,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 5151
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0066824203062745,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 5152
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.033218134550746,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 5153
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2494747820422316,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 5154
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7374956098522836,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 5155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.053977466507937,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 5156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.311390996526452,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 5157
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9221283040166832,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 5158
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7383303563045107,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 5159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1382955977850866,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 5160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7550440710462794,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 5161
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7474188395376107,
+      "learning_rate": 2e-05,
+      "loss": 0.0787,
+      "step": 5162
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7113489071503589,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 5163
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.951088471622478,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 5164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.173115766942858,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 5165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.215084942270635,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 5166
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9251360634611636,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 5167
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.729585137498749,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 5168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.194861929511628,
+      "learning_rate": 2e-05,
+      "loss": 0.1213,
+      "step": 5169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.27845773343412,
+      "learning_rate": 2e-05,
+      "loss": 0.1057,
+      "step": 5170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.463381487744438,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 5171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8426264964385304,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 5172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9560535029558864,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 5173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14080005584711963,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 5174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2159428636160254,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 5175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.244721897631214,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 5176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.213336950073198,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 5177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2439001979903472,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 5178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4713327924784227,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 5179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5475376746309834,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 5180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3752833421247395,
+      "learning_rate": 2e-05,
+      "loss": 0.0729,
+      "step": 5181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7369603445572365,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 5182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.741501540549969,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 5183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3826384353998857,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 5184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9561181519809596,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 5185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0794341390430526,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 5186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8734471566633029,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 5187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.393659944575723,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 5188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.525841197026051,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 5189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7214584511366715,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 5190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.686866500058548,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 5191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7856439243122267,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 5192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.904091310184758,
+      "learning_rate": 2e-05,
+      "loss": 0.1032,
+      "step": 5193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8344263878664115,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 5194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.774258415694392,
+      "learning_rate": 2e-05,
+      "loss": 0.1224,
+      "step": 5195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.277844933537286,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 5196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2116175168557146,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 5197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5665963159142784,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 5198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.000980495444431,
+      "learning_rate": 2e-05,
+      "loss": 0.1394,
+      "step": 5199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.615733328839465,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 5200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8773173317258893,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 5201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4071684236450306,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 5202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.875300184417741,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 5203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5633084952346143,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 5204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6402909175350179,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 5205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5315382051638293,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 5206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.65174052102247,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 5207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8469676164956637,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 5208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8629446945294668,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 5209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1289210335050617,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 5210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.159006557997574,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 5211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4015644851721545,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 5212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8346058013092377,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 5213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2059275880551192,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 5214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.013024448626947,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 5215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9845675277427626,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 5216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.027755078169811,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 5217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7728833916750686,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 5218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4053866809338158,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 5219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8231231970503656,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 5220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6691520155846886,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 5221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2452420684874843,
+      "learning_rate": 2e-05,
+      "loss": 0.11,
+      "step": 5222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.101181814050706,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 5223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5669314516230353,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 5224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7598558544605305,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 5225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3751698735813511,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 5226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.740767580284613,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 5227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6354048342202384,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 5228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.83792439786554,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 5229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6297573467354615,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 5230
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7995512561923126,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 5231
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4833084330913651,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 5232
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.251125692082418,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 5233
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9217349770451796,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 5234
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.94842057700827,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 5235
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9359593572606548,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 5236
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4144175246816757,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 5237
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.044110843272454,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 5238
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1637121607839427,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 5239
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.202266789852318,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 5240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.01411265662243,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 5241
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4391143539165876,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 5242
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.114252147268679,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 5243
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7920200605589434,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 5244
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6144111145210016,
+      "learning_rate": 2e-05,
+      "loss": 0.1083,
+      "step": 5245
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9710776644479817,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 5246
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.12456806761853,
+      "learning_rate": 2e-05,
+      "loss": 0.0919,
+      "step": 5247
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.826045577336629,
+      "learning_rate": 2e-05,
+      "loss": 0.0967,
+      "step": 5248
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1930384685657875,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 5249
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8492760015173129,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 5250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.151385793064741,
+      "learning_rate": 2e-05,
+      "loss": 0.096,
+      "step": 5251
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7264788162871791,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 5252
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0424699288780888,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 5253
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8118288722921356,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 5254
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.023797223281014,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 5255
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7699824888365319,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 5256
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0829780529988553,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 5257
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9847581148146265,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 5258
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6549766251909992,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 5259
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8254406775827192,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 5260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8141722216797316,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 5261
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7474343702574164,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 5262
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7355284884569855,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 5263
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7358172197495745,
+      "learning_rate": 2e-05,
+      "loss": 0.1223,
+      "step": 5264
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.405260134715556,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 5265
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.179890831654959,
+      "learning_rate": 2e-05,
+      "loss": 0.0964,
+      "step": 5266
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7362710945850903,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 5267
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.698950665252177,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 5268
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9634736870506834,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 5269
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2592737959367395,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 5270
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8922878550622302,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 5271
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.4353929272051165,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 5272
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.892288233046829,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 5273
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.676928037782811,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 5274
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0248185978242814,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 5275
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4235919171120935,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 5276
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.039036308500369,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 5277
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4420916160706418,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 5278
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.610271852936652,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 5279
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.731347808322725,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 5280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3884144798141679,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 5281
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1317668130428955,
+      "learning_rate": 2e-05,
+      "loss": 0.141,
+      "step": 5282
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5433537642662503,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 5283
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0511634688458487,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 5284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7153280316654399,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 5285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.332380277039685,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 5286
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1363661229250104,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 5287
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.982278711038685,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 5288
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2507464971908098,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 5289
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9465193611655067,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 5290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6051224755115143,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 5291
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5321378663979266,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 5292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1697113456882686,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 5293
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0445590129945352,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 5294
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2031053318546336,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 5295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8214199607538375,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 5296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.754966239653114,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 5297
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.763015032782235,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 5298
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7371904903673013,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 5299
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8010566182350454,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 5300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4476871262791842,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 5301
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7208346122878573,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 5302
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5053689525820366,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 5303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.543143207151537,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 5304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.401047689740444,
+      "learning_rate": 2e-05,
+      "loss": 0.1187,
+      "step": 5305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.71023636468567,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 5306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8428825581831334,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 5307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.581014135636082,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 5308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.463040312377858,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 5309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5569972990854954,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 5310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.901845661562024,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 5311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8473333032172778,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 5312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.987298462266724,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 5313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.693356123854471,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 5314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8860400043338723,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 5315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9766607311435133,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 5316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.73595651913172,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 5317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2029268432674587,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 5318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.974264671385721,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 5319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.25203054554528054,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 5320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2042603523390698,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 5321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4455895800113034,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 5322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4206670492390154,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 5323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5648876258309674,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 5324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.891899435983809,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 5325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1362965117483153,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 5326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3739595378015352,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 5327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6817067628408624,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 5328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.108588197758418,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 5329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4649055976526881,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 5330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9964290568689682,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 5331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5043494175753762,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 5332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5296563797685354,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 5333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.582675222835491,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 5334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6001184253131773,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 5335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6294187475959303,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 5336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.717751242807101,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 5337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4649589499562627,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 5338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6386384867768923,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 5339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.461076229968669,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 5340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3987137080212446,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 5341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5614495270226583,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 5342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.16164479947044,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 5343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9525920128120866,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 5344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.295908420123771,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 5345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.307620438171673,
+      "learning_rate": 2e-05,
+      "loss": 0.1239,
+      "step": 5346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5482908993054273,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 5347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.795074025399151,
+      "learning_rate": 2e-05,
+      "loss": 0.0671,
+      "step": 5348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.313158302428151,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 5349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8213240426650692,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 5350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.771248568825703,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 5351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.953128356930709,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 5352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.510829453674756,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 5353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.060919172117171,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 5354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8064760127914011,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 5355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.624416558957983,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 5356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.692345622238145,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 5357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0759771045752657,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 5358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3040228281225423,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 5359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.671143541518889,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 5360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.401073684534069,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 5361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.679304562600643,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 5362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3322158412560215,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 5363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2559271952824966,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 5364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.164540262194579,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 5365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9492321696197172,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 5366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.317163172096745,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 5367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5198239238297278,
+      "learning_rate": 2e-05,
+      "loss": 0.0876,
+      "step": 5368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5255863782095527,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 5369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5384108433072238,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 5370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6421220034995698,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 5371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5175681488219794,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 5372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3758251575263247,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 5373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.907072546424666,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 5374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8252828300479815,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 5375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.978828182288963,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 5376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7865228283034487,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 5377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5393426039040157,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 5378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4286167845677826,
+      "learning_rate": 2e-05,
+      "loss": 0.1006,
+      "step": 5379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2893454731571397,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 5380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0895225777228923,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 5381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0748133839717022,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 5382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.235502479595896,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 5383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3927757690050688,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 5384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.253247857670881,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 5385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.81831958042938,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 5386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8230462237140939,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 5387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5024599725656544,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 5388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.980207959285865,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 5389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0001704620197325,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 5390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5422309753412036,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 5391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.240976090802609,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 5392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8184143123665952,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 5393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7081978005951557,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 5394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3066466199779208,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 5395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.503724840926481,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 5396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8820779087590056,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 5397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6762016021899333,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 5398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4787499613052233,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 5399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.3408812785721524,
+      "learning_rate": 2e-05,
+      "loss": 0.09,
+      "step": 5400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.556894087886911,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 5401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6483613489288251,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 5402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5330193069812545,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 5403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7316267753172516,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 5404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9775704027271939,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 5405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5254633930336188,
+      "learning_rate": 2e-05,
+      "loss": 0.134,
+      "step": 5406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.140579223143348,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 5407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8573764350218596,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 5408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0246730257892613,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 5409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.815922170143365,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 5410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5832222682078838,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 5411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.834633794760815,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 5412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3835050983378427,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 5413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9858544090160954,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 5414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.602861601290474,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 5415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.875110458242383,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 5416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.394404512061333,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 5417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.843392821297482,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 5418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.887601405066056,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 5419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2130634344759863,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 5420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8039910760350064,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 5421
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5742237479080867,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 5422
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0161334780189293,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 5423
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7768666322338809,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 5424
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0669710652947284,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 5425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.836320846266387,
+      "learning_rate": 2e-05,
+      "loss": 0.1057,
+      "step": 5426
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.450333580351618,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 5427
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.710914455215987,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 5428
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0516457791935685,
+      "learning_rate": 2e-05,
+      "loss": 0.1158,
+      "step": 5429
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1306039261422312,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 5430
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5108206927502354,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 5431
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.187994853862701,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 5432
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4232549858108876,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 5433
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1562571006166262,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 5434
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5417733613013076,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 5435
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.319862243804576,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 5436
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.514768456746549,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 5437
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3466363046408674,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 5438
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5259616380317143,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 5439
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3407572309820766,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 5440
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4923542162124077,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 5441
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.62939223945367,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 5442
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5141991349648849,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 5443
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2781860468969874,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 5444
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7539754223083523,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 5445
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5585322547196336,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 5446
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.670189390732725,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 5447
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1595418377034292,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 5448
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1273346094935988,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 5449
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.460353624622062,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 5450
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6743967877458799,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 5451
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.223710146510221,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 5452
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2856478170539565,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 5453
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.504957114334195,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 5454
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5165566009714888,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 5455
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2291095286619464,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 5456
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.507739682984068,
+      "learning_rate": 2e-05,
+      "loss": 0.0821,
+      "step": 5457
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9511183580699824,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 5458
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.291290316082599,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 5459
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8501178140896055,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 5460
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3427285148735544,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 5461
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2209656455996951,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 5462
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6478289163264763,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 5463
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6695680477672925,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 5464
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3395525112531659,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 5465
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.53169039439102,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 5466
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2892859063037025,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 5467
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.866004637486179,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 5468
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.642836327636521,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 5469
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.07924295054859,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 5470
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5746059651433713,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 5471
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.766896153186089,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 5472
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.35710037046628,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 5473
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6332760514112553,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 5474
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.68724884707363,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 5475
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.359535821631439,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 5476
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0801467812539767,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 5477
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6333906570224899,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 5478
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0596008732618833,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 5479
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9213851661850989,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 5480
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1213055985358864,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 5481
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.202628706538736,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 5482
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9850593648340302,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 5483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.859686445945384,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 5484
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.677147381293728,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 5485
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.734662024556198,
+      "learning_rate": 2e-05,
+      "loss": 0.0754,
+      "step": 5486
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9951820240106073,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 5487
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3519052530092832,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 5488
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8561277868389254,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 5489
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2190873870077215,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 5490
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.139869546405282,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 5491
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.148649447997141,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 5492
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7877278700949017,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 5493
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.021010885093513,
+      "learning_rate": 2e-05,
+      "loss": 0.1546,
+      "step": 5494
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3340836204618129,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 5495
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6499829031318602,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 5496
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.0050991668482,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 5497
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.578541286085492,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 5498
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2147155890371029,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 5499
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2374011260327644,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 5500
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3738236776752504,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 5501
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2156741065306385,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 5502
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6712307515787423,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 5503
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7882520929820536,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 5504
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9756584866023026,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 5505
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7649510203869965,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 5506
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3210242796261376,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 5507
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0631317558281967,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 5508
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.617841041157075,
+      "learning_rate": 2e-05,
+      "loss": 0.1175,
+      "step": 5509
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.304418101972012,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 5510
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0552912926456275,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 5511
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8942641237056348,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 5512
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6094647020695243,
+      "learning_rate": 2e-05,
+      "loss": 0.0773,
+      "step": 5513
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9666222473020523,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 5514
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2478863011371653,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 5515
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.201144764699877,
+      "learning_rate": 2e-05,
+      "loss": 0.1198,
+      "step": 5516
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2277538695537853,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 5517
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44554674111433995,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 5518
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6823294510352835,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 5519
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.934669429482536,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 5520
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7767685850348789,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 5521
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5558751616178685,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 5522
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9684110786942697,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 5523
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3006031116376007,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 5524
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9679527383035507,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 5525
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.498075522257053,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 5526
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5353495989924797,
+      "learning_rate": 2e-05,
+      "loss": 0.0853,
+      "step": 5527
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3772568269672303,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 5528
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8921234876041098,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 5529
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8864140596252512,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 5530
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7708847999574583,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 5531
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.527746707716027,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 5532
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.569791993778519,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 5533
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0116362619617525,
+      "learning_rate": 2e-05,
+      "loss": 0.1036,
+      "step": 5534
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2945663380655097,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 5535
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.991384316300217,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 5536
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3694620013750451,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 5537
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.082483296230044,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 5538
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.731975488710486,
+      "learning_rate": 2e-05,
+      "loss": 0.0894,
+      "step": 5539
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0344910409612258,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 5540
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9100129555467538,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 5541
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6563704016907085,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 5542
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.168386498145701,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 5543
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8522332375943387,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 5544
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.045208090579982,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 5545
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.704238317735253,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 5546
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.052166535129842,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 5547
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9806438760782419,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 5548
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.960475676510148,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 5549
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.15516967247098,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 5550
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0232805709132298,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 5551
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9226810036762685,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 5552
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.381267922918281,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 5553
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.924204506568332,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 5554
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7502811069870896,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 5555
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.328416332475807,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 5556
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9629672297694247,
+      "learning_rate": 2e-05,
+      "loss": 0.0956,
+      "step": 5557
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.953141759506213,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 5558
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2898246876272292,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 5559
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6421226348046245,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 5560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4583592639615908,
+      "learning_rate": 2e-05,
+      "loss": 0.0906,
+      "step": 5561
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1424031703843633,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 5562
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0333067824501097,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 5563
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0382775712524333,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 5564
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3816066004353154,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 5565
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8478542487932659,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 5566
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0901827813490956,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 5567
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4100782978429143,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 5568
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0730962499787666,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 5569
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.251743407853487,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 5570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0858541772285595,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 5571
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9965257792318036,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 5572
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1439009054470564,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 5573
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.453525124109806,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 5574
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5584037419010874,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 5575
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9100367988802827,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 5576
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.152867367620458,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 5577
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.165868891975803,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 5578
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9897208108056272,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 5579
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2147166341724276,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 5580
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0035457334831333,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 5581
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0466037797461223,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 5582
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9194131244501778,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 5583
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.680571563180971,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 5584
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4598148497066266,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 5585
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8348960120019826,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 5586
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7538246995401807,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 5587
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.811940455407237,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 5588
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2708934342394456,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 5589
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.147144763132904,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 5590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6635364917531376,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 5591
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3636653674206193,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 5592
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1014304115696354,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 5593
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3135211854932987,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 5594
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9730259755637856,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 5595
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.339644148366668,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 5596
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.3356275842930305,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 5597
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.193548722556666,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 5598
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9593880295510382,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 5599
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4200090174321511,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 5600
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.52277849077606,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 5601
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9960670544057892,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 5602
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.36218247318811386,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 5603
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.273473785215697,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 5604
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.459390081261179,
+      "learning_rate": 2e-05,
+      "loss": 0.1017,
+      "step": 5605
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.738444597091602,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 5606
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44323854858593703,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 5607
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0284421835361157,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 5608
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.123167886720008,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 5609
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.641450075039273,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 5610
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.23087115636929628,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 5611
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5869958662809729,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 5612
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.592176988731405,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 5613
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6727948306456435,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 5614
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4246380709415413,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 5615
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.162092705615084,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 5616
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7951834721499664,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 5617
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7723522373240037,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 5618
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7639242437362157,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 5619
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3446691109618607,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 5620
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7445645676867034,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 5621
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.80923416412689,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 5622
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7669734025776591,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 5623
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0927184111480823,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 5624
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.83943246184591,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 5625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.202055046857583,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 5626
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.401201972381834,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 5627
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.535132410383638,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 5628
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1760217779647615,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 5629
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.451878096328061,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 5630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.405314745815195,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 5631
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5343985108438605,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 5632
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2870916772956438,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 5633
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9077360895524782,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 5634
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5659098384594077,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 5635
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.964630320309126,
+      "learning_rate": 2e-05,
+      "loss": 0.1393,
+      "step": 5636
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3743520857430394,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 5637
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5782175320228828,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 5638
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3069242389147189,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 5639
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.378844720357405,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 5640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7224019195879655,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 5641
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.35049840227668,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 5642
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2846341075373906,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 5643
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0776975693180177,
+      "learning_rate": 2e-05,
+      "loss": 0.131,
+      "step": 5644
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.092151430145279,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 5645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.225202296809718,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 5646
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1635116195328012,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 5647
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.781886401886617,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 5648
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7647486751329182,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 5649
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.14510564462954692,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 5650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4462666429305535,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 5651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.907973260198872,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 5652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.775515787622328,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 5653
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.229032960810328,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 5654
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9668946379198498,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 5655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3031289213191157,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 5656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.131928017876328,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 5657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.562688320603303,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 5658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.234771258030418,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 5659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.774653053661273,
+      "learning_rate": 2e-05,
+      "loss": 0.1341,
+      "step": 5660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0404193207864583,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 5661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3887675878748444,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 5662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8673564642823393,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 5663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0995692901333465,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 5664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.499260688082438,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 5665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4118673663293928,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 5666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.74970079316453,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 5667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5047478489095383,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 5668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3212006278839654,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 5669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7906078344457685,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 5670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.171602275949381,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 5671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.007054184089455,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 5672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.894282156539019,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 5673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8876420125526214,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 5674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.269602566286503,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 5675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.050926272114734,
+      "learning_rate": 2e-05,
+      "loss": 0.0883,
+      "step": 5676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9779202704478065,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 5677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.650115975055412,
+      "learning_rate": 2e-05,
+      "loss": 0.0896,
+      "step": 5678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.89795740475097,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 5679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.178472802567404,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 5680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7662434802842808,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 5681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.824072362574236,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 5682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3761145552774785,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 5683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1464822533551198,
+      "learning_rate": 2e-05,
+      "loss": 0.094,
+      "step": 5684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3424441959732527,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 5685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.981244747389159,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 5686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2562818783861838,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 5687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5742088298627028,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 5688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2863569186463732,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 5689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.819764636296257,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 5690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.007902545324228,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 5691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8180470382694639,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 5692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6957218361582125,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 5693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.800378768599566,
+      "learning_rate": 2e-05,
+      "loss": 0.1412,
+      "step": 5694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7685518456499967,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 5695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.732822788988805,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 5696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1993584867975677,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 5697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.707978230207156,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 5698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.042365311438836,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 5699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.278641659687398,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 5700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.866555292346476,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 5701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9313540655942976,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 5702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1188831399300385,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 5703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6212271394198756,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 5704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.000376518781529,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 5705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8565333461427549,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 5706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7731033884317393,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 5707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.005389937727236,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 5708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.388345965955201,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 5709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9837533992497067,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 5710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9396200686009843,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 5711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1589750191182662,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 5712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7243602351650353,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 5713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.33198636817522315,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 5714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1507723241718506,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 5715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1572302709592477,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 5716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7303182222155626,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 5717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7673268176409205,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 5718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2968533054287414,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 5719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.665597342141639,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 5720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.556520496667057,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 5721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.828311892915054,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 5722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0500020608659013,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 5723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2939215891840603,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 5724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7903771814776026,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 5725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2514805690834208,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 5726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2885648948197894,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 5727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1879544467263727,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 5728
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1206417269486915,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 5729
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9159009003525778,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 5730
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7886245947652524,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 5731
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1738665115517986,
+      "learning_rate": 2e-05,
+      "loss": 0.1052,
+      "step": 5732
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9510259716938985,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 5733
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0580179469639313,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 5734
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3137236749133014,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 5735
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3682136955172557,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 5736
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.406904576862159,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 5737
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7022508250097392,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 5738
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.032033680314488,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 5739
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.856499216565666,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 5740
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.69152379990369,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 5741
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.549987502814848,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 5742
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9339721273126735,
+      "learning_rate": 2e-05,
+      "loss": 0.11,
+      "step": 5743
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7216709632241378,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 5744
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5276534212922568,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 5745
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.582124693964384,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 5746
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5489718971772425,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 5747
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.650018304665497,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 5748
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2476495578994857,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 5749
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5639506663032083,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 5750
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5689439463664354,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 5751
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1161868542520297,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 5752
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.089137106860782,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 5753
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0414533957852394,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 5754
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.102430540155365,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 5755
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4366935458106487,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 5756
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6163172937532078,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 5757
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7992753964568653,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 5758
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.82867502426182,
+      "learning_rate": 2e-05,
+      "loss": 0.1171,
+      "step": 5759
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7877394269103266,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 5760
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.046543181377162,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 5761
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8435389763195835,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 5762
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.478643575596964,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 5763
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.25836479388962824,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 5764
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5939550080931761,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 5765
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4962278320059873,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 5766
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5099181845047687,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 5767
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.492060120476635,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 5768
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0206652174822475,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 5769
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3397693654051834,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 5770
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.294143921498834,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 5771
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.842457041412821,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 5772
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6630971960985586,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 5773
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6742470011220454,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 5774
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8526341037370848,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 5775
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5223617276697603,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 5776
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.244547304572533,
+      "learning_rate": 2e-05,
+      "loss": 0.1282,
+      "step": 5777
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7208929218733724,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 5778
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.199578048113175,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 5779
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8786061431537143,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 5780
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.496808079523093,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 5781
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1358364033859916,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 5782
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.570175612266243,
+      "learning_rate": 2e-05,
+      "loss": 0.0963,
+      "step": 5783
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.149608611801055,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 5784
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.294087341472156,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 5785
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.039914081386101,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 5786
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.341021460230942,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 5787
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.503562963225359,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 5788
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8798264373649203,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 5789
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.573680835279508,
+      "learning_rate": 2e-05,
+      "loss": 0.0994,
+      "step": 5790
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6126948808112993,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 5791
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4287023433236474,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 5792
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6861225330463412,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 5793
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4707350657211573,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 5794
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6803471648203345,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 5795
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2057147207143124,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 5796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8506567304854746,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 5797
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5962083495632706,
+      "learning_rate": 2e-05,
+      "loss": 0.1014,
+      "step": 5798
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.194419597801619,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 5799
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0327327305054674,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 5800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5423061831489056,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 5801
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9787351104311075,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 5802
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4185746004686277,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 5803
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4856887318874574,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 5804
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3722362499532044,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 5805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5164871123508665,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 5806
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.096694935348561,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 5807
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.049583681213428,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 5808
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5008019052222956,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 5809
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.923886835582889,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 5810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.48777923717397526,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 5811
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8400203075678974,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 5812
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.724293023889131,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 5813
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.815051431469438,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 5814
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3502287773616253,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 5815
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7185213803916672,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 5816
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7348542374469454,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 5817
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3229940313163207,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 5818
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7931364661936778,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 5819
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.542096490365899,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 5820
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8879101350492395,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 5821
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7046758502750042,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 5822
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.713188446153721,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 5823
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.675561157976173,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 5824
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.3358554591759875,
+      "learning_rate": 2e-05,
+      "loss": 0.0785,
+      "step": 5825
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7431518896196119,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 5826
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2132978477084593,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 5827
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3496330910571137,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 5828
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9644667394518707,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 5829
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2584240359101218,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 5830
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.624973076817419,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 5831
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.993109419107657,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 5832
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.26912927925454,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 5833
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.068486190189263,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 5834
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.08876537415504,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 5835
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6557149922677263,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 5836
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.066524278698875,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 5837
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9074342017301333,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 5838
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2998956638428492,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 5839
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2373757652870907,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 5840
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6981297115621151,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 5841
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.830700747041888,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 5842
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.17612084857191668,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 5843
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44588104561679276,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 5844
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6411872263726242,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 5845
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.034079825321221,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 5846
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2130954294526535,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 5847
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4223308503391188,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 5848
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3585455545362861,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 5849
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4943033765526859,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 5850
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.28996094701747166,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 5851
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.7823586767984905,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 5852
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1940872794412736,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 5853
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33874377551113805,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 5854
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2125933247467586,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 5855
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8517455964157752,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 5856
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42131418463621,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 5857
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5506024389689874,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 5858
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6232497742901518,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 5859
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6356761538250097,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 5860
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.871473388484798,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 5861
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7296546030082741,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 5862
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2925511069303157,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 5863
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.719418135739426,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 5864
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9740075387352505,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 5865
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0166563812631932,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 5866
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6385604107850483,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 5867
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5486676396353056,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 5868
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3660432223160421,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 5869
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0559973676680166,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 5870
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.064988571209258,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 5871
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2277235257556345,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 5872
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.322275406816777,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 5873
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.414778165373474,
+      "learning_rate": 2e-05,
+      "loss": 0.1034,
+      "step": 5874
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3219194181384757,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 5875
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9947240350445108,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 5876
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.831624068648871,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 5877
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.032379701810125,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 5878
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1649935282580666,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 5879
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.405456730575686,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 5880
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.561899642114737,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 5881
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3398386340697017,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 5882
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3625497410154725,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 5883
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.113280309641254,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 5884
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3452663140458163,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 5885
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6412968304497915,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 5886
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39024966327401,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 5887
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3491046011327357,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 5888
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3921461037112337,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 5889
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9880138003274457,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 5890
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.69139021515437,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 5891
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.24230531389629,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 5892
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0049632997596631,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 5893
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.46145876250798,
+      "learning_rate": 2e-05,
+      "loss": 0.0882,
+      "step": 5894
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7172753091218009,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 5895
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40991290912016237,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 5896
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.642707742655769,
+      "learning_rate": 2e-05,
+      "loss": 0.102,
+      "step": 5897
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.37680169260835433,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 5898
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.619276061703665,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 5899
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8931675631708553,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 5900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1430433534660615,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 5901
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.750184429362217,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 5902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9348673284591142,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 5903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.249295608417765,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 5904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5968718527580865,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 5905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.618370029231162,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 5906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6032990508006353,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 5907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1178857981986248,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 5908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.7741803000653125,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 5909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2636378324731887,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 5910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2586839307570368,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 5911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.897073347591193,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 5912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.893476904223685,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 5913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.481091038298661,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 5914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1858427630594883,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 5915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6620221343775214,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 5916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9157075624580617,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 5917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.267227538691133,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 5918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6568994189881274,
+      "learning_rate": 2e-05,
+      "loss": 0.1013,
+      "step": 5919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2906344914491432,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 5920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51788024721611,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 5921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.808320733861836,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 5922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6129680634356436,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 5923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.961096122740346,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 5924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1055929681657637,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 5925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.559228906304998,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 5926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.877522028691505,
+      "learning_rate": 2e-05,
+      "loss": 0.1115,
+      "step": 5927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.100716119099202,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 5928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3637288113667037,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 5929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8042616226452137,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 5930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5816934569487247,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 5931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3999364480852956,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 5932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5865081361837916,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 5933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.922640613472272,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 5934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1839945362218347,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 5935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.396527054460268,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 5936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8473739569671574,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 5937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1467888344974586,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 5938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4993350098242146,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 5939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2825489644641697,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 5940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8679584821888596,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 5941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0612483225203737,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 5942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.482762323140164,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 5943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9715733002953706,
+      "learning_rate": 2e-05,
+      "loss": 0.1378,
+      "step": 5944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.06440621532969534,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 5945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1305256628237603,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 5946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8653626731486277,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 5947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.325411069876482,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 5948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9979936619138801,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 5949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7572170351013325,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 5950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7035205529604434,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 5951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6626186381583157,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 5952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1611543722372886,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 5953
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3634386644484151,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 5954
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5999314519208965,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 5955
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9890953450813615,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 5956
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3203284810018872,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 5957
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.604474876157182,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 5958
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2672660409164536,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 5959
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4238148964684691,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 5960
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.8307638049463355,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 5961
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.653653659877647,
+      "learning_rate": 2e-05,
+      "loss": 0.0707,
+      "step": 5962
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.071898691263434,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 5963
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.485753076921896,
+      "learning_rate": 2e-05,
+      "loss": 0.0837,
+      "step": 5964
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2771142875981543,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 5965
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6948761960045573,
+      "learning_rate": 2e-05,
+      "loss": 0.0858,
+      "step": 5966
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0457892413803784,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 5967
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.451652799904189,
+      "learning_rate": 2e-05,
+      "loss": 0.094,
+      "step": 5968
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9885473992997547,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 5969
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4246331596010302,
+      "learning_rate": 2e-05,
+      "loss": 0.0946,
+      "step": 5970
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2924524757458056,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 5971
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5952078939099068,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 5972
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.137503127209064,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 5973
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1860978128410333,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 5974
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2717356479994875,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 5975
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.74206733983752,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 5976
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.334916243122406,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 5977
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.544722788614181,
+      "learning_rate": 2e-05,
+      "loss": 0.1093,
+      "step": 5978
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.42162416643082,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 5979
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.618067726877692,
+      "learning_rate": 2e-05,
+      "loss": 0.0757,
+      "step": 5980
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4960963289489317,
+      "learning_rate": 2e-05,
+      "loss": 0.0721,
+      "step": 5981
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.661206456836486,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 5982
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.606718729058249,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 5983
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.205806966503216,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 5984
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5352668647888722,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 5985
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7440400128908817,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 5986
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8759799939371922,
+      "learning_rate": 2e-05,
+      "loss": 0.0912,
+      "step": 5987
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8772067753053197,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 5988
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.245686088881831,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 5989
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5866105478978398,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 5990
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.506744718231924,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 5991
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2911327961023298,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 5992
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7332548567295344,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 5993
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5374523589653948,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 5994
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.635641592039375,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 5995
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7063095666864804,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 5996
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.914720067345042,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 5997
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.232086143907041,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 5998
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.372868918051728,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 5999
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.547414318246487,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 6000
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.463052968513755,
+      "learning_rate": 2e-05,
+      "loss": 0.1097,
+      "step": 6001
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41879758920271704,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 6002
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.201009209755183,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 6003
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5851498204936117,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 6004
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9506517543748558,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 6005
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.996673440088534,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 6006
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.661086836409438,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 6007
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.10031363097996,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 6008
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2045636183872512,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 6009
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4457946797828027,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 6010
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4746165333902676,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 6011
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6768893599814165,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 6012
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.61371693489364,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 6013
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2311024208309105,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 6014
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8922425020971165,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 6015
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.079974685665947,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 6016
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.101660277701522,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 6017
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5125878048257793,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 6018
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.06838230702089068,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 6019
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7759494552464978,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 6020
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0342690750697765,
+      "learning_rate": 2e-05,
+      "loss": 0.1042,
+      "step": 6021
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2232711269628802,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 6022
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5955493614516243,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 6023
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6273491092470938,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 6024
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15830619421522726,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 6025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6869224810308119,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 6026
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6787071173032101,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 6027
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5807056804451824,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 6028
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9560966329391944,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 6029
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0596024938956856,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 6030
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.154102861111966,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 6031
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.239074992366636,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 6032
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.256636369333717,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 6033
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7292927542162931,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 6034
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3026197380866322,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 6035
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6578502481748871,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 6036
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.286084428368087,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 6037
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5471007289963197,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 6038
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7501802351644538,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 6039
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1595853691800544,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 6040
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5555590377804448,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 6041
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.334469328433512,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 6042
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6222538117957572,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 6043
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3886714028573608,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 6044
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8779828822337905,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 6045
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.354492735023233,
+      "learning_rate": 2e-05,
+      "loss": 0.1011,
+      "step": 6046
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.339245549380272,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 6047
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0488608242221997,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 6048
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.306643725852594,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 6049
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2097037279895066,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 6050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.681321028446323,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 6051
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.553567361932031,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 6052
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.019954262468525,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 6053
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7090761676536186,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 6054
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3065591476505634,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 6055
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4131650014835813,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 6056
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6381251590500179,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 6057
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.390103067052212,
+      "learning_rate": 2e-05,
+      "loss": 0.1351,
+      "step": 6058
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.83081146279699,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 6059
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.46779063915886104,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 6060
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.339094395442346,
+      "learning_rate": 2e-05,
+      "loss": 0.116,
+      "step": 6061
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1647365078084249,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 6062
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5911616081625415,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 6063
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.382668015109525,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 6064
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3340513091918993,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 6065
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8508270992269673,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 6066
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7264315870547182,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 6067
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.93960180413793,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 6068
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7087935934502723,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 6069
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5712349114835344,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 6070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.724974510446831,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 6071
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.143774622250873,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 6072
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5502147648828346,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 6073
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.725374805808877,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 6074
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8405439660153655,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 6075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5985472775161007,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 6076
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5963705621056838,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 6077
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.456647643453191,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 6078
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2337195672298449,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 6079
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5043930556482055,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 6080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6097610158838105,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 6081
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8365504755713173,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 6082
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.203048921177826,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 6083
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.896760690782228,
+      "learning_rate": 2e-05,
+      "loss": 0.0792,
+      "step": 6084
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.945053205865606,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 6085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0992137136120297,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 6086
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1202555856120655,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 6087
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6704917961359886,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 6088
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7622939564275546,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 6089
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3519119075652268,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 6090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8283384515871562,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 6091
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7096329684408245,
+      "learning_rate": 2e-05,
+      "loss": 0.0976,
+      "step": 6092
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.183652439087031,
+      "learning_rate": 2e-05,
+      "loss": 0.0809,
+      "step": 6093
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.268694276008794,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 6094
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.434101233193075,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 6095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.902814162791027,
+      "learning_rate": 2e-05,
+      "loss": 0.1373,
+      "step": 6096
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9836052200824215,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 6097
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.0055746333467,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 6098
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5628912192244737,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 6099
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.0055198718027,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 6100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.516633973701276,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 6101
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.43719553833016,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 6102
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4285587895251237,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 6103
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9202200048108214,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 6104
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5133605709941507,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 6105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2166560843194747,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 6106
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.077285264779669,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 6107
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4105854363294212,
+      "learning_rate": 2e-05,
+      "loss": 0.0732,
+      "step": 6108
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.511707738967894,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 6109
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.721792337472291,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 6110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.536003209180261,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 6111
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.039892715769534,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 6112
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7771444400627981,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 6113
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1321621416784846,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 6114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.153875834475819,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 6115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.539800744746619,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 6116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3244045151274912,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 6117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.36903953451135,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 6118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.776654679759859,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 6119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0989561156239964,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 6120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4219505373833403,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 6121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.295961034969038,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 6122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.050626738040912,
+      "learning_rate": 2e-05,
+      "loss": 0.1182,
+      "step": 6123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.630736387941653,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 6124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9729097448621093,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 6125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5714402218891211,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 6126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.181693517851433,
+      "learning_rate": 2e-05,
+      "loss": 0.0809,
+      "step": 6127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1487842996351916,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 6128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.364696904831414,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 6129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7084606748070833,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 6130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9154636371064147,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 6131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.747357285831858,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 6132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1077149526899746,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 6133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.150572376535126,
+      "learning_rate": 2e-05,
+      "loss": 0.0928,
+      "step": 6134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.553486044598013,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 6135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9902053618083863,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 6136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3558931437276502,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 6137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5856990646660987,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 6138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2262561252687507,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 6139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2262246460714645,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 6140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.208891022746659,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 6141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5401363038864226,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 6142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.121241555587694,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 6143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9959952327043524,
+      "learning_rate": 2e-05,
+      "loss": 0.0855,
+      "step": 6144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1856887455978393,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 6145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2450294373950215,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 6146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.885450767035706,
+      "learning_rate": 2e-05,
+      "loss": 0.1124,
+      "step": 6147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.477733154873921,
+      "learning_rate": 2e-05,
+      "loss": 0.1142,
+      "step": 6148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7057282258950658,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 6149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0865012153411542,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 6150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.18674924774,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 6151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1497913606706283,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 6152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.760534747833165,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 6153
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2058779333533325,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 6154
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.664151649325969,
+      "learning_rate": 2e-05,
+      "loss": 0.1346,
+      "step": 6155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.508300543620414,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 6156
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.417852896168908,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 6157
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5954510729298397,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 6158
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.755145648767366,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 6159
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2791359917379763,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 6160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6417536383270135,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 6161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0453463538921643,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 6162
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7636035386340503,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 6163
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6844374806098843,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 6164
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5466189702396724,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 6165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1499209563290278,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 6166
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.806252920797227,
+      "learning_rate": 2e-05,
+      "loss": 0.1124,
+      "step": 6167
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6226471721987777,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 6168
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.892393323930863,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 6169
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7020104581741279,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 6170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1104340331959395,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 6171
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.141154164704291,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 6172
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3506157265644438,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 6173
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.827345722747038,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 6174
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2547434451123338,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 6175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5892810622464055,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 6176
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9355037735027499,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 6177
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4381085999231405,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 6178
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5836979855924023,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 6179
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.517070964575614,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 6180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.47723642756674495,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 6181
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.377127172413844,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 6182
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.243166422431737,
+      "learning_rate": 2e-05,
+      "loss": 0.1242,
+      "step": 6183
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.74688223185797,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 6184
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6508526939606036,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 6185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5116800308197578,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 6186
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1751789058271038,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 6187
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5243487056336487,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 6188
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.278221738553991,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 6189
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4974910735424727,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 6190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1504457277865714,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 6191
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5677766584375394,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 6192
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4925487101089168,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 6193
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0649883402940894,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 6194
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.418422477686356,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 6195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4698576554892604,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 6196
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9446647884100756,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 6197
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8875310261098407,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 6198
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4824217823352802,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 6199
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30644494391589977,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 6200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6444416340867325,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 6201
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.939246380761153,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 6202
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.219657390035269,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 6203
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.954229562051306,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 6204
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7316333305055516,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 6205
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.536779967011363,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 6206
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.556571976283139,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 6207
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4516189968840945,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 6208
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7471329742564468,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 6209
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5296250888768013,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 6210
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0200567686960884,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 6211
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.616055007280417,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 6212
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2792881623182557,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 6213
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.778453732266446,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 6214
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.37271874661484394,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 6215
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3773172016648396,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 6216
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1454620286990296,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 6217
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.436724417124301,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 6218
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8276910511351907,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 6219
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5941798808455323,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 6220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.778885854539516,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 6221
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7384123105576912,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 6222
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.65649125742609,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 6223
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0883106918913956,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 6224
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.697874300415738,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 6225
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3652394952262149,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 6226
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3316186618308539,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 6227
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.704459682238129,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 6228
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5847567295781567,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 6229
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18258860563297444,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 6230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1514831040747544,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 6231
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1801378885459815,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 6232
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2702357758159388,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 6233
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.920293838596419,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 6234
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.339253193453474,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 6235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1178481087969647,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 6236
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5103801152860055,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 6237
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.023428966920563,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 6238
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1547650738888273,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 6239
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.164339564245669,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 6240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0007083374477137,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 6241
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.350839091711644,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 6242
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.25808481864328,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 6243
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6697015628154657,
+      "learning_rate": 2e-05,
+      "loss": 0.0808,
+      "step": 6244
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36866287317459856,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 6245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6083895836748616,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 6246
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7560926373666008,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 6247
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.299161000957925,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 6248
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.915440604213282,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 6249
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.18691055931916,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 6250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1674720328294499,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 6251
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.729147057042516,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 6252
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1456207193730656,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 6253
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6147463217057556,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 6254
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4740780713227553,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 6255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6442501495503262,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 6256
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5460671861815765,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 6257
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9320787766973093,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 6258
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9663122330298763,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 6259
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.55957271782572,
+      "learning_rate": 2e-05,
+      "loss": 0.0958,
+      "step": 6260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6009949322739144,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 6261
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.902714178502339,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 6262
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.17724086028640562,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 6263
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.220517162796517,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 6264
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.615089418184414,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 6265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7609821818250178,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 6266
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4925138923464476,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 6267
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.521800809280187,
+      "learning_rate": 2e-05,
+      "loss": 0.0808,
+      "step": 6268
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.009263676914548,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 6269
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1424226660998773,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 6270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.550196741901898,
+      "learning_rate": 2e-05,
+      "loss": 0.0981,
+      "step": 6271
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2741253171306803,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 6272
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.136219483733,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 6273
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.60578978361509,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 6274
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.442753108278878,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 6275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2696875896014514,
+      "learning_rate": 2e-05,
+      "loss": 0.1158,
+      "step": 6276
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1867783009779025,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 6277
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6624865750972795,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 6278
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9268561839355205,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 6279
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8147404076805798,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 6280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2082290878491009,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 6281
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1751494373498885,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 6282
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.39949872027087,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 6283
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8817264324139499,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 6284
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0172786455000085,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 6285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6082680004416123,
+      "learning_rate": 2e-05,
+      "loss": 0.1085,
+      "step": 6286
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8393068603379388,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 6287
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8184726245242672,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 6288
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.20858801930302842,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 6289
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6537819233160957,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 6290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0196258968692122,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 6291
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.019547560988867,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 6292
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.104915171021689,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 6293
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3724813658743242,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 6294
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4450331134367935,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 6295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1727669072151472,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 6296
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.217154852805136,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 6297
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.343683596012671,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 6298
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.047976014496147,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 6299
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.433064949789923,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 6300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8738301426355566,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 6301
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.700775624943294,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 6302
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.924994776794465,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 6303
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.903982446885457,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 6304
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1010482345187873,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 6305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6296217821718297,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 6306
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0316352702546745,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 6307
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9667925285472847,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 6308
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4047650796274884,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 6309
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.939012751554529,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 6310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.365447477654868,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 6311
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8815260168218613,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 6312
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1068565903893026,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 6313
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.49974288326253,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 6314
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0228554808622112,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 6315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.953566093111395,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 6316
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5392793965398165,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 6317
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1103412816804763,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 6318
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8743460318720393,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 6319
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.698506110408697,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 6320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.45037564181626,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 6321
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.178413439361794,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 6322
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.20802204101325275,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 6323
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.150406368797385,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 6324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7710912767287184,
+      "learning_rate": 2e-05,
+      "loss": 0.1215,
+      "step": 6325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.037114099537632,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 6326
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0567891211303237,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 6327
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5223323894122722,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 6328
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.458610930074663,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 6329
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0291860573702,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 6330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6783040581618267,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 6331
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0088138780375893,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 6332
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2174002611852566,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 6333
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.244738951435384,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 6334
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4607112719987785,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 6335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.120360344484854,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 6336
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.171986089925325,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 6337
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8414402087466528,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 6338
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.479781260435255,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 6339
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1868685499099383,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 6340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.46373118557065,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 6341
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.323993678374893,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 6342
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.346494884970948,
+      "learning_rate": 2e-05,
+      "loss": 0.1089,
+      "step": 6343
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.792930592096203,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 6344
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4870090604094335,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 6345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.54900213853899,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 6346
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1342487039726907,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 6347
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.8731703626011855,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 6348
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.458040209175162,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 6349
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4766273282467952,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 6350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6462996527229827,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 6351
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.2284317254956045,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 6352
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.833262629655266,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 6353
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.574951160301063,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 6354
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.575730779675162,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 6355
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7633017049066946,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 6356
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8327520051495139,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 6357
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.969936376696418,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 6358
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.113126817998317,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 6359
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.201677536784773,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 6360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.548197958497997,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 6361
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.103286530336592,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 6362
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.253640387801059,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 6363
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9221116207193256,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 6364
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4467077170904166,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 6365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1270476356160342,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 6366
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8544990013121927,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 6367
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6877654820151791,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 6368
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.16453103872432348,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 6369
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7126150774463172,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 6370
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7814822805327591,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 6371
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7350643952921594,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 6372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.148169902209607,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 6373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5040762668964747,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 6374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.836810056391138,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 6375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.873881705899569,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 6376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5424049326659763,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 6377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6118918552746246,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 6378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10018357102155902,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 6379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8006128327562059,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 6380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4304318184202447,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 6381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7868203946958046,
+      "learning_rate": 2e-05,
+      "loss": 0.1086,
+      "step": 6382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1891344767978522,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 6383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5087957148242865,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 6384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.423682608326898,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 6385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5969157473148723,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 6386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1664684156047573,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 6387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.5425535351475235,
+      "learning_rate": 2e-05,
+      "loss": 0.0945,
+      "step": 6388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.780229036567585,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 6389
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.8222260718811265,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 6390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.30049281719703036,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 6391
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.510503543178882,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 6392
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8780288046280668,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 6393
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8769787837248515,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 6394
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1936943634020607,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 6395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6360812527548394,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 6396
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3608603878135037,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 6397
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0652948533709015,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 6398
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.531851047186972,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 6399
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.101353231268519,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 6400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2738265864863927,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 6401
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.390261454147506,
+      "learning_rate": 2e-05,
+      "loss": 0.0993,
+      "step": 6402
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.125619597999559,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 6403
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8483710075330695,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 6404
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4646339387661444,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 6405
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.757984688793132,
+      "learning_rate": 2e-05,
+      "loss": 0.1058,
+      "step": 6406
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8532704391144876,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 6407
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8928982016795767,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 6408
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0647299300369717,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 6409
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7554595441760318,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 6410
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9376208574426985,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 6411
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.8174000887234545,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 6412
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.510600124214052,
+      "learning_rate": 2e-05,
+      "loss": 0.0998,
+      "step": 6413
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5601768771180007,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 6414
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9771442386530231,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 6415
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.698452476403454,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 6416
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.405349752469709,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 6417
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7067536361832172,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 6418
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.09880469227654,
+      "learning_rate": 2e-05,
+      "loss": 0.0858,
+      "step": 6419
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3779769789015618,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 6420
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.606093310717825,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 6421
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0986612139626564,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 6422
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.3967704798483425,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 6423
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2285435480078535,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 6424
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.037924748332163,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 6425
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.038198470885104,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 6426
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7454600026802163,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 6427
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3919758688752473,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 6428
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.824751872397656,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 6429
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.717670708762986,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 6430
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8534234038578323,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 6431
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.838154571148075,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 6432
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6258528012208833,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 6433
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2728991628505792,
+      "learning_rate": 2e-05,
+      "loss": 0.0858,
+      "step": 6434
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2660968513173028,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 6435
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.126126222322708,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 6436
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2624090060487814,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 6437
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.846552882306524,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 6438
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.853724380491044,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 6439
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3836280012926157,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 6440
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7927143541266655,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 6441
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6346517908348748,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 6442
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.200313507312883,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 6443
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3265466652376787,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 6444
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3393487933608257,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 6445
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.830760423381845,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 6446
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2937789083628872,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 6447
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6863241761147714,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 6448
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0064278069253016,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 6449
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.961596071167694,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 6450
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.609323009955261,
+      "learning_rate": 2e-05,
+      "loss": 0.1116,
+      "step": 6451
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9689531372871436,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 6452
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2230974565230182,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 6453
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2032958070463684,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 6454
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4194109680481817,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 6455
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2012628288258829,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 6456
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.14912687389110926,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 6457
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.733445363733937,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 6458
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6416446968325478,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 6459
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1040994866351266,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 6460
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.964847271252589,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 6461
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.717315646237857,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 6462
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.420648950775903,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 6463
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5687808775148167,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 6464
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.104584572838244,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 6465
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5205511676190901,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 6466
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.68268945043971,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 6467
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3129355752338576,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 6468
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0134041470109274,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 6469
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9136134944039423,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 6470
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.614839406369831,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 6471
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.002583821545656,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 6472
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.0031527968669245,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 6473
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.170078273502434,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 6474
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.964951945427704,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 6475
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1585776933200123,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 6476
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.9944051454446825,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 6477
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2613807366025256,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 6478
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9634980612845679,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 6479
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.728407482492803,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 6480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.435631955358173,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 6481
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1281871456085337,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 6482
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7347171898007177,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 6483
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7712041031892529,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 6484
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.20174653648632043,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 6485
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.691723095684014,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 6486
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.782930016886287,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 6487
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6896741072058872,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 6488
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.881149007542968,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 6489
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.366419069416508,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 6490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5571951501557764,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 6491
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3197419463711704,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 6492
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6500644364613377,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 6493
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4264356809570309,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 6494
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8664555596645604,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 6495
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5658913247947677,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 6496
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09256410070016616,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 6497
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3570785967832246,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 6498
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.45461052705153676,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 6499
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.695465291956352,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 6500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0937783538264965,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 6501
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3454802166607216,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 6502
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5080635037341288,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 6503
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.245655089851704,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 6504
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0290354445555963,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 6505
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0258128374589766,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 6506
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.295314880126205,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 6507
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.89300122959361,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 6508
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38236383498378207,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 6509
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9945078338229174,
+      "learning_rate": 2e-05,
+      "loss": 0.1709,
+      "step": 6510
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4625430337166228,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 6511
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.174918533098242,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 6512
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.627346068888245,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 6513
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2401009131588814,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 6514
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6887727105952262,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 6515
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.626674658212346,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 6516
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9288195306652227,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 6517
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.713066122861332,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 6518
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6325121151531545,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 6519
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4578303987825376,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 6520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8781131331438612,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 6521
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.202315594404207,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 6522
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6357414587218526,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 6523
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2651426656899463,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 6524
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6501155252286845,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 6525
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.086567872768796,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 6526
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7938783224861177,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 6527
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.616371428021033,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 6528
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2366011407106483,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 6529
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.594972804814542,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 6530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38778579770607186,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 6531
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.117100265404961,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 6532
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7776663351229727,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 6533
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6252360464869002,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 6534
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5941178394219353,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 6535
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5482894860750022,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 6536
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2164671621096432,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 6537
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5380194296788283,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 6538
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3986250902338853,
+      "learning_rate": 2e-05,
+      "loss": 0.0925,
+      "step": 6539
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1363346514773913,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 6540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.37251581853426,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 6541
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7969763822203177,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 6542
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1504827957922714,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 6543
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5740906163927548,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 6544
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.120271695558916,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 6545
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.490299952167939,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 6546
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.507356564423603,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 6547
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9804801484315495,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 6548
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9564863653256257,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 6549
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9424954203790358,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 6550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1812720362308304,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 6551
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2332675563558175,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 6552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9951465247896578,
+      "learning_rate": 2e-05,
+      "loss": 0.0998,
+      "step": 6553
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.910980168291578,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 6554
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.122926231206715,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 6555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3891151032072526,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 6556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.988140525913192,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 6557
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2838010763669403,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 6558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.559906266296473,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 6559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.144852264485092,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 6560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8491108427541594,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 6561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8815119117412351,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 6562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0806608369973212,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 6563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.981119174747,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 6564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.638539320117275,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 6565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0823699200818593,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 6566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2346493517115316,
+      "learning_rate": 2e-05,
+      "loss": 0.1168,
+      "step": 6567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.668064169003155,
+      "learning_rate": 2e-05,
+      "loss": 0.0886,
+      "step": 6568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3894586647546947,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 6569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3785581537642972,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 6570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5711509602875657,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 6571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4694057169843222,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 6572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.902038311706078,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 6573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7184196848343882,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 6574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.128868709586811,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 6575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4695846121070236,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 6576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9465887423857609,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 6577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7266140062753632,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 6578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5341131705280193,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 6579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4049160458470824,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 6580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9395886515029894,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 6581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.202165651782773,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 6582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5269176932825574,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 6583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7721902056978156,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 6584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.437616910871898,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 6585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5666873517936704,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 6586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.602755334555954,
+      "learning_rate": 2e-05,
+      "loss": 0.1143,
+      "step": 6587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6032562576035283,
+      "learning_rate": 2e-05,
+      "loss": 0.0945,
+      "step": 6588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8292065161964693,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 6589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8004622766493752,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 6590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1427633923503837,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 6591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6872094797641273,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 6592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0937926692813946,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 6593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.222466258604083,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 6594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.800415663939237,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 6595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.008179389889035,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 6596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2817601901151992,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 6597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3772765042428252,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 6598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3289305992519602,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 6599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0371153439135345,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 6600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6752604202955084,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 6601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5635167310865963,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 6602
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2651126436735016,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 6603
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.816888056374189,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 6604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7123072467000309,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 6605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9753124521065688,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 6606
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.320613854358639,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 6607
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.19486838392291025,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 6608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.222455015513158,
+      "learning_rate": 2e-05,
+      "loss": 0.1477,
+      "step": 6609
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.92127848577261,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 6610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3769302163550197,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 6611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7160547798634125,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 6612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.5673984861262005,
+      "learning_rate": 2e-05,
+      "loss": 0.1196,
+      "step": 6613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.769419550791279,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 6614
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7672694827220174,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 6615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.036019940210159,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 6616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0864076666654188,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 6617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.613954417477252,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 6618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.668062869033414,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 6619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3936147803841066,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 6620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.782518405589746,
+      "learning_rate": 2e-05,
+      "loss": 0.109,
+      "step": 6621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6131408379984205,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 6622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2912900326192496,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 6623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.335244174221042,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 6624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8723697639479828,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 6625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.483475912297945,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 6626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9310224549508028,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 6627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1712140062922558,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 6628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8872564903973372,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 6629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3756836976699943,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 6630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.729134558121944,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 6631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.592279541230325,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 6632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4255195021890192,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 6633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6138529472929464,
+      "learning_rate": 2e-05,
+      "loss": 0.0942,
+      "step": 6634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8118610080049518,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 6635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.673676516620169,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 6636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8243002467067969,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 6637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.00105219508907,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 6638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.770330926272327,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 6639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1235933516714796,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 6640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.332000272152134,
+      "learning_rate": 2e-05,
+      "loss": 0.0787,
+      "step": 6641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5392584622396,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 6642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5856642464188992,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 6643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.250894262832529,
+      "learning_rate": 2e-05,
+      "loss": 0.0721,
+      "step": 6644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.828867597334908,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 6645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8497517728621826,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 6646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8239809389224483,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 6647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.333624370918408,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 6648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0626144347554685,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 6649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4214050821514554,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 6650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8838960425572457,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 6651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5396984066968975,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 6652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.926660673512748,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 6653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.963431971662268,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 6654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1838195383982342,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 6655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1240202027911481,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 6656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3411086981610414,
+      "learning_rate": 2e-05,
+      "loss": 0.0787,
+      "step": 6657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.032605587113891,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 6658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4761095083975797,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 6659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.176834597793056,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 6660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8064087541404948,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 6661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1515134762146428,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 6662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0935559236819623,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 6663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9741556032421119,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 6664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7978921244878254,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 6665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6056551246512547,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 6666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.114591889810876,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 6667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6816168771234403,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 6668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9099634615414027,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 6669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5180704896515522,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 6670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8484792625307364,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 6671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.663048514875034,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 6672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.733774167025121,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 6673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.837350540594158,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 6674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.140960580212663,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 6675
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.960059892848705,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 6676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.970984554926916,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 6677
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6892515205533403,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 6678
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5357527088963694,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 6679
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.376079785581605,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 6680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8345926106975228,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 6681
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8889559479883462,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 6682
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5225651517561287,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 6683
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6749258992899305,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 6684
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6147184089845885,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 6685
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0217273197144285,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 6686
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5847003662763224,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 6687
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5949134219916006,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 6688
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9737432223415634,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 6689
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.708900165455643,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 6690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.285132642720463,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 6691
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6854352289563781,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 6692
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.153401603708596,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 6693
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.905695412430619,
+      "learning_rate": 2e-05,
+      "loss": 0.0925,
+      "step": 6694
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.381816613159932,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 6695
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4734799797212371,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 6696
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3799093356189887,
+      "learning_rate": 2e-05,
+      "loss": 0.1085,
+      "step": 6697
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.668659576366407,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 6698
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.667341705727411,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 6699
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4583643319604662,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 6700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.373663643022218,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 6701
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5398042799657361,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 6702
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9948961725648122,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 6703
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2216610987138694,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 6704
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.97997878294472,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 6705
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2812132333784634,
+      "learning_rate": 2e-05,
+      "loss": 0.0874,
+      "step": 6706
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.873529456941425,
+      "learning_rate": 2e-05,
+      "loss": 0.0929,
+      "step": 6707
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1787465871557221,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 6708
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8548124096144114,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 6709
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8850755822082619,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 6710
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.987026397997507,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 6711
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2587374490909606,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 6712
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1914780548897763,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 6713
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0403649841351896,
+      "learning_rate": 2e-05,
+      "loss": 0.0855,
+      "step": 6714
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.049793754065053,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 6715
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4948866048971374,
+      "learning_rate": 2e-05,
+      "loss": 0.09,
+      "step": 6716
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8887234850977693,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 6717
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.361256492185918,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 6718
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4728280850632278,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 6719
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.001778035174518,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 6720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.4478293536130975,
+      "learning_rate": 2e-05,
+      "loss": 0.1036,
+      "step": 6721
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9305535353810865,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 6722
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5048297138441186,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 6723
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8207231966953064,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 6724
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8608468905665574,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 6725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.231958360920438,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 6726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9333998666334496,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 6727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5126059439889727,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 6728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.576801571703049,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 6729
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6751955516023789,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 6730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.130328130210856,
+      "learning_rate": 2e-05,
+      "loss": 0.0775,
+      "step": 6731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6781728425394484,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 6732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.529002946543457,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 6733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7701788950845477,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 6734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9708989842820195,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 6735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1429114677638728,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 6736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4057403448651914,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 6737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.139774492026268,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 6738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4768006470795574,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 6739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2821712438383577,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 6740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5806691815524827,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 6741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.969640199747436,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 6742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5288195882259967,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 6743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0692591809459915,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 6744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.744719312839152,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 6745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.736838832794408,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 6746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0156113971576484,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 6747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4398485155694245,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 6748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9816890433253476,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 6749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09873480292069203,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 6750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4568829757029187,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 6751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5856752752479033,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 6752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39985239091892705,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 6753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.077519514363438,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 6754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.526306702073353,
+      "learning_rate": 2e-05,
+      "loss": 0.1202,
+      "step": 6755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.234541435771369,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 6756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.800354671834357,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 6757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.647330334634675,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 6758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8828515357706026,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 6759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6824393852620794,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 6760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3367717435890123,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 6761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.684542869852233,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 6762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8827871926290083,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 6763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2536578644820073,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 6764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.057098318772725,
+      "learning_rate": 2e-05,
+      "loss": 0.136,
+      "step": 6765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1338012063807814,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 6766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2909278201380228,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 6767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4833769459405,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 6768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6179819296263482,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 6769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.13268074344209654,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 6770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0009093495869674,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 6771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2723824035781511,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 6772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2995487365243328,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 6773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.146885203596852,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 6774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6811418190363043,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 6775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0126410444943,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 6776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.326472769234172,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 6777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.925491307117913,
+      "learning_rate": 2e-05,
+      "loss": 0.0997,
+      "step": 6778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.648412799196885,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 6779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2381313967359002,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 6780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36484020811399753,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 6781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2257897925580679,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 6782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2777621635687515,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 6783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.66092297767892,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 6784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6512593604749547,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 6785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.498557508480953,
+      "learning_rate": 2e-05,
+      "loss": 0.0941,
+      "step": 6786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5287201332783513,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 6787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0188521935074237,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 6788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.013724088685633,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 6789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4317976590952286,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 6790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.566623587049862,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 6791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4769600007782379,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 6792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6803243218841722,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 6793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5770521388460548,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 6794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9132912612637716,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 6795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7318906765113194,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 6796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.841825565656455,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 6797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4645942170275483,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 6798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.102959974707425,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 6799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4998430328892003,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 6800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1536299946245223,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 6801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4532614288306656,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 6802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.730433213292607,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 6803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.703967104965911,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 6804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.793746477163257,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 6805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8859923492084087,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 6806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.84091006598998,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 6807
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.924457323013128,
+      "learning_rate": 2e-05,
+      "loss": 0.057,
+      "step": 6808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6956249599548177,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 6809
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2841236638949964,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 6810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1021845581161127,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 6811
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.302936784620354,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 6812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.892675713664397,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 6813
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3211809270554415,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 6814
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8151347217201392,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 6815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20159051907395004,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 6816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3537504534663376,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 6817
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.978208956318026,
+      "learning_rate": 2e-05,
+      "loss": 0.2073,
+      "step": 6818
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09520028320878192,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 6819
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9074820911384007,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 6820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.140129087601185,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 6821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0822489275064526,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 6822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2583674278861314,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 6823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3566681048031509,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 6824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.109400520568445,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 6825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.194323333853435,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 6826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.703369658759574,
+      "learning_rate": 2e-05,
+      "loss": 0.0904,
+      "step": 6827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0295925714320013,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 6828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.466049796482992,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 6829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.462562823583434,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 6830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.298803664816082,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 6831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7534837331019464,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 6832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3330457260353317,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 6833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.630042688243724,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 6834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.139673731700958,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 6835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3350532414576537,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 6836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0588737764933496,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 6837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1417674553553887,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 6838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9593019389717532,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 6839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5958156289229946,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 6840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.605551835854756,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 6841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.941432497459934,
+      "learning_rate": 2e-05,
+      "loss": 0.1024,
+      "step": 6842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.42663960335808,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 6843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.166068489061663,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 6844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.14252840895777,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 6845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.640889067152776,
+      "learning_rate": 2e-05,
+      "loss": 0.1134,
+      "step": 6846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.627784303312266,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 6847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7845415584431703,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 6848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.400569403177641,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 6849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.377655679882972,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 6850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6595940569238177,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 6851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2893580242450344,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 6852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8037147641529527,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 6853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.768316048053914,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 6854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0079083252835326,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 6855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2393596772308624,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 6856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.090063350940254,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 6857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7926209887502325,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 6858
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8076157438854983,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 6859
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0674473928564674,
+      "learning_rate": 2e-05,
+      "loss": 0.0955,
+      "step": 6860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7840139820745287,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 6861
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3959499948482559,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 6862
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.085229052524478,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 6863
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.906587351981087,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 6864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4467425720382756,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 6865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5564982541788606,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 6866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7082603097830917,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 6867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8016967979534035,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 6868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0517635363013065,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 6869
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.41345552881529,
+      "learning_rate": 2e-05,
+      "loss": 0.131,
+      "step": 6870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8663750356515402,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 6871
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6434070795133964,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 6872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2736445533598542,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 6873
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9298323603940948,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 6874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3533855777593824,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 6875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7406878243813795,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 6876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9566272310943547,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 6877
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8499635641273136,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 6878
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.230167324113793,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 6879
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6728910406991586,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 6880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.41229390662441,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 6881
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7813067510833382,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 6882
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.578604720161765,
+      "learning_rate": 2e-05,
+      "loss": 0.0815,
+      "step": 6883
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8624855275252926,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 6884
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.622619027589108,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 6885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.55984503432524,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 6886
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2803645048397114,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 6887
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.582323049362282,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 6888
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4516933779483808,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 6889
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2194337462908877,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 6890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7265286593810134,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 6891
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0128649254130375,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 6892
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0614089579373838,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 6893
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1540592257612055,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 6894
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.352362291373133,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 6895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.6688853108128034,
+      "learning_rate": 2e-05,
+      "loss": 0.1272,
+      "step": 6896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1348960676066695,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 6897
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7952575268785242,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 6898
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.268093369306226,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 6899
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.594106881794241,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 6900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8318960957400856,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 6901
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3348033619481388,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 6902
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.904971075333402,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 6903
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6666367369189665,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 6904
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3836168561880613,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 6905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.839303771321869,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 6906
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3017250339519416,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 6907
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7664017143465273,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 6908
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3667404751889713,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 6909
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.712748644521876,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 6910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5233771730646174,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 6911
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4390674835350734,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 6912
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8632253913586624,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 6913
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7071149997742455,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 6914
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.86329290744131,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 6915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3699739379093665,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 6916
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7329585847082691,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 6917
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5307142844622543,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 6918
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.330946010495819,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 6919
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.23736433370610036,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 6920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1018650446941587,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 6921
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8072397810146511,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 6922
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.738934614855263,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 6923
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8149710943479187,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 6924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7066692669126368,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 6925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4741828232064154,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 6926
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0738087049456255,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 6927
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.304628636529806,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 6928
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.287586314817206,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 6929
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.49732475676407,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 6930
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.451766385744772,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 6931
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.370944077670973,
+      "learning_rate": 2e-05,
+      "loss": 0.1183,
+      "step": 6932
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.203792022758199,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 6933
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3999784813993653,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 6934
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.334060197632966,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 6935
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5394312661484444,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 6936
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5381174059887877,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 6937
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.483754110424094,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 6938
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9603092965516877,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 6939
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.085106846521781,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 6940
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.464842782723166,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 6941
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.775367006562504,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 6942
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6578548202445553,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 6943
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0150407724694563,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 6944
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1633874356679166,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 6945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4146058951120335,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 6946
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2486190797530055,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 6947
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.87915494228633,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 6948
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9383818611370832,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 6949
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8682487111827578,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 6950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8539613402902697,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 6951
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4676128611795696,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 6952
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.091695255301384,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 6953
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9922391398599755,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 6954
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7706299231494316,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 6955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9434787698183897,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 6956
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8652135398460021,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 6957
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.385205077625214,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 6958
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.350924059830144,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 6959
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8956927441470475,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 6960
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.198747425503159,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 6961
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.163213680447822,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 6962
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9976713810281659,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 6963
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.529695016583616,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 6964
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.963278456509855,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 6965
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4630178235807205,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 6966
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7806416861494412,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 6967
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8811330095377081,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 6968
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5995390227966364,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 6969
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.497975254460525,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 6970
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7578654136818406,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 6971
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5985562937930973,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 6972
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3347382544618587,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 6973
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8518700700422721,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 6974
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.537254827091685,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 6975
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5902821829048606,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 6976
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1259229371004933,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 6977
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7004885731091138,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 6978
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0434272234647515,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 6979
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8522517856790608,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 6980
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4564386957731135,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 6981
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5269554795905516,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 6982
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.3632058433362095,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 6983
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0440905245379994,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 6984
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.219695091444028,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 6985
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5349577564304104,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 6986
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.23129132133608835,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 6987
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3510132208305983,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 6988
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.490413390040384,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 6989
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.6795441033522955,
+      "learning_rate": 2e-05,
+      "loss": 0.1207,
+      "step": 6990
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.535509613753729,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 6991
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4107851860827172,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 6992
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3187410113828957,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 6993
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2301425996013986,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 6994
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5098011983685272,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 6995
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4449685592206236,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 6996
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9054344427550696,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 6997
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0162740216521855,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 6998
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9420093236752436,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 6999
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0261061143079353,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 7000
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5586428572040756,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 7001
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8211758021798894,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 7002
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7439827338491708,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 7003
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.363348740804116,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 7004
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7490644580495592,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 7005
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7011813532366649,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 7006
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7503635399620238,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 7007
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0475528843259916,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 7008
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3006164421283068,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 7009
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.368588476788547,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 7010
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.404599429075768,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 7011
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.139174032264257,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 7012
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.006732924304001,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 7013
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40267071817477695,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 7014
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7922913910377223,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 7015
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.422399490008832,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 7016
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.714561989224241,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 7017
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5568496008615471,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 7018
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3340252580982415,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 7019
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4601666053661944,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 7020
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7404677694957751,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 7021
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4637250774030717,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 7022
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7551709481473476,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 7023
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6810545701161463,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 7024
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3422107897246187,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 7025
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.593524438156929,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 7026
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9592447335339994,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 7027
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5185172835751275,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 7028
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.22695011646263,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 7029
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3953865911584467,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 7030
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8132713912570948,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 7031
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.089673647714562,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 7032
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.7419232684679935,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 7033
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.727792638235728,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 7034
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8595490019379249,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 7035
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6933300676801912,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 7036
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5084888739853608,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 7037
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.965442631032476,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 7038
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1490851412138365,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 7039
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.657197820168009,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 7040
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.491751642133251,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 7041
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.396780457391855,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 7042
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15694737376854692,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 7043
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.004451025012449,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 7044
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.574154760229191,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 7045
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0043852142213403,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 7046
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5684473148456122,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 7047
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.196609711128142,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 7048
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.953003892815099,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 7049
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.117686637720258,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 7050
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.641467999612204,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 7051
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.609757009265458,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 7052
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.74644925894361,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 7053
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2896913324016757,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 7054
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.959135528833198,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 7055
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8301766123779195,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 7056
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2801542359816884,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 7057
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8112509369808011,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 7058
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.335299503632774,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 7059
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7360049312758834,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 7060
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4148863053839333,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 7061
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1488936160765284,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 7062
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0424923407553037,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 7063
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49427377467457717,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 7064
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1556377923928194,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 7065
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4856691039860446,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 7066
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6833596384247858,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 7067
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49241617733568527,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 7068
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6282151034384817,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 7069
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.542147028893476,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 7070
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3085860601505231,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 7071
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.217948661409555,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 7072
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.143112188218425,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 7073
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.982215924664846,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 7074
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4504611301519896,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 7075
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.572906483824396,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 7076
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2354433771926605,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 7077
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7276614735018851,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 7078
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.603824125014039,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 7079
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2814963112071767,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 7080
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7553589364459112,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 7081
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.762843987164532,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 7082
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6445938062209546,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 7083
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4701526518469659,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 7084
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.158690001660045,
+      "learning_rate": 2e-05,
+      "loss": 0.1601,
+      "step": 7085
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9012121951465915,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 7086
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6446631009486254,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 7087
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2671046144069236,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 7088
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3042559720806595,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 7089
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.38568235541033,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 7090
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.061435193405339,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 7091
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.946070242226303,
+      "learning_rate": 2e-05,
+      "loss": 0.1056,
+      "step": 7092
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9911355027245035,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 7093
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3663627411604769,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 7094
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4512283865808855,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 7095
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.33858254695785,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 7096
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4722274844157546,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 7097
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8840016959366594,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 7098
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 29.043080919840552,
+      "learning_rate": 2e-05,
+      "loss": 0.1219,
+      "step": 7099
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2147514956467425,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 7100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9095335700879228,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 7101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3758116719706092,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 7102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.043429954937952,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 7103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.071146440789042,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 7104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4168415914357091,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 7105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7795809237458566,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 7106
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7796516609099904,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 7107
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9710293994797908,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 7108
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.220199406769766,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 7109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6722128135883856,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 7110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.606347075517318,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 7111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9474836742025337,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 7112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.765101410177072,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 7113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9696657073569996,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 7114
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0194870511701766,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 7115
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.155894328796046,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 7116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1243738691583927,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 7117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.582953215027591,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 7118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2038302505807223,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 7119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7191986885484722,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 7120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0890744683416689,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 7121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.397414197291328,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 7122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0295742754124169,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 7123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5903626890802625,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 7124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1134585942512,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 7125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.681528152371112,
+      "learning_rate": 2e-05,
+      "loss": 0.0721,
+      "step": 7126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4546524551710223,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 7127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.200362307952304,
+      "learning_rate": 2e-05,
+      "loss": 0.1425,
+      "step": 7128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5883588719955424,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 7129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8109131147373714,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 7130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.660913473283249,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 7131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4503327861084191,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 7132
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0383651534825367,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 7133
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.114295013857681,
+      "learning_rate": 2e-05,
+      "loss": 0.0776,
+      "step": 7134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.0611753211371795,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 7135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.924216975692569,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 7136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0268309427908133,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 7137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0797166967026306,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 7138
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.827158680677628,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 7139
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3124876279757727,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 7140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.707606310954314,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 7141
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.359645105193273,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 7142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.751385821263229,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 7143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2152242769847499,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 7144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.85080172204176,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 7145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5881639016528637,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 7146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6034853673349958,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 7147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1205687602056797,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 7148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7982605609305258,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 7149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6079334358131052,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 7150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7722068097746404,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 7151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5655705608108774,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 7152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.014522517120338,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 7153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.359454627303874,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 7154
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2977397923482545,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 7155
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.654393882747175,
+      "learning_rate": 2e-05,
+      "loss": 0.0729,
+      "step": 7156
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.978023636624558,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 7157
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2191908770247498,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 7158
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7080219977700725,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 7159
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.552277657379035,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 7160
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4210780498860005,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 7161
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8389594001084038,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 7162
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1137671958910995,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 7163
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5407361896578347,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 7164
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1651853228176643,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 7165
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9720850726680399,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 7166
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4498233819141961,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 7167
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3171068359147788,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 7168
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3371875573829914,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 7169
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9726413235241282,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 7170
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1215537332642156,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 7171
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3501965422409277,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 7172
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8191231701983956,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 7173
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3917701322215232,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 7174
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7753181206442563,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 7175
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.19766991540378,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 7176
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7517020420301312,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 7177
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8167501573962389,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 7178
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5040878384435479,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 7179
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6258159936201526,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 7180
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.615949843312009,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 7181
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7999014536054578,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 7182
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.397636923423068,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 7183
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1118247041925127,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 7184
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0429613850299995,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 7185
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.409194729761512,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 7186
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4814214804584016,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 7187
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0511124407051122,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 7188
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.12944044522025,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 7189
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.532185793606201,
+      "learning_rate": 2e-05,
+      "loss": 0.1041,
+      "step": 7190
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6159024500250088,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 7191
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0213024997439626,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 7192
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.078484984460422,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 7193
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1523690431049651,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 7194
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4858153830653371,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 7195
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9411785521810594,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 7196
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7183010555359777,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 7197
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.396656779435534,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 7198
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4989524998483967,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 7199
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7174880508118027,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 7200
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3044608827328161,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 7201
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8236064353954531,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 7202
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.543712399130187,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 7203
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.285712268781339,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 7204
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6766683724598836,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 7205
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5554785947222287,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 7206
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6025416660694496,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 7207
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4446752908173947,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 7208
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9000510611884358,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 7209
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.145383655332589,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 7210
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2599701916577124,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 7211
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4146535070408266,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 7212
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4113119464529127,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 7213
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2043849243483662,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 7214
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3797824226464093,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 7215
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.42557983502293,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 7216
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9315762141075394,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 7217
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5489524223798046,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 7218
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.74552371103319,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 7219
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.881537183865305,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 7220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.44640383924368054,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 7221
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0048765844035312,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 7222
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.021838878562458,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 7223
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.403060679883154,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 7224
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2729513690745735,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 7225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.226650599334711,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 7226
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.537242436608868,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 7227
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.140083077324416,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 7228
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8390737863133216,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 7229
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1703422132398384,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 7230
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.903897757153202,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 7231
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.378474254251495,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 7232
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.022289524891367,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 7233
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.602092712166562,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 7234
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9229562090620302,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 7235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4563851045137945,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 7236
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25731906487589556,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 7237
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4693664717984531,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 7238
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.14751940951095482,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 7239
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.953355577209855,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 7240
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.039677435860484,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 7241
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.375489033995898,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 7242
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.067973647563973,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 7243
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8508181210511636,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 7244
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.564683402676271,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 7245
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8247942621713515,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 7246
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8042999776287867,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 7247
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.819941336812893,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 7248
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1637893970839723,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 7249
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.141836280938849,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 7250
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2526360849396883,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 7251
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8978642721990777,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 7252
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0353478020551345,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 7253
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9107800748447409,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 7254
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.361692935195461,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 7255
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4586951785372048,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 7256
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3887810644045493,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 7257
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.091680000816321,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 7258
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1188096340800238,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 7259
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1494726070510148,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 7260
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5992596489199884,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 7261
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.847939789679667,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 7262
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.64646713426814,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 7263
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38014463213294597,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 7264
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2563940070060662,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 7265
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.758493412504716,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 7266
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2543854889864954,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 7267
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7386086087311576,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 7268
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3569028661106726,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 7269
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9107173415440977,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 7270
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.132604882264706,
+      "learning_rate": 2e-05,
+      "loss": 0.1238,
+      "step": 7271
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3852562551965146,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 7272
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4912006938223845,
+      "learning_rate": 2e-05,
+      "loss": 0.0894,
+      "step": 7273
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5628428346131527,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 7274
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.479759487364317,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 7275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4247183525662348,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 7276
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2607243466867426,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 7277
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.0218224190960346,
+      "learning_rate": 2e-05,
+      "loss": 0.104,
+      "step": 7278
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.877941614774846,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 7279
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3417162919591163,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 7280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3651545390634294,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 7281
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0481685219488124,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 7282
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5745534491133233,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 7283
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1820366345484943,
+      "learning_rate": 2e-05,
+      "loss": 0.1039,
+      "step": 7284
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8297813932204328,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 7285
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8449003007861846,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 7286
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5757411834151068,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 7287
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5846225191551144,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 7288
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8282786172862041,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 7289
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4958669789970695,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 7290
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3392002795989362,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 7291
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.367053948167763,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 7292
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5190923264118004,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 7293
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6607278714891458,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 7294
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4010547462429546,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 7295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4567945354535148,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 7296
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6446323680824262,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 7297
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.952903603269118,
+      "learning_rate": 2e-05,
+      "loss": 0.172,
+      "step": 7298
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4513582612512293,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 7299
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.429152462181686,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 7300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4671024253156926,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 7301
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4033397087993584,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 7302
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.804454880951729,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 7303
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1922114179941867,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 7304
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0856152612120866,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 7305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1684856483025376,
+      "learning_rate": 2e-05,
+      "loss": 0.0855,
+      "step": 7306
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.754607726401622,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 7307
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7236938811552378,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 7308
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8313282248611039,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 7309
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5146451190815484,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 7310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.032884458496485,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 7311
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.877294535156515,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 7312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.762605704989264,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 7313
