diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,31104 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 4434,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006765899864682003,
+      "grad_norm": 1294.7799740470043,
+      "learning_rate": 2.2522522522522524e-08,
+      "loss": 11.9131,
+      "step": 1
+    },
+    {
+      "epoch": 0.0013531799729364006,
+      "grad_norm": 1298.4331558214524,
+      "learning_rate": 4.504504504504505e-08,
+      "loss": 11.9163,
+      "step": 2
+    },
+    {
+      "epoch": 0.0020297699594046007,
+      "grad_norm": 1325.2434867763561,
+      "learning_rate": 6.756756756756757e-08,
+      "loss": 11.9102,
+      "step": 3
+    },
+    {
+      "epoch": 0.0027063599458728013,
+      "grad_norm": 1366.1141393954351,
+      "learning_rate": 9.00900900900901e-08,
+      "loss": 11.9533,
+      "step": 4
+    },
+    {
+      "epoch": 0.0033829499323410014,
+      "grad_norm": 1425.1438127122265,
+      "learning_rate": 1.1261261261261262e-07,
+      "loss": 11.928,
+      "step": 5
+    },
+    {
+      "epoch": 0.0040595399188092015,
+      "grad_norm": 1395.5312633979067,
+      "learning_rate": 1.3513513513513515e-07,
+      "loss": 11.9208,
+      "step": 6
+    },
+    {
+      "epoch": 0.004736129905277402,
+      "grad_norm": 1438.0298310330422,
+      "learning_rate": 1.5765765765765766e-07,
+      "loss": 11.8879,
+      "step": 7
+    },
+    {
+      "epoch": 0.005412719891745603,
+      "grad_norm": 1636.1424356180714,
+      "learning_rate": 1.801801801801802e-07,
+      "loss": 11.8713,
+      "step": 8
+    },
+    {
+      "epoch": 0.006089309878213802,
+      "grad_norm": 1638.3207575482402,
+      "learning_rate": 2.0270270270270273e-07,
+      "loss": 11.7031,
+      "step": 9
+    },
+    {
+      "epoch": 0.006765899864682003,
+      "grad_norm": 1796.5344880486534,
+      "learning_rate": 2.2522522522522524e-07,
+      "loss": 11.6355,
+      "step": 10
+    },
+    {
+      "epoch": 0.007442489851150203,
+      "grad_norm": 1140.5545815120106,
+      "learning_rate": 2.477477477477478e-07,
+      "loss": 11.1436,
+      "step": 11
+    },
+    {
+      "epoch": 0.008119079837618403,
+      "grad_norm": 1054.7170118427175,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 11.0597,
+      "step": 12
+    },
+    {
+      "epoch": 0.008795669824086604,
+      "grad_norm": 1026.8961411097375,
+      "learning_rate": 2.927927927927928e-07,
+      "loss": 10.9774,
+      "step": 13
+    },
+    {
+      "epoch": 0.009472259810554804,
+      "grad_norm": 1032.8878691346852,
+      "learning_rate": 3.153153153153153e-07,
+      "loss": 10.8926,
+      "step": 14
+    },
+    {
+      "epoch": 0.010148849797023005,
+      "grad_norm": 982.9642790210778,
+      "learning_rate": 3.378378378378379e-07,
+      "loss": 10.0553,
+      "step": 15
+    },
+    {
+      "epoch": 0.010825439783491205,
+      "grad_norm": 1205.4064168126627,
+      "learning_rate": 3.603603603603604e-07,
+      "loss": 9.9305,
+      "step": 16
+    },
+    {
+      "epoch": 0.011502029769959404,
+      "grad_norm": 889.7008770631086,
+      "learning_rate": 3.828828828828829e-07,
+      "loss": 9.8475,
+      "step": 17
+    },
+    {
+      "epoch": 0.012178619756427604,
+      "grad_norm": 797.7351269703299,
+      "learning_rate": 4.0540540540540546e-07,
+      "loss": 9.6404,
+      "step": 18
+    },
+    {
+      "epoch": 0.012855209742895805,
+      "grad_norm": 588.7198873727792,
+      "learning_rate": 4.27927927927928e-07,
+      "loss": 9.5483,
+      "step": 19
+    },
+    {
+      "epoch": 0.013531799729364006,
+      "grad_norm": 312.9505155993464,
+      "learning_rate": 4.504504504504505e-07,
+      "loss": 8.9424,
+      "step": 20
+    },
+    {
+      "epoch": 0.014208389715832206,
+      "grad_norm": 365.06295714206897,
+      "learning_rate": 4.7297297297297305e-07,
+      "loss": 8.8538,
+      "step": 21
+    },
+    {
+      "epoch": 0.014884979702300407,
+      "grad_norm": 278.7861480433846,
+      "learning_rate": 4.954954954954956e-07,
+      "loss": 8.7556,
+      "step": 22
+    },
+    {
+      "epoch": 0.015561569688768605,
+      "grad_norm": 224.60328456473022,
+      "learning_rate": 5.180180180180181e-07,
+      "loss": 8.7774,
+      "step": 23
+    },
+    {
+      "epoch": 0.016238159675236806,
+      "grad_norm": 194.31032323395746,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 8.6965,
+      "step": 24
+    },
+    {
+      "epoch": 0.016914749661705007,
+      "grad_norm": 171.4227734694513,
+      "learning_rate": 5.630630630630631e-07,
+      "loss": 8.6245,
+      "step": 25
+    },
+    {
+      "epoch": 0.017591339648173207,
+      "grad_norm": 181.2827273623685,
+      "learning_rate": 5.855855855855856e-07,
+      "loss": 8.5409,
+      "step": 26
+    },
+    {
+      "epoch": 0.018267929634641408,
+      "grad_norm": 316.2362978808448,
+      "learning_rate": 6.081081081081082e-07,
+      "loss": 8.3035,
+      "step": 27
+    },
+    {
+      "epoch": 0.018944519621109608,
+      "grad_norm": 199.10162323615126,
+      "learning_rate": 6.306306306306306e-07,
+      "loss": 8.1246,
+      "step": 28
+    },
+    {
+      "epoch": 0.01962110960757781,
+      "grad_norm": 158.8747267496112,
+      "learning_rate": 6.531531531531532e-07,
+      "loss": 7.9727,
+      "step": 29
+    },
+    {
+      "epoch": 0.02029769959404601,
+      "grad_norm": 142.90489272104213,
+      "learning_rate": 6.756756756756758e-07,
+      "loss": 7.9089,
+      "step": 30
+    },
+    {
+      "epoch": 0.02097428958051421,
+      "grad_norm": 112.95667777019665,
+      "learning_rate": 6.981981981981982e-07,
+      "loss": 7.8067,
+      "step": 31
+    },
+    {
+      "epoch": 0.02165087956698241,
+      "grad_norm": 99.97524285645189,