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.555970409371271,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 7314
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6040931618764416,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 7315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25533559928157096,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 7316
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.623144856637507,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 7317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6332612949496375,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 7318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1165376920850811,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 7319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2618842451460557,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 7320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8341705910228925,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 7321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2370047743974228,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 7322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5480869054130733,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 7323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4491855175392852,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 7324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8731660356210476,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 7325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.005218254824447,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 7326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6089490308257641,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 7327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7199974844146864,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 7328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9294698812404976,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 7329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.906042768411776,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 7330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.243638488381612,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 7331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8364483712604134,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 7332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.714165252866329,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 7333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6196285382357511,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 7334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5495460405369,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 7335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1628881585784065,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 7336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.90184318332842,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 7337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4008037810629204,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 7338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5961871492802289,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 7339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3865036918544213,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 7340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5078997023609939,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 7341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8210772434633729,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 7342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7413454197649045,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 7343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3359438355055822,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 7344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.48713636245788,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 7345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9207400805451735,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 7346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.771937942059728,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 7347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.67686203469956,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 7348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5290992060807618,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 7349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.836912634776492,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 7350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1695446438456627,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 7351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.7997102093639015,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 7352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3488490612073713,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 7353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5046355621786074,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 7354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5671941978845503,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 7355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.302298545853814,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 7356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0763166991332946,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 7357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5713981328674029,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 7358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9366137448108953,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 7359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.540319700771717,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 7360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9203390791615627,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 7361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9173607882380794,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 7362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8565036176568703,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 7363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1951134210493846,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 7364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7411428226741317,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 7365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5109391255779974,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 7366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.204498751174465,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 7367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8135667547117774,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 7368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.503172387521073,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 7369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6240658687430463,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 7370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5135619187246716,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 7371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.036193891497377,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 7372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2541804504059537,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 7373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7834031760415544,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 7374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.874656415904331,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 7375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4631458258501224,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 7376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.0201334907443025,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 7377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.304581588599078,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 7378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.368617345006351,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 7379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.859080148305324,
+      "learning_rate": 2e-05,
+      "loss": 0.1313,
+      "step": 7380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0645608720478235,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 7381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.231295996997726,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 7382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7530932336876972,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 7383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8644088044607505,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 7384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.331285555254609,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 7385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2095608688364443,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 7386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9434691738375947,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 7387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5539218292412781,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 7388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4935260780440793,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 7389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.747282630025648,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 7390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.770466201722487,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 7391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.694347053953086,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 7392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8239459664974706,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 7393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.504212604221638,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 7394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9093424224261768,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 7395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37942689342751484,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 7396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7083520495910385,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 7397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0593488189252802,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 7398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.163997124562545,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 7399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5356088995021566,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 7400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.284674174786485,
+      "learning_rate": 2e-05,
+      "loss": 0.0861,
+      "step": 7401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7732355608425387,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 7402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1906604279661277,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 7403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3321480797687544,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 7404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.628621101416865,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 7405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.784462840384541,
+      "learning_rate": 2e-05,
+      "loss": 0.0991,
+      "step": 7406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.039577725967425,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 7407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.441975324301429,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 7408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7001907795396224,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 7409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.048755577015252,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 7410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3385835732555968,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 7411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2801149235589486,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 7412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.147084894538922,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 7413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7412529130592165,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 7414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6175609858895534,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 7415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6920936098853474,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 7416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2827858907005136,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 7417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.296322814584334,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 7418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8398768218869914,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 7419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7079089808455916,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 7420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2826414518537086,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 7421
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0021619953665626,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 7422
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.2063565789106985,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 7423
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.381078851030149,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 7424
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.355042680001296,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 7425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5046572353571626,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 7426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.927115496959743,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 7427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.544835309468834,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 7428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8875085034715713,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 7429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6152237664170939,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 7430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6125533796633853,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 7431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.656969040431507,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 7432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4179484048019588,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 7433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.865507028628145,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 7434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.076428992474722,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 7435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8010637680851964,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 7436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2271509296470535,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 7437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.314325513389355,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 7438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7097113307062908,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 7439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.071431597371519,
+      "learning_rate": 2e-05,
+      "loss": 0.1277,
+      "step": 7440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.356979313445609,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 7441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8671377646234704,
+      "learning_rate": 2e-05,
+      "loss": 0.115,
+      "step": 7442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9748984709126716,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 7443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8628610581019194,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 7444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9338097653958195,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 7445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.045094538324528,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 7446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6027673422732451,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 7447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4829100636789756,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 7448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3488312836937884,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 7449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3980758028749412,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 7450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.399617427256591,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 7451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9146330314000066,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 7452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2776164550821942,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 7453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.772560088295074,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 7454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.57629575561618,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 7455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5477826137757336,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 7456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.318870434075331,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 7457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4408950636709124,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 7458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1104846297085014,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 7459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6266351689092868,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 7460
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.36698605526941,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 7461
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6157782122276618,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 7462
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.127711529553954,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 7463
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.1842999986922464,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 7464
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7270026521183737,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 7465
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.524101699409466,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 7466
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9952374835766451,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 7467
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.263745073874478,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 7468
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8949918521313396,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 7469
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.667684818283808,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 7470
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.761894252871422,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 7471
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6373797069449824,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 7472
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.238731479902428,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 7473
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4983947368538297,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 7474
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2071028999864513,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 7475
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1065409410874247,
+      "learning_rate": 2e-05,
+      "loss": 0.0776,
+      "step": 7476
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.210453502775169,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 7477
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3443388868346942,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 7478
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8131637063222412,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 7479
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9288006470192656,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 7480
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.640208916923053,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 7481
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7653043425792316,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 7482
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.854993801564778,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 7483
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0516162943577743,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 7484
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8292289429349565,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 7485
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9258348473594594,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 7486
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5797058349342041,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 7487
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6489841493425397,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 7488
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5305391335380425,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 7489
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0722128155042576,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 7490
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.153815512255806,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 7491
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7089345695867535,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 7492
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7955284713664215,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 7493
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.443263551340918,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 7494
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8967632782364869,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 7495
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.313032431368034,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 7496
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7817506330516217,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 7497
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.028477463858205,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 7498
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8092324871540523,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 7499
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.612196898331103,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 7500
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5238911271242053,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 7501
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.405349867215576,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 7502
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2365015760967792,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 7503
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.981949833636348,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 7504
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.147656440943005,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 7505
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.167117609878704,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 7506
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7023744241612808,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 7507
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.702655553019852,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 7508
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7791895574249437,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 7509
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7876350005061967,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 7510
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4061059483856955,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 7511
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4036449038396905,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 7512
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0992546030211923,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 7513
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2181072109924687,
+      "learning_rate": 2e-05,
+      "loss": 0.1562,
+      "step": 7514
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6045429165311665,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 7515
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8725883423313521,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 7516
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3211322023107153,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 7517
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.578352029466346,
+      "learning_rate": 2e-05,
+      "loss": 0.0921,
+      "step": 7518
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0566698447539695,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 7519
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.38521717242377,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 7520
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8467749970644207,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 7521
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0670482114537803,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 7522
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.751685844405134,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 7523
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28991496195835187,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 7524
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9356663892378858,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 7525
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4814927879979345,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 7526
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.974589223404909,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 7527
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.737016964321406,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 7528
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5247938329447854,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 7529
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1887922557191875,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 7530
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4815263417966547,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 7531
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8912008331348011,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 7532
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.907693092812841,
+      "learning_rate": 2e-05,
+      "loss": 0.107,
+      "step": 7533
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2747879474450063,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 7534
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6006903580861638,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 7535
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8059309601751976,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 7536
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6387487673081942,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 7537
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.000185710828864,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 7538
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.49356591499096875,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 7539
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.228025374027863,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 7540
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.270219836327109,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 7541
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8523400283018935,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 7542
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7482047409293586,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 7543
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37127811469924804,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 7544
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5749823494980243,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 7545
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1242851850215985,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 7546
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8116340999680722,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 7547
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4064769885518213,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 7548
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4053945604623785,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 7549
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6808582614489065,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 7550
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.47072833931530766,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 7551
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9570787009084951,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 7552
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2468383382125598,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 7553
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1110253214830441,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 7554
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6590994552952109,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 7555
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7768574252484595,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 7556
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5107692526209675,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 7557
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6084392012043687,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 7558
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4126167782996757,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 7559
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.076810148353351,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 7560
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.499845496803663,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 7561
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2429852710616425,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 7562
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.6743412716502135,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 7563
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1576870931439198,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 7564
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7598631585792894,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 7565
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.109497628796471,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 7566
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6615638420382237,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 7567
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2633800611854116,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 7568
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.800230143643035,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 7569
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5498511588844943,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 7570
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9484247031323139,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 7571
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.269126663295588,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 7572
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.562902932665626,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 7573
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8266039005805526,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 7574
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8811090366976715,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 7575
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6524593457298287,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 7576
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0695944834951374,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 7577
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.07196838919946,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 7578
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0220346190283975,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 7579
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5613222773598426,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 7580
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.493257308852662,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 7581
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4957107958188852,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 7582
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8441484005211145,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 7583
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6311881276524145,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 7584
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9565292194276704,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 7585
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2097010305834965,
+      "learning_rate": 2e-05,
+      "loss": 0.0703,
+      "step": 7586
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3896196996380672,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 7587
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.64401454063387,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 7588
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.920014323772808,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 7589
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7574449021221438,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 7590
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.34052940346597,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 7591
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8888762261854914,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 7592
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9682085412842443,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 7593
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38237700700809923,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 7594
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3152794501620164,
+      "learning_rate": 2e-05,
+      "loss": 0.0858,
+      "step": 7595
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5225026197013447,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 7596
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0263049889908706,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 7597
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.134410861790958,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 7598
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3108948474442594,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 7599
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0895022058766597,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 7600
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3029742859230846,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 7601
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.212743553937497,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 7602
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5083016506401266,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 7603
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7365009210075715,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 7604
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9148268438594445,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 7605
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2834650292055263,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 7606
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0454198892142803,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 7607
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5012986624815285,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 7608
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1870160248652906,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 7609
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0009969271937567,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 7610
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1326191178049003,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 7611
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.491640153202468,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 7612
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.467284431510259,
+      "learning_rate": 2e-05,
+      "loss": 0.1082,
+      "step": 7613
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.56750862850622,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 7614
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.537102084887811,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 7615
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.804168348524743,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 7616
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8205282282019417,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 7617
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7211691956269917,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 7618
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6496415453620688,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 7619
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.589484959584794,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 7620
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3489981784115448,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 7621
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.632014031833708,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 7622
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4056360812023911,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 7623
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.31165676791405206,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 7624
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.253034579850974,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 7625
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7534131428829162,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 7626
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7663073371498936,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 7627
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7470958317640826,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 7628
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.456516192646155,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 7629
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4818439789625866,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 7630
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.933851221755472,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 7631
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.660514803293271,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 7632
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.06678020146501347,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 7633
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8862074667727746,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 7634
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.380976562707174,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 7635
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5308208337231672,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 7636
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4618785091910906,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 7637
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0170495023930477,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 7638
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1593854079560884,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 7639
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7014204769344502,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 7640
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.859103497398772,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 7641
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.314012526811242,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 7642
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5506321841131827,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 7643
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.814396162340018,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 7644
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.83915497820227,
+      "learning_rate": 2e-05,
+      "loss": 0.1059,
+      "step": 7645
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3287280025047274,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 7646
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.701051172435213,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 7647
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.522146999412948,
+      "learning_rate": 2e-05,
+      "loss": 0.0815,
+      "step": 7648
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9060162209833423,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 7649
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.195446512605004,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 7650
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28237104842303173,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 7651
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7337924188573177,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 7652
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1253899903136464,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 7653
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8487920929888316,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 7654
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8940489355749652,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 7655
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2040972558754341,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 7656
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5283514214627337,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 7657
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8981747543551646,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 7658
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3615589126501267,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 7659
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7915753303509071,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 7660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.2695946858850995,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 7661
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.615613536781741,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 7662
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6777996944204493,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 7663
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6992814019973521,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 7664
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7587277890492665,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 7665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.100360627886572,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 7666
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3032600766947695,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 7667
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.499687652149141,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 7668
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9458345714460377,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 7669
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.226937556232114,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 7670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3395624007536675,
+      "learning_rate": 2e-05,
+      "loss": 0.0996,
+      "step": 7671
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.49493690835209514,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 7672
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3692618627920627,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 7673
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5578423628306599,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 7674
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8799174175078424,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 7675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.023268125073891,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 7676
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6325001517184394,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 7677
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.7235890405456145,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 7678
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.477073929503878,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 7679
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8821114989300657,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 7680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2542101659728875,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 7681
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.279349246242176,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 7682
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.39928846143076,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 7683
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5866272343582724,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 7684
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5610833279191585,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 7685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1405569082341815,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 7686
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9367459443561689,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 7687
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3516811252640697,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 7688
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4962903085376533,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 7689
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1490038558689422,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 7690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.138085857212591,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 7691
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8643135074470931,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 7692
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1492883034929022,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 7693
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6660686706718173,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 7694
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8257102984822215,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 7695
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.42789434763693,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 7696
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8164818532653133,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 7697
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.074646664299904,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 7698
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7731525973274467,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 7699
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.948958467189697,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 7700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.031933866942445,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 7701
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1683602585911956,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 7702
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3803624537775074,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 7703
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1660251134863215,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 7704
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0966560351774897,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 7705
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.321439335227899,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 7706
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5154555938327624,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 7707
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.463890039042187,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 7708
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24539595781839094,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 7709
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6336093641960653,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 7710
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.882191164621776,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 7711
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3759143563498135,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 7712
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3520378570169878,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 7713
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16030116619327237,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 7714
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2826323827853194,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 7715
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.545122174614313,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 7716
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6977550345902315,
+      "learning_rate": 2e-05,
+      "loss": 0.0938,
+      "step": 7717
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.498290757043576,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 7718
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.204049337450283,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 7719
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5374864399108816,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 7720
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.41518638142144,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 7721
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5598320881635852,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 7722
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4348019697870793,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 7723
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9641386227717921,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 7724
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.5704684561302855,
+      "learning_rate": 2e-05,
+      "loss": 0.0884,
+      "step": 7725
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.087772662235634,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 7726
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44712260929073205,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 7727
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.075058237160114,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 7728
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.339282425981461,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 7729
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9373968773747483,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 7730
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3189883391882165,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 7731
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9216832388715965,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 7732
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5364129356099931,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 7733
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3106006959544747,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 7734
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.180613987403673,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 7735
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.51296474720772,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 7736
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.244795077174547,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 7737
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4027071385393284,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 7738
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6983702188938543,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 7739
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.907111989226137,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 7740
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6354360473262015,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 7741
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.681813771177606,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 7742
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.112357214214228,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 7743
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.256795820592505,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 7744
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2353496426515345,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 7745
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.575884313491287,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 7746
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2532430165528723,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 7747
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5238965142742276,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 7748
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.469705131320123,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 7749
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7197816700204862,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 7750
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2115469048890675,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 7751
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8321477206363996,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 7752
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.36628394861385016,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 7753
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.20972347290123,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 7754
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2081956880485727,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 7755
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6455044190841331,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 7756
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0882348828603723,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 7757
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0974574500986582,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 7758
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2282956749739076,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 7759
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.768965270620684,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 7760
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2555258205619337,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 7761
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2119254880361643,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 7762
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3629601089330365,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 7763
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1389012879952167,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 7764
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5879798103990062,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 7765
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.12615125210046518,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 7766
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.530226177438606,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 7767
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9229372103117104,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 7768
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.565252825823273,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 7769
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.212645106502807,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 7770
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0810077148177175,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 7771
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7341337293912351,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 7772
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8958102415494582,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 7773
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.507932272784023,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 7774
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0874976059459773,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 7775
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.37120632814142,
+      "learning_rate": 2e-05,
+      "loss": 0.1064,
+      "step": 7776
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2313414854345386,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 7777
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.209309334533541,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 7778
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0505566283236156,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 7779
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.833942789331175,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 7780
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1020548376330697,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 7781
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7876020575778921,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 7782
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7963042306378862,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 7783
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.19923898182777477,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 7784
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.465366443203273,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 7785
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3031530402517255,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 7786
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.674499223286236,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 7787
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5623326317487515,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 7788
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3515576150342854,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 7789
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0640411419622504,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 7790
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5768304935423434,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 7791
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0465946948887566,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 7792
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1949036341481738,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 7793
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7662363530945989,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 7794
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4066703168240113,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 7795
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.094793094787731,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 7796
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.979622777066876,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 7797
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.48087836701207,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 7798
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.788975732987054,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 7799
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3466393526431135,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 7800
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7341796939948291,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 7801
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.724359977065795,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 7802
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.57067061599188,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 7803
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3660513364780759,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 7804
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9459078197974757,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 7805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.156211852689902,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 7806
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6782532001875698,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 7807
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7455439444495,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 7808
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.57596718983286,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 7809
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.075510932263848,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 7810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9662446929801818,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 7811
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5487068965846735,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 7812
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7982785921129363,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 7813
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.080872994691664,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 7814
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.573304239648076,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 7815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.682595235195264,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 7816
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.918929222757769,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 7817
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.24335026912711,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 7818
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2970438008585214,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 7819
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1738285637072785,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 7820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8229803092382768,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 7821
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.788346313504114,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 7822
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.866391437832232,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 7823
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.697668046922738,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 7824
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.714992855602097,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 7825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7862847033363775,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 7826
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.035898164581686,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 7827
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.408622067928732,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 7828
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5116617200249616,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 7829
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2900715274630055,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 7830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9284734574920186,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 7831
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3799192780341305,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 7832
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6887450864654683,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 7833
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.535719616928595,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 7834
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.652027647872728,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 7835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.864651669860768,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 7836
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0084633570114043,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 7837
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.57937802818243,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 7838
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.197145714955177,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 7839
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2936861788449665,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 7840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.949371051013693,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 7841
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3487722882831856,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 7842
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.092399617183463,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 7843
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9500401714264053,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 7844
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3347159955371972,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 7845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5858567691794536,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 7846
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22393175825018594,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 7847
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3871803482534732,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 7848
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.67532242760459,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 7849
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4503340765861987,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 7850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.561013698243404,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 7851
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9468292083988274,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 7852
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2344966257419463,
+      "learning_rate": 2e-05,
+      "loss": 0.0881,
+      "step": 7853
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.976774753802425,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 7854
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8962194727211317,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 7855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6550650766491708,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 7856
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7542823439246633,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 7857
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8401591928383337,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 7858
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5241385387469142,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 7859
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5020416299642885,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 7860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5774070896160404,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 7861
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.489258495793034,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 7862
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.505192819487464,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 7863
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.679635999635806,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 7864
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.387959930570802,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 7865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3866637742871295,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 7866
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8768920877196059,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 7867
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.877680730946797,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 7868
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5358796490070863,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 7869
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4473602368390606,
+      "learning_rate": 2e-05,
+      "loss": 0.1445,
+      "step": 7870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6322833309430632,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 7871
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2245784017200134,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 7872
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.03890268526409,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 7873
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7759466292173087,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 7874
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.886828263953124,
+      "learning_rate": 2e-05,
+      "loss": 0.0838,
+      "step": 7875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2468643591447117,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 7876
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3351054160850504,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 7877
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0230047875280297,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 7878
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9749349841424035,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 7879
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.166566137280252,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 7880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3388717575465945,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 7881
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.168378300374055,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 7882
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9200334487425372,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 7883
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2062220887750126,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 7884
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10234712620618992,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 7885
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7955339818047649,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 7886
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9021686863035941,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 7887
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0034014863636207,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 7888
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0929493494237157,
+      "learning_rate": 2e-05,
+      "loss": 0.1191,
+      "step": 7889
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0944377851281963,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 7890
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1149702143432714,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 7891
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.72054897141763,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 7892
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7108037190847976,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 7893
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2011145601322066,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 7894
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4065736801639262,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 7895
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.802973061210642,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 7896
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2502592771567733,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 7897
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.749423613314054,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 7898
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8426015243146153,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 7899
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2919486016906547,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 7900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2118077258382391,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 7901
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5353899535950476,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 7902
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1837671310023623,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 7903
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.067507723438634,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 7904
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2590019561209225,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 7905
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5507372033653758,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 7906
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0771036147391555,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 7907
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1666262207168665,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 7908
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.678427708721056,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 7909
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8482425964621503,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 7910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.255734870414302,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 7911
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0143852296361595,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 7912
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9408685872255824,
+      "learning_rate": 2e-05,
+      "loss": 0.095,
+      "step": 7913
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.042839185117797,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 7914
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1786998631321366,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 7915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8831023946747973,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 7916
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.912640101468481,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 7917
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5305698207568146,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 7918
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7772973588298653,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 7919
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7610522445216015,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 7920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6046330732677228,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 7921
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.02998888265056,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 7922
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5428658366589103,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 7923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.683631012349165,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 7924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3696018372715009,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 7925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3794282784173835,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 7926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38974735849548536,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 7927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2001015769743737,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 7928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5078686609757488,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 7929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.6911616964387175,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 7930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.731943411719959,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 7931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.313392381203995,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 7932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1877053735724368,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 7933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0192539164854786,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 7934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8585623626942138,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 7935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9380415498001664,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 7936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5511992921618085,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 7937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8677098268799934,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 7938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.777141699053045,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 7939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.022612785424013,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 7940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5604435097695595,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 7941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09734988794829744,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 7942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1393075494723166,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 7943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.093884250265406,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 7944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.727981591579269,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 7945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6683633592147609,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 7946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.130766617445759,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 7947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6499304600354483,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 7948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3361021241599358,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 7949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.707787876352999,
+      "learning_rate": 2e-05,
+      "loss": 0.0906,
+      "step": 7950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6173505470531806,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 7951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9588999317571294,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 7952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9579476776070206,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 7953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5489515954289605,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 7954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9855237417636253,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 7955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6829412837855633,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 7956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7676563951029021,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 7957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.482939819401006,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 7958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.411276206152497,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 7959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0353311067124509,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 7960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.063368442388278,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 7961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.210956586883436,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 7962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.327694641890605,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 7963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9843615786265616,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 7964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2425192623832815,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 7965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2155829630744592,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 7966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17400491489602338,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 7967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.4812092234433285,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 7968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.444523363932596,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 7969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6937739690794928,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 7970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45270388865084726,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 7971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9859814025003248,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 7972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5175874057359002,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 7973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6369327232317652,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 7974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3794357968799882,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 7975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9881272619127572,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 7976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.27313845074534,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 7977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2240861769925857,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 7978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6098003840205086,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 7979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.31248053947149,
+      "learning_rate": 2e-05,
+      "loss": 0.139,
+      "step": 7980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.222624294323048,
+      "learning_rate": 2e-05,
+      "loss": 0.1897,
+      "step": 7981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3914862976523557,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 7982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7034839636806716,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 7983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.728263418820777,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 7984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44404733528589335,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 7985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2016024678382726,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 7986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.369816788732663,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 7987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0803549139126662,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 7988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.803565014450631,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 7989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.003708857579764,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 7990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9841887018656281,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 7991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.028498046469706,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 7992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3005987895668176,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 7993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5170523817380777,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 7994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4370687054265585,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 7995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.780115736831255,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 7996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.3693018738155835,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 7997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.160001283574606,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 7998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9409646545769,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 7999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8432804496553534,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 8000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2891889171815833,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 8001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.073010976078371,
+      "learning_rate": 2e-05,
+      "loss": 0.0923,
+      "step": 8002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6911754758361526,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 8003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.054262646747019,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 8004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8983276963514168,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 8005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.22879460970388,
+      "learning_rate": 2e-05,
+      "loss": 0.1013,
+      "step": 8006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4642773970258074,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 8007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.138178631763054,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 8008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8554736968990458,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 8009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0122544843766157,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 8010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.040766473266364,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 8011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7023932608643135,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 8012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.882854492006935,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 8013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2592598306326308,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 8014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.24236126019288962,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 8015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.819160896334089,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 8016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6041928161207228,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 8017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7376259243568006,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 8018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6470187879705838,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 8019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8057324189751514,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 8020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5024714651395907,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 8021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3133146845676358,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 8022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7407245070170929,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 8023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.584196582249115,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 8024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.633997439170133,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 8025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7597802099960616,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 8026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8286998529657452,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 8027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.434457576963417,
+      "learning_rate": 2e-05,
+      "loss": 0.1125,
+      "step": 8028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.65589067609262,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 8029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3778359542683174,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 8030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9868999674213484,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 8031
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0993925324603981,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 8032
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9641142869432733,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 8033
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5430449837059131,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 8034
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3126515767667413,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 8035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.298415078511474,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 8036
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.46168567910393,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 8037
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.478858917967452,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 8038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1009528454792012,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 8039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7970518770177057,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 8040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.65146133508605,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 8041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.565954438593702,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 8042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7816012927848994,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 8043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7273623999904824,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 8044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.341729170303338,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 8045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9145553108099089,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 8046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0508934443254911,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 8047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3053378475916063,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 8048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46312739759022775,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 8049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3499307135249583,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 8050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5195563142923425,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 8051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5739658048373741,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 8052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.678200025473144,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 8053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.46500798810241,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 8054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3480322337041193,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 8055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4304336518544956,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 8056
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6494066833775785,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 8057
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1930549095832004,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 8058
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5303813339815374,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 8059
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.21015902201869957,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 8060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3181936654392583,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 8061
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4851779588888726,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 8062
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0353188477213218,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 8063
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.913133591034302,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 8064
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2632211532877557,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 8065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 24.748461126401374,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 8066
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.003249279180571,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 8067
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8259589902444593,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 8068
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9182773815693028,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 8069
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1368420441533287,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 8070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1458167970793487,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 8071
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2762784646977607,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 8072
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6999459443138796,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 8073
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.428195863130094,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 8074
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4632796011048337,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 8075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.543814077856038,
+      "learning_rate": 2e-05,
+      "loss": 0.0862,
+      "step": 8076
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1507460519589685,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 8077
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.47097363751634175,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 8078
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5080364634155385,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 8079
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2840914376210585,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 8080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.668917226796671,
+      "learning_rate": 2e-05,
+      "loss": 0.1075,
+      "step": 8081
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1380088275942477,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 8082
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8507471878638713,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 8083
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0315576173694043,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 8084
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.746152021890215,
+      "learning_rate": 2e-05,
+      "loss": 0.1021,
+      "step": 8085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6106108336040178,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 8086
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.185964150172089,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 8087
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.225550692698934,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 8088
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.509139634173964,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 8089
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.878935688763659,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 8090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 23.352605363130124,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 8091
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.700328312222007,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 8092
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.374849371360656,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 8093
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5643844974129753,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 8094
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2351573927196644,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 8095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.34674966669994856,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 8096
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5140271532266003,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 8097
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.234894058652369,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 8098
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17666652598371693,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 8099
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8463727109880794,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 8100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2439151960917811,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 8101
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5429089329682313,
+      "learning_rate": 2e-05,
+      "loss": 0.0979,
+      "step": 8102
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9453393747737753,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 8103
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.492103614246743,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 8104
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.912487270318689,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 8105
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9542448957995866,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 8106
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8837983221363364,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 8107
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8971198329301029,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 8108
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9238990731881147,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 8109
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7502033852013827,
+      "learning_rate": 2e-05,
+      "loss": 0.1255,
+      "step": 8110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2034161203090517,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 8111
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.187224432309275,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 8112
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6060959797302127,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 8113
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0387963813433805,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 8114
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0513887214129682,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 8115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.099079579778926,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 8116
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9810491720838549,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 8117
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5363921692748232,
+      "learning_rate": 2e-05,
+      "loss": 0.0893,
+      "step": 8118
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3542346888354964,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 8119
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2188859032393728,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 8120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3846633476800867,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 8121
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.055808570113746,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 8122
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.415415067050894,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 8123
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46791383551638155,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 8124
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7209991304394279,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 8125
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5874897803777216,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 8126
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5320284674666376,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 8127
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5455782300241265,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 8128
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.280263698026253,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 8129
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1205747246940059,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 8130
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.5166683905529,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 8131
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.04918579115026,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 8132
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.078430841642023,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 8133
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2773009791101393,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 8134
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6237927867723083,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 8135
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2983586660385835,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 8136
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1123357268798393,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 8137
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.443524760898013,
+      "learning_rate": 2e-05,
+      "loss": 0.0838,
+      "step": 8138
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8228988161978235,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 8139
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.388707414003311,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 8140
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7902745301689298,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 8141
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.560991321815742,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 8142
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3206620386580328,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 8143
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.563404889343276,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 8144
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0467244509723788,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 8145
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6014336286949619,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 8146
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.175587178038238,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 8147
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0509898606164239,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 8148
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.38910283238197,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 8149
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9200053230847638,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 8150
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.41189102234749,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 8151
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9113416472079634,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 8152
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2244777578294124,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 8153
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5940822876148578,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 8154
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7767227222476882,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 8155
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5898806654196447,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 8156
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.472106426754467,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 8157
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9709966204335951,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 8158
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2702245609113836,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 8159
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4151562693534505,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 8160
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2215435583442167,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 8161
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8764649117557425,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 8162
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6445855551296904,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 8163
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.85865931496736,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 8164
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4623967191304605,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 8165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2649265292849345,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 8166
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5654294503167006,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 8167
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8883357491797352,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 8168
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.333035711199086,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 8169
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.792443723317625,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 8170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4011475601069825,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 8171
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.171750176457611,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 8172
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.250314139929652,
+      "learning_rate": 2e-05,
+      "loss": 0.1179,
+      "step": 8173
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9341084486021952,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 8174
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7572674314949026,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 8175
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6193812137486017,
+      "learning_rate": 2e-05,
+      "loss": 0.0763,
+      "step": 8176
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.026959156755937,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 8177
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.44296094436055,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 8178
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.476879319745963,
+      "learning_rate": 2e-05,
+      "loss": 0.0873,
+      "step": 8179
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.0732637113054,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 8180
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7224552789870486,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 8181
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.034594448996775,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 8182
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.12703595760345254,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 8183
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.632479411712156,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 8184
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0092179541078603,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 8185
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.048730845593288,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 8186
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.20266742703417776,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 8187
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2312865353112166,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 8188
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.147507899917087,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 8189
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2208306670992974,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 8190
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.151809432210494,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 8191
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2973528330414394,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 8192
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1171488721878635,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 8193
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.261647698735921,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 8194
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45888933861717884,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 8195
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7387008332048055,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 8196
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.06576919721607043,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 8197
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9165456251222607,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 8198
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9735830403331065,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 8199
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4034517325773677,
+      "learning_rate": 2e-05,
+      "loss": 0.0884,
+      "step": 8200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0438057988100318,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 8201
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9350662327552346,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 8202
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.777846581730315,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 8203
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6128303692113637,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 8204
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.082205297666216,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 8205
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6982609296469742,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 8206
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.898771775703839,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 8207
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4681508689543916,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 8208
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.008683987484869,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 8209
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.400426103033257,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 8210
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.176397998038558,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 8211
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.072627777467692,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 8212
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4758673272151848,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 8213
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4534485264078194,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 8214
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8028760070104863,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 8215
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2543744901609464,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 8216
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6944020019315902,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 8217
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.312222951077194,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 8218
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.05603122648876,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 8219
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2847610001395746,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 8220
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.574120550624514,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 8221
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7534114420043523,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 8222
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5062543291919357,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 8223
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4566499309391048,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 8224
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.208380331024315,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 8225
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8413792966246402,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 8226
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6644045005198643,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 8227
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0890424510363688,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 8228
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3948448961094475,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 8229
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2011958422246842,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 8230
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.892925934419361,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 8231
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9534119346089331,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 8232
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6787912506011855,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 8233
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7877281576877941,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 8234
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1177006856627663,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 8235
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5438449363894584,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 8236
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7127316283848444,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 8237
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5961235921182697,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 8238
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3019607282771912,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 8239
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.709744669049594,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 8240
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.326070690478665,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 8241
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6461756227011803,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 8242
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4892824109649303,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 8243
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0705629598676807,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 8244
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.247529401642928,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 8245
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3696420886411504,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 8246
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8249469749376908,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 8247
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 16.70504786786252,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 8248
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7920935926888343,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 8249
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.225039881076754,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 8250
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6909060542507257,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 8251
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.183808072240007,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 8252
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1645293468284468,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 8253
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1980986691856523,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 8254
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7548134440025825,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 8255
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0869891198262613,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 8256
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2212805487958382,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 8257
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0141839000289854,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 8258
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.962318997631365,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 8259
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3850371435060125,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 8260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.580621991198034,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 8261
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.180370809726276,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 8262
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2809686356612002,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 8263
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.952191402060824,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 8264
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3182441920628643,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 8265
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9778003976724556,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 8266
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.338747738463186,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 8267
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.234238484233675,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 8268
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.48780605844647074,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 8269
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.217726995468158,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 8270
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2565392156538826,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 8271
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.313502761799255,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 8272
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.316450096088508,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 8273
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7500743855906404,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 8274
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.014929481389342,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 8275
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2914775535620724,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 8276
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5943188330797198,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 8277
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8548542759616301,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 8278
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9994414127387365,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 8279
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7672924607795433,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 8280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2625799301950065,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 8281
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.385554416453278,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 8282
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7831781224715555,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 8283
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8815688428951473,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 8284
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.349117515136744,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 8285
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3668906625304704,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 8286
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0396461281830507,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 8287
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2390328421777432,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 8288
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6172146468391535,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 8289
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.717275636284424,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 8290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.0184149187277045,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 8291
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.313537732369775,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 8292
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0740169266196256,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 8293
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7425718011935873,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 8294
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5765568626975928,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 8295
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7542489809529441,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 8296
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6014177222482027,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 8297
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6390312886719144,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 8298
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7317212632156668,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 8299
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.993376373391968,
+      "learning_rate": 2e-05,
+      "loss": 0.0903,
+      "step": 8300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.279823798013301,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 8301
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8804783578777,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 8302
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1729018734980974,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 8303
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.101610729696408,
+      "learning_rate": 2e-05,
+      "loss": 0.1217,
+      "step": 8304
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7872211819493535,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 8305
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0072905459464376,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 8306
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4230179437402704,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 8307
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9414114417728338,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 8308
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7990450816249697,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 8309
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.245803703585166,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 8310
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5703570727344787,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 8311
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9984117996913524,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 8312
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37792562635344235,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 8313
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9277807922959016,
+      "learning_rate": 2e-05,
+      "loss": 0.0838,
+      "step": 8314
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.470595802960839,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 8315
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8123791095274773,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 8316
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8332540827296944,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 8317
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.157333875605763,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 8318
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.785667429717993,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 8319
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.266685606580646,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 8320
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.522227387317288,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 8321
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6354250654584006,
+      "learning_rate": 2e-05,
+      "loss": 0.0967,
+      "step": 8322
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8653598542096264,
+      "learning_rate": 2e-05,
+      "loss": 0.0996,
+      "step": 8323
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3021874267611255,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 8324
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9762985596068479,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 8325
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9557417609747478,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 8326
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.329980967105834,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 8327
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8274085346393096,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 8328
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7938735440870108,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 8329
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8680561629399188,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 8330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2906633551488775,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 8331
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7145104573159797,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 8332
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2114372452758906,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 8333
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7261208453211405,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 8334
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7634550941801663,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 8335
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2698259677340153,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 8336
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0594037032217747,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 8337
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5500237958268711,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 8338
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1812934806935116,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 8339
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8465039203888746,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 8340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3851490009396374,
+      "learning_rate": 2e-05,
+      "loss": 0.1142,
+      "step": 8341
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2115583243505954,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 8342
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.43690414970909,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 8343
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7160326197264817,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 8344
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1305766140945535,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 8345
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2132506257732239,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 8346
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6304932231478373,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 8347
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.248387076506352,
+      "learning_rate": 2e-05,
+      "loss": 0.0825,
+      "step": 8348
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7302857440034995,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 8349
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8273029926120041,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 8350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5136137320575007,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 8351
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9209604414356412,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 8352
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.525672418932259,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 8353
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.225170361654456,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 8354
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9519168015986643,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 8355
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.127158678245442,
+      "learning_rate": 2e-05,
+      "loss": 0.1317,
+      "step": 8356
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9610099765148591,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 8357
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.287672460405565,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 8358
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5676668569289058,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 8359
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7685097278460082,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 8360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1066968876538006,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 8361
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4047923268548526,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 8362
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6894948259620433,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 8363
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.247373417625166,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 8364
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.43424997724731,
+      "learning_rate": 2e-05,
+      "loss": 0.0957,
+      "step": 8365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0731528321756363,
+      "learning_rate": 2e-05,
+      "loss": 0.1077,
+      "step": 8366
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5411928347998571,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 8367
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2353514660017255,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 8368
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7803225925512425,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 8369
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.575987064705139,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 8370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3798753991087858,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 8371
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3634362337612347,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 8372
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0279267342028,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 8373
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0449679070431699,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 8374
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8047307657447305,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 8375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4925170854009087,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 8376
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0040812573967104,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 8377
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5530775103683757,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 8378
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.20447646393028604,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 8379
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27360149642722636,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 8380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1448955281667668,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 8381
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.229971434405966,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 8382
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0608235765734157,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 8383
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.265963426718561,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 8384
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4866015934888344,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 8385
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.034431196570457,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 8386
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0803322794821908,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 8387
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7030555690817462,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 8388
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.222285773934469,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 8389
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6815107276612293,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 8390
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7100131974538391,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 8391
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3188167866693677,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 8392
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.171443351085814,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 8393
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2311618366724226,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 8394
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5392854754610715,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 8395
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9238075760174693,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 8396
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.328448672534607,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 8397
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5892722862496875,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 8398
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.060489660022908,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 8399
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.401733618691421,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 8400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37188742240224615,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 8401
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2768640029226987,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 8402
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.159415291398158,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 8403
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.338696714493044,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 8404
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7034299401140128,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 8405
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9662368848041353,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 8406
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8957364682893423,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 8407
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.507658196096208,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 8408
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7565748339036413,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 8409
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.070249600530567,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 8410
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9530015146277481,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 8411
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5547179092883092,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 8412
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.19206080137160203,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 8413
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4655405319967326,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 8414
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.686094970702683,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 8415
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0845955164035517,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 8416
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5581154878627284,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 8417
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6870281831459604,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 8418
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.930763962392618,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 8419
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5571023407060864,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 8420
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.224147776559011,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 8421
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9913146617719845,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 8422
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5486861136177896,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 8423
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2526847141986488,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 8424
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.773870471952885,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 8425
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6982505407670854,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 8426
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08340723764057056,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 8427
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6054292973616119,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 8428
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4690489133815136,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 8429
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2919268605855048,
+      "learning_rate": 2e-05,
+      "loss": 0.1118,
+      "step": 8430
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8847886474468254,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 8431
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6849584760243226,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 8432
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5564122317045372,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 8433
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5217852387248425,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 8434
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9378215532637104,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 8435
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.269784541177245,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 8436
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.769937767508067,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 8437
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9888392143655436,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 8438
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.870413320018176,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 8439
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31005850964642684,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 8440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8905990663734715,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 8441
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5471890833874888,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 8442
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.640828863998616,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 8443
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.846987469638124,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 8444
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.496322883989888,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 8445
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.340204430803878,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 8446
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7052351452166876,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 8447
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3864047599003837,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 8448
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.685596734963773,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 8449
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3616518333560443,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 8450
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9865664354740658,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 8451
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.02859461532109,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 8452
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7006363377602318,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 8453
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.29349449032887315,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 8454
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.437589062129354,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 8455
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0334905502226894,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 8456
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.090616414917794,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 8457
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9679984900190856,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 8458
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1051363709192894,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 8459
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4452983133805437,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 8460
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8113019620362204,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 8461
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9961970556207602,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 8462
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.826425283969887,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 8463
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.172470998976521,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 8464
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2131502650665547,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 8465
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8704683855531372,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 8466
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0006029888969348,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 8467
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7109384754474246,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 8468
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.355481668632747,
+      "learning_rate": 2e-05,
+      "loss": 0.1453,
+      "step": 8469
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1056499226318905,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 8470
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.330907925395315,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 8471
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1301044058883731,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 8472
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6899509343206867,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 8473
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.868896640196176,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 8474
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8509177120400195,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 8475
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4436962847368619,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 8476
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.895013624370678,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 8477
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.24858226217296367,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 8478
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.598349107922998,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 8479
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8337902092246987,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 8480
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41771469238210013,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 8481
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1435438460203422,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 8482
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.73718788423088,
+      "learning_rate": 2e-05,
+      "loss": 0.1337,
+      "step": 8483
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.92387003795318,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 8484
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.997084757932169,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 8485
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2867817903234013,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 8486
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8838456987298415,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 8487
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18365669186110387,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 8488
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5279472700177954,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 8489
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.762697728878406,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 8490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8577444386536824,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 8491
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3456695065223097,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 8492
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0684017937104504,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 8493
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.475163494836629,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 8494
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.10824136294678,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 8495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2605125887393056,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 8496
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5025819174068569,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 8497
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2407395404228128,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 8498
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.128353736069404,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 8499
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.027323621964129,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 8500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1938685253480177,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 8501
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0777759644571496,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 8502
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1758180360434825,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 8503
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1921638456425265,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 8504
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0597078998731626,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 8505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.665521842865614,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 8506
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0742147171788985,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 8507
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1699243714269327,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 8508
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5094036627969776,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 8509
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.558730353283243,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 8510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7873350240696317,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 8511
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7522924673124954,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 8512
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8585092147709545,
+      "learning_rate": 2e-05,
+      "loss": 0.1008,
+      "step": 8513
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9165059588961937,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 8514
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09973449889122608,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 8515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.304776959301856,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 8516
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5118269187227581,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 8517
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9640116636453085,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 8518
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5823391838380767,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 8519
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.183757556621275,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 8520
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0323666393617223,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 8521
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7672629396777306,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 8522
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.316847760740389,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 8523
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.970129077424558,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 8524
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7371601046717904,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 8525
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0588884119493287,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 8526
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.693012262208079,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 8527
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5006779967704476,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 8528
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.071947744600588,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 8529
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.837781523852289,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 8530
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6408746330891524,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 8531
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9107426403878149,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 8532
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.239162274602947,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 8533
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4878945042607925,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 8534
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.38533112040773,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 8535
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7195289233970588,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 8536
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.852675091580982,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 8537
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5239455720789163,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 8538
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.861351030179251,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 8539
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.157612798400929,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 8540
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7758070118584053,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 8541
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9737773393899087,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 8542
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5191759807212561,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 8543
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2600153712440096,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 8544
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8964948134136639,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 8545
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8587868024598468,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 8546
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2083932502336959,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 8547
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0105254452279695,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 8548
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0725515849372371,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 8549
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3637343421659738,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 8550
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8265484263560747,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 8551
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3818023041137035,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 8552
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3296731788948832,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 8553
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8964185836020637,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 8554
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7710602409245266,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 8555
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2799355887634003,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 8556
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.210087225003619,
+      "learning_rate": 2e-05,
+      "loss": 0.1399,
+      "step": 8557
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.233334019330864,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 8558
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.423171748354104,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 8559
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.268204920521954,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 8560
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5783113426252893,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 8561
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3013860997419258,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 8562
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.69580078125,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 8563
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.449819106295335,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 8564
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1705553380928466,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 8565
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2550258665440976,
+      "learning_rate": 2e-05,
+      "loss": 0.0924,
+      "step": 8566
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.345820212002833,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 8567
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.222993886195582,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 8568
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1211088120292794,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 8569
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0397145799390004,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 8570
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.990986802108227,
+      "learning_rate": 2e-05,
+      "loss": 0.1176,
+      "step": 8571
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6256186999252434,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 8572
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0716557670580826,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 8573
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3587720564216266,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 8574
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.544617352458011,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 8575
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.196577040542362,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 8576
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3698857975641017,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 8577
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.599777570861176,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 8578
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1708894781767323,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 8579
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.581840974644522,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 8580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.175312751996881,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 8581
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8737792286602376,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 8582
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.637852607756966,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 8583
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8364039037975497,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 8584
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2193509112099679,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 8585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8363959223043541,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 8586
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.401043894340093,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 8587
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0480755427047912,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 8588
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2772830171308438,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 8589
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.096131312260963,
+      "learning_rate": 2e-05,
+      "loss": 0.1141,
+      "step": 8590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.886031169552352,
+      "learning_rate": 2e-05,
+      "loss": 0.168,
+      "step": 8591
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0136408628102984,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 8592
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4791356025263576,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 8593
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0568189770125962,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 8594
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6609159997533574,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 8595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.934289585151753,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 8596
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7822091666762674,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 8597
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7747385235394013,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 8598
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1834439669519323,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 8599
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9982017799995953,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 8600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6952686250957956,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 8601
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.26761720191021204,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 8602
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.422265995922937,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 8603
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.21130589704963887,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 8604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2110135454480293,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 8605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.5547360766063125,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 8606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7901004724954752,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 8607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.457411245852063,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 8608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3057539202166573,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 8609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0898216175083872,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 8610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9984355489734273,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 8611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7708724286930505,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 8612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.547633514930539,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 8613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3235118656532707,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 8614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.263546220189965,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 8615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7574356951345418,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 8616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9279443121841378,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 8617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.006388588271724,
+      "learning_rate": 2e-05,
+      "loss": 0.138,
+      "step": 8618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3299587829197723,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 8619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.053285069373184,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 8620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1697646403170006,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 8621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.029783130424898,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 8622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2769036424349327,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 8623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4113022224082794,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 8624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8024517220684434,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 8625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4897580966822261,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 8626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.7838979133124395,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 8627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9627664251557362,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 8628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.508639570617863,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 8629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.640340114465517,
+      "learning_rate": 2e-05,
+      "loss": 0.1044,
+      "step": 8630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.663831886565973,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 8631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5791107724781666,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 8632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6012130728282066,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 8633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.425045042831886,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 8634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8347764132300246,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 8635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.293304077182951,
+      "learning_rate": 2e-05,
+      "loss": 0.0947,
+      "step": 8636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.026909166036925,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 8637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5507354498544226,
+      "learning_rate": 2e-05,
+      "loss": 0.1104,
+      "step": 8638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.49481568194550724,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 8639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4212547877892847,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 8640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2444583046676188,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 8641
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1000286271965178,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 8642
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8428867064710117,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 8643
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2544945498424502,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 8644
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5287877410694064,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 8645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2643726887778457,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 8646
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6021563661726281,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 8647
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6692274180974712,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 8648
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4973218989388374,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 8649
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7965446610247215,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 8650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5272785767994366,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 8651
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9548388222298976,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 8652
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9655976844059787,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 8653
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9956079831917121,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 8654
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.19921338317224369,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 8655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.409495486841896,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 8656
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9934049276046686,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 8657
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.138277276633717,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 8658
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0610203185081644,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 8659
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.143235000719834,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 8660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25391688691379954,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 8661
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.004995901923546,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 8662
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0427150917913945,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 8663
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1977087199741938,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 8664
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3641228808692687,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 8665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.351206371531345,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 8666
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7932242417154776,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 8667
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.487270640279399,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 8668
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8351129506452666,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 8669
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6659619589924057,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 8670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5377765065728766,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 8671
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7671146691489799,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 8672
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.36806333553643,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 8673
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6711929926711193,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 8674
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8417045108940218,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 8675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6336801456735306,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 8676
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.152770916703821,
+      "learning_rate": 2e-05,
+      "loss": 0.095,
+      "step": 8677
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9824281889716262,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 8678
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1998828453254962,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 8679
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4821400732647856,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 8680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.839174898224515,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 8681
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7636924799410146,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 8682
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.315548026954308,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 8683
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6325346057723216,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 8684
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4180161649484733,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 8685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5260785218978626,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 8686
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5070838983177898,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 8687
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.171239856087944,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 8688
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.585765360378216,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 8689
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3070272150336515,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 8690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.990962083847222,
+      "learning_rate": 2e-05,
+      "loss": 0.1017,
+      "step": 8691
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1303571939643247,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 8692
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.455279614496556,
+      "learning_rate": 2e-05,
+      "loss": 0.1121,
+      "step": 8693
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.092010686580868,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 8694
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2135068154412798,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 8695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6282683296595166,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 8696
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5329909968143602,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 8697
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.497863798819264,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 8698
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.058198970212764,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 8699
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2350269479166773,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 8700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3007335772260005,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 8701
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0935385298334532,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 8702
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8289476809560963,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 8703
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9471806753548908,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 8704
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5850583920313717,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 8705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0168759508015435,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 8706
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7282748469527343,
+      "learning_rate": 2e-05,
+      "loss": 0.057,
+      "step": 8707
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2444514151408226,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 8708
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.65016931199363,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 8709
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.844731854548994,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 8710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.038348354342325,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 8711
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5642666556758167,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 8712
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2879218863944337,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 8713
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2891459582506122,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 8714
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2968779874100804,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 8715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0411387758968371,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 8716
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4170408019554485,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 8717
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7812450027305995,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 8718
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9615021762685108,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 8719
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.166397099984081,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 8720
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4935849022705496,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 8721
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1884815738591663,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 8722
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3873681806415945,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 8723
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.624139281350368,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 8724
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.474903931156242,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 8725
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2101447679284005,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 8726
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.402671280517872,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 8727
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8475386922911627,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 8728
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45801708480461256,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 8729
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0116629566730009,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 8730
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.165780218522633,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 8731
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.470511182483718,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 8732
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.619734796977385,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 8733
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2195498676855228,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 8734
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.30364583170768,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 8735
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.104094963184327,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 8736
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.706582300518719,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 8737
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.76109407071614,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 8738
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.970273472931102,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 8739
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6862728046665192,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 8740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.395634407962486,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 8741
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0821440049236852,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 8742
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4030948848083837,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 8743
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6378618495750663,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 8744
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.745822941977983,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 8745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.324007938345189,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 8746
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7324640287348607,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 8747
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.579840217408258,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 8748
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9810242485168297,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 8749
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6649248637753031,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 8750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2094690533816115,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 8751
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.789724138834107,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 8752
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3284087382890966,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 8753
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3422083772005946,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 8754
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6851415060581788,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 8755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.976575796342924,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 8756
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25761952547367417,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 8757
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.402932173769093,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 8758
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3828127335528226,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 8759
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15347144095287668,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 8760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7535128580045547,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 8761
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40149629099024636,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 8762
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.333220969479506,
+      "learning_rate": 2e-05,
+      "loss": 0.0978,
+      "step": 8763
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7328858538797705,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 8764
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8939700209713286,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 8765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9939426907073584,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 8766
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3875105772732814,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 8767
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9081305362036465,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 8768
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40517538332235786,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 8769
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2293788061683704,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 8770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6808527962965685,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 8771
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7392956863043416,
+      "learning_rate": 2e-05,
+      "loss": 0.1002,
+      "step": 8772
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28525028246047324,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 8773
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.423948730851416,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 8774
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0753306761550365,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 8775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9712582081431943,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 8776
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.335301373879649,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 8777
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1376896512105494,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 8778
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.012954954292384,
+      "learning_rate": 2e-05,
+      "loss": 0.0929,
+      "step": 8779
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9432155586383068,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 8780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.200690707526896,
+      "learning_rate": 2e-05,
+      "loss": 0.1257,
+      "step": 8781
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.437220059180019,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 8782
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3688872123976346,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 8783
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1655790959132528,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 8784
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8001041090845527,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 8785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.441073416384432,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 8786
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.298678837862672,
+      "learning_rate": 2e-05,
+      "loss": 0.1151,
+      "step": 8787
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.356090548882408,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 8788
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.619260784690226,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 8789
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7446105337486615,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 8790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.813733826314977,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 8791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.227870076241147,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 8792
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.830568491325516,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 8793
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7797087308908393,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 8794
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9306707309729298,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 8795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4239063805967665,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 8796
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.386254298194783,
+      "learning_rate": 2e-05,
+      "loss": 0.0971,
+      "step": 8797
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5704032150560174,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 8798
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1561410848780675,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 8799
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7897171755009451,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 8800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.266943821366849,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 8801
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.138065129663108,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 8802
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6177138727678438,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 8803
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9148558548941963,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 8804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4234355391566071,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 8805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.026179710610766,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 8806
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.075523937849703,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 8807
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3933647091607235,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 8808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7557931022331656,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 8809
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.410309801697131,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 8810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3514068496236176,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 8811
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.628462861148532,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 8812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3213009767784274,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 8813
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0308177504496268,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 8814
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1099989356811153,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 8815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5256168524975036,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 8816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.359810593165056,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 8817
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9930700688749043,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 8818
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2128155283967126,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 8819
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1747264353725226,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 8820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.873551699456217,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 8821
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.808247806216429,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 8822
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4932507626378124,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 8823
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0810804647409284,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 8824
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9678536448448845,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 8825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0219474020223864,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 8826
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1011228292692599,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 8827
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2833081439052119,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 8828
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6708762492362625,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 8829
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.561798472275575,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 8830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0752199812856078,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 8831
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4945238606758322,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 8832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.055447861997748,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 8833
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4069378421672448,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 8834
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2538684589716218,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 8835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.386875877889635,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 8836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4620177776689,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 8837
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.4153070096404265,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 8838
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.07470513147323,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 8839
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0952479271505515,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 8840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6830434887489822,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 8841
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.716654265751843,
+      "learning_rate": 2e-05,
+      "loss": 0.0924,
+      "step": 8842
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.900590531282182,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 8843
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9326007650743144,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 8844
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2456117851907664,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 8845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.239801401058458,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 8846
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.608130542416009,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 8847
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6002269107134415,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 8848
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8799101119299625,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 8849
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2122131519597954,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 8850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7121889595267796,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 8851
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8335698871010586,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 8852
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7245223000742635,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 8853
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4780204544738207,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 8854
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8382412719431076,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 8855
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7026171114660364,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 8856
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.394215797940815,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 8857
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.718529010784218,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 8858
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6502349599526507,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 8859
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4327536467900324,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 8860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3597326137148678,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 8861
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33647456597458136,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 8862
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.811841713638068,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 8863
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8461994129603554,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 8864
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8905023858511234,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 8865
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7953872255845242,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 8866
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9468957057972338,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 8867
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4553410849324744,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 8868
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.392053706015511,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 8869
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1666691364372066,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 8870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5932339383124025,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 8871
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.143673139701272,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 8872
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7364247306772733,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 8873
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.226354559970597,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 8874
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.289742954874512,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 8875
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9870055423936743,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 8876
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3107237831475471,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 8877
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6929893473388543,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 8878
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7048800499307811,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 8879
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.315870045654502,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 8880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.864478554398835,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 8881
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9285596529769413,
+      "learning_rate": 2e-05,
+      "loss": 0.0785,
+      "step": 8882
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9380760648240863,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 8883
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2056193198663,
+      "learning_rate": 2e-05,
+      "loss": 0.0819,
+      "step": 8884
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8134871135270132,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 8885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4201351954645467,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 8886
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9903802406878214,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 8887
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9728214610195536,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 8888
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0406739163341108,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 8889
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8190884383409558,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 8890
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.61618879251196,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 8891
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6404831526539727,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 8892
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7848088269188381,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 8893
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4658446645647008,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 8894
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5166643649213043,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 8895
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9603381885453056,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 8896
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.999693727884492,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 8897
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.148040367807985,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 8898
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8845932478909869,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 8899
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1049365784793057,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 8900
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6521508622674229,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 8901
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3198836992004381,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 8902
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0804792202281326,
+      "learning_rate": 2e-05,
+      "loss": 0.1202,
+      "step": 8903
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4852781753347384,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 8904
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.180135982727788,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 8905
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.212688018496803,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 8906
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7494969341332434,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 8907
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6551712115576612,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 8908
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3588595235106609,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 8909
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.430924177524375,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 8910
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3051883908527722,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 8911
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6005586036456845,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 8912
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.058869838265303,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 8913
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.401829251700712,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 8914
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.785536823238317,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 8915
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.873110108551675,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 8916
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.174464647153217,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 8917
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5505110267411903,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 8918
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4211864857659311,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 8919
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37886068719065635,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 8920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4923622763643904,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 8921
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5886889863622753,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 8922
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6155788365594841,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 8923
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.152204583746583,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 8924
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17796040928212664,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 8925
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.200416571530692,
+      "learning_rate": 2e-05,
+      "loss": 0.0919,
+      "step": 8926
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.534767627083071,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 8927
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16891889420715359,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 8928
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.66488455788964,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 8929
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34390273819013834,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 8930
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.875435460704219,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 8931
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.536619636895034,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 8932
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.24484519036706,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 8933
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2157214756513366,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 8934
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.373896593465795,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 8935
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2758911926402063,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 8936
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4068644135058632,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 8937
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5463106301092346,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 8938
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.754105908612616,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 8939
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.667682718747133,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 8940
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.019753047909233,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 8941
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0265664905752923,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 8942
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.887386231666113,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 8943
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3172040582666338,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 8944
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6938100740822088,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 8945
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.007415811340938,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 8946
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0204835044906813,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 8947
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.460446841762712,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 8948
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9930051434647205,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 8949
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5280935819862497,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 8950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9332509069716806,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 8951
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8911882932773438,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 8952
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0643317757127282,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 8953
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8724282481685899,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 8954
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6763988825625253,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 8955
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2579033386410574,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 8956
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4193517072774513,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 8957
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.783949589778627,
+      "learning_rate": 2e-05,
+      "loss": 0.1312,
+      "step": 8958
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5157835537245,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 8959
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2100391517381714,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 8960
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6102821463600363,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 8961
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8548747748419125,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 8962
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9205740137736659,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 8963
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5702561949154448,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 8964
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7262995254254541,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 8965
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9693876444461075,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 8966
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.152244530799539,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 8967
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.104878457066282,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 8968
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1550522864260526,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 8969
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9199485046031224,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 8970
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2104713773372682,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 8971
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16213360570877983,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 8972
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.286611761855691,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 8973
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9672815822514109,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 8974
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9870314035057928,
+      "learning_rate": 2e-05,
+      "loss": 0.0844,
+      "step": 8975
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9769739428669765,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 8976
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.343479395185539,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 8977
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.095234841866114,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 8978
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4633857558110224,
+      "learning_rate": 2e-05,
+      "loss": 0.0995,
+      "step": 8979
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4500896304990262,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 8980
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6571251338565332,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 8981
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5370286638132687,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 8982
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2429440191986747,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 8983
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1719520034921993,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 8984
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34953141009105826,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 8985
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2325708534812976,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 8986
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.058504106281698,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 8987
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9866558172675486,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 8988
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1595807017025677,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 8989
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3034435824954348,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 8990
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5835804872313703,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 8991
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.722703711644697,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 8992
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0028291854349958,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 8993
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6772455251545615,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 8994
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8160402647083551,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 8995
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3927006607395933,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 8996
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6239052482996273,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 8997
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7160040605259416,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 8998
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.70382355059555,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 8999
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6053454858145428,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 9000
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7637467213511917,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 9001
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.108515924782124,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 9002
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0467004772868536,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 9003
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45072391284196006,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 9004
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2075265405237832,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 9005
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9727674105057953,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 9006
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9592710792844134,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 9007
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.539285480135772,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 9008
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5717252383167877,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 9009
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.647742608575994,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 9010
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.231311324957106,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 9011
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7070279200354291,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 9012
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.331309909099366,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 9013
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1586226271299354,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 9014
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0793883420185386,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 9015
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0520596580233568,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 9016
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.558753740972292,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 9017
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0821136973591208,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 9018
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.924186241219913,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 9019
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.651502608272076,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 9020
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8149286638643354,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 9021
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.79413541846638,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 9022
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5405044144569375,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 9023
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6760585895140516,
+      "learning_rate": 2e-05,
+      "loss": 0.1084,
+      "step": 9024
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6340630228082256,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 9025
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33032781422896673,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 9026
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9861264046685587,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 9027
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.896070866875977,
+      "learning_rate": 2e-05,
+      "loss": 0.0874,
+      "step": 9028
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8690944223858215,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 9029
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7494052830099212,
+      "learning_rate": 2e-05,
+      "loss": 0.0808,
+      "step": 9030
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7075311203853214,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 9031
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5095945905405552,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 9032
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3915104441732726,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 9033
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.17838422474696,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 9034
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.177801594744521,
+      "learning_rate": 2e-05,
+      "loss": 0.0969,
+      "step": 9035
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.480418960191728,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 9036
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.787558975012373,
+      "learning_rate": 2e-05,
+      "loss": 0.1346,
+      "step": 9037
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8962758634207895,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 9038
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5866795228322539,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 9039
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5221685143791275,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 9040
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9545251848805875,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 9041
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0825566815084664,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 9042
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8915646598674767,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 9043
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6152597630637777,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 9044
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.823287628212362,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 9045
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2257626106007198,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 9046
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7320447026892429,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 9047
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.308466719022493,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 9048
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3520624890233055,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 9049
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5109198062644507,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 9050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2073628979202877,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 9051
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6687891637656946,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 9052
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.34140202213551,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 9053
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.402193588898218,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 9054
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7597570008017973,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 9055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0829467756237716,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 9056
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0762152412109456,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 9057
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.276829233866477,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 9058
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7431589334877273,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 9059
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.268840612027203,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 9060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.36770678385695366,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 9061
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.767490255747517,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 9062
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6492105035912273,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 9063
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3181318072545254,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 9064
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7024712626493472,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 9065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5131122957599936,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 9066
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8107936324735456,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 9067
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5875282014759824,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 9068
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0488440673906076,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 9069
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.0715070205434944,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 9070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1497979313129165,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 9071
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.007684329850273,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 9072
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.018962137986749,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 9073
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2932837065207552,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 9074
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.145106823294185,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 9075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.794975383276162,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 9076
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.24299417849748,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 9077
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4852864729585131,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 9078
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.342717159306065,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 9079
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8064064984767234,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 9080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9386729104944154,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 9081
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1923659339661636,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 9082
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.162937904958621,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 9083
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5343364827280308,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 9084
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.047274040919191,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 9085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.361529559805374,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 9086
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6686342950354198,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 9087
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.567609298890559,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 9088
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9648592210771065,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 9089
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2724233696428084,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 9090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.892924945447122,
+      "learning_rate": 2e-05,
+      "loss": 0.0876,
+      "step": 9091
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6765228266790329,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 9092
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.252385485852601,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 9093
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6937563619516394,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 9094
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3582071132562956,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 9095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7988520387833511,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 9096
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0234945281840573,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 9097
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.554322284584277,
+      "learning_rate": 2e-05,
+      "loss": 0.0754,
+      "step": 9098
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9059388303796414,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 9099
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.324747930697531,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 9100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2285666644678863,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 9101
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2469306537165132,
+      "learning_rate": 2e-05,
+      "loss": 0.0773,
+      "step": 9102
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9763729064004723,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 9103
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6228206306876,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 9104
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26550981885368885,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 9105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3943863908344325,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 9106
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1059365512728496,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 9107
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.449327490497967,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 9108
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.760741933913261,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 9109
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3535015438180937,
+      "learning_rate": 2e-05,
+      "loss": 0.1124,
+      "step": 9110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4199897088765925,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 9111
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6798511834199591,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 9112
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2722783551441528,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 9113
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3646571166056547,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 9114
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.047918825775581,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 9115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38837450239347926,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 9116
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4283596188334755,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 9117
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.619229387279038,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 9118
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5659835915950175,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 9119
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.142317052504953,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 9120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5679762432867785,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 9121
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5043645779637351,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 9122
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8981968101192548,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 9123
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.990039580061764,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 9124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4369776239381427,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 9125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6203289387971704,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 9126
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.059317355835419,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 9127
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.9098472285539865,
+      "learning_rate": 2e-05,
+      "loss": 0.082,
+      "step": 9128
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.609148301196292,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 9129
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6679458795270456,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 9130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4686915609218454,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 9131
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9825439983931413,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 9132
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9759898223189951,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 9133
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6247452756334944,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 9134
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7206666932033966,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 9135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8572788684085884,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 9136
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3192470246686812,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 9137
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.422923780714862,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 9138
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.890790923875205,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 9139
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.603229380996244,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 9140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1994084012547555,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 9141
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.733796856720751,
+      "learning_rate": 2e-05,
+      "loss": 0.0703,
+      "step": 9142
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5613696698986,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 9143
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4634982117275814,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 9144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4715138019269025,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 9145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40868094190465654,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 9146
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0636045660884266,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 9147
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5058199303761305,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 9148
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.655524261270052,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 9149
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.27716827832177,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 9150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.268224534053364,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 9151
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8855776733912042,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 9152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3678949236935845,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 9153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5779957310676163,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 9154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.218223602805406,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 9155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6033302785457026,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 9156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2711920382001045,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 9157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9467847530701428,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 9158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6255006766811495,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 9159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.306788944446893,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 9160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.669831517301162,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 9161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0725722162528504,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 9162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.011762522335997,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 9163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7261916232614198,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 9164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.037315832136244,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 9165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.719188698638344,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 9166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4164102242872163,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 9167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3003282646061503,
+      "learning_rate": 2e-05,
+      "loss": 0.1008,
+      "step": 9168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4901376302026215,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 9169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.621143804294877,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 9170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.18922763855561528,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 9171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7710317322005896,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 9172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4574370036570727,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 9173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0576758174934797,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 9174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5769670075416752,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 9175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8602034303786121,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 9176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8878964367806066,
+      "learning_rate": 2e-05,
+      "loss": 0.093,
+      "step": 9177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0504712069277138,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 9178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.021007886222848,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 9179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2259819454836418,
+      "learning_rate": 2e-05,
+      "loss": 0.089,
+      "step": 9180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.821612367627668,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 9181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6576979443864308,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 9182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.241207816107282,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 9183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.968948838622723,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 9184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.223978785495982,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 9185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.062945577867936,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 9186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6650270344140332,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 9187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8508518298309049,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 9188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.088331095696603,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 9189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6706504752234579,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 9190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6079444857807315,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 9191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5991005919108514,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 9192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1824312449784373,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 9193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4381890830709328,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 9194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5053839883005975,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 9195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7519845931169183,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 9196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.437414664347802,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 9197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6579396876201167,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 9198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2540857774792344,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 9199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7139133172323072,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 9200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3696260340043414,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 9201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0184441979262049,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 9202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.019904205547359,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 9203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.166413563566792,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 9204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8188454361250277,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 9205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5544928040630845,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 9206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5275080107130361,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 9207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.250561008449379,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 9208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4539809628516747,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 9209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2591659680251817,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 9210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4133400942232859,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 9211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1867759596253142,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 9212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.291459974406883,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 9213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2414441089218373,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 9214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5657577275874672,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 9215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4013073419957744,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 9216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.022116879175635,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 9217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5948268020251106,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 9218
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3511539416460082,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 9219
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.915153460451315,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 9220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2429990994251363,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 9221
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8984900749414091,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 9222
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0769854144970905,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 9223
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9572179806769618,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 9224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1609033304887484,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 9225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.960477889641015,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 9226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.092888428390867,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 9227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26564934562862424,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 9228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.29381110844204483,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 9229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.144238728235112,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 9230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.24778066452290123,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 9231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.769771841010589,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 9232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6886260756305527,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 9233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.12077880050751,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 9234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.510370407273043,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 9235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8467192307643185,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 9236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.053662345026049,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 9237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1523897840712416,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 9238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.234787182182728,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 9239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3235924172175704,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 9240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.15548161623448,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 9241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6861055051796634,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 9242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.610690515782673,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 9243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7539871095731048,
+      "learning_rate": 2e-05,
+      "loss": 0.0813,
+      "step": 9244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6967889359990704,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 9245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5197762793465717,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 9246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.3232489057298,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 9247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6557671950758512,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 9248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.403942843444815,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 9249
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.694016330734352,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 9250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3544842217341171,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 9251
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.157356531140278,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 9252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.48382606394935,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 9253
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.772006203624032,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 9254
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6028311566096134,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 9255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4386450312127483,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 9256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9106513319862037,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 9257
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8174751213976803,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 9258
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2112301380587915,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 9259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2233012877501874,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 9260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.966401626949008,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 9261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1666668426422713,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 9262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.023608614394322,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 9263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1804983623837426,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 9264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8319999080002257,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 9265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1454906413554775,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 9266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2548113349433394,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 9267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3975042287117678,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 9268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0993905264672668,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 9269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6264270464822204,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 9270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4840002197105813,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 9271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.121033582202957,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 9272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2463302627016208,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 9273
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4875164319581344,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 9274
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6972590930316986,
+      "learning_rate": 2e-05,
+      "loss": 0.0936,
+      "step": 9275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1717699155641528,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 9276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9964047542392358,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 9277
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2620358154267666,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 9278
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.056592515054875,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 9279
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.48819789938007896,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 9280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.572074319325129,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 9281
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2603657683049128,
+      "learning_rate": 2e-05,
+      "loss": 0.1003,
+      "step": 9282
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7166122319871586,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 9283
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.134483828360265,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 9284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.356114415777521,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 9285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.449529234238523,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 9286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3171017221777976,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 9287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3086174806179682,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 9288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.882991101669866,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 9289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9082424117996633,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 9290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7355550659177418,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 9291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.734402197575188,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 9292
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5089018559660863,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 9293
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7672856715934555,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 9294
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5068387907195486,
+      "learning_rate": 2e-05,
+      "loss": 0.0983,
+      "step": 9295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5783117053311093,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 9296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.174642323449358,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 9297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7574585550131182,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 9298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8639167414369867,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 9299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4502503869487402,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 9300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5228300697945102,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 9301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.511397321907445,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 9302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38000286333360556,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 9303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.489144790735486,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 9304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.404618044241161,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 9305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2240559910982696,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 9306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.913404215203779,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 9307
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2615444926596187,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 9308
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.036264472809469,
+      "learning_rate": 2e-05,
+      "loss": 0.0987,
+      "step": 9309
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9890206014485394,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 9310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7469653812333674,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 9311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7033484633642848,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 9312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5507160445770652,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 9313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3272291966296183,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 9314
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9081748055585579,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 9315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15563224026869835,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 9316
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3509700193517715,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 9317
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.66710786348598,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 9318
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0840338618349228,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 9319
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1060893147468067,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 9320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5104561197674833,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 9321
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.933316975831298,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 9322
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8163894578584818,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 9323
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.457177593344934,
+      "learning_rate": 2e-05,
+      "loss": 0.0928,
+      "step": 9324
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.986486376658852,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 9325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2501706960477805,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 9326
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09237287782475546,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 9327
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4421931240844175,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 9328
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6026385158477372,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 9329
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.570751674433052,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 9330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1538460518584674,
+      "learning_rate": 2e-05,
+      "loss": 0.1207,
+      "step": 9331
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8068554503346776,
+      "learning_rate": 2e-05,
+      "loss": 0.0953,
+      "step": 9332
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7017186829985895,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 9333
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4807619897717395,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 9334
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.498265713910557,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 9335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9514689840561834,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 9336
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5018494440851433,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 9337
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.812078624282698,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 9338
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2936580213352316,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 9339
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8878521838343971,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 9340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4406614098171433,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 9341
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5244910340961186,
+      "learning_rate": 2e-05,
+      "loss": 0.0741,
+      "step": 9342
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6909696670574785,
+      "learning_rate": 2e-05,
+      "loss": 0.1183,
+      "step": 9343
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5074671216324202,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 9344
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.546479453945978,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 9345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44129273972182576,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 9346
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6879364438272753,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 9347
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.799222457689779,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 9348
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.076570412837875,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 9349
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9991023087527636,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 9350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2565690048341605,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 9351
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2516969606176578,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 9352
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1122464670385983,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 9353
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0359130971574622,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 9354
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.864636868965477,
+      "learning_rate": 2e-05,
+      "loss": 0.0941,
+      "step": 9355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7866649304329327,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 9356
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3724689176242673,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 9357
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.778974095413388,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 9358
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8137850898442536,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 9359
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.733734331280091,
+      "learning_rate": 2e-05,
+      "loss": 0.1035,
+      "step": 9360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7737452706293415,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 9361
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0194945341273796,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 9362
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9094323032602236,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 9363
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2270803633925416,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 9364
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1970530110109301,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 9365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.629483435924445,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 9366
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0242687664142385,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 9367
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9084732474323068,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 9368
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3033558336902773,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 9369
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.267400602773433,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 9370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.199697011282155,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 9371
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.364770858467583,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 9372
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6330961985793264,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 9373
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6064997794242872,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 9374
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1667040353421596,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 9375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0070106806108154,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 9376
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5103809359422886,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 9377
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6093691556601378,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 9378
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.456837331840481,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 9379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5059045767524772,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 9380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.439528733730239,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 9381
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1188964748195938,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 9382
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.71835794312213,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 9383
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4187460206097826,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 9384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.30280972881942,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 9385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3681518806065545,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 9386
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6332388873268915,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 9387
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.180652047246813,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 9388
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6469417493347798,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 9389
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8510957191014504,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 9390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.483109828455023,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 9391
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.672945488536248,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 9392
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4993291785365086,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 9393
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7884115423343697,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 9394
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2945300123523051,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 9395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.503108062965949,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 9396
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5825387930743133,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 9397
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4902922892784858,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 9398
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.880688026831436,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 9399
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1313431258005973,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 9400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3121631280825574,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 9401
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1018420466679237,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 9402
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7159334166086908,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 9403
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.320752736852986,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 9404
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4014584514134063,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 9405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4253474163386215,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 9406
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.108621622510753,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 9407
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9948455170174902,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 9408
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.4497985194167295,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 9409
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.796515500284033,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 9410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5288712066236552,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 9411
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9057876151199804,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 9412
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.661335056599954,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 9413
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8224103351502796,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 9414
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1777313890308592,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 9415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0174105520166647,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 9416
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.091170894119544,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 9417
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4000020146355439,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 9418
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0823277366778579,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 9419
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.786144225270229,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 9420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0273919660240654,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 9421
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8577070047441453,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 9422
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2553765378838007,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 9423
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.3631834506059475,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 9424
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9801457393550328,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 9425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7917985220742844,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 9426
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.018581137674903,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 9427
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1810796241405046,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 9428
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5311860051698416,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 9429
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.721717430851869,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 9430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3267786719295382,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 9431
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.749242070215555,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 9432
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32854420132809303,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 9433
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8009022570684285,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 9434
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.093929340026585,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 9435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8302665945276773,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 9436
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.928466348712008,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 9437
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6446530983335155,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 9438
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.944687250347401,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 9439
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.320957903255482,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 9440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9309272193917271,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 9441
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.728742629638448,
+      "learning_rate": 2e-05,
+      "loss": 0.0792,
+      "step": 9442
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0817923632448436,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 9443
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.89546705906731,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 9444
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.598726227374726,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 9445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5466604372973343,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 9446
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.590814336793029,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 9447
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.355384887435346,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 9448
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2812075724206167,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 9449
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2732083444364655,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 9450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7204711521354485,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 9451
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8272341652540407,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 9452
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.661579128278738,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 9453
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8837696349317357,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 9454
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6069881589487354,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 9455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8791863228346719,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 9456
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1881423016231052,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 9457
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.768660779664017,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 9458
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0794417591577483,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 9459
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9993008345179815,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 9460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7836051141987097,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 9461
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.152327798829277,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 9462
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.594879647620179,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 9463
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9163431615673407,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 9464
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.221262543907878,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 9465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6658177677564028,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 9466
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4367659738760064,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 9467
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8752927361940968,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 9468
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.485059056184748,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 9469
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.831650881873557,
+      "learning_rate": 2e-05,
+      "loss": 0.1118,
+      "step": 9470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.57268313579841,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 9471
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.191467682885068,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 9472
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.302199129210132,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 9473
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.550089722774499,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 9474
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4819833863391498,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 9475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.869858495284848,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 9476
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.597132011928764,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 9477
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.037463699980733,
+      "learning_rate": 2e-05,
+      "loss": 0.1207,
+      "step": 9478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.42376167022478906,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 9479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9365986191843296,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 9480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9762468946649787,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 9481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.374455019430785,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 9482
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2095701484879062,
+      "learning_rate": 2e-05,
+      "loss": 0.1127,
+      "step": 9483
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.220046458779549,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 9484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2534695225752723,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 9485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.088889786650192,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 9486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1925390467275996,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 9487
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4188841500066651,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 9488
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.359172584996087,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 9489
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2195967642606385,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 9490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1106632972687547,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 9491
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.627735110292573,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 9492
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.065897990359702,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 9493
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1975111624962578,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 9494
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2816993111635273,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 9495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4883871250593983,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 9496
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.85823383093422,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 9497
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.196312944512219,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 9498
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4906618237615112,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 9499
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.972660469239176,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 9500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5938568771943701,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 9501
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3725942292175564,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 9502
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7904445191080423,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 9503
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8957448664312372,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 9504
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8163466363341888,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 9505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8535339547508118,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 9506
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0773525092954195,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 9507
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2540170971132312,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 9508
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.22929170657933995,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 9509
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.490775829615832,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 9510
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.34184139362515,
+      "learning_rate": 2e-05,
+      "loss": 0.1294,
+      "step": 9511
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2749259534539512,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 9512
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7915522147002685,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 9513
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4969403709738417,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 9514
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6802385003096951,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 9515
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13562527431293292,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 9516
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9190139663784833,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 9517
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1723088990545008,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 9518
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.377431280564741,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 9519
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3366222299925077,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 9520
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8669345246157308,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 9521
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3095508775968612,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 9522
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0651720891185874,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 9523
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.293700220966109,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 9524
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1063687600220693,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 9525
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7097514393070645,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 9526
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7331920998932795,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 9527
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4966064117058055,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 9528
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.997392564063814,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 9529
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.084331614920889,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 9530
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1374937931318554,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 9531
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.350306859286038,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 9532
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.49243056638086996,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 9533
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8125928657681767,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 9534
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.653257906356192,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 9535
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.679092514760941,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 9536
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.12582239431739203,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 9537
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.735548977195349,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 9538
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2538455936580053,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 9539
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9713127018291533,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 9540
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2532930389903736,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 9541
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.419378415384555,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 9542
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8467645996514834,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 9543
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.037867844753570384,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 9544
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8146653211178703,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 9545
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.669232383377873,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 9546
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1276457528416146,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 9547
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.876053161119157,
+      "learning_rate": 2e-05,
+      "loss": 0.1132,
+      "step": 9548
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.330880750569001,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 9549
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.552053417627331,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 9550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5430181430456553,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 9551
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5847872617173995,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 9552
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0133065978609908,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 9553
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0951316281464132,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 9554
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7093293457629448,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 9555
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.082615993883802,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 9556
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.47532140436302733,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 9557
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6496247992041748,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 9558
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.869023890497996,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 9559
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1269960565499075,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 9560
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.033945527918888,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 9561
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.64201494247259,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 9562
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6756712953571746,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 9563
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1184951854054184,
+      "learning_rate": 2e-05,
+      "loss": 0.1239,
+      "step": 9564
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7775795581037108,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 9565
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.369952129381761,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 9566
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.775437399272554,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 9567
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3806497935481763,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 9568
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3338048319820728,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 9569
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5879848569457619,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 9570
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.024130095141942,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 9571
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9863984035721891,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 9572
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2828550287554559,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 9573
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.027585045827251,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 9574
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.822274241714787,
+      "learning_rate": 2e-05,
+      "loss": 0.1416,
+      "step": 9575
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.428200024166222,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 9576
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2455378999705555,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 9577
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.972364040534202,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 9578
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2851998544040565,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 9579
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.216021785358984,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 9580
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.438520815415285,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 9581
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4905438383067384,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 9582
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3205679471446434,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 9583
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37594291239177424,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 9584
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5784745357983105,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 9585
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0133892725363105,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 9586
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4059015478118988,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 9587
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4019552410577918,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 9588
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5256592975676937,
+      "learning_rate": 2e-05,
+      "loss": 0.1131,
+      "step": 9589
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3614262418265262,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 9590
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.101374622871542,
+      "learning_rate": 2e-05,
+      "loss": 0.0804,
+      "step": 9591
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2878365437911228,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 9592
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1674081228728137,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 9593
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.241363636051258,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 9594
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6509089423630382,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 9595
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4495693718322944,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 9596
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.768337234488197,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 9597
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4171942495166416,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 9598
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9282296177302816,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 9599
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9301287189114544,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 9600
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5039647474925866,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 9601
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6563670278985931,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 9602
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3674177139463888,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 9603
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5475280486145353,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 9604
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6890796755782895,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 9605
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.384594460483123,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 9606
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7878856759017568,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 9607
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4263123241718816,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 9608
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3102803307443405,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 9609
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2014416399469028,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 9610
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2927193024753192,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 9611
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.13459947324983,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 9612
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5562457373809364,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 9613
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.40512273353382877,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 9614
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9513266748813737,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 9615
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.409340812308898,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 9616
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1257298009738927,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 9617
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.109904488571884,
+      "learning_rate": 2e-05,
+      "loss": 0.0721,
+      "step": 9618
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0956332379732031,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 9619
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4693174106911486,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 9620
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6835555848931532,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 9621
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6618847489833435,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 9622
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.297417920361805,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 9623
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.238614607012373,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 9624
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5360866172358267,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 9625
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0835046754955322,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 9626
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4602508897170443,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 9627
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0650646563514927,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 9628
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.017817199520083,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 9629
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.34496089762083887,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 9630
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5103556729791094,
+      "learning_rate": 2e-05,
+      "loss": 0.0732,
+      "step": 9631
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.271587361795596,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 9632
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1296958933580452,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 9633
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7487435504998996,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 9634
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7222582571707087,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 9635
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2056655844599815,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 9636
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.154219740997597,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 9637
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.689630772659568,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 9638
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2119158330592426,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 9639
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0407282162360572,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 9640
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7757041931298668,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 9641
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3515836898508025,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 9642
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2950117441526676,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 9643
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.960687693406248,
+      "learning_rate": 2e-05,
+      "loss": 0.1051,
+      "step": 9644
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4811459707554634,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 9645
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5330989340593024,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 9646
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9863400679297734,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 9647
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.231424692996236,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 9648
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.629054072448289,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 9649
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.813492567588052,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 9650
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5463719176055948,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 9651
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7686584880326781,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 9652
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5494340694295657,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 9653
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.688000664994124,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 9654
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7866145393576832,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 9655
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0763529713425084,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 9656
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.627440893503332,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 9657
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.470746259350499,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 9658
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0079624698900775,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 9659
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2244017210493732,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 9660
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8855064608350599,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 9661
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2691183584588144,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 9662
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2973393523010812,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 9663
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0217199844001983,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 9664
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.097048310850752,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 9665
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.524386144818633,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 9666
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6379673692884573,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 9667
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.318415526147964,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 9668
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14434230542621448,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 9669
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0034487383132051,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 9670
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6519421237682483,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 9671
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7708227568665849,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 9672
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1726506229173195,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 9673
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5268754928785668,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 9674
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8387630066453762,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 9675
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.263507861853755,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 9676
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0212482011608777,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 9677
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.560728010624882,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 9678
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.345894901736756,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 9679
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8720553440633433,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 9680
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6943566256202829,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 9681
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23126467967479677,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 9682
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0124681779755664,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 9683
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.396931625950649,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 9684
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3098993421422884,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 9685
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6645622841023784,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 9686
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.108711760060588,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 9687
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6222504455285874,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 9688
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3838314031074224,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 9689
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.298169971728447,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 9690
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6467077094216519,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 9691
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3424236436467902,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 9692
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.631212463531661,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 9693
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5242912674389272,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 9694
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1238471057762083,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 9695
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.254520032194666,
+      "learning_rate": 2e-05,
+      "loss": 0.0897,
+      "step": 9696
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.032948755331323,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 9697
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3933367634612913,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 9698
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8219358891775292,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 9699
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4568074645136604,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 9700
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0891798725120492,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 9701
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5999315255328663,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 9702
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3344135180593628,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 9703
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.523620882250407,
+      "learning_rate": 2e-05,
+      "loss": 0.0952,
+      "step": 9704
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9804499096698467,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 9705
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.859363203251706,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 9706
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7317231774791855,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 9707
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4715390446083263,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 9708
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7232375822402242,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 9709
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5061084620100179,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 9710
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.686246180258725,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 9711
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6542220115534766,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 9712
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1407191616540728,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 9713
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.895179253861029,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 9714
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2572936889497326,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 9715
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3290179341632792,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 9716
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2452329334988068,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 9717
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6078168805118974,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 9718
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6241279239964231,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 9719
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.467393456194093,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 9720
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39259010030976194,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 9721
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.737950942985621,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 9722
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9657411610383752,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 9723
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.663010242148473,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 9724
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.752739397096591,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 9725
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9066685006237649,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 9726
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.072371930443669,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 9727
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.496300513264375,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 9728
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9250645499002144,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 9729
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.263093057472373,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 9730
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3079032575340923,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 9731
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08382836811549475,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 9732
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.384316491925305,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 9733
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8752667837791918,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 9734
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6545211117580003,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 9735
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1266131491525042,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 9736
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.081996748311718,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 9737
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1902556319509139,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 9738
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9995007460774112,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 9739
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2529919123396493,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 9740
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.823961717684874,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 9741
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4716437014660935,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 9742
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5649083291857656,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 9743
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.483039908089308,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 9744
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0208176138050333,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 9745
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2136036225358513,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 9746
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6637578437650307,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 9747
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0339162633329406,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 9748
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7363522917328325,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 9749
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.5196971022129295,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 9750
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3649027453000095,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 9751
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0452006030669665,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 9752
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9944864507205762,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 9753
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2852300391681327,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 9754
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4736363718686305,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 9755
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3602644904683503,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 9756
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0395174067138573,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 9757
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5569375032825832,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 9758
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.236188459407831,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 9759
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7988083100030043,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 9760
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5895185521484928,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 9761
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.200497867698317,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 9762
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8554414370283091,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 9763
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0268963321784852,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 9764
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7517560824549989,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 9765
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.945875768042228,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 9766
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.688469643520743,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 9767
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.357112380652313,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 9768
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9938558137850768,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 9769
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.288905115543282,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 9770
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.123709602964844,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 9771
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1779482316784053,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 9772
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.061817903920211,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 9773
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5202355015418878,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 9774
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.843418050711489,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 9775
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3455003383526896,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 9776
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9246043851496228,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 9777
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6950062418009741,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 9778
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8089750929465276,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 9779
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0272631037645312,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 9780
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8832952743105575,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 9781
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7514237334438616,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 9782
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8974008685739174,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 9783
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.292355300400044,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 9784
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1344416059207836,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 9785
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.160844989503354,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 9786
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.568380505335163,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 9787
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.833644440383447,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 9788
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1867343040185325,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 9789
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28020152056517655,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 9790
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.579916684040036,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 9791
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8062914445855841,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 9792
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8446649124078345,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 9793
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1044824445105097,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 9794
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.75425927347534,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 9795
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7514334331885625,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 9796
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1855289843715915,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 9797
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33732559943301743,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 9798
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.940067866111652,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 9799
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4569051519690813,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 9800
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0552419264863446,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 9801
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6969967491985353,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 9802
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1987844668655192,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 9803
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7241212320341208,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 9804
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.072547713628942,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 9805
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7658050481106343,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 9806
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.771535517281338,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 9807
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8758223955312017,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 9808
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4144647257004213,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 9809
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.756822933217925,
+      "learning_rate": 2e-05,
+      "loss": 0.1196,
+      "step": 9810
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0151378930669013,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 9811
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0686244776697045,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 9812
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.004317337890354,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 9813
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.754023989061137,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 9814
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9472292606857196,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 9815
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7512549260442531,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 9816
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1134249945748507,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 9817
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5876837075938288,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 9818
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4060383302208437,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 9819
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.685950786386841,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 9820
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4417862985443235,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 9821
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0099507794803475,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 9822
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.614419826309078,
+      "learning_rate": 2e-05,
+      "loss": 0.0785,
+      "step": 9823
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1766103565063086,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 9824
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.316384232591803,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 9825
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4864327071862904,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 9826
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34910246259153493,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 9827
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8731718051524378,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 9828
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5066501703175073,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 9829
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7373268878358137,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 9830
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.35223996268389274,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 9831
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6227725975850407,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 9832
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.340387350862038,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 9833
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.256623780805365,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 9834
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0657650203948357,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 9835
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5001549778606011,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 9836
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.694711390310538,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 9837
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.351207840973761,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 9838
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.179494791025717,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 9839
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7713723543857087,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 9840
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.331310780732561,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 9841
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4159877590199916,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 9842
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9543114682403186,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 9843
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9126737419991864,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 9844
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5904057889424281,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 9845
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4810753496009834,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 9846
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4209080017397182,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 9847
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.165358503053669,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 9848
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3987604590093623,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 9849
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1937169694723333,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 9850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.85748034469136,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 9851
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8727998540799207,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 9852
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4578218743941793,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 9853
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2970893406976503,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 9854
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6121867533573166,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 9855
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7059559311134588,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 9856
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4803505137832746,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 9857
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.13438988277095,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 9858
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4987943603180858,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 9859
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0206577902652296,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 9860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0748124646846615,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 9861
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.009811178419602,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 9862
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40948649628001305,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 9863
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4361476665646937,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 9864
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.062028342536301,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 9865
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.840684013467122,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 9866
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5548260497652742,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 9867
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.086884120312814,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 9868
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5643863735174162,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 9869
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0109012202219383,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 9870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4132783517549488,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 9871
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4700873758986195,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 9872
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4768022615049659,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 9873
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.308477986389637,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 9874
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0115790546843275,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 9875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1774562425035418,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 9876
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.753794652963822,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 9877
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2057366008850945,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 9878
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6042339367125305,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 9879
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.756062997043753,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 9880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.233610638279308,
+      "learning_rate": 2e-05,
+      "loss": 0.1606,
+      "step": 9881
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.518723196463226,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 9882
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5335058826945134,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 9883
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7133802281850485,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 9884
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8149297403697108,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 9885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.110390212058206,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 9886
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1352561638922445,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 9887
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2084880312660267,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 9888
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.127269182752814,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 9889
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3268587906257765,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 9890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.706957801276121,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 9891
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.850768099955922,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 9892
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8836548424064811,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 9893
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2444654486838251,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 9894
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3259281555233757,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 9895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.067800489162976,
+      "learning_rate": 2e-05,
+      "loss": 0.0706,
+      "step": 9896
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.018723235396685,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 9897
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8582032972430272,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 9898
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7145341667582257,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 9899
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.4443107240481785,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 9900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.347270203901914,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 9901
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.85534210661162,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 9902
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.642645757830694,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 9903
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2540530827652616,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 9904
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6457589611520516,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 9905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9043180986389695,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 9906
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0683150203779772,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 9907
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6212246403083648,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 9908
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5466665262326362,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 9909
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5479270896597119,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 9910
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8693975034919705,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 9911
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.34991813104144,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 9912
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6105940478177034,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 9913
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8964266623409918,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 9914
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8636076297269256,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 9915
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0462040815223823,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 9916
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45503714381031296,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 9917
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0222641987160586,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 9918
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2793730501980964,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 9919
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.19298811633887966,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 9920
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.499671914507138,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 9921
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4048254577048094,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 9922
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.359986005708525,
+      "learning_rate": 2e-05,
+      "loss": 0.1,
+      "step": 9923
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5949592490655708,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 9924
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3702469311359153,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 9925
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6389756404237201,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 9926
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5106336430876867,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 9927
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8348080021668953,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 9928
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3368949000119762,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 9929
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.56361273244659,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 9930
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0311277699160346,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 9931
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.828308242678507,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 9932
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.227811040686843,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 9933
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.819974165723708,
+      "learning_rate": 2e-05,
+      "loss": 0.1137,
+      "step": 9934
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.953488540518949,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 9935
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.754186131749547,
+      "learning_rate": 2e-05,
+      "loss": 0.0881,
+      "step": 9936
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1827747797526103,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 9937
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1587535635622768,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 9938
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4943050518441403,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 9939
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2983909846488206,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 9940
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.834950193815656,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 9941
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.399801754711014,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 9942
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7045987031443404,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 9943
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8446662695075131,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 9944
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5695383129214178,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 9945
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9217135401137339,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 9946
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8055453552870886,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 9947
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.944006914534654,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 9948
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1107870437314284,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 9949
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.908296018621558,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 9950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0816204531782656,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 9951
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2070969252018104,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 9952
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1116509530149792,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 9953
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.573937639508078,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 9954
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8513562377421175,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 9955
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7013722083077478,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 9956
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8708032051951267,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 9957
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.123167611971158,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 9958
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0005034133265738,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 9959
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.887397810593942,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 9960
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6097186564216839,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 9961
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0494983314802584,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 9962
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6609429181104569,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 9963
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8464386297006279,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 9964
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.294271959065102,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 9965
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0297038135651175,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 9966
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.082477800816507,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 9967
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4525471020150733,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 9968
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.72138909051452,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 9969
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.744823438736792,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 9970
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5496547016400504,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 9971
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6541005708315601,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 9972
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7252584259900774,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 9973
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.042484236899077,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 9974
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4513579258310751,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 9975
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.678852955672725,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 9976
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37850996314381846,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 9977
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9486410320713121,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 9978
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1864857261043038,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 9979
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6918891763533583,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 9980
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8493619952787228,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 9981
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5493041990831133,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 9982
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7059009388556077,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 9983
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1981572626656884,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 9984
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.975533515272175,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 9985
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7749720584658089,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 9986
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8126540038069165,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 9987
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4243231137429248,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 9988
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.325654017844824,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 9989
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.098109929698539,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 9990
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6118805279439126,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 9991
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1352396252461971,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 9992
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.262182438036414,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 9993
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3436129300491597,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 9994
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9259346984020915,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 9995
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.090903781612727,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 9996
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0759868705118645,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 9997
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.418391981619446,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 9998
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1767163841250383,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 9999
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0050360115203931,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 10000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4083776273289006,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 10001
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.460840907547565,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 10002
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2073223664157613,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 10003
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15847451867602025,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 10004
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.857504572887269,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 10005
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2751686639114899,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 10006
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0464496898678943,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 10007
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.341821003869907,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 10008
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7902126030427272,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 10009
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0714795547938567,
+      "learning_rate": 2e-05,
+      "loss": 0.0904,
+      "step": 10010
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.108551149309208,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 10011
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9867072029693253,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 10012
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.889968389241215,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 10013
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0319183423838583,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 10014
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0059976056541995,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 10015
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.369597665300578,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 10016
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.116630095030083,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 10017
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9751672456645153,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 10018
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.688067886324082,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 10019
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4666780507483976,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 10020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9492124507702886,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 10021
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8465130033265018,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 10022
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.30818748289381703,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 10023
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4791641324876739,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 10024
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2977654950240716,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 10025
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.270499214810977,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 10026
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8257224503949647,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 10027
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5919294287469974,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 10028
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6497064147515113,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 10029
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.47943357276426934,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 10030
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.040688520438916,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 10031
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1813439367172214,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 10032
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8239728370930014,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 10033
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9810173657909107,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 10034
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2914170923061246,
+      "learning_rate": 2e-05,
+      "loss": 0.1052,
+      "step": 10035
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.241181895747295,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 10036
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.178592704399259,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 10037
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3081570935857505,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 10038
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3530558456122646,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 10039
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9342109080586396,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 10040
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8236956631419282,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 10041
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.873204043489585,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 10042
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16450679315618147,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 10043
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5736057097150155,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 10044
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0061901126218045,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 10045
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0575140482733703,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 10046
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9821430515933154,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 10047
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1668839536138447,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 10048
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7311062402508212,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 10049
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.507433425917126,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 10050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.906100689733938,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 10051
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5166359901819184,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 10052
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8396809175845632,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 10053
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7198324089684887,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 10054
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.451322050927475,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 10055
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.895037863942769,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 10056
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.706564827124325,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 10057
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5049075273765462,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 10058
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27127772637803227,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 10059
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.630735973170583,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 10060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.964287247221378,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 10061
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.820977048890584,
+      "learning_rate": 2e-05,
+      "loss": 0.0992,
+      "step": 10062
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37724664358740345,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 10063
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4271551332255341,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 10064
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.256757637436294,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 10065
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.091340799771715,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 10066
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.331419594577743,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 10067
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.696451210908129,
+      "learning_rate": 2e-05,
+      "loss": 0.1631,
+      "step": 10068
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.62012153684037,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 10069
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7761196867456417,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 10070
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8483058779758395,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 10071
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0727451829377628,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 10072
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5694949678813273,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 10073
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26457303370055363,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 10074
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5250344753299139,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 10075
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9568467681059487,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 10076
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8336633426178508,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 10077
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1352348570637822,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 10078
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4495100772187617,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 10079
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.860484269999046,
+      "learning_rate": 2e-05,
+      "loss": 0.1478,
+      "step": 10080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.805650211118132,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 10081
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.806219336123401,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 10082
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.108515715591327,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 10083
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9625573190738275,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 10084
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.070945763702919,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 10085
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.22140220887420595,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 10086
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.356539000740645,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 10087
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3984376869949201,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 10088
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0442442791512434,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 10089
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.344890367918068,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 10090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.38906959348016,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 10091
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.342285120582998,
+      "learning_rate": 2e-05,
+      "loss": 0.0992,
+      "step": 10092
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8037385633797358,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 10093
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.305998557901039,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 10094
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7847417771801495,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 10095
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3962948451375436,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 10096
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.268836225232013,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 10097
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.899703871086503,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 10098
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6792185250670888,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 10099
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.24772335197554,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 10100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2293318567429181,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 10101
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7341917093301955,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 10102
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2359629689906764,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 10103
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.785771236873908,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 10104
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.232902179352289,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 10105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9490321725591325,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 10106
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.04002383029693,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 10107
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6301168962052373,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 10108
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.20749787655617,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 10109
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4057552738976093,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 10110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4871047769149266,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 10111
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4873757144415682,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 10112
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2070722356954822,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 10113
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4777654359695716,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 10114
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4827153615216362,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 10115
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0912042691675838,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 10116
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3345974899733721,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 10117
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36439666964969836,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 10118
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7565387363790627,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 10119
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5517312372610016,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 10120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.48483101083368163,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 10121
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3573930257023132,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 10122
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7356581881258477,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 10123
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2474279167280793,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 10124
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1195656067713475,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 10125
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8449958504343568,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 10126
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.094311749869961,
+      "learning_rate": 2e-05,
+      "loss": 0.1113,
+      "step": 10127
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9318359937889708,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 10128
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.511205285649552,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 10129
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6184222027398422,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 10130
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.732467683703976,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 10131
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.757546503558374,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 10132
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5943430376905328,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 10133
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.197406654701709,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 10134
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2148671040636043,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 10135
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9171650556449065,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 10136
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7526230639204875,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 10137
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3853896958252239,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 10138
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.217565966850967,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 10139
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7272722612726279,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 10140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7368868260428838,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 10141
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4250893782309428,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 10142
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6635881921054254,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 10143
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4897812220590005,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 10144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1762618373393554,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 10145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5174255561618204,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 10146
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1390712031279204,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 10147
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.553190279195862,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 10148
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.32538783471787563,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 10149
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9220407305955718,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 10150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9763789195122589,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 10151
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.929175452260676,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 10152
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1295537961213733,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 10153
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.899524263636772,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 10154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.239495870031668,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 10155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0654325733209746,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 10156
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.176330402155655,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 10157
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4709470605864268,
+      "learning_rate": 2e-05,
+      "loss": 0.0978,
+      "step": 10158
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9883690835254773,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 10159
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4708519208529505,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 10160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.19935154878511,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 10161
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8097631904425941,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 10162
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2170266994370738,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 10163
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8320226534547099,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 10164
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6208180388421183,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 10165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5067747033187409,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 10166
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.752036068248936,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 10167
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6054401627197472,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 10168
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1645696162653996,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 10169
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3944973695012183,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 10170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3200464375592729,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 10171
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7995414626777801,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 10172
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0432796294984645,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 10173
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.676282369466968,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 10174
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.47116613765026105,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 10175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.110851308106301,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 10176
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.231821113483347,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 10177
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1994613253014288,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 10178
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7028998756287558,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 10179
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5315255851017784,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 10180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.196821818153761,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 10181
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8767647657758154,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 10182
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4216031814133383,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 10183
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5354037097537765,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 10184
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3006169670769803,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 10185
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6834850585972367,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 10186
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7725976564968747,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 10187
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1413401837956982,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 10188
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.637922705026457,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 10189
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7274007704151517,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 10190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.688668969505789,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 10191
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4235414862956826,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 10192
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3719514071082939,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 10193
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0702036781400388,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 10194
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.52008031732237,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 10195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.012973434529203,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 10196
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.782921461817108,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 10197
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9228741714247657,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 10198
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5603692104410283,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 10199
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.577109353013064,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 10200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2520046372312523,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 10201
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.052164524120744,
+      "learning_rate": 2e-05,
+      "loss": 0.0706,
+      "step": 10202
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0868818718735205,
+      "learning_rate": 2e-05,
+      "loss": 0.1285,
+      "step": 10203
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3813397968812535,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 10204
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.364640478509046,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 10205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.125473291795899,
+      "learning_rate": 2e-05,
+      "loss": 0.0862,
+      "step": 10206
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5747246715419547,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 10207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.809900746104636,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 10208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5016472890887536,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 10209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4484784891034406,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 10210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7600946682436884,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 10211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5366566545762397,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 10212
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.813800688724836,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 10213
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6091149129238018,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 10214
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2067981157978998,
+      "learning_rate": 2e-05,
+      "loss": 0.0867,
+      "step": 10215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.281874562392785,
+      "learning_rate": 2e-05,
+      "loss": 0.0809,
+      "step": 10216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.363349446975734,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 10217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.373688201424098,
+      "learning_rate": 2e-05,
+      "loss": 0.0978,
+      "step": 10218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7858080080414135,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 10219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2083308559699257,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 10220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5579984200140758,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 10221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3117099133961094,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 10222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4461716031757466,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 10223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5852742338535952,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 10224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0548101848313112,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 10225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5110554977464916,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 10226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.177110679835022,
+      "learning_rate": 2e-05,
+      "loss": 0.0874,
+      "step": 10227
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7526683625998142,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 10228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5649118902067611,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 10229
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9983833477396767,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 10230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.838456040901435,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 10231
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.367261425916109,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 10232
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.607463387719632,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 10233
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0115688103584692,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 10234
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5488631447757115,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 10235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1562944867622964,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 10236
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1302938414307933,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 10237
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2448836522717621,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 10238
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7189211434320407,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 10239
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.515344639668893,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 10240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8499436524547688,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 10241
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.05848164762137242,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 10242
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3856489756380685,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 10243
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.40790629862979,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 10244
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.366205697992548,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 10245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7713492255025534,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 10246
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5933421027855625,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 10247
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.389118891986669,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 10248
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3114278215133957,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 10249
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2583436235686514,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 10250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3183889428107916,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 10251
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.142342624410879,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 10252
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6016876041075054,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 10253
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8296431408203224,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 10254
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1475825338663956,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 10255
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14758089377985026,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 10256
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.270871841617449,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 10257
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.663437951573696,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 10258
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4485030964932513,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 10259
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.231336114327588,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 10260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1678304430874715,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 10261
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1433279332254926,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 10262
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9517965282673452,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 10263
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5357842042367373,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 10264
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25938915995357653,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 10265
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.172309419302359,
+      "learning_rate": 2e-05,
+      "loss": 0.1186,
+      "step": 10266
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0374410328201495,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 10267
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.806967300951977,
+      "learning_rate": 2e-05,
+      "loss": 0.0751,
+      "step": 10268
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1851306165665303,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 10269
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5589289484037638,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 10270
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8068935257431638,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 10271
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15809611834184592,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 10272
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0711367141487302,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 10273
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.375217508831531,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 10274
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8585784955060127,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 10275
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.445294106856148,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 10276
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7028986248541815,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 10277
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4775177625695388,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 10278
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7854041797582605,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 10279
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3837780708695937,
+      "learning_rate": 2e-05,
+      "loss": 0.1202,
+      "step": 10280
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0749134689121136,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 10281
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0552374077293922,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 10282
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2032896152191197,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 10283
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.621036996758956,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 10284
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4532571482508803,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 10285
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8735342637655754,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 10286
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2617981593413812,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 10287
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.696395612170581,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 10288
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.368408250880669,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 10289
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0209602892870615,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 10290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.614358457498259,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 10291
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.407549828817836,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 10292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8790599773328747,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 10293
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5921269548489043,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 10294
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0552049850807403,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 10295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.690315865898863,
+      "learning_rate": 2e-05,
+      "loss": 0.0881,
+      "step": 10296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4276623887846673,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 10297
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.331950679757494,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 10298
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.8435954407674515,
+      "learning_rate": 2e-05,
+      "loss": 0.1371,
+      "step": 10299
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30395126646784876,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 10300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.879321603679531,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 10301
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4016968887443007,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 10302
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.326673919774649,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 10303
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.244647972186946,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 10304
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1832917197340973,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 10305
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9783024660425846,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 10306
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.42463508815870427,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 10307
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.819044594267532,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 10308
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9884303548547502,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 10309
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.4695267435740265,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 10310
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.635447475590635,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 10311
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0099649233374266,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 10312
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2267751491792884,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 10313
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.944674363646416,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 10314
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.43899267870070435,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 10315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6176323762523137,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 10316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.41309574918164355,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 10317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8256260577187127,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 10318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6252469528593774,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 10319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2734271090030909,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 10320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9187186769792908,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 10321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7870468045313457,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 10322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.837875512088335,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 10323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7382112903851896,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 10324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25841743128393924,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 10325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4518139056828856,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 10326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.055018873290481,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 10327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7416186580094846,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 10328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5920655629437457,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 10329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6314494453194242,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 10330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3643587601512251,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 10331
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0218553875262315,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 10332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.864789661721817,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 10333
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9089147983998784,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 10334
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2772464540155706,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 10335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8339746249491227,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 10336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.28895341816227,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 10337
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8284360180895898,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 10338
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5725030755284348,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 10339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2215015943269323,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 10340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.301754698924915,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 10341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.359973775596785,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 10342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6273104308513309,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 10343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.296167709328257,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 10344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9151660745545056,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 10345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5680382044624062,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 10346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5803518164347128,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 10347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7896753452555034,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 10348
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2311248967471522,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 10349
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4636244449925848,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 10350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7507710120783953,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 10351
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1042125020620275,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 10352
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5555174639367386,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 10353
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5568046546347003,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 10354
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.955423419391987,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 10355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4022539795431253,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 10356
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1938679242296293,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 10357
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.57532713989405,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 10358
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3072892248528563,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 10359
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.839138469155148,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 10360
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.141975384162467,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 10361
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8541878992248482,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 10362
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6208505694239038,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 10363
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.056324458460443,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 10364
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4813298360821643,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 10365
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5301564055938477,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 10366
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2968762799285603,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 10367
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2461305810154957,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 10368
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7805452123663246,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 10369
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.178454853842956,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 10370
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.505181731346275,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 10371
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6922283134741827,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 10372
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.891720583138287,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 10373
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4046690637314279,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 10374
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.232002225532926,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 10375
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.417168004954373,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 10376
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9984008101764932,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 10377
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7492953009773325,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 10378
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.890303537009672,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 10379
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8459768931387373,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 10380
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.642260072739666,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 10381
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.966866467923484,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 10382
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16843931132565282,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 10383
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8989216702360505,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 10384
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.01665665595142,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 10385
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6798763232098156,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 10386
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9754611709742685,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 10387
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5801568539814331,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 10388
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8146989571676244,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 10389
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2720339858319061,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 10390
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7888116785661121,
+      "learning_rate": 2e-05,
+      "loss": 0.1373,
+      "step": 10391
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1108124245414093,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 10392
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.181167140989304,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 10393
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7451696489386526,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 10394
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3125273930941282,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 10395
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.245466116090171,
+      "learning_rate": 2e-05,
+      "loss": 0.1065,
+      "step": 10396
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5340902333508435,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 10397
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8454584908766247,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 10398
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5726698456511494,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 10399
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6261006809093141,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 10400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.23993833646258347,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 10401
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.428706682137588,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 10402
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3387897303990409,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 10403
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.45946503490001234,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 10404
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6014477944584564,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 10405
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.249099019681805,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 10406
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.340524208333189,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 10407
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.837181876385445,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 10408
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3948380987758116,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 10409
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.476069776886097,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 10410
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7370400101177514,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 10411
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.36603288850474,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 10412
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.393371101849994,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 10413
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5615470789153965,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 10414
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5910073411006107,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 10415
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.098953024081502,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 10416
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.4462046607888075,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 10417
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.44041752428400177,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 10418
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2944952030262449,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 10419
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.62967538228108,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 10420
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8571480072525253,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 10421
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7716579692044245,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 10422
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.131872536937239,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 10423
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7254627888270475,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 10424
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9373151014147905,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 10425
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5788030484133726,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 10426
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5971975872838697,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 10427
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2060670921521737,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 10428
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7584710160352387,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 10429
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1215015693071928,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 10430
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.883513577312306,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 10431
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2020449422414092,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 10432
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.44016477818296407,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 10433
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5486428533626202,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 10434
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4356684419189796,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 10435
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2172151011388177,
+      "learning_rate": 2e-05,
+      "loss": 0.0703,
+      "step": 10436
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.407521295978164,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 10437
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.517834727635662,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 10438
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13467032061895942,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 10439
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34463867277433075,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 10440
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6622070831929077,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 10441
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5085834647034968,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 10442
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5246681133647595,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 10443
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.34486254159681,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 10444
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7725522692980873,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 10445
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0828649630167093,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 10446
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5430438037840897,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 10447
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.745274298048758,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 10448
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.182680870680284,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 10449
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38993572457385195,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 10450
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9149949623969812,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 10451
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.27022778997368846,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 10452
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8870960434900914,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 10453
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9237236302577672,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 10454
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0578253515637686,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 10455
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1457374676040417,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 10456
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9787269773125924,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 10457
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5945005706422442,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 10458
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9368704589469035,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 10459
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.372622052380527,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 10460
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38288110001599496,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 10461
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4518105607283496,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 10462
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0919666465374611,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 10463
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.548525689579719,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 10464
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9210121885207898,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 10465
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6038775067326725,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 10466
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.885270876356264,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 10467
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5234205978629685,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 10468
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.556757409160323,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 10469
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2120802658997714,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 10470
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1576750096211326,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 10471
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.054667500805683,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 10472
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.246571102281835,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 10473
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7925899329828394,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 10474
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8249087611344996,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 10475
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40954184153518775,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 10476
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.44124924559704315,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 10477
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7440679031456154,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 10478
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.770837365810423,
+      "learning_rate": 2e-05,
+      "loss": 0.0913,
+      "step": 10479
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3753275037207475,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 10480
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2592538666505806,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 10481
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0540519501989745,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 10482
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4593830346021077,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 10483
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1049664552998952,
+      "learning_rate": 2e-05,
+      "loss": 0.1069,
+      "step": 10484
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.505723765691746,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 10485
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.232662691892858,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 10486
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0195765831469012,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 10487
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.616120351466915,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 10488
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.0568985101738235,
+      "learning_rate": 2e-05,
+      "loss": 0.0945,
+      "step": 10489
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5230290574516032,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 10490
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9079341367033109,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 10491
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7899424983836334,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 10492
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.793134436327875,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 10493
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4990036357354575,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 10494
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.22618905893248684,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 10495
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4827205070684424,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 10496
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5267213299528128,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 10497
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.011672265594005,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 10498
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.10400709407022156,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 10499
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8107653220084842,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 10500
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.104560829680667,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 10501
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7768264856199691,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 10502
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1833393591516939,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 10503
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.202300546472022,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 10504
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.717493155130294,
+      "learning_rate": 2e-05,
+      "loss": 0.0898,
+      "step": 10505
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6691281660457156,
+      "learning_rate": 2e-05,
+      "loss": 0.1002,
+      "step": 10506
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.291346828196847,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 10507
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3257767056245084,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 10508
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.527776762314486,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 10509
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0043338799975765,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 10510
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.726519692006786,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 10511
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.547129754560848,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 10512
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.798784192318795,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 10513
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0858790978843675,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 10514
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4993540346234925,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 10515
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9084035016584915,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 10516
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3261445931731044,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 10517
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1186594895916886,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 10518
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9382840692760255,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 10519
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3802123111691524,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 10520
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5217686424836401,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 10521
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9337379312309814,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 10522
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9872435614255661,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 10523
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.900189002926754,
+      "learning_rate": 2e-05,
+      "loss": 0.0876,
+      "step": 10524
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5170225798272379,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 10525
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2726869770272464,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 10526
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9374820707514073,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 10527
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0253158687407156,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 10528
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.194577708067532,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 10529
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.44065615969925404,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 10530
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6204406893574415,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 10531
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4406741324553139,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 10532
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3955314003692436,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 10533
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8660478750236839,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 10534
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.518763523952143,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 10535
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.35186344617301524,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 10536
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3340236239209155,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 10537
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.22500026524051292,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 10538
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7894564108046415,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 10539
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7379495918055,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 10540
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8024858892988734,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 10541
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.254892338288339,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 10542
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1001903493079896,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 10543
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2845094740116459,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 10544
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26333093146142217,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 10545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7949593989421342,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 10546
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.183112185200611,
+      "learning_rate": 2e-05,
+      "loss": 0.0943,
+      "step": 10547
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1779445960831234,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 10548
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7324480917064018,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 10549
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.144599638959348,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 10550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07197859181912021,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 10551
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2952444557935683,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 10552
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0045743129825673,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 10553
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6127767813248723,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 10554
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2820294237504997,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 10555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.598963264068114,
+      "learning_rate": 2e-05,
+      "loss": 0.1311,
+      "step": 10556
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.671242082240106,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 10557
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3129154637673324,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 10558
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.543676603040991,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 10559
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.98201542591017,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 10560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2102055887240306,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 10561
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8945372494390225,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 10562
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9682548142399363,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 10563
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0679962824388873,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 10564
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.052344787300662,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 10565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6888899191298352,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 10566
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.17265816014299107,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 10567
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.528817794806231,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 10568
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4566681028369193,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 10569
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.143225211367351,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 10570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6865456239428296,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 10571
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3942609077421033,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 10572
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7143308122907551,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 10573
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.216505731170449,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 10574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.868462369761687,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 10575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2399977006653075,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 10576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4223926985082325,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 10577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7817257697517348,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 10578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.04352464131340873,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 10579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2845684891027824,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 10580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1172541389138958,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 10581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8361694156839024,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 10582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9407856676607134,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 10583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9198587898142792,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 10584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.621070261577098,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 10585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.557249318773815,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 10586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4224680407911257,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 10587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.12648036225504,
+      "learning_rate": 2e-05,
+      "loss": 0.1049,
+      "step": 10588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5120382403588305,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 10589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2031025211196757,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 10590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8597502842686642,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 10591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.457226005283908,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 10592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8576013777165064,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 10593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.632052660038362,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 10594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6835246236389003,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 10595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.385842756410672,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 10596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4015393088302848,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 10597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.035967357816524506,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 10598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30286210349662085,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 10599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.500314978784382,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 10600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.071069693196097,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 10601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.10709620440007862,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 10602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4111708269641758,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 10603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.708038098798364,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 10604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.915570058231676,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 10605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.1519742866490486,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 10606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7010029028670519,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 10607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.021588584587885,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 10608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.848573356272812,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 10609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3650334855747743,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 10610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9272406619587316,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 10611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6156008337583962,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 10612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5268749555178224,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 10613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0715792253515093,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 10614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1862266103933936,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 10615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9096622354241795,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 10616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.955922580567009,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 10617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.771381331329579,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 10618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7806820549599311,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 10619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.421225577621691,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 10620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.680568304442642,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 10621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1081720066562317,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 10622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1974999875013332,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 10623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7974523528977062,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 10624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9429679990700652,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 10625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.166145750333284,
+      "learning_rate": 2e-05,
+      "loss": 0.1249,
+      "step": 10626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2806433660438516,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 10627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.098848091964224,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 10628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0964823789072162,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 10629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9665080245879312,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 10630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8738934748651568,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 10631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.09975937418407,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 10632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4330024851200467,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 10633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2527569408662196,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 10634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7385576007067876,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 10635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.293286766633516,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 10636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0416228793795694,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 10637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4283130480209767,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 10638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.23194004250646266,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 10639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.576166808178415,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 10640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.702347604257913,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 10641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4819566964637674,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 10642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2113740847174217,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 10643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9501684776297528,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 10644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4875610691645633,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 10645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.423181994411769,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 10646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3795836948558289,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 10647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.928330785883679,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 10648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8517973481979126,
+      "learning_rate": 2e-05,
+      "loss": 0.1189,
+      "step": 10649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.783986582795396,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 10650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39622993009264296,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 10651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9308435961386099,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 10652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.961614119680221,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 10653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6124594704200277,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 10654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9943910865529424,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 10655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.344505408225455,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 10656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5061634909452224,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 10657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.275558269038834,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 10658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7597717834066163,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 10659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.522560845523197,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 10660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.304052543458234,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 10661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4179464711522436,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 10662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4599907737822553,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 10663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4401837879357053,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 10664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.06120733731676,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 10665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3199224403743657,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 10666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5258132017103772,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 10667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9162703690600154,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 10668
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8051027504284195,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 10669
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4157107494209224,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 10670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5844769458890107,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 10671
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3978357944340836,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 10672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8429498390951329,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 10673
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8160834153258835,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 10674
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0456542823063986,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 10675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.783822527602536,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 10676
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7022721372910414,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 10677
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.308198979593946,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 10678
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.465165166560246,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 10679
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8365829307220727,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 10680
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.293106944402118,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 10681
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3162470179053276,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 10682
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3642307680216912,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 10683
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8794454997202211,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 10684
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.4462700635371055,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 10685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9436084070181274,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 10686
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3635096754843603,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 10687
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.11674829672065144,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 10688
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7554028079758537,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 10689
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3956604988172012,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 10690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0114579030175805,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 10691
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.7429208706892965,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 10692
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6041648879441888,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 10693
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3248476860443277,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 10694
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3729834508016592,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 10695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.758079091090955,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 10696
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.149427048071879,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 10697
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1099793875171082,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 10698
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9138437574306961,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 10699
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5269977148322473,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 10700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2390663714546823,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 10701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2563474190360007,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 10702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7718869860872924,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 10703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4866497879752285,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 10704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6636876807221164,
+      "learning_rate": 2e-05,
+      "loss": 0.1748,
+      "step": 10705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4043973799222291,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 10706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7609784563701987,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 10707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6560930563951874,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 10708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9595372612632421,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 10709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0881817357994774,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 10710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.673388749435592,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 10711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.669998053398436,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 10712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17461527786688708,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 10713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3715429614522145,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 10714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7969814678200965,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 10715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7403976944693267,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 10716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.528363679995726,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 10717
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3114710437877352,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 10718
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5749525454883406,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 10719
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7770621611156414,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 10720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6403078045071973,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 10721
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9508047158974482,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 10722
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7809638747246785,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 10723
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9314587813475631,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 10724
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.46199333365593237,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 10725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8764998320075057,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 10726
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6745246497384825,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 10727
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6234147471623708,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 10728
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.591857534551644,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 10729
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8317989687778164,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 10730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.131670830362682,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 10731
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0515542163086047,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 10732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.949075570316393,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 10733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.38966189865532386,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 10734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.48551710901453127,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 10735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7990977370584527,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 10736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0016776555792015,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 10737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9815671384161837,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 10738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.496133141943646,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 10739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9927514244524236,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 10740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7963419608724887,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 10741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.216455771560945,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 10742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1890290518406004,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 10743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5793184356765104,
+      "learning_rate": 2e-05,
+      "loss": 0.1071,
+      "step": 10744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0383105317658954,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 10745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.291954080284469,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 10746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15776124513743772,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 10747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.240238951301723,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 10748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9500631650203759,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 10749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4107763331838874,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 10750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.268160118573274,
+      "learning_rate": 2e-05,
+      "loss": 0.0857,
+      "step": 10751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5377077410917729,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 10752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.36281679057262956,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 10753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0777503205654841,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 10754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0824949737502618,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 10755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.344912736485035,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 10756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0757812234051602,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 10757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8506998507879127,
+      "learning_rate": 2e-05,
+      "loss": 0.0755,
+      "step": 10758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0171029047288047,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 10759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44442912371706617,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 10760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7760700063224981,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 10761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8025467553585838,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 10762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8318265206349645,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 10763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9608788348537542,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 10764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.417292830518007,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 10765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.089385991000736,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 10766
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4473066439974056,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 10767
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8825925704990855,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 10768
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2176594807798766,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 10769
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9693838322505801,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 10770
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2150310114446783,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 10771
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0837887699183861,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 10772
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9768660583297732,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 10773
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9979662359370973,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 10774
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9059584347734944,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 10775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1065548794561293,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 10776
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8432504093226851,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 10777
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.680777103007089,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 10778
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.06139712764596558,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 10779
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8655812276521588,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 10780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.04355115299721,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 10781
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7484136334717303,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 10782
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1185278436892676,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 10783
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9062401556032866,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 10784
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.478823671287381,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 10785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.41880515181086,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 10786
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7072114569031327,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 10787
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6091526951999042,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 10788
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.697246279416504,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 10789
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3797620762869326,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 10790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.784508243201415,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 10791
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6605178977647297,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 10792
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.499363436718668,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 10793
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0959634998501753,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 10794
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9813875551451933,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 10795
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.07703078760967,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 10796
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.232142259870177,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 10797
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0031203723681434,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 10798
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9772149618673558,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 10799
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.419449311329496,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 10800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6507605797826853,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 10801
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4118781322365936,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 10802
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2944893855285322,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 10803
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3048554729217823,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 10804
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.657286138963701,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 10805
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7144391108035778,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 10806
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3999714882535907,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 10807
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8932700115389016,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 10808
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1117038728491944,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 10809
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8253643408593967,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 10810
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4017170645414838,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 10811
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.716861067140139,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 10812
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3945118738836356,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 10813
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6966537868603543,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 10814
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.21061724617379685,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 10815
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.333796862200073,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 10816
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5876767796148794,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 10817
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.568666724924156,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 10818
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6835150101136025,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 10819
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9541529277435135,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 10820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14535702822998056,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 10821
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2132786423458022,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 10822
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.79843317079517,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 10823
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.544642600414655,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 10824
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.21447841620225183,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 10825
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7383038808656748,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 10826
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9323174918805027,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 10827
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6756031283380417,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 10828
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5809299733182014,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 10829
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.667694977588895,
+      "learning_rate": 2e-05,
+      "loss": 0.0736,
+      "step": 10830
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3740268293217395,
+      "learning_rate": 2e-05,
+      "loss": 0.0675,
+      "step": 10831
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8769395333551206,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 10832
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.62725760244652,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 10833
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8310263909632787,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 10834
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.009840706075938,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 10835
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.49995887110351883,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 10836
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1830812861875437,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 10837
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4524155521889689,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 10838
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0740088240369312,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 10839
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0312464787379336,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 10840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.247273963541632,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 10841
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.549580683441372,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 10842
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.574637824520572,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 10843
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.928144452256809,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 10844
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.01306878452587,
+      "learning_rate": 2e-05,
+      "loss": 0.1008,
+      "step": 10845
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4500062397468674,
+      "learning_rate": 2e-05,
+      "loss": 0.1073,
+      "step": 10846
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.082085684819466,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 10847
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6712920685143393,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 10848
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6312077099806324,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 10849
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6756112371855725,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 10850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5793336142827191,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 10851
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8946610735509019,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 10852
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40379556729432986,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 10853
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2296199753607313,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 10854
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2039283448406732,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 10855
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8528247542460132,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 10856
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.540771472645365,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 10857
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4774990441953404,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 10858
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8740485638419464,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 10859
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.719511789396297,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 10860
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4936931211634366,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 10861
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.036195629060145,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 10862
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8514563936501154,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 10863
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0499336766685943,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 10864
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8935976111936785,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 10865
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8283854894000091,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 10866
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9057862640921183,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 10867
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9660959064805599,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 10868
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.166060644151462,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 10869
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.180675623279447,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 10870
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7872435892318617,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 10871
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3251554467140347,
+      "learning_rate": 2e-05,
+      "loss": 0.0985,
+      "step": 10872
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8391947968507685,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 10873
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.29406207480891394,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 10874
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1161838638346402,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 10875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9184513203404321,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 10876
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1451208419372723,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 10877
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.783497312915383,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 10878
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1102445176729436,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 10879
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3102299657522027,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 10880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0232844350060484,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 10881
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.988227336117221,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 10882
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7441521758479845,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 10883
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8660715844739906,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 10884
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1825502253080984,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 10885
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0601367785233857,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 10886
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.123142440905143,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 10887
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.535975865994524,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 10888
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6805067715276393,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 10889
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5859587758083435,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 10890
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4939036461885549,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 10891
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8240085189437147,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 10892
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9588615769692637,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 10893
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5449831528203294,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 10894
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1253236728866745,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 10895
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7740437992256378,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 10896
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.549318042894155,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 10897
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9959908475685129,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 10898
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0879163756139438,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 10899
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9426982450560144,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 10900
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0806296033914324,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 10901
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.06586598295879,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 10902
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.826056133009133,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 10903
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9014038541368214,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 10904
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8385982630792532,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 10905
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.637055719689182,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 10906
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4875497697470041,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 10907
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7146634852175666,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 10908
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.33927777176179374,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 10909
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2998430689211398,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 10910
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4742529656827124,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 10911
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5934996314765018,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 10912
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0592559770096308,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 10913
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6072613956057165,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 10914
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.11866081662792802,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 10915
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3324182121698454,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 10916
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.208294380435763,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 10917
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.277386558004965,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 10918
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.230682452185518,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 10919
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9600124532926768,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 10920
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2257186027549913,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 10921
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.775399215437972,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 10922
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0032728995269895,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 10923
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.137517589391178,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 10924
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.058012964444469,
+      "learning_rate": 2e-05,
+      "loss": 0.0729,
+      "step": 10925
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.120014801268186,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 10926
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.629369006303104,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 10927
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6818249254771005,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 10928
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.699666184137077,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 10929
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5306892146655082,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 10930
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8879318677501133,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 10931
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3811490673626998,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 10932
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0576215833746943,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 10933
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3857850642457993,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 10934
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5359938959457387,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 10935
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6734785655360804,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 10936
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9005012754432515,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 10937
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9464359385982557,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 10938
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7655767308803726,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 10939
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.205833879948015,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 10940
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.886608095082596,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 10941
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1989736399707824,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 10942
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2928776633579857,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 10943
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.41116188374789,
+      "learning_rate": 2e-05,
+      "loss": 0.1104,
+      "step": 10944
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6967380568743747,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 10945
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2058463097413896,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 10946
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.562820554276589,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 10947
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.813769586458669,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 10948
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8967671332805662,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 10949
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.45044699561774576,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 10950
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1312255793512092,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 10951
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7448832890209904,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 10952
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6219126503056633,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 10953
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0675390398670395,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 10954
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0095061036936703,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 10955
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1075703085804967,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 10956
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8508031416669554,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 10957
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4352769662821827,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 10958
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.099559069623682,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 10959
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6635119231388155,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 10960
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1729083703363217,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 10961
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1087569005809508,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 10962
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5351264375835205,
+      "learning_rate": 2e-05,
+      "loss": 0.0972,
+      "step": 10963
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.18886926575507,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 10964
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.598718285487206,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 10965
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0315942478697009,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 10966
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9751415346833376,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 10967
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.027885937818135,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 10968
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8615342842825422,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 10969
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6321231242433597,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 10970
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.472319131331706,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 10971
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0420340017493175,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 10972
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5169996087406677,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 10973
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9331579468636215,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 10974
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.579071742939882,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 10975
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.304799035176395,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 10976
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.712136794444286,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 10977
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0860242504125703,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 10978
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8260814716486555,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 10979
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5956168690753921,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 10980
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5340857927571654,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 10981
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1022813223742343,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 10982
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5894477980507622,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 10983
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8534650285681664,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 10984
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.333282186310908,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 10985
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.665569368464364,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 10986
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.29738175909451314,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 10987
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.472287078644433,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 10988
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8230829159433273,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 10989
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.899793117641683,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 10990
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3605198528405467,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 10991
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4938945492829945,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 10992
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5745092838530916,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 10993
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.210035375941987,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 10994
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4739840969175,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 10995
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9362387688861675,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 10996
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9356178247591622,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 10997
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.438696836422938,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 10998
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3326609414724406,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 10999
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1271785728750439,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 11000
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5480055834652084,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 11001
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5678304157902416,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 11002
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.774355680085736,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 11003
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5881934162442667,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 11004
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.20713821383026312,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 11005
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.679789897177152,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 11006
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.383256414368206,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 11007
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1031596568061317,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 11008
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5115019652122862,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 11009
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8630070834514056,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 11010
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6480274142327682,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 11011
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.321444537280786,
+      "learning_rate": 2e-05,
+      "loss": 0.1195,
+      "step": 11012
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2116089958693192,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 11013
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.733238725265463,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 11014
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1303318618326856,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 11015
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.19437769340599323,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 11016
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5753238141052397,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 11017
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1563416522018142,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 11018
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5549411730906275,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 11019
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.633546359061007,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 11020
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.288155639978067,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 11021
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6075322981180806,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 11022
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.234529500763379,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 11023
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8785542180124073,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 11024
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8145225034550414,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 11025
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4608134985898165,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 11026
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9995795642488439,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 11027
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5641614166768663,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 11028
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6465364187445994,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 11029
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.274343932574604,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 11030
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.829278956669063,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 11031
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3045592191523476,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 11032
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5127455898177127,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 11033
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.508160090694856,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 11034
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.465085266548277,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 11035
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2742294695245205,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 11036
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8166805335898178,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 11037
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.50981070980448,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 11038
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.559150844815157,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 11039
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6589376952004508,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 11040
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3300143547287318,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 11041
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3518341761618444,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 11042
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1263266794095286,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 11043
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.280011925415172,
+      "learning_rate": 2e-05,
+      "loss": 0.0871,
+      "step": 11044
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5819171628905866,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 11045
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7695038301646467,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 11046
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2382874593970152,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 11047
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.577365832828252,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 11048
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.170343288123604,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 11049
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5285393768055067,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 11050
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.658434740477186,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 11051
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.555781837119229,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 11052
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4567203138818297,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 11053
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.13766251545404,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 11054
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.20721681843390816,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 11055
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.748357021938351,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 11056
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6912276564475412,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 11057
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8290183089550204,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 11058
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6057086592330033,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 11059
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8792886642248143,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 11060
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5983839155464913,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 11061
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2207153807978512,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 11062
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2912006482325866,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 11063
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7075518549673225,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 11064
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8759580204084633,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 11065
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1011936841890992,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 11066
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9080048910074303,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 11067
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9181093473508941,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 11068
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.45368150540879537,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 11069
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4976381940554198,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 11070
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1473317022845424,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 11071
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5127850844489612,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 11072
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7467452163117574,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 11073
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.306350339228267,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 11074
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2683956699742356,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 11075
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.250019073405485,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 11076
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.378237675768309,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 11077
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4040728776362736,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 11078
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6945862510974314,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 11079
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.836009831728406,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 11080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7494411257300515,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 11081
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8910274600154857,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 11082
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.891966316618424,
+      "learning_rate": 2e-05,
+      "loss": 0.1003,
+      "step": 11083
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9371408728346352,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 11084
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8535545916358894,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 11085
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8552858340002433,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 11086
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3003291931085603,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 11087
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8509528430081126,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 11088
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.239380784449132,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 11089
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9870704306611986,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 11090
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.706817573589196,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 11091
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1836995999685471,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 11092
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2828602523954466,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 11093
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3544152277706454,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 11094
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7988856884798375,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 11095
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0459383859850047,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 11096
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2685222122554447,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 11097
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4990917091922645,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 11098
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0796637680161625,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 11099
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1262077938243786,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 11100
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.180315620997966,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 11101
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.626027153569758,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 11102
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.209223512530976,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 11103
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.21377368084751797,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 11104
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5128610321940754,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 11105
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.678428865907733,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 11106
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.015033406947148,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 11107
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.26793298009791,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 11108
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.176262574573103,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 11109
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7496090861084128,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 11110
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1509619795849115,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 11111
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.368651186208102,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 11112
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.728916893196701,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 11113
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7467046091578973,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 11114
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3528127050244163,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 11115
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4192153033204493,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 11116
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9038032087478343,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 11117
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1786306136442697,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 11118
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.782237398297539,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 11119
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9194026023596686,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 11120
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6838567807733784,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 11121
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.11730897488477678,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 11122
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5629887388705985,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 11123
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.290444126936398,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 11124
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.271778219229019,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 11125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.771047855921767,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 11126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.032931661729095,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 11127
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7101378640710925,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 11128
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.54168292200265,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 11129
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4087984535716553,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 11130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3911085252556687,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 11131
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.013621559143317,
+      "learning_rate": 2e-05,
+      "loss": 0.0902,
+      "step": 11132
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2732200095092683,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 11133
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1253671576804951,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 11134
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.228272433974029,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 11135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4510999349732332,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 11136
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8768093280310016,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 11137
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5542471306209941,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 11138
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7443324765375413,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 11139
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6362815010876298,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 11140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3740403554562148,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 11141
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.887154325138409,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 11142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0574474291846783,
+      "learning_rate": 2e-05,
+      "loss": 0.0815,
+      "step": 11143
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4634873516608957,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 11144
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9215827037315552,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 11145
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0291266273553576,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 11146
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7267809236584153,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 11147
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.764010890481033,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 11148
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.515401686114586,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 11149
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8280020778443862,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 11150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3083307984633135,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 11151
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9023583721136545,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 11152
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.650698300185368,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 11153
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7081139011010356,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 11154
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1251563968326037,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 11155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0854111637749664,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 11156
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2870971030777592,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 11157
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9473724934179494,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 11158
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9751502239210412,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 11159
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1730947631451099,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 11160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.323868651795459,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 11161
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1953984647269025,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 11162
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6401471850012141,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 11163
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2363574369550507,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 11164
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1654679454755454,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 11165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4059247806446,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 11166
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.01436060792313,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 11167
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0727951326639587,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 11168
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.09016966372949,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 11169
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.511820125165519,
+      "learning_rate": 2e-05,
+      "loss": 0.0986,
+      "step": 11170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.12814964429587933,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 11171
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2729169113337655,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 11172
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4415942227164205,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 11173
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5408688287543816,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 11174
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3046110864333402,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 11175
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.094981699845595,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 11176
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1569913600971606,
+      "learning_rate": 2e-05,
+      "loss": 0.0837,
+      "step": 11177
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5795502977565166,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 11178
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.181117422747522,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 11179
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5999785041932144,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 11180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1973321345464176,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 11181
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4643382961278775,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 11182
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40469694384303917,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 11183
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1376337638396117,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 11184
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6825572494130743,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 11185
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.921161911386738,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 11186
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5315403565691792,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 11187
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.234407224622848,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 11188
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5936482290701903,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 11189
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2880795566280527,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 11190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.513779146622062,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 11191
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4316183091072814,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 11192
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.59328970152411,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 11193
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4031963134243486,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 11194
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8753521270709785,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 11195
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0851159656065175,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 11196
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4966281504160195,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 11197
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3262354737369311,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 11198
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.325559675108832,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 11199
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.891202631792012,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 11200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.049539593424198,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 11201
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.431729553312391,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 11202
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2318938023724066,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 11203
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5594300631907925,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 11204
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3750927633725802,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 11205
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6004705660972731,
+      "learning_rate": 2e-05,
+      "loss": 0.0941,
+      "step": 11206
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7284884520727857,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 11207
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.401730671705321,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 11208
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.11996226671153856,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 11209
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5899630798490885,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 11210
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.074267918970281,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 11211
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30304125336843907,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 11212
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4879063411893125,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 11213
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.944693502852993,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 11214
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.168001767715658,
+      "learning_rate": 2e-05,
+      "loss": 0.1184,
+      "step": 11215
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2104596087158714,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 11216
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.12938138899809,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 11217
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7924148428980156,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 11218
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.38698953723317636,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 11219
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0286674529431301,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 11220
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3197760684364406,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 11221
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5153925978834027,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 11222
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.174363172410692,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 11223
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30944931361146494,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 11224
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4837635587925992,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 11225
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.413992377685548,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 11226
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4395487871953266,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 11227
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3660261616833629,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 11228
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1438267285375905,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 11229
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.345811858857178,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 11230
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.603896826433171,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 11231
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.302524599157221,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 11232
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5269469569745198,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 11233
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.36034784082491256,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 11234
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8915594786690729,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 11235
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0351162884754128,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 11236
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3832199444700264,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 11237
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6903492234539605,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 11238
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.176892529268355,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 11239
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7623507091521116,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 11240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1159073223989224,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 11241
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9119446828362029,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 11242
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7088403897111005,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 11243
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1207007274156615,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 11244
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0543584804902812,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 11245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.378076665813532,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 11246
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4737223904246654,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 11247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2100631018018884,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 11248
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3787586334264987,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 11249
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9319518267292871,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 11250
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39770599608669577,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 11251
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.363761533064288,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 11252
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4832943646887846,
+      "learning_rate": 2e-05,
+      "loss": 0.1303,
+      "step": 11253
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.125307285306367,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 11254
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4507079172165667,
+      "learning_rate": 2e-05,
+      "loss": 0.1041,
+      "step": 11255
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8122257723566175,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 11256
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5227031796437278,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 11257
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.6480813282665014,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 11258
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0416219148467163,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 11259
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1863820303909196,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 11260
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0627589190783422,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 11261
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.772335725531999,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 11262
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.617019159698225,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 11263
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.365384680018591,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 11264
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.972356671539774,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 11265
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.88909607739327,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 11266
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2625153096847639,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 11267
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7585440914645147,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 11268
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.172062159096681,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 11269
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1775445488688883,
+      "learning_rate": 2e-05,
+      "loss": 0.1089,
+      "step": 11270
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0297775159745022,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 11271
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0344531280423082,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 11272
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6332407120633687,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 11273
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.455949263082793,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 11274
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8274819738608611,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 11275
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4821387863759852,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 11276
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2159659958581956,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 11277
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9518794909848327,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 11278
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5391435147650098,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 11279
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.666706593711818,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 11280
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6237945614494207,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 11281
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0489098998287123,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 11282
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5071327377010604,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 11283
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6073115011285504,
+      "learning_rate": 2e-05,
+      "loss": 0.1033,
+      "step": 11284
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2538887094001678,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 11285
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2689971739434653,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 11286
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.824694879709338,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 11287
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.114396060361739,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 11288
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.338595284106766,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 11289
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7483320364979791,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 11290
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2823712567593124,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 11291
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2225823425770157,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 11292
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0545187510834197,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 11293
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0190570798516863,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 11294
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3507966983955675,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 11295
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7821187684339104,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 11296
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.067869809973651,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 11297
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.868907439575378,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 11298
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.525153365057126,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 11299
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.526138324789048,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 11300
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0261630972380456,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 11301
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.899174763565036,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 11302
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9194350229053407,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 11303
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.12170981008290434,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 11304
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9487481226041812,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 11305
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.177055352927761,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 11306
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.339488391072659,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 11307
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9286033014444609,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 11308
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15315394273991972,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 11309
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.43723962051848875,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 11310
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.053099503477351,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 11311
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.774879188313449,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 11312
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7627827573460743,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 11313
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2659577297307227,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 11314
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6734979677601536,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 11315
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.464939938248173,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 11316
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.94430056069885,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 11317
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.137039857827381,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 11318
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6395638076543764,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 11319
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24692356893364772,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 11320
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1555793466829105,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 11321
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.744170425633338,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 11322
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.23513377072688202,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 11323
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0558166758389564,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 11324
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0931323896259728,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 11325
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8333421508004685,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 11326
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1706265220489627,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 11327
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.3085461432628955,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 11328
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.943451630976022,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 11329
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.068523929251323,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 11330
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.218676821079461,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 11331
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.675177464292027,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 11332
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4565593374562136,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 11333
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2276289677987826,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 11334
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8023724790854705,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 11335
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4496697193948909,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 11336
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.19524329866895473,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 11337
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3975229096283523,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 11338
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6452193315320174,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 11339
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9184204288958607,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 11340
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0220856735160404,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 11341
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3430864780701086,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 11342
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3098539984004178,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 11343
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0992156790085144,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 11344
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.57157305848924,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 11345
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.180031336251667,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 11346
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.993918135028974,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 11347
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.292731043697213,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 11348
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6192524575701732,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 11349
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4421042947390794,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 11350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2802033153897677,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 11351
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3813380948489358,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 11352
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3156271632280907,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 11353
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5339024206406884,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 11354
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.173930184343606,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 11355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0377077944873036,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 11356
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7650408833553322,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 11357
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.147801644896814,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 11358
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0562858595041273,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 11359
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8878544663712363,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 11360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2627404862043399,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 11361
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.146594541562299,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 11362
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9802064886829128,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 11363
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4044683341633008,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 11364
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.169405734175015,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 11365
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7426370564233175,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 11366
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4242691511790513,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 11367
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7305097890670376,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 11368
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7731535431542254,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 11369
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.008905035053602,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 11370
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5248043682381276,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 11371
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3652171166570245,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 11372
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2633454612166277,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 11373
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.888938908600327,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 11374
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1815291423069019,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 11375
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0979492437516,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 11376
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.573123426115175,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 11377
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7685587349783023,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 11378
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0355941134297673,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 11379
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.512943313154305,
+      "learning_rate": 2e-05,
+      "loss": 0.1441,
+      "step": 11380
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2452592709363404,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 11381
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.843073446954698,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 11382
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5295598545114824,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 11383
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.812550777400786,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 11384
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.301521812659463,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 11385
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.111501830994835,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 11386
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26161184547313643,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 11387
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7081179508739917,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 11388
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.881276710592439,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 11389
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.23314251295738708,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 11390
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1795757821848722,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 11391
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.418443190567407,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 11392
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1886711255408766,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 11393
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.195547691593715,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 11394
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0288646003431574,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 11395
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2189018937360565,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 11396
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7987051889697718,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 11397
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4123043830901656,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 11398
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.346991848085732,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 11399
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.284160377450176,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 11400
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8765063262728557,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 11401
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3181881005074134,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 11402
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8683320130822023,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 11403
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8880110082823713,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 11404
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6398371527288615,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 11405
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5486430952355983,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 11406
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5480774058662893,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 11407
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1736034108397748,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 11408
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8369893381813828,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 11409
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5330350164108282,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 11410
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.854435390621227,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 11411
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9046711871705817,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 11412
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.156593379613724,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 11413
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9054826536527647,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 11414
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.547559854292707,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 11415
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6945096413345677,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 11416
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8215349806699752,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 11417
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.631014258059845,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 11418
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8777261907639898,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 11419
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7522566145748406,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 11420
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7833056299572922,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 11421
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6234091539559585,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 11422
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.204875635123598,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 11423
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6300731842602378,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 11424
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5472657451904066,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 11425
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2371840564962182,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 11426
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.39265726300655,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 11427
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.21339763847200088,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 11428
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0660531713597872,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 11429
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.198225050952197,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 11430
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.879514472842784,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 11431
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8755267124708592,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 11432
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.741769525130945,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 11433
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.6990074814182625,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 11434
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2649675710388901,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 11435
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8517974833172148,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 11436
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1074451262440057,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 11437
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.020871739086755,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 11438
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8527559063999637,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 11439
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5994289511941163,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 11440
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.652664983018882,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 11441
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.215324037065035,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 11442
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6498793673488947,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 11443
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2177618862849307,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 11444
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1836836374728303,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 11445
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.992078650997712,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 11446
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.873575038010161,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 11447
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.21874407425122,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 11448
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.523159999515729,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 11449
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.19578862330868185,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 11450
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7265775401402874,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 11451
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.791649958806166,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 11452
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4847372234561,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 11453
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4297602158486718,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 11454
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1442699459571968,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 11455
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5665145473186213,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 11456
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.113600729663385,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 11457
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9927383558846952,
+      "learning_rate": 2e-05,
+      "loss": 0.1094,
+      "step": 11458
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2146911214053646,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 11459
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5940444524686013,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 11460
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.474809686053392,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 11461
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5661923857927875,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 11462
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4735527244328075,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 11463
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.48421613564693,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 11464
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6797070169423902,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 11465
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30117899123120234,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 11466
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5229499274995377,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 11467
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0674703131127985,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 11468
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8084185866660228,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 11469
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0785928070586968,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 11470
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.255896027719652,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 11471
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4488865162820297,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 11472
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4825843197766323,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 11473
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.765492369393876,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 11474
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5034515528115913,
+      "learning_rate": 2e-05,
+      "loss": 0.0946,
+      "step": 11475
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8207300485863716,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 11476
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5848544729209316,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 11477
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7869964215658961,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 11478
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7903020960800639,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 11479
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3859888439279946,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 11480
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0191007560280239,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 11481
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0677585621005747,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 11482
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5251283443049783,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 11483
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4175565594491964,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 11484
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.596497178455532,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 11485
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6165811775804717,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 11486
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3712095119372298,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 11487
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.412640637974312,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 11488
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5504131504843552,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 11489
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.36174346467937,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 11490
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5960964059084073,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 11491
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5578572533701704,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 11492
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.722701522472644,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 11493
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4889193390979392,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 11494
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6243963001510617,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 11495
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0276970244982837,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 11496
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9232375288096772,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 11497
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9985009178138518,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 11498
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3539919764986759,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 11499
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.927488301658507,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 11500
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.397076646262919,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 11501
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.178810542314251,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 11502
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.178886322527385,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 11503
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3119827341129897,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 11504
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7330969382506638,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 11505
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2811824036421522,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 11506
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2407028635030874,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 11507
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6268670844413293,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 11508
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.389589265344728,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 11509
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.698492045480795,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 11510
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3891749752855103,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 11511
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.984281293253824,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 11512
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.059530953274934,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 11513
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.10970961950888095,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 11514
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7869523036031868,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 11515
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0341995712832668,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 11516
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.317945355997457,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 11517
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2494156396517004,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 11518
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0697746297912345,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 11519
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6466851343954585,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 11520
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8476542108045388,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 11521
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8542509600218486,
+      "learning_rate": 2e-05,
+      "loss": 0.1009,
+      "step": 11522
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6331762751746998,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 11523
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9002330436336454,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 11524
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0256737309141208,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 11525
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8212748590892216,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 11526
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9588620574583975,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 11527
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4446694988147293,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 11528
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0677835701586749,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 11529
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7583241652538851,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 11530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6836662542392606,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 11531
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27969694494570885,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 11532
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0508890203069536,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 11533
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5055580480882595,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 11534
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4555812129046286,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 11535
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.05595023644886772,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 11536
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.196401627051413,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 11537
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9733179132228588,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 11538
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0947098471244214,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 11539
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3315541659580603,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 11540
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1027281525531247,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 11541
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1548280311881367,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 11542
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5054695443955253,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 11543
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7866144256972191,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 11544
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.012167068882233,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 11545
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4474592606999717,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 11546
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8407016058169516,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 11547
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4068039544773137,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 11548
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.147381854032843,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 11549
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0461018112589942,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 11550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8459532553105524,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 11551
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7521833587519805,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 11552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2662348485525383,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 11553
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2128865308462786,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 11554
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4947587473719124,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 11555
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.655088193042878,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 11556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6755881031474595,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 11557
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.008703798751054,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 11558
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1463469076843922,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 11559
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.209663590248043,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 11560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.151801046535657,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 11561
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4548203617641808,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 11562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1188693527768443,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 11563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.623361418438995,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 11564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.033463196698347,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 11565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4953188968788689,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 11566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.516189322348797,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 11567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7620639505114843,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 11568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2846795215326463,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 11569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8205507931420069,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 11570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.602901904379422,
+      "learning_rate": 2e-05,
+      "loss": 0.1172,
+      "step": 11571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7107974478477153,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 11572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5763484878117783,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 11573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.572614207039778,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 11574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.20862689594278122,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 11575
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.256487984201743,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 11576
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1916635285869206,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 11577
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.070141133827557,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 11578
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8882490629610262,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 11579
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8163190380740195,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 11580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.547769307584898,
+      "learning_rate": 2e-05,
+      "loss": 0.0869,
+      "step": 11581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.22931302970633743,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 11582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.451693906882503,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 11583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4575444462036954,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 11584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0651447813813646,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 11585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6862383449933567,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 11586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4722194681601686,
+      "learning_rate": 2e-05,
+      "loss": 0.0804,
+      "step": 11587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8582870762745785,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 11588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.35204124702315426,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 11589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8080030117025996,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 11590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5036343415414348,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 11591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33956363797361355,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 11592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.29302832951310304,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 11593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.797389280280923,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 11594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.421742946654885,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 11595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3242130335681603,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 11596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.47841006301854433,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 11597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5030702323215026,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 11598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9080649010658335,
+      "learning_rate": 2e-05,
+      "loss": 0.1498,
+      "step": 11599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.1240454727966265,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 11600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1285547928786377,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 11601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8476112795754491,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 11602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6323998011750263,
+      "learning_rate": 2e-05,
+      "loss": 0.0939,
+      "step": 11603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.071135044441683,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 11604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6812474461967915,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 11605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2117398956834566,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 11606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5818827995068827,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 11607
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3870123753837864,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 11608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2832577330378334,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 11609
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4118872925593473,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 11610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6360840943974824,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 11611
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5282787378280955,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 11612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.259060258558015,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 11613
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.895680690391313,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 11614
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.850205515709852,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 11615
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.738206607335218,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 11616
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6338670601385195,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 11617
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2645575804124882,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 11618
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1949186684523427,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 11619
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.279254196270443,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 11620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.19224810846387838,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 11621
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7754274450827656,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 11622
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33015373312067064,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 11623
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3128622554221125,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 11624
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6114577362644673,
+      "learning_rate": 2e-05,
+      "loss": 0.1025,
+      "step": 11625
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9084811926330882,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 11626
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9606409623278602,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 11627
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5854620761191394,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 11628
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0580000673052472,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 11629
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2500764461700518,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 11630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.662651516180563,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 11631