+      "learning_rate": 7.207207207207208e-07,
+      "loss": 7.7026,
+      "step": 32
+    },
+    {
+      "epoch": 0.022327469553450607,
+      "grad_norm": 130.4578778756448,
+      "learning_rate": 7.432432432432434e-07,
+      "loss": 7.6579,
+      "step": 33
+    },
+    {
+      "epoch": 0.023004059539918808,
+      "grad_norm": 168.55159538067102,
+      "learning_rate": 7.657657657657658e-07,
+      "loss": 7.5495,
+      "step": 34
+    },
+    {
+      "epoch": 0.02368064952638701,
+      "grad_norm": 82.56038654646147,
+      "learning_rate": 7.882882882882883e-07,
+      "loss": 7.4503,
+      "step": 35
+    },
+    {
+      "epoch": 0.02435723951285521,
+      "grad_norm": 153.7208230043944,
+      "learning_rate": 8.108108108108109e-07,
+      "loss": 7.3843,
+      "step": 36
+    },
+    {
+      "epoch": 0.02503382949932341,
+      "grad_norm": 91.97910168568909,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 7.2279,
+      "step": 37
+    },
+    {
+      "epoch": 0.02571041948579161,
+      "grad_norm": 69.76268667391646,
+      "learning_rate": 8.55855855855856e-07,
+      "loss": 7.2079,
+      "step": 38
+    },
+    {
+      "epoch": 0.02638700947225981,
+      "grad_norm": 77.57527495553569,
+      "learning_rate": 8.783783783783785e-07,
+      "loss": 7.0677,
+      "step": 39
+    },
+    {
+      "epoch": 0.02706359945872801,
+      "grad_norm": 52.26613705638054,
+      "learning_rate": 9.00900900900901e-07,
+      "loss": 7.0129,
+      "step": 40
+    },
+    {
+      "epoch": 0.02774018944519621,
+      "grad_norm": 53.80168795569409,
+      "learning_rate": 9.234234234234235e-07,
+      "loss": 6.9802,
+      "step": 41
+    },
+    {
+      "epoch": 0.028416779431664412,
+      "grad_norm": 59.68642115374619,
+      "learning_rate": 9.459459459459461e-07,
+      "loss": 6.8825,
+      "step": 42
+    },
+    {
+      "epoch": 0.029093369418132613,
+      "grad_norm": 91.42597837015056,
+      "learning_rate": 9.684684684684686e-07,
+      "loss": 6.8503,
+      "step": 43
+    },
+    {
+      "epoch": 0.029769959404600813,
+      "grad_norm": 101.20418159876743,
+      "learning_rate": 9.909909909909911e-07,
+      "loss": 6.8065,
+      "step": 44
+    },
+    {
+      "epoch": 0.030446549391069014,
+      "grad_norm": 43.005131526460474,
+      "learning_rate": 1.0135135135135136e-06,
+      "loss": 6.7757,
+      "step": 45
+    },
+    {
+      "epoch": 0.03112313937753721,
+      "grad_norm": 42.64828727931448,
+      "learning_rate": 1.0360360360360361e-06,
+      "loss": 6.6841,
+      "step": 46
+    },
+    {
+      "epoch": 0.031799729364005415,
+      "grad_norm": 97.23527419575188,
+      "learning_rate": 1.0585585585585587e-06,
+      "loss": 6.6109,
+      "step": 47
+    },
+    {
+      "epoch": 0.03247631935047361,
+      "grad_norm": 169.8678960637773,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 6.5767,
+      "step": 48
+    },
+    {
+      "epoch": 0.033152909336941816,
+      "grad_norm": 62.92191915861204,
+      "learning_rate": 1.1036036036036037e-06,
+      "loss": 6.5223,
+      "step": 49
+    },
+    {
+      "epoch": 0.03382949932341001,
+      "grad_norm": 136.66699482372732,
+      "learning_rate": 1.1261261261261262e-06,
+      "loss": 6.493,
+      "step": 50
+    },
+    {
+      "epoch": 0.03450608930987822,
+      "grad_norm": 81.89062026113558,
+      "learning_rate": 1.148648648648649e-06,
+      "loss": 6.3889,
+      "step": 51
+    },
+    {
+      "epoch": 0.035182679296346414,
+      "grad_norm": 118.4391312567866,
+      "learning_rate": 1.1711711711711712e-06,
+      "loss": 6.4154,
+      "step": 52
+    },
+    {
+      "epoch": 0.03585926928281461,
+      "grad_norm": 73.87596492845805,
+      "learning_rate": 1.1936936936936937e-06,
+      "loss": 6.361,
+      "step": 53
+    },
+    {
+      "epoch": 0.036535859269282815,
+      "grad_norm": 89.03923620297563,
+      "learning_rate": 1.2162162162162164e-06,
+      "loss": 6.2851,
+      "step": 54
+    },
+    {
+      "epoch": 0.03721244925575101,
+      "grad_norm": 73.1383569774123,
+      "learning_rate": 1.2387387387387387e-06,
+      "loss": 6.2606,
+      "step": 55
+    },
+    {
+      "epoch": 0.037889039242219216,
+      "grad_norm": 44.97513849432748,
+      "learning_rate": 1.2612612612612613e-06,
+      "loss": 6.2387,
+      "step": 56
+    },
+    {
+      "epoch": 0.03856562922868741,
+      "grad_norm": 79.41788212317785,
+      "learning_rate": 1.2837837837837838e-06,
+      "loss": 6.181,
+      "step": 57
+    },
+    {
+      "epoch": 0.03924221921515562,
+      "grad_norm": 58.612598455402114,
+      "learning_rate": 1.3063063063063065e-06,
+      "loss": 6.1738,
+      "step": 58
+    },
+    {
+      "epoch": 0.039918809201623814,
+      "grad_norm": 63.80154957894249,
+      "learning_rate": 1.328828828828829e-06,
+      "loss": 6.0666,
+      "step": 59
+    },
+    {
+      "epoch": 0.04059539918809202,
+      "grad_norm": 31.593281308336145,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 6.0429,
+      "step": 60
+    },
+    {
+      "epoch": 0.041271989174560215,
+      "grad_norm": 74.39180683057303,
+      "learning_rate": 1.373873873873874e-06,
+      "loss": 6.0485,
+      "step": 61
+    },
+    {
+      "epoch": 0.04194857916102842,
+      "grad_norm": 74.28723111978216,
+      "learning_rate": 1.3963963963963963e-06,
+      "loss": 5.9679,
+      "step": 62
+    },
+    {
+      "epoch": 0.04262516914749662,
+      "grad_norm": 71.70645602731112,
+      "learning_rate": 1.418918918918919e-06,
+      "loss": 5.9602,
+      "step": 63
+    },
+    {
+      "epoch": 0.04330175913396482,
+      "grad_norm": 61.89579228047879,
+      "learning_rate": 1.4414414414414416e-06,
+      "loss": 5.8796,
+      "step": 64
+    },