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.580055340389262,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 11632
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0890990416691666,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 11633
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8149990018007683,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 11634
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3326547154632623,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 11635
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3485762499490759,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 11636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8460232170765469,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 11637
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7339625430074963,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 11638
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3271019243017412,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 11639
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.89700224266716,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 11640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5369069452056736,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 11641
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7075096175898943,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 11642
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.061712335374936,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 11643
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6089222522147212,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 11644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6623802193736399,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 11645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2416467313577284,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 11646
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1597414359995817,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 11647
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2847704961199464,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 11648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.136137442170911,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 11649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.728209317605618,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 11650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.704877259445831,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 11651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9536636581435667,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 11652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.18032542271643195,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 11653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8484528485731138,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 11654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7956966589026824,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 11655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.079142712114879,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 11656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5127148562326906,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 11657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0420720482948798,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 11658
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3672058306195183,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 11659
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2762351168400792,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 11660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.893385941044129,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 11661
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6511782439343183,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 11662
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.370757077493052,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 11663
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0419612464403216,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 11664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42527669356483594,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 11665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6449140394579324,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 11666
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.500898654555046,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 11667
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.864976539566529,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 11668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9484181484024448,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 11669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3663048647154235,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 11670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7954244421714654,
+      "learning_rate": 2e-05,
+      "loss": 0.1227,
+      "step": 11671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3408507084886838,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 11672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3187131957934237,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 11673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8348957109049715,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 11674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5205511676190901,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 11675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6726809546594121,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 11676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.013988878960353,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 11677
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8869574137402298,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 11678
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7518174482533638,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 11679
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.752380843946479,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 11680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6780822256986434,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 11681
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7396756280253336,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 11682
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3995368749094674,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 11683
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.119658994779677,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 11684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1554467736921668,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 11685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7987867717152182,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 11686
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5504301318250161,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 11687
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30520867742621044,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 11688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4143939393815563,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 11689
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4749790157829779,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 11690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9591978951316296,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 11691
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1431930061246522,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 11692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4228358629232716,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 11693
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3893321596199857,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 11694
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.702100265942206,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 11695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9713667629267247,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 11696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5358369456451135,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 11697
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7475607250095733,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 11698
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2412129543957355,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 11699
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4653558675116867,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 11700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9059575085806926,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 11701
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.171333080716109,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 11702
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9384222580583064,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 11703
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7653994714642911,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 11704
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0718008121362832,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 11705
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4120825716168393,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 11706
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9261656835917211,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 11707
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4383709446861147,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 11708
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5230339885282935,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 11709
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1605522248114473,
+      "learning_rate": 2e-05,
+      "loss": 0.0992,
+      "step": 11710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14298880233221586,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 11711
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0211865663846948,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 11712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2171203646129904,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 11713
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7494456706317058,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 11714
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.243865976550613,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 11715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.452308124775222,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 11716
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9437687274975164,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 11717
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2463910532475087,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 11718
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7259009979825135,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 11719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3732275810002017,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 11720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.517441359240053,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 11721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2175521841005197,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 11722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7501922996677005,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 11723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0199208300850702,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 11724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1218069645645565,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 11725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.43247089538880046,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 11726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.369438132233741,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 11727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.590895622445833,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 11728
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.394489810458005,
+      "learning_rate": 2e-05,
+      "loss": 0.0857,
+      "step": 11729
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41169653005557655,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 11730
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0181782247120748,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 11731
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2523772522555505,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 11732
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5204626528698337,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 11733
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.307812813561697,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 11734
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9447175352600996,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 11735
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4729599410548322,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 11736
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5315573234961946,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 11737
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.576109205912524,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 11738
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.3889222713981555,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 11739
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36157234804923943,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 11740
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6575613038823436,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 11741
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3197014201480421,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 11742
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8536603286260724,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 11743
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26494518408101386,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 11744
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.088522463081589,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 11745
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.528283346943946,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 11746
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7495838441223247,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 11747
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0979864635154253,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 11748
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7478999876066137,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 11749
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5622612923522383,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 11750
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1276295089436097,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 11751
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.100250504739685,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 11752
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0768715026054776,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 11753
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1694934296082953,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 11754
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.19836427659254718,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 11755
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0732140128134446,
+      "learning_rate": 2e-05,
+      "loss": 0.0902,
+      "step": 11756
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5233206262635255,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 11757
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6136865730142622,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 11758
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3855811382366632,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 11759
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.18183970053084653,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 11760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.332788112546954,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 11761
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9398909551904677,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 11762
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.167940018214354,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 11763
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.073885156275163,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 11764
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5180112645809672,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 11765
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6134504825096703,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 11766
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7395489379889666,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 11767
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2127051465026963,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 11768
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4557606576588116,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 11769
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.087968844455362,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 11770
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4298510848715846,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 11771
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9727796459317466,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 11772
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1425691115851317,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 11773
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7290035383219726,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 11774
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5089853387737152,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 11775
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.96498431212763,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 11776
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9604550562257426,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 11777
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9061652505643846,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 11778
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.35128104281446937,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 11779
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28419634845566955,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 11780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6159139585061966,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 11781
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.245232086759203,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 11782
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2630200373242513,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 11783
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1561759718832618,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 11784
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6776023292399165,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 11785
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2022572192452619,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 11786
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.988167750148695,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 11787
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.238485710520543,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 11788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8895325942055186,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 11789
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4273532510074796,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 11790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7431075680036083,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 11791
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6505139341800326,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 11792
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1843416574359225,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 11793
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2569445346014976,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 11794
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7698052806753481,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 11795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0564907049396246,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 11796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.267133120347182,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 11797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0188510423412234,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 11798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8617938592571146,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 11799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.130979752269122,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 11800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7507546522369191,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 11801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0439668943995173,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 11802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0690065756328806,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 11803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9375142414282885,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 11804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.091887113769273,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 11805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09459618316585927,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 11806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.32028578412124,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 11807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4094726978092194,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 11808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1778271387111043,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 11809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7785618119002118,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 11810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.048193087564235,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 11811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5300345754768228,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 11812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27152749960907596,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 11813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9703299827919352,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 11814
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5867914462101878,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 11815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8428544484053856,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 11816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3777630444047433,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 11817
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6746287261791593,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 11818
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0654154775705118,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 11819
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0269608034880577,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 11820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.46719386565295584,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 11821
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7886041445549625,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 11822
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.409778633397931,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 11823
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0833117898608478,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 11824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.261676056612107,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 11825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.034442911065815,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 11826
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.745205167041103,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 11827
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.999735346622021,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 11828
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.441683089772825,
+      "learning_rate": 2e-05,
+      "loss": 0.0928,
+      "step": 11829
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6589861550389045,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 11830
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3758044057323853,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 11831
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4455389721742071,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 11832
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8595493486583133,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 11833
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.726866337205308,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 11834
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2350745539850205,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 11835
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.956324867123149,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 11836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1035885947078847,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 11837
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5929964565469972,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 11838
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.130014897325742,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 11839
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.185278172888205,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 11840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1269939025165616,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 11841
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.085840769350256,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 11842
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4967471948034737,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 11843
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8403212136314793,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 11844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5394742191731283,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 11845
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.641101357668003,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 11846
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1147195192127826,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 11847
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6331020681703848,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 11848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.634776198559182,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 11849
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.746487368160348,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 11850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9829373230380044,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 11851
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7443165808878194,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 11852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0865106510916531,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 11853
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5337260721169808,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 11854
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8239885793545876,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 11855
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4643720757540308,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 11856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0143324035801566,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 11857
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0528854764573357,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 11858
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.14656523534001,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 11859
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5030950563109022,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 11860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2355466318405317,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 11861
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2433010364476298,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 11862
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5686822276275123,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 11863
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9269721815341216,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 11864
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9424614908056516,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 11865
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.671354622829427,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 11866
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8657579256431471,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 11867
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1636761529849613,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 11868
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.45843443694648345,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 11869
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3406036475252269,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 11870
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4559486151445711,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 11871
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.001192454804883,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 11872
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.540816795551365,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 11873
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.887067381764828,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 11874
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4905200365314113,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 11875
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9022978403541841,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 11876
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6703035251754653,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 11877
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8077271199815454,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 11878
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1057435879758217,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 11879
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.608622093994944,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 11880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.248035556222515,
+      "learning_rate": 2e-05,
+      "loss": 0.057,
+      "step": 11881
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.240241228951786,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 11882
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41449015503179143,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 11883
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4211824525393985,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 11884
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1491894325134284,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 11885
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1844928209788523,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 11886
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.2913398741709345,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 11887
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5135803588766272,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 11888
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5386853764801182,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 11889
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6298250310402709,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 11890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29834091760719944,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 11891
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6499658291344512,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 11892
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6418511821038496,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 11893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34447646321767267,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 11894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1376567340132207,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 11895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9259082047214814,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 11896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4922963047803162,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 11897
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6386546561343568,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 11898
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2685038943323028,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 11899
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3516063793314061,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 11900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6602024987627924,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 11901
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.172507237154835,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 11902
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.551861134780969,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 11903
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30552166678003584,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 11904
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7596416617420021,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 11905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.269896886862548,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 11906
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.283885952581461,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 11907
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3226565257937961,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 11908
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2967574744228847,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 11909
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.745407494431643,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 11910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4378034977881295,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 11911
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.492602625118903,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 11912
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1525458207196526,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 11913
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.927515796648077,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 11914
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4747357244866033,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 11915
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6076538363540658,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 11916
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6619033718061863,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 11917
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.341188073341819,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 11918
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7772360135628587,
+      "learning_rate": 2e-05,
+      "loss": 0.0707,
+      "step": 11919
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.516778243246212,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 11920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0888034134736324,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 11921
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.035407607030778,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 11922
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8829592354956647,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 11923
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8364481213153312,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 11924
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6339752583679155,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 11925
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2127273130195895,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 11926
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7635037668402611,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 11927
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6269291668173518,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 11928
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1391808760974644,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 11929
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5675328643371333,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 11930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7707779221186633,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 11931
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1545604527515752,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 11932
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6810724245720391,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 11933
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8690271539231837,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 11934
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9926246313079556,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 11935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9959944310273054,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 11936
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1443549531423514,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 11937
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7571088950987725,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 11938
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6674647297988384,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 11939
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1637778222090491,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 11940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33725089222370774,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 11941
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.167259576671205,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 11942
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9311881576075921,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 11943
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.414845953662352,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 11944
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8664894235568312,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 11945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2430329234918045,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 11946
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7476184305254154,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 11947
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7318995224790423,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 11948
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.424368380847066,
+      "learning_rate": 2e-05,
+      "loss": 0.1553,
+      "step": 11949
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.610930200628695,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 11950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.60927070353798,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 11951
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2435792649019934,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 11952
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4451100811038797,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 11953
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4437985300602203,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 11954
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.550118937243064,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 11955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.504630412348791,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 11956
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.48498329298502824,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 11957
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.925465710770743,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 11958
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.392820862079434,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 11959
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15040442168128937,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 11960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.01274968369519,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 11961
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7849236468272394,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 11962
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.18892172387974984,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 11963
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5062059134841748,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 11964
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2131641625475251,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 11965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2312504636270154,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 11966
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9965020226922836,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 11967
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1656160165067275,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 11968
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0174318526970283,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 11969
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4929210198087235,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 11970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5564684755496043,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 11971
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6252530777667473,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 11972
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.360894461125327,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 11973
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26792385570487853,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 11974
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8580924346975118,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 11975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8999700443516454,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 11976
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.248742085713498,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 11977
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.370293119146366,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 11978
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.816709799665499,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 11979
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4559127973862789,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 11980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9953669271154919,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 11981
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5200665077917979,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 11982
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8610844807910019,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 11983
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4107136562546054,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 11984
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7769535177852199,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 11985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.459961213131516,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 11986
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4756755254592404,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 11987
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4678016259678737,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 11988
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.45640417276635,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 11989
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6978371768341765,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 11990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.890703849748058,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 11991
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28551735940708195,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 11992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.942225452413655,
+      "learning_rate": 2e-05,
+      "loss": 0.1661,
+      "step": 11993
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5310012459628406,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 11994
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0429757793061052,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 11995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.113505409748201,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 11996
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6003901959512483,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 11997
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.414535935566314,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 11998
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1747447162646545,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 11999
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4856028851318315,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 12000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7073901116310024,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 12001
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0335477603936993,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 12002
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.432345197089919,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 12003
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.916291902076257,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 12004
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5653599314815396,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 12005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.92458456092353,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 12006
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3823331244784624,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 12007
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0506423801706946,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 12008
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9015140434478836,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 12009
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8600749773890268,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 12010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.425578230944028,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 12011
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6377225194415526,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 12012
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9344275940383886,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 12013
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40339890792610916,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 12014
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0655410772637794,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 12015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.184068003898756,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 12016
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6806744392731489,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 12017
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4419402546700345,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 12018
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.005372492580595,
+      "learning_rate": 2e-05,
+      "loss": 0.0819,
+      "step": 12019
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.157925242753383,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 12020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.570493001890126,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 12021
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.581098957991214,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 12022
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.638130861016916,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 12023
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7056417050267805,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 12024
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.188349204760035,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 12025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5532296862917326,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 12026
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9427793627091033,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 12027
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9131622065577907,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 12028
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.91791175400213,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 12029
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8340968868368337,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 12030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2816879688138845,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 12031
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.608605876949151,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 12032
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.619070986647625,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 12033
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1369806098015869,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 12034
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3674858857564949,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 12035
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9235110232651059,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 12036
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0289115144154017,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 12037
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.223872436961087,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 12038
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3973601896062662,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 12039
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2093602968527306,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 12040
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6562321499996258,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 12041
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5177517069777755,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 12042
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.10425445215348933,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 12043
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4214774660634891,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 12044
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.912491445584582,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 12045
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.980321878036683,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 12046
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8263148287546833,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 12047
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.20121738696773414,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 12048
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.516355515080806,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 12049
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7358024153181761,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 12050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09821916940006645,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 12051
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4961260236920334,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 12052
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0977623504729266,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 12053
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4273803104923544,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 12054
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.737699870033635,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 12055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7875433024648166,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 12056
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6176709563176176,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 12057
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1517307634013427,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 12058
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7733888433782066,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 12059
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0293014570373558,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 12060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7959953389534515,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 12061
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9562113772801537,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 12062
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.966779581551062,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 12063
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.18415911853937686,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 12064
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5240398186886159,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 12065
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8856951144831072,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 12066
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9793501183331448,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 12067
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.067904973724525,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 12068
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7987296442996847,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 12069
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.833041138635421,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 12070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6467833628098876,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 12071
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.618086293312999,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 12072
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5651322889631265,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 12073
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0410835194833346,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 12074
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6520014148554822,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 12075
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9471399999441037,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 12076
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8276663963034276,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 12077
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.649675637033269,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 12078
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0119492656594087,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 12079
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2721159460737559,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 12080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.809726815590954,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 12081
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0677168787500197,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 12082
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2989118199438807,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 12083
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6235220129872132,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 12084
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1859412776518647,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 12085
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.261622894708662,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 12086
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.571880104949053,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 12087
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.177761552032129,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 12088
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9163584118622929,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 12089
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5953864129977897,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 12090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2513142032565487,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 12091
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0735401021239146,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 12092
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.74900695904253,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 12093
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2121897192264513,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 12094
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.087096486770041,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 12095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4582842228204274,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 12096
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6745370367401704,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 12097
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5556501159234384,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 12098
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0841534589765613,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 12099
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3292214748040299,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 12100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5656660834298005,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 12101
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.11354698891196,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 12102
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9923213539782432,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 12103
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7148720157225489,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 12104
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7105132670039949,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 12105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3432260867030892,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 12106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3858501235028927,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 12107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0273528008007924,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 12108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5436964030542715,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 12109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3759575890558884,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 12110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7422509818785287,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 12111
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5853647611782785,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 12112
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2942023265250513,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 12113
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3816761355634495,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 12114
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9911070520241141,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 12115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.007051698227016,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 12116
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7183182705128228,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 12117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2305547961524987,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 12118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9979319609320825,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 12119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5341360181861785,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 12120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9260491882289954,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 12121
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.346171336331302,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 12122
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.226275929662258,
+      "learning_rate": 2e-05,
+      "loss": 0.0963,
+      "step": 12123
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9211927614142783,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 12124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0388333634437161,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 12125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4332006263659145,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 12126
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6910971473232582,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 12127
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2963220878004917,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 12128
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9225885894541132,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 12129
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8314569129699132,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 12130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5689136908845291,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 12131
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8155051556230708,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 12132
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2588302096558728,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 12133
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7944985652116,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 12134
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5834521282537446,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 12135
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9418390316034483,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 12136
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0808334606663716,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 12137
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2767338157485757,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 12138
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0263240380476937,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 12139
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3774517914908926,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 12140
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4838397215379655,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 12141
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6357852577118839,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 12142
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.763707465003486,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 12143
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.120069505883671,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 12144
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.898479286571256,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 12145
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8818229512517346,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 12146
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4581037976505309,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 12147
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.530982489746313,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 12148
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1970088614961993,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 12149
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.931676580027167,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 12150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.859696368486543,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 12151
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7216315124965194,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 12152
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.42585524519753337,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 12153
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5085772220569478,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 12154
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.321998828471901,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 12155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7299764677722712,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 12156
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0792829686154355,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 12157
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3527897289688915,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 12158
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.900486541736438,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 12159
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7030016684980385,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 12160
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.881953569822565,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 12161
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.073976926489899,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 12162
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3209006851623277,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 12163
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3347001173262152,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 12164
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3675394603408021,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 12165
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.437341930157272,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 12166
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0478513349691039,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 12167
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3334395942465922,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 12168
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.15093907221901,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 12169
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3342276891651716,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 12170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2464818251967786,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 12171
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6965461008809073,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 12172
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2825162325863673,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 12173
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5800414389601785,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 12174
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5810450364513964,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 12175
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7075004666399356,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 12176
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8206724557860914,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 12177
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8002556884150247,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 12178
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5129412456684657,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 12179
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8110729302894352,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 12180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8312399239754021,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 12181
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5949270496147414,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 12182
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.094727966413694,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 12183
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.645289618026451,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 12184
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1232007791962053,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 12185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.936923031248775,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 12186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7717597183858833,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 12187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5401327371831903,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 12188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3617503354316733,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 12189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.69641602202696,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 12190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5991094039361042,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 12191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.20029844093380184,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 12192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.935420982046692,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 12193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.476672630388448,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 12194
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.734231135269731,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 12195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8495795983694094,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 12196
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5157778675712175,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 12197
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2716934805643327,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 12198
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9012710416689034,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 12199
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.623908263048965,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 12200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9451603129199655,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 12201
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2167378467919279,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 12202
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.133815873491944,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 12203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5555425956989347,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 12204
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6117381466729318,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 12205
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.517858994163443,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 12206
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8273840414435865,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 12207
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1483070662769645,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 12208
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9445380082401893,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 12209
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.813979893002008,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 12210
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0552740680235715,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 12211
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8236682094220242,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 12212
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4915588812361575,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 12213
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.975703238067829,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 12214
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3724443840365743,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 12215
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6239671359048944,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 12216
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.053188727034912,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 12217
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4784087894597457,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 12218
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4645414727180925,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 12219
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5108172797293677,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 12220
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.958398986032631,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 12221
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.230990635232763,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 12222
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2476973305318615,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 12223
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5568275539338425,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 12224
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6116461124873211,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 12225
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1626305701648674,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 12226
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.46733400733253394,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 12227
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0666396895116532,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 12228
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1711666763940451,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 12229
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.898163059047707,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 12230
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2953686525991257,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 12231
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.299375995013584,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 12232
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.524215186744017,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 12233
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9429223264049624,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 12234
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3329502191083857,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 12235
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7779668700797195,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 12236
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.919518739863716,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 12237
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.686654338163369,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 12238
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6384376112904966,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 12239
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8036675151065562,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 12240
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2783224942742963,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 12241
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5488601349425104,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 12242
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7087587951601866,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 12243
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.066322574895892,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 12244
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.49304610248542025,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 12245
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2634450246024027,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 12246
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5313719876024414,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 12247
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1058166692880538,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 12248
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9286674224574767,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 12249
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.577539769803373,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 12250
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.32989845798907,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 12251
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0499429868932495,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 12252
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2525567947617448,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 12253
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4383442099827812,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 12254
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.24809664129400025,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 12255
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5163800433339703,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 12256
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6739418486672686,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 12257
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.42836523346666505,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 12258
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.242246195061456,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 12259
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3619301204503027,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 12260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8003715769789738,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 12261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8413424936152057,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 12262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7562319282878045,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 12263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2385691599873037,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 12264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0357173910235375,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 12265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7114828344755458,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 12266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4813299171011542,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 12267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.49537056302805976,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 12268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8551006784408153,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 12269
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.675877352717653,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 12270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2194705202669354,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 12271
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8926625952619486,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 12272
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.421910801395484,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 12273
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.558942325521992,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 12274
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9312075292367816,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 12275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0104025744736855,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 12276
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2365830188730484,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 12277
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3048881579168166,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 12278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8146307686855077,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 12279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6391718362858619,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 12280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.662740440418146,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 12281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2465844219706315,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 12282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7845232272269356,
+      "learning_rate": 2e-05,
+      "loss": 0.1082,
+      "step": 12283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6737882783342153,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 12284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.138072921979545,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 12285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0467656230094953,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 12286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15238195942725746,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 12287
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.910896770296779,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 12288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.022791436140815,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 12289
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7283634712114694,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 12290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.190164088777215,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 12291
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6901065334622858,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 12292
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4853222846556786,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 12293
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.607412861092982,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 12294
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3815035867706076,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 12295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7917470305627727,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 12296
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4432856188604017,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 12297
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4245193319684764,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 12298
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6856748813013853,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 12299
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4749731158376576,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 12300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1195607087651196,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 12301
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.142972897627457,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 12302
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9854172500123817,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 12303
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.569849581832434,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 12304
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4911023733094952,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 12305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9164498932601481,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 12306
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1391443021728154,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 12307
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4070792171747823,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 12308
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.308881175800698,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 12309
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.10525771060735663,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 12310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.256828539277468,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 12311
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8005933587688173,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 12312
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2637984670821412,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 12313
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09322412841740582,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 12314
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0201910814183737,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 12315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.821992769977581,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 12316
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3499779145381043,
+      "learning_rate": 2e-05,
+      "loss": 0.0942,
+      "step": 12317
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.774098354798217,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 12318
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.340046337742793,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 12319
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1492713003830266,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 12320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4225219428554764,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 12321
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5623941238464507,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 12322
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8323447929113268,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 12323
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.847283576693383,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 12324
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5324352892955369,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 12325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.876203278043067,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 12326
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5346994329800872,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 12327
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7436221662022672,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 12328
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5015738301732404,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 12329
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0783235842249996,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 12330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5899803054076385,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 12331
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4202603924765365,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 12332
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.316480640154573,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 12333
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.217636235269777,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 12334
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8904834434045363,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 12335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.143834187779306,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 12336
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37461931058339293,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 12337
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31765085492522677,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 12338
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.183113370822087,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 12339
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.349565519070255,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 12340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9311339402042932,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 12341
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.545770269925798,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 12342
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6899712104596267,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 12343
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3648263971595025,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 12344
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3371321038496708,
+      "learning_rate": 2e-05,
+      "loss": 0.1045,
+      "step": 12345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3883801890309146,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 12346
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.399746913281218,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 12347
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8110291662705942,
+      "learning_rate": 2e-05,
+      "loss": 0.0763,
+      "step": 12348
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.120170828036593,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 12349
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.0353750961649215,
+      "learning_rate": 2e-05,
+      "loss": 0.0906,
+      "step": 12350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9522153032844014,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 12351
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8380423958489065,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 12352
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5577489159512383,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 12353
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4725786183585257,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 12354
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.660732537270379,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 12355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7225932760686953,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 12356
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7622744070920748,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 12357
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2125640385729803,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 12358
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6969183422756566,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 12359
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0747479054478717,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 12360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7860789512493072,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 12361
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3077326888111087,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 12362
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.193997975502662,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 12363
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.156896925707615,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 12364
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.178944043872088,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 12365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8518195857876331,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 12366
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8444301837666035,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 12367
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8416836116281825,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 12368
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2140741394349097,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 12369
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9105416061974188,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 12370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8634123102146107,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 12371
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4122526272388714,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 12372
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7507810893094875,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 12373
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.503964427920797,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 12374
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9591141692926843,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 12375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9163565453896143,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 12376
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4349140497418227,
+      "learning_rate": 2e-05,
+      "loss": 0.0997,
+      "step": 12377
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6377876087401484,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 12378
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0180972600053413,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 12379
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0676429667738376,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 12380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.846724874265376,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 12381
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7142033761664957,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 12382
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7133262367447906,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 12383
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1380637689633972,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 12384
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.919540034457164,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 12385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08823503459783856,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 12386
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2286229287474404,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 12387
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8425293676198898,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 12388
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.388049022204624,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 12389
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8272366472098107,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 12390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2672162079585906,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 12391
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.779376481203577,
+      "learning_rate": 2e-05,
+      "loss": 0.0755,
+      "step": 12392
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.556667428162829,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 12393
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7268448557880345,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 12394
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8110253082541026,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 12395
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5042884323630803,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 12396
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1839981143749532,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 12397
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5438921133258379,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 12398
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7953425613447797,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 12399
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3546570533758804,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 12400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0534376863102053,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 12401
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7632285780288539,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 12402
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2106611859492322,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 12403
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.055234545075805,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 12404
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6863543783669623,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 12405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14881235399563486,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 12406
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2540891044602709,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 12407
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5770701291511238,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 12408
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.437523572758862,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 12409
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7014380669637831,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 12410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6125473565732995,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 12411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.745004571527129,
+      "learning_rate": 2e-05,
+      "loss": 0.1033,
+      "step": 12412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.132941257481795,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 12413
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7871165518965646,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 12414
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3894762696148822,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 12415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.31944353288951,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 12416
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41057735805423146,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 12417
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7601753444174929,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 12418
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.296926255569861,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 12419
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.807992921891325,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 12420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1552930944710027,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 12421
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4441182744363126,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 12422
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.22802712319994436,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 12423
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.833540478782236,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 12424
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.380518330243369,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 12425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.893242052761953,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 12426
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4521723335011,
+      "learning_rate": 2e-05,
+      "loss": 0.1188,
+      "step": 12427
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.318884104837648,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 12428
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.10234838194957632,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 12429
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.844179326785002,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 12430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6844339069966847,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 12431
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3764107143101485,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 12432
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9146047760543229,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 12433
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.244655088679009,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 12434
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8304992647187779,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 12435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.220946368117082,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 12436
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6275865070656546,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 12437
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8297563355477819,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 12438
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4037063057163435,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 12439
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.978481763152667,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 12440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1261501565553207,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 12441
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37824102387174935,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 12442
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.011021554404157,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 12443
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.027254841348877,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 12444
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2124495980076846,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 12445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.483102745807824,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 12446
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5863279416390177,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 12447
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2104130866592832,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 12448
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4967399203626997,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 12449
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.674541657599301,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 12450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.866035602324976,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 12451
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1155524393201097,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 12452
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.567703453404946,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 12453
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8505125006042455,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 12454
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2468838473892867,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 12455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.33375698848437,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 12456
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5546244699554386,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 12457
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.961475376847381,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 12458
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7879967456453834,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 12459
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8857680614523857,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 12460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2393631886560497,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 12461
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3321917314340661,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 12462
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1311058510277974,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 12463
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4817719775185976,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 12464
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.50598713967353,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 12465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4576025269892736,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 12466
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5113272588968774,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 12467
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4562023013800253,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 12468
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1733751486151018,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 12469
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08197911058728546,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 12470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0692821471737886,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 12471
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.633263795473734,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 12472
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.952988764779881,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 12473
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.495868702565168,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 12474
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5038425479111552,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 12475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.272410666439073,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 12476
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7978234414489926,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 12477
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.457287312614677,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 12478
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1649150926322045,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 12479
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.478452169651554,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 12480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3323125877197373,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 12481
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.184612628781603,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 12482
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0697570807571872,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 12483
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.880824895841811,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 12484
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12816297939346308,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 12485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4470423064094209,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 12486
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7869634753383228,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 12487
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9160558515100456,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 12488
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.386962998062342,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 12489
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4973671376892637,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 12490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9028669549635613,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 12491
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.761961114625378,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 12492
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.136030626813221,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 12493
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1688946030707763,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 12494
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.816308473537893,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 12495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.084701800714047,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 12496
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.768769250023965,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 12497
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9280115295816144,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 12498
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5854243001971884,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 12499
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7742367862636846,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 12500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9235354195717743,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 12501
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0517217895308666,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 12502
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6571082452209434,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 12503
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4772584340476183,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 12504
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.49981551939364843,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 12505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.916692740497981,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 12506
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1618546529910223,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 12507
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2782759691536625,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 12508
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4517008868423993,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 12509
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.35445889403315034,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 12510
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.679119211678529,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 12511
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5460739280114026,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 12512
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14984674696366618,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 12513
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9255446119515057,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 12514
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3950025583146037,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 12515
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9303062157500064,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 12516
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09092060314750147,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 12517
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.45277364303908,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 12518
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1637414579149683,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 12519
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8595925852945507,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 12520
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.946512883288773,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 12521
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5903440995617855,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 12522
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3957579117112604,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 12523
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.602246358458744,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 12524
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8111260608343417,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 12525
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5296872007469065,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 12526
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6736574030269484,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 12527
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.175404356191095,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 12528
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9279384331113054,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 12529
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3687338666966662,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 12530
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3521195515080886,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 12531
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0803286932630114,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 12532
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4960067327909874,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 12533
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0630986827035176,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 12534
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7319526623971466,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 12535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2755747920635663,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 12536
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6233177194280226,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 12537
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.246674305066142,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 12538
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.013895057996216,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 12539
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5674050968285407,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 12540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2642731687879316,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 12541
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7791414416576738,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 12542
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8916841803032671,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 12543
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.383791086953243,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 12544
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.085901322213761,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 12545
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.34682222553351,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 12546
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7860833563196443,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 12547
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5280797409501077,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 12548
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7245179810171724,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 12549
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.444706247149442,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 12550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1032480101041224,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 12551
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4332348806345117,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 12552
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.539533647693236,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 12553
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2189548026914856,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 12554
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.425225470925514,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 12555
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2035288647172631,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 12556
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 23.320254920364427,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 12557
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.148773152126623,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 12558
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7331762209014883,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 12559
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13617928203018836,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 12560
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.23557616632128903,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 12561
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7436028888010154,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 12562
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7680960723570975,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 12563
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4118791452352641,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 12564
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.80858982806112,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 12565
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.034600047266996,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 12566
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9869629884600054,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 12567
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2722474136431194,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 12568
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.183031559240237,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 12569
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.718174231064317,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 12570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.005454017813901,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 12571
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.726862150411607,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 12572
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7120448281502593,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 12573
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8340831075951263,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 12574
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7686265396865288,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 12575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.670777932419458,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 12576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.32314828536192,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 12577
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6494430440069683,
+      "learning_rate": 2e-05,
+      "loss": 0.0773,
+      "step": 12578
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2361178588444892,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 12579
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0409982189562657,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 12580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8965383608983254,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 12581
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5720590017326286,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 12582
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8259098007054809,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 12583
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5209650470241427,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 12584
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7785643765678859,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 12585
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.257156853547683,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 12586
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.884014584818957,
+      "learning_rate": 2e-05,
+      "loss": 0.0991,
+      "step": 12587
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5638178117656418,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 12588
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8946827023977137,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 12589
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8174204699877174,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 12590
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.912070970471066,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 12591
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2265859710974616,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 12592
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.726231193688644,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 12593
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4287959586533807,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 12594
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2913637018789337,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 12595
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4599473687100972,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 12596
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.059629697201297,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 12597
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6036830198671144,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 12598
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.81022356942688,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 12599
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8820111480690807,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 12600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3530382864325903,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 12601
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8441609515372916,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 12602
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.084850728335697,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 12603
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.21738237746484,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 12604
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1803269326981713,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 12605
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3891021019146728,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 12606
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7450861493166844,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 12607
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6112943888513809,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 12608
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.86584437117294,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 12609
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.936946851463007,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 12610
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9399107771998225,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 12611
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1559119374711966,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 12612
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7004604928123288,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 12613
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6943867348743364,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 12614
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8801777239251378,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 12615
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9114772230942119,
+      "learning_rate": 2e-05,
+      "loss": 0.0932,
+      "step": 12616
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.674960579982754,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 12617
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7218444018405903,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 12618
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1423161132875574,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 12619
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.171689471546896,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 12620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2257288632889998,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 12621
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.306126971132067,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 12622
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2215982557886358,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 12623
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0519557748292607,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 12624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6612262068219429,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 12625
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.328668825267392,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 12626
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0533406452838603,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 12627
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9981945949028892,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 12628
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8176089663156794,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 12629
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0740664438770864,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 12630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.524918994940909,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 12631
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0741372532468523,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 12632
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9941358584906146,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 12633
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6671886727900906,
+      "learning_rate": 2e-05,
+      "loss": 0.1083,
+      "step": 12634
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4150156651654218,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 12635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1441858009920747,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 12636
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0958240541691318,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 12637
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3949196330943368,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 12638
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2990389809284328,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 12639
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6670504684405938,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 12640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1790897800374376,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 12641
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9042784522138831,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 12642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4072076291909226,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 12643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1133937845740824,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 12644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7051178358419823,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 12645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9565248356220576,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 12646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8028136594454192,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 12647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7507179320883177,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 12648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.290475558913525,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 12649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5100207193481545,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 12650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0935165502870245,
+      "learning_rate": 2e-05,
+      "loss": 0.0867,
+      "step": 12651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2962579753045493,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 12652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5848672196279783,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 12653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.279013075346813,
+      "learning_rate": 2e-05,
+      "loss": 0.1492,
+      "step": 12654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4564962365579532,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 12655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30318996234759693,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 12656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.55479099635659,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 12657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2340288280415088,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 12658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6441814567079869,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 12659
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8839696541579048,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 12660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.211957833900567,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 12661
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3992689846782655,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 12662
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3950886614116906,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 12663
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3274054429309798,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 12664
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5916828972736523,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 12665
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8185175194353815,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 12666
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6059620046147147,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 12667
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4724735663603937,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 12668
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3822480251127731,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 12669
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.117760950108151,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 12670
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9875389296668635,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 12671
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.984130377857378,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 12672
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31479704157803623,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 12673
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5538296580660008,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 12674
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.390720165104589,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 12675
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0853652454887333,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 12676
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4034507441315167,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 12677
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6202659607350114,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 12678
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30132373588041755,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 12679
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.27179986474568535,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 12680
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.057327105670392,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 12681
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2386653231097413,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 12682
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2723126856924507,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 12683
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6553050231143226,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 12684
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25937884798987276,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 12685
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7267763859759915,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 12686
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8058969647660064,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 12687
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0974466513941032,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 12688
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.046089243366143,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 12689
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.164267299615126,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 12690
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3463965320729621,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 12691
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9510740194566257,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 12692
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5391505647046648,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 12693
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9128608630204975,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 12694
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9952009800306628,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 12695
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41455221906037465,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 12696
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.356956962172836,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 12697
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1909524021902962,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 12698
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.19413572570022403,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 12699
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2301641662490241,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 12700
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3679992954672413,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 12701
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8657386980544539,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 12702
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7674962745103908,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 12703
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6775986593278235,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 12704
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2947057483224034,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 12705
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3315768541370874,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 12706
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5774243465015079,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 12707
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5298769017002285,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 12708
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.158098701800182,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 12709
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4807088420344203,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 12710
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7099097661978595,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 12711
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.419164148456507,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 12712
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5759248409939066,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 12713
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5868126754654843,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 12714
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0657626155470803,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 12715
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5389044517948074,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 12716
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.075482313586219,
+      "learning_rate": 2e-05,
+      "loss": 0.0923,
+      "step": 12717
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1325097701006415,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 12718
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.461674238069137,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 12719
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7752306410393848,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 12720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14659510559897856,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 12721
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16463123245215994,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 12722
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4126517608820195,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 12723
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7459864514273673,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 12724
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0636541605359342,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 12725
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1306612305664192,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 12726
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7048418683193851,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 12727
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3092459348739673,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 12728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5241655723558333,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 12729
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.192703841199564,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 12730
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.714073728989001,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 12731
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.194142528748395,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 12732
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.22585643717578094,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 12733
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8934667369135951,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 12734
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.009607417036763,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 12735
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3299322090601517,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 12736
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.140470451493693,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 12737
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6422509991449252,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 12738
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0460076277917874,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 12739
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4847947179788739,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 12740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.0476575411507545,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 12741
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7257155330332252,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 12742
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4322403592254473,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 12743
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3973017934501712,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 12744
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8204507257170417,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 12745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.416433388414465,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 12746
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8180247325662697,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 12747
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14175730723303873,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 12748
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.50835096795046,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 12749
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1834368436120895,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 12750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7866379914641213,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 12751
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7436600783027887,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 12752
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7468175921862314,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 12753
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3337356483542466,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 12754
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4819069365053537,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 12755
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9777840318950013,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 12756
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32877114799055307,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 12757
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0182936014664614,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 12758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.884457324809214,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 12759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2398269402545457,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 12760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2429795411427325,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 12761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6247538600253006,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 12762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5835157423185455,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 12763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.575589221834516,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 12764
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.895146572895662,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 12765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6996070806400712,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 12766
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7239607712592555,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 12767
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.492098900526378,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 12768
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9456799018483342,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 12769
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5151973209914362,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 12770
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9658363822966192,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 12771
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3908222787203268,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 12772
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.18827527420564447,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 12773
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.823776695932455,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 12774
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.12251675032055187,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 12775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1533404598240714,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 12776
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.315581700208235,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 12777
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7520715001056812,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 12778
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0023425211997545,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 12779
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4185609027695403,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 12780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.510714878497318,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 12781
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2286038861862956,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 12782
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.384318091837061,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 12783
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6369619751822606,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 12784
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.48791077864645277,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 12785
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6230441574275911,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 12786
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5854753093258607,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 12787
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6921087457016925,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 12788
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.300572206999223,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 12789
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9931273927173635,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 12790
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3568158230732734,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 12791
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.36976036646730914,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 12792
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0516263265071997,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 12793
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0493790546736608,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 12794
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.38611908860774113,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 12795
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6385364447209552,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 12796
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1997554252966003,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 12797
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.475538756691114,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 12798
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.623661754305747,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 12799
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4114160939108813,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 12800
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.567373384013273,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 12801
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.671491182655778,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 12802
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0074920736337103,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 12803
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4246813013945106,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 12804
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6598721608952711,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 12805
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.469134530167193,
+      "learning_rate": 2e-05,
+      "loss": 0.0757,
+      "step": 12806
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6274971903779859,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 12807
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3481642622526082,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 12808
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.119194065246394,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 12809
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3165604486427232,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 12810
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8273181943755819,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 12811
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9614968881422797,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 12812
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5917412309340653,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 12813
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.229375621473666,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 12814
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6944169874346542,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 12815
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.28563361039248,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 12816
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.059007172448374,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 12817
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.23363485589128,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 12818
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7409693132574104,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 12819
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16126230411937525,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 12820
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.745873354116994,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 12821
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.33707112507507225,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 12822
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.653810412444889,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 12823
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8067452790161234,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 12824
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.26476621337138,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 12825
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0343839677416713,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 12826
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5986909011046446,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 12827
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28445544415046886,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 12828
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9785831059568193,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 12829
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2603054102616178,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 12830
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2216759795232497,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 12831
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1318059431542324,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 12832
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8529895409438112,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 12833
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6566596334347758,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 12834
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6680683485587561,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 12835
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3933546651174686,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 12836
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2748014677844908,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 12837
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9450899287733313,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 12838
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.176396883806698,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 12839
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6937524323956243,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 12840
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.09404467291692,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 12841
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7313680162319677,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 12842
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1590984929868544,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 12843
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5241031115542365,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 12844
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8938282698924666,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 12845
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7894269413792605,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 12846
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9525165233747264,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 12847
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8987294925053103,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 12848
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8100467839682192,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 12849
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.411234577014091,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 12850
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4324216551465696,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 12851
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6600706056195228,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 12852
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27274197266529154,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 12853
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.458157510603491,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 12854
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.79246949919943,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 12855
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1400376544599627,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 12856
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4949220652832051,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 12857
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6520478725439888,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 12858
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0386266717615626,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 12859
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3633001153366338,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 12860
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4142473637416546,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 12861
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8750934550922339,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 12862
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.643897742794344,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 12863
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0266110781315727,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 12864
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4542682211463795,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 12865
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4969074635977286,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 12866
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08356664420898949,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 12867
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.16278169561938,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 12868
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.333769337472348,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 12869
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6772210543513518,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 12870
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44221488342147497,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 12871
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07432631490470848,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 12872
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5008532872512087,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 12873
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8330349110540407,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 12874
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4424344636856307,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 12875
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7084308912375341,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 12876
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4243781167403915,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 12877
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1547794269314797,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 12878
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1350964377516657,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 12879
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3297249411163578,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 12880
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6281221181071543,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 12881
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.451809523539214,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 12882
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4625775397626084,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 12883
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.03927805106063,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 12884
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.635051378731916,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 12885
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.34191911963635774,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 12886
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7266346065057493,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 12887
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.05647548698932058,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 12888
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7452436867222256,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 12889
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7761073197169335,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 12890
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9610784783884191,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 12891
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.796103946707885,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 12892
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7199542130094616,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 12893
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1670898396643063,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 12894
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2067950860249352,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 12895
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.22253155981416461,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 12896
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07098706173540673,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 12897
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1810021559253883,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 12898
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4120751678588651,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 12899
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.517171719407142,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 12900
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2840935874176864,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 12901
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3782114954610974,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 12902
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2867658635672807,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 12903
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.38922905401788943,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 12904
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.307280358142377,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 12905
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7231392148664089,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 12906
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1621285220409198,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 12907
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9819108304989861,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 12908
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.02030779372609,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 12909
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.712751275643793,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 12910
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1762882292937131,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 12911
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9881342925376506,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 12912
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.46640208640228975,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 12913
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.645121340933939,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 12914
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1802821397500014,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 12915
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.504411714806885,
+      "learning_rate": 2e-05,
+      "loss": 0.0828,
+      "step": 12916
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7464765195967216,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 12917
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.639238531962359,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 12918
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9365324865640505,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 12919
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.095145031226664,
+      "learning_rate": 2e-05,
+      "loss": 0.0897,
+      "step": 12920
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4627167989339793,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 12921
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1845346871575124,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 12922
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1504363348164084,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 12923
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4855164644963733,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 12924
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4620275663602802,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 12925
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6758972696899046,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 12926
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2411760254758892,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 12927
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6705578758492294,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 12928
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2403249414656954,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 12929
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7672059998932997,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 12930
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6701083252089155,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 12931
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9484508278397794,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 12932
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2631104886183037,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 12933
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0417269625696925,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 12934
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5639891585342576,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 12935
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7309025550093062,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 12936
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.342768909793503,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 12937
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2275220191518956,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 12938
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7577269201032908,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 12939
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1351861844858946,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 12940
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5731065479722577,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 12941
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1847788099692895,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 12942
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8571332649988509,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 12943
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5369734158533278,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 12944
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.248530859673141,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 12945
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6462088282915693,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 12946
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.473096139307099,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 12947
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0177532491985828,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 12948
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4322183023794859,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 12949
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35509393654714083,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 12950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8246922664642145,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 12951
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6846520868890696,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 12952
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5214434754110484,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 12953
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.122526945023838,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 12954
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.74561638520318,
+      "learning_rate": 2e-05,
+      "loss": 0.1104,
+      "step": 12955
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4521792215497558,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 12956
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1542590759594722,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 12957
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8514312826410926,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 12958
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.207852821614685,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 12959
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2802600638868493,
+      "learning_rate": 2e-05,
+      "loss": 0.1169,
+      "step": 12960
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3325097843035606,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 12961
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0945197794989854,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 12962
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6641776767567245,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 12963
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2468842872852499,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 12964
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5893833217169466,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 12965
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8389344124602944,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 12966
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.231997264134478,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 12967
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.557553984220499,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 12968
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4003315039249473,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 12969
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.277067452317468,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 12970
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3538477546571557,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 12971
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7388548216669824,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 12972
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.471825627389695,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 12973
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9192409384443992,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 12974
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6124474999251496,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 12975
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.727272129515038,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 12976
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6354512663838009,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 12977
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.291784757403622,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 12978
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.191558127806671,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 12979
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3128462780255968,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 12980
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.533713848832842,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 12981
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6114290015994706,
+      "learning_rate": 2e-05,
+      "loss": 0.0754,
+      "step": 12982
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5509455629385123,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 12983
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4872127512058397,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 12984
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2650884267938456,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 12985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2509475720950705,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 12986
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8108287867515191,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 12987
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.31170256030964,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 12988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.285205390597435,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 12989
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3849305791768187,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 12990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5049955732793641,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 12991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4142858845158814,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 12992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1918995320716474,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 12993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5811781855623646,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 12994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1805120958793502,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 12995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4445818729721505,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 12996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3243869460873638,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 12997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3597334856718979,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 12998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.20305651647347195,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 12999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6418760922951027,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 13000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7433117715779824,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 13001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9907366469698267,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 13002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0750883709269807,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 13003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8454396539173559,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 13004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5404913606385946,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 13005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6929482329039637,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 13006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9574924922911194,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 13007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7451986113489202,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 13008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9585747863833762,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 13009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0954136865651969,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 13010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4782269211853887,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 13011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.10268794558957196,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 13012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15561257519588023,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 13013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.669873117549351,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 13014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6756443921463601,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 13015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2020481789846111,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 13016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3978421842362625,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 13017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3731467589163666,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 13018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7597232182650553,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 13019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.769093983381577,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 13020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2114184993231363,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 13021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.484422512185311,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 13022
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5330274243519151,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 13023
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8664267039465638,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 13024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32941895752586386,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 13025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6357805405462655,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 13026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5661719853364087,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 13027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5472242431835985,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 13028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9630843861742857,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 13029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.33441606028855386,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 13030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.357959752322545,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 13031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.444949382109802,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 13032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.456055669956912,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 13033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14810687185571364,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 13034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7128201930181812,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 13035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0694372090235222,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 13036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0257997695634018,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 13037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1937002141666788,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 13038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7678465240479329,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 13039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7663681356847245,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 13040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0791894384998477,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 13041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.117436852822608,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 13042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4161814506894281,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 13043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35462379639710057,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 13044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4207800537292283,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 13045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7810578011956726,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 13046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6595472228482326,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 13047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.244199056503768,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 13048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.259161734111846,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 13049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1267168178245486,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 13050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3256740883236476,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 13051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8370984890887208,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 13052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.351264503835817,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 13053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6429464797546789,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 13054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.460903636519508,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 13055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7938075773126897,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 13056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5266911119616742,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 13057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7345857115036465,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 13058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2085150145046177,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 13059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3318822077419257,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 13060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.509702456634538,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 13061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3585452035453178,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 13062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42587802382152296,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 13063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.924538265768469,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 13064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41290030129750355,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 13065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6707696916076205,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 13066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29677542472136065,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 13067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9903316469523324,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 13068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.23640758187523306,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 13069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9526077806370312,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 13070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0147678415673154,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 13071
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.963566963789539,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 13072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8664982373453847,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 13073
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0121758212822891,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 13074
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.515857946379299,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 13075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1543903345267807,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 13076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.04301738019111812,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 13077
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7263641496857902,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 13078
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0924066195527318,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 13079
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3760038093503282,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 13080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.418116765768289,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 13081
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1955956017827765,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 13082
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4338390538988404,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 13083
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8351355566528331,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 13084
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.716712714168019,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 13085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9205109011880228,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 13086
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7033105393493166,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 13087
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6076948037230908,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 13088
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.460399508228078,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 13089
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.729601296031618,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 13090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.105706448829974,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 13091
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6799963378866336,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 13092
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.620675149839629,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 13093
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.72337205887567,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 13094
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4699353244073952,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 13095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0968172389632425,
+      "learning_rate": 2e-05,
+      "loss": 0.0675,
+      "step": 13096
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.021462438561691,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 13097
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.821612569186606,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 13098
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.944261415383249,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 13099
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2243811914840665,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 13100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.141750355365682,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 13101
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1707227897339716,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 13102
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5350183183885168,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 13103
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2978844532880327,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 13104
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5633258849513074,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 13105
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.152547475618945,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 13106
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9318688916369267,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 13107
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2275056681791003,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 13108
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7936430018608467,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 13109
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4765687165306058,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 13110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1808214106436814,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 13111
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5534446003701154,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 13112
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.10655808380909981,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 13113
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0954651599673404,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 13114
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1486470556778772,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 13115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0776217432076325,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 13116
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5951186642406994,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 13117
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8173391679166085,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 13118
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4868761062364657,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 13119
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9537915122514857,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 13120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.044605524979283,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 13121
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.245298135242954,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 13122
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9641076106290565,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 13123
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7146942838211696,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 13124
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.11509176902402311,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 13125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1131575130831264,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 13126
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4131092939695256,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 13127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7690631171439206,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 13128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7783100122705946,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 13129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.472948525968643,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 13130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.357914238866665,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 13131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0429893805931536,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 13132
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2664350107493558,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 13133
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.24512214959581896,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 13134
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8165864882392823,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 13135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.406337984136161,
+      "learning_rate": 2e-05,
+      "loss": 0.123,
+      "step": 13136
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.04971213951543064,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 13137
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.10052037185437836,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 13138
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.363242686335243,
+      "learning_rate": 2e-05,
+      "loss": 0.1146,
+      "step": 13139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3610568083957495,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 13140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3470081042305775,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 13141
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2445883041463755,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 13142
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0669698937377574,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 13143
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5602624512248797,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 13144
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1931625678421027,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 13145
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6102762013805787,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 13146
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7637224107494816,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 13147
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8623370016573205,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 13148
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1689741123284927,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 13149
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2130815105170675,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 13150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.488718604358136,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 13151
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5607063679838786,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 13152
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.49544894748791146,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 13153
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.62258054858677,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 13154
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4368152539080049,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 13155
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.64567732663927,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 13156
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7909686333035806,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 13157
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.960369600246026,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 13158
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.623878693643496,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 13159
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1789944362331148,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 13160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1532018020815602,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 13161
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.206896283650016,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 13162
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1674227762107126,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 13163
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1362988486653576,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 13164
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2584026402645627,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 13165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27256607433565505,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 13166
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7693500983316685,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 13167
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5099781806336094,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 13168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4839783791152068,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 13169
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5879094970073897,
+      "learning_rate": 2e-05,
+      "loss": 0.0841,
+      "step": 13170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2969698779298184,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 13171
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6526226886415349,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 13172
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7919082442183734,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 13173
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8485597494269796,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 13174
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3098001093594784,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 13175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7632137476651673,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 13176
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7171191882013286,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 13177
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5977927151997031,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 13178
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3212801304957242,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 13179
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9383776689041496,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 13180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.037417160877704,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 13181
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0675107385421334,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 13182
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26392853572201075,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 13183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31702998577033176,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 13184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7798455708196108,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 13185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.21937825889286078,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 13186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3428695848171546,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 13187
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6618179760323695,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 13188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.594793086981235,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 13189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6546000053176133,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 13190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.056323040786998194,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 13191
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.949172913441441,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 13192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.19171005583731815,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 13193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2150278227960285,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 13194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.38447069589562255,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 13195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.333306505412322,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 13196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.162126403170938,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 13197
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5465576348590053,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 13198
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8303965956167949,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 13199
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0525286018671478,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 13200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.487512585242109,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 13201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.038180105522475,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 13202
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.668993739926053,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 13203
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.300391932240749,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 13204
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7510885445351118,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 13205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2108968420278752,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 13206
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3077309112439146,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 13207
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2016013765432598,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 13208
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6774897076159956,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 13209
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5841522046959446,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 13210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.171715943806797,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 13211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4991333365150128,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 13212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.002994560481897,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 13213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7864103786845964,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 13214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3562946366906168,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 13215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2463443229128213,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 13216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.822546246050929,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 13217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6333471271684994,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 13218
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9136586288675823,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 13219
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4812127687246126,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 13220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25757470868858845,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 13221
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1294017087293546,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 13222
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3123079341048312,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 13223
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.160068033377312,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 13224
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.112496807295912,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 13225
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5611682066757028,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 13226
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.542924684298024,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 13227
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9699098107465696,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 13228
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.44347210806474,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 13229
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.568187359869548,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 13230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8204518517704792,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 13231
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5258606150963203,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 13232
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6865249785708499,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 13233
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2264099026170086,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 13234
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2715785235822357,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 13235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3031562453807595,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 13236
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6096802460489701,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 13237
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6099713155904132,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 13238
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.06433642386364,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 13239
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9465041444012314,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 13240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5753615548740261,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 13241
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6443334900021873,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 13242
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9260196809010706,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 13243
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.317706065673692,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 13244
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0139019361437416,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 13245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.809442256678293,
+      "learning_rate": 2e-05,
+      "loss": 0.0675,
+      "step": 13246
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3676204758738288,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 13247
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.184070842124955,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 13248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8403375371990798,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 13249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2984587729978223,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 13250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.735668639300589,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 13251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27348304096814763,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 13252
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9697855675024428,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 13253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7963917332718814,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 13254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5230635746531733,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 13255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.745283719801385,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 13256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7711322222139714,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 13257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9781404612149396,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 13258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5384163785169183,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 13259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6420684462778544,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 13260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2769111429308909,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 13261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.880425360527315,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 13262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8032779886539994,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 13263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.846889299355278,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 13264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0727783533712991,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 13265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.285879584704549,
+      "learning_rate": 2e-05,
+      "loss": 0.0828,
+      "step": 13266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3137818629396848,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 13267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1418960027942275,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 13268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1566519682900327,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 13269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4246289470692415,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 13270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0828121961452835,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 13271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9265532580770446,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 13272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0812873900018054,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 13273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0847321551513025,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 13274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29009277154719876,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 13275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5006499238780969,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 13276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.923033296012409,
+      "learning_rate": 2e-05,
+      "loss": 0.1166,
+      "step": 13277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0786783760843555,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 13278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8013342459869689,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 13279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.095291832398031,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 13280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.162185217149803,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 13281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4193029090919802,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 13282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.664682502781601,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 13283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.996471408234753,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 13284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.331255948918454,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 13285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.18107344150633567,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 13286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5314549163096404,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 13287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7033494573950991,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 13288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.143409986975551,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 13289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2587266810823778,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 13290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8116624146487332,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 13291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.069867006575314,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 13292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8888375526196415,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 13293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4063079398193044,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 13294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6367747367048553,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 13295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5403263643805962,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 13296
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.19068587261457007,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 13297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.273995380607852,
+      "learning_rate": 2e-05,
+      "loss": 0.1041,
+      "step": 13298
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1964645319335323,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 13299
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9195576785127049,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 13300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1667192083970657,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 13301
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7272017753193317,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 13302
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30889085157236457,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 13303
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.372543915704379,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 13304
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2122278121506347,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 13305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6996920810973727,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 13306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8202482547801415,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 13307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8100777980742345,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 13308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2509167170745783,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 13309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.084312650443611,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 13310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.45276499291697836,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 13311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.654673780000773,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 13312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2527911495407666,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 13313
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3757227818609126,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 13314
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1375489979378299,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 13315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8004819729435313,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 13316
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2390616091019073,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 13317
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9071847107332236,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 13318
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.866271010109694,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 13319
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0139589583049662,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 13320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9443494762250957,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 13321
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0006649667453071,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 13322
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8338622957524748,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 13323
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6295710608926446,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 13324
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.20203558678925115,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 13325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.655669204943723,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 13326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7456664614409276,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 13327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39417647471870315,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 13328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9813538283335705,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 13329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5429843563787486,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 13330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.820071631446302,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 13331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.308407332224646,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 13332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.292643065217609,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 13333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6790990600440763,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 13334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.008740399948377,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 13335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7745620182121734,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 13336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9122170950364366,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 13337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7863221482694747,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 13338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5303111194689286,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 13339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8198486106383862,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 13340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.371618272560618,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 13341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8490149863282982,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 13342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.532414674608291,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 13343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.199538740054398,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 13344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2875583120115803,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 13345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9204440769457274,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 13346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9822010376033999,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 13347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.564788639445572,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 13348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.770596613921927,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 13349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0066225555657873,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 13350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.890443538434362,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 13351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2360296145135061,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 13352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6509455543565654,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 13353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.22578619453269455,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 13354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.00298456184734,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 13355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.062702631388812,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 13356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.315756283640869,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 13357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.144372930745322,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 13358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.170130686187329,
+      "learning_rate": 2e-05,
+      "loss": 0.1166,
+      "step": 13359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9529372468350548,
+      "learning_rate": 2e-05,
+      "loss": 0.1065,
+      "step": 13360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27329199870690807,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 13361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4567400828079865,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 13362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.828814848984897,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 13363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27138635526223015,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 13364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.637197545308885,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 13365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.587687013649624,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 13366
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5853989161108841,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 13367
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.5622991621983715,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 13368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1586447214169646,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 13369
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.149182551557497,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 13370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2435033294823359,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 13371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9964516090380459,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 13372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5528510681551199,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 13373
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.22388610488340216,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 13374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.960975398114939,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 13375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2879257276042082,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 13376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.032862214192312,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 13377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8096639214733357,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 13378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0947666348040461,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 13379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0505518612408777,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 13380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7940373478808864,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 13381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1621058463160003,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 13382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.310229981145747,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 13383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.47103316263820205,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 13384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4032944455878678,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 13385
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3638419935490766,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 13386
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7503065113305032,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 13387
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8776330499772926,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 13388
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9894738822774753,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 13389
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2387955136218818,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 13390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6746330129893745,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 13391
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7698079497628791,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 13392
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7862925846477473,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 13393
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0401991743319563,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 13394
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0424685761177614,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 13395
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5860631426571438,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 13396
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7575874014945065,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 13397
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9712249151694716,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 13398
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8455469458972042,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 13399
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0526160915076157,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 13400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.37095165543835334,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 13401
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4761955976031131,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 13402
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.193911310850934,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 13403
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7305441811958107,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 13404
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.879511774998962,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 13405
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41373138864153197,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 13406
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0765161534757166,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 13407
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5006063348450598,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 13408
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.325698612219905,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 13409
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6371581732709461,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 13410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5707232942352745,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 13411
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9093567771003409,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 13412
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.47137022430157394,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 13413
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6297754862479312,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 13414
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3958757369117483,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 13415
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.115817808806507,
+      "learning_rate": 2e-05,
+      "loss": 0.1026,
+      "step": 13416
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1134304264566026,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 13417
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.671940317526723,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 13418
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1308733235227222,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 13419
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4797071445534682,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 13420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6768253266408144,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 13421
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2580908226681335,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 13422
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8396633220523714,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 13423
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5942140539411027,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 13424
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6094349073648881,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 13425
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.132881144772803,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 13426
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.149203687767413,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 13427
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.197699762139954,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 13428
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0792741594329902,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 13429
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8752927216491293,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 13430
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.114275404083859,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 13431
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.629711729548018,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 13432
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8710257328318429,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 13433
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.538466851417834,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 13434
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6198231638323478,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 13435
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9157319107393517,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 13436
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6955363745470369,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 13437
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.553872089450919,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 13438
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39194673085523973,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 13439
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5323942427765994,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 13440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2880445680489458,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 13441
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2207756705178654,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 13442
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0628784851928694,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 13443
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.32652235385998,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 13444
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7653639990791479,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 13445
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3097477834405171,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 13446
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.332547308778015,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 13447
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6678688323135542,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 13448
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3883868433305606,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 13449
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8623047911837246,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 13450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.477487183723438,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 13451
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0032749194717954,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 13452
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9508611488988374,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 13453
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.125374096024432,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 13454
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0418524235930144,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 13455
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5330740803934305,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 13456
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1158037293190777,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 13457
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3691882072158195,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 13458
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.961083818432737,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 13459
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7344674397179213,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 13460
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9743642532234873,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 13461
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9533994839782327,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 13462
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1660731826965984,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 13463
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.913238643177761,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 13464
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.526024386853354,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 13465
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0967737634787358,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 13466
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8155059961487952,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 13467
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.283577450235031,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 13468
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.666865321082273,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 13469
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3698752223791219,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 13470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8950782851301553,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 13471
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0448431748600497,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 13472
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.200003560382606,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 13473
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8813851557227544,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 13474
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8222572665739603,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 13475
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23833130482719606,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 13476
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.797063655587113,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 13477
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.406553075950527,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 13478
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.03804814009795572,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 13479
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8859607632928169,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 13480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7755704426232312,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 13481
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.151346242060715,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 13482
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4598472481110094,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 13483
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.074342144901485,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 13484
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1193665281451224,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 13485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9029571297989178,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 13486
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.405383839356269,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 13487
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.091881786799299,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 13488
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29450806898371074,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 13489
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3976536688880117,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 13490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.767779180776483,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 13491
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6449837685546154,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 13492
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.533616631017024,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 13493
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.763753550032686,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 13494
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6413052599038825,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 13495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1816109151025785,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 13496
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0615809897900423,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 13497
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1313036637196898,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 13498
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2660087590164759,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 13499
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3365450510621773,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 13500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.014341869262077,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 13501
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3362032482375515,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 13502
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.915931194284466,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 13503
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9261588138531263,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 13504
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.67530727130207,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 13505
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.027477125205588,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 13506
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1147949635791925,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 13507
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.668304265513864,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 13508
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0020659028828405,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 13509
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.486712429794827,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 13510
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.951725768686805,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 13511
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6991734369783287,
+      "learning_rate": 2e-05,
+      "loss": 0.0964,
+      "step": 13512
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3107675289864202,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 13513
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.221990343132678,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 13514
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8701936664643645,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 13515
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.176779718233204,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 13516
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9285449532517801,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 13517
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.078498720273535,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 13518
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17504908350203652,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 13519
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9358312126582673,
+      "learning_rate": 2e-05,
+      "loss": 0.0671,
+      "step": 13520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0908042678296037,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 13521
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2560655767597196,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 13522
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6972203221217685,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 13523
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2474400474351226,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 13524
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3627271865281598,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 13525
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.604920602302249,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 13526
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8583715996653373,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 13527
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0874834103250672,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 13528
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.703732067989379,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 13529
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.415700977593759,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 13530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2558582033591175,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 13531
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9219526807740164,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 13532
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9863640135877205,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 13533
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5866883870686332,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 13534
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.93437077486381,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 13535
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0358365684350321,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 13536
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7236371780243234,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 13537
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9996430236710352,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 13538
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.21034105180781698,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 13539
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0974480916914542,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 13540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2395889645403155,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 13541
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.485490150079858,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 13542
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.695369023269419,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 13543
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.663516890935103,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 13544
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9769028337629463,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 13545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0777817884733247,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 13546
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5689907696181411,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 13547
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.454801116853889,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 13548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.81283809533978,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 13549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23631203021265373,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 13550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.406711583638569,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 13551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8105041492645405,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 13552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2861441204063448,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 13553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4474825751489955,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 13554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7885374969538645,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 13555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9580440375194834,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 13556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3587998236651864,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 13557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0784323433924365,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 13558
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.962553171786029,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 13559
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.649871708477421,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 13560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8178117462151393,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 13561
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.22710393468051007,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 13562
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.18675716595516859,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 13563
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.331288105235734,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 13564
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.981706151775813,
+      "learning_rate": 2e-05,
+      "loss": 0.1229,
+      "step": 13565
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8384946055662434,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 13566
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.199526866626131,
+      "learning_rate": 2e-05,
+      "loss": 0.0876,
+      "step": 13567
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12618893457828204,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 13568
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26824640786520215,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 13569
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.911759435452855,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 13570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1440643747330133,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 13571
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3017916492548887,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 13572
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3907871366127063,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 13573
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0956150502320525,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 13574
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.36707503036969663,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 13575
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5535425158411342,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 13576
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1677208066348415,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 13577
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8892964274909583,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 13578
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.131281956595392,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 13579
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3901334148930723,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 13580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39998546663346424,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 13581
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.781291492715059,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 13582
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4885140785061493,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 13583
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3342588383014067,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 13584
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2073060250710723,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 13585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3337058897866727,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 13586
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0549550035110424,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 13587
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5607156865143887,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 13588
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7541660040025826,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 13589
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.476985468476687,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 13590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.261646750560863,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 13591
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9208053465464139,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 13592
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.45907768761925694,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 13593
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4518556651749184,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 13594
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3542040550743621,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 13595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3647606372911347,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 13596
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.75126922024277,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 13597
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9558410680244374,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 13598
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9702893248971611,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 13599
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1119395419467568,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 13600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7881479382768692,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 13601
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.170390566282558,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 13602
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2073448786008865,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 13603
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.48503735078944477,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 13604
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5345396401717819,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 13605
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6338337893842256,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 13606
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6217526713299703,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 13607
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9671543860886684,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 13608
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6638947667731117,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 13609
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.21835212627246522,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 13610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9851960587834288,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 13611
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8633354135498064,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 13612
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0989575886941982,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 13613
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0915696892117996,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 13614
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3658267415562393,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 13615
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6252439022994714,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 13616
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.659250528440956,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 13617
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30400053243606145,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 13618
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.354388947461271,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 13619
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5398409930323025,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 13620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5404338256103994,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 13621
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.828078636168384,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 13622
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5299765392096596,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 13623
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3412406428070764,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 13624
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6258788129211537,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 13625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7308699097653901,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 13626
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.026685192479405,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 13627
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5264012709642854,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 13628
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.111715078448347,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 13629
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3740672482683864,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 13630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1910598921790285,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 13631
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.005711457570546,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 13632
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5474231969622241,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 13633
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.010743257509901,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 13634
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.141838787697751,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 13635
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2034295054274335,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 13636
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8486568652894093,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 13637
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8302364057002944,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 13638
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.636606714199626,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 13639
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0206492056840772,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 13640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1394007168114872,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 13641
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.611386569726056,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 13642
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8033048940003762,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 13643
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.490953040302813,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 13644
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.0320858468403245,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 13645
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8738363225532765,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 13646
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6125782138166155,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 13647
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0548410104474146,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 13648
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5648796555541878,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 13649
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.556807939845398,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 13650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5613669268550617,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 13651
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1209866501786356,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 13652
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16417770087750813,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 13653
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4697462938252856,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 13654
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7700212149776475,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 13655
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1146403800174485,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 13656
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0346292752586495,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 13657
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8223763796772584,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 13658
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3458161465798395,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 13659
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.912238951143871,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 13660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4928461189149809,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 13661
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.569466237359312,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 13662
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6349263357197414,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 13663
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.884250772154899,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 13664
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7345869570797565,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 13665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.499781036289823,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 13666
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9893430351159069,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 13667
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7680339673001174,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 13668
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1939718669407668,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 13669
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5745291959903058,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 13670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.153488263232937,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 13671
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6921896295105703,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 13672
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8234085874168532,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 13673
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2597521297993511,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 13674
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1373022620234043,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 13675
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7692649253527224,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 13676
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4451342443434487,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 13677
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.975454388399248,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 13678
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9082024030785101,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 13679
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.055979400178174765,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 13680
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14679153028648395,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 13681
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8233121461119661,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 13682
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9105122733527511,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 13683
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7698140665459433,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 13684
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3525615814207275,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 13685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9225984094951291,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 13686
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5118902537919572,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 13687
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5847956112268906,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 13688
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2927789820789184,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 13689
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0853967893081666,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 13690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5918607546854529,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 13691
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.674279185355255,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 13692
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.728290918259547,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 13693
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2431775352305348,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 13694
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5547090176989036,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 13695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5008372806619903,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 13696
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.47333816329692385,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 13697
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9988696361090719,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 13698
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4401433256922953,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 13699
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9294253067736411,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 13700
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3576672446584579,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 13701
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7242219690691234,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 13702
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.263831622301691,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 13703
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0133159333965587,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 13704
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.949647374963976,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 13705
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9590237797801031,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 13706
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.956133120521716,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 13707
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9579253710505743,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 13708
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.779465701890346,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 13709
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.246441145858903,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 13710
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8377227615053775,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 13711
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38587323630584514,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 13712
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.72298250114867,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 13713
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.495894476469905,
+      "learning_rate": 2e-05,
+      "loss": 0.1176,
+      "step": 13714
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9881486829769972,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 13715
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9402192733052667,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 13716
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7016795878417436,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 13717
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3170945604523948,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 13718
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25518980575312117,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 13719
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9032638614645706,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 13720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7751535202067498,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 13721
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7782365170657757,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 13722
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6562878506961868,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 13723
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5660024321540904,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 13724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0552932130216341,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 13725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4511810566021577,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 13726
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3894753579545245,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 13727
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3153616272354316,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 13728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6611788124923108,
+      "learning_rate": 2e-05,
+      "loss": 0.0872,
+      "step": 13729
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.956616650518878,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 13730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9851757304771124,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 13731
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4089280377645481,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 13732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.62520752095241,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 13733
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5345993488321468,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 13734
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0448858807505292,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 13735
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3248918111285062,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 13736
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.096162118269509,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 13737
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8437057766097973,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 13738
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8444624118468967,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 13739
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.105622890765829,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 13740
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.423921077645788,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 13741
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.108550289019913,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 13742
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.568081791265938,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 13743
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.939977292685,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 13744
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7529727117019039,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 13745
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8692939194496769,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 13746
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.767495304085241,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 13747
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6032104201732,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 13748
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1863498388823683,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 13749
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8158562737909787,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 13750
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1713757959781417,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 13751
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7773242737136906,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 13752
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3567296814214804,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 13753
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.626478696161177,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 13754
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.472936777264315,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 13755
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5733382647377792,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 13756
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.454605130542591,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 13757
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.884822020971056,
+      "learning_rate": 2e-05,
+      "loss": 0.0963,
+      "step": 13758
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7826972526339291,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 13759
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.3549843831006765,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 13760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1061495783857074,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 13761
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25071405716809175,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 13762
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8386091161734152,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 13763
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.279423668632748,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 13764
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5719170414969457,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 13765
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9065668436853473,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 13766
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.588799425380693,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 13767
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0636458248506657,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 13768
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7755361655833047,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 13769
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.83406210480435,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 13770
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.860246493865908,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 13771
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4187549271687054,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 13772
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4536511688881677,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 13773
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0308485116563295,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 13774
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5319532998041296,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 13775
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5351899509759022,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 13776
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0008805244003964,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 13777
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6703839946401607,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 13778
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4481649757633237,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 13779
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.163949589725771,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 13780
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.11275296513323098,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 13781
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3632639398860213,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 13782
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6959585553030172,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 13783
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8736110289680943,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 13784
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2359964850971403,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 13785
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7339422289983385,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 13786
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3736929750278293,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 13787
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1461416638844046,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 13788
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.442113025928697,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 13789
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7997182493104984,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 13790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.767755309467626,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 13791
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13931649778559085,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 13792
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3423724099306624,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 13793
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5183541806433694,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 13794
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5891047768637622,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 13795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.580342990851904,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 13796
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0121547981766954,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 13797
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7383336447233602,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 13798
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6417796498899234,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 13799
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.767773552397004,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 13800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.311541792947644,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 13801
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3911668452660817,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 13802
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4178249822197022,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 13803
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3268352204801988,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 13804
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7372786813197223,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 13805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4340529888720006,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 13806
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41485629377061006,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 13807
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1206192979811165,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 13808
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4342555760117299,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 13809
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7130023189469515,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 13810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9188864159028637,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 13811
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43300402979043773,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 13812
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.42205856003257564,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 13813
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6684480299141695,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 13814
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.33368885060515924,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 13815
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.22818475358218,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 13816
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1823756934511858,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 13817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2190492455070925,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 13818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15232265155141078,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 13819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.36080042477913654,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 13820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5336361223101194,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 13821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.507703075490255,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 13822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3855159217252786,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 13823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5965101986208974,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 13824
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7378199513700131,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 13825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0880499155428693,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 13826
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3991667861115364,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 13827
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2975293997031683,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 13828
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.220899935306307,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 13829
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.131889842352732,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 13830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3892143895938152,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 13831
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.160348823644527,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 13832
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1178279344441817,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 13833
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0681672017650706,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 13834
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8359885350702558,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 13835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.34105429501107326,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 13836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6137158498626514,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 13837
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.541834143035195,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 13838
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.832151341614309,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 13839
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6430271546435244,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 13840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5868662853384796,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 13841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8873040708337783,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 13842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.43633065139179,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 13843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9309171350562351,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 13844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5531979426672684,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 13845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6160678575213314,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 13846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1979304547829395,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 13847
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.952333403007746,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 13848
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.377805038671187,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 13849
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9056616954079155,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 13850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9314860729562631,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 13851
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6237599703370327,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 13852
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.176186104438479,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 13853
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.713312284118502,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 13854
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5794339181155144,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 13855
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4551618932812198,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 13856
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.834950172753417,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 13857
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.529282220256253,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 13858
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.004699522917801,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 13859
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.938851961545087,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 13860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.82923214157099,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 13861
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.080604508110838,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 13862
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2700654830600128,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 13863
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3109972427001773,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 13864
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7287900200415856,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 13865
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1889191730351905,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 13866
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2025728791569446,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 13867
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8454909429749085,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 13868
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3413130360151887,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 13869
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44861632048170985,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 13870
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5157356615075717,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 13871
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.18176470384583254,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 13872
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7877590978681939,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 13873
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2784874604989824,
+      "learning_rate": 2e-05,
+      "loss": 0.1057,
+      "step": 13874
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.722584860084929,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 13875
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2334487433606083,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 13876
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.18444959537759537,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 13877
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4737697610994327,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 13878
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7856752337063537,
+      "learning_rate": 2e-05,
+      "loss": 0.1166,
+      "step": 13879
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6161196219010836,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 13880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3933384682365956,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 13881
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7443580209763822,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 13882
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.181451773943448,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 13883
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5660389728877507,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 13884
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2635850832346993,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 13885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9657371492905965,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 13886
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1086611533021198,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 13887
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.671017409968603,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 13888
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5078799990307696,
+      "learning_rate": 2e-05,
+      "loss": 0.0995,
+      "step": 13889
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.10944365577524867,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 13890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8940609691064095,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 13891
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7060522393359614,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 13892
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4056961664390029,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 13893
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8982616704197464,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 13894
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6876664177009192,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 13895
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1354889919565456,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 13896
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.42739366870951484,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 13897
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9798027299346549,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 13898
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.816462036088241,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 13899
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.256860446507982,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 13900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43176286787440005,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 13901
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9880568821788185,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 13902
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30576536066158005,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 13903
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0761064069880666,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 13904
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2604099725382616,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 13905
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.068650270520153,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 13906
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.18302183148439935,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 13907
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4816060910061125,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 13908
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9880617825326167,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 13909
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.539591103328828,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 13910
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.871310185798695,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 13911
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09527660714163803,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 13912
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4590487495892298,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 13913
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.485842742144503,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 13914
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4336043004307447,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 13915
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4891148737203571,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 13916
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6431003606725638,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 13917
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8161015736997954,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 13918
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4187576999539382,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 13919
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1713927739103345,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 13920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.322238212917745,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 13921
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9152602204693823,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 13922
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9849823183321937,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 13923
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14524969466467852,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 13924
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0905885509969435,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 13925
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1592584597682216,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 13926
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.600194178748533,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 13927
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4183452638603584,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 13928
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3321185041471297,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 13929
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9137760920002163,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 13930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7306152033740512,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 13931
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1505803986317145,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 13932
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5330310585820933,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 13933
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8364226101237866,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 13934
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6616501611505563,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 13935
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.002415422120462,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 13936
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0397826259196816,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 13937
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.20440961031327,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 13938
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4216116357388335,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 13939
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.10376784798053688,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 13940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.932786208964006,
+      "learning_rate": 2e-05,
+      "loss": 0.1072,
+      "step": 13941
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.953575753650276,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 13942
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6307740101416015,
+      "learning_rate": 2e-05,
+      "loss": 0.0869,
+      "step": 13943
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3983927884759386,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 13944
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.69495988833668,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 13945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0239064547234245,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 13946
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8654050981510291,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 13947
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.919456412771134,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 13948
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.604230290932301,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 13949
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.410017457795919,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 13950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.623917763229058,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 13951
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.803754689295774,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 13952
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0781160437170447,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 13953
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9863850492159604,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 13954
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3665201792868829,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 13955
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2163714159496295,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 13956
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0844441856278004,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 13957
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.502105660027863,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 13958
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5662974940570922,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 13959
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.10952566103304,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 13960
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.255611944459427,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 13961
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8603558557816273,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 13962
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3203459698938133,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 13963
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4099992574865672,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 13964
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.059807493788802,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 13965
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.23705125156018686,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 13966
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2848766733081077,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 13967
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.48485811812772145,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 13968
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.105012995181762,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 13969
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.37956565817177507,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 13970
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1341619412344528,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 13971
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0991635155234536,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 13972
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.614453062989083,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 13973
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6720812946587434,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 13974
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5026332471161234,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 13975
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.122364000545788,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 13976
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9463227288681139,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 13977
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2737070390008001,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 13978
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2439222398661494,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 13979
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1738457828980158,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 13980
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6151150309929114,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 13981
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9148390967741828,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 13982
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7960479408688459,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 13983
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.231305857906567,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 13984
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.135216667920055,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 13985
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.280041862778893,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 13986
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5827264459231456,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 13987
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6780910351801714,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 13988
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.46128642889448973,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 13989
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8053028753591449,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 13990
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.703072519840315,
+      "learning_rate": 2e-05,
+      "loss": 0.0732,
+      "step": 13991
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7908188186522112,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 13992
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6664127720718268,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 13993
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0147330842676976,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 13994
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7535904138765279,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 13995
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.312739849881692,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 13996
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5861302630919778,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 13997
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.22580921798451628,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 13998
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.796161078559617,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 13999
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3409815409647928,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 14000
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4063827676186498,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 14001
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.324858276161974,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 14002
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.823463684915253,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 14003
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9249822653539526,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 14004
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0322560113984502,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 14005
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.729547879890999,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 14006
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3067129013506643,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 14007
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0114647388091733,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 14008
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.5304345811297715,
+      "learning_rate": 2e-05,
+      "loss": 0.1502,
+      "step": 14009
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3698744609841103,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 14010
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.35381388859020657,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 14011
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2922914290352137,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 14012
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0181274257159783,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 14013
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.88317604150418,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 14014
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4706174463288049,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 14015
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0155511095604473,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 14016
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1738029896207829,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 14017
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2772083269012093,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 14018
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6855435014183358,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 14019
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7936513433946492,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 14020
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.88989149004155,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 14021
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6028892639323304,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 14022
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3715643876524055,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 14023
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5527455898265982,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 14024
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.646149632206344,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 14025
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.995864083490853,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 14026
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6892460628296065,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 14027
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4977255011280954,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 14028
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3929973896738803,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 14029
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9997103109364223,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 14030
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5309377079356377,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 14031
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.138129903324396,
+      "learning_rate": 2e-05,
+      "loss": 0.0921,
+      "step": 14032
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8837680628574214,
+      "learning_rate": 2e-05,
+      "loss": 0.1068,
+      "step": 14033
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3968699041925522,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 14034
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4988555515972288,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 14035
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4940317949540083,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 14036
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.935002015615158,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 14037
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0488324306213697,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 14038
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7463523451373435,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 14039
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.08452866094907,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 14040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7601989822178437,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 14041
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2273951820169573,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 14042
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9662843212531518,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 14043
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5508466370640503,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 14044
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4709460956987956,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 14045
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.771689819794651,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 14046
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27856549623045007,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 14047
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.10336586177132331,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 14048
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.97728694862824,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 14049
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6471977792289176,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 14050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6051289989541816,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 14051
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5011866961622793,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 14052
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.486189927281915,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 14053
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.18734113599148236,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 14054
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.357158402808644,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 14055
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.065708253354693,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 14056
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5919021915606609,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 14057
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3824874906421663,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 14058
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1109701013010113,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 14059
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7822718939631432,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 14060
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4766472686170882,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 14061
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7844984315847665,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 14062
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0003879614153814,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 14063
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8536376912930626,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 14064
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0377259532870124,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 14065
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.23605804143022516,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 14066
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.286707372868435,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 14067
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.298620956075322,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 14068
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.691878249273622,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 14069
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.755137214381259,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 14070
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3321185582876502,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 14071
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.366954674650937,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 14072
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1716386839384007,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 14073
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.808302656109798,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 14074
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.366222149773617,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 14075
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7684816279700288,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 14076
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6106055867868228,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 14077
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.46640401932328307,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 14078
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6945659910011373,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 14079
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3182075017412052,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 14080
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3133440302018864,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 14081
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3039584378540432,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 14082
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.485191733372657,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 14083
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5790576371360847,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 14084
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3444066338779814,
+      "learning_rate": 2e-05,
+      "loss": 0.1092,
+      "step": 14085
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.400620875048471,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 14086
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4039523533727103,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 14087
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4596630392146388,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 14088
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2580678455543004,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 14089
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3406690274636228,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 14090
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.102725142136498,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 14091
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0634644305355025,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 14092
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6134138087934775,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 14093
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.88175826382748,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 14094
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0389668700303052,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 14095
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.591844487556825,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 14096
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8199840069583555,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 14097
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.644267806383774,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 14098
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.122524876563149,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 14099
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1916917802703089,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 14100
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9230353924153709,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 14101
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.258874527757738,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 14102
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2911523697495157,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 14103
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2689273313046434,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 14104
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3103146752399246,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 14105
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.975006695259777,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 14106
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25576884000170846,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 14107
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9365841206450187,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 14108
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3367495832579452,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 14109
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0830620768430277,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 14110
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8113916230892413,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 14111
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.582983998295231,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 14112
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8551232250230316,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 14113
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8872245992580774,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 14114
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.23103654553218764,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 14115
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.199540616863373,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 14116
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8203239123676216,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 14117
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9898260354351478,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 14118
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06806380884039154,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 14119
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.64346386598813,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 14120
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6752165440889226,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 14121
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8491037552364712,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 14122
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1717716425774405,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 14123
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8344348358325224,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 14124
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26947424465462516,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 14125
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.78286456274116,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 14126
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4091467797342216,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 14127
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.880580958474383,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 14128
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.40102790154522,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 14129
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.725734727133926,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 14130
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.184249206326264,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 14131
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9583296741965448,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 14132
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2439001979903472,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 14133
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9319198062049483,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 14134
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.602880470541627,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 14135
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6664228578894333,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 14136
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5659866357175874,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 14137
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5642324101905409,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 14138
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4730758271650937,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 14139
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.135974382412656,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 14140
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.706911060981775,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 14141
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5155232647031512,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 14142
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7881985457468156,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 14143
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6880629271669465,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 14144
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.642610914327147,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 14145
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3647272698975745,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 14146
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8216962133966805,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 14147
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5137471481114145,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 14148
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2243090783093344,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 14149
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4594655375886929,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 14150
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8725721190631228,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 14151
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5945870873275474,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 14152
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4912631114982577,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 14153
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45529436532209305,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 14154
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0806686690835758,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 14155
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.652840037023632,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 14156
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4641482724007904,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 14157
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1134298138369005,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 14158
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2147194163980177,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 14159
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6055163448360676,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 14160
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2422455588154084,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 14161
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3840301239092467,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 14162
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6018290486526127,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 14163
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3509683967455492,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 14164
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.012622815288213,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 14165
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0103840668823736,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 14166
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3288550494038347,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 14167
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3447794851888506,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 14168
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0897592491287726,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 14169
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.698351743412962,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 14170
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9599708409152656,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 14171
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.800512331000055,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 14172
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6379078480901939,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 14173
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6194297983253554,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 14174
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.987664930826682,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 14175
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.033264434646799,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 14176
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4430345058690677,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 14177
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.63114400908413,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 14178
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.61267957944196,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 14179
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3897893302088506,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 14180
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1486348092954184,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 14181
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9642275380523728,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 14182
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.003805734256598,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 14183
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7228795299463951,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 14184
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2566666685470325,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 14185
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9053305532797132,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 14186
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.849566965714635,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 14187
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2148943279454705,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 14188
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1494212186812596,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 14189
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7932235104614482,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 14190
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.548898821312738,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 14191
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.777160896044212,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 14192
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8115148073327401,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 14193
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1352503360158914,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 14194
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.626300951242371,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 14195
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.626755866174247,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 14196
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.46431630126838974,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 14197
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0719937116986498,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 14198
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8047412002570605,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 14199
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4129383222631073,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 14200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2584792555689537,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 14201
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0617244358205231,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 14202
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4976982098881675,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 14203
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.630412843528975,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 14204
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4442230242375285,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 14205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7935849130404322,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 14206
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4255554755532151,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 14207
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.386494535130689,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 14208
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7581303116878849,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 14209
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5194784513190134,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 14210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.684920281808188,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 14211
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6956810349604485,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 14212
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2327749218458037,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 14213
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7722009725635384,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 14214
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6415710627929632,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 14215
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.96631096813183,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 14216
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5821417875879848,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 14217
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.283767924271861,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 14218
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8615589135406309,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 14219
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.986334404264022,
+      "learning_rate": 2e-05,
+      "loss": 0.1546,
+      "step": 14220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6559600189399915,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 14221
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.23942690732223507,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 14222
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4575259925788524,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 14223
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.129007565924919,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 14224
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7232726896005639,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 14225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9507901720241178,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 14226
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.205614779830521,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 14227
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9401788062911867,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 14228
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.49007007730067,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 14229
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3842884532287169,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 14230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9323110667222348,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 14231
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2488795504036092,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 14232
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9991851407290061,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 14233
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.21736897769853422,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 14234
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.536002045029034,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 14235
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.05697306341222589,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 14236
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6156597200127203,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 14237
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9640174447139266,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 14238
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.260604663068708,
+      "learning_rate": 2e-05,
+      "loss": 0.1102,
+      "step": 14239
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9688845202662714,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 14240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3619337543690926,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 14241
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5509176964535502,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 14242
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0957473501164263,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 14243
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.162429579100858,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 14244
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08659533393960642,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 14245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.19293368325939336,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 14246
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7726140063898328,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 14247
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3385794321316138,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 14248
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9603703073220359,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 14249
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.19059004267676927,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 14250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.125023364778124,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 14251
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.154464735292982,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 14252
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6926704598126774,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 14253
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3972879725272276,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 14254
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.24310946330315233,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 14255
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.059586625918484,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 14256
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3333484032891902,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 14257
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6414452818233738,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 14258
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.631034803062467,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 14259
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5330855597839816,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 14260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.80672681592758,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 14261
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.766581790912542,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 14262
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.66116474710724,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 14263
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3011348553181362,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 14264
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7136241326860538,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 14265
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0372195427781716,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 14266
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3664697978337426,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 14267
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1708345499179442,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 14268
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5812850887943821,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 14269
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0164259466656864,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 14270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.183419439070037,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 14271
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7457650144461394,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 14272
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0522525414223303,
+      "learning_rate": 2e-05,
+      "loss": 0.0775,
+      "step": 14273
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31865039933570344,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 14274
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9357189426389434,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 14275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9833497960657626,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 14276
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9430455557550435,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 14277
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0935430577243027,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 14278
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.49828750182973025,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 14279
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.817681773701626,
+      "learning_rate": 2e-05,
+      "loss": 0.1349,
+      "step": 14280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.074761903270163,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 14281
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9965133082586917,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 14282
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4765827612400151,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 14283
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.20340487565762927,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 14284
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8600225609769806,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 14285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8650950713639334,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 14286
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7735404841685254,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 14287
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2669071918922476,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 14288
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0941688915665395,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 14289
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3998737312322727,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 14290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.22101063382622,
+      "learning_rate": 2e-05,
+      "loss": 0.1597,
+      "step": 14291
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1894100025024135,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 14292
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4874708167601787,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 14293
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6811534073887502,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 14294
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.41236692176091,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 14295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5999618140625972,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 14296
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.829405383358444,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 14297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5245142335011448,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 14298
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7723496247114918,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 14299
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.574452498808792,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 14300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8640917602330737,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 14301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.798684842412088,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 14302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4151294010159545,
+      "learning_rate": 2e-05,
+      "loss": 0.0757,
+      "step": 14303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3292661812977422,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 14304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5607844280531733,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 14305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9513321571111146,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 14306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.644678070376421,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 14307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.343479667917645,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 14308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1811659182540815,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 14309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8087345314020569,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 14310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0737386520070773,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 14311
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.798877454893401,
+      "learning_rate": 2e-05,
+      "loss": 0.0828,
+      "step": 14312
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6533563030918317,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 14313
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9040907754705747,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 14314
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3160076456276749,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 14315
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1048085835608417,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 14316
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4629263981722103,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 14317
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5983459880769268,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 14318
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4460908185948729,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 14319
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5075371050973303,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 14320
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5781440216234378,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 14321
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.205240286212314,
+      "learning_rate": 2e-05,
+      "loss": 0.0878,
+      "step": 14322
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5103742747003186,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 14323
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9507985723918874,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 14324
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2688284155199143,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 14325
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.958687589085367,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 14326
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7213014842786157,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 14327
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7984757300853222,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 14328
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.609999895936226,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 14329
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5509426655892873,
+      "learning_rate": 2e-05,
+      "loss": 0.0671,
+      "step": 14330
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.165415189130872,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 14331
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6879527579562295,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 14332
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3048311257257001,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 14333
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16561510915732935,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 14334
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.561526567504876,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 14335
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6756753936670434,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 14336
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9339228732266848,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 14337
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4351967313204046,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 14338
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8023289308600201,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 14339
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4307088070274752,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 14340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9428009528646886,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 14341
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2773327841565087,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 14342
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.101385385371149,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 14343
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0081852073441455,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 14344
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0822977777528229,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 14345
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3165458441906345,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 14346
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8540426412340074,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 14347
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2989304122228504,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 14348
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3972000955599653,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 14349
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.717688780880346,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 14350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1459399202652572,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 14351
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7380424421196194,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 14352
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9164840053862949,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 14353
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4519890888266533,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 14354
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0170258471224858,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 14355
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9487776182575148,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 14356
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.320272426066506,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 14357
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3016612424179812,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 14358
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.508409609192007,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 14359
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1819215046625977,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 14360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8637991685025558,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 14361
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4373173673085481,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 14362
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.465046192688117,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 14363
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3011199957928725,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 14364
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0008346484951964,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 14365
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.540291118606937,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 14366
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9314892346000416,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 14367
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3079225749987658,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 14368
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0553185729534624,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 14369
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0781908982634496,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 14370
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4452470919268519,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 14371
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9650451511730644,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 14372
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.540127380973611,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 14373
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8460705952104546,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 14374
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9882236810736151,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 14375
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1933991820443663,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 14376
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0908677489072356,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 14377
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7878805195658467,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 14378
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6271112545636806,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 14379
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.778024195275814,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 14380
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5007312105048898,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 14381
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2268207820758315,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 14382
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2096998719726102,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 14383
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7570476389889937,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 14384
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.223535288755081,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 14385
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5503175826436394,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 14386
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3363718502951456,
+      "learning_rate": 2e-05,
+      "loss": 0.0751,
+      "step": 14387
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9723568497648417,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 14388
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9006705922852962,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 14389
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.2770804474355,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 14390
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1317197066917521,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 14391
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4174926220647411,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 14392
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.580971458498947,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 14393
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.579698520049753,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 14394
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3780583401711433,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 14395
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.325619982022267,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 14396
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2165057154951144,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 14397
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2820949296120947,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 14398
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8024195171911388,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 14399
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.23366019048679,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 14400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.777694229353009,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 14401
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8189789958871536,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 14402
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27480616674180985,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 14403
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7682484812921669,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 14404
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1870138779514248,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 14405
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4886876665547574,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 14406
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6028771730124545,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 14407
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4131354143382358,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 14408
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4486112218235983,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 14409
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6264994379107691,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 14410
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3469616777361856,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 14411
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.031690109697653,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 14412
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.858058772417782,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 14413
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7284958315648142,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 14414
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5785648889062704,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 14415
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6158747171409458,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 14416
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5166377900370436,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 14417
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7505609633765133,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 14418
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9299894531738943,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 14419
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.87231138107002,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 14420
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7354680712477739,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 14421
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.642997525285775,
+      "learning_rate": 2e-05,
+      "loss": 0.1178,
+      "step": 14422
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7046271490741336,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 14423
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.592253931446315,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 14424
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5229652608909502,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 14425
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.24792264095021468,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 14426
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4733898995016154,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 14427
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9566703583040242,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 14428
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7759339089220259,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 14429
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.457332958248145,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 14430
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.913006407690755,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 14431
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0698654466325732,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 14432
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14911675637363178,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 14433
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1647144092692769,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 14434
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.305211101472287,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 14435
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.302195333114438,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 14436
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.84568301725909,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 14437
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0792963842419696,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 14438
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18660910870819652,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 14439
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5127184024474447,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 14440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.317976108930719,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 14441
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8822216832278902,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 14442
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1743695451659915,
+      "learning_rate": 2e-05,
+      "loss": 0.1336,
+      "step": 14443
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.618612965933768,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 14444
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4434962733618706,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 14445
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.12048611159588792,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 14446
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.292986866803896,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 14447
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9317268635564315,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 14448
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2139975616358498,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 14449
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9207198777938465,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 14450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5828058463101495,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 14451
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.151319070170306,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 14452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.553318036555227,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 14453
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8954195417615978,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 14454
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8644004035878885,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 14455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.80795712679719,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 14456
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.109851734723951,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 14457
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.007582714424272,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 14458
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7380642445857236,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 14459
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.082808759132225,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 14460
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.108888358035801,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 14461
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2568234174055886,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 14462
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.604680128813194,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 14463
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.661276482035346,
+      "learning_rate": 2e-05,
+      "loss": 0.1144,
+      "step": 14464
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9309941242667936,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 14465
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1515449731478427,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 14466
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9327444966595237,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 14467
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7016572840049289,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 14468
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7108660818026262,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 14469
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6596917995889217,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 14470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6419870272511135,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 14471
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.702203563540383,
+      "learning_rate": 2e-05,
+      "loss": 0.0878,
+      "step": 14472
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9224429240054032,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 14473
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.19483003153530853,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 14474
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2704572940297703,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 14475
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5620298059626059,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 14476
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.350239409305065,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 14477
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5171968626760786,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 14478
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.211621294462031,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 14479
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5068909425968187,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 14480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6501539332342365,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 14481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.351463878342722,
+      "learning_rate": 2e-05,
+      "loss": 0.0732,
+      "step": 14482
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.11862865258573826,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 14483
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2200711743120842,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 14484
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6541192365840136,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 14485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.774701848308787,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 14486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.643700867938345,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 14487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1008800401282311,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 14488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2953244327230211,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 14489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5280201382533265,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 14490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.20423101125682722,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 14491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.274747937004796,
+      "learning_rate": 2e-05,
+      "loss": 0.1074,
+      "step": 14492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9103603137020315,
+      "learning_rate": 2e-05,
+      "loss": 0.0751,
+      "step": 14493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5714923203620663,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 14494
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6511352303966524,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 14495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9256444580485057,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 14496
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1418712026682218,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 14497
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3239151147991106,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 14498
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.108928944084824,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 14499
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6434988365192462,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 14500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.213904341363633,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 14501
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5863956484951833,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 14502
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4339096819467785,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 14503
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.956960620998697,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 14504
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8591829729697373,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 14505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1064913168599648,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 14506
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.22005425495377817,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 14507
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6277604653729186,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 14508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3779554136524646,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 14509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0431694676136565,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 14510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9844877087691338,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 14511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7638334215621791,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 14512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5505360267104099,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 14513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7107465298890302,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 14514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.610254732439786,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 14515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4439494566240936,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 14516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.190763709102347,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 14517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3244949336479334,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 14518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.455962659806689,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 14519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8163925334453537,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 14520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4140760726698849,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 14521
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.276783578282112,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 14522
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4654214366432385,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 14523
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.165394837670265,
+      "learning_rate": 2e-05,
+      "loss": 0.0775,
+      "step": 14524
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.21801233231917688,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 14525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.762630116955687,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 14526
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6109049663348034,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 14527
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8079939109100498,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 14528
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0170595254053163,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 14529
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7143630665677754,
+      "learning_rate": 2e-05,
+      "loss": 0.0923,
+      "step": 14530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.052318823617594,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 14531
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.699083790858675,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 14532
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3497628964872717,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 14533
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4748677784306676,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 14534
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30158007427257794,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 14535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2815270451621643,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 14536
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.24413779447822062,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 14537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5787208682446378,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 14538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4685582881533133,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 14539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7371216066932187,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 14540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2502119361500486,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 14541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5142605321260927,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 14542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2783466566619142,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 14543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9442106749776114,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 14544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0274715787539486,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 14545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0449375044464873,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 14546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0971004950401415,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 14547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6621646715014062,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 14548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.694228287061097,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 14549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.566649171431063,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 14550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1991290926231977,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 14551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.822996453549889,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 14552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5914219326918877,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 14553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.78278171201796,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 14554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9412676760671645,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 14555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5376884615962805,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 14556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3734901045694836,
+      "learning_rate": 2e-05,
+      "loss": 0.0857,
+      "step": 14557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.36986216916244796,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 14558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7401497825257854,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 14559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1211587175957263,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 14560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8899269391627087,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 14561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0546723541302532,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 14562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8553307130133715,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 14563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2317378854728113,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 14564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0814590318197808,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 14565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7258223999315333,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 14566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3868668168227491,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 14567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.839850099718674,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 14568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.153748318787205,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 14569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6591581328610203,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 14570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3225412403951145,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 14571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.513372024862154,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 14572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6883398014985302,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 14573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6476976241090118,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 14574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2789702253238793,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 14575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6769679680378519,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 14576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8419266886090243,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 14577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.650834999177588,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 14578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.461198774727861,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 14579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7549029876109685,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 14580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5310012571639644,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 14581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7947935333496677,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 14582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8232996367368588,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 14583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7871659402897111,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 14584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.874156545381043,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 14585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9180451706263498,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 14586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1443419892690352,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 14587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.938225814871513,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 14588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8436986766323279,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 14589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7261458364518194,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 14590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5783909488596344,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 14591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6275669313731556,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 14592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1337276373421829,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 14593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.75409576540998,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 14594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5638335896793151,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 14595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9461895365779727,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 14596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.46582984514699055,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 14597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.316926997352219,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 14598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5114875445858373,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 14599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3852546377250667,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 14600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2447746284453212,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 14601
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9212071213907862,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 14602
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8670901348952385,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 14603
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7293287530635941,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 14604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5936635857992464,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 14605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0103959519532677,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 14606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8590860401201827,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 14607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8877189809331776,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 14608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6222581804288769,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 14609
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14631655935027968,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 14610
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9222396876742809,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 14611
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1819515931039764,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 14612
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.195601484483223,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 14613
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7595493038113781,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 14614
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7662878475830043,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 14615
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3886326436913998,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 14616
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4422904096692784,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 14617
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8080814039589825,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 14618
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.11765016775633998,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 14619
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3318238306621093,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 14620
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5556796947778615,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 14621
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2878235382665986,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 14622
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.20361780028965798,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 14623
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1413637146801792,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 14624
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4133894356502956,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 14625
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1738053424778453,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 14626
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7149590988064674,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 14627
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6438957549334232,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 14628
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6186812132135147,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 14629
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31200197589688755,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 14630
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6384609664118337,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 14631
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3754860279090524,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 14632
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9842064252723408,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 14633
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8348084070634134,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 14634
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4285864855755265,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 14635
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.092793018024929,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 14636
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35074949566838176,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 14637
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7879595838411125,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 14638
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.470844973386093,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 14639
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08893847488236277,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 14640
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3972089939019236,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 14641
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.192371311207869,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 14642
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6434792921621708,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 14643
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0736827831079208,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 14644
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.99040211718353,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 14645
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6406674952485122,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 14646
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.400687311416958,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 14647
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.618734329650171,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 14648
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9473972818965788,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 14649
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.276854427885127,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 14650
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9615638986303988,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 14651
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3406009678718526,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 14652
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.641101139638042,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 14653
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1553720964534876,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 14654
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4224423125625725,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 14655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6166661845858665,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 14656
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.22590160165147605,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 14657
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8874547987159382,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 14658
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1246362549540662,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 14659
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1095809745309884,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 14660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8299533145973852,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 14661
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.524905438081359,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 14662
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7395777978256994,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 14663
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9006933667338157,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 14664
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.07014441736636,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 14665
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2042677910914947,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 14666
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.700152892361924,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 14667
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.420633046118611,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 14668
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.22826973001550535,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 14669
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3152588301843017,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 14670
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5771200105279166,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 14671
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.38457629585164343,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 14672
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.227513053504723,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 14673
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.781020931993995,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 14674
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.154129506739747,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 14675
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.186059422940891,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 14676
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6832320520161397,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 14677
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9760319898734058,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 14678
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.493084361897725,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 14679
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3500124030073282,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 14680
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.339980591098467,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 14681
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8382641153281981,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 14682
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8457056974940125,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 14683
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6842275608057715,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 14684
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.690751083202263,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 14685
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8103652089323095,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 14686
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4151291944172093,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 14687
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3674274431666911,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 14688
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4290021285435528,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 14689
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.296188874819847,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 14690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2269601845378777,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 14691
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8758679606245297,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 14692
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3880691455573701,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 14693
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9215118210065907,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 14694
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.02307798171554,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 14695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0020199877940164,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 14696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.630880580893099,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 14697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9842788709902485,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 14698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6114795241401694,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 14699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3451820549542985,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 14700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9025819142117698,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 14701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6476809677505615,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 14702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.519026948927434,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 14703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.309888309949254,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 14704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.778811073353851,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 14705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4993051827219616,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 14706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4961762750442427,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 14707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07326072458290823,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 14708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.685776855055705,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 14709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.958809963688779,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 14710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.857061064775558,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 14711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0515623512471386,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 14712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0620307388530494,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 14713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8255863005634803,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 14714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4098839747936942,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 14715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9662809481310224,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 14716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.239252808303665,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 14717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6167386066881977,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 14718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5867678873504176,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 14719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6659098790140723,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 14720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1172558769814973,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 14721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.163560609423881,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 14722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6736944978320286,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 14723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8492403791572616,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 14724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.37768617813659267,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 14725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27274316096625884,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 14726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8253427227109631,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 14727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2233352562563855,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 14728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.247555488239163,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 14729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.927419929683755,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 14730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8909940791420345,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 14731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.679326069204648,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 14732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8298657672987324,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 14733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0790115802238933,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 14734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.137938581657104,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 14735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6858225428503744,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 14736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.60004781899127,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 14737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8480380996434476,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 14738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.148819321459543,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 14739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8150568042667947,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 14740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.990813176244915,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 14741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0510263185819437,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 14742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4851418371569765,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 14743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9698265923023612,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 14744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1548650743558175,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 14745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.64253373542263,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 14746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4161939206177514,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 14747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6853052631366385,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 14748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12558044581768293,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 14749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.694514811412911,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 14750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.282498811007912,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 14751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.47103246666642246,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 14752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6919903846951957,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 14753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5489099391961434,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 14754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.848988542623389,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 14755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6017452556881099,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 14756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7473970783147954,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 14757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.580861214876628,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 14758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2714969165567193,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 14759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1605698076394353,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 14760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0920678418363465,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 14761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8649296498280687,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 14762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7148627606975202,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 14763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.903094062041179,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 14764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9671838374024246,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 14765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5707410540741136,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 14766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.376666596280224,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 14767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18111370939403468,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 14768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7163264440528008,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 14769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3353242332776087,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 14770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6438506614119592,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 14771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4006555099157434,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 14772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.197111865279842,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 14773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2276181094651413,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 14774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.226710316686488,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 14775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.764584395022863,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 14776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6973110671374647,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 14777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5213504678467273,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 14778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2747160221321387,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 14779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1833143250635627,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 14780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6650595386845786,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 14781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06290922452609145,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 14782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.021906092786427,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 14783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7755340904672322,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 14784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1137534227757462,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 14785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7619419260022421,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 14786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2048834376148328,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 14787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.432507997904715,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 14788
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.46290496610807097,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 14789
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5199521876325348,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 14790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.609528245191057,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 14791
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.746965376229165,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 14792
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39961977933015685,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 14793
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.25246625589809574,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 14794
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.24512298546731165,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 14795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2853705632896263,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 14796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.407088517162627,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 14797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7675217469455193,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 14798
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.994952883596565,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 14799
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.179554403605155,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 14800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4581752230559344,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 14801
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0821758511441385,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 14802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6432132279777036,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 14803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5035555306565841,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 14804
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06718445690604628,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 14805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0266427088600876,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 14806
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.687891535258335,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 14807
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9967205459487543,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 14808
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.111569880390353,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 14809
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8686674629279592,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 14810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4666770754069112,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 14811
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1967437531293545,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 14812
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39315575004844816,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 14813
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.505785732123747,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 14814
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.546358581036436,
+      "learning_rate": 2e-05,
+      "loss": 0.0736,
+      "step": 14815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7584606820026675,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 14816
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.78888896327464,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 14817
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.843081207531272,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 14818
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5195624263271912,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 14819
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.336024106860345,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 14820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.103130820369765,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 14821
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29120686027389714,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 14822
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.5497593071961475,
+      "learning_rate": 2e-05,
+      "loss": 0.1001,
+      "step": 14823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6086781013609995,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 14824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7097377085090013,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 14825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18366190477226862,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 14826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9544934528282805,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 14827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9386848274191387,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 14828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26662960533800334,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 14829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.384646650966004,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 14830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9321971646510285,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 14831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5046368115837899,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 14832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8689775978379007,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 14833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1064528958993654,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 14834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8745329495935383,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 14835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4233421140827693,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 14836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5772728955224925,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 14837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3044882391770902,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 14838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18019767566175762,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 14839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3518774846770136,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 14840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.059057425989015,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 14841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6239247617634747,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 14842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5595899115786711,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 14843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.397164370341053,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 14844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9119742577270175,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 14845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4813451945687395,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 14846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.015269173131807,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 14847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2164221093326143,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 14848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08876297677898765,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 14849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2521075868197515,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 14850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7383651829370876,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 14851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.030163595484406135,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 14852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.434334960910572,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 14853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4949206342011687,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 14854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8929097500698009,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 14855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08097463124277511,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 14856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.24322900326224373,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 14857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13600891994008144,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 14858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6401031043031183,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 14859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12366379007768204,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 14860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9962761304813617,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 14861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5377510815072634,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 14862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2652094088313093,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 14863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.678970374415942,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 14864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2196338089920675,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 14865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.477510339794351,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 14866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.666810010797513,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 14867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7708300169452431,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 14868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06628338231964757,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 14869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9022597151413072,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 14870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.249233176671019,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 14871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0974063632580145,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 14872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5392314217905216,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 14873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0095654995789345,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 14874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.74570467757972,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 14875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8547780247026326,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 14876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7906956154872165,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 14877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7743320878695279,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 14878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3572865226407145,
+      "learning_rate": 2e-05,
+      "loss": 0.0939,
+      "step": 14879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.059464503935125,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 14880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1431753739705366,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 14881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30462815856157016,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 14882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4388008037118414,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 14883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6096050121792171,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 14884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4032041183722576,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 14885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.172025280754218,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 14886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5757291224160846,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 14887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.964328582002565,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 14888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.80267712471472,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 14889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0728286349035803,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 14890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.6010814243167,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 14891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.427051269628613,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 14892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.290390942428591,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 14893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6955871359579904,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 14894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1977018025414232,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 14895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9277973675674259,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 14896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.47915982849659705,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 14897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3942144124819582,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 14898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6981338310195423,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 14899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1141271373748663,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 14900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9851912187485418,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 14901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6151384281320311,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 14902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7738492621009948,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 14903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5043144190948308,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 14904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5961181445648457,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 14905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.806394774646744,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 14906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9490574515230105,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 14907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6261954319188169,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 14908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.132036228530567,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 14909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6295041969622173,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 14910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8867428666321766,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 14911
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5624794956284824,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 14912
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7113559159018716,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 14913
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0847783100270467,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 14914
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5987191456499352,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 14915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.398285680427103,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 14916
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.007727183360606,
+      "learning_rate": 2e-05,
+      "loss": 0.1042,
+      "step": 14917
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3360337962151044,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 14918
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2229549696630513,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 14919
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2264838223826955,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 14920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7234668206685192,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 14921
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.56609004666016,
+      "learning_rate": 2e-05,
+      "loss": 0.0998,
+      "step": 14922
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6152025655545648,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 14923
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5521876188665584,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 14924
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7001309456601739,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 14925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2826065752189485,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 14926
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12012322340895966,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 14927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3163523677133326,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 14928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.72518088123234,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 14929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.456048280620371,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 14930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9889314646044602,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 14931
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0658066570144067,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 14932
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.173890859241905,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 14933
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.476067520485731,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 14934
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.992381111720566,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 14935
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.097832933643485,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 14936
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2540418129723032,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 14937
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17107920214670772,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 14938
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.625768185792729,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 14939
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8633312711421045,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 14940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.542882478750847,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 14941
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7232389610664145,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 14942
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6961924952759226,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 14943
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1206262125326767,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 14944
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.003943965326429,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 14945
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4809652260019273,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 14946
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6630665122223216,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 14947
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5463200887326023,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 14948
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.185312539799844,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 14949
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1208169942749815,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 14950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.611515582439205,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 14951
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7380472851959345,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 14952
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5690023546024936,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 14953
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.598506934185816,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 14954
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6412907142057895,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 14955
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6046223538813832,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 14956
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3671891566675232,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 14957
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.36732608127978766,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 14958
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6025712720759504,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 14959
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5094837947201776,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 14960
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4733815317052033,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 14961
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2680532380230183,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 14962
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3315041133861485,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 14963
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.45343015680124,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 14964
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.442337931888282,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 14965
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.076781933554991,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 14966
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.6232785294595775,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 14967
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.239589685803119,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 14968
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.180492404420516,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 14969
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1541031676542877,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 14970
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0820060313151543,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 14971
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8084207248311283,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 14972
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.576523449719432,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 14973
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12069930160047024,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 14974
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9729643024090229,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 14975
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4899964462009794,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 14976
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.696722261920698,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 14977
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.361838059309338,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 14978
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1509383423734216,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 14979
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.086137596113627,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 14980
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8445550821701673,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 14981
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4425911654834773,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 14982
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3569569299586503,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 14983
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.627249700984907,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 14984
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4236588338278446,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 14985
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4509196620602995,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 14986
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41711500802445706,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 14987
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6861631927736016,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 14988
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2404670557139539,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 14989
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1625973297665729,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 14990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9514407614019502,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 14991
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0336790379838487,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 14992
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8016934293481731,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 14993
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8125005868762905,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 14994
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.225521339459893,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 14995
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.677770612987608,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 14996
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.915425572418296,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 14997
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.44742784515122513,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 14998
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5718924701546484,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 14999
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6954084799884432,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 15000
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.7977962044244284,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 15001
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0688692372157478,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 15002
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8350090860683379,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 15003
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.97637579279117,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 15004
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4250619900045836,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 15005
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.378406290469302,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 15006
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3842875933343843,
+      "learning_rate": 2e-05,
+      "loss": 0.1276,
+      "step": 15007
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40612224257270724,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 15008
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06401157492106258,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 15009
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.108532207392288,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 15010
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.422993653884619,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 15011
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.738482909309042,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 15012
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5688363344061202,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 15013
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3294032267902314,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 15014
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6869729949879106,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 15015
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0728882474129202,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 15016
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4004658698113468,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 15017
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.821369471112088,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 15018
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0187552342775117,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 15019
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5872571453392188,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 15020
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8187217547371268,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 15021
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4528324283866751,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 15022
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4564690606272843,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 15023
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.165137031829462,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 15024
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4140099362883487,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 15025
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.287854745497035,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 15026
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5098365646636925,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 15027
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4271223749996365,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 15028
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.636893251717133,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 15029
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2906706517899549,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 15030
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9818644724084162,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 15031
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8909898311856739,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 15032
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.122680632472621,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 15033
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1091901060952787,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 15034
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7908411657778258,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 15035
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5613274008176163,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 15036
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4861959431204408,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 15037
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5076868667047736,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 15038
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3071022931866378,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 15039
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.590971964205579,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 15040
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.564538229892045,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 15041
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.573621017684942,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 15042
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.332893659402486,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 15043
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5789162191447168,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 15044
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.02124590632359,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 15045
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.38487459600795,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 15046
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.763715972017485,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 15047
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0481694886621535,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 15048
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.22110723110655994,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 15049
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30562502334454217,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 15050
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5773500196970784,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 15051
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1311526535891345,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 15052
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0888810366446908,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 15053
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.16793344829789,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 15054
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.500073590062917,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 15055
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6925907348472986,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 15056
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.306210719201543,
+      "learning_rate": 2e-05,
+      "loss": 0.1087,
+      "step": 15057
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.670931326956589,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 15058
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.329985623949749,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 15059
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29442224433242914,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 15060
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.347604262871459,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 15061
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4328815866410398,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 15062
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.033711933694384,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 15063
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2642086722566699,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 15064
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8709313171526096,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 15065
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.177375360156097,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 15066
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.486290909931603,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 15067
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06396014072513219,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 15068
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7536392355121153,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 15069
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8014329816793059,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 15070
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1045737549447967,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 15071
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5919780642675789,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 15072
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4662070478959663,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 15073
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1678824504495013,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 15074
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.017448133064266,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 15075
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.443649067465109,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 15076
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7685013019965732,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 15077
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6189268206322481,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 15078
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.943041034952498,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 15079
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5400464021208122,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 15080
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9369037721717723,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 15081
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.439609983729199,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 15082
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0102595112028845,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 15083
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07067708687242308,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 15084
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2185598371775703,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 15085
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6047668104371735,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 15086
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1896732922069622,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 15087
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9662392364614765,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 15088
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8079108314510839,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 15089
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8882300315993792,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 15090
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1847637675818208,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 15091
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.182724008347297,
+      "learning_rate": 2e-05,
+      "loss": 0.0976,
+      "step": 15092
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.374662594354052,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 15093
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6361353159813947,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 15094
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8535684879090433,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 15095
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2815827700493205,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 15096
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1722086622157946,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 15097
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.011287902323159,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 15098
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7529133004509289,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 15099
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7941263000897565,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 15100
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.651256086062493,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 15101
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3707080005665384,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 15102
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6587621961776249,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 15103
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8624567862750627,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 15104
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.443705416738023,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 15105
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5738950637969744,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 15106
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8966334055557996,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 15107
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6168182246293274,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 15108
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2624103760223724,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 15109
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7748347675397463,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 15110
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2281645499114688,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 15111
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9809386810363679,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 15112
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1758368089292452,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 15113
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16936835756332783,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 15114
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6854658160419685,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 15115
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8122329585138677,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 15116
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7555993788724197,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 15117
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.646693314839903,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 15118
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6294284047694552,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 15119
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.579787755034333,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 15120
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1982746197602503,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 15121
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.530021641896091,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 15122
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.782620646587978,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 15123
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0270607562159715,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 15124
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7542963155372002,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 15125
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8961238759571614,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 15126
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5164700541371083,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 15127
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.11001181703124,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 15128
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1234783902638683,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 15129
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.69440795925038,
+      "learning_rate": 2e-05,
+      "loss": 0.1009,
+      "step": 15130
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9002756637033935,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 15131
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.026657782250724,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 15132
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2153684610150752,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 15133
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8229599510752398,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 15134
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.258445255091511,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 15135
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.290799490745853,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 15136
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.441746201041725,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 15137
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6601523006616627,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 15138
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.934090228835032,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 15139
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.098092543481913,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 15140
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1772700417676445,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 15141
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.855831017279037,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 15142
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8354264595681768,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 15143
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9276698757771358,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 15144
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14175149937707965,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 15145
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.066556929269643,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 15146
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6636692902512985,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 15147
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2702381689417488,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 15148
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9800574723967335,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 15149
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0695872062010445,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 15150
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.46359092114970496,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 15151
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6724538781345526,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 15152
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6439635685598988,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 15153
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1404819025367937,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 15154
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3512706892369795,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 15155
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8039127685111898,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 15156
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8599180413160092,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 15157
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4930114812210446,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 15158
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.174598290767895,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 15159
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6926898974857423,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 15160
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9331056603831239,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 15161
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6926367204022502,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 15162
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26090642492773,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 15163
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.37155131189235086,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 15164
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.946818043471923,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 15165
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7577121282107544,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 15166
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1550724360001437,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 15167
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1455926903606074,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 15168
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.284088635933575,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 15169
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8888337658649084,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 15170
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7710402743610331,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 15171
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26253467739480263,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 15172
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7716973505723809,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 15173
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8449646133296098,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 15174
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9198549412311757,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 15175
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.539460136381597,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 15176
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6912211154124811,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 15177
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7441085047560475,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 15178
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.708626985453948,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 15179
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.448502285215259,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 15180
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8121269763010304,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 15181
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.385531615662419,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 15182
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1748374623636337,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 15183
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3727628107454453,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 15184
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9076204220418338,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 15185
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1337089471158426,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 15186
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1845399203258027,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 15187
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.449834070906185,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 15188
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.084450212178697,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 15189
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6199300610786427,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 15190
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.862868712553865,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 15191
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7824607883329664,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 15192
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.38562137305286753,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 15193
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4651613833214977,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 15194
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8298116458176744,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 15195
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6767488068921352,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 15196
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41769728356615793,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 15197
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5074782224971816,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 15198
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8425788563817425,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 15199
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3407886361344965,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 15200
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.487775100786002,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 15201
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2329005913199931,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 15202
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.208942916890212,
+      "learning_rate": 2e-05,
+      "loss": 0.0942,
+      "step": 15203
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5663604499991948,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 15204
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.889955516649103,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 15205
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0006468588105213,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 15206
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4371199678391666,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 15207
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7154851142139553,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 15208
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8403752618959771,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 15209
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8434631641785266,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 15210
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8123679787408249,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 15211
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.382062486017192,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 15212
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7011341520485059,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 15213
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8037542266690438,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 15214
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.263801524915063,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 15215
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.171758415331454,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 15216
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1575258322856716,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 15217
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.116847950638937,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 15218
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8444858957456118,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 15219
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1122372199036177,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 15220
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.737231450974351,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 15221
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8844793199580447,
+      "learning_rate": 2e-05,
+      "loss": 0.0956,
+      "step": 15222
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6436031479599678,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 15223
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5529153480408656,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 15224
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2874081202330947,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 15225
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.408859121298642,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 15226
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9353627638943667,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 15227
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2020089558951652,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 15228
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.121718282565317,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 15229
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8532162977280561,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 15230
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0193119324728115,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 15231
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6244109238684998,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 15232
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1356196382487334,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 15233
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3232565239899231,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 15234
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9831956732023865,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 15235
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4601443026638092,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 15236
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9568552906440877,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 15237
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3715867632490664,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 15238
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.327232599572249,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 15239
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2644050154319573,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 15240
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.410270008141411,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 15241
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3349158980761735,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 15242
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.750622765510482,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 15243
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7031402587207092,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 15244
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2703055030845896,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 15245
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.34438523206377697,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 15246
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9347773116417779,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 15247
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5079152516800925,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 15248
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.882478431351635,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 15249
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4618356186059844,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 15250
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8166016176226003,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 15251
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8048950340204792,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 15252
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.887596399588929,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 15253
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8232609032871256,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 15254
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4156595558513072,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 15255
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9544640795319166,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 15256
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6800459368419935,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 15257
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0177388307245496,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 15258
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.106494434333944,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 15259
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8789759968527955,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 15260
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.50559645493057,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 15261
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.32019442033335094,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 15262
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.173806899604012,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 15263
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9569021955396556,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 15264
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.20916083564502,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 15265
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28814971993624866,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 15266
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5454219582635844,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 15267
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3778052324504397,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 15268
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16854560292716006,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 15269
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7349838013126406,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 15270
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1586771054677432,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 15271
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.773281274295667,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 15272
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5130627397083782,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 15273
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7089134332373874,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 15274
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.494785784008972,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 15275
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.425472516302266,
+      "learning_rate": 2e-05,
+      "loss": 0.1306,
+      "step": 15276
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.341243135447147,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 15277
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11297489538664848,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 15278
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.3020678405236445,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 15279
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8626249105055257,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 15280
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.12311411780688,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 15281
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9832243699774272,
+      "learning_rate": 2e-05,
+      "loss": 0.0978,
+      "step": 15282
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2932102865171653,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 15283
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6829340994516249,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 15284
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4042755342462652,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 15285
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3259062979759335,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 15286
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6056531814115558,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 15287
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3742205696061327,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 15288
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0425798927083472,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 15289
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7426871776019929,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 15290
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1372607658755046,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 15291
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2629049379440028,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 15292
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.285390770507742,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 15293
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0761280073123096,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 15294
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.96957803522241,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 15295
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.18109948551551455,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 15296
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7733748439494559,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 15297
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3969113343430433,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 15298
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.88570937805239,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 15299
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0929633445142533,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 15300
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2914358983029588,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 15301
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2228721452414733,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 15302
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7830957733935577,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 15303
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.872657393294818,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 15304
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3533611347055141,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 15305
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2998905238758574,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 15306
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.336581240338156,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 15307
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9325471382397045,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 15308
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.434227512792235,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 15309
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24289849728366414,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 15310
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5224857585222116,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 15311
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8013135675932417,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 15312
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7536594995095725,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 15313
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1436066522038533,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 15314
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.388945218639369,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 15315
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8165483487958157,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 15316
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6419005614344324,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 15317
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2823338460092402,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 15318
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.062269233605024,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 15319
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7601780887301266,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 15320
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42381182872188566,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 15321
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7989900606728382,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 15322
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8929129875974801,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 15323
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7710834861930295,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 15324
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0555420980655268,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 15325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.077016232661307,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 15326
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1996583949326256,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 15327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8878334198165097,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 15328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0276372222043633,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 15329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2440231004238316,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 15330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9395970101972064,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 15331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.79562566696776,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 15332
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30845645071402683,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 15333
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4936803570078616,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 15334
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4106288128667368,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 15335
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8422861180441834,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 15336
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2361320352152567,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 15337
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0825081886175607,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 15338
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7958215693938921,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 15339
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4989104917628473,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 15340
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.2291608287351306,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 15341
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.828117478534442,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 15342
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.157586056969965,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 15343
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1772253481491421,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 15344
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8443754844748845,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 15345
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3387272578711102,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 15346
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.566934853115496,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 15347
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9145682476308616,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 15348
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.657862184518866,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 15349
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.431487617524443,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 15350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.039087054225364,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 15351
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5612356792736908,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 15352
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3241419587051793,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 15353
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.381247084408074,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 15354
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.271186159587691,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 15355
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0930737175752754,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 15356
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3041614225645735,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 15357
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8095473779758987,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 15358
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7908673936391576,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 15359
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.48645186477815,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 15360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6598205615807999,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 15361
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.620233089627647,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 15362
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9186260394583723,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 15363
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8451677464260998,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 15364
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41379693349999097,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 15365
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.299277273485673,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 15366
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.007477638147916,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 15367
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7193548725669773,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 15368
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6192590833523723,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 15369
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.693896023715573,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 15370
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.610300977423213,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 15371
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2726739571910222,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 15372
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4098870609585021,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 15373
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3723557458722637,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 15374
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.46203704974685,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 15375
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.911076673054477,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 15376
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.476546807946392,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 15377
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6167011276058517,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 15378
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5171416076606334,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 15379
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17570455255433104,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 15380
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.132446257617288,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 15381
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3067865447033282,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 15382
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1096427464982748,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 15383
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5532232702470494,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 15384
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0321744763126977,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 15385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9265472204997411,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 15386
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8268576947604972,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 15387
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.15978217033221,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 15388
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7250652410414196,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 15389
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.114787863876185,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 15390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4123356145482777,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 15391
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.604722676991284,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 15392
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.33910057319959963,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 15393
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.346718726975393,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 15394
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7306428223281523,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 15395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1345593893714265,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 15396
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5153691784263674,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 15397
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4147930471096832,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 15398
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26565244476802424,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 15399
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.658069403011417,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 15400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.66517638247241,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 15401
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0706376881992212,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 15402
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.938889871937866,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 15403
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2574735860154966,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 15404
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.332285523965895,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 15405
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9142488794964043,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 15406
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2357807122669602,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 15407
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8939878987425638,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 15408
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17423305067890343,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 15409
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7175249357243485,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 15410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6678174417017602,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 15411
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1007038313943065,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 15412
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.641999459139959,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 15413
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6444262794788264,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 15414
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6231751504228592,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 15415
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0490558171109132,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 15416
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.35201704522337035,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 15417
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8922298094008675,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 15418
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8022707228963425,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 15419
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.14394161085030235,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 15420
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5246646502610365,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 15421
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6175780931505344,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 15422
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5696039893816962,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 15423
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8782353648212333,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 15424
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0969181289444396,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 15425
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6393208724643686,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 15426
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.90471017900692,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 15427
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.369978060780718,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 15428
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1112243925733563,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 15429
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1767051263210955,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 15430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2379596472503582,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 15431
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11520125026152744,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 15432
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24846769665195062,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 15433
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3254532813114115,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 15434
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1311626653522275,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 15435
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7203254957945875,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 15436
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.105671119094162,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 15437
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2339384279363634,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 15438
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15554302731961248,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 15439
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.544174688818993,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 15440
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3962247003643524,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 15441
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4761293291411124,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 15442
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6266428812650042,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 15443
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0585186997111653,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 15444
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.150900576738875,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 15445
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5868284699394577,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 15446
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5355200818550058,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 15447
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2892055888096356,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 15448
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6153476587721176,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 15449
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.179427151281167,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 15450
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3155775638852413,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 15451
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8351554777821609,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 15452
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.15740084688823,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 15453
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.35227733628971886,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 15454
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13527723999363134,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 15455
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0014026460718144,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 15456
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2530480415894876,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 15457
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2068357127843057,
+      "learning_rate": 2e-05,
+      "loss": 0.0804,
+      "step": 15458
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6438484274050843,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 15459
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8671494484060738,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 15460
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9559573006052333,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 15461
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.9152877625859097,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 15462
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.32930141685675096,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 15463
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2553547606264397,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 15464
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.00638457252844,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 15465
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2480427200176583,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 15466
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.188388241296975,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 15467
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7865169245659236,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 15468
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7120084556093575,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 15469
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49755299333859004,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 15470
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1713860050489324,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 15471
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6904939301005812,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 15472
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9758230978060567,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 15473
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3340113779495629,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 15474
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7045911725010934,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 15475
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4337287860022632,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 15476
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8377891781444728,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 15477
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9821035646351406,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 15478
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9174568209765096,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 15479
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.21989275263500577,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 15480
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.610709097520468,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 15481
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.29208817706306944,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 15482
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5541668449726631,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 15483
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.187680700195598,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 15484
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9310736700467156,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 15485
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4053212383695017,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 15486
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0969505330564995,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 15487
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0086957033427568,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 15488
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.986737978486936,
+      "learning_rate": 2e-05,
+      "loss": 0.095,
+      "step": 15489
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.481008512599259,
+      "learning_rate": 2e-05,
+      "loss": 0.1246,
+      "step": 15490
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6946744641100369,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 15491
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6576894006147929,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 15492
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7633225991051001,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 15493
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9405712999495832,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 15494
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.579722146737923,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 15495
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1585360944193532,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 15496
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2652588733263931,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 15497
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.387877332224858,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 15498
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4624825533616903,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 15499
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0619088379585992,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 15500
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0955256082004479,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 15501
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5915789564919842,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 15502
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2184170488244277,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 15503
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.602299993684627,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 15504
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.560203267312675,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 15505
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6521822081615134,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 15506
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0234904042990376,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 15507
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0621943314622457,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 15508
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4075681319646862,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 15509
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8385002916847808,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 15510
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2611710382756812,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 15511
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3964908199561725,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 15512
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0844550133463282,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 15513
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8010051135258321,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 15514
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6817985575700287,
+      "learning_rate": 2e-05,
+      "loss": 0.107,
+      "step": 15515
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.539906992223197,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 15516
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.256787895701255,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 15517
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1835152880153017,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 15518
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8766683150036327,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 15519
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43893208450437726,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 15520
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5580968997041864,
+      "learning_rate": 2e-05,
+      "loss": 0.1147,
+      "step": 15521
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8704216592431556,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 15522
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4674171253430383,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 15523
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7672226254877014,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 15524
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.162498046344911,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 15525
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7936290002945268,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 15526
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3242236562092233,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 15527
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.22751466677620916,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 15528
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9801498745638411,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 15529
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.468480430376518,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 15530
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7525754097728028,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 15531
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.255742344772456,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 15532
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4898394740122827,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 15533
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6285930670947506,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 15534
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.711234670094371,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 15535
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.175394732731184,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 15536
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1281318516404317,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 15537
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8809888791043794,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 15538
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2437301159964747,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 15539
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.37289475619826,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 15540
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2859281256968242,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 15541
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0560844517072954,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 15542
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8302034171296705,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 15543
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.608883806480167,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 15544
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5122650957242981,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 15545
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0859198808955954,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 15546
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6693522308566043,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 15547
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7613680674477603,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 15548
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.471643152058105,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 15549
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8457005172583568,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 15550
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9503495905582222,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 15551
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0325820296355783,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 15552
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2656717539760487,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 15553
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2573744206452317,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 15554
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.233383705747763,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 15555
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.086646198672819,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 15556
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4709333059827743,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 15557
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8319396922280584,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 15558
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6827697861828264,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 15559
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.504582399241051,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 15560
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.20092613989010794,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 15561
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6060541337969921,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 15562
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.29138055744378083,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 15563
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4200259751643987,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 15564
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.021188553989055,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 15565
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4913185562946513,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 15566
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3713853333121238,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 15567
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9580671342887865,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 15568
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31623289280635525,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 15569
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6301757604414852,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 15570
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9845015414108254,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 15571
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2502268585339553,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 15572
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.154263230698299,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 15573
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3618948913366506,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 15574
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2259971733009927,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 15575
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5457321723404733,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 15576
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.713153559972276,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 15577
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7197115312970975,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 15578
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7613954689457201,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 15579
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7265743922108134,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 15580
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2386300712254084,
+      "learning_rate": 2e-05,
+      "loss": 0.1104,
+      "step": 15581
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6156937594670135,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 15582
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16032274245099282,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 15583
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2957365715395288,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 15584
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.34531322410127835,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 15585
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5430677937456185,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 15586
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45874426222287074,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 15587
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7767274068055947,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 15588
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2229577332323235,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 15589
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5501340961901537,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 15590
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8929470324132507,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 15591
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.891841158333939,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 15592
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5253344651168083,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 15593
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5147408334870545,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 15594
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.44274660472265,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 15595
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1460716635481292,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 15596
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.356476736889363,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 15597
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4225460606846048,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 15598
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9402096373057831,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 15599
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3390538859674528,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 15600
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3382393134819178,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 15601
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8422533199856552,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 15602
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8648289404396247,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 15603
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4953558701273657,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 15604
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.818470445798414,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 15605
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.028507462029982,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 15606
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.678906778270697,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 15607
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.729698565759684,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 15608
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1465154741971327,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 15609
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6385650545434857,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 15610
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1636591235844125,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 15611
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8853505203372216,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 15612
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0694270125670708,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 15613
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0958208914994472,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 15614
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.625074164825554,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 15615
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.544939590001202,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 15616
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0752265225765172,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 15617
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7695601404643856,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 15618
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0773605179398118,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 15619
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8371705804452243,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 15620
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8661249561155218,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 15621
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7195602493986507,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 15622
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.508665802003953,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 15623
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.47451561523646507,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 15624
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8722555512710567,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 15625
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2877312002363155,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 15626
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.274078823135689,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 15627
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7825849571664779,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 15628
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.517574526338099,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 15629
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3940595582419206,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 15630
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7796230159536721,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 15631
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1043360676575715,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 15632
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6994705634789944,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 15633
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.777755052533867,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 15634
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1636983779680725,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 15635
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.701046436472098,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 15636
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9527673832116595,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 15637
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.303819619064785,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 15638
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.437506162195486,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 15639
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.18471684769728805,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 15640
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1970308034245423,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 15641
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0129551903715748,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 15642
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.067192654017769,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 15643
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.718915203131155,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 15644
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.22571333883539174,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 15645
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.718403857266026,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 15646
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6005273218227176,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 15647
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7975751007976324,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 15648
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2220777139629698,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 15649
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.592425364318473,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 15650
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6739903552781328,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 15651
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0640687467378034,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 15652
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.619681605027114,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 15653
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2341228718434326,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 15654
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.687704446044524,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 15655
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4986484319458826,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 15656
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4573690303289064,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 15657
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5309517239036188,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 15658
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2447116080623606,
+      "learning_rate": 2e-05,
+      "loss": 0.1556,
+      "step": 15659
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.602995881430921,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 15660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4468415083852455,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 15661
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4354975477813534,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 15662
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9236617795451719,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 15663
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31946804801955586,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 15664
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5156629891894524,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 15665
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.162265292975076,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 15666
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.18367785698395261,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 15667
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8676378609996026,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 15668
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8592612104508687,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 15669
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1944973881554506,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 15670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.871285701454456,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 15671
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3062510213875513,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 15672
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6745474303605437,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 15673
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.922302541205134,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 15674
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.37199116663545867,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 15675
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6504971622187254,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 15676
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4570164311280693,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 15677
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.004885789097589,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 15678
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7493987057255487,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 15679
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4812590139956794,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 15680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0385683648680228,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 15681
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0493008452122549,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 15682
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2970283141916916,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 15683
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06345044112101014,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 15684
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0808177988254137,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 15685
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5263494586952242,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 15686
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4227986291610524,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 15687
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2414899208841396,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 15688
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11089557977254802,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 15689
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3777637449356588,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 15690
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0037670421768006,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 15691
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.150415753800913,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 15692
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6112819595659402,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 15693
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9521930149440745,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 15694
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7798551485508582,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 15695
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3567532216672251,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 15696
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.038037647979466,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 15697
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6811393490443933,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 15698
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.064162860456295,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 15699
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.519674024763379,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 15700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7252658445644764,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 15701
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2171740240503413,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 15702
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4710745576729916,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 15703
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3987028413651676,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 15704
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.475579474464668,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 15705
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.38892149225308,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 15706
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5719108653677776,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 15707
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9492633950366111,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 15708
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7358458783890115,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 15709
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0385665582752774,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 15710
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9676596028836986,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 15711
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.47485997997286944,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 15712
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.25961547470159135,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 15713
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.063960742968927,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 15714
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3956893685338494,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 15715
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6742097741299768,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 15716
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.309354343760694,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 15717
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5193187888703132,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 15718
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6576787744726491,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 15719
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9202926115467921,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 15720
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6002384327954029,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 15721
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.005178335793078,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 15722
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.021429766963199,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 15723
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8164517330659864,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 15724
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7030590639562346,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 15725
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.30672359915863,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 15726
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0831118809719409,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 15727
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4548872354963139,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 15728
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.36885511872022275,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 15729
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7140082407233572,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 15730
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5897793271133818,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 15731
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1814444512642113,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 15732
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1996892447234113,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 15733
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5851455740759501,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 15734
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27704324695237614,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 15735
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0105972259800486,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 15736
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.16242086900039,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 15737
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7980348336856006,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 15738
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8915832235050942,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 15739
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.71591134794946,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 15740
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1329705489826558,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 15741
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7097486432813787,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 15742
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1623444647978463,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 15743
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.07240126702484,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 15744
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.224683869402149,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 15745
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2581833476348208,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 15746
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.045160398265461,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 15747
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.497497345035137,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 15748
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4963703735950585,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 15749
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.947366010096299,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 15750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.09565344318906638,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 15751
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8345205115831065,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 15752
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3933268670262104,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 15753
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2708435706365138,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 15754
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4346147150222133,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 15755
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.36900857417849,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 15756
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5550548177053842,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 15757
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6113761386075731,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 15758
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8568808341018987,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 15759
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0519049213072724,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 15760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.319743513489883,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 15761
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2734888920623442,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 15762
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9961457925625927,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 15763
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5397789541359816,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 15764
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31711849063271824,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 15765
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8022377594801877,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 15766
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.954195995422514,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 15767
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6745557594481153,
+      "learning_rate": 2e-05,
+      "loss": 0.1033,
+      "step": 15768
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.145169928299429,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 15769
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7368981055082251,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 15770
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.24322106179750358,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 15771
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.567839235773929,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 15772
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.37772651750978176,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 15773
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44862861016862454,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 15774
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7234531030146357,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 15775
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.91359395465095,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 15776
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8504243997324744,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 15777
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.803153306550452,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 15778
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6710456564460048,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 15779
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6828534504637558,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 15780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.9836373575577815,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 15781
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.36263969130475776,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 15782
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.759429716818413,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 15783
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9823715429070827,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 15784
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4360695231351823,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 15785
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4944304871315566,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 15786
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4070220303258294,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 15787
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.155723812898278,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 15788
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.23268769394392344,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 15789
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.22424442662614083,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 15790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4273935641139339,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 15791
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4050152337955524,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 15792
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4939946121976595,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 15793
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5941200859738214,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 15794
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.153032310502928,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 15795
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8684905289049544,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 15796
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.32201305086257487,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 15797
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0717077141391156,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 15798
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1647501201085089,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 15799
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3019148161991369,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 15800
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.573423274562137,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 15801
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5059920473939976,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 15802
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.598417707071098,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 15803
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.72125646926989,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 15804
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2264756092954043,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 15805
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2202973446655894,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 15806
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2188928072461525,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 15807
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3360602815440292,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 15808
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5115580313164803,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 15809
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.116756719140422,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 15810
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07807690510933347,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 15811
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1780463718576256,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 15812
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4235823837381951,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 15813
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6281779831595748,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 15814
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1156598296725773,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 15815
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.068774255960285,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 15816
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5047334490665627,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 15817
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4796060348369109,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 15818
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7689737853810024,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 15819
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3225203794124238,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 15820
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7301861221338353,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 15821
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6278057421206387,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 15822
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9938247268699029,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 15823
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3666163532175788,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 15824
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.066378973867059,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 15825
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2998950461739141,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 15826
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3476813341980933,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 15827
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7461282212847223,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 15828
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2355447851760948,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 15829
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.880151216568749,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 15830
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6372046354975462,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 15831
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9628964127991462,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 15832
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4535446689975364,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 15833
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14860905596526683,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 15834
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7048525251190293,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 15835
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.374705713456129,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 15836
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4385975503897206,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 15837
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.393487134109977,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 15838
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6459551737951634,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 15839
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1366408803164612,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 15840
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3572907546043803,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 15841
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6791007508929296,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 15842
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4674269601557084,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 15843
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.49190714586875633,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 15844
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6267305842463693,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 15845
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9544165041487045,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 15846
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.536002665909798,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 15847
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6666560743404367,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 15848
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8279211045216491,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 15849
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.895295797925046,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 15850
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8986985532230911,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 15851
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2704980468539984,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 15852
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0506597966338267,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 15853
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4476607757514086,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 15854
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.23059816283152162,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 15855
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9557821725114677,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 15856
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3528330128877593,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 15857
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.455550226764175,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 15858
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8671902493270044,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 15859
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5013929892613476,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 15860
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7713668154225258,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 15861
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.04346677351695,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 15862
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1305005388726794,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 15863
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9729714859433453,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 15864
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9995887792401574,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 15865
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5362575938671077,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 15866
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8869694450423509,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 15867
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9209563640608003,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 15868
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9070467933376625,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 15869
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9986528389452178,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 15870
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.435432926990624,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 15871
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3861790351289427,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 15872
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0179339645192247,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 15873
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3421641462112825,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 15874
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3743140782125377,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 15875
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.818393466502707,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 15876
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3814148817428051,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 15877
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.203679093424491,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 15878
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1573655700892713,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 15879
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.483631630524907,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 15880
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5431337448926182,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 15881
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.025371790534414,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 15882
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.098787912228964,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 15883
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8210704390014306,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 15884
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6737150014375534,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 15885
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1451485176881977,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 15886
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48168485922539334,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 15887
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.832347147257757,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 15888
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3777159397329253,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 15889
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.594993043025401,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 15890
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0388391598981226,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 15891
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7487773029849114,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 15892
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1592806713463635,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 15893
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.818974736017792,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 15894
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2793288179284323,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 15895
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48293772619712866,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 15896
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6558101178882801,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 15897
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9896910340349805,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 15898
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9867873227657753,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 15899
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4769075182330578,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 15900
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48456851108620563,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 15901
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1943412856703136,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 15902
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.087578555810521,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 15903
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9729175446456662,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 15904
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.24585001092175268,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 15905
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8457175279489288,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 15906
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3835615303819201,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 15907
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.801795935280023,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 15908
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33765212224637675,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 15909
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9879146695180523,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 15910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.45700997523161846,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 15911
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6607255027028875,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 15912
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.19074643734774,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 15913
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.144725797684501,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 15914
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.55292430171961,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 15915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3828070257908285,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 15916
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.776771797702525,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 15917
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06295379167564964,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 15918
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.19796082773592855,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 15919
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4852194542356325,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 15920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14775395491070398,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 15921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.193887929308176,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 15922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3911758789032302,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 15923
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5050814033277996,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 15924