+    {
+      "epoch": 0.04397834912043302,
+      "grad_norm": 90.73272911379827,
+      "learning_rate": 1.463963963963964e-06,
+      "loss": 5.8541,
+      "step": 65
+    },
+    {
+      "epoch": 0.044654939106901215,
+      "grad_norm": 45.77250463669355,
+      "learning_rate": 1.4864864864864868e-06,
+      "loss": 5.7697,
+      "step": 66
+    },
+    {
+      "epoch": 0.04533152909336942,
+      "grad_norm": 91.74739934484515,
+      "learning_rate": 1.5090090090090093e-06,
+      "loss": 5.7921,
+      "step": 67
+    },
+    {
+      "epoch": 0.046008119079837616,
+      "grad_norm": 93.61184773344448,
+      "learning_rate": 1.5315315315315316e-06,
+      "loss": 5.7767,
+      "step": 68
+    },
+    {
+      "epoch": 0.04668470906630582,
+      "grad_norm": 39.84854834313949,
+      "learning_rate": 1.5540540540540541e-06,
+      "loss": 5.5705,
+      "step": 69
+    },
+    {
+      "epoch": 0.04736129905277402,
+      "grad_norm": 61.450451753613216,
+      "learning_rate": 1.5765765765765766e-06,
+      "loss": 5.5836,
+      "step": 70
+    },
+    {
+      "epoch": 0.04803788903924222,
+      "grad_norm": 134.76740317296654,
+      "learning_rate": 1.5990990990990993e-06,
+      "loss": 5.5517,
+      "step": 71
+    },
+    {
+      "epoch": 0.04871447902571042,
+      "grad_norm": 54.398999052693796,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 5.4925,
+      "step": 72
+    },
+    {
+      "epoch": 0.04939106901217862,
+      "grad_norm": 68.95008079459163,
+      "learning_rate": 1.6441441441441444e-06,
+      "loss": 5.4054,
+      "step": 73
+    },
+    {
+      "epoch": 0.05006765899864682,
+      "grad_norm": 152.54862412840902,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 5.3895,
+      "step": 74
+    },
+    {
+      "epoch": 0.05074424898511502,
+      "grad_norm": 59.01945826720156,
+      "learning_rate": 1.6891891891891894e-06,
+      "loss": 5.3621,
+      "step": 75
+    },
+    {
+      "epoch": 0.05142083897158322,
+      "grad_norm": 64.87355979260256,
+      "learning_rate": 1.711711711711712e-06,
+      "loss": 5.2075,
+      "step": 76
+    },
+    {
+      "epoch": 0.052097428958051424,
+      "grad_norm": 84.64730212336376,
+      "learning_rate": 1.7342342342342344e-06,
+      "loss": 5.1949,
+      "step": 77
+    },
+    {
+      "epoch": 0.05277401894451962,
+      "grad_norm": 55.40660875297934,
+      "learning_rate": 1.756756756756757e-06,
+      "loss": 5.1719,
+      "step": 78
+    },
+    {
+      "epoch": 0.05345060893098782,
+      "grad_norm": 149.80942207452873,
+      "learning_rate": 1.7792792792792792e-06,
+      "loss": 5.1801,
+      "step": 79
+    },
+    {
+      "epoch": 0.05412719891745602,
+      "grad_norm": 86.22324212364583,
+      "learning_rate": 1.801801801801802e-06,
+      "loss": 5.0994,
+      "step": 80
+    },
+    {
+      "epoch": 0.05480378890392422,
+      "grad_norm": 99.81963307964912,
+      "learning_rate": 1.8243243243243245e-06,
+      "loss": 5.0621,
+      "step": 81
+    },
+    {
+      "epoch": 0.05548037889039242,
+      "grad_norm": 80.3364116036936,
+      "learning_rate": 1.846846846846847e-06,
+      "loss": 4.9882,
+      "step": 82
+    },
+    {
+      "epoch": 0.05615696887686062,
+      "grad_norm": 88.89892195412031,
+      "learning_rate": 1.8693693693693697e-06,
+      "loss": 4.9389,
+      "step": 83
+    },
+    {
+      "epoch": 0.056833558863328824,
+      "grad_norm": 113.17311584350547,
+      "learning_rate": 1.8918918918918922e-06,
+      "loss": 4.9489,
+      "step": 84
+    },
+    {
+      "epoch": 0.05751014884979702,
+      "grad_norm": 136.83618643424379,
+      "learning_rate": 1.9144144144144145e-06,
+      "loss": 4.8699,
+      "step": 85
+    },
+    {
+      "epoch": 0.058186738836265225,
+      "grad_norm": 75.51256109039457,
+      "learning_rate": 1.9369369369369372e-06,
+      "loss": 4.7703,
+      "step": 86
+    },
+    {
+      "epoch": 0.05886332882273342,
+      "grad_norm": 96.23041293661558,
+      "learning_rate": 1.9594594594594595e-06,
+      "loss": 4.7873,
+      "step": 87
+    },
+    {
+      "epoch": 0.05953991880920163,
+      "grad_norm": 161.37019408814405,
+      "learning_rate": 1.9819819819819822e-06,
+      "loss": 4.7921,
+      "step": 88
+    },
+    {
+      "epoch": 0.060216508795669824,
+      "grad_norm": 71.3560661853614,
+      "learning_rate": 2.0045045045045045e-06,
+      "loss": 4.7185,
+      "step": 89
+    },
+    {
+      "epoch": 0.06089309878213803,
+      "grad_norm": 69.99168283497767,
+      "learning_rate": 2.0270270270270273e-06,
+      "loss": 4.5539,
+      "step": 90
+    },
+    {
+      "epoch": 0.061569688768606225,
+      "grad_norm": 114.51118147162934,
+      "learning_rate": 2.0495495495495496e-06,
+      "loss": 4.5638,
+      "step": 91
+    },
+    {
+      "epoch": 0.06224627875507442,
+      "grad_norm": 178.4518511158092,
+      "learning_rate": 2.0720720720720723e-06,
+      "loss": 4.6301,
+      "step": 92
+    },
+    {
+      "epoch": 0.06292286874154263,
+      "grad_norm": 68.22787157402642,
+      "learning_rate": 2.0945945945945946e-06,
+      "loss": 4.4462,
+      "step": 93
+    },
+    {
+      "epoch": 0.06359945872801083,
+      "grad_norm": 160.54020098575876,
+      "learning_rate": 2.1171171171171173e-06,
+      "loss": 4.5235,
+      "step": 94
+    },
+    {
+      "epoch": 0.06427604871447902,
+      "grad_norm": 132.99722363643008,
+      "learning_rate": 2.13963963963964e-06,
+      "loss": 4.5069,
+      "step": 95
+    },
+    {
+      "epoch": 0.06495263870094722,
+      "grad_norm": 90.41930171832493,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 4.4033,
+      "step": 96
+    },
+    {
+      "epoch": 0.06562922868741543,
+      "grad_norm": 155.6113006075671,