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1026079884449216,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 15925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3730006586291736,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 15926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3123891420322733,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 15927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1040429939730125,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 15928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7352538024845524,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 15929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0081290997151628,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 15930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.843356430150711,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 15931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.114589685202114,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 15932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6319283796604356,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 15933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1388307856389779,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 15934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3353216598929665,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 15935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.408460130412381,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 15936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5521361884009064,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 15937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7714345771494373,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 15938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8962775607737112,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 15939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8386027904212974,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 15940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.536317730520988,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 15941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2889423013317054,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 15942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.005385826090256,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 15943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7077763882955582,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 15944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.11141033171366,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 15945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.811110513567285,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 15946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9615772877703773,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 15947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.099123696756691,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 15948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.32548372462494,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 15949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8184940931524756,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 15950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9419602153176168,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 15951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5673010236977322,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 15952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9905033800354444,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 15953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.380409955347022,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 15954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.6138745087158295,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 15955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9098073865454452,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 15956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43040084580219573,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 15957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07357613091937154,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 15958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4628365968880277,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 15959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5758664497444244,
+      "learning_rate": 2e-05,
+      "loss": 0.1018,
+      "step": 15960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4020835813050502,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 15961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4037026114884996,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 15962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6410505923527698,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 15963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.93238385034708,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 15964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.775406229549176,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 15965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5632233280582848,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 15966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43983126160151914,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 15967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6948736651215797,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 15968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1729897347023033,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 15969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4737890930426525,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 15970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.19916606377292,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 15971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.333743206246403,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 15972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7071954104545382,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 15973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7460468537136902,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 15974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6625668963977234,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 15975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2225287130217424,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 15976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8808425791654997,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 15977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4384588278035242,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 15978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5277252145223055,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 15979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9752223481189826,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 15980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3494915834908563,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 15981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2727838722716562,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 15982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2377162040380485,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 15983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7391535535535874,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 15984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5944965588026495,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 15985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4157581485112263,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 15986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.108387545879049,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 15987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4929049273737805,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 15988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3108391854998165,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 15989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6864065013722654,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 15990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5511117579322184,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 15991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8481799850125329,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 15992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.562728254812292,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 15993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.057254867811231,
+      "learning_rate": 2e-05,
+      "loss": 0.1045,
+      "step": 15994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.688413420412712,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 15995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5333738584277068,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 15996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8303724059007406,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 15997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3240588954588963,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 15998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.378853833352382,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 15999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9791501057353846,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 16000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9708063332149766,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 16001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.101488088488671,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 16002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3459835100806158,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 16003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3008680961373085,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 16004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6767122107440653,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 16005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.172994612864346,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 16006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34342854815297813,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 16007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.053840862533441,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 16008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1120905883617203,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 16009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5083523115084667,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 16010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34519371388906855,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 16011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2980285937123688,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 16012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.623925216080157,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 16013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9533609817292528,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 16014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3590831193240904,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 16015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.14364813887812,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 16016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3407191041568279,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 16017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1950968696470688,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 16018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7575380916455852,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 16019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3111761773841257,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 16020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3652434934406009,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 16021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.066499935110457,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 16022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.042721771215279,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 16023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5078836537909447,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 16024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.266681398549544,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 16025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.23608376346171475,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 16026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5073324428622384,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 16027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2774834348270752,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 16028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7313400382855981,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 16029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7173065535519986,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 16030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6943517326350394,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 16031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1881282449200827,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 16032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.326806503267281,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 16033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.23885852426834212,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 16034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.20510717821953575,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 16035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9445845583377854,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 16036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9590039085867166,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 16037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4747973997986574,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 16038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6867727902666343,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 16039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1010681558175648,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 16040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.235171836204013,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 16041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3268676024879578,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 16042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.448455362713901,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 16043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2908444765400955,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 16044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2717913419686457,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 16045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.283403608071406,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 16046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5637116121479007,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 16047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9683684327465427,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 16048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6876270220195788,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 16049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2614258700833898,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 16050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2369005561726625,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 16051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4297562045906391,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 16052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.481405534842043,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 16053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8780360119853514,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 16054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3462547201571515,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 16055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4104254093108595,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 16056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.76562359919293,
+      "learning_rate": 2e-05,
+      "loss": 0.0969,
+      "step": 16057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7939030447168631,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 16058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5921095839313617,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 16059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6056215647156152,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 16060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5278749924668076,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 16061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4217467642966958,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 16062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5497409112776621,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 16063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7636963429660022,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 16064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9757652385244032,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 16065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7326310736393528,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 16066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1889299858332884,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 16067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9123088610354989,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 16068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.368393415223118,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 16069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.47919465285437,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 16070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.19241779996806116,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 16071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06652942271587889,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 16072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9739799666079025,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 16073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6363961040044894,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 16074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7300538297297886,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 16075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3002739008610489,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 16076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9734423674487243,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 16077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2390671892305902,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 16078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9092021738584404,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 16079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5130339034846088,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 16080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.35495582715420826,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 16081
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2889523545804709,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 16082
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.552601373734912,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 16083
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.318126787932383,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 16084
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.408090700696059,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 16085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1185028762234714,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 16086
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3234252872096297,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 16087
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4068968238985058,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 16088
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.885072532795747,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 16089
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.35457481913462124,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 16090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.237139114052758,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 16091
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13879346889390598,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 16092
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3179243778507642,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 16093
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.743916480720314,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 16094
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8538043059898377,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 16095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5193885090857915,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 16096
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.10559434936654,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 16097
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.392535396607724,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 16098
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.784290824980906,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 16099
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.066051777252705,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 16100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34000247230402275,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 16101
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5616141287570977,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 16102
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6886429624709773,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 16103
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6158100103406996,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 16104
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9200792965507856,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 16105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.530131253828458,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 16106
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.576444656610512,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 16107
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9409628492657256,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 16108
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.358277829958274,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 16109
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8595967786892779,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 16110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8009655376146934,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 16111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.374647008495082,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 16112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9642822237809399,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 16113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6690755205159007,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 16114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9667073506926323,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 16115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5054128259348856,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 16116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.474695217748208,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 16117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3998044541578176,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 16118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.386099448735762,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 16119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2871751774747919,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 16120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.313927056298319,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 16121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.14601614707283,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 16122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9432737260246383,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 16123
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28405490246087417,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 16124
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6190166401372705,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 16125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8669375328111724,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 16126
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9039117639846572,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 16127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.998252880406007,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 16128
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8061765662448808,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 16129
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.994675553949259,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 16130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5630223180402529,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 16131
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.752180991136566,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 16132
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06021599767249297,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 16133
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.435490097448665,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 16134
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5198501213115096,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 16135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2337600045151205,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 16136
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.089164495255563,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 16137
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7725515295159768,
+      "learning_rate": 2e-05,
+      "loss": 0.0862,
+      "step": 16138
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7222416697562193,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 16139
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9743345251479567,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 16140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.791620956350329,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 16141
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8053928728004739,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 16142
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.194829873964949,
+      "learning_rate": 2e-05,
+      "loss": 0.116,
+      "step": 16143
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6282908056842846,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 16144
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5688778848761424,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 16145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6017709735933892,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 16146
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.519020576982753,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 16147
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7170696103476545,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 16148
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.839519330033653,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 16149
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4599370131557738,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 16150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7411795809624004,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 16151
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6044117222293606,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 16152
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3620836741782612,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 16153
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.707067284105515,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 16154
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6760762403216314,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 16155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2176933109258554,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 16156
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.638437980730097,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 16157
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0287114311580658,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 16158
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3322030233561055,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 16159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.055772270199841,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 16160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5493022187838006,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 16161
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8485695274294116,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 16162
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.35356462254998916,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 16163
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.007427965441846,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 16164
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.225451275784283,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 16165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7309131615380626,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 16166
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3538203251928587,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 16167
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9553912633219421,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 16168
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.099387840618251,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 16169
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3747438278850517,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 16170
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1419676625237953,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 16171
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.57194449420905,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 16172
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3813667048387053,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 16173
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6563087164632926,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 16174
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.22191956435200327,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 16175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.527786720334657,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 16176
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9820542824075313,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 16177
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6426729045652462,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 16178
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6617404281145883,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 16179
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5305327734985156,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 16180
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7099733466586429,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 16181
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7367469547572529,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 16182
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9152418590432885,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 16183
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.015168150456107,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 16184
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.05613060230654,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 16185
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5359744968876847,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 16186
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7783672523220604,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 16187
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.932195292603192,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 16188
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6684022578668155,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 16189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8865885702017298,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 16190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5060892486267543,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 16191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13171729648145594,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 16192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5894003959903917,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 16193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4313041010463994,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 16194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9656645196238054,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 16195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.822278839423215,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 16196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.296146199346726,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 16197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0213429807173293,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 16198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7447403477955068,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 16199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.623094321110823,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 16200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.393772848651896,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 16201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.430932514084051,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 16202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7363678076964961,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 16203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8702396362694713,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 16204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8807914301720416,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 16205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7939790692421431,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 16206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14412997576030906,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 16207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.314080909382746,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 16208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8821894041845545,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 16209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8785106853214713,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 16210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0448212548743685,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 16211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9139858308345379,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 16212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.077312224494298,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 16213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9427447162723901,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 16214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6585639767663036,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 16215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7126210608468033,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 16216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9205998151950954,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 16217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0974498839873144,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 16218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7489524917127133,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 16219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0678666842508697,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 16220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9308519440819925,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 16221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9083067147848911,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 16222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5095747695049597,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 16223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8979099549280908,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 16224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.636368871874761,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 16225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4013161916891227,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 16226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.362937422311673,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 16227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6084353954668683,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 16228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3191731067981423,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 16229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0991008195876293,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 16230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.10292511418197,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 16231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4590014516706873,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 16232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9957721864343803,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 16233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0561535612640363,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 16234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9468272974569144,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 16235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.005529711556001,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 16236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.21650490560546165,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 16237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3516475256781147,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 16238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.551732412218799,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 16239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4938487974141993,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 16240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5291238025969203,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 16241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8103161586880425,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 16242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.135669204085961,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 16243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.47023973567433924,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 16244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7301393878148065,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 16245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0694299651323135,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 16246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11925428472124802,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 16247
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2724137897784925,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 16248
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3060310892169764,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 16249
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5602329592502324,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 16250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7734327219805857,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 16251
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2915248711385146,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 16252
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17577138449217106,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 16253
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6747441354350444,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 16254
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.061903665047906,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 16255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7929548440850813,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 16256
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1134016540735703,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 16257
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6031611614554044,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 16258
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9747619509476761,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 16259
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.36217224908535645,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 16260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2833798082201608,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 16261
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8898616330139123,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 16262
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5189464116829633,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 16263
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8098774020510016,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 16264
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4327586267868688,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 16265
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1956882260281334,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 16266
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4889583806152,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 16267
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.351518531040018,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 16268
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1683887249707843,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 16269
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.350285509994037,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 16270
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9582661218134176,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 16271
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0024728984165607,
+      "learning_rate": 2e-05,
+      "loss": 0.0959,
+      "step": 16272
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8267851012385661,
+      "learning_rate": 2e-05,
+      "loss": 0.1039,
+      "step": 16273
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2554007682127013,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 16274
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6719521477983716,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 16275
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2041390651378845,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 16276
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3152578250681515,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 16277
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9403474682854966,
+      "learning_rate": 2e-05,
+      "loss": 0.121,
+      "step": 16278
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.584198519713284,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 16279
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40115963658104403,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 16280
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7395713876825128,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 16281
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3861445137939379,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 16282
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8467244224031951,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 16283
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5664861568322648,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 16284
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1702772310898908,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 16285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.31460070256835615,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 16286
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.385329590357906,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 16287
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.827293770552949,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 16288
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7504344076890392,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 16289
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7897500545722674,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 16290
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6962781479931072,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 16291
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6331153844235693,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 16292
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.810144933303803,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 16293
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1774874755776437,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 16294
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0150390695465834,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 16295
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8168268899679685,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 16296
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7558389113857212,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 16297
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.972152736347154,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 16298
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5802702922600987,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 16299
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3248341490955395,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 16300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.36680528348764746,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 16301
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2925179147862265,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 16302
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14613699731899393,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 16303
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.818369280406938,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 16304
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.608819365886903,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 16305
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5343900890339289,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 16306
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.155347368675323,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 16307
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8839431544060129,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 16308
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4791368114341883,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 16309
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.519250054315965,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 16310
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5251716329578282,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 16311
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.254112012861502,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 16312
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.46334957757553713,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 16313
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4251841286488153,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 16314
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.35705338958276633,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 16315
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1570985103532643,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 16316
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.533827112644374,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 16317
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6590288710010634,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 16318
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5593078082793266,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 16319
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3760217425020282,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 16320
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7503936188665188,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 16321
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9426136109762184,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 16322
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.34073615869323,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 16323
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9287621837961725,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 16324
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4441438532219609,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 16325
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7989196975326185,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 16326
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.507484866217358,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 16327
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.5310968927432524,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 16328
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0784702024804602,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 16329
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.319547171804382,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 16330
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4826682467783825,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 16331
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6715541068176774,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 16332
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1266076233818225,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 16333
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0544710290364883,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 16334
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6465011778892296,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 16335
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.22572614012247,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 16336
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.668823093670677,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 16337
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.195272457474337,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 16338
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2650137151222696,
+      "learning_rate": 2e-05,
+      "loss": 0.1118,
+      "step": 16339
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.748450069706165,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 16340
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5687034296061138,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 16341
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1102880025233828,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 16342
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7288408141698363,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 16343
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4313612349150389,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 16344
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1595245593892802,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 16345
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13242486173924292,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 16346
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0895409480544265,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 16347
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.35477369076267695,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 16348
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1789609174648954,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 16349
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4077389464420852,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 16350
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5765614837552473,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 16351
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3115073492067497,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 16352
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.620585953803457,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 16353
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2625191817768394,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 16354
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2895784284859055,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 16355
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6741643187432211,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 16356
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9689559828767385,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 16357
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9133704978155723,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 16358
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.196536771477352,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 16359
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16722051726733175,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 16360
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5808842021028529,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 16361
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.588653854772696,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 16362
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4600025796214269,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 16363
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.326249267624804,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 16364
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1897066594658503,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 16365
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.311788684100806,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 16366
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.309579101262674,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 16367
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12319090149401044,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 16368
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3234394672473757,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 16369
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1941636991408524,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 16370
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1496910756088256,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 16371
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.196025538644244,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 16372
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9728477707070455,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 16373
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9521123473441384,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 16374
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1380999062634891,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 16375
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1672464644361569,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 16376
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.368193641762923,
+      "learning_rate": 2e-05,
+      "loss": 0.089,
+      "step": 16377
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3119536579716844,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 16378
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8202459723485878,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 16379
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24498586442369208,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 16380
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.252064922962109,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 16381
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7147259856434975,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 16382
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28578629847468645,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 16383
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3657087988983982,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 16384
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5500471335150245,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 16385
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3097822384510185,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 16386
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.23475619946186227,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 16387
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2913821994136605,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 16388
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1995777996731045,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 16389
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7505831835277101,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 16390
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.196635217726111,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 16391
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7488877075242907,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 16392
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5176904506672733,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 16393
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.793099990728824,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 16394
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.307808798482949,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 16395
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1162586218660073,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 16396
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7778777392236637,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 16397
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7080884928593356,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 16398
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.140981157286755,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 16399
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.504243110932879,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 16400
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7747366427986724,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 16401
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3654360642664178,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 16402
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9740760452750261,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 16403
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3443642568383718,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 16404
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.370415601717791,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 16405
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28338563234199143,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 16406
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8433121498812279,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 16407
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.239090423461762,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 16408
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7588089894675558,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 16409
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.049860284277588,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 16410
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3143377154218618,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 16411
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.651962320264337,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 16412
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8967909943053586,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 16413
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2284156261312278,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 16414
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7492509916526181,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 16415
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6750221840427275,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 16416
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9478615370987473,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 16417
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2115877333085256,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 16418
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1703070259730175,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 16419
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5499915122753447,
+      "learning_rate": 2e-05,
+      "loss": 0.0953,
+      "step": 16420
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4758870835570872,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 16421
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39098772374318114,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 16422
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.10483022628892259,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 16423
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7135541528767178,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 16424
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1909239246052128,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 16425
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1764725588683846,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 16426
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6200599407362584,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 16427
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.44863382487998665,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 16428
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4619303710680185,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 16429
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.705468271895869,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 16430
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9282120930302915,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 16431
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2195224759203507,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 16432
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4823110970633304,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 16433
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3122457076278837,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 16434
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0417025305614072,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 16435
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2623701103467244,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 16436
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.05952812678800039,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 16437
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2225854523818986,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 16438
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6212149207958891,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 16439
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.50915323027093,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 16440
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5140488260415865,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 16441
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4259026462660127,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 16442
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.465651632689208,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 16443
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1028616528732957,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 16444
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.35368046157960165,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 16445
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6318291436264545,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 16446
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8372806262418375,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 16447
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7700482098857502,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 16448
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2042922760723747,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 16449
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1696534015924858,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 16450
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7960120741304506,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 16451
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7416505979733193,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 16452
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3022291585009145,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 16453
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7983097549041258,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 16454
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8015767158946587,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 16455
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9205849236222843,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 16456
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2514506505547576,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 16457
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.04752398539802217,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 16458
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.523196654739775,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 16459
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.21634039203804442,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 16460
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.47583796972676035,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 16461
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2472265947676384,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 16462
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2841117362768415,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 16463
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2081515279343047,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 16464
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4606557243431637,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 16465
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0703414578940762,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 16466
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5147995844465859,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 16467
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4650837414974455,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 16468
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2711663148655363,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 16469
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.114094741585476,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 16470
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0404921713245354,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 16471
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17087423970709809,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 16472
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.744776696288,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 16473
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5051495849689904,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 16474
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.149962211070944,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 16475
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1821469561348695,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 16476
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9762833158056576,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 16477
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6016310367927312,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 16478
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.99408006716249,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 16479
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.754996659582714,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 16480
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3130421346672524,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 16481
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5301774007943427,
+      "learning_rate": 2e-05,
+      "loss": 0.1008,
+      "step": 16482
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3915138622073284,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 16483
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3275104671719353,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 16484
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8950743850360983,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 16485
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9755776103924234,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 16486
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6152129154398541,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 16487
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2744599843545745,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 16488
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.550637210441887,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 16489
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.52366286024017,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 16490
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3366517951467825,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 16491
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5081479188684968,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 16492
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2976328003089168,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 16493
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1498686861257896,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 16494
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1615546946331317,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 16495
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4557970971742145,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 16496
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.178172360839842,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 16497
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9348863408165515,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 16498
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8209357301063467,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 16499
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.13304310205247,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 16500
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4952831014296324,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 16501
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6174698735162854,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 16502
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0642167415104273,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 16503
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7768204651032273,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 16504
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7342171735411855,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 16505
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6294998807061367,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 16506
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1834981646827742,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 16507
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2839968854770922,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 16508
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5057291872264833,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 16509
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1625332374484714,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 16510
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6561368507746128,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 16511
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0633524672561545,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 16512
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5810634757527833,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 16513
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.961307299410465,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 16514
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.434525937717791,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 16515
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1681160722271977,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 16516
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6533027005304333,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 16517
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3961200266150695,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 16518
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1160423972577416,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 16519
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5058450586362491,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 16520
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.35675507763602,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 16521
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6571339376174484,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 16522
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.14536247404602,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 16523
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.861082559179936,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 16524
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6385739452631578,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 16525
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.573290378393669,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 16526
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1707648041151864,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 16527
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3197610690838269,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 16528
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.23923286775058497,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 16529
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5053295702134752,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 16530
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7740764483295103,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 16531
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8434813998851227,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 16532
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.204930672056017,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 16533
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.27979211076994,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 16534
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5481223458864162,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 16535
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8982586560563786,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 16536
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6627195579260181,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 16537
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06981857983954567,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 16538
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8028372074540818,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 16539
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4431388387103263,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 16540
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.017971949440033,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 16541
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4008927342931365,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 16542
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.633202317354729,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 16543
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6167859328340018,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 16544
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.523048394499949,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 16545
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8076550475982612,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 16546
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.375645139374342,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 16547
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8366832760557734,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 16548
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8660751632025625,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 16549
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5420296320996918,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 16550
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.33321096462719,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 16551
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8642485020022863,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 16552
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9578318458582308,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 16553
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9390129900447626,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 16554
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.23478296060394668,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 16555
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.555717625387331,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 16556
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7297865185749779,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 16557
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6516752872685482,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 16558
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7688049141723816,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 16559
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.267182698458266,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 16560
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5247280729810844,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 16561
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0685856364898076,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 16562
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8344809090553318,
+      "learning_rate": 2e-05,
+      "loss": 0.0763,
+      "step": 16563
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7084683728989614,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 16564
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5928143356769395,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 16565
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7441783104512817,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 16566
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8214843513396874,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 16567
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.162882571693856,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 16568
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6264360618966813,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 16569
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5484682308183914,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 16570
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.614976932007719,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 16571
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1669705128971803,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 16572
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9768251387814344,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 16573
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9179070923134138,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 16574
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8147440205964387,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 16575
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7064693495377106,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 16576
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.22946448463269103,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 16577
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.409680089176078,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 16578
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5566748357599812,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 16579
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.31882964338988,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 16580
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0687782819251694,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 16581
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5805748534779678,
+      "learning_rate": 2e-05,
+      "loss": 0.0775,
+      "step": 16582
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3071242550755406,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 16583
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.972812276278537,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 16584
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0141945963022585,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 16585
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8672521068803625,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 16586
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3696997107313519,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 16587
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.560969873699601,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 16588
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.278811573339999,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 16589
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9423165886098209,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 16590
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.515917798891152,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 16591
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8229668383278914,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 16592
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0949093343354767,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 16593
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.541061600335903,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 16594
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7417866477642959,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 16595
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7979653131268005,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 16596
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0074668113978322,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 16597
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0635870813928734,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 16598
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1612446107934373,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 16599
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8871251321969613,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 16600
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43949142289173576,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 16601
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32463234637676375,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 16602
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4689716415368699,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 16603
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7023193830394023,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 16604
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3425576886332853,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 16605
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1121914532155208,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 16606
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0882203511590365,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 16607
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3671502845247713,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 16608
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8471783464219383,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 16609
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6829396415377338,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 16610
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5504286050409493,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 16611
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1386028288490673,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 16612
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9720296638880757,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 16613
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.937533423872276,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 16614
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.124963086873499,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 16615
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.875284489296168,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 16616
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.981680294446658,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 16617
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5046392033855147,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 16618
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9413182704225131,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 16619
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5979567127795488,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 16620
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2124288766312867,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 16621
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.211467773685679,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 16622
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.37545391267756334,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 16623
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8800717741880582,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 16624
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5676477440748814,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 16625
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6601285985087584,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 16626
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4710169968464055,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 16627
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7911398281387063,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 16628
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8008686300993686,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 16629
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.529369832119791,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 16630
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8981692905678638,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 16631
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3597735392707508,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 16632
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.570970084446418,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 16633
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0460538917794036,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 16634
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4684608203256771,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 16635
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.45339459587619735,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 16636
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15662312659606833,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 16637
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9030090125554044,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 16638
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.061924542704334826,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 16639
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06060430471831134,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 16640
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2169041942859975,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 16641
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7047083934046442,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 16642
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.229912283878843,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 16643
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.794951402958656,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 16644
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7877274908676668,
+      "learning_rate": 2e-05,
+      "loss": 0.0773,
+      "step": 16645
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2019103718599644,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 16646
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.498506576795092,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 16647
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7516624780722877,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 16648
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0615154641561753,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 16649
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4046847714612527,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 16650
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.36888965774571547,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 16651
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9580358720844552,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 16652
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.118705339435646,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 16653
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0338304198233912,
+      "learning_rate": 2e-05,
+      "loss": 0.0943,
+      "step": 16654
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.283601771836576,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 16655
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5919318203553985,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 16656
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2615018959409154,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 16657
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0878953916489773,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 16658
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1404727677435085,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 16659
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.579177641549946,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 16660
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4805445991350872,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 16661
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1334401463682013,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 16662
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3146268116145214,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 16663
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1441808691530175,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 16664
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3391486497405691,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 16665
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2464799910193,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 16666
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7906490437489497,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 16667
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3973463692609094,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 16668
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9672685185580104,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 16669
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4488041664681091,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 16670
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6180680427885216,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 16671
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7016480510519597,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 16672
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.924768755334569,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 16673
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0247249557768114,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 16674
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0011540548619215,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 16675
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6785691907688054,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 16676
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5337183492687942,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 16677
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6576935694594899,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 16678
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9749458970440018,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 16679
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1717697655858786,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 16680
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6889064014444275,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 16681
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1445024382414535,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 16682
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.603451986594993,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 16683
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6047890077699598,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 16684
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8680940605963152,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 16685
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2079184040389817,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 16686
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2893094837668069,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 16687
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0087162076449514,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 16688
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7035514174299539,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 16689
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14037832506628833,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 16690
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9733472161620893,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 16691
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.259823112369441,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 16692
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.208958903063795,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 16693
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5823266293408627,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 16694
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7497730107915237,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 16695
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9805532584133205,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 16696
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3697816053052416,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 16697
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8221022563663464,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 16698
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0876817589854393,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 16699
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0968435951366917,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 16700
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4734757782471922,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 16701
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0114012236137118,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 16702
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7127618960856614,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 16703
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5128485821772863,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 16704
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.176486136705369,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 16705
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9944700583356529,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 16706
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2676553805515314,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 16707
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5124017074578102,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 16708
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.985766942936396,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 16709
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2937157078128105,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 16710
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.20575910764828956,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 16711
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0847009438286317,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 16712
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6776373127936766,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 16713
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9149724881147837,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 16714
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2574431965420554,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 16715
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0709882955579426,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 16716
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.72995449028353,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 16717
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6224668271945284,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 16718
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.32602718136249775,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 16719
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0059344937857206,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 16720
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.770541201993018,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 16721
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8274968121861946,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 16722
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6822194023870174,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 16723
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9069349590465998,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 16724
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6960777981296582,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 16725
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4002301150260371,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 16726
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.864186725326088,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 16727
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.084729187916299,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 16728
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8812380577386048,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 16729
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3991916264088775,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 16730
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3436081390018506,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 16731
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.868509439148002,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 16732
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.700869109534629,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 16733
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.394484472446324,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 16734
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.259475172760705,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 16735
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7279511817862079,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 16736
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2217678983174953,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 16737
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4966283097198296,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 16738
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5505266340884243,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 16739
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9784178940339774,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 16740
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.459210180166713,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 16741
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2525231765742418,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 16742
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3703257192191771,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 16743
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8330885209655985,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 16744
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.114918749535942,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 16745
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7831075329366267,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 16746
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1720680077874515,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 16747
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06700664022375817,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 16748
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4632035819141633,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 16749
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2829188201744157,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 16750
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9499171321263263,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 16751
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.48464343108598906,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 16752
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.495418103509824,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 16753
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.050985210151307,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 16754
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.177760102671738,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 16755
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2320975998697473,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 16756
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3917101794234108,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 16757
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8205283834579704,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 16758
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8221597127593926,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 16759
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3641718540578728,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 16760
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3499233912877746,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 16761
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1683356193103599,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 16762
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.228910203892191,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 16763
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5026589563052031,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 16764
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8630955699237255,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 16765
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.815537486690373,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 16766
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3029660599674222,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 16767
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.478222998179091,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 16768
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5071011182141938,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 16769
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6897877429592102,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 16770
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.10083728238062549,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 16771
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.969415208676374,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 16772
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7350465260915506,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 16773
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.09137657511165012,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 16774
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7887245436488517,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 16775
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4541221531947541,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 16776
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7781725253388823,
+      "learning_rate": 2e-05,
+      "loss": 0.1189,
+      "step": 16777
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.138178631763054,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 16778
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6455197659293208,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 16779
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.195253956676286,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 16780
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6897601559753016,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 16781
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6904605065445215,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 16782
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6703766594261104,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 16783
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.32234218728319547,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 16784
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0410788202312835,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 16785
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.957039635267173,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 16786
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1834293666788773,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 16787
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.557728717827096,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 16788
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4276470681141027,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 16789
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12148488234754397,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 16790
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18430528332916862,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 16791
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3916721473327047,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 16792
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.964306527621134,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 16793
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3590889114476545,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 16794
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.670831557581508,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 16795
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6499248818860918,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 16796
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9038246190460367,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 16797
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.014231562224243,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 16798
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9660418095416161,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 16799
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2530823850163666,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 16800
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.474132334365267,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 16801
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.151901061127087,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 16802
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4219814721790363,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 16803
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.249743816826686,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 16804
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3189203986306963,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 16805
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.288669595661443,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 16806
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3855152334079965,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 16807
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8514376207530951,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 16808
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7094195343146583,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 16809
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1731985628601127,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 16810
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6959867762598897,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 16811
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.525437383534431,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 16812
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8620812788644163,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 16813
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.184148346107181,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 16814
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.343464134584835,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 16815
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6930095306822213,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 16816
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.04518423486226559,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 16817
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9489277410289478,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 16818
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4281508103825523,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 16819
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1388017666709285,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 16820
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4094553127370162,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 16821
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.33798014827489237,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 16822
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4068293755486976,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 16823
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7851939124343539,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 16824
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7821386028165844,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 16825
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1265079533338316,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 16826
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5675204683029131,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 16827
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9341979685180437,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 16828
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.082354837146072,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 16829
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8012861195485447,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 16830
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8602838045732413,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 16831
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.037292981463292,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 16832
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.085130170971352,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 16833
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3508565304784514,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 16834
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7467807981706261,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 16835
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3781379526096198,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 16836
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.616185048060204,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 16837
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.19571192906234233,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 16838
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07457073719109546,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 16839
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1879222814546804,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 16840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2083316945470461,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 16841
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.48359277074958124,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 16842
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2279874370542655,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 16843
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7956197238935966,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 16844
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5093634342358624,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 16845
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3025503152782605,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 16846
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.957160735224353,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 16847
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7918970566114896,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 16848
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.374696481323734,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 16849
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2529308768048963,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 16850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5068593248837234,
+      "learning_rate": 2e-05,
+      "loss": 0.1049,
+      "step": 16851
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1812028983640626,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 16852
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6636664658455778,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 16853
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4435954368750037,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 16854
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.386635533374575,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 16855
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4332306470063134,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 16856
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9710775532066097,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 16857
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8681691430268244,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 16858
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.033743243028175,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 16859
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5669078316060605,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 16860
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3194864260388792,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 16861
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9846181417848424,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 16862
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6270728462197128,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 16863
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7249602050613835,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 16864
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.22704987064302024,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 16865
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5872062425530392,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 16866
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6873642180478727,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 16867
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.282894325748119,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 16868
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.768838968109509,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 16869
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5833511557015534,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 16870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2985615668923103,
+      "learning_rate": 2e-05,
+      "loss": 0.111,
+      "step": 16871
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9266111300027453,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 16872
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4812400760019777,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 16873
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8690965276954048,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 16874
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.162552060252947,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 16875
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.828641874126868,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 16876
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.058065232201854076,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 16877
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8430486735324283,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 16878
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9504026714020892,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 16879
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3884698262080095,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 16880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1191556440214399,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 16881
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2490972119839039,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 16882
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4749866446073122,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 16883
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.094046079688442,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 16884
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0462849220819175,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 16885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.384877895052554,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 16886
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9983596631506289,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 16887
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.792163652810255,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 16888
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.906777938964412,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 16889
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4448287534128856,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 16890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7454597215779375,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 16891
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1649903049802333,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 16892
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5973585826144443,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 16893
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2653657153115945,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 16894
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4525338067537614,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 16895
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1608421789700842,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 16896
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.034081242732429,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 16897
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3994909832668534,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 16898
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.073735353771857,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 16899
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2891111401282225,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 16900
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1015167798596077,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 16901
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3033031040029122,
+      "learning_rate": 2e-05,
+      "loss": 0.0876,
+      "step": 16902
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0286102031218778,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 16903
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3447563483901626,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 16904
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4235811192194014,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 16905
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3347418030928764,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 16906
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6905138158992419,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 16907
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8985087159168701,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 16908
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.391974448054475,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 16909
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6029098581625196,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 16910
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.570311892684895,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 16911
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6866346895757607,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 16912
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.214203811044887,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 16913
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.48461139201928777,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 16914
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9272773339672201,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 16915
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7954146149137428,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 16916
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5579384648188764,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 16917
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5436827541292842,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 16918
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7874655446819927,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 16919
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7413491583851476,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 16920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.440739578986302,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 16921
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2346870897003241,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 16922
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0101180439538027,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 16923
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.231195919185268,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 16924
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7578400334215185,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 16925
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7440588398041913,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 16926
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7182966544072198,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 16927
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8349839351688835,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 16928
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6601223457252725,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 16929
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7490155593875543,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 16930
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8863631556360679,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 16931
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2044830465376513,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 16932
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1757283246368064,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 16933
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.675435777623752,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 16934
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.203356138543384,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 16935
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1752937801851167,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 16936
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.22720514608063636,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 16937
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3923438402705239,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 16938
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2658604672140755,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 16939
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7666427919585987,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 16940
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8588788420905347,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 16941
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.04874882935072102,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 16942
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.156559206883943,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 16943
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4064123916324682,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 16944
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3260128504560522,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 16945
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.499283724572727,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 16946
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.742511257315877,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 16947
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2667533805547153,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 16948
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3094512879110777,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 16949
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5639792700569637,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 16950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.946259941931965,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 16951
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3191691758418524,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 16952
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4883740909051452,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 16953
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.142957741502129,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 16954
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.703261343721829,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 16955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.295202183403075,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 16956
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7947937208343265,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 16957
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7694960842714578,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 16958
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6877413456121653,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 16959
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4519683993082362,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 16960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9249960873495295,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 16961
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0146754006996992,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 16962
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6447622540167193,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 16963
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2996390192518262,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 16964
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.207479597357642,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 16965
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4603733054493022,
+      "learning_rate": 2e-05,
+      "loss": 0.082,
+      "step": 16966
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.215731624447929,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 16967
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5308057105284716,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 16968
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.728366996280355,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 16969
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.298148103755597,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 16970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8283209299032603,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 16971
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9864308678565779,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 16972
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2788584183479352,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 16973
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2364598305020136,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 16974
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.507810788733353,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 16975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.23045599627363036,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 16976
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.20152194849814753,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 16977
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4469904876164326,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 16978
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6871290723483244,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 16979
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.188553407177159,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 16980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.08991767538642,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 16981
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.015179843686087,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 16982
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.24895690328328293,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 16983
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5286279935285231,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 16984
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2899484075290058,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 16985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5148006438977948,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 16986
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2237425052463973,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 16987
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6749191297547493,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 16988
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9539491498518579,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 16989
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9759593474293783,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 16990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5352651944611807,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 16991
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4411368596448182,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 16992
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2661007099506981,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 16993
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7914992749445203,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 16994
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30332255906097794,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 16995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1867636355060127,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 16996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.756771276849738,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 16997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.13631272028516,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 16998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1154501438571687,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 16999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0446611596112247,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 17000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.179882956871299,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 17001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9068952625239615,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 17002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.065412904098445,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 17003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7589798421130484,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 17004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.45964843112001846,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 17005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26782839967141514,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 17006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5197789743163463,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 17007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.153060433758141,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 17008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7574000574077415,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 17009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5519420209555095,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 17010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.150184614550097,
+      "learning_rate": 2e-05,
+      "loss": 0.0997,
+      "step": 17011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.110886862786733,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 17012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.804464681502229,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 17013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4957505187697012,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 17014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7264992471049136,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 17015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.153653399555318,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 17016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2933385602881255,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 17017
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.04174751426401648,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 17018
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3060884549211234,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 17019
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5975170097962481,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 17020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.194932029036677,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 17021
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0436177435495337,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 17022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.087606670365918,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 17023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9268064656967658,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 17024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.49499264868180726,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 17025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41937097297360926,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 17026
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.52119201699084,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 17027
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1888859340793418,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 17028
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0631488613924687,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 17029
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6038708682461524,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 17030
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8383389853924625,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 17031
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.05693478084131456,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 17032
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7128383669577005,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 17033
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.35754663343320964,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 17034
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3271306972146033,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 17035
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4500889547898166,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 17036
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5623606403227743,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 17037
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5252582425810254,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 17038
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2943943612028919,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 17039
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.20170686010102318,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 17040
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1861300289110028,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 17041
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.38199590663813854,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 17042
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0732239712114207,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 17043
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.009711766239581,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 17044
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3832745399773407,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 17045
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6382907114996894,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 17046
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18268803125030816,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 17047
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43831470110253684,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 17048
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0934621159567408,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 17049
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11450316910413226,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 17050
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.036565077017488,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 17051
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.530630002352047,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 17052
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0664415554787587,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 17053
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.115542482069626,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 17054
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1848789702951696,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 17055
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3661688361180737,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 17056
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7719228048452242,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 17057
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.977671339100192,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 17058
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8114605096249736,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 17059
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5614512088320267,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 17060
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0954605350776716,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 17061
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5049788972255924,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 17062
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.267515106085046,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 17063
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.775028365139091,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 17064
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0572860009481584,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 17065
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.240520748336057,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 17066
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8640572352878434,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 17067
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5334205044596073,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 17068
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5403819590766149,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 17069
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.473592304188514,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 17070
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.605166418735371,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 17071
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3068231294655287,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 17072
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1761151723581849,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 17073
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.904417711423693,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 17074
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3493861056529415,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 17075
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.34104829830502376,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 17076
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8121689708984158,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 17077
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.826225021906551,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 17078
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2419557655154265,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 17079
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6536718491637925,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 17080
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1742930578500077,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 17081
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4185366356554976,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 17082
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.464313528642399,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 17083
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5559256350186725,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 17084
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8568293581436728,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 17085
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4079207350164187,
+      "learning_rate": 2e-05,
+      "loss": 0.0909,
+      "step": 17086
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12495061152020615,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 17087
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5110252031437208,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 17088
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.21686661088422196,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 17089
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28957477489466965,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 17090
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.786745200143433,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 17091
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5712635523651354,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 17092
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1775888128459324,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 17093
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8554294300932472,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 17094
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6196350888768791,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 17095
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5623016280638606,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 17096
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6462595038227574,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 17097
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4804501878827955,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 17098
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43018418129590363,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 17099
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6354736135843971,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 17100
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.081148171439702,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 17101
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42216799416366596,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 17102
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5985710025010997,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 17103
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9271408270392337,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 17104
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0011177515845455,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 17105
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.514906123022724,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 17106
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1475431894962893,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 17107
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0024412873310329,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 17108
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.491821952650397,
+      "learning_rate": 2e-05,
+      "loss": 0.0937,
+      "step": 17109
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8070335262595145,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 17110
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.216393221648804,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 17111
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8200344151607093,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 17112
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.727266062381577,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 17113
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.960499383990019,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 17114
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6800238840263004,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 17115
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7995166003547762,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 17116
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.527143467509749,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 17117
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5585678201272444,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 17118
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7316771930409504,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 17119
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.386719351754931,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 17120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3414180089625374,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 17121
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.036544355704104,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 17122
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5874703051575736,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 17123
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.24672580103608,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 17124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.376587168443282,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 17125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5828654109410083,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 17126
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5570851915263513,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 17127
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.19103639251836257,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 17128
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2198239509790505,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 17129
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1635876909725476,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 17130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4253801173468124,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 17131
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9793691143972443,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 17132
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6265091871924784,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 17133
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1151498739159955,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 17134
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6339594267076618,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 17135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1620876791074806,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 17136
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1565332323569788,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 17137
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4785504386849625,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 17138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9155578377479732,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 17139
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.422749712093668,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 17140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6138326510776799,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 17141
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6140289365414321,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 17142
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8612727890515328,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 17143
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4859218367945422,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 17144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7114907093044678,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 17145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6679635744842443,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 17146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2157294672228933,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 17147
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7331305572091009,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 17148
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07434748164056074,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 17149
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7476747149795212,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 17150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4573670682926023,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 17151
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4950311537593655,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 17152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5144600279047314,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 17153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2041008694809947,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 17154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2216915919808682,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 17155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.002086905607226,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 17156
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8135497783123555,
+      "learning_rate": 2e-05,
+      "loss": 0.112,
+      "step": 17157
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0374235837439567,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 17158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2288241582711026,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 17159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5393241026414191,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 17160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9334286340242002,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 17161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8000642124870215,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 17162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3622882363478008,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 17163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5901658390337414,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 17164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9806725339631887,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 17165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1005449632437396,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 17166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3281138245300927,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 17167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2368166727208694,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 17168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1669385860211112,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 17169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7950172460832684,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 17170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1474978696682838,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 17171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0218499953450875,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 17172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3656058181358295,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 17173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4186248522442533,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 17174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9425815194895234,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 17175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.184796116024969,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 17176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6982353785039965,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 17177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5994598728690297,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 17178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.778639905555718,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 17179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3990247122250057,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 17180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8562698080630784,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 17181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8921772283248504,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 17182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6331550202382488,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 17183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.144461208512967,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 17184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.764963954633118,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 17185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7659282592682874,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 17186
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1264603028977063,
+      "learning_rate": 2e-05,
+      "loss": 0.106,
+      "step": 17187
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5974342546381166,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 17188
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.22567124838511932,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 17189
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5019212016467,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 17190
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07491538656637813,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 17191
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5021497586209558,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 17192
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1903322976434654,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 17193
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4406845810019467,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 17194
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.175864747121501,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 17195
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3152735333551555,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 17196
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1600456553752251,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 17197
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8705972873305464,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 17198
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.507109926866847,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 17199
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6859233941645941,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 17200
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.000530936438198,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 17201
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3984415491141406,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 17202
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.45335932953653696,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 17203
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.31749313784490185,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 17204
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8248702814250758,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 17205
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9100904973063687,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 17206
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4749594569695736,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 17207
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.689171210786152,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 17208
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7608103641083672,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 17209
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1858743688653468,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 17210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.350845994900196,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 17211
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7242473510230238,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 17212
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1325342525670585,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 17213
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3478478336987438,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 17214
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.296779189414598,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 17215
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7025999761080763,
+      "learning_rate": 2e-05,
+      "loss": 0.0946,
+      "step": 17216
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33919597143359664,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 17217
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0588770170236073,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 17218
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5407087941624218,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 17219
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4675888178915335,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 17220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5551957876215114,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 17221
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3428332728145669,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 17222
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4437257872033262,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 17223
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0589065458887075,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 17224
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3510383515720485,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 17225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8507153908536033,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 17226
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0927036004956814,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 17227
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3253850269393737,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 17228
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7561219626891964,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 17229
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.967731709740306,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 17230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2395055594727653,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 17231
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5819806879699845,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 17232
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9270132052582833,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 17233
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0021003481133532,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 17234
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0747621029059,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 17235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1574857364546111,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 17236
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.715083576833891,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 17237
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37300285359229407,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 17238
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5495100570119379,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 17239
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.115657577376522,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 17240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5831832647313266,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 17241
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.681591143971143,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 17242
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5591219434877175,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 17243
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.38227771851389575,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 17244
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.6525625781003805,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 17245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6521992932000473,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 17246
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6794547696009487,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 17247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7702087048693342,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 17248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1703449687859235,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 17249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4771498203452735,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 17250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2365728681436712,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 17251
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.045986942326536,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 17252
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.177239410495339,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 17253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.903228052916434,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 17254
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8138340611501786,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 17255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.113150605684181,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 17256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4604913524391464,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 17257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.21701850388176264,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 17258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0352504471439592,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 17259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.283715132896338,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 17260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9370858231224906,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 17261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9437825772050129,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 17262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.819481000783777,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 17263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0306137896796177,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 17264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.02390278729919,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 17265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6194314363945119,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 17266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6591301567401132,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 17267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.46638173433470564,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 17268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.678264509520404,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 17269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1729892193347915,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 17270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6719722541259154,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 17271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1939828495793905,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 17272
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5182506195524381,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 17273
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4503087461475254,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 17274
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16781562656636934,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 17275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6124830944297314,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 17276
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1362724345628936,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 17277
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2623595036806663,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 17278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2123750582945825,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 17279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8715987174523184,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 17280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14053834788572142,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 17281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3723453296302908,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 17282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0228941278824553,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 17283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.155250981312838,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 17284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4036568445324017,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 17285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6428652772050485,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 17286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6620013618700131,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 17287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42746142383440755,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 17288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6694597034663108,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 17289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4110180474887699,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 17290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5587291419628477,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 17291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5613740672594508,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 17292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0762872928267988,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 17293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5010877935652058,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 17294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2025587557675939,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 17295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.028129138456102,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 17296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2514691302538274,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 17297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0045501685730933,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 17298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7334289658106745,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 17299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4144162741861885,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 17300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8345645397204702,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 17301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1727887469518143,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 17302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2916104171645189,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 17303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4797960239377634,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 17304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0833090388190418,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 17305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9640251424466675,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 17306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.785768759897017,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 17307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6011664744967122,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 17308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.52042039290383,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 17309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.559819727563358,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 17310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5208711575895149,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 17311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1964154171079016,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 17312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.31335023610368046,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 17313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5904949082572188,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 17314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.307538251136163,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 17315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.449895390341817,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 17316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8157348061963656,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 17317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.524072717147344,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 17318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2422139729717414,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 17319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.662874168322101,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 17320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1026859371453361,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 17321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6184112392129626,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 17322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1822572722732636,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 17323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8481502587669125,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 17324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7265437954371591,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 17325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5653453010007636,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 17326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.45722761581801463,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 17327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3140127456081723,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 17328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.883462475951391,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 17329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0901508576451507,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 17330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9246347466594969,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 17331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7670261596099044,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 17332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.933255191023993,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 17333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06779517242995499,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 17334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8564087045429413,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 17335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4564748729577353,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 17336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5271364440345107,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 17337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9956809137661042,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 17338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.18390588402205282,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 17339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.111622427804095,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 17340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.470907301564213,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 17341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0633283639710003,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 17342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6574929090207233,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 17343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.057135153697830655,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 17344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.782248242857131,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 17345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16976185092689589,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 17346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7267066584684887,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 17347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1396054976479668,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 17348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3413822694033486,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 17349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8035595679270893,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 17350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5967055199076212,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 17351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11487201518009206,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 17352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3025721576830125,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 17353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9393865359493301,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 17354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8980718780926309,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 17355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6988334020134986,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 17356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.145282746015685,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 17357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.186100968048581,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 17358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7832003091585525,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 17359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8753641019009097,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 17360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06957219397063892,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 17361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3482725916419887,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 17362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.45046762104402377,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 17363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.793462792297088,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 17364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.810490632173745,
+      "learning_rate": 2e-05,
+      "loss": 0.0991,
+      "step": 17365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.192640771854116,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 17366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.410142335201559,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 17367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6159440572233432,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 17368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8364663634779775,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 17369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07031196355615064,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 17370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6192024272314202,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 17371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6460868382682223,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 17372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.056098785981234724,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 17373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2911651109264477,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 17374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3301722594668965,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 17375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0408372426907455,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 17376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8272406336397359,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 17377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1360834712671968,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 17378
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.1817608120718575,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 17379
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7278729467086553,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 17380
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.04409633620500823,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 17381
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8347562483250646,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 17382
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.584047837695957,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 17383
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.780139229807651,
+      "learning_rate": 2e-05,
+      "loss": 0.0913,
+      "step": 17384
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9058965454108121,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 17385
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37432253758170797,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 17386
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9875051293058816,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 17387
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1192166785027755,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 17388
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.824326044896311,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 17389
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9683991689072595,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 17390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2001359643998506,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 17391
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7163113025627628,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 17392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9337829497278967,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 17393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7110702965830926,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 17394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.698750942724346,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 17395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.053007097374993,
+      "learning_rate": 2e-05,
+      "loss": 0.1228,
+      "step": 17396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7937020192391577,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 17397
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9748998908465147,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 17398
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9719397726166468,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 17399
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9649870916600206,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 17400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5613267690019443,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 17401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1932562800701105,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 17402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4404293564618265,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 17403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2264534506948439,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 17404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15390785284224334,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 17405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6428190434799254,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 17406
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41418224978281754,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 17407
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2635030346128135,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 17408
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.921354517975883,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 17409
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1078278942679574,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 17410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.35256030344896,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 17411
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1547872345242536,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 17412
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.23266575944372045,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 17413
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3791899564837593,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 17414
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7184783173291149,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 17415
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42504423556402104,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 17416
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13969792642873766,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 17417
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.173847217372809,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 17418
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3699737713834947,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 17419
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3731978049964373,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 17420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5094041378472056,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 17421
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6183519846124542,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 17422
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7530227781703283,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 17423
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5162399203860866,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 17424
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8041049366978179,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 17425
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9562531714293304,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 17426
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3487589743658766,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 17427
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7422814560697475,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 17428
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2576411823626499,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 17429
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.348902417607817,
+      "learning_rate": 2e-05,
+      "loss": 0.0707,
+      "step": 17430
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.59997710867242,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 17431
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.782815623324744,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 17432
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0271498488810504,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 17433
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.310106046985216,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 17434
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3807640738951292,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 17435
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9868297862392,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 17436
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.108474957349724,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 17437
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0964604716588462,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 17438
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3224929771885878,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 17439
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7786020373141547,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 17440
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2748713501284155,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 17441
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.590058796491503,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 17442
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9847754127037374,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 17443
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9787392673454391,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 17444
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4142946792107183,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 17445
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8025918727348245,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 17446
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0248747327700656,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 17447
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2172722783270122,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 17448
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1229214293727747,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 17449
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4450480696614549,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 17450
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.612230577219785,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 17451
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.071354157270556,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 17452
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.992901032233628,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 17453
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1461839964668346,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 17454
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.20263571681498815,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 17455
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3039109813063385,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 17456
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5016622869418534,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 17457
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5659348993015197,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 17458
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6057747590566136,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 17459
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11198233075070418,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 17460
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1214916092169128,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 17461
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8149178918808997,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 17462
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.997933457829934,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 17463
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.463682717723316,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 17464
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3960341361307211,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 17465
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6543971079756807,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 17466
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7613049287819615,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 17467
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7635821111569965,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 17468
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5567757082424962,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 17469
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5618902451310852,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 17470
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.889485387619469,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 17471
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9199896768451372,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 17472
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0003791924612444,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 17473
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.237972744028208,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 17474
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8812223818960065,
+      "learning_rate": 2e-05,
+      "loss": 0.083,
+      "step": 17475
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.337164413107577,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 17476
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.917362093883967,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 17477
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2851161855462214,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 17478
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8501693055290114,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 17479
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9565639990091028,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 17480
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.123760175756848,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 17481
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9828255887516976,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 17482
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7851868428964863,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 17483
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6141099625558117,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 17484
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.339240937896196,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 17485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0587223020078906,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 17486
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.625259204964247,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 17487
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4349506826931282,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 17488
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0329156197966676,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 17489
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.012099617808345,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 17490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.292669273458546,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 17491
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4908478000565146,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 17492
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.626503835420868,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 17493
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6319206759117497,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 17494
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3636189193767942,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 17495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7591636646209999,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 17496
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8004678383126136,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 17497
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07447576567858732,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 17498
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.363240238654935,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 17499
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.220794702686566,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 17500
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9624828313574625,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 17501
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.759517931233474,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 17502
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1494996930353087,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 17503
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6213684674067141,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 17504
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.243920791835187,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 17505
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7092429352397132,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 17506
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5288980536903435,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 17507
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4804254002425092,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 17508
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0120248506122387,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 17509
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6508750838208073,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 17510
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2766456679633316,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 17511
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34245220031766604,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 17512
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6507688551661754,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 17513
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5501627036807106,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 17514
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.49902499025107006,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 17515
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8440230951923342,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 17516
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0719624631024727,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 17517
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2442751856960426,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 17518
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4422065939673009,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 17519
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8722081061540738,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 17520
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.05027465376063,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 17521
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.695847256451067,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 17522
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8110008714791165,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 17523
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.47879603986253666,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 17524
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9053936269793494,
+      "learning_rate": 2e-05,
+      "loss": 0.1024,
+      "step": 17525
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4153991573218668,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 17526
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9136134291633923,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 17527
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.069549995933644,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 17528
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8729258203747281,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 17529
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0475324587470314,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 17530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3115713626745065,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 17531
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1852548308643519,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 17532
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7627329798737145,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 17533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.556576614714957,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 17534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7218200381935163,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 17535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.38331804175906004,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 17536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.20899444627113392,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 17537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.569729206234693,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 17538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.08799384293497,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 17539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6512995374513633,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 17540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4691775756056656,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 17541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.236520761243846,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 17542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08998768099610198,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 17543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9695495314091986,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 17544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6738761110772491,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 17545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5147858930156188,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 17546
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7482907940198942,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 17547
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5801746918884525,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 17548
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5284726954127374,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 17549
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0136555282656532,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 17550
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34850037353911556,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 17551
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.637333894207134,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 17552
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4485544767109313,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 17553
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.10198967760724109,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 17554
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.09549352016266521,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 17555
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9128479999454926,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 17556
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5703656158795551,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 17557
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7763787301198655,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 17558
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.285047300410731,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 17559
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41585442728743094,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 17560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7887075022207508,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 17561
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5623176323730674,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 17562
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9997251702625913,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 17563
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3863350350773915,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 17564
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8590926909708445,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 17565
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7621333632220687,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 17566
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8044424179584626,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 17567
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5130083758770945,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 17568
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2775654731590111,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 17569
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1634409224229056,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 17570
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8332241264947817,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 17571
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2582103029109715,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 17572
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0542875570939952,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 17573
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4438533531285564,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 17574
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8842965202421826,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 17575
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.340852363353387,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 17576
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6991615271995103,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 17577
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.025594261475279,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 17578
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5691600765117573,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 17579
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25153569084566707,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 17580
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07054987548285553,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 17581
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.578146934093187,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 17582
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3180248749896082,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 17583
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4325728737804644,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 17584
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4446306656844405,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 17585
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.673563950834006,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 17586
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2338872018140625,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 17587
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.059178827627636,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 17588
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0011504826993265,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 17589
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.640078939555621,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 17590
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.727231599404104,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 17591
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8664054988765362,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 17592
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1203027307542537,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 17593
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2859747231195886,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 17594
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.24597220219634716,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 17595
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.089638475344175,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 17596
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.553996047958626,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 17597
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.587164395432524,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 17598
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5972965647032467,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 17599
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.995865353856091,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 17600
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8851697390438431,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 17601
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4959462543818696,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 17602
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4658746513195897,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 17603
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8830010879419176,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 17604
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8803844990399946,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 17605
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.481280336747057,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 17606
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.18010916163164734,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 17607
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0498369930530773,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 17608
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.998455126672314,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 17609
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0699121627968506,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 17610
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.2137308286526665,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 17611
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4960472956285327,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 17612
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6170820556574166,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 17613
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1442479570113764,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 17614
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.135922287284224,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 17615
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.714676469238873,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 17616
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2882895186771015,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 17617
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4909571756248203,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 17618
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5054645342290236,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 17619
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7880682771918585,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 17620
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6776114140640258,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 17621
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0995018264456962,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 17622
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7705651710347003,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 17623
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0794272367188176,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 17624
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8900937862907234,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 17625
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.567914035029695,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 17626
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.444940040852787,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 17627
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.36957082959856535,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 17628
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9638857620096861,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 17629
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2881881532531363,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 17630
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.23107258847464152,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 17631
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.484212605477415,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 17632
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.722471211739213,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 17633
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6634537792860815,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 17634
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3544450418527213,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 17635
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6051702291466081,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 17636
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3313972366344704,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 17637
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5380152513519212,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 17638
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8859477787666198,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 17639
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.779699889182623,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 17640
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8288315691729401,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 17641
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08009278058642329,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 17642
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7274261661987087,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 17643
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5834119408867471,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 17644
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2183771296629604,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 17645
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16252328701238786,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 17646
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6657646678889644,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 17647
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4424988550236,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 17648
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.18679845218160138,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 17649
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5563269936999005,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 17650
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.40172631377166107,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 17651
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.402868991901646,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 17652
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.399525150371553,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 17653
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5292357237002615,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 17654
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0272816686402242,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 17655
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.165054747003,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 17656
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7422405023220934,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 17657
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11624403490251792,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 17658
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.931821057903055,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 17659
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8410688115022235,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 17660
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4176107327707237,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 17661
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12285100014926308,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 17662
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.46896773685476284,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 17663
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7379323952576572,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 17664
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3482087702715033,
+      "learning_rate": 2e-05,
+      "loss": 0.0931,
+      "step": 17665
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3949445693440858,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 17666
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8800565524786506,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 17667
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3890182985840285,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 17668
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.132967765103119,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 17669
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6555785649804542,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 17670
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3259855653572656,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 17671
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1983342092631981,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 17672
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6644270401218256,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 17673
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9989528895137172,
+      "learning_rate": 2e-05,
+      "loss": 0.1437,
+      "step": 17674
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3992012968186106,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 17675
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.043839744907254,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 17676
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29007864531067884,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 17677
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4614124036202787,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 17678
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16760790122415425,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 17679
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8323481108830786,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 17680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7625230113450072,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 17681
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3501040983478347,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 17682
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5772370659282444,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 17683
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6695260972455691,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 17684
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3120487890766876,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 17685
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2994334186454384,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 17686
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9608375993608017,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 17687
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.230685729581313,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 17688
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7358267212769031,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 17689
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3260978487054134,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 17690
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2845349632855525,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 17691
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8137122427623301,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 17692
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9130604133506389,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 17693
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.518385137062614,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 17694
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.34858506674490963,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 17695
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1518115867674465,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 17696
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4680142668685028,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 17697
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7782637935779931,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 17698
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9736662986146971,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 17699
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.791212098172533,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 17700
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1111448971591915,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 17701
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.375973703504615,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 17702
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5195814409606457,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 17703
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9681115661634715,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 17704
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1291308328120402,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 17705
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5103007463922561,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 17706
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9099153394061033,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 17707
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.806992719529978,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 17708
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3086867114397684,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 17709
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1093302362665317,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 17710
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5113778925214851,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 17711
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2817431640531995,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 17712
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9573495452341853,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 17713
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2994146514401903,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 17714
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8291057965100284,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 17715
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5524042894282026,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 17716
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.20431517415121445,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 17717
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5699060777646283,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 17718
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6947635455257516,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 17719
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3219933729626043,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 17720
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5981673006667487,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 17721
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5607510769630053,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 17722
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6458889030378656,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 17723
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.14329512479031675,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 17724
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.458696586366157,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 17725
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.140832709934396,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 17726
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6380547144488418,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 17727
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2475345138943226,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 17728
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.740874064790572,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 17729
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6726799265267323,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 17730
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0535428299896865,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 17731
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1233564663287006,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 17732
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4840812794052842,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 17733
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.439537014839457,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 17734
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3303875009384984,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 17735
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2014827667092611,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 17736
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.589016392519193,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 17737
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2653292557654354,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 17738
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.996543633087124,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 17739
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.184365779218736,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 17740
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8452024321192977,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 17741
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3451185186475935,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 17742
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5309497984900466,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 17743
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8799184438453451,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 17744
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9910103106114392,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 17745
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2352657243059408,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 17746
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6685937556342338,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 17747
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18155671304156976,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 17748
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4038406501095355,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 17749
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2552969388982613,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 17750
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.363293754279823,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 17751
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.604451783637158,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 17752
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11701297877764608,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 17753
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6767015753209382,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 17754
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08306958480608792,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 17755
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.462403811038496,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 17756
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2116632563207628,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 17757
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1642070910881541,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 17758
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0923445254456665,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 17759
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7442727760056452,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 17760
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.920403265922517,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 17761
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8262655468748312,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 17762
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3984425511815908,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 17763
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.545017903935037,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 17764
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5447433069427614,
+      "learning_rate": 2e-05,
+      "loss": 0.0969,
+      "step": 17765
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5540710569905367,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 17766
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5964240572038038,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 17767
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9712494630947126,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 17768
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5070033298751808,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 17769
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3171064738809553,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 17770
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.069211128650324,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 17771
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1978406268578128,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 17772
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2946535410698585,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 17773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1324789683726475,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 17774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5567332854028346,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 17775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8170747638134404,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 17776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.713527095775241,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 17777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.954901495863074,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 17778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0108011171321523,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 17779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.696389422678351,
+      "learning_rate": 2e-05,
+      "loss": 0.1131,
+      "step": 17780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5390862169438749,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 17781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0552351483436597,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 17782
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.435800791507927,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 17783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5518381516023413,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 17784
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45906171755792197,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 17785
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3005856203627926,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 17786
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7802953609590201,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 17787
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9220325206081348,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 17788
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24709263425570055,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 17789
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.186743495298937,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 17790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.47855666626852833,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 17791
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1204981520918387,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 17792
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5036169300905088,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 17793
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6346257216994379,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 17794
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5622274542573655,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 17795
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6715395582004353,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 17796
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5928864473463882,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 17797
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4913122933327512,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 17798
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18183471195779038,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 17799
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.203146427131507,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 17800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3653459626137738,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 17801
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18402807117734668,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 17802
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9628588014366803,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 17803
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12996385670242078,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 17804
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.32113284223431326,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 17805
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8317278710443838,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 17806
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0535547883680496,
+      "learning_rate": 2e-05,
+      "loss": 0.0751,
+      "step": 17807
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.229987899336892,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 17808
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0715138730652085,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 17809
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2739475843465278,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 17810
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6629432173466605,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 17811
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6996159624037739,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 17812
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18888380101853244,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 17813
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4895736527669498,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 17814
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4887720134910842,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 17815
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.48849171489534277,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 17816
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49978145651215855,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 17817
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.14294372349973547,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 17818
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0078177247925697,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 17819
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3410983400643803,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 17820
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1975259196650498,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 17821
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1693788519611865,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 17822
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.429031888799523,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 17823
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5271703909518013,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 17824
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.521909251427975,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 17825
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0318286951726763,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 17826
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6502922831467421,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 17827
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5198581690460586,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 17828
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.128469439849224,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 17829
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9391844240273913,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 17830
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.023244324954505388,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 17831
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0042555383211882,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 17832
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6858371435278189,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 17833
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5076984277394692,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 17834
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2044985710976679,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 17835
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5575101251922919,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 17836
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7588344739678112,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 17837
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0607356962434875,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 17838
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6233993776755214,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 17839
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.98203590281252,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 17840
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9798556550710944,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 17841
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5014106951618085,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 17842
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7578954927313917,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 17843
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8645156734296726,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 17844
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6976143742175723,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 17845
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0189845748999873,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 17846
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0533678629402423,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 17847
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1724873786523504,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 17848
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.802445201968805,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 17849
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9414674731089004,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 17850
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7595940949802473,
+      "learning_rate": 2e-05,
+      "loss": 0.0811,
+      "step": 17851
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9600271110216667,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 17852
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9144989992911077,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 17853
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0555743974264424,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 17854
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.686811801310256,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 17855
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3763310752083924,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 17856
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36194327288532596,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 17857
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.586404810052205,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 17858
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8981773995103701,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 17859
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7261024577351521,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 17860
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.705861817462491,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 17861
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0084434722354725,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 17862
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6504219729363823,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 17863
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1572171367792556,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 17864
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.770713048250549,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 17865
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.673498053348794,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 17866
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2784070063974765,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 17867
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5363447330526723,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 17868
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5302046496834074,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 17869
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4241843594262977,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 17870
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2412815528802037,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 17871
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2742886867010843,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 17872
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3295179048370977,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 17873
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5498545270049557,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 17874
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.504278146438411,
+      "learning_rate": 2e-05,
+      "loss": 0.083,
+      "step": 17875
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.792596250580826,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 17876
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.23498147540177666,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 17877
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.786363389705453,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 17878
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30034476973182206,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 17879
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.449077342247158,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 17880
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7838299685235721,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 17881
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7217948522290639,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 17882
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6942600686301664,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 17883
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8379107969760098,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 17884
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25732335012676566,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 17885
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7350430276026143,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 17886
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2257035514428747,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 17887
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9463037010825248,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 17888
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3367475318151074,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 17889
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.107040696057513,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 17890
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8340257146399146,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 17891
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9659798608955216,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 17892
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.378878239247191,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 17893
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.23734542905101833,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 17894
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.532389547652223,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 17895
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7964250747637966,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 17896
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.359964562327815,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 17897
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24381949094417396,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 17898
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3030654450539177,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 17899
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8191970878686805,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 17900
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3452973439141551,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 17901
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5477663034730489,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 17902
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5106062990964033,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 17903
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.715859149456422,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 17904
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3969857549603262,
+      "learning_rate": 2e-05,
+      "loss": 0.1247,
+      "step": 17905
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.922481038132751,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 17906
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1822522810818983,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 17907
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5784482310086454,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 17908
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.125473902054976,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 17909
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.628967430374589,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 17910
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2511076084103543,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 17911
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4537622018631657,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 17912
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2809273272827097,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 17913
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5964976530964192,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 17914
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1795932151050836,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 17915
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2390029200195436,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 17916
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7140555355880887,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 17917
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7321949222026483,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 17918
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.20940563490207578,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 17919
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5103573825109535,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 17920
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.40966884171912354,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 17921
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7089900497335525,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 17922
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9553610727766668,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 17923
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18783098055321654,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 17924
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1131064859214987,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 17925
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1824736378092195,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 17926
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.365955525694323,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 17927
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0463872883733283,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 17928
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.507723000197552,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 17929
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3955369349931064,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 17930
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.548842026069504,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 17931
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1927916929385554,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 17932
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.33636462980464615,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 17933
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.21711038139689282,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 17934
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7147390987316115,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 17935
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0895704247978546,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 17936
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1083114253773387,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 17937
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.952469303157101,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 17938
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8080404849013216,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 17939
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.411858796275363,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 17940
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8632036407245439,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 17941
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9208379505151035,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 17942
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4957131481855983,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 17943
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.570184351308701,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 17944
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7662463170238354,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 17945
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9373544897960766,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 17946
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.848194132735144,
+      "learning_rate": 2e-05,
+      "loss": 0.0868,
+      "step": 17947
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4413504876444367,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 17948
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6252913580635908,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 17949
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08701839239470585,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 17950
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0282385232972657,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 17951
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5829104241542932,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 17952
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3091143912057941,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 17953
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4741945709405773,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 17954
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.50841396417332,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 17955
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29405867966330007,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 17956
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9130313509153374,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 17957
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8704221385883031,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 17958
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7810647363343007,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 17959
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3617502266883836,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 17960
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4578179493237011,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 17961
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1087745330917593,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 17962
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7067826397742958,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 17963
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.674818047089004,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 17964
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3070507004790373,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 17965
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6722249960272898,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 17966
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6969020333627591,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 17967
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08826533022112319,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 17968
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7047099866354236,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 17969
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8533211493056916,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 17970
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.080928205109263,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 17971
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4038057670134965,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 17972
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3916185142291732,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 17973
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0630917658595638,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 17974
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7402923883336651,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 17975
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0015335245449117,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 17976
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6756654982025866,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 17977
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.150815253911043,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 17978
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.923238102125309,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 17979
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.2282675750230885,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 17980
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2562185779938049,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 17981
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5859931156466873,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 17982
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6125354727379632,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 17983
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.051596450736565914,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 17984
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2804958776674269,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 17985
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.040142956941871,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 17986
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1122339160084236,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 17987
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3959103985289065,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 17988
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.206590047154793,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 17989
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5712817033651083,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 17990
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2173124785403937,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 17991
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4467215581367349,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 17992
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7293638144504092,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 17993
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.784882747189554,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 17994
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9508411372937416,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 17995
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3445972167106985,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 17996
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6609444060990636,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 17997
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2013704286062845,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 17998
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7568640969072795,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 17999
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.619751965525842,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 18000
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0952511351453919,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 18001
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28012677917953815,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 18002
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2630472185347505,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 18003
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.836290106494437,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 18004
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.22205810345560648,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 18005
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11932467487088307,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 18006
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0735620329247852,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 18007
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31912541814314227,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 18008
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.204703952404738,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 18009
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3991993744509847,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 18010
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5536913994465468,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 18011
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0129145327040427,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 18012
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.824665292812311,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 18013
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.868471714043776,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 18014
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8516492886894775,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 18015
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.647209670562004,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 18016
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.44498327700663876,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 18017
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.033337151674948,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 18018
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7771715519783108,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 18019
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3134865447264329,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 18020
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7420959416181725,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 18021
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3367818651398349,
+      "learning_rate": 2e-05,
+      "loss": 0.1072,
+      "step": 18022
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5445869537715724,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 18023
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4902001373237708,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 18024
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.047275409702577,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 18025
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7526738663619915,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 18026
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2002056919080533,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 18027
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1479077317478021,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 18028
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7062878419361678,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 18029
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.91914758272176,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 18030
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8158761512267951,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 18031
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6979827185352373,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 18032
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4404545979043064,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 18033
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7642777966199767,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 18034
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9416100176867066,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 18035
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6465814062408483,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 18036
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1038166547027168,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 18037
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2344020405958298,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 18038
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6689954975544021,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 18039
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6559997945407612,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 18040
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.039926155659845,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 18041
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9805803881308368,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 18042
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1138602374116033,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 18043
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7995857026895004,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 18044
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8757115603350274,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 18045
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8665330496514866,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 18046
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8244684211832382,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 18047
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15474520507657474,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 18048
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5352082888508035,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 18049
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.213355670913722,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 18050
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2840583046071337,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 18051
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0340229092094717,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 18052
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7707967276945668,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 18053
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9196788366247393,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 18054
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18881066511096628,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 18055
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7489546651008432,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 18056
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1089494063927576,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 18057
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.35129985535266023,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 18058
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0940699518075807,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 18059
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7622606746117566,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 18060
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.376145492591014,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 18061
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.495211109323393,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 18062
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4923148375498174,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 18063
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7692553805118767,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 18064
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7230474289655495,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 18065
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8890130362804616,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 18066
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0935450770641735,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 18067
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.05586801567138613,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 18068
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6969309873108573,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 18069
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6358978576030885,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 18070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.037639130604927,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 18071
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08793549187193363,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 18072
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7315585663617747,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 18073
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0667618378759498,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 18074
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7536029510950053,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 18075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2268788603976006,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 18076
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1108304000285387,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 18077
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5906700510675432,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 18078
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4293043946982942,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 18079
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6119443815910172,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 18080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2459404351798091,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 18081
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.713597776968767,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 18082
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8025839735342344,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 18083
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2106855068732918,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 18084
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9901591066621016,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 18085
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3831579739630198,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 18086
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9968148407676308,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 18087
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8089571150198215,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 18088
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5902350435183745,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 18089
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9948756612784622,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 18090
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6254827809262227,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 18091
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.34643875633542,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 18092
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9737109093784324,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 18093
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8025371374170754,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 18094
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5806959757287093,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 18095
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0149750610632673,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 18096
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.257914634669513,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 18097
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4953466933087596,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 18098
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12646417359870477,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 18099
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.19439541083985537,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 18100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9551919604838544,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 18101
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.70320619162119,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 18102
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.001512278043052,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 18103
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2986041332816984,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 18104
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2078902848307949,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 18105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0316770998520175,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 18106
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.094166758019898,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 18107
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7732073508882669,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 18108
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7261578243018734,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 18109
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.941869805949473,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 18110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.234559574481376,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 18111
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5386089683797437,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 18112
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6174275860817348,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 18113
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.51643443857184,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 18114
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8117446732869555,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 18115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9114533051851637,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 18116
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27683112627824635,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 18117
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8817164606354269,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 18118
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24770289321799055,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 18119
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.33717059997263776,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 18120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6993562573809791,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 18121
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8138073777128139,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 18122
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.092903518093498,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 18123
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.739921224827885,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 18124
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1378243495845286,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 18125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6487752884877138,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 18126
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.612174112029256,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 18127
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8737971992485307,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 18128
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5389343606989473,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 18129
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1294615016926637,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 18130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2161266752881008,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 18131
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11150155863292616,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 18132
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.41676124055757424,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 18133
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.689238316354529,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 18134
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9137145138071584,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 18135
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.334473626480268,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 18136
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.002641483218811,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 18137
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.055874411004569,
+      "learning_rate": 2e-05,
+      "loss": 0.107,
+      "step": 18138
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06780598282260772,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 18139
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.046211601837942,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 18140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8078168679384894,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 18141
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9893203519764662,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 18142
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0679345635736186,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 18143
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.133812415040999,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 18144
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8066476852691435,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 18145
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8148108910859149,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 18146
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9373433618024889,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 18147
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4904365852325316,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 18148
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.718874259827617,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 18149
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9245512634246431,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 18150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6960848044560254,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 18151
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.04080862063296,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 18152
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3388980831307171,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 18153
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4483381613829063,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 18154
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.121487350897264,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 18155
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3122089305904847,
+      "learning_rate": 2e-05,
+      "loss": 0.0775,
+      "step": 18156
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8784752449374604,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 18157
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7181287236527996,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 18158
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.41601970831787866,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 18159
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9604265983673195,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 18160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17286708999115907,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 18161
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.374579191838586,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 18162
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3198825462676531,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 18163
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4150154677185958,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 18164
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7544416196873549,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 18165
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4078789918631769,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 18166
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3539897314045715,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 18167
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6664325606059248,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 18168
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3912556135670685,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 18169
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6333627780753712,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 18170
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.925148502068788,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 18171
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.104131984303593,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 18172
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.266368976964645,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 18173
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4879278128762048,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 18174
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.717406094550031,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 18175
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.04527258800601573,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 18176
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.202832249880154,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 18177
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4285996263990857,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 18178
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.332201858889656,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 18179
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9309717492304992,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 18180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42223217643783983,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 18181
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36815973173068295,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 18182
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.154894859068,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 18183
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.30674153993874037,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 18184
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4973237841416753,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 18185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8848254791965512,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 18186
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.559336587860025,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 18187
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7787663729377219,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 18188
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5456029111870306,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 18189
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14360386142475062,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 18190
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6083012927339969,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 18191
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9510838944128235,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 18192
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.550068884069812,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 18193
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2728142646685532,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 18194
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9130241169120613,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 18195
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5001292955594985,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 18196
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8422346282398674,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 18197
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9989240624308184,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 18198
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6852065978043234,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 18199
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.954558372958923,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 18200
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8974426400495509,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 18201
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3080530741738033,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 18202
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.337315300639868,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 18203
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1559735560511013,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 18204
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.082878584503769,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 18205
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.40654796896907947,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 18206
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1790917009892554,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 18207
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4023240834586896,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 18208
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.96714546611757,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 18209
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5884490038638144,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 18210
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2513483881115044,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 18211
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8232323407255124,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 18212
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5842792952803177,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 18213
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.532877232606953,
+      "learning_rate": 2e-05,
+      "loss": 0.0902,
+      "step": 18214
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4514366103524672,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 18215
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5680922569362516,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 18216
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.721590365442098,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 18217
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0521296248622682,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 18218
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9520742221176504,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 18219
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7028007548643505,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 18220
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24644379470655295,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 18221
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7190275900140523,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 18222
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4273362968095722,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 18223
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3908770060636284,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 18224
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9001923915430063,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 18225
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.326807737653594,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 18226
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1228580899778313,
+      "learning_rate": 2e-05,
+      "loss": 0.091,
+      "step": 18227
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1265739206176897,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 18228
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32661387482042475,
+      "learning_rate": 2e-05,
+      "loss": 0.0707,
+      "step": 18229
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28118270492789493,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 18230
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.128345970851236,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 18231
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.240310893793178,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 18232
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7783420233389413,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 18233
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3384410310708013,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 18234
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4280806720289838,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 18235
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0823935442727046,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 18236
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4227330810409584,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 18237
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0533212360057607,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 18238
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5211543492621474,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 18239
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7031664942047758,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 18240
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.30657428582699403,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 18241
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3991659275053397,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 18242
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3384953600917417,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 18243
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6771262350917944,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 18244
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9607686460434128,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 18245
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14615362975051677,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 18246
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.28377711727099,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 18247
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0718014238645988,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 18248
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5942214800450707,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 18249
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5710277541614925,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 18250
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1780997701113802,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 18251
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32370677418250865,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 18252
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0814591420498312,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 18253
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24200491790907308,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 18254
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8063183895943024,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 18255
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9800579540363896,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 18256
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.77975979189466,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 18257
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3437389458699598,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 18258
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8680993861682814,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 18259
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.002957678889905,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 18260
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.529755622196355,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 18261
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9568188336918764,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 18262
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4793480930461138,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 18263
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28858480417939947,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 18264
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45009886728439574,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 18265
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5608597134407476,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 18266
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5025647167078247,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 18267
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29058053948144447,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 18268
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.955178782666522,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 18269
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7951030852703063,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 18270
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5635991325008267,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 18271
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.383345847552686,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 18272
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38214374048426336,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 18273
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9166250086141055,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 18274
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4161947623757365,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 18275
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0801843518519887,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 18276
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7115760283258439,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 18277
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1548832728699352,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 18278
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7359614538410986,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 18279
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6724433934307765,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 18280
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6205768542451595,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 18281
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2036740315585974,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 18282
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.333368528191484,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 18283
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7411307591835241,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 18284
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8336645223218064,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 18285
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9055345536545367,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 18286
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5230890509172545,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 18287
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8415616231149552,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 18288
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7634143834277898,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 18289
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.41149878806008733,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 18290
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7084765568063438,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 18291
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.829056722454916,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 18292
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.782941450157902,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 18293
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3855507048105618,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 18294
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6876948208847362,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 18295
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8208499419441023,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 18296
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3996741868434825,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 18297
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4260427730214364,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 18298
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7956511879901483,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 18299
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5197197134891983,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 18300
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38989267367156977,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 18301
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.099462771953018,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 18302
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.800932423485431,
+      "learning_rate": 2e-05,
+      "loss": 0.0825,
+      "step": 18303
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7140242622769657,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 18304
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.20640927580377844,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 18305
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1756624180762332,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 18306
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4745665006829967,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 18307
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.019509677001017,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 18308
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6750449430550499,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 18309
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1637126217584988,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 18310
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.695983753828294,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 18311
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9610581361054212,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 18312
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.099056938532603,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 18313
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.23204473944347812,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 18314
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0542357996309222,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 18315
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.044531142716192,
+      "learning_rate": 2e-05,
+      "loss": 0.083,
+      "step": 18316
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.315291613932717,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 18317
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8822238951637021,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 18318
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0403930061796043,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 18319
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8312449433595978,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 18320
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6097653434796756,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 18321
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9038162706016104,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 18322
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5999168314775813,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 18323
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0672068207891034,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 18324
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0610503679883143,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 18325
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7825379650893451,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 18326
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.12406479853039,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 18327
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.453683233195721,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 18328
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16263629694756035,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 18329
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3030235754990296,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 18330
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5007557819335462,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 18331
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.582206500824989,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 18332
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2252378572347613,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 18333
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.297697546489659,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 18334
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8063313977337383,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 18335
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5786196245136244,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 18336
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7381407263029607,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 18337
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.215793839131964,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 18338
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.108432022117011,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 18339
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.775878863305299,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 18340
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3609313435875798,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 18341
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.141533633414599,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 18342
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.304381242992003,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 18343
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4514004512267843,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 18344
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0127415103413484,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 18345
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5253772662992243,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 18346
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1107014530038093,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 18347
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.035074397374011,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 18348
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1310901315346562,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 18349
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6494356997907995,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 18350
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.753913871447778,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 18351
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2922210836601398,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 18352
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.993528027270727,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 18353
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.260737772700621,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 18354
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.33443793791854726,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 18355
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9046007473846593,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 18356
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8836238812099405,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 18357
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1660862171267656,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 18358
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3918030281798097,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 18359
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5885804903396601,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 18360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7823925437514424,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 18361
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9890562795530812,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 18362
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18448655180649623,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 18363
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7188305187734824,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 18364
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0714194739193634,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 18365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8335097881101684,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 18366
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.543368514839051,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 18367
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14346435844972266,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 18368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49489960413369893,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 18369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7145533884452853,
+      "learning_rate": 2e-05,
+      "loss": 0.0776,
+      "step": 18370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5217882906668406,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 18371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.595825414745847,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 18372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09998438404015847,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 18373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0678434643094903,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 18374
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6173017334561632,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 18375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9684698723000758,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 18376
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5872576245295007,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 18377
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3229737407157227,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 18378
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1047353839343341,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 18379
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6890332356193027,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 18380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.670191541223455,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 18381
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.292389846697293,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 18382
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9155801417067522,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 18383
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.19499268096246367,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 18384
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4567576297202107,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 18385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.521043980611537,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 18386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2812403173899213,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 18387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.341997948493995,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 18388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09200786631095836,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 18389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9876860986468039,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 18390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.245657837292428,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 18391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5624842641536755,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 18392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18637148674455759,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 18393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.627872915054551,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 18394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.356260049732744,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 18395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8530652350169088,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 18396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.730189877075934,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 18397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.295851107942573,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 18398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6790087149570618,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 18399
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3292003990046464,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 18400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0243250369071046,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 18401
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8834526644182127,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 18402
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4072485901800567,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 18403
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1452177937527293,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 18404
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.911774698484086,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 18405
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6087494060257956,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 18406
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.33633810376329204,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 18407
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.827421656232619,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 18408
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9357907290001681,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 18409
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0293099398795476,
+      "learning_rate": 2e-05,
+      "loss": 0.0945,
+      "step": 18410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5216271994238383,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 18411
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2702461795263352,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 18412
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3920595720322584,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 18413
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4162936289149267,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 18414
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6320566134310321,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 18415
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3323325529126812,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 18416
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.268452695711567,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 18417
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7692606131642425,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 18418
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2044246144312392,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 18419
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.791301120052849,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 18420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12951867248432697,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 18421
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8366574547236445,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 18422
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16797410047126762,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 18423
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4438328772863427,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 18424
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3136299566095421,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 18425
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9949479479716454,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 18426
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1973864943958863,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 18427
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3259335749527894,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 18428
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7626708902314865,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 18429
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8778481785645722,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 18430
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7198324345601254,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 18431
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4662393888079084,
+      "learning_rate": 2e-05,
+      "loss": 0.0841,
+      "step": 18432
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28040498045754353,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 18433
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2040214666626372,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 18434
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2896030102010765,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 18435
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.607826343565048,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 18436
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8195949867853872,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 18437
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5743217411732577,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 18438
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9557618918561428,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 18439
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6061068218303597,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 18440
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.045531785835956,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 18441
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26551696038717093,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 18442
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9429393330736561,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 18443
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8116627928140719,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 18444
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.089231483398416,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 18445
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4441389716259353,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 18446
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.381458035089908,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 18447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6120561654266616,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 18448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.052403681654641,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 18449
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9716818958046515,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 18450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.567037769657516,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 18451
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.100194202966332,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 18452
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4178361058325145,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 18453
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9460828180413909,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 18454
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3511230989305694,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 18455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.729307175396088,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 18456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3427170620267834,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 18457
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0360162006017284,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 18458
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3570501683068898,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 18459
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0028571939206006,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 18460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1099815344341466,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 18461
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5030362507041378,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 18462
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16943202718840528,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 18463
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1696111046610944,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 18464
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45170035518427903,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 18465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8794758462695964,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 18466
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.272120584683226,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 18467
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.10811162067886687,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 18468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9627206712133303,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 18469
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5460609917372624,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 18470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6117797499602873,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 18471
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.008187985519019,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 18472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.271352640995213,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 18473
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1835608650901677,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 18474
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8768104078196137,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 18475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16699861212970324,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 18476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7359301556587239,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 18477
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6775893570153635,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 18478
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7382839931509255,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 18479
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1867575050579669,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 18480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16228441761384527,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 18481
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3815167158559385,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 18482
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3322846739314305,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 18483
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7037183095625354,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 18484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8523613219018921,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 18485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.564009204608144,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 18486
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3692541805578217,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 18487
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.585357334398565,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 18488
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5065867375500375,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 18489
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6033232174952374,
+      "learning_rate": 2e-05,
+      "loss": 0.1196,
+      "step": 18490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.630002505736439,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 18491
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.051343873030804774,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 18492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7957416777899167,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 18493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.786993771342057,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 18494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.252688920442398,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 18495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3513856498813976,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 18496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29169050471211344,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 18497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5386180150890912,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 18498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1628780611619758,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 18499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.841170262209272,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 18500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5640264039893215,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 18501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5077725614834661,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 18502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.33642724290151005,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 18503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1844725230962734,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 18504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5472414423890808,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 18505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.68209637747485,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 18506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7561870328052205,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 18507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3881521204635408,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 18508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.62132632873668,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 18509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3583327604814135,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 18510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.19428220801679455,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 18511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8737266676029072,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 18512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.092772974966416,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 18513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4270270434555187,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 18514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.056961574598099814,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 18515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.362009560677917,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 18516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.843700873828595,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 18517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.304714316817753,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 18518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2593136945242955,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 18519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.041159408982417,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 18520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7102414461526136,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 18521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7882622919742428,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 18522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7245501470908787,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 18523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.306555924592968,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 18524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5517561920952423,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 18525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07776225335325478,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 18526
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.648399597885968,
+      "learning_rate": 2e-05,
+      "loss": 0.0881,
+      "step": 18527
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1646075612371494,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 18528
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8734543977853272,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 18529
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3065517321196887,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 18530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32878968490645005,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 18531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3075422806756354,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 18532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1443765164609536,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 18533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8738873764774627,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 18534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3097903127029813,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 18535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5325780425394114,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 18536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.074676513671875,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 18537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8041025510089177,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 18538
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1564827392580765,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 18539
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.062138612470912,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 18540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2215358975788602,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 18541
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4042425810251668,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 18542
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15886291625727525,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 18543
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6079349185764078,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 18544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9392974470743295,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 18545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7804613207820679,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 18546
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9400614396774006,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 18547
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09184934759945315,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 18548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4247324809550066,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 18549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.720938625987959,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 18550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8012283937632722,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 18551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7934230386348393,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 18552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.81318113768085,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 18553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4002992851464404,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 18554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3242843296612847,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 18555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07895759444039907,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 18556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.44548737287982754,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 18557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9197254099790517,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 18558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5388958991415148,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 18559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2515677528543739,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 18560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1849732370211403,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 18561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6322357705016989,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 18562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4939529600283081,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 18563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5295484498740106,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 18564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.31164780289805605,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 18565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6952665823769715,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 18566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1352526986720728,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 18567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9988829513021379,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 18568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1524083288789084,
+      "learning_rate": 2e-05,
+      "loss": 0.0821,
+      "step": 18569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5280668599105294,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 18570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0246087281155625,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 18571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3094705005593544,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 18572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7163298538059005,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 18573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6704303902782205,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 18574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.772854338035476,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 18575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0222654740921326,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 18576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.604652438574677,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 18577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7701630510617554,
+      "learning_rate": 2e-05,
+      "loss": 0.0775,
+      "step": 18578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5865572639845399,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 18579
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.62594232379961,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 18580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6727354214696804,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 18581
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.22135698466283663,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 18582
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5629084244038451,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 18583
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28154024406646905,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 18584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.225460810573166,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 18585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.252418807383787,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 18586
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.21936525096315007,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 18587
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7764887108851094,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 18588
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4811557262527064,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 18589
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6804642411672064,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 18590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8969820257786704,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 18591
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26005175614115167,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 18592
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.20978224491364772,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 18593
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3525768729202838,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 18594
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4080268442728711,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 18595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7323409516401476,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 18596
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9178980049306115,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 18597
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5959717691987215,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 18598
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.273290675295345,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 18599
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2776550007081942,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 18600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5322682722594085,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 18601
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15029382686853357,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 18602
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3419409421023425,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 18603
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9119047470417659,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 18604
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.75130847665983,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 18605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7964231294486355,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 18606
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6245435660576817,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 18607
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24322479898968097,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 18608
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5126883775664146,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 18609
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1573038152122086,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 18610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.412411226166713,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 18611
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2656413418692019,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 18612
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0949833591712148,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 18613
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3231112820108333,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 18614
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3250975926743815,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 18615
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7294060951142265,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 18616
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45378196708791196,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 18617
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3167649813965242,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 18618
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.47892799514763473,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 18619
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0019663313502833,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 18620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1907240328956373,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 18621
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2339623159747581,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 18622
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3623925402900188,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 18623
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2885519622636861,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 18624
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4292212123682181,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 18625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.063946534565851,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 18626
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4839331549757696,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 18627
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0552932645947792,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 18628
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.239309025015391,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 18629
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9640028502177516,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 18630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1110131968959969,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 18631
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.730073354081005,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 18632
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.062616678451014,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 18633
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.403791385320158,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 18634
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.40770650246564,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 18635
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.38693214171573,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 18636
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.46392956766004584,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 18637
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4321224805766604,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 18638
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4141272548207696,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 18639
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6442419847974943,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 18640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6788044597492716,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 18641
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.200938082637045,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 18642
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6229673290384186,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 18643
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4236623408499625,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 18644
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3789887092765555,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 18645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7670516521190638,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 18646
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6347063711827179,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 18647
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32527107066574595,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 18648
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6399283246915801,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 18649
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.22890532106074343,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 18650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.041420147107346,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 18651
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.126768440012672,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 18652
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9205368157214087,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 18653
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3730210699772912,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 18654
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5124188068086917,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 18655
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3611359426391536,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 18656
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3513169278665078,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 18657
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.276541252796354,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 18658
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3126170154555554,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 18659
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6627136638125577,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 18660
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6582150588038622,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 18661
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7115702485679142,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 18662
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.365670638504623,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 18663
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16041307094391813,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 18664
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.243982746091935,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 18665
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5426546573852862,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 18666
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3741108565596516,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 18667
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.284015703451283,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 18668
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9735331431714441,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 18669
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0123960959726612,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 18670
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.514670065727564,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 18671
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7489825340750526,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 18672
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.549558513269852,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 18673
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32971728086428975,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 18674
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3023965365635685,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 18675
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9371889541504828,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 18676
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.289929277714056,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 18677
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13424462353251523,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 18678
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.428838452717808,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 18679
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11547900758279607,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 18680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45318609680791877,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 18681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7441604342034123,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 18682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4427100593565751,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 18683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.776014307589859,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 18684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.188554781971748,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 18685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5005780395408672,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 18686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.020882090909497,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 18687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5417543678425947,
+      "learning_rate": 2e-05,
+      "loss": 0.1032,
+      "step": 18688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.09773594941899652,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 18689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3340094038301918,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 18690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4045291968434712,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 18691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3097928602673214,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 18692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1379290344402477,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 18693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.646240505923945,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 18694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8753553758747792,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 18695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4354304341903311,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 18696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3486209835321726,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 18697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49893587841031806,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 18698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6620132753598615,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 18699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.062194053038662996,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 18700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.429730460695129,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 18701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.510281592147582,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 18702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9320947591882497,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 18703
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5350698136742261,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 18704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.246212951150831,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 18705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3699729228998505,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 18706
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7136214990387386,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 18707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2638995282777861,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 18708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3837757096490992,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 18709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.016684548621807,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 18710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9843607023429923,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 18711
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2417528748037046,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 18712
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4553366323868652,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 18713
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.358285842698553,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 18714
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2094939949577286,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 18715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2367944786998442,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 18716
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8408837200363085,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 18717
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8256948898712784,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 18718
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6007689902613589,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 18719
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0040160241577034,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 18720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11544957917411611,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 18721
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0730953299302857,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 18722
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1106749539076957,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 18723
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.22445462742537325,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 18724
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7405804119871227,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 18725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5833422844617029,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 18726
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.171755780577363,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 18727
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2052812451480777,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 18728
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10892663318204432,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 18729
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3099389633918026,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 18730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0704902445648423,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 18731
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.981344523402568,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 18732
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8261859097680944,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 18733
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.453963909204981,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 18734
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9575809156080763,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 18735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6885869799242001,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 18736
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.22634806844364522,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 18737
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9092042735617687,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 18738
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1454930328846316,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 18739
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4215131026612444,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 18740
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.03954815105244,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 18741
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6149757447150493,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 18742
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0896644581534791,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 18743
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6744922405924891,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 18744
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16470524926603008,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 18745
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5736755494072614,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 18746
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8033465721759542,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 18747
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.41511534943429007,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 18748
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9417614985980498,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 18749
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.09356099189111115,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 18750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.15080103052674754,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 18751
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5372578065741904,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 18752
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3273427566964675,
+      "learning_rate": 2e-05,
+      "loss": 0.0755,
+      "step": 18753
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.550732043177592,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 18754
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1895383606987173,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 18755
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25286948470692944,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 18756
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0774392974287972,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 18757
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2486753611630373,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 18758
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6388680632753365,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 18759
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2639974068086232,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 18760
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14499905631152907,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 18761
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8008160701537983,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 18762
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6493639037758943,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 18763
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1161379469377963,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 18764
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4473921376890717,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 18765
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3472643797501335,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 18766
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0483140875642223,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 18767
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14146479875099596,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 18768
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0663151964153865,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 18769
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.08018348322312646,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 18770
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1293077645820848,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 18771
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.39375669155792714,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 18772
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4782148531485846,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 18773
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0119681621639687,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 18774
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2465746199869046,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 18775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.615653964876889,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 18776
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5846578484687238,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 18777
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1040856707110755,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 18778
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8946157690373883,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 18779
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.127911760623202,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 18780
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.38048552724831325,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 18781
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.177468979063802,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 18782
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.036578087893528,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 18783
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5220127643416733,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 18784
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1681838045040286,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 18785
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2381544078278752,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 18786
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9500220411091482,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 18787
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.006696695291067,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 18788
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9787335486265674,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 18789
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6990137705490215,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 18790
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7878383545376261,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 18791
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3262628668224625,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 18792
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5842133650387439,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 18793
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.020569196251216,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 18794
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5063096329217363,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 18795
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4062186343615748,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 18796
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4021686246899112,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 18797
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8400382238160812,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 18798
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8583117757022107,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 18799
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3550160996159808,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 18800
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.015267628770449,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 18801
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1139956674953841,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 18802
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3494230326681385,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 18803
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3338569710425765,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 18804
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.389687168334894,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 18805
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12399978111713073,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 18806
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.550831829946952,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 18807
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0462421865902485,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 18808
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1188521354554335,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 18809
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.471213689078886,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 18810
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3024395519002274,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 18811
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30519777752759525,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 18812
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5240366222256636,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 18813
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2146068967667691,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 18814
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.23902979514863348,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 18815
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1949034221530266,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 18816
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4683585520539357,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 18817
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.172433575491572,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 18818
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16343225610208828,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 18819
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1139695566150745,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 18820
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2702290327025363,
+      "learning_rate": 2e-05,
+      "loss": 0.09,
+      "step": 18821
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.617835585838673,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 18822
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3479106165403661,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 18823
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6851036566077884,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 18824
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4137917839226566,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 18825
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24199560469338294,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 18826
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0998141977139153,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 18827
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1789424834561757,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 18828
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8011457028924938,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 18829
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5252298726052428,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 18830
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6506702864292498,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 18831
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4758395892799911,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 18832
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4966375493117932,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 18833
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2740989586427983,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 18834
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.145010942048675,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 18835
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6390027930015094,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 18836
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.665060970576178,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 18837
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30210208697587826,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 18838
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14211025943449426,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 18839
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8990963717521815,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 18840
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.395528596337743,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 18841
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.530152297200368,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 18842
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4683041809051605,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 18843
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5077186204256564,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 18844
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8856168445865835,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 18845
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.096523099612368,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 18846
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4380326320887675,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 18847
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.21308304026843006,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 18848
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.378395363957326,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 18849
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.623730163335585,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 18850
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9868111862297795,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 18851
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12812275913651483,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 18852
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6251393771920897,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 18853
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7861855814975797,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 18854
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.0856685946288875,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 18855
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6054585011979836,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 18856
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5743874584884754,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 18857
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3691614538704477,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 18858
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18297060259460182,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 18859
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0606033389775598,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 18860
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4201484069148927,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 18861
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4130946266524056,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 18862
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.106964557211045,
+      "learning_rate": 2e-05,
+      "loss": 0.1151,
+      "step": 18863
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5519475290024235,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 18864
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36585969483815495,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 18865
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7526152053270017,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 18866
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1772658176849784,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 18867
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1624042040824345,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 18868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.480370188630446,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 18869
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.057294576779653,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 18870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4100880698842952,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 18871
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.038327581033545,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 18872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1141361099899587,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 18873
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7388367868864067,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 18874
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.287016178515897,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 18875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3959283547372693,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 18876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0041162887113688,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 18877
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7154206829665557,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 18878
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9914046676057974,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 18879
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.33267739703834,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 18880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5668674038648307,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 18881
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3926406734608764,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 18882
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3617498765234193,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 18883
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.15159157149731806,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 18884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07304709518623825,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 18885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8139414481583471,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 18886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25380731268952444,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 18887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4113706865597443,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 18888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1003593204745772,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 18889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7856440639312692,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 18890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9034620019261507,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 18891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.732882748878578,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 18892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.38025938215558,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 18893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.298678920851349,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 18894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3287061132211879,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 18895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.691975590837657,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 18896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4004827705829759,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 18897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33534466753599423,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 18898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1063128911476656,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 18899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.407656344482832,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 18900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5072788223039204,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 18901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9298758436205073,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 18902
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2645642755558335,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 18903
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17356847272582027,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 18904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3819902893449906,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 18905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.484711224537897,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 18906
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6525956631439782,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 18907
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7335354259250118,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 18908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32542389634332,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 18909
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.641134055953006,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 18910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.3332830328345,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 18911
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.344765080140649,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 18912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.04316988248771069,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 18913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30060804011908415,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 18914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8907251469038093,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 18915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3137972371002845,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 18916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.09959722426111683,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 18917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.488137216305226,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 18918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6013634835525928,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 18919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4080792289234322,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 18920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3304255310727966,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 18921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.22339473361770057,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 18922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8128401996065986,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 18923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5531742515835087,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 18924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7096540162801664,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 18925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.018174208026069,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 18926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.209434856737371,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 18927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0963539897531627,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 18928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.09546864258280953,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 18929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6069580002087218,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 18930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.698998038125457,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 18931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.15357652140684555,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 18932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7567224429771204,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 18933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7483402005781806,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 18934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.511813196094402,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 18935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.336544991843014,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 18936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9539854671949102,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 18937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3853933567868209,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 18938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3723697647813473,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 18939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32123762374074577,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 18940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5521400321827907,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 18941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6508213942145405,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 18942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7222134444934153,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 18943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4875996948941423,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 18944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9290553917277685,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 18945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7853874131951628,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 18946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2841700916046597,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 18947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3850897448689914,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 18948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.57044931528801,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 18949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5291166381245214,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 18950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7603799253671395,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 18951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9216750299845734,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 18952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0481653374750297,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 18953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5418125117968149,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 18954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3080671930640393,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 18955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6554310051295222,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 18956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5996415070612924,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 18957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.553356194894121,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 18958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3972035510208345,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 18959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.41267791509178176,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 18960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.144497341311784,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 18961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5284772488236227,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 18962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9766188948560206,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 18963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.891478038687577,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 18964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.531000000144515,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 18965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9292951795095936,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 18966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7551504365180697,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 18967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10765957975290316,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 18968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37961258875304343,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 18969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.47308848108022,
+      "learning_rate": 2e-05,
+      "loss": 0.0853,
+      "step": 18970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25106897396869543,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 18971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0206587246370524,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 18972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8239974678005242,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 18973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8063132889621566,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 18974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07775224631194652,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 18975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10363994636382473,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 18976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11437233459253988,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 18977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9767875717197232,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 18978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.22728620239971628,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 18979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.687393960577833,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 18980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0609910242217473,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 18981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2284473103025026,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 18982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1052045497139877,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 18983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9033362804754212,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 18984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0128951401514588,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 18985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.669727109043446,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 18986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6398820920753492,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 18987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9557817047951218,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 18988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7224763801040122,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 18989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3844245328234146,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 18990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.48938894132631006,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 18991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1805112375405555,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 18992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6591680318441764,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 18993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3234015900154819,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 18994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7137976741735778,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 18995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.360616529685893,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 18996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2773233581221823,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 18997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34447663624728764,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 18998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0550731377764453,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 18999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.747288928129031,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 19000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5368018418282667,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 19001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.834180892135451,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 19002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37855735918786415,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 19003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2343927696011325,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 19004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.509644446457865,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 19005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9407987731913431,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 19006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4286449268589501,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 19007
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.903983900317159,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 19008
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2752844789820528,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 19009
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8514094783940874,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 19010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5650980140535482,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 19011
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3120581483398095,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 19012
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1219839887887562,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 19013
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1204110436081478,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 19014
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3479585612575158,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 19015
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0170968761008685,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 19016
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3596827599008463,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 19017
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1318913872986804,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 19018
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.291260929537012,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 19019
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2483091358817857,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 19020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3726819612964218,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 19021
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5426125417054357,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 19022
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0794089039313561,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 19023
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.9626168748243966,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 19024
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9706649226052844,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 19025
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7304535944772284,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 19026
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7312029450008988,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 19027
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.57691523950787,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 19028
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5777008695319668,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 19029
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.056666041705963364,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 19030
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1582325389974399,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 19031
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6965182890149337,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 19032
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.723693120403872,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 19033
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0297833619479686,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 19034
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4374941246455638,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 19035
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4021474674281703,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 19036
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0252865994462512,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 19037
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0569735582485538,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 19038
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9543076863543585,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 19039
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.332175212580945,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 19040
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1184916050956826,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 19041
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8653559545514768,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 19042
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36663074082602115,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 19043
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.367334499630802,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 19044
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0745247699763485,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 19045
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.478442090724219,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 19046
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.034784030373359,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 19047
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.188266292958077,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 19048
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6259078346646438,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 19049
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.505242883242411,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 19050
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7539073963848352,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 19051
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6003475801702045,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 19052
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5366222877940395,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 19053
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6552925221291035,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 19054
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.261622800219814,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 19055
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07375639024019223,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 19056
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3971943790997643,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 19057
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6460276702233008,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 19058
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9862314006855032,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 19059
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4333571568808379,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 19060
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12543640845773715,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 19061
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.42220148945748526,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 19062
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13249416628626925,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 19063
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.048092103716941115,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 19064
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.790692534275644,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 19065
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1968009507501876,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 19066
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.790304471808912,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 19067
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7605384093422293,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 19068
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3502982957971736,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 19069
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.915216223178139,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 19070
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5713356442656348,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 19071
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.15308005953145856,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 19072
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5146363345610504,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 19073
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2913967427602515,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 19074
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.22716635767014726,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 19075
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26276904351269514,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 19076
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.39534852387710057,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 19077
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8433056086003146,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 19078
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.517965725851077,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 19079
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7638784902204372,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 19080
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4114858956680845,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 19081
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37927165551617287,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 19082
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.22963083015759225,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 19083
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.573569046484653,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 19084
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.083470384455866,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 19085
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6240537871859053,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 19086
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6436408163273246,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 19087
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6132078673565365,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 19088
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4924733951567963,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 19089
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1724174425880585,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 19090
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28921642586935037,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 19091
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.812094745046986,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 19092
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9765421140450319,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 19093
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4131547321901843,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 19094
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9857748765024407,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 19095
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.35866673792619125,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 19096
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2708513644247328,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 19097
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6440788607280883,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 19098
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.751263388833369,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 19099
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.412303027650092,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 19100
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4053275579723516,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 19101
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9761156021735384,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 19102
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2280682111428023,
+      "learning_rate": 2e-05,
+      "loss": 0.091,
+      "step": 19103
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.264766451518906,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 19104
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7406644351938776,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 19105
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4041850538921529,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 19106
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1965755197529515,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 19107
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5944347448637101,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 19108
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4538121393359986,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 19109
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6857343890019127,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 19110
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6646138538977817,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 19111
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8879020290758433,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 19112
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16392068751145644,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 19113
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8827987607082668,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 19114
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.31992304781930503,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 19115
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5827315676032656,
+      "learning_rate": 2e-05,
+      "loss": 0.0903,
+      "step": 19116
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1425006546753191,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 19117
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0809507518175467,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 19118
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.179429183871616,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 19119
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49515450541994555,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 19120
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6361924318901374,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 19121
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6921236013019767,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 19122
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2093295835498545,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 19123
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2244264992383367,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 19124
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0651360517111348,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 19125
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9357991289961534,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 19126
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0803582654224315,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 19127
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.484953195495774,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 19128
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12678478524138243,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19129
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6720602534194585,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 19130
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4849790068143452,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 19131
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8147601800957025,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 19132
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.155645109159503,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 19133
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.81477613013084,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 19134
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0464478945238358,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 19135
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3042285102959586,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 19136
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.63290187785861,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 19137
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.180469589504214,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 19138
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7077989750472282,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 19139
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4920270743687643,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 19140
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1711537710948674,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 19141
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1871714395469761,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 19142
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.893797994538757,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 19143
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6876037000653088,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 19144
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6720866131776109,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 19145
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2583069865940073,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 19146
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9833519478560847,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 19147
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8097988258180109,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 19148
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6289258392193635,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 19149
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.095319840047313,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 19150
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9357123497378208,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 19151
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.41279602699419377,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 19152
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25822913577002776,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 19153
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2928297538611055,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 19154
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.813539141354308,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 19155
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8186164780727208,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 19156
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6664630136765743,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 19157
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.09866379732728683,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 19158
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.008942495459772,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 19159
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5077593551185915,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 19160
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.712461933353324,
+      "learning_rate": 2e-05,
+      "loss": 0.1342,
+      "step": 19161
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.005904054112375,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 19162
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33713066767594907,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 19163
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8523475458076355,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 19164
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2288296393725222,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 19165
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8201012877690756,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 19166
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2675280450882118,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 19167
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0656015902546023,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 19168
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.425108053692364,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 19169
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.39490480018175106,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 19170
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5856162060533974,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 19171
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.577371096173226,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 19172
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.212140673062126,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 19173
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4128443070450851,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 19174
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.798917521720833,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 19175
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.500640017344154,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 19176
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4151852191976921,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 19177
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.530796373087601,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 19178
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1627716998001019,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 19179
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3929679506891153,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 19180
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.204245492418641,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 19181
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3976149748440503,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 19182
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0507397838479806,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 19183
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.85308601376982,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 19184
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.76777921959275,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 19185
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.459874056779459,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 19186
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5715084078696759,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 19187
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2153580242487039,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 19188
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.309022711905242,
+      "learning_rate": 2e-05,
+      "loss": 0.1115,
+      "step": 19189
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7578384736579313,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 19190
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.576157986748334,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 19191
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1195998389775146,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 19192
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3424022762600325,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 19193
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.22094947434475265,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 19194
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9626918617926044,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 19195
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8946236308674864,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 19196
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6158699990749463,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 19197
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4409455180725839,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 19198
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9932849551640778,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 19199
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26645063662554047,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 19200
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6401864853753799,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 19201
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9701145832025266,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 19202
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.293326614761861,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 19203
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.09755432054632865,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 19204
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9594070217988924,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 19205
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7598892490263418,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 19206
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.612095135865529,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 19207
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.153078040509203,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 19208
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8283091286423901,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 19209
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8162367795425634,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 19210
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1889188221008422,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 19211
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.505863455342333,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 19212
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2291743435831473,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 19213
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4262674185409145,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 19214
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.578237397374268,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 19215
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.129974371435714,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 19216
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.078134063471483,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 19217
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.24972929305169958,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19218
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7224571159861926,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 19219
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.716366630964595,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 19220
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3349205295611553,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 19221
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8845319692036109,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 19222
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2386535721798844,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 19223
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.335894836240305,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 19224
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5594051992668514,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 19225
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.792028612082403,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 19226
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9911856511841628,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 19227
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9658001795882856,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 19228
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.806425834465787,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 19229
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7775455002181644,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 19230
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0523996475655557,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 19231
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5212804929487258,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 19232
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7303094726006905,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 19233
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3753620451341595,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 19234
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4356416505069743,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 19235
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2679060715310221,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 19236
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5288307699553532,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 19237
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.316945093275393,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 19238
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9441981186463364,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 19239
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7963309086235664,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 19240
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.73310639517048,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 19241
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.405381182086123,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 19242
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2837327149037637,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 19243
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6720517835249563,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 19244
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5114681421439347,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 19245
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.23743068819298624,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 19246
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.938759302231333,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 19247
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8544048155665865,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 19248
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30767113348829045,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19249
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0961217399943526,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 19250
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.183862877708679,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 19251
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5370687360614168,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 19252
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2673215262240568,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 19253
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4265062970792275,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 19254
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3351453696980016,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 19255
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6913882538775653,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 19256
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9032419201626501,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 19257
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6416634315936491,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 19258
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8183397072815244,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 19259
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5889283966086336,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 19260
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.296966549503026,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 19261
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3120509242229172,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 19262
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5837466671540161,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 19263
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.777601255671011,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 19264
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5098610042236322,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 19265
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7295279604175495,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 19266
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06942032326396398,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 19267
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.637870801918969,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 19268
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3129375146984437,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 19269
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7194313882954734,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 19270
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2841946181737042,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 19271
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3395252349657951,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 19272
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7418547838215581,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 19273
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7003073077368956,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 19274
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9061647572382973,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 19275
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3658233218825389,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 19276
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5227332715587404,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 19277
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.14130806596574375,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 19278
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8877641674233687,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 19279
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7538863263437636,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 19280
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7172265004478975,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 19281
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2620312094207262,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 19282
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2167556290348378,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 19283
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4294397506865859,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 19284
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.38293643813548695,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 19285
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3956602852815694,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 19286
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1918623886419986,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 19287
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7173795904081475,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 19288
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.579671389620674,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 19289
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3859345703782144,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 19290
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5267583069818458,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 19291
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5910394183485771,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 19292
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0425471335900531,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 19293
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8362688864040884,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 19294
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2960281364659776,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 19295
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.22078477986261608,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 19296
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.75334544863433,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 19297
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4380674088701451,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 19298
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2190985607316343,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 19299
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.576470611197035,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 19300
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.942235006982811,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 19301
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7490106016574223,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 19302
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.670334158060669,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 19303
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.09554722310218812,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 19304
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46426462915509065,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 19305
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39341888735181535,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 19306
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1645378831832902,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 19307
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8840727960732622,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 19308
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4261808488616259,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 19309
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07503201042470756,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 19310
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.650991912754635,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 19311
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5054512586913237,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19312
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.823436243492907,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 19313
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.04347215212595558,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 19314
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7772047649219047,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 19315
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0483116479779824,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 19316
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4864020693532238,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 19317
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.483136272556373,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 19318
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3689240244841476,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 19319
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5667927179900512,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 19320
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3085987071647285,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 19321
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3003928305145225,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 19322
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9246211771177543,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 19323
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3328028296914867,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 19324
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8470744821127196,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 19325
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.833109805804907,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 19326
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0662740548607919,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 19327
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3110998616367855,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 19328
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4132607125408385,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 19329
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2610081091731846,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 19330
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9289801323073545,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 19331
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5656520162632279,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 19332
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.45795511947125156,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 19333
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.045718533429192,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 19334
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5452195373899753,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 19335
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.060724013972146,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 19336
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.345085500601779,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 19337
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.769455467828034,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 19338
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5420840348784793,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 19339
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3829378389967484,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 19340
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.823473025612208,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 19341
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6571198103251531,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 19342
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8676680580926919,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 19343
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8427883248544779,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 19344
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1391268781305137,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 19345
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1957009874715459,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 19346
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9874266899034352,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 19347
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4167130219598378,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 19348
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34462481501964853,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 19349
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.095111431771956,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 19350
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4982377668008942,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 19351
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6779425495211115,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 19352
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4407192386466794,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 19353
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0087386614458882,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 19354
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17626741783164304,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 19355
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.28405547405691,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 19356
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1055435988333593,
+      "learning_rate": 2e-05,
+      "loss": 0.057,
+      "step": 19357
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.49606156432505844,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 19358
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6229421593373274,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 19359
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6798446226929384,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 19360
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2532174100720956,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 19361
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1428567139164938,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 19362
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.641764395153723,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 19363
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9805085679438086,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 19364
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.09275648564511232,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 19365
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1176906599701826,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 19366
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6583991056092029,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 19367
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5283138384697144,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 19368
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3396527202878006,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 19369
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.919558335960286,
+      "learning_rate": 2e-05,
+      "loss": 0.0813,
+      "step": 19370
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4578437856537003,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 19371
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8628895418946915,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 19372
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3761586162914863,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 19373
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1256591137636518,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 19374
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5947397415379624,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 19375
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3078512943961313,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 19376
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3469286602201858,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 19377
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2157156412806438,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 19378
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13919626403301802,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 19379
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07201668674847521,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 19380
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7912566771383723,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 19381
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8217687677641567,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 19382
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3525975038902863,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 19383
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4376092952145434,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 19384
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.736557008154702,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 19385
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0832914319860467,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 19386
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08608330744208344,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 19387
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7059116002182706,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 19388
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5680624560840744,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 19389
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5809015152299255,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 19390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.710610066614555,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 19391
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0668551384704323,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 19392
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9568466631320709,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 19393
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3260999231236728,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 19394
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5957089520911811,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 19395
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.95357255723134,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 19396
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5824767334429463,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 19397
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15382080029688244,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 19398
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6780899829876775,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 19399
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0383014616710444,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 19400
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.42378788418623164,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 19401
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.678120862788778,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 19402
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0837386182120636,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 19403
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8679373995581114,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 19404
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2818521834652469,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 19405
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9974156005069214,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 19406
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5233677774760337,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 19407
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1804193411887791,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 19408
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.19975032197284173,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 19409
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3426931261806594,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 19410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.47618209790341054,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 19411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06803170760508544,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 19412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4946565821769962,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 19413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2887650454304668,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 19414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06745484551565455,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 19415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6302108738325212,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 19416
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9197822351429352,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 19417
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.238772514438909,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 19418
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.25178060613395764,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 19419
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1925593564917862,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 19420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.41407892806374486,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 19421
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11630506870491476,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 19422
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4238317990473404,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 19423
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1812091748529529,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 19424
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.089107852905448,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 19425
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3088872371215263,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 19426
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.094445753504678,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 19427
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5585080077791995,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 19428
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06618423725318093,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 19429
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5760531545623686,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 19430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4762957215649486,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 19431
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1764929255650984,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 19432
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3432193691995178,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 19433
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8100196045925578,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 19434
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3288791126409216,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 19435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.03342728644596655,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 19436
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8412721773259672,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 19437
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0488662570720368,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 19438
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7992620783436613,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 19439
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39318264018632854,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 19440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.41794390025200545,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19441
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.469341483632403,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 19442
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8095808776239705,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 19443
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1192507598104107,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19444
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.371273102304568,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 19445
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8671275492345116,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 19446
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0073666801220837,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 19447
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1532642890314537,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 19448
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0376806033021937,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 19449
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.40809147964384657,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 19450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7338023286320359,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 19451
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5352802566257693,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 19452
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.25832704291581116,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 19453
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9498124374759895,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 19454
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9370944306215796,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 19455
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7132671459261755,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 19456
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9016838137558636,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 19457
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.20694014506030692,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 19458
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.381354948518773,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 19459
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4035469323685999,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 19460
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.20004735922373057,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 19461
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4143266799905723,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 19462
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.035431389860683,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 19463
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4552744004195613,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 19464
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6482692108307673,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 19465
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.643106843710295,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 19466
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.760573414476363,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 19467
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1680285746250267,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 19468
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4702009785079762,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 19469
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07879424147636815,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 19470
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.304925314421347,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 19471
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9215271503631927,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 19472
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7130496331123214,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 19473
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2214787086914114,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 19474
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1944580557057429,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19475
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.848791374584399,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 19476
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3010832289976655,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 19477
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5621951919801292,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 19478
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6067679463810384,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 19479
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2640685950829686,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 19480
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2018426774350799,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 19481
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7230499432383787,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 19482
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.580496050291507,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 19483
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5129707927822043,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 19484
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1673448101698072,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 19485
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.365232934255208,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 19486
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.275122368843683,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 19487
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.25122722769254535,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 19488
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12267201331464496,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 19489
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33570140482191263,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 19490
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4122172210019532,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 19491
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.05560199514659963,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 19492
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7683785529096974,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 19493
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.610100285587505,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 19494
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9678818596733904,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 19495
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4206641091449353,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 19496
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.95154882086081,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 19497
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3693075255184748,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 19498
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.0635934284669295,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 19499
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3999402374045984,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 19500
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8872822616999154,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 19501
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7872454818969588,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 19502
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.940267030695,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 19503
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.605348124518753,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 19504
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3506414397180957,
+      "learning_rate": 2e-05,
+      "loss": 0.1399,
+      "step": 19505
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7839523118803536,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 19506
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.921842536943894,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 19507
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3531018349864916,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 19508
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7191283763196914,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 19509
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.340100411168794,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 19510
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.386011748951651,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 19511
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6763377299382285,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 19512
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7501035040240627,
+      "learning_rate": 2e-05,
+      "loss": 0.1033,
+      "step": 19513
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.718182754584581,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 19514
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.37456609642104,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 19515
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9492500012172589,
+      "learning_rate": 2e-05,
+      "loss": 0.1074,
+      "step": 19516
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9311436381180613,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 19517
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0531058425745594,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 19518
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6466445935400549,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 19519
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3123765582419399,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 19520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8989794550181565,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 19521
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6968967752482111,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 19522
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26415324598712,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 19523
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3874003636026129,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 19524
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4401241896641706,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 19525
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27633982276908337,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 19526
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8697070044394344,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 19527
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5426961537287542,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 19528
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.09559721341208051,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 19529
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.108196552802292,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 19530
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2937550530818722,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 19531
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1389356197960527,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 19532
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1142643920811164,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 19533
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46853890434164386,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 19534
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3312937343264854,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 19535
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7045419826100401,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 19536
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.484932403310374,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 19537
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.858966452848326,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 19538
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7496141792370028,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 19539
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.36066295509438,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 19540
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.078775403197734,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 19541
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6847178997984926,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 19542
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3747435677435655,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 19543
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46702996025022225,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 19544
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2370922931906887,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 19545
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6959823425056383,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 19546
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9679435018551759,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 19547
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.438448898700171,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 19548
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.787838505849475,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 19549
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1339795645028004,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 19550
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3048987303278108,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 19551
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.18500152553431973,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 19552
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2985578901204753,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 19553
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16393952074743373,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 19554
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7054386643661795,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 19555
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5464759851484895,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 19556
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.834613078847518,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 19557
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.833796254870268,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 19558
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5484156025484608,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 19559
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3924478306748598,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 19560
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.20360088539354929,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 19561
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.647183271072277,
+      "learning_rate": 2e-05,
+      "loss": 0.0841,
+      "step": 19562
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6235631161293181,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 19563
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4295824356246765,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 19564
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28227302142476807,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 19565
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3629653131106327,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 19566
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1550979545472775,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 19567
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.43406033520609,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 19568
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2808286902122135,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 19569
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6194580650447472,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 19570
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.534626721402227,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 19571
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5829667704378463,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 19572
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0535386787016634,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 19573
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27499138861958233,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 19574
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1489410365177952,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 19575
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.305817851080502,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 19576
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.407221437401211,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 19577
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7664716282450396,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 19578
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8562959453406755,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 19579
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3112869920748755,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 19580
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1535234766332128,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 19581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.38985737717534624,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 19582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7060126226836962,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 19583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5629033920460347,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 19584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8305123304270464,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 19585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.47712150994299374,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 19586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0572017512430731,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 19587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9157694015947057,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 19588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13139097215649925,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4996514595685331,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 19590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.328755845983785,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 19591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7694011907512275,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 19592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1799942022116519,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 19593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3671944394368987,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 19594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7727873796019918,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 19595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4861693128238769,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 19596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9161243496201703,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 19597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.090360512128721,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 19598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6299895930981887,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 19599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.829623182201007,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 19600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.766221506277212,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 19601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6202423941617687,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 19602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34466585700044833,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 19603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.405692791517534,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 19604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6461567812228963,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 19605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.585706716228546,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 19606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15710118430329856,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 19607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3034413948449033,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 19608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1721616048178984,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 19609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.38164767333972843,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 19610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4665994106800382,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 19611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0794820675652497,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 19612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5489915951524809,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 19613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4934527015156362,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 19614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.509956845372267,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 19615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1008838159063712,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 19616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2995338301032744,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 19617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7647671857002116,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 19618
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2861516280597391,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 19619
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6680471363140724,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 19620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.024017788399223125,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 19621
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2733928916511468,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 19622
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.03423105195099444,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 19623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.057273815455325935,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 19624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.994359739847053,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 19625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6021860537030468,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 19626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.362784139510653,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 19627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8683673811256236,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 19628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9290619356350639,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 19629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.597821259830453,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 19630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7930095061280495,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 19631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5181538828728494,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 19632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0263053374524018,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 19633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1162670585294303,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 19634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.546660054063152,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 19635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2958185320943793,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 19636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.558563471687106,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 19637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3389632554372097,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 19638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8016958169169863,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 19639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.964016084465606,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 19640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.713389899152538,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 19641
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.868474221692533,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 19642
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7740368607485804,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 19643
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8281537896676078,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 19644
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.40312805396409174,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 19645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6044497374880591,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 19646
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5172265640211541,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 19647
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5956673402170567,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 19648
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3915055459288633,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 19649
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.416140662966086,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 19650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.42905125763702917,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 19651
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.20229470678599323,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 19652
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.545610932480115,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 19653
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3937967241074705,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 19654
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.05738538041351403,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 19655
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9234651977722723,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 19656
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7084393889372975,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 19657
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8980631172666008,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 19658
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4692654618258925,
+      "learning_rate": 2e-05,
+      "loss": 0.1049,
+      "step": 19659
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5128275673446768,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 19660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3873454222619115,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 19661
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7147845050879578,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 19662
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.201550382198349,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 19663
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6835381660225193,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 19664
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4256165291728111,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 19665
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4303522170332714,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 19666
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6609481486010534,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 19667
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4531409970295122,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 19668
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7759080980977874,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 19669
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3033333792446613,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 19670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.653808099629094,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 19671
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4523888231132949,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 19672
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2244490377055757,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 19673
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7772579049867621,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 19674
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.03946676130742024,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 19675
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.36286783793573,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 19676
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.195826258313746,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 19677
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0855062088875087,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 19678
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13398151060996263,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 19679
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.959194022327665,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 19680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3304401308658086,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 19681
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2349959067741807,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 19682
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2158032519510853,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 19683
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.19782870726457896,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 19684
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4191302121927802,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 19685
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.635581994254773,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 19686
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8425751383184237,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 19687
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5127614005954253,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 19688
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2024324018343124,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 19689
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6484802415848081,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 19690
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6799842142979587,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 19691
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5314724818941645,
+      "learning_rate": 2e-05,
+      "loss": 0.0864,
+      "step": 19692
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.22236810410085536,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19693
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5541966644093805,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 19694
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8931892022774701,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 19695
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6372054878303866,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 19696
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.41026266624870616,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 19697
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5152251512522577,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 19698
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6944657421684413,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 19699
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9589426377298108,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 19700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7649264557975566,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 19701
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1014232682780203,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 19702
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8123669148535992,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 19703
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.15926623185793554,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 19704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2597947228198387,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 19705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.590054152395822,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 19706
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7271658330491882,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 19707
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9203203938308755,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 19708
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.047824372208094,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 19709
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4923893196859269,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 19710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06910981997981235,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 19711
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7904071115597594,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 19712
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6285347403967703,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 19713
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.309923914697588,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 19714
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.636639361902815,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 19715
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.49630462489973726,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 19716
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1650332301234458,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 19717
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.413292058478462,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 19718
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0398576031914952,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 19719
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9415471660966932,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 19720
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1116590778643007,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 19721
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7282829424435966,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 19722
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4938916923786865,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 19723
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0122107410500938,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 19724
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5590423473799782,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 19725
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3735511777674914,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 19726
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12263098322856487,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19727
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.920490446838209,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 19728
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3291513852682968,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 19729
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7745851004953654,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 19730
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0001741496079908,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 19731
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.568027489399172,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 19732
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3453096165997203,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 19733
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.24873803183024618,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 19734
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11563531178810242,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 19735
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5249372036026169,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 19736
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1331695520108327,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 19737
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3033111073451713,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 19738
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.55821754077549,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 19739
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7000146163985261,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 19740
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.24234591194728147,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 19741
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07121772332014274,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 19742
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3076884758746297,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 19743
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8830221821730159,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 19744
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1937307875090564,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 19745
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.278967848536955,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 19746
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17073753411683473,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 19747
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.350857236455767,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 19748
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6014780941579243,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 19749
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6907083469690101,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 19750
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.278697950947877,
+      "learning_rate": 2e-05,
+      "loss": 0.0732,
+      "step": 19751
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.133271667463417,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 19752
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5388089663427527,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 19753
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7812621687895065,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 19754
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6703075298238947,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 19755
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.163681051741349,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 19756
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9273184624270538,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 19757
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2325394439060897,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 19758
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1092625950564003,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 19759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0996991352962804,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 19760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2899564475462582,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 19761
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.723603971850603,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 19762
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0142479265914193,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 19763
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9665408379770914,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 19764
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9722991655571549,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 19765
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.20386273953931708,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 19766
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7995943605854396,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 19767
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3959563851087815,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 19768
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7190391990803624,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 19769
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.268090143227607,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 19770
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9374993324277409,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 19771
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18881231258348938,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 19772
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27490130029619214,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 19773
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.70328640318123,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 19774
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.292229663018011,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 19775
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8499443358539078,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 19776
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2818301627875425,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 19777
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3813264264672958,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 19778
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6648935615356909,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 19779
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.414392675136785,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 19780
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.752432787470429,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 19781
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28399792655923334,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 19782
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3555386862234072,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 19783
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4561116691638,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 19784
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2591055626794267,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 19785
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16693547607624606,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 19786
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8580856334545297,
+      "learning_rate": 2e-05,
+      "loss": 0.0785,
+      "step": 19787
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9846491356255211,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 19788
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5050766534038916,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 19789
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.37086115328649,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 19790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5305350485539486,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 19791
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.272691008819651,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 19792
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6741064990814425,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 19793
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1413040230239184,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 19794
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5038607350704316,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 19795
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.886732713477864,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 19796
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.436015814138908,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 19797
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1159208105012475,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 19798
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7619517779122074,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 19799
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2323303470888747,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 19800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1786552611999548,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 19801
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7914316096107358,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 19802
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6991539159294793,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 19803
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4439673685365315,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 19804
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5871937318707698,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 19805
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.0802665144695585,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 19806
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.20186355345784965,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 19807
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0928871312589825,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 19808
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.779840948234134,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 19809
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3449428054816253,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 19810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0031084699737227,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 19811
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7756948955067464,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 19812
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8675693324303475,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 19813
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25602176841649715,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 19814
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6763823597567001,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 19815
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.560611692875454,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 19816
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10225667215621194,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 19817
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.08522337632321576,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 19818
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5967726065401312,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 19819
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7400027997376344,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 19820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.367299999421696,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 19821
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.86734988709723,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 19822
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4616362399559387,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 19823
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.808157688078669,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 19824
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.121687697709566,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 19825
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.24217500962073957,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 19826
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7678345307794232,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 19827
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2273100500500504,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 19828
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8428457500955511,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 19829
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.22109314544379408,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 19830
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16600515108166677,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 19831
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8782757001917637,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 19832
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1815067638885284,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 19833
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4229236969371541,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 19834
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.122239328375725,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 19835
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6129112828543598,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 19836
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4304950581454143,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 19837
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9464142422456017,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 19838
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1766520979000092,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 19839
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0107596175710665,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 19840
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9613460819655195,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 19841
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9436764865723786,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 19842
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.807172881380593,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 19843
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0768722437106657,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 19844
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4643671493585451,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19845
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7129908661063352,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 19846
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3447235798909347,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 19847
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.367007207771676,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 19848
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.515946572749535,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 19849
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.307843948668848,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 19850
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.02949551804786,
+      "learning_rate": 2e-05,
+      "loss": 0.1231,
+      "step": 19851
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7949463561691154,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 19852
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6990084310868467,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 19853
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 39.91578659542174,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 19854
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17579136925286615,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 19855
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45775589186096993,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 19856
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1558301212781825,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 19857
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9702529619918532,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 19858
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7449594233170798,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 19859
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.479165663741224,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 19860
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.44105395884525883,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 19861
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3274006831918153,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 19862
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3125234833476933,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 19863
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.32477754673332604,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 19864
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2772848239916406,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 19865
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8541884688027075,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 19866
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.742370338539263,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 19867
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7772912987586307,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 19868
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5334285497110082,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 19869
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13899627539177906,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 19870
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5760912305354169,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 19871
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0696845346704635,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 19872
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5989848164737521,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 19873
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7282904354327644,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 19874
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.396522552310919,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 19875
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6240123574671647,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 19876
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1234077205610664,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 19877
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8477670623493192,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 19878
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38013189235411016,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 19879
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4121832216458689,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 19880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8777178457892405,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 19881
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6752767448117798,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 19882
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.40069172400254965,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 19883
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.756490188529166,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 19884
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.449915140233867,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 19885
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7011741173923223,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 19886
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0986063419907506,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 19887
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.072509291090806,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 19888
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.197583201876257,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 19889
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47806315925942683,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 19890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.487082251223,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 19891
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.338491574944538,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 19892
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5958814510398774,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 19893
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9407003454910703,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 19894
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3122185450726773,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 19895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2725441347375046,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 19896
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9614561898540659,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 19897
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.887613727106374,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 19898
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.691720840764271,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 19899
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.978794652657789,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 19900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0176329497287002,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 19901
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4879964226636545,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 19902
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.15734951471831804,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 19903
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6534345196452125,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 19904
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1952571351121124,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 19905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.01989686921409,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 19906
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38056805546427436,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 19907
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1454835126532006,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 19908
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5753705754365135,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 19909
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8375747191846737,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 19910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.22586273440537,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 19911
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2408888163992018,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 19912
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7753476532143705,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 19913
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.22494597713632078,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19914
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3994237996525709,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 19915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4011896302213747,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 19916
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18579640925795515,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 19917
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2660853800014423,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 19918
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7612910391372774,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 19919
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3294803804951509,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 19920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3646698206294476,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 19921
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3034402866990993,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 19922
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7810017000441175,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 19923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.90905573996232,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 19924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8659370097560746,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 19925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8290526394737563,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 19926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9018391090639442,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 19927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.88087629744383,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 19928
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.324785876729302,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 19929
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.04482084814293169,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 19930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3764428891057956,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 19931
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5604417283600698,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 19932
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7372807985187261,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 19933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.145531175249955,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 19934
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5194971836359747,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 19935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9396176897916303,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 19936
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29789510940149705,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 19937
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5590559374454478,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 19938
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5424771455280362,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 19939
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3120386054235413,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 19940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.04959059494970056,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 19941
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.31385973510250553,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 19942
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1367799621880478,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 19943
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0913864210398683,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 19944
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4829554948102834,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 19945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7147425315372833,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 19946
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5759119814581088,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 19947
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.14867978024375852,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 19948
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.104198745130504,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 19949
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8548295696964288,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 19950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0611025090615107,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 19951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4952759196735747,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 19952
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.680917263059121,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 19953
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06788654631751868,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 19954
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.886952387715991,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 19955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1432250710010663,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 19956
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3522364088416765,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 19957
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6870662664842702,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 19958
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3037632716187656,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 19959
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.49363261710728534,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 19960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06924423681775457,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 19961
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4481453884546324,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 19962
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9170349891106687,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 19963
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.229931126886649,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 19964
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.294009570559525,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 19965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.526690487294049,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 19966
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4120578782860278,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 19967
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.533231736168015,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 19968
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.163504691079632,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 19969
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37304879232982413,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 19970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6481427114415281,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 19971
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2443064012368853,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19972
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2924925050942433,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 19973
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1606037548382433,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 19974
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.212552372249486,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 19975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.023786528963835,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 19976
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.599669905028772,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 19977
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.53850921135179,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 19978
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.23621342768238918,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 19979
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1844746047035462,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 19980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6689164553714981,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 19981
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4081458980873724,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 19982
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6023981625203056,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 19983
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6883935323658533,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 19984
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.386242852016282,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 19985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2791462284858492,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 19986
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.777163920855868,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 19987
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8777688699148236,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 19988
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5888653402473488,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 19989
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.385890486078533,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 19990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.32150304329901563,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 19991
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0305720325968135,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 19992
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.941793686534242,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 19993
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4638165898082986,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 19994
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1898519915209536,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 19995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1136493468048996,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 19996
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0708210841796824,
+      "learning_rate": 2e-05,
+      "loss": 0.0941,
+      "step": 19997
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.219995068725058,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 19998
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5186840529703938,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 19999
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1151479497180798,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 20000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.380288987054099,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 20001
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5841923023352305,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 20002
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30465254216290955,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 20003
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2909946034175164,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 20004
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.452919403148624,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 20005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7488167031950208,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 20006
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1741745321814898,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 20007
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8750537106514152,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 20008
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5376732935001414,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 20009
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2327139579348496,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 20010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7847184436185742,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 20011
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28414404209653,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 20012
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8969346123000901,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 20013
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7429874106262744,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 20014
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5129388818744383,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 20015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3895110908881043,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 20016
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5950375199709572,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 20017
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4079278629287379,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 20018
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.477765862214053,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 20019
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3493126617051776,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 20020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5741814841290285,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 20021
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.89976055866815,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 20022
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8270077636428138,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 20023
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0230120291352254,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 20024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3812865588455803,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 20025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8849069809160166,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 20026
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3537017121865444,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 20027
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3281203791883355,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 20028
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9646519591992682,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 20029
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3710365351352234,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 20030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6391384508659931,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 20031
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5805793033388316,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 20032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3214909807137216,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 20033
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8284997800431029,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 20034
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8320911368009077,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 20035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2068489702480716,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 20036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2623128950616787,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 20037
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9395273222937134,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 20038
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7473685939697994,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 20039
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0509112216998653,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 20040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1748882463757406,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 20041
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3328654524049726,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 20042
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6955397802353337,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 20043
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0469367734988406,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 20044
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.077060948378462,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 20045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3456990500777382,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 20046
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9660252429870535,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 20047
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25166625317851155,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 20048
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5484768160560032,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 20049
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.018029111183637,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 20050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.163216334096349,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 20051
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1322784133402416,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 20052
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6473770202292891,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 20053
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1928986254593739,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 20054
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.850331177427499,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 20055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3145639902778476,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 20056
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3611667269202299,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 20057
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7587174184420642,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 20058
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7566619082299845,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 20059
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16031600381320096,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 20060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9079169192469838,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 20061
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6119056873984244,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 20062
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1037745349269692,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 20063
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.649759232422073,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 20064
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2244904078673075,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 20065
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7172740210908152,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 20066
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1706484670165707,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 20067
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.004168457960992,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 20068
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6497435254835455,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 20069
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.684519431410031,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 20070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.160901328097085,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 20071
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.836974285462826,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 20072
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0990926229618603,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 20073
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.035442212030812,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 20074
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.267404639256726,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 20075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7970508306656069,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 20076
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0312974514303432,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 20077
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8468351474407936,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 20078
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4274330471268166,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 20079
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.112828439031665,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 20080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9489662556713614,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 20081
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2517095796075115,
+      "learning_rate": 2e-05,
+      "loss": 0.0918,
+      "step": 20082
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37477850333029317,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 20083
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4311576746851413,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 20084
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.316585650966387,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 20085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4254159119637566,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 20086
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2726566752602257,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 20087
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.963407768362401,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 20088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16738462346987948,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 20089
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5471162808712493,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 20090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9078070648608442,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 20091
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4126375619223625,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 20092
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.20131695710109473,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 20093
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.44572754174371,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 20094
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.152549000003998,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 20095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9632961200887437,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 20096
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.984263199187369,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 20097
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.091820240335483,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 20098
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1166377282070035,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 20099
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2497093816521814,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 20100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5762832275376554,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 20101
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4488928086442978,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 20102
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.21063834629166142,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 20103
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3932501200113427,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 20104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.08566592665860522,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 20105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8148444171145699,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 20106
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.671121859228239,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 20107
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2735950863929113,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 20108
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0713967992666706,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 20109
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.19952989946622,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 20110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6128015525186284,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 20111
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.24353017158242,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 20112
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.50223631972111,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 20113
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.22528640876416417,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 20114
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06172950234489782,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 20115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3126580673042949,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 20116
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.116849534352814,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 20117
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.515414744471975,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 20118
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6219995000230062,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 20119
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.353962633411646,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 20120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6389460374096059,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 20121
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6273075896182393,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 20122
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2698902834831201,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 20123
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16818220968832084,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 20124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47562246946045134,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 20125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3051241754960925,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 20126
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8100817713258558,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 20127
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0805914488526625,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 20128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18900060192143497,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 20129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5375775048466196,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 20130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7270597991486574,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 20131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1223897852664624,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 20132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7640706847830963,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 20133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1967188222795209,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 20134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9935083383345615,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 20135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6356134073606796,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 20136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0051617916315547,
+      "learning_rate": 2e-05,
+      "loss": 0.0963,
+      "step": 20137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4137224564503336,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 20138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7620176853570632,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 20139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7508164174301516,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 20140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.034655237412086,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 20141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2327554435725834,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 20142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4396068564262805,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 20143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2567914526594717,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 20144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.193346988095492,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 20145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10624761333309453,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 20146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.22484258900621662,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 20147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3941412105550253,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 20148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1672083866138866,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 20149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5023941973443327,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 20150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.15042453854133295,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 20151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6122482979588432,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 20152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0177713456001518,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 20153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11211234497495973,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 20154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.324233153498363,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 20155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7654856398956449,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 20156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2647143726024754,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 20157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0026946121968914,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 20158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17429646610364224,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 20159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6287911588192263,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 20160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.361848006731559,
+      "learning_rate": 2e-05,
+      "loss": 0.0792,
+      "step": 20161
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7115313805705632,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 20162
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0703602801051078,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 20163
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1586387291149483,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 20164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.15956945617520718,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 20165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5015806010655061,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 20166
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.524902296536574,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 20167
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.085853573417027,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 20168
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4430562321736073,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 20169
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.910478068419116,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 20170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7858554895890675,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 20171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2646181787825526,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 20172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8424433374452419,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 20173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3540690998239653,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 20174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.291901715744749,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 20175
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.143316317362697,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 20176
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7456248542548882,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 20177
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.706077330545153,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 20178
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.593428675388574,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 20179
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.458944815756058,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 20180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2527474250744157,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 20181
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6347105501197526,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 20182
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9143487205040416,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 20183
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.754817598433516,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 20184
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0272514969050552,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 20185
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.556239747775828,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 20186
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9576867175005369,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 20187
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40717408644861286,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 20188
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5917896855665863,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 20189
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9314940715177282,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 20190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5224560828679015,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 20191
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9892774683000434,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 20192
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4026743039701657,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 20193
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5528418500493611,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 20194
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1989357580229851,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 20195
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.20742064868973983,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 20196
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08178354796441513,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 20197
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5940742611109646,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 20198
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4283249666489063,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 20199
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.022331730994692,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 20200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3432890445690732,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 20201
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6172511876046323,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 20202
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0211794454606093,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 20203
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0015560203497376,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 20204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5538191916592405,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 20205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.249183188631315,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 20206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27911511088669805,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 20207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.352028835353424,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 20208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.808558684105932,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 20209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.045980099465027,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 20210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.44946897420749615,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 20211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5919389951673613,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 20212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0221876607230755,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 20213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5791192189557965,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 20214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3990109935132745,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 20215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.208387384630789,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 20216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3786749713365967,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 20217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1666923872065256,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 20218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9170261494580323,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 20219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2439283898060076,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 20220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7805484260116142,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 20221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5994747395908492,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 20222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4771665114785357,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 20223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1425128102992972,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 20224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.644680714101272,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 20225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.253403420778473,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 20226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.410968815897143,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 20227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.783594428506179,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 20228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.828141823715657,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 20229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9047619691591158,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 20230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8440991491660765,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 20231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8017174435301748,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 20232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7460975846596483,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 20233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6251664655251697,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 20234
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.998150414284988,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 20235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2692048302601564,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 20236
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5590111777809437,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 20237
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.078484061825038,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 20238
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3747045893212844,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 20239
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4959272420599908,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 20240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.34454166759908156,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 20241
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7816820857134346,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 20242
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1945660249811692,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 20243
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6292773508349959,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 20244
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.23754948677685217,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 20245
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6051585693699097,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 20246
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4660583343505222,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 20247