+      "learning_rate": 2.1846846846846846e-06,
+      "loss": 4.3758,
+      "step": 97
+    },
+    {
+      "epoch": 0.06630581867388363,
+      "grad_norm": 122.2472173078203,
+      "learning_rate": 2.2072072072072073e-06,
+      "loss": 4.2974,
+      "step": 98
+    },
+    {
+      "epoch": 0.06698240866035182,
+      "grad_norm": 111.61521534772868,
+      "learning_rate": 2.22972972972973e-06,
+      "loss": 4.3209,
+      "step": 99
+    },
+    {
+      "epoch": 0.06765899864682003,
+      "grad_norm": 122.18880976249555,
+      "learning_rate": 2.2522522522522524e-06,
+      "loss": 4.26,
+      "step": 100
+    },
+    {
+      "epoch": 0.06833558863328823,
+      "grad_norm": 80.45807204609973,
+      "learning_rate": 2.274774774774775e-06,
+      "loss": 4.1875,
+      "step": 101
+    },
+    {
+      "epoch": 0.06901217861975643,
+      "grad_norm": 109.02033184349753,
+      "learning_rate": 2.297297297297298e-06,
+      "loss": 4.1745,
+      "step": 102
+    },
+    {
+      "epoch": 0.06968876860622462,
+      "grad_norm": 110.10640416298837,
+      "learning_rate": 2.31981981981982e-06,
+      "loss": 4.1548,
+      "step": 103
+    },
+    {
+      "epoch": 0.07036535859269283,
+      "grad_norm": 86.96841790122264,
+      "learning_rate": 2.3423423423423424e-06,
+      "loss": 4.1614,
+      "step": 104
+    },
+    {
+      "epoch": 0.07104194857916103,
+      "grad_norm": 97.15417103395869,
+      "learning_rate": 2.364864864864865e-06,
+      "loss": 4.0024,
+      "step": 105
+    },
+    {
+      "epoch": 0.07171853856562922,
+      "grad_norm": 136.2349346627184,
+      "learning_rate": 2.3873873873873874e-06,
+      "loss": 4.1771,
+      "step": 106
+    },
+    {
+      "epoch": 0.07239512855209743,
+      "grad_norm": 70.38037906411675,
+      "learning_rate": 2.40990990990991e-06,
+      "loss": 4.0541,
+      "step": 107
+    },
+    {
+      "epoch": 0.07307171853856563,
+      "grad_norm": 90.75425332026454,
+      "learning_rate": 2.432432432432433e-06,
+      "loss": 4.0113,
+      "step": 108
+    },
+    {
+      "epoch": 0.07374830852503383,
+      "grad_norm": 78.1950389368104,
+      "learning_rate": 2.454954954954955e-06,
+      "loss": 4.0502,
+      "step": 109
+    },
+    {
+      "epoch": 0.07442489851150202,
+      "grad_norm": 70.23321966900136,
+      "learning_rate": 2.4774774774774775e-06,
+      "loss": 3.9199,
+      "step": 110
+    },
+    {
+      "epoch": 0.07510148849797023,
+      "grad_norm": 92.15274678991364,
+      "learning_rate": 2.5e-06,
+      "loss": 3.9936,
+      "step": 111
+    },
+    {
+      "epoch": 0.07577807848443843,
+      "grad_norm": 147.6498031425845,
+      "learning_rate": 2.5225225225225225e-06,
+      "loss": 3.9897,
+      "step": 112
+    },
+    {
+      "epoch": 0.07645466847090664,
+      "grad_norm": 71.75018072076564,
+      "learning_rate": 2.5450450450450452e-06,
+      "loss": 3.8258,
+      "step": 113
+    },
+    {
+      "epoch": 0.07713125845737483,
+      "grad_norm": 144.25923235185607,
+      "learning_rate": 2.5675675675675675e-06,
+      "loss": 4.0061,
+      "step": 114
+    },
+    {
+      "epoch": 0.07780784844384303,
+      "grad_norm": 90.10313216179375,
+      "learning_rate": 2.5900900900900907e-06,
+      "loss": 3.8232,
+      "step": 115
+    },
+    {
+      "epoch": 0.07848443843031123,
+      "grad_norm": 62.702162623266574,
+      "learning_rate": 2.612612612612613e-06,
+      "loss": 3.7921,
+      "step": 116
+    },
+    {
+      "epoch": 0.07916102841677942,
+      "grad_norm": 128.51448612306217,
+      "learning_rate": 2.6351351351351353e-06,
+      "loss": 3.787,
+      "step": 117
+    },
+    {
+      "epoch": 0.07983761840324763,
+      "grad_norm": 106.85834399979501,
+      "learning_rate": 2.657657657657658e-06,
+      "loss": 3.8209,
+      "step": 118
+    },
+    {
+      "epoch": 0.08051420838971583,
+      "grad_norm": 74.34385902839736,
+      "learning_rate": 2.6801801801801803e-06,
+      "loss": 3.6591,
+      "step": 119
+    },
+    {
+      "epoch": 0.08119079837618404,
+      "grad_norm": 74.135577478339,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 3.6711,
+      "step": 120
+    },
+    {
+      "epoch": 0.08186738836265223,
+      "grad_norm": 85.95726780507401,
+      "learning_rate": 2.7252252252252253e-06,
+      "loss": 3.7639,
+      "step": 121
+    },
+    {
+      "epoch": 0.08254397834912043,
+      "grad_norm": 102.2908279358922,
+      "learning_rate": 2.747747747747748e-06,
+      "loss": 3.7085,
+      "step": 122
+    },
+    {
+      "epoch": 0.08322056833558863,
+      "grad_norm": 74.03973608536255,
+      "learning_rate": 2.7702702702702703e-06,
+      "loss": 3.7246,
+      "step": 123
+    },
+    {
+      "epoch": 0.08389715832205684,
+      "grad_norm": 85.12172994853982,
+      "learning_rate": 2.7927927927927926e-06,
+      "loss": 3.7063,
+      "step": 124
+    },
+    {
+      "epoch": 0.08457374830852503,
+      "grad_norm": 54.05878326063868,
+      "learning_rate": 2.8153153153153158e-06,
+      "loss": 3.5608,
+      "step": 125
+    },
+    {
+      "epoch": 0.08525033829499323,
+      "grad_norm": 65.2074038079349,
+      "learning_rate": 2.837837837837838e-06,
+      "loss": 3.5081,
+      "step": 126
+    },
+    {
+      "epoch": 0.08592692828146144,
+      "grad_norm": 44.86678222702895,
+      "learning_rate": 2.860360360360361e-06,
+      "loss": 3.5391,
+      "step": 127
+    },
+    {
+      "epoch": 0.08660351826792964,
+      "grad_norm": 71.466489787615,
+      "learning_rate": 2.882882882882883e-06,
+      "loss": 3.4926,
+      "step": 128
+    },
+    {
+      "epoch": 0.08728010825439783,
+      "grad_norm": 39.11904869837415,
+      "learning_rate": 2.9054054054054054e-06,
+      "loss": 3.4806,
+      "step": 129
+    },
+    {
+      "epoch": 0.08795669824086604,
+      "grad_norm": 124.47175823597068,
+      "learning_rate": 2.927927927927928e-06,
+      "loss": 3.6277,
+      "step": 130
+    },
+    {
+      "epoch": 0.08863328822733424,
+      "grad_norm": 67.66619348885503,
+      "learning_rate": 2.9504504504504504e-06,
+      "loss": 3.5235,
+      "step": 131
+    },
+    {
+      "epoch": 0.08930987821380243,
+      "grad_norm": 67.72976849710265,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 3.4433,
+      "step": 132
+    },
+    {
+      "epoch": 0.08998646820027063,
+      "grad_norm": 59.48885329779129,
+      "learning_rate": 2.995495495495496e-06,
+      "loss": 3.4593,
+      "step": 133
+    },
+    {
+      "epoch": 0.09066305818673884,
+      "grad_norm": 41.831980152297795,
+      "learning_rate": 3.0180180180180186e-06,
+      "loss": 3.3989,
+      "step": 134
+    },
+    {
+      "epoch": 0.09133964817320704,
+      "grad_norm": 78.06010995786896,
+      "learning_rate": 3.040540540540541e-06,
+      "loss": 3.4448,
+      "step": 135
+    },
+    {
+      "epoch": 0.09201623815967523,
+      "grad_norm": 51.0196629836936,
+      "learning_rate": 3.063063063063063e-06,
+      "loss": 3.3393,
+      "step": 136
+    },
+    {
+      "epoch": 0.09269282814614344,
+      "grad_norm": 52.123543615948236,
+      "learning_rate": 3.085585585585586e-06,
+      "loss": 3.3557,
+      "step": 137
+    },
+    {
+      "epoch": 0.09336941813261164,
+      "grad_norm": 50.51804959908098,
+      "learning_rate": 3.1081081081081082e-06,
+      "loss": 3.3341,
+      "step": 138
+    },
+    {
+      "epoch": 0.09404600811907984,
+      "grad_norm": 59.51105112522959,
+      "learning_rate": 3.130630630630631e-06,
+      "loss": 3.2939,
+      "step": 139
+    },
+    {
+      "epoch": 0.09472259810554803,
+      "grad_norm": 56.287180220643776,
+      "learning_rate": 3.1531531531531532e-06,
+      "loss": 3.341,
+      "step": 140
+    },
+    {
+      "epoch": 0.09539918809201624,
+      "grad_norm": 75.52518208966724,
+      "learning_rate": 3.1756756756756755e-06,
+      "loss": 3.3611,
+      "step": 141
+    },
+    {
+      "epoch": 0.09607577807848444,
+      "grad_norm": 48.910828651220605,
+      "learning_rate": 3.1981981981981987e-06,
+      "loss": 3.2481,
+      "step": 142
+    },
+    {
+      "epoch": 0.09675236806495263,
+      "grad_norm": 47.218392810236764,
+      "learning_rate": 3.220720720720721e-06,
+      "loss": 3.2752,
+      "step": 143
+    },
+    {
+      "epoch": 0.09742895805142084,
+      "grad_norm": 63.94134891175671,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 3.3645,
+      "step": 144
+    },
+    {
+      "epoch": 0.09810554803788904,
+      "grad_norm": 54.666789257562534,
+      "learning_rate": 3.265765765765766e-06,
+      "loss": 3.2376,
+      "step": 145
+    },
+    {
+      "epoch": 0.09878213802435724,
+      "grad_norm": 36.94989947017251,
+      "learning_rate": 3.2882882882882887e-06,
+      "loss": 3.1951,
+      "step": 146
+    },
+    {
+      "epoch": 0.09945872801082543,
+      "grad_norm": 48.33680686944214,
+      "learning_rate": 3.310810810810811e-06,
+      "loss": 3.1377,
+      "step": 147
+    },
+    {
+      "epoch": 0.10013531799729364,
+      "grad_norm": 35.74504153701644,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 3.2506,
+      "step": 148
+    },
+    {
+      "epoch": 0.10081190798376184,
+      "grad_norm": 49.239566709083704,
+      "learning_rate": 3.3558558558558565e-06,
+      "loss": 3.1739,
+      "step": 149
+    },
+    {
+      "epoch": 0.10148849797023005,
+      "grad_norm": 49.98074107148794,
+      "learning_rate": 3.3783783783783788e-06,
+      "loss": 3.2584,
+      "step": 150
+    },
+    {
+      "epoch": 0.10216508795669824,
+      "grad_norm": 57.12006421383496,
+      "learning_rate": 3.4009009009009015e-06,
+      "loss": 3.1443,
+      "step": 151
+    },
+    {
+      "epoch": 0.10284167794316644,
+      "grad_norm": 51.83187954914922,
+      "learning_rate": 3.423423423423424e-06,
+      "loss": 3.2512,
+      "step": 152
+    },
+    {
+      "epoch": 0.10351826792963464,
+      "grad_norm": 41.928571108063004,
+      "learning_rate": 3.445945945945946e-06,
+      "loss": 3.0448,
+      "step": 153
+    },
+    {
+      "epoch": 0.10419485791610285,
+      "grad_norm": 44.516486827766585,
+      "learning_rate": 3.468468468468469e-06,
+      "loss": 3.1117,
+      "step": 154
+    },
+    {
+      "epoch": 0.10487144790257104,
+      "grad_norm": 42.27354179773472,
+      "learning_rate": 3.490990990990991e-06,
+      "loss": 3.129,
+      "step": 155
+    },
+    {
+      "epoch": 0.10554803788903924,
+      "grad_norm": 48.86935574613096,
+      "learning_rate": 3.513513513513514e-06,
+      "loss": 3.0613,
+      "step": 156
+    },
+    {
+      "epoch": 0.10622462787550745,
+      "grad_norm": 37.05452600718779,
+      "learning_rate": 3.536036036036036e-06,
+      "loss": 2.9485,
+      "step": 157
+    },
+    {
+      "epoch": 0.10690121786197564,
+      "grad_norm": 42.975006587033576,
+      "learning_rate": 3.5585585585585584e-06,
+      "loss": 3.1076,
+      "step": 158
+    },
+    {
+      "epoch": 0.10757780784844384,
+      "grad_norm": 36.19174264324649,
+      "learning_rate": 3.5810810810810816e-06,
+      "loss": 3.1296,
+      "step": 159
+    },
+    {
+      "epoch": 0.10825439783491204,
+      "grad_norm": 51.57163331423397,
+      "learning_rate": 3.603603603603604e-06,
+      "loss": 2.9326,
+      "step": 160
+    },
+    {
+      "epoch": 0.10893098782138025,
+      "grad_norm": 28.926129651651678,
+      "learning_rate": 3.6261261261261266e-06,
+      "loss": 2.8705,
+      "step": 161
+    },
+    {
+      "epoch": 0.10960757780784844,