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2941339590829506,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 20248
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08551181135590383,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 20249
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.576057212468855,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 20250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.37186562422144875,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 20251
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.985611424789566,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 20252
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3830561349159146,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 20253
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.079583939425024,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 20254
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.22549634037674735,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 20255
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6436259843001022,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 20256
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.451854692774178,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 20257
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7817552084313426,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 20258
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3265483048763667,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 20259
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2370834574319776,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 20260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9254109411849623,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 20261
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1966248576951285,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 20262
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.709851760908239,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 20263
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.44023058465715403,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 20264
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3732514943669753,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 20265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09986287471868298,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 20266
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4439794430785194,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 20267
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9713484769898276,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 20268
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6938416037474742,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 20269
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6453642346424274,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 20270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.120945259643227,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 20271
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4183567661691099,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 20272
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9201528983978987,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 20273
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6118435632731205,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 20274
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3681211498863308,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 20275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0711675065993735,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 20276
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5919702609540869,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 20277
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8451355722544562,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 20278
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0418207798097248,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 20279
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0747397248063906,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 20280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0436287100262023,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 20281
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1508073668023288,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 20282
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.552684500337666,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 20283
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4975241255050706,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 20284
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9925767929179745,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 20285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6151875311562669,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 20286
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.41734913327291745,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 20287
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6205739419224048,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 20288
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9506666655128054,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 20289
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8329358175667172,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 20290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3106943416145286,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 20291
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5202242596008496,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 20292
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2711560928633847,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 20293
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7625624068314616,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 20294
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7765264383765907,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 20295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0423636771271756,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 20296
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9368517621872212,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 20297
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5629919867546637,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 20298
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7911466513540137,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 20299
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.887639982977267,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 20300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8164186093803885,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 20301
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.286873545424819,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 20302
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.62442237865171,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 20303
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2527556096491134,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 20304
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.520283961248753,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 20305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25290310105032365,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 20306
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6649260809816924,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 20307
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21550965714058448,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 20308
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4531066310705083,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 20309
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8948907254815968,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 20310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4175050255688046,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 20311
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6651282341729164,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 20312
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0655570060862143,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 20313
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7633953363494277,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 20314
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0465359494613717,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 20315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0336885231493214,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 20316
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9209970075618743,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 20317
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3902342233846374,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 20318
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2506778309733178,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 20319
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.176444186762881,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 20320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45271667630190776,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 20321
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1411886099740642,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 20322
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9102344397493352,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 20323
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8584365923043178,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 20324
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2879481267197728,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 20325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5367241150025899,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 20326
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9955981067172166,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 20327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.232624248366328,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 20328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6658847266292962,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 20329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6387624431402488,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 20330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1557933317952862,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 20331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5281384052086489,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 20332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6551473293822092,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 20333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7965098797903127,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 20334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8967788420536094,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 20335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5356848118258064,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 20336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27760156549138054,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 20337
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0324492561672771,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 20338
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5526452144246836,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 20339
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.46975252710224,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 20340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.567409545564357,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 20341
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.21000576475047,
+      "learning_rate": 2e-05,
+      "loss": 0.0994,
+      "step": 20342
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.297085291779023,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 20343
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.19089799664319432,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 20344
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4800765444897924,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 20345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1966164405572197,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 20346
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.565153843688486,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 20347
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2417353160920013,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 20348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.04869458056223713,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 20349
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0945370158684173,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 20350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3526127433676307,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 20351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.749840991885141,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 20352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.365269781916917,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 20353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5462216698132945,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 20354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.588320365457878,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 20355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.008331047139584,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 20356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.281473047333964,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 20357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.832058578970603,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 20358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.251505041053608,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 20359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2127530669561046,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 20360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5312130203452838,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 20361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.15538433001786708,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 20362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7766406334812055,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 20363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1590271152783702,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 20364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5941036827345033,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 20365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.83631973796917,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 20366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7145731262575372,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 20367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3139195149924287,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 20368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.137888005190382,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 20369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6985094532610231,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 20370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5801689583942453,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 20371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4862598999457177,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 20372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8474656504740208,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 20373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7775594134184016,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 20374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.040687585779339,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 20375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.128083964440946,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 20376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3725532941328813,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 20377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.43764411732404757,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 20378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7938597211690936,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 20379
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0763992097147888,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 20380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6062519320473956,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 20381
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3767360991474546,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 20382
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21614329305940852,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 20383
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29615816356784985,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 20384
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30757948629258736,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 20385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3798352535013427,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 20386
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0840677746974716,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 20387
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9053043166802341,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 20388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9939448963919268,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 20389
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7034927677929643,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 20390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.130027545697113,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 20391
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6394107503188786,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 20392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2684639656112145,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 20393
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0559469544460758,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 20394
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0044764224333427,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 20395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1874393146468298,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 20396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.58978551615054,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 20397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0084637452485101,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 20398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.033567506183964,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 20399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.984644692902506,
+      "learning_rate": 2e-05,
+      "loss": 0.0951,
+      "step": 20400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.967250887177155,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 20401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9730597659309013,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 20402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9417588772538887,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 20403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1030352331709332,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 20404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6059296403534076,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 20405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2342302744561957,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 20406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8914973019394882,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 20407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3835118002418385,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 20408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6533968119715015,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 20409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5312705333489127,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 20410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9326939096898381,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 20411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6548554649511056,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 20412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9719155550495064,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 20413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24926964799131812,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 20414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17037525748732288,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 20415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.870629259470141,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 20416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2531631026337606,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 20417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0339692994140286,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 20418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7917957368004122,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 20419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1797134435846557,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 20420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9340575913433968,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 20421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9249185977377975,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 20422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7668838739324104,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 20423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5883677986758102,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 20424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.110852936049221,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 20425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4170481317835106,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 20426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1746052281744486,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 20427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9836387605798965,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 20428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4997515359084938,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 20429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.903948492251707,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 20430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6314922028631809,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 20431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5389167987772958,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 20432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7411582636555928,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 20433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0082875396376278,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 20434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7431101635349403,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 20435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2392516079650384,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 20436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2181224250504842,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 20437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8589098625252867,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 20438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.898883758152664,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 20439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5176246766082704,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 20440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1066573799176438,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 20441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21112746056364506,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 20442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7606541554513949,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 20443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0836596678414545,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 20444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0486473971096943,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 20445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.516127226180091,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 20446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45696704772202246,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 20447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6137933911104898,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 20448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3297273815132655,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 20449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.030249833579421,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 20450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5786204406037825,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 20451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24387390826482927,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 20452
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0011154152931363,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 20453
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5939045251471067,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 20454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6143017456715825,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 20455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6772953777354325,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 20456
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.43011482837471054,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 20457
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.19517400599717993,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 20458
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.835347582433217,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 20459
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6705825928418926,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 20460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1573128841090699,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 20461
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.815768616030597,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 20462
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9774305525785214,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 20463
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6972997747868676,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 20464
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0938593128619598,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 20465
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6590043350389728,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 20466
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9539333654304485,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 20467
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.19544964743504573,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 20468
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.034610077044102586,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 20469
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.806540469491019,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 20470
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4517651830121996,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 20471
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3314071938563183,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 20472
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.44490914732668946,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 20473
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8229908406754928,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 20474
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7653502925141303,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 20475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.613087371796125,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 20476
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3360758430041326,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 20477
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4453029322951898,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 20478
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6477258752962944,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 20479
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0670640327563965,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 20480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.348314701132234,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 20481
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5740740846846977,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 20482
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7643111887112732,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 20483
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7928854987880564,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 20484
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1701154264234098,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 20485
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.783918151024998,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 20486
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3766415074307032,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 20487
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1909742228667592,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 20488
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.49548090229030645,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 20489
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8696461782394826,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 20490
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.369229083938023,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 20491
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.977956950795266,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 20492
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.974378851150421,
+      "learning_rate": 2e-05,
+      "loss": 0.0864,
+      "step": 20493
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.522196272531226,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 20494
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.71149620096296,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 20495
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8635895122591818,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 20496
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1747026533939071,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 20497
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6254093022032562,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 20498
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2390895952814308,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 20499
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.927772055412838,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 20500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.001147686454815,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 20501
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8058578857574163,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 20502
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25580891992583504,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 20503
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.103430049363348,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 20504
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17906904559795897,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 20505
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2915469062821923,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 20506
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.71363902488232,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 20507
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.471502804020688,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 20508
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3401098404176262,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 20509
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8469897755751388,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 20510
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2751267298297845,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 20511
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.741883331233483,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 20512
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7383166511565553,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 20513
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.250612326997011,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 20514
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9332278823602418,
+      "learning_rate": 2e-05,
+      "loss": 0.1232,
+      "step": 20515
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4921392841034412,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 20516
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1481989650514504,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 20517
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.208913346083203,
+      "learning_rate": 2e-05,
+      "loss": 0.1271,
+      "step": 20518
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1326763403051947,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 20519
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24643264626703862,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 20520
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3219479245390782,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 20521
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8567471641046114,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 20522
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4859362482625166,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 20523
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5017673094736044,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 20524
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4886422107536696,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 20525
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.466974583720378,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 20526
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.540940741011662,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 20527
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8416040469628927,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 20528
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1757315184762689,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 20529
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7370124288181247,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 20530
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.507218159761792,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 20531
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6508713970051635,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 20532
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1875549604848012,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 20533
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.987676563654517,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 20534
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11005610181395666,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 20535
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4810077311326721,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 20536
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7466497821504934,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 20537
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7736876497429545,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 20538
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3155932109434207,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 20539
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1683886347700243,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 20540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.760857628423635,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 20541
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7819064754300562,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 20542
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5428310670869327,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 20543
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.516986825009432,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 20544
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7230968884647629,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 20545
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6437930083359493,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 20546
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9759347347497623,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 20547
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8989226979913004,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 20548
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.769782176973685,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 20549
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.22532883562399592,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 20550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8864415949770122,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 20551
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1601516153083651,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 20552
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.419628642706122,
+      "learning_rate": 2e-05,
+      "loss": 0.1125,
+      "step": 20553
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4133334489246689,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 20554
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.990892120352019,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 20555
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5540027739943167,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 20556
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.309178980466862,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 20557
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4769385126496313,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 20558
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.38995998996519005,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 20559
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8074172871343666,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 20560
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4240169192550693,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 20561
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5889301930593349,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 20562
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4350152984209055,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 20563
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4344476552609349,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 20564
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0174760822231244,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 20565
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4528718953536377,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 20566
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21956373327361156,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 20567
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7246698729562001,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 20568
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.019454667332320683,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 20569
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7202580858264251,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 20570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.14649681674254109,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 20571
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1210912511711317,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 20572
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.282598855445407,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 20573
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9797084337641473,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 20574
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27359673087229797,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 20575
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2290669986704638,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 20576
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.651324740269808,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 20577
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2089886642158112,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 20578
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2809756845970217,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 20579
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4102532942499624,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 20580
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.622745764028445,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 20581
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0030114605782277,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 20582
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5721118446972946,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 20583
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5098035282582907,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 20584
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4773929420322225,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 20585
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4198966043734333,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 20586
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.119629862340691,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 20587
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7193542805362398,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 20588
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2534626684179924,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 20589
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.554819521169529,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 20590
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1659875509422736,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 20591
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.792230599075555,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 20592
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2643747639035012,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 20593
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17452356440254638,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 20594
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5295441136553323,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 20595
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2945019593586532,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 20596
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.32215676058363263,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 20597
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8274939988607877,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 20598
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8156223823242275,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 20599
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9476090585353742,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 20600
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0160729814096625,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 20601
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2994603962839018,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 20602
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.110537286482823,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 20603
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7655818304230738,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 20604
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.180073037455148,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 20605
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6519904153421114,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 20606
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7756887482567077,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 20607
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2092241040308558,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 20608
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5747953620903385,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 20609
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3714393549662633,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 20610
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4212992322147686,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 20611
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5583983946629427,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 20612
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.61619957865155,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 20613
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.03300232510126518,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 20614
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6655977715705454,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 20615
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.12814305256694,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 20616
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6520975756283219,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 20617
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4569286484170627,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 20618
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8127738482677256,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 20619
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8816926994037921,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 20620
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4057419176949708,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 20621
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8985436178305022,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 20622
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.777475943208085,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 20623
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8785913539082449,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 20624
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8353260335171986,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 20625
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.947377935614129,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 20626
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.362498256137749,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 20627
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.41104594420643303,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 20628
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16105300198343098,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 20629
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.472015905396555,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 20630
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5290906043542556,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 20631
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3759377284704017,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 20632
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5086771013020122,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 20633
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28931826797491017,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 20634
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8308112468379862,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 20635
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3677830026834243,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 20636
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7147560274257609,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 20637
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3036110371807461,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 20638
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08085047320418072,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 20639
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9807256536814202,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 20640
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6999860061742726,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 20641
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.522201566102771,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 20642
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9917523675566706,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 20643
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.14441157876712024,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 20644
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7048602405481947,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 20645
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2865152333555256,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 20646
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.37951764204459937,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 20647
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1508479350878833,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 20648
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1987380763665163,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 20649
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6521800172647962,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 20650
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2654755115528573,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 20651
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.918303566226598,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 20652
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4324280367579072,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 20653
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.935005365754097,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 20654
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6132101782013866,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 20655
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3803886247836616,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 20656
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6022015187206013,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 20657
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7783733048341694,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 20658
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08828941841229278,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 20659
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.230926296761878,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 20660
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3129382362612976,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 20661
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27789675959180044,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 20662
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0624262279256103,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 20663
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07626394979199784,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 20664
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2009149137543806,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 20665
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4195509225789959,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 20666
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.729785926585924,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 20667
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1516019340812922,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 20668
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9482176411179111,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 20669
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0503722461980758,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 20670
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27507617935667,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 20671
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45364456972222855,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 20672
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07298763864370604,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 20673
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9765065901964111,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 20674
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8919457797947294,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 20675
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4602947061680904,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 20676
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5660056966959978,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 20677
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1060347439239175,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 20678
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7162631591866273,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 20679
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6140301256648562,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 20680
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9361666736933891,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 20681
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2815884399021285,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 20682
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5296804860868789,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 20683
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6245791210221683,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 20684
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2683220844374119,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 20685
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7134320611500415,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 20686
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7984010270146327,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 20687
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8426456995550607,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 20688
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1626668154074051,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 20689
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0242719669905405,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 20690
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7480130897641405,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 20691
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3443004146261541,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 20692
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7728654384122544,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 20693
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.092425140719003,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 20694
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8382755631309935,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 20695
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.863893420344512,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 20696
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5644565832962695,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 20697
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7373480979821956,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 20698
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4907489674664681,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 20699
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3532646794279677,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 20700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6032916891187289,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 20701
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9378394466038423,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 20702
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5373485195765197,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 20703
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4938562148434213,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 20704
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.037764892573192786,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 20705
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17494408680705323,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 20706
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4913895966538706,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 20707
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.541828748413654,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 20708
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5456646894950032,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 20709
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.464039167057433,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 20710
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.660633691686727,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 20711
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.44069126663553,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 20712
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.37720318690801,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 20713
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9230086904304685,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 20714
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3086505934482366,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 20715
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4694704764178976,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 20716
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7601801273559553,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 20717
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7477294009147561,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 20718
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3339620883934402,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 20719
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.785428416667126,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 20720
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.031788945251712,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 20721
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2743259443314494,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 20722
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2933608038571605,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 20723
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0459253568858484,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 20724
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9382463027836533,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 20725
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6878126907175052,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 20726
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2292973831413292,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 20727
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2114901980744115,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 20728
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0412103925226603,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 20729
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7463025815197475,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 20730
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5120621322617479,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 20731
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2094788657049425,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 20732
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4070852977786792,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 20733
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5453411389894385,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 20734
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9249698343922272,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 20735
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.610930268739717,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 20736
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8839310505654715,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 20737
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9459149986391067,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 20738
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2368569486856014,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 20739
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6610097390235059,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 20740
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.56670653177263,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 20741
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.83365687492034,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 20742
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0951309372365522,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 20743
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6016045840516732,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 20744
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.43917516874104,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 20745
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3864020196889038,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 20746
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9592079476908537,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 20747
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4654388465252297,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 20748
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2769598894473966,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 20749
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8318855265489454,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 20750
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.540001892029065,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 20751
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2881763639612881,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 20752
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.1937155072631445,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 20753
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6663696824625318,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 20754
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6349873149067036,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 20755
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9056738633478411,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 20756
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9111482193705434,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 20757
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1548788343180731,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 20758
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.096478682429503,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 20759
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9091759061544913,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 20760
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5825559146398038,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 20761
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0679674283893614,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 20762
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4280914402864608,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 20763
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3538450529227306,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 20764
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2215800073149303,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 20765
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18199424776874937,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 20766
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23129840002811525,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 20767
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2793720384308528,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 20768
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4664973331377231,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 20769
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8822504127763068,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 20770
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10098888646666236,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 20771
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8777252736673542,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 20772
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5440301648637422,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 20773
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.01749834027519657,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 20774
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5911958366949466,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 20775
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.955753508928306,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 20776
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.240373586615432,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 20777
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.65162849051925,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 20778
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4325976744098118,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 20779
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.09006956477487249,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 20780
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4635898767678461,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 20781
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7187400123999585,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 20782
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9377553592041612,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 20783
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.39447309990827845,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 20784
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.39835738330260123,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 20785
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5296709911168653,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 20786
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.22297214385178427,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 20787
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.623847909889601,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 20788
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7100888311531942,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 20789
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1433872064422865,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 20790
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4846079015426072,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 20791
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1194904839067366,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 20792
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.720043882969678,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 20793
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.034943748023027625,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 20794
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0423152308661323,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 20795
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1124045684687722,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 20796
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.856471948499928,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 20797
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.14659984488005778,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 20798
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8842024874164908,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 20799
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8423342660049128,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 20800
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07897298865685388,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 20801
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.444632326833363,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 20802
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7195177419272754,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 20803
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3823864165027095,
+      "learning_rate": 2e-05,
+      "loss": 0.1041,
+      "step": 20804
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1604101215819086,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 20805
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06163821035570322,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 20806
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3132825970571895,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 20807
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.271595845998127,
+      "learning_rate": 2e-05,
+      "loss": 0.0819,
+      "step": 20808
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46686741762757705,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 20809
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.092845664511329,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 20810
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7970623469311072,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 20811
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2841420492829315,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 20812
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6879229545450727,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 20813
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6834409427554509,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 20814
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3201613457250194,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 20815
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1330347189923464,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 20816
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3371070972246327,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 20817
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.35619518962644886,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 20818
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8774880729582021,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 20819
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0531865495303159,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 20820
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2149722408530148,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 20821
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.42227309482400566,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 20822
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7412844656814648,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 20823
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4884762489046343,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 20824
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9528263749876154,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 20825
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.044042590255666,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 20826
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.459223593591383,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 20827
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8505201347116161,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 20828
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.996981608092649,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 20829
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.598397927777164,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 20830
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.419670082719341,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 20831
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7977406904489691,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 20832
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.585877662727024,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 20833
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8986009200368417,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 20834
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2335732185921298,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 20835
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7657928282853473,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 20836
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2641487252628005,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 20837
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7583180696569183,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 20838
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2775999014612059,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 20839
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4946675651333914,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 20840
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.05514359379294649,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 20841
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.371868251705734,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 20842
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.65730748515689,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 20843
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.872185164489134,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 20844
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.129678481871935,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 20845
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7675202326018223,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 20846
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9160654017867103,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 20847
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.04571442036553678,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 20848
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1028397102058216,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 20849
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8934431162734698,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 20850
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.712049976057196,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 20851
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.64633464319277,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 20852
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9060294127169928,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 20853
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0885730338337396,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 20854
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0620935212913496,
+      "learning_rate": 2e-05,
+      "loss": 0.1177,
+      "step": 20855
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.588362621541045,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 20856
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8858563434761915,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 20857
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2636693742421557,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 20858
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.556244343820121,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 20859
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2102563031049227,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 20860
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.855519086669568,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 20861
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.14591649287185507,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 20862
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5982651534646615,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 20863
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.759695511076354,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 20864
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1149209779697413,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 20865
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6609039812178488,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 20866
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4018246000552523,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 20867
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8269436271376551,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 20868
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4207476663726377,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 20869
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6781300206152351,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 20870
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5426056639897847,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 20871
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.044319392698014,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 20872
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9648975249220076,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 20873
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8037400825564956,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 20874
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.316162387971741,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 20875
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8532066222329547,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 20876
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9568576888917334,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 20877
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3289691534169433,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 20878
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7193469595394839,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 20879
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7618987417312053,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 20880
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5453357560409804,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 20881
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3483379196058239,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 20882
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8426628880139427,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 20883
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23541478934144328,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 20884
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2775065253323936,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 20885
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8662661907890312,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 20886
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.451983400662426,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 20887
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7246291165605911,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 20888
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.153169135926366,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 20889
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0799382759045972,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 20890
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.235086866836774,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 20891
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9786453313388048,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 20892
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24614099018876456,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 20893
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5134248767779104,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 20894
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6373798939752106,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 20895
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1092974138664926,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 20896
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.596537452451996,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 20897
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4290717307734966,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 20898
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6137899245194974,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 20899
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0888029537083695,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 20900
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.20761300580755,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 20901
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3657406852159221,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 20902
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.33424377406576783,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 20903
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5519292244457638,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 20904
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.383269757032202,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 20905
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7198357360603225,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 20906
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8520295403145253,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 20907
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8878487264512441,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 20908
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.513835943440523,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 20909
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0963011045828885,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 20910
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0452027223646136,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 20911
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9285632113125242,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 20912
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4623535033780047,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 20913
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0642380803421716,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 20914
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1396571194171767,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 20915
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13929479009081058,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 20916
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9814953236150276,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 20917
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5522581509784552,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 20918
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5432970440433609,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 20919
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.549031963031138,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 20920
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7743722978971057,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 20921
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.276568069352476,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 20922
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.889360470876568,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 20923
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9310447658732786,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 20924
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0179973770999657,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 20925
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4938155752409011,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 20926
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.920559154157627,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 20927
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3106666190129627,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 20928
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2381458444933777,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 20929
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.717589821405807,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 20930
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.04568211834838084,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 20931
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.290042758794856,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 20932
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8958505288772908,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 20933
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7830022996768821,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 20934
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6004426395989744,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 20935
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5655500017122825,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 20936
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4181059463586125,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 20937
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23380592718779458,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 20938
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2458092998351614,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 20939
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6860942636884484,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 20940
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3847560453196808,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 20941
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.15153032556166474,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 20942
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6352302376796893,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 20943
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9451059793657911,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 20944
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5212075573865135,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 20945
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3469972579741905,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 20946
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0176992509033829,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 20947
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8617473801588801,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 20948
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.460631576139177,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 20949
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1940998084143297,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 20950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6187781283611642,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 20951
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1901237769441811,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 20952
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6263127526411014,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 20953
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6267780047498759,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 20954
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2015526144863689,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 20955
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6115444397830209,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 20956
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6087301461352375,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 20957
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.612340190325336,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 20958
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9913102555961384,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 20959
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7264175699432383,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 20960
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7862877796788048,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 20961
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0813700173862566,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 20962
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47162594696402804,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 20963
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0721279257376068,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 20964
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8031647356942481,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 20965
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.440009176472861,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 20966
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7699094511023543,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 20967
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.05955033279930706,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 20968
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8984215776442359,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 20969
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.912302258809164,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 20970
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4234561078533195,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 20971
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.254478110279421,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 20972
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4873664026612208,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 20973
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1251480067489465,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 20974
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.38439538250356886,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 20975
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2321803806598663,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 20976
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0924804403793182,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 20977
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1678903100456492,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 20978
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7401542519559944,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 20979
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5160125374250425,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 20980
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0299532279718786,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 20981
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1261446445121891,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 20982
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5918248991421222,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 20983
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6091446318818137,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 20984
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6084469832486716,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 20985
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.04564167296960693,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 20986
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3911996238927554,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 20987
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7945041944192806,
+      "learning_rate": 2e-05,
+      "loss": 0.0944,
+      "step": 20988
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.140325260756476,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 20989
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9292013814726234,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 20990
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.49610585475781754,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 20991
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.56645008510984,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 20992
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5474727373845014,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 20993
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0871064734685778,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 20994
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8091978048945083,
+      "learning_rate": 2e-05,
+      "loss": 0.0873,
+      "step": 20995
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7707045414782416,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 20996
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.782721731297466,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 20997
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0358784010011826,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 20998
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7456016444451115,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 20999
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5748622137683495,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 21000
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0717853519774847,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 21001
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7249241065718545,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 21002
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4097367275583492,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 21003
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.19128635234894,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 21004
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.448928351495454,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 21005
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.018324202368027,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 21006
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.35635205446740015,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 21007
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5966689487434715,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 21008
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7777748254500472,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 21009
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8838084084920843,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 21010
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3511643457216724,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 21011
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5722708067154884,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 21012
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6449626750274574,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 21013
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.310719456108862,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 21014
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5274061887812762,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 21015
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.412976793251771,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 21016
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5152870666672964,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 21017
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0151981337016862,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 21018
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5103794049457093,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 21019
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24158471988573077,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 21020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2192751413318388,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 21021
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6892876575091039,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 21022
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9145771762116379,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 21023
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1377553200334816,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 21024
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7849563799723651,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 21025
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5751636026576188,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 21026
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8867256876819543,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 21027
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5266329743627938,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 21028
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4573517678056171,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 21029
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9686611700453833,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 21030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0047365310119254,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 21031
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.14833280359554887,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 21032
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5134248790415576,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 21033
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3145749629379408,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 21034
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.307498339749328,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 21035
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5888941222699697,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 21036
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7355829791929236,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 21037
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0670632649819791,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 21038
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.399250501585933,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 21039
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.502130902190103,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 21040
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0478512212036373,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 21041
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9668611139090648,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 21042
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07969862285290841,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 21043
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4112540856770065,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 21044
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8446042363673237,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 21045
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1855240997188117,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 21046
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2690966086535576,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 21047
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4287268563966835,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 21048
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.697519808611234,
+      "learning_rate": 2e-05,
+      "loss": 0.1115,
+      "step": 21049
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9407991850010916,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 21050
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8791711188331977,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 21051
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9083800440918958,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 21052
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8906166009339604,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 21053
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1488422914116745,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 21054
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.16748003500834985,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 21055
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.021399645819751,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 21056
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8744819333591435,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 21057
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17949088369068233,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 21058
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.445428340287907,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 21059
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.284816987033674,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 21060
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.607400531945798,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 21061
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4070120971663929,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 21062
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.226506925190475,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 21063
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5699292292914594,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 21064
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1261196392963637,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 21065
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8613534521414489,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 21066
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6452705766854719,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 21067
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3837000267012816,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 21068
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4710840923274645,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 21069
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.047410744550175486,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 21070
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5363952050209502,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 21071
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9175213636670767,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 21072
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2196456505891892,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 21073
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0040888712397225,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 21074
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8257007664385254,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 21075
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.498585968663867,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 21076
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2086058691697508,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 21077
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0465785503552032,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 21078
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8849424812785045,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 21079
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46967369305487755,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 21080
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3848346644571257,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 21081
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.25514936576848907,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 21082
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3824381779649608,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 21083
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07096869518512236,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 21084
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5617210239581901,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 21085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23344853394256976,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 21086
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.846622283678563,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 21087
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0475094809213683,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 21088
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.87160921947234,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 21089
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1376429729946964,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 21090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3787126702436785,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 21091
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.04545024158800579,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 21092
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6875315062795953,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 21093
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.924668842161476,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 21094
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.760274290689009,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 21095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3877309624148335,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 21096
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8901094316327205,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 21097
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3894666089753029,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 21098
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7814116698864213,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 21099
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.958317134330252,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 21100
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.004795319035585,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 21101
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4614451893756482,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 21102
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1521950835973453,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 21103
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.518611391101451,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 21104
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6117339082401252,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 21105
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2702658068719446,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 21106
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7789094578591792,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 21107
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9321137512477744,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 21108
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12812113814232653,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 21109
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6466803584992554,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 21110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0692221106298454,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 21111
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9237795407704069,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 21112
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5986698257429137,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 21113
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.029195329397537082,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 21114
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.173489526253495,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 21115
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.205741415380408,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 21116
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6817164622208547,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 21117
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0633769063083867,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 21118
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.05378895208974905,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 21119
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6447273418487673,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 21120
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.040971468945203,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 21121
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6825877146649753,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 21122
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.48001837362449484,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 21123
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.98268506036618,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 21124
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8849429032357619,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 21125
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6232728460948858,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 21126
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6076583892478571,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 21127
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.562412679040409,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 21128
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2344073037857923,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 21129
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5062322183106636,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 21130
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4242374789313832,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 21131
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.271894493700379,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 21132
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41000079887591434,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 21133
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2142690298793481,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 21134
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6187286405001047,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 21135
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47806746068059647,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 21136
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6773391362685397,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 21137
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46168664088000794,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 21138
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7075870078945414,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 21139
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6446507660999923,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 21140
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.45878460371596114,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 21141
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7814322068408059,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 21142
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.412238699450654,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 21143
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.49542906680147564,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 21144
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.230966587291766,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 21145
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9984933112220816,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 21146
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6919784548842558,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 21147
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1966186567644668,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 21148
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.958626276332977,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 21149
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1878685776958915,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 21150
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8997660568167163,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 21151
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4215371580544843,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 21152
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2793898686978529,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 21153
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6249265654183636,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 21154
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.350571638515684,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 21155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0638593504988043,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 21156
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.098620279876434,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 21157
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7689606976046357,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 21158
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.275598763135787,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 21159
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1276304962866608,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 21160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1084673023506406,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 21161
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8355902160736244,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 21162
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.32147627596659417,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 21163
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.150756583053303,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 21164
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4521456178734784,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 21165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.060081140813446,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 21166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9532640074374314,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 21167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2154974941665635,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 21168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9376274975864317,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 21169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.380814784947382,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 21170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23920527290039226,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 21171
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3486465132982108,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 21172
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7277559324330203,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 21173
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0136470607888257,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 21174
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1883474638759965,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 21175
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17590388682457514,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 21176
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2602305732845601,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 21177
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4510986136533338,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 21178
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.35136970955927965,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 21179
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0182052700558981,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 21180
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7781036343379708,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 21181
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.389688086410204,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 21182
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9979706840120934,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 21183
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9668381809274806,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 21184
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0865476252777717,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 21185
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6740305941922219,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 21186
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07612470547947182,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 21187
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7141772389151582,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 21188
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.850925899471195,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 21189
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4050409841838523,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 21190
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.481818237881499,
+      "learning_rate": 2e-05,
+      "loss": 0.1216,
+      "step": 21191
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.194245863297879,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 21192
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.594928604847997,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 21193
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.540675136531876,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 21194
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.935026222904692,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 21195
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2038999761557811,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 21196
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3248507418299785,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 21197
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4385406685166164,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 21198
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4612682416715103,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 21199
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0385994448098184,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 21200
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.968809648953921,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 21201
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.41977080355626495,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 21202
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9687183922546436,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 21203
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5590499936762716,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 21204
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6132508531557502,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 21205
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3297415824318193,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 21206
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1844415429578563,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 21207
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6776674440957692,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 21208
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4260824345651885,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 21209
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.22177214521919233,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 21210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.229658142753684,
+      "learning_rate": 2e-05,
+      "loss": 0.1061,
+      "step": 21211
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.524355126732944,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 21212
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2281282107898034,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 21213
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.657548923418907,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 21214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6060297428250646,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 21215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5080699854742915,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 21216
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1281591669123496,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 21217
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8873687915791701,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 21218
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1978705639013581,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 21219
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3094433659661615,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 21220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9215418003327844,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 21221
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.36211824406022,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 21222
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0672233134401843,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 21223
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.299461864083583,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 21224
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.46468892444220394,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 21225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9908865476862454,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 21226
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9768783973932527,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 21227
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8779249265613402,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 21228
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3331639591687012,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 21229
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.457741408336994,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 21230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2351006415128798,
+      "learning_rate": 2e-05,
+      "loss": 0.0872,
+      "step": 21231
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.501231939049464,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 21232
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6775180790308368,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 21233
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8765001018228358,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 21234
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0165617193022203,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 21235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6512264144333484,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 21236
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5786308704390757,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 21237
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6495443668694922,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 21238
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.20984181418133124,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 21239
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30996340995239957,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 21240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.48492398991222685,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 21241
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0659073441622922,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 21242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8085202538448746,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 21243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6573010157319762,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 21244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0969914889608141,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 21245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.576280357363872,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 21246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3146299137698701,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 21247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6156316257637,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 21248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.493578996007802,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 21249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9786658866469509,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 21250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5483071783701764,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 21251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.618864774173769,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 21252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2981450495385851,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 21253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.278645669931144,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 21254
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6837414457674515,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 21255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12512916358752427,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 21256
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7118907448465066,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 21257
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6635891293049776,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 21258
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6530307216477937,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 21259
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6198527816218324,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 21260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30233600616395157,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 21261
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.32786966001063,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 21262
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.691051952729041,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 21263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6409246395239969,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 21264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.23265066826749198,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 21265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3021873577279325,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 21266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.93974287041844,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 21267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07204618192156118,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 21268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5168520492747597,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 21269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.387617223100834,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 21270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9716657321253667,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 21271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6235549433711908,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 21272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3896369263136551,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 21273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5114687533011757,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 21274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2163917605510563,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 21275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.291008059512117,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 21276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7831108438440789,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 21277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5677062276632007,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 21278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2610596998830324,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 21279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9817584852391553,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 21280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.40236100835728994,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 21281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.790904096111283,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 21282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7309761078594426,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 21283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9273859422238149,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 21284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.806770476752391,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 21285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.549743892906376,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 21286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6049047735241245,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 21287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0772318803383671,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 21288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1682860794977739,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 21289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.135012878450597,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 21290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8856478667018575,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 21291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8804964882073123,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 21292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5981310037134221,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 21293
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2244983481615081,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 21294
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3133609309580672,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 21295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5805356338234726,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 21296
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2536200964839046,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 21297
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29462718767139695,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 21298
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6672762577240119,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 21299
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9034874343927146,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 21300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8598191828668706,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 21301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.35834201295823703,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 21302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28059818978731743,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 21303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.642968937633935,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 21304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.705112379558267,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 21305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9620931470648273,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 21306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9271861172965667,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 21307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.39776427295043093,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 21308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9504312478557928,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 21309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8993094996524318,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 21310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.338903558794928,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 21311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.57075425951567,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 21312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6922454913024545,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 21313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.482362039656777,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 21314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2697317581293437,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 21315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38177922991183605,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 21316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9722177754681862,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 21317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.23015565942039778,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 21318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9654679229283344,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 21319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4486245745382898,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 21320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.577222403032256,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 21321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.59785365914666,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 21322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33023807780759873,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 21323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7642377625079302,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 21324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8098338104409415,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 21325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6668577402577955,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 21326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6902315848997294,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 21327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.052071442237859,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 21328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3649970245328897,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 21329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5285382295219532,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 21330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0780702659976968,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 21331
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0233278288574126,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 21332
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.853802697208538,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 21333
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3284585189024236,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 21334
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2036160720652567,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 21335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1531226677599635,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 21336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4738924620594192,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 21337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.028909714218079,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 21338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7551692501738918,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 21339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3571796448102784,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 21340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1207843951349991,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 21341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1527014018084114,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 21342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6890498214655185,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 21343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26892990256537147,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 21344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2730164283073002,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 21345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.701884085986194,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 21346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9807195152797787,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 21347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8711965525910973,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 21348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3000236802511973,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 21349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6771662856908439,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 21350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18575997412845685,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 21351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28604579031866717,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 21352
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.507887922387043,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 21353
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4183619067260617,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 21354
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1071612877344477,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 21355
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.341782437931225,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 21356
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8819208621746196,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 21357
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.640750408016203,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 21358
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7793747330017367,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 21359
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2510516569971206,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 21360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1933133729849483,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 21361
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8773498433459102,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 21362
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0739884563159545,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 21363
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5624778796191067,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 21364
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0726878406737712,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 21365
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.586701617332505,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 21366
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2061346977350853,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 21367
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8416403427980468,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 21368
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0704331156020057,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 21369
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0738287200315513,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 21370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30534586324908214,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 21371
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0791164759645364,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 21372
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.484047286630029,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 21373
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.155990262112637,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 21374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2011899634280625,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 21375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8032162436632864,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 21376
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.20364350385057148,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 21377
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6615127981595716,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 21378
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29430287814011336,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 21379
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12397162789556734,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 21380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5166251522286416,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 21381
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2111758089915168,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 21382
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8106925373869343,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 21383
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3133839400863874,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 21384
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.378770729732221,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 21385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6491332630194978,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 21386
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7108198151547542,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 21387
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6544768754501904,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 21388
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2497190371876787,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 21389
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.112916960426405,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 21390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.401877748212021,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 21391
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.775607984234781,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 21392
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8439957649319053,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 21393
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9462270758622706,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 21394
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8955881760559357,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 21395
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8538786162069613,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 21396
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3893032035049067,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 21397
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2686695200158834,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 21398
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.40354469837132406,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 21399
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2161020968129155,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 21400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5289881602914797,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 21401
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.262059665841873,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 21402
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.239841948961878,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 21403
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.387045580374033,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 21404
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.453978981985957,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 21405
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0471830341877963,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 21406
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.21826985149891528,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 21407
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1531226347721145,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 21408
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2239883877674509,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 21409
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.932316090171537,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 21410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.579050811071796,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 21411
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.22754623365105,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 21412
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9713463292900857,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 21413
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7622828033535278,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 21414
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16463780578099468,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 21415
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0331461265143906,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 21416
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17697243279894614,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 21417
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.951653376250889,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 21418
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7191298766319765,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 21419
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6785640639966706,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 21420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5774243465015079,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 21421
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9870314101967403,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 21422
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.284910635504954,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 21423
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9986158566869868,
+      "learning_rate": 2e-05,
+      "loss": 0.0821,
+      "step": 21424
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9188022395176767,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 21425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.47901015904629696,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 21426
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7519874465838118,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 21427
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3195781430855402,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 21428
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.999963879259089,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 21429
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16054356077351395,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 21430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5139360751032065,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 21431
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3784324977648217,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 21432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1649690720541783,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 21433
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4032817030913274,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 21434
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.730631732369633,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 21435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2652104310810848,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 21436
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3330895332007241,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 21437
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5618104947311426,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 21438
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.170086271170163,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 21439
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2899642053663483,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 21440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.226923755986726,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 21441
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0826579458405365,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 21442
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0688507791040491,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 21443
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0049667397494428,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 21444
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9344929155391073,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 21445
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2940058855957657,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 21446
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1394571047279554,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 21447
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3050950835655277,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 21448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7533633001522548,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 21449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8021816754844835,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 21450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7488975369356259,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 21451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.034365196988308,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 21452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.830713261256423,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 21453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6704657657283953,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 21454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.670433771137471,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 21455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8783303981046715,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 21456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.08146691324006,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 21457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4167308479787148,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 21458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.431333027576664,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 21459
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.579335636743761,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 21460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3719973724634795,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 21461
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8209503758899794,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 21462
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3231454117940187,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 21463
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1340604868984465,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 21464
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.25852564234619196,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 21465
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8495034688953976,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 21466
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4595865560466466,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 21467
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.20577234205460354,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 21468
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.364617218985984,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 21469
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6931034314211122,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 21470
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6479878992581508,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 21471
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6073678120525923,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 21472
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1588103839732942,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 21473
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2124430987190824,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 21474
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.038783329912112,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 21475
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1130993777157143,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 21476
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12758441369932097,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 21477
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9159611927769818,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 21478
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7864954520722743,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 21479
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7769216416981958,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 21480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4210202508996121,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 21481
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.09374476190168,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 21482
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6721462849638729,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 21483
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6592089295125831,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 21484
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9617961675749616,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 21485
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7395669953238565,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 21486
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4898214082626999,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 21487
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09574362812491426,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 21488
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3878059470789834,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 21489
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7042167034435916,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 21490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.43050414422773026,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 21491
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.067294355323667,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 21492
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2031335492340216,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 21493
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8305912720525478,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 21494
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11019823511974736,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 21495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1456860129700661,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 21496
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8749041504814125,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 21497
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0535693422624746,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 21498
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4259586589647038,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 21499
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.77936589979841,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 21500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.956758203286398,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 21501
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1798413915235477,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 21502
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1191051004839805,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 21503
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4275578547577048,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 21504
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9186412539319134,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 21505
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3407273117897083,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 21506
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9821234718071965,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 21507
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30121278151040576,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 21508
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.156120550796472,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 21509
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6191470303656097,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 21510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4147322949428878,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 21511
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.488741425586418,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 21512
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1168793473375243,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 21513
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.136591691194767,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 21514
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2817283509618367,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 21515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.12134206327978,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 21516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6740782281507354,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 21517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.87259172356401,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 21518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7575897975605264,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 21519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4585008062378462,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 21520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1695425558705685,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 21521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5735442163604956,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 21522
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9078174715677457,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 21523
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.210133328795955,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 21524
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34746357582810344,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 21525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.817459565543559,
+      "learning_rate": 2e-05,
+      "loss": 0.1148,
+      "step": 21526
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0843943327573788,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 21527
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0589776275656202,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 21528
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8706460666230055,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 21529
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5236064328160768,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 21530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0029246953228967,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 21531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2746776322873412,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 21532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3688406962050381,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 21533
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13886284774458932,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 21534
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8900337933971817,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 21535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1943322444431352,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 21536
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.582089943278178,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 21537
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6628321975378227,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 21538
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.407381280868939,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 21539
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2321722439801768,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 21540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8588586939910399,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 21541
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33654734238235645,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 21542
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.488820289620961,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 21543
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08699256855350623,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 21544
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6237322586150952,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 21545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.42260766869531713,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 21546
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5855354710054356,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 21547
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1582942109853915,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 21548
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9403657980634347,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 21549
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6855279208005635,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 21550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5189009837383297,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 21551
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0336084612249414,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 21552
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4373519069971166,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 21553
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7611139488476697,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 21554
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6371581544802265,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 21555
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4491627533883148,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 21556
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9315020103592253,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 21557
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1622398219950174,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 21558
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.047012385683781,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 21559
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18886826878803248,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 21560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.44155973955980243,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 21561
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6758796532049022,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 21562
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4714839621179223,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 21563
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10350961937667201,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 21564
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3301194866544024,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 21565
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6373009568010028,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 21566
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6937786929752392,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 21567
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8548543456865458,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 21568
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16938984550716865,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 21569
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6822081119151245,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 21570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.960230218134852,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 21571
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5991100008682975,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 21572
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.503424391904419,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 21573
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.557141073626652,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 21574
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3696253989291369,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 21575
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.428014737525971,
+      "learning_rate": 2e-05,
+      "loss": 0.1253,
+      "step": 21576
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7015373459314989,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 21577
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7720405601831037,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 21578
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4464961849331543,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 21579
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.287519330950467,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 21580
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5042888996761774,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 21581
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5792729862198052,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 21582
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6509537192387485,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 21583
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.53196450515544,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 21584
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6047902643378235,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 21585
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3775350200438629,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 21586
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.24753083594487074,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 21587
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1323239555539528,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 21588
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3688336734366442,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 21589
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.14112855241239075,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 21590
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.354541428378454,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 21591
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1744982735055208,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 21592
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7494611591738427,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 21593
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1616444009069752,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 21594
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.507946976905787,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 21595
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1292861700668577,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 21596
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1701198913410538,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 21597
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30782416510383936,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 21598
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3005214287318387,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 21599
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2353589445956559,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 21600
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06309469608057637,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 21601
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7311747601918155,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 21602
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7957613603842513,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 21603
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6063900274254046,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 21604
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0100721119169056,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 21605
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.586323674418251,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 21606
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1704023135753614,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 21607
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0026675646005154,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 21608
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8472157894905281,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 21609
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12613600613735143,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 21610
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16886964658084577,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 21611
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1348605540564267,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 21612
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7364553398498594,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 21613
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0272440698804808,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 21614
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3969611134557722,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 21615
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19928226206526825,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 21616
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9192433521334005,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 21617
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7717042090504458,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 21618
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2813705477394068,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 21619
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.968357037244824,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 21620
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9383664577649715,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 21621
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3170099430621287,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 21622
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.330417628938538,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 21623
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4595014378677718,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 21624
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1169916184782478,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 21625
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.017417531755076775,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 21626
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5418797006810157,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 21627
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5554695258155266,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 21628
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.40152839337684443,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 21629
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.32179888825366776,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 21630
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.217758998464043,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 21631
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.20595475797452228,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 21632
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0711145667481596,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 21633
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.117023194190945,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 21634
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0147506828453685,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 21635
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8234511400250333,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 21636
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.969079250348268,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 21637
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.833976298661821,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 21638
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0066406771060692,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 21639
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5581247187888194,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 21640
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0153419687133918,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 21641
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17837680786596044,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 21642
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5981689199029817,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 21643
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5844549004686952,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 21644
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8272083951571823,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 21645
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8233460129617194,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 21646
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.279965396825514,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 21647
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5405923433692237,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 21648
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.849030606678561,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 21649
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1361638448384612,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 21650
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9203266460063371,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 21651
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.819723336811734,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 21652
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0540660581463661,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 21653
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4059841680094202,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 21654
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6107096117760215,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 21655
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9819917599764394,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 21656
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5988216382376439,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 21657
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.20054835382502556,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 21658
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8590873688127598,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 21659
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9079121441412639,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 21660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.939949026028013,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 21661
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0707890951649452,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 21662
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.888282115091076,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 21663
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1931725572467405,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 21664
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.2824487052201565,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 21665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8036215139877891,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 21666
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4726437653713386,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 21667
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4732543219546832,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 21668
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.990330692563628,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 21669
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40896806729949753,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 21670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2079615788412834,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 21671
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3271684680182207,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 21672
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24865742132359978,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 21673
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7238777656591078,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 21674
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24935637264666047,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 21675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.327480056668538,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 21676
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.86203133385834,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 21677
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1970126458840913,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 21678
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5799073690013103,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 21679
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1467888344974586,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 21680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7667396761159085,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 21681
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.44715588485209135,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 21682
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8395464549646863,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 21683
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7934080429303494,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 21684
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.394612020534673,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 21685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7493098182741599,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 21686
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2634759245483729,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 21687
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.810452508717142,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 21688
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4650615072520807,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 21689
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0590541162325033,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 21690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.312182660567464,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 21691
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.35995393229812683,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 21692
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8510778605415485,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 21693
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.818718005421542,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 21694
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.687750339092461,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 21695
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2289109009040047,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 21696
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5118439208249984,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 21697
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8767732295539534,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 21698
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.436574222134824,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 21699
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.220704365617084,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 21700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.19163706544850842,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 21701
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3640271526358068,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 21702
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9906401365653847,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 21703
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.896352055909359,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 21704
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.126779576997543,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 21705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8979172484323368,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 21706
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0494799163571398,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 21707
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5107949924341535,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 21708
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0120545643695755,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 21709
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3491483969312414,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 21710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.610712344541787,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 21711
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1002495157578367,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 21712
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07956124886141842,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 21713
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6206414597202369,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 21714
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.20731327353221898,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 21715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.46238353977015656,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 21716
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.10569399610598,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 21717
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4400237775534936,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 21718
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28363632744251754,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 21719
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3267977197093304,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 21720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09099397667892554,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 21721
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4969681820374761,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 21722
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.693769911246038,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 21723
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2904710829832413,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 21724
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.384333619818936,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 21725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7357737024827037,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 21726
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5545520804907653,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 21727
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0071361307943922,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 21728
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5860787314683915,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 21729
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.251090384791224,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 21730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4220955646181506,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 21731
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0962088058891923,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 21732
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8100998703935227,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 21733
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1986233105903676,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 21734
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.203364798840078,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 21735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9812622286542527,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 21736
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3794036054734715,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 21737
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.688868828888623,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 21738
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9211122946130195,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 21739
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5341160821792896,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 21740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.041054990835978,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 21741
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32494893635726735,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 21742
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8985365108168418,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 21743
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1185555671709535,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 21744
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6871178639141737,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 21745
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5005247521486865,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 21746
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8855809713765118,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 21747
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2750281660871242,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 21748
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.33435455241919776,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 21749
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.281978958058368,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 21750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3615697484833622,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 21751
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1907377274782023,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 21752
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4691906606146735,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 21753
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.034081127452032,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 21754
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8123677953120922,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 21755
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7993074205717248,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 21756
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.373303660875201,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 21757
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6961547165533689,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 21758
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.49565229523530513,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 21759
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9114693922552685,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 21760
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5647419121455675,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 21761
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5610463819437237,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 21762
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.47817495247190517,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 21763
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4454272331112366,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 21764
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.796277286019546,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 21765
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2520010666736043,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 21766
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1583755930597421,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 21767
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9074032450293603,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 21768
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9072471920168644,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 21769
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8799993541563439,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 21770
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8041184093667104,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 21771
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.047248174464632337,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 21772
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1261865927812558,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 21773
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2553579892839952,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 21774
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5969513619592455,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 21775
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16732984837723797,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 21776
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.23087270540068905,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 21777
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.797753563674213,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 21778
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7040037597791562,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 21779
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6804460261127749,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 21780
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31235337393791607,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 21781
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.575665814583418,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 21782
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6066429565542087,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 21783
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.02897456616551771,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 21784
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09061858631326836,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 21785
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5023428150126772,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 21786
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5044103000684848,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 21787
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9451512601181192,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 21788
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1928085329481022,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 21789
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37503015873752377,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 21790
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8317620285100032,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 21791
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0741039422846757,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 21792
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.774353597363435,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 21793
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.21790756064025882,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 21794
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7257029874097004,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 21795
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7968370765189066,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 21796
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0029793244474137,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 21797
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1219114681401177,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 21798
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9091762532937464,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 21799
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9021155369038247,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 21800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8646668236800232,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 21801
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7547008863772413,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 21802
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.15075666914092808,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 21803
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8240176855059673,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 21804
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2749660112126624,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 21805
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5872806920946271,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 21806
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3527388284091599,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 21807
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10320496727875075,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 21808
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8349216978740814,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 21809
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16749268534509973,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 21810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08152886514275459,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 21811
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3656937876819129,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 21812
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5245107494140032,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 21813
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8961871949007854,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 21814
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0359810476226508,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 21815
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4792433227313744,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 21816
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4071005207651584,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 21817
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.46473952335042484,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 21818
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.674456592874371,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 21819
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.052527625051468435,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 21820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.416918909766412,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 21821
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.884792695320495,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 21822
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1606671975328683,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 21823
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4682156378036585,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 21824
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9064140683695445,
+      "learning_rate": 2e-05,
+      "loss": 0.0913,
+      "step": 21825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5210425760319906,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 21826
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1893215263241077,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 21827
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1006833078253555,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 21828
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1775662885825335,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 21829
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09384329445315831,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 21830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8550005114704409,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 21831
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37828723248836993,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 21832
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4343052067949185,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 21833
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3853789398487037,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 21834
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4642182909499,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 21835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.148797997186955,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 21836
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.33759278196524267,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 21837
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.257237341907656,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 21838
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1232552542613503,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 21839
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8660510064977195,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 21840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5535905759450275,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 21841
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2078722644148425,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 21842
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13621036832453684,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 21843
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7153950632677872,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 21844
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9923921966107367,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 21845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.588981525829898,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 21846
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.699819664646437,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 21847
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.850396106707964,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 21848
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9310920107678492,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 21849
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8005327536978372,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 21850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5265619640078474,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 21851
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5356692944396477,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 21852
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9121769000818687,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 21853
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2002703004632724,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 21854
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2820635129909131,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 21855
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8153114827077711,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 21856
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2649053614472003,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 21857
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.05217381629533869,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 21858
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7779705645035735,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 21859
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7754716806781031,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 21860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0073563469477937,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 21861
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.535640106437385,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 21862
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.17167499106983,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 21863
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.963467003762916,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 21864
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28824984526733455,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 21865
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08951614088454822,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 21866
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1107779751830122,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 21867
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27154120529667886,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 21868
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9627107300571809,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 21869
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6332809927502422,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 21870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4649271623386746,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 21871
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7705730222079064,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 21872
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9566726812934172,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 21873
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1036289933296912,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 21874
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5416789011426846,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 21875
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.107988807842039,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 21876
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.883713018191234,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 21877
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34555629262870285,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 21878
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07903321869394085,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 21879
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7537224552223685,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 21880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4244318080088316,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 21881
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4026123469704748,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 21882
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.615767145452201,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 21883
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7054026694203749,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 21884
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3750019073473099,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 21885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.385837930674372,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 21886
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4177807559822821,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 21887
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7804512900231413,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 21888
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7484933899974284,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 21889
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.787274950428067,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 21890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.19959846474230525,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 21891
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9754816047547714,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 21892
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9407167495931728,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 21893
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5160383532801079,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 21894
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6003820305769437,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 21895
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.261244686131986,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 21896
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.239764940224512,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 21897
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4238958096383114,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 21898
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.15589347337290582,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 21899
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3449287119569124,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 21900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8932293742692148,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 21901
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6847679516784867,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 21902
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9363623392015312,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 21903
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0490343399063708,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 21904
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.594887725224071,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 21905
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4746473477718153,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 21906
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9385069525618158,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 21907
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5689329093899682,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 21908
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6729216728450984,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 21909
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.041903119271271,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 21910
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.125802866233827,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 21911
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.113630509205277,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 21912
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7192486401632248,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 21913
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0183712143935928,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 21914
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.743945826882648,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 21915
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9504652910936058,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 21916
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.22551704775046194,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 21917
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2551321366801091,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 21918
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7857235808875436,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 21919
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30497517820882014,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 21920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0853985334017777,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 21921
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25252145055265274,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 21922
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5684808325382222,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 21923
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3264463376469022,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 21924
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2135026404309242,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 21925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8779139446168003,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 21926
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2663397948650907,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 21927
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6776797897817906,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 21928
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.274881330476169,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 21929
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.20728984912637802,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 21930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8869034260153,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 21931
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.729320618850117,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 21932
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9677241338802505,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 21933
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5563027889996128,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 21934
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5425223841641774,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 21935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1547235258725845,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 21936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7311484290765001,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 21937
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8617880389765586,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 21938
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5395731720526813,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 21939
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.025783989871725,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 21940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0045514241523326,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 21941
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6310727493785193,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 21942
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6933757081592961,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 21943
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9161066851873023,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 21944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5837200166080043,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 21945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8172668958208629,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 21946
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7772011500572726,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 21947
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9146240033578417,
+      "learning_rate": 2e-05,
+      "loss": 0.1217,
+      "step": 21948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0142846197339828,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 21949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1074246199639426,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 21950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.623039239913497,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 21951
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0544153498157465,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 21952
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18687631271692248,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 21953
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.19748382163141,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 21954
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2854123728955887,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 21955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6705786596749802,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 21956
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6449158185909005,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 21957
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2724384062873555,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 21958
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7259520405176363,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 21959
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9745471452863492,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 21960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1796418422243982,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 21961
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8083772599846538,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 21962
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7845018214533024,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 21963
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.658784977674199,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 21964
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.712474768290273,
+      "learning_rate": 2e-05,
+      "loss": 0.0902,
+      "step": 21965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5593414903971589,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 21966
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1471879134152085,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 21967
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.173678269555823,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 21968
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17716020547122047,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 21969
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1749742154727331,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 21970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.36288126593635556,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 21971
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0398054406770865,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 21972
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7021324569729941,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 21973
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8189071666019523,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 21974
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7998391291468538,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 21975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.557100981668585,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 21976
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2454998790305187,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 21977
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6137437230816676,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 21978
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07168835358893066,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 21979
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8502972673861446,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 21980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2269306069921252,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 21981
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.33503648890151416,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 21982
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.303832946507596,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 21983
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4094572157467675,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 21984
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.19221431110970408,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 21985
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24406358647806795,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 21986
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.274349924649186,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 21987
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.35308435383367787,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 21988
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2341214397870529,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 21989
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.160197291915194,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 21990
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8574761306176458,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 21991
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9728685321461352,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 21992
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.20190091112981287,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 21993
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29549812709469775,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 21994
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6004092072767612,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 21995
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7501460171656913,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 21996
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5246563070757326,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 21997
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9714046927007005,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 21998
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8871876153411736,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 21999
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.193302333683975,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 22000
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2798664652887766,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 22001
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.051099877813835,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 22002
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.03417288508162984,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 22003
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4729322778194928,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 22004
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.14353364673484645,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 22005
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2201880039005756,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 22006
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5360316918054102,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 22007
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6992942087384517,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 22008
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13637611506362005,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 22009
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3171004097996188,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 22010
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.628106867198101,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 22011
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.47501992886548083,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 22012
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40110213166571324,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 22013
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2384600066577445,
+      "learning_rate": 2e-05,
+      "loss": 0.0754,
+      "step": 22014
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4767937857518822,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 22015
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4781312917158253,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 22016
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.158773297977895,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 22017
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4842893957588523,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 22018
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.177497751432244,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 22019
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.778661230319313,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 22020
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1344391292383573,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 22021
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10496460107160092,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 22022
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4180821565197732,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 22023
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9192497243790035,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 22024
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.683908101272238,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 22025
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4440673854969865,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 22026
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7250571294177409,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 22027
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5418876627018228,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 22028
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7873031339864278,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 22029
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.288221391367973,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 22030
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6785961465707612,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 22031
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6196448261689852,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 22032
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7814008567118971,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 22033
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6088444844608363,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 22034
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0151157179325794,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 22035
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2685654100531639,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 22036
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.624308585746294,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 22037
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.398046854535572,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 22038
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.285014878123795,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 22039
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5392281690012524,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 22040
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.14574210614498898,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 22041
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06268078020606142,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 22042
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6385570271206976,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 22043
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1480685341524494,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 22044
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6383327785502813,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 22045
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06437446750841921,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 22046
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5236091673479684,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 22047
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9256197414803076,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 22048
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25152587150086814,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 22049
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.423138738070836,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 22050
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7767211494298762,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 22051
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40515298547102613,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 22052
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7691050089682805,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 22053
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7665238845249247,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 22054
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1896676808042328,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 22055
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.30604053155989,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 22056
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5387556612553726,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 22057
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.644822796256148,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 22058
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07560966942208174,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 22059
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9106440403504876,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 22060
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4818839774237187,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 22061
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6428971315338785,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 22062
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2913651678493652,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 22063
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8869668890655299,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 22064
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5873501343010992,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 22065
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7839927593062275,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 22066
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8947022036473122,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 22067
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2508995633992372,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 22068
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.734605022686499,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 22069
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1192486296460673,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 22070
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4025631463824909,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 22071
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.073378570570125,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 22072
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5078036087431503,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 22073
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1685450226251612,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 22074
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6356376045158274,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 22075
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7918560755954127,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 22076
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.411563312642679,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 22077
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7162714339638363,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 22078
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3357391783561496,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 22079
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9608408383562084,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 22080
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3820727460358789,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 22081
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1851236787618213,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 22082
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.562724974648377,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 22083
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8400814607459057,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 22084
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.19274243276488542,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 22085
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28849095468125735,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 22086
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9763665269672794,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 22087
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7341683066985285,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 22088
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6802763306618916,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 22089
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.89656043129663,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 22090
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4061357545360136,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 22091
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0403507120713194,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 22092
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.12218194126717,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 22093
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.224232446804486,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 22094
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9098831324748418,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 22095
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5390776038033869,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 22096
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.243280656015661,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 22097
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6168041728906448,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 22098
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9493265600949555,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 22099
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.748756307354692,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 22100
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9539993528127975,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 22101
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.764334684764066,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 22102
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5494038112065152,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 22103
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5830685829203264,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 22104
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5328716332775734,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 22105
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3824943076439441,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 22106
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1626359991738215,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 22107
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.36950011869431,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 22108
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8821959579121782,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 22109
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0829632322528309,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 22110
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.630633077538819,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 22111
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6298484852034858,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 22112
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17587123792707773,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 22113
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0269254695960075,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 22114
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6253117260792302,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 22115
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4602418763193197,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 22116
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7356680783731928,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 22117
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0247120964632566,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 22118
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.20403013850496682,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 22119
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.281747854192461,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 22120
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.942321684322817,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 22121
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7290665777569592,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 22122
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9873887808253732,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 22123
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8768349887217408,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 22124
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7336021170987039,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 22125
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.560544190084693,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 22126
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6722485133977285,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 22127
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2597174635676164,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 22128
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.20637234613663472,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 22129
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2993102366230405,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 22130
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4855631284686355,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 22131
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9821961524672079,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 22132
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9582730391788318,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 22133
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.081263796767925,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 22134
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6910371172178944,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 22135
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.21460949764348,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 22136
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7024480740100655,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 22137
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09049745825628268,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 22138
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5338866441415433,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 22139
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.191108234561393,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 22140
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.046316085806119,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 22141
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.893376834033254,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 22142
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1962239284194602,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 22143
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6458686229286468,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 22144
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7157862968741552,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 22145
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08966842939573838,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 22146
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4262738057953137,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 22147
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5480808043632067,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 22148
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5088239054684323,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 22149
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1660020654414123,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 22150
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.01952481907245,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 22151
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.51449779004817,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 22152
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6310776250389152,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 22153
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5686011308199909,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 22154
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2631731117441998,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 22155
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.516500517003652,
+      "learning_rate": 2e-05,
+      "loss": 0.1163,
+      "step": 22156
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.942005640601188,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 22157
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.056743895519847,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 22158
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1912633090961946,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 22159
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.189483384537548,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 22160
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4475056272934288,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 22161
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.175298451627453,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 22162
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2747602776625273,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 22163
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3916446538413942,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 22164
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5119429231406865,
+      "learning_rate": 2e-05,
+      "loss": 0.0776,
+      "step": 22165
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0611267141806637,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 22166
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8447812101294769,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 22167
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1817979944278925,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 22168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7203086721323997,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 22169
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5165723219372085,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 22170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.20367389572598524,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 22171
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.22838309104515406,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 22172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.182265531130012,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 22173
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.654580336316418,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 22174
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3501219817965655,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 22175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4748136449711914,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 22176
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.37336310120387656,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 22177
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29537222122131573,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 22178
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0952587076383893,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 22179
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7739562901656667,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 22180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6336698334689226,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 22181
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7382916982236476,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 22182
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0757255390177853,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 22183
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.845096807447056,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 22184
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.231773620325568,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 22185
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9865047485538894,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 22186
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1807178267360816,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 22187
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5220932727610674,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 22188
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.85914235653869,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 22189
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0170587250137761,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 22190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2985929118403765,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 22191
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.05523926323798183,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 22192
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.478055330533884,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 22193
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5828065764507117,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 22194
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.592983734823693,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 22195
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.155665885203542,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 22196
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6303723466005293,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 22197
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6765656313906931,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 22198
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3485588080669121,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 22199
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2378660164242724,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 22200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1116547874391536,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 22201
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4841544972245634,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 22202
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.16374972592760403,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 22203
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.238745300061965,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 22204
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6262726700842566,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 22205
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8223301261779203,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 22206
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9181407843543485,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 22207
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.962563584043717,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 22208
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.11674545680882038,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 22209
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4894499377235206,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 22210
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6113514271155489,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 22211
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6404149478706455,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 22212
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1485389612752899,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 22213
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10037001795358298,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 22214
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5334947442361762,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 22215
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4237983187680956,
+      "learning_rate": 2e-05,
+      "loss": 0.0982,
+      "step": 22216
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.204122331121691,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 22217
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2438520939079055,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 22218
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5254374361365242,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 22219
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9689179551749847,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 22220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.970258050790674,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 22221
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6266828293018069,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 22222
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06596016769362091,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 22223
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9906325311682986,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 22224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3935508065233483,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 22225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09748042968584125,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 22226
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.967966317315016,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 22227
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9030717713878325,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 22228
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5344721310826521,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 22229
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6070263424236799,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 22230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9520899045510153,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 22231
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0332893090458037,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 22232
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8285508050783107,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 22233
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2647226095952841,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 22234
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23245102034478388,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 22235
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.42170714750365,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 22236
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.22573299486276105,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 22237
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8508346994330003,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 22238
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1244606738611367,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 22239
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7230861313115717,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 22240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2546325273220474,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 22241
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49211353170041655,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 22242
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5185115817844778,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 22243
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4817834014448965,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 22244
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2688345958947508,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 22245
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2904914641464022,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 22246
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.154460883446999,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 22247
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2197748427859896,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 22248
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2940652582967804,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 22249
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.484533483973092,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 22250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.618987109518777,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 22251
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.990147701992915,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 22252
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9967653889918824,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 22253
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49925484862573405,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 22254
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5021183632387147,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 22255
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7444196363209299,
+      "learning_rate": 2e-05,
+      "loss": 0.0732,
+      "step": 22256
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.38048540975766704,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 22257
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4147568584758034,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 22258
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06373825781451747,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 22259
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9891333412326815,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 22260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4136991508236862,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 22261
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3259260814857474,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 22262
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.670463629504717,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 22263
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6959729219285654,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 22264
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3465677537339282,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 22265
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.44768554431872093,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 22266
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7614919080172813,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 22267
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2234326417322272,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 22268
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3438961845287233,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 22269
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6234332952142478,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 22270
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9461520856643506,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 22271
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.564030165057008,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 22272
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1007408606355287,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 22273
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7788467445715451,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 22274
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8843504750655807,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 22275
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0492297208188304,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 22276
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6880605261197685,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 22277
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0185354931376962,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 22278
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3634455016253508,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 22279
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.141542927563513,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 22280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5345157131460765,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 22281
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0418607276712408,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 22282
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.14325696213099492,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 22283
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.52765919581587,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 22284
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5316226437657942,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 22285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8646981906874096,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 22286
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.001213063477587,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 22287
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5177312157589646,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 22288
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5631408749289263,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 22289
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.060179705697852,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 22290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.158855698259463,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 22291
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.050613848090035,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 22292
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5719322846366222,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 22293
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7359504844346193,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 22294
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6175971060242199,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 22295
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4087147623172105,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 22296
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3387118526975539,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 22297
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.882417900558573,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 22298
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6047301186222379,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 22299
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.21547257557627114,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 22300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7739389123614934,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 22301
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.510279882564124,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 22302
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06002347198345322,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 22303
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8501398644658729,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 22304
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.43528825411671956,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 22305
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6603015762270426,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 22306
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5909234327379825,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 22307
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2172712514478683,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 22308
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7364537608775454,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 22309
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5886798516374129,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 22310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3267848827323935,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 22311
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7474083783930692,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 22312
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8143634932105517,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 22313
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5754427742129797,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 22314
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.22331600153687067,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 22315
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3139810716519222,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 22316
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9612647329027341,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 22317
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.309755419453951,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 22318
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.102548469118143,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 22319
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2114663230695943,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 22320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5444840070023895,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 22321
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2064780044510983,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 22322
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6471865239849146,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 22323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0064133387142582,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 22324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2532674047363788,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 22325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.566054162397767,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 22326
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2942203869523403,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 22327
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3300059087180963,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 22328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7945141695750414,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 22329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1534797544043082,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 22330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4309260810406939,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 22331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6925659152792657,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 22332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.052267144996694,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 22333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28816554375911974,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 22334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5993777376354976,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 22335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8375296839031043,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 22336
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3669585117945349,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 22337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.764977192825924,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 22338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.499737327839617,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 22339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9810441596992611,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 22340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.255168101614378,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 22341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2039266411681249,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 22342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.928703644171337,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 22343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6794737923869414,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 22344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.800967646048979,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 22345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6776192207237751,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 22346
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8572694821150584,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 22347
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4402087025931853,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 22348
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4865325506439964,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 22349
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0999582203372403,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 22350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.045202484955956,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 22351
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.05331040506925798,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 22352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5856538716275062,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 22353
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8768111625920594,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 22354
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6075738060625927,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 22355
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24084721683349694,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 22356
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7123964033604712,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 22357
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9729151791892912,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 22358
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6563369602534547,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 22359
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0008189900747113,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 22360
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3931198029658647,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 22361
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1729209083035386,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 22362
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.532840787298888,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 22363
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0404865573746114,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 22364
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2742998964930446,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 22365
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.14779739017358473,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 22366
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8339264825028669,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 22367
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5285199799684,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 22368
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12205750588949732,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 22369
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3678993690669348,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 22370
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6627522356095439,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 22371
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9576802348329179,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 22372
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8758298821264834,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 22373
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33528787433609203,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 22374
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1486997759012751,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 22375
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.023674048719876,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 22376
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5805420224753517,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 22377
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.950902630495418,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 22378
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.00851772406319,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 22379
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5406894507864486,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 22380
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7343398082015242,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 22381
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6530640788161902,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 22382
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3425717099849459,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 22383
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10403825523995398,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 22384
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45407519520328665,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 22385
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.374068638343452,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 22386
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9492566450427442,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 22387
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5603839278341218,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 22388
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9990160989351982,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 22389
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9741445897072384,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 22390
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5504548478678741,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 22391
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2371279289741383,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 22392
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.014080550036699,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 22393
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.08246341467587595,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 22394
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0916533819735754,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 22395
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1753463701288593,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 22396
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2699113736131752,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 22397
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2297946112587428,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 22398
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.576831476349492,
+      "learning_rate": 2e-05,
+      "loss": 0.1218,
+      "step": 22399
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5778115943748924,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 22400
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 99630,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 200,
+  "total_flos": 1755100520448000.0,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}