+      "grad_norm": 37.471976631295426,
+      "learning_rate": 3.648648648648649e-06,
+      "loss": 3.1038,
+      "step": 162
+    },
+    {
+      "epoch": 0.11028416779431664,
+      "grad_norm": 40.95517313070317,
+      "learning_rate": 3.6711711711711716e-06,
+      "loss": 2.8534,
+      "step": 163
+    },
+    {
+      "epoch": 0.11096075778078485,
+      "grad_norm": 30.52496431382125,
+      "learning_rate": 3.693693693693694e-06,
+      "loss": 2.9543,
+      "step": 164
+    },
+    {
+      "epoch": 0.11163734776725305,
+      "grad_norm": 25.34559414829517,
+      "learning_rate": 3.7162162162162162e-06,
+      "loss": 2.9391,
+      "step": 165
+    },
+    {
+      "epoch": 0.11231393775372124,
+      "grad_norm": 41.00047666306257,
+      "learning_rate": 3.7387387387387394e-06,
+      "loss": 2.9402,
+      "step": 166
+    },
+    {
+      "epoch": 0.11299052774018944,
+      "grad_norm": 38.40358833951516,
+      "learning_rate": 3.7612612612612612e-06,
+      "loss": 2.8399,
+      "step": 167
+    },
+    {
+      "epoch": 0.11366711772665765,
+      "grad_norm": 25.778136989342435,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 2.8529,
+      "step": 168
+    },
+    {
+      "epoch": 0.11434370771312584,
+      "grad_norm": 43.89518290972547,
+      "learning_rate": 3.8063063063063067e-06,
+      "loss": 2.8869,
+      "step": 169
+    },
+    {
+      "epoch": 0.11502029769959404,
+      "grad_norm": 35.78579942027662,
+      "learning_rate": 3.828828828828829e-06,
+      "loss": 2.8886,
+      "step": 170
+    },
+    {
+      "epoch": 0.11569688768606225,
+      "grad_norm": 38.32163511272516,
+      "learning_rate": 3.851351351351352e-06,
+      "loss": 2.8628,
+      "step": 171
+    },
+    {
+      "epoch": 0.11637347767253045,
+      "grad_norm": 24.390093584623898,
+      "learning_rate": 3.8738738738738744e-06,
+      "loss": 2.8215,
+      "step": 172
+    },
+    {
+      "epoch": 0.11705006765899864,
+      "grad_norm": 33.2833424715733,
+      "learning_rate": 3.896396396396397e-06,
+      "loss": 2.7963,
+      "step": 173
+    },
+    {
+      "epoch": 0.11772665764546685,
+      "grad_norm": 33.3311496772107,
+      "learning_rate": 3.918918918918919e-06,
+      "loss": 2.7788,
+      "step": 174
+    },
+    {
+      "epoch": 0.11840324763193505,
+      "grad_norm": 30.752650997097287,
+      "learning_rate": 3.941441441441442e-06,
+      "loss": 2.8082,
+      "step": 175
+    },
+    {
+      "epoch": 0.11907983761840325,
+      "grad_norm": 36.63901906403704,
+      "learning_rate": 3.9639639639639645e-06,
+      "loss": 2.8628,
+      "step": 176
+    },
+    {
+      "epoch": 0.11975642760487144,
+      "grad_norm": 27.97375841977239,
+      "learning_rate": 3.986486486486487e-06,
+      "loss": 2.6965,
+      "step": 177
+    },
+    {
+      "epoch": 0.12043301759133965,
+      "grad_norm": 21.067527031604392,
+      "learning_rate": 4.009009009009009e-06,
+      "loss": 2.7417,
+      "step": 178
+    },
+    {
+      "epoch": 0.12110960757780785,
+      "grad_norm": 27.504039817885566,
+      "learning_rate": 4.031531531531531e-06,
+      "loss": 2.6387,
+      "step": 179
+    },
+    {
+      "epoch": 0.12178619756427606,
+      "grad_norm": 40.89949692772889,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 2.7225,
+      "step": 180
+    },
+    {
+      "epoch": 0.12246278755074425,
+      "grad_norm": 29.287860242320992,
+      "learning_rate": 4.076576576576577e-06,
+      "loss": 2.6841,
+      "step": 181
+    },
+    {
+      "epoch": 0.12313937753721245,
+      "grad_norm": 23.16703424289219,
+      "learning_rate": 4.099099099099099e-06,
+      "loss": 2.5594,
+      "step": 182
+    },
+    {
+      "epoch": 0.12381596752368065,
+      "grad_norm": 29.543799764885705,
+      "learning_rate": 4.121621621621622e-06,
+      "loss": 2.6742,
+      "step": 183
+    },
+    {
+      "epoch": 0.12449255751014884,
+      "grad_norm": 34.95327011699371,
+      "learning_rate": 4.1441441441441446e-06,
+      "loss": 2.6456,
+      "step": 184
+    },
+    {
+      "epoch": 0.12516914749661706,
+      "grad_norm": 29.66519736577202,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 2.6835,
+      "step": 185
+    },
+    {
+      "epoch": 0.12584573748308525,
+      "grad_norm": 24.40015340294758,
+      "learning_rate": 4.189189189189189e-06,
+      "loss": 2.5426,
+      "step": 186
+    },
+    {
+      "epoch": 0.12652232746955344,
+      "grad_norm": 35.7957191425465,
+      "learning_rate": 4.2117117117117115e-06,
+      "loss": 2.609,
+      "step": 187
+    },
+    {
+      "epoch": 0.12719891745602166,
+      "grad_norm": 32.11830023640179,
+      "learning_rate": 4.234234234234235e-06,
+      "loss": 2.6174,
+      "step": 188
+    },
+    {
+      "epoch": 0.12787550744248985,
+      "grad_norm": 27.713420304792045,
+      "learning_rate": 4.256756756756757e-06,
+      "loss": 2.6129,
+      "step": 189
+    },
+    {
+      "epoch": 0.12855209742895804,
+      "grad_norm": 25.13598174696013,
+      "learning_rate": 4.27927927927928e-06,
+      "loss": 2.4842,
+      "step": 190
+    },
+    {
+      "epoch": 0.12922868741542626,
+      "grad_norm": 20.315219130144893,
+      "learning_rate": 4.301801801801802e-06,
+      "loss": 2.5572,
+      "step": 191
+    },
+    {
+      "epoch": 0.12990527740189445,
+      "grad_norm": 22.887320533614375,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 2.5483,
+      "step": 192
+    },
+    {
+      "epoch": 0.13058186738836267,
+      "grad_norm": 20.914247782101484,
+      "learning_rate": 4.346846846846847e-06,
+      "loss": 2.5147,
+      "step": 193
+    },
+    {
+      "epoch": 0.13125845737483086,
+      "grad_norm": 20.697866625432855,
+      "learning_rate": 4.369369369369369e-06,
+      "loss": 2.5196,
+      "step": 194
+    },
+    {
+      "epoch": 0.13193504736129905,
+      "grad_norm": 23.49583918293928,
+      "learning_rate": 4.391891891891892e-06,
+      "loss": 2.4808,
+      "step": 195
+    },
+    {
+      "epoch": 0.13261163734776726,
+      "grad_norm": 32.74575243265221,
+      "learning_rate": 4.414414414414415e-06,
+      "loss": 2.4672,
+      "step": 196
+    },
+    {
+      "epoch": 0.13328822733423545,
+      "grad_norm": 22.23921840689608,
+      "learning_rate": 4.436936936936938e-06,
+      "loss": 2.452,
+      "step": 197
+    },
+    {
+      "epoch": 0.13396481732070364,
+      "grad_norm": 21.013548498079665,
+      "learning_rate": 4.45945945945946e-06,
+      "loss": 2.3407,
+      "step": 198
+    },
+    {
+      "epoch": 0.13464140730717186,
+      "grad_norm": 31.878428050622567,
+      "learning_rate": 4.4819819819819824e-06,
+      "loss": 2.469,
+      "step": 199
+    },
+    {
+      "epoch": 0.13531799729364005,
+      "grad_norm": 22.974412989264575,
+      "learning_rate": 4.504504504504505e-06,
+      "loss": 2.4717,
+      "step": 200
+    },
+    {
+      "epoch": 0.13599458728010824,
+      "grad_norm": 17.882363303347567,
+      "learning_rate": 4.527027027027027e-06,
+      "loss": 2.4163,
+      "step": 201
+    },
+    {
+      "epoch": 0.13667117726657646,
+      "grad_norm": 22.493322907839186,
+      "learning_rate": 4.54954954954955e-06,
+      "loss": 2.4467,
+      "step": 202
+    },
+    {
+      "epoch": 0.13734776725304465,
+      "grad_norm": 19.548539758212335,
+      "learning_rate": 4.5720720720720725e-06,
+      "loss": 2.3677,
+      "step": 203
+    },
+    {
+      "epoch": 0.13802435723951287,
+      "grad_norm": 22.72777213575744,
+      "learning_rate": 4.594594594594596e-06,
+      "loss": 2.4086,
+      "step": 204
+    },
+    {
+      "epoch": 0.13870094722598106,
+      "grad_norm": 22.499189227034087,
+      "learning_rate": 4.617117117117118e-06,
+      "loss": 2.3678,
+      "step": 205
+    },
+    {
+      "epoch": 0.13937753721244925,
+      "grad_norm": 26.867693885650482,
+      "learning_rate": 4.63963963963964e-06,
+      "loss": 2.2966,
+      "step": 206
+    },
+    {
+      "epoch": 0.14005412719891747,
+      "grad_norm": 18.98888773696992,
+      "learning_rate": 4.6621621621621625e-06,
+      "loss": 2.2594,
+      "step": 207
+    },
+    {
+      "epoch": 0.14073071718538566,
+      "grad_norm": 20.923260812895332,
+      "learning_rate": 4.684684684684685e-06,
+      "loss": 2.3373,
+      "step": 208
+    },
+    {
+      "epoch": 0.14140730717185385,
+      "grad_norm": 22.43708159337281,
+      "learning_rate": 4.707207207207208e-06,
+      "loss": 2.349,
+      "step": 209
+    },
+    {
+      "epoch": 0.14208389715832206,
+      "grad_norm": 15.791501622620217,
+      "learning_rate": 4.72972972972973e-06,
+      "loss": 2.2408,
+      "step": 210
+    },
+    {
+      "epoch": 0.14276048714479025,
+      "grad_norm": 27.9477620074452,
+      "learning_rate": 4.7522522522522526e-06,
+      "loss": 2.3142,
+      "step": 211
+    },
+    {
+      "epoch": 0.14343707713125844,
+      "grad_norm": 20.28378726114416,
+      "learning_rate": 4.774774774774775e-06,
+      "loss": 2.2812,
+      "step": 212
+    },
+    {
+      "epoch": 0.14411366711772666,
+      "grad_norm": 23.301416792731526,
+      "learning_rate": 4.797297297297297e-06,
+      "loss": 2.2495,
+      "step": 213
+    },
+    {
+      "epoch": 0.14479025710419485,
+      "grad_norm": 20.477578746933535,
+      "learning_rate": 4.81981981981982e-06,
+      "loss": 2.2285,
+      "step": 214
+    },
+    {
+      "epoch": 0.14546684709066307,
+      "grad_norm": 15.941850024131293,
+      "learning_rate": 4.842342342342343e-06,
+      "loss": 2.2056,
+      "step": 215
+    },
+    {
+      "epoch": 0.14614343707713126,
+      "grad_norm": 32.42025261478585,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 2.3471,
+      "step": 216
+    },
+    {
+      "epoch": 0.14682002706359945,
+      "grad_norm": 18.338599158407256,
+      "learning_rate": 4.887387387387388e-06,
+      "loss": 2.2433,
+      "step": 217
+    },
+    {
+      "epoch": 0.14749661705006767,
+      "grad_norm": 20.454382479996063,
+      "learning_rate": 4.90990990990991e-06,
+      "loss": 2.2717,
+      "step": 218
+    },
+    {
+      "epoch": 0.14817320703653586,
+      "grad_norm": 16.14586789867164,
+      "learning_rate": 4.932432432432433e-06,
+      "loss": 2.2331,
+      "step": 219
+    },
+    {
+      "epoch": 0.14884979702300405,
+      "grad_norm": 22.113495495267212,
+      "learning_rate": 4.954954954954955e-06,
+      "loss": 2.1726,
+      "step": 220
+    },
+    {
+      "epoch": 0.14952638700947227,
+      "grad_norm": 15.95883139356251,
+      "learning_rate": 4.977477477477478e-06,
+      "loss": 2.1805,
+      "step": 221
+    },
+    {
+      "epoch": 0.15020297699594046,
+      "grad_norm": 18.79118628674278,
+      "learning_rate": 5e-06,
+      "loss": 2.148,
+      "step": 222
+    },
+    {
+      "epoch": 0.15087956698240865,
+      "grad_norm": 14.403072472289919,
+      "learning_rate": 5.022522522522523e-06,
+      "loss": 2.1814,
+      "step": 223
+    },
+    {
+      "epoch": 0.15155615696887687,
+      "grad_norm": 16.137035847145082,
+      "learning_rate": 5.045045045045045e-06,
+      "loss": 2.1728,
+      "step": 224
+    },
+    {
+      "epoch": 0.15223274695534506,
+      "grad_norm": 18.356637411169537,
+      "learning_rate": 5.067567567567568e-06,
+      "loss": 2.1512,
+      "step": 225
+    },
+    {
+      "epoch": 0.15290933694181327,
+      "grad_norm": 16.504673195570593,
+      "learning_rate": 5.0900900900900905e-06,
+      "loss": 2.1588,
+      "step": 226
+    },
+    {
+      "epoch": 0.15358592692828146,
+      "grad_norm": 13.899118228878194,
+      "learning_rate": 5.112612612612613e-06,
+      "loss": 2.0877,
+      "step": 227
+    },
+    {
+      "epoch": 0.15426251691474965,
+      "grad_norm": 17.912245295524656,
+      "learning_rate": 5.135135135135135e-06,
+      "loss": 2.1466,
+      "step": 228
+    },
+    {
+      "epoch": 0.15493910690121787,
+      "grad_norm": 13.534422946138394,
+      "learning_rate": 5.157657657657657e-06,
+      "loss": 2.0077,
+      "step": 229
+    },
+    {
+      "epoch": 0.15561569688768606,
+      "grad_norm": 22.62031309389824,
+      "learning_rate": 5.180180180180181e-06,
+      "loss": 2.1199,
+      "step": 230
+    },
+    {
+      "epoch": 0.15629228687415425,
+      "grad_norm": 15.743397945734683,
+      "learning_rate": 5.202702702702704e-06,
+      "loss": 2.0383,
+      "step": 231
+    },
+    {
+      "epoch": 0.15696887686062247,
+      "grad_norm": 19.395021397972336,
+      "learning_rate": 5.225225225225226e-06,
+      "loss": 1.985,
+      "step": 232
+    },
+    {
+      "epoch": 0.15764546684709066,
+      "grad_norm": 13.009112032469817,
+      "learning_rate": 5.247747747747748e-06,
+      "loss": 2.0359,
+      "step": 233
+    },
+    {
+      "epoch": 0.15832205683355885,
+      "grad_norm": 20.359644663731412,
+      "learning_rate": 5.2702702702702705e-06,
+      "loss": 2.0488,
+      "step": 234
+    },
+    {
+      "epoch": 0.15899864682002707,
+      "grad_norm": 15.53405031855752,
+      "learning_rate": 5.292792792792794e-06,
+      "loss": 2.0,
+      "step": 235
+    },
+    {
+      "epoch": 0.15967523680649526,
+      "grad_norm": 14.809102908639488,
+      "learning_rate": 5.315315315315316e-06,
+      "loss": 2.0319,
+      "step": 236
+    },
+    {
+      "epoch": 0.16035182679296348,
+      "grad_norm": 19.88444381670935,
+      "learning_rate": 5.337837837837838e-06,
+      "loss": 1.9792,
+      "step": 237
+    },
+    {
+      "epoch": 0.16102841677943167,
+      "grad_norm": 13.118123019183166,
+      "learning_rate": 5.360360360360361e-06,
+      "loss": 1.9924,
+      "step": 238
+    },
+    {
+      "epoch": 0.16170500676589986,
+      "grad_norm": 18.482436970789802,
+      "learning_rate": 5.382882882882884e-06,
+      "loss": 2.0255,
+      "step": 239
+    },
+    {
+      "epoch": 0.16238159675236807,
+      "grad_norm": 17.198808199707006,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 1.9914,
+      "step": 240
+    },
+    {
+      "epoch": 0.16305818673883626,
+      "grad_norm": 18.95392851638832,
+      "learning_rate": 5.427927927927928e-06,
+      "loss": 2.0717,
+      "step": 241
+    },
+    {
+      "epoch": 0.16373477672530445,
+      "grad_norm": 13.5701694050597,
+      "learning_rate": 5.450450450450451e-06,
+      "loss": 2.0124,
+      "step": 242
+    },
+    {
+      "epoch": 0.16441136671177267,
+      "grad_norm": 15.414340952802318,
+      "learning_rate": 5.472972972972973e-06,
+      "loss": 1.9557,
+      "step": 243
+    },
+    {
+      "epoch": 0.16508795669824086,
+      "grad_norm": 17.440870830742558,
+      "learning_rate": 5.495495495495496e-06,
+      "loss": 1.9832,
+      "step": 244
+    },
+    {
+      "epoch": 0.16576454668470908,
+      "grad_norm": 14.239022528198992,
+      "learning_rate": 5.518018018018018e-06,
+      "loss": 1.9448,
+      "step": 245
+    },
+    {
+      "epoch": 0.16644113667117727,
+      "grad_norm": 17.431495370774222,
+      "learning_rate": 5.540540540540541e-06,
+      "loss": 1.9717,
+      "step": 246
+    },
+    {
+      "epoch": 0.16711772665764546,
+      "grad_norm": 14.79637144533563,
+      "learning_rate": 5.563063063063063e-06,
+      "loss": 1.9688,
+      "step": 247
+    },
+    {
+      "epoch": 0.16779431664411368,
+      "grad_norm": 18.333738504558415,
+      "learning_rate": 5.585585585585585e-06,
+      "loss": 2.0003,
+      "step": 248
+    },
+    {
+      "epoch": 0.16847090663058187,
+      "grad_norm": 13.005075190265611,
+      "learning_rate": 5.608108108108109e-06,
+      "loss": 1.96,
+      "step": 249
+    },
+    {
+      "epoch": 0.16914749661705006,
+      "grad_norm": 12.997802259515195,
+      "learning_rate": 5.6306306306306316e-06,
+      "loss": 1.8585,
+      "step": 250
+    },
+    {
+      "epoch": 0.16982408660351828,
+      "grad_norm": 11.014175768405828,
+      "learning_rate": 5.653153153153154e-06,
+      "loss": 1.8642,
+      "step": 251
+    },
+    {
+      "epoch": 0.17050067658998647,
+      "grad_norm": 18.048619090207655,
+      "learning_rate": 5.675675675675676e-06,
+      "loss": 1.9065,
+      "step": 252
+    },
+    {
+      "epoch": 0.17117726657645466,
+      "grad_norm": 15.060560783435594,
+      "learning_rate": 5.6981981981981985e-06,
+      "loss": 1.9563,
+      "step": 253
+    },
+    {
+      "epoch": 0.17185385656292287,
+      "grad_norm": 10.891591875795623,
+      "learning_rate": 5.720720720720722e-06,
+      "loss": 1.8022,
+      "step": 254
+    },
+    {
+      "epoch": 0.17253044654939106,
+      "grad_norm": 12.582318972666503,
+      "learning_rate": 5.743243243243244e-06,
+      "loss": 1.9277,
+      "step": 255
+    },
+    {
+      "epoch": 0.17320703653585928,
+      "grad_norm": 17.020939171544274,
+      "learning_rate": 5.765765765765766e-06,
+      "loss": 1.8127,
+      "step": 256
+    },
+    {
+      "epoch": 0.17388362652232747,
+      "grad_norm": 10.908213123889828,
+      "learning_rate": 5.7882882882882885e-06,
+      "loss": 1.7682,
+      "step": 257
+    },
+    {
+      "epoch": 0.17456021650879566,
+      "grad_norm": 12.609343066655516,
+      "learning_rate": 5.810810810810811e-06,
+      "loss": 1.8415,
+      "step": 258
+    },
+    {
+      "epoch": 0.17523680649526388,
+      "grad_norm": 12.051612217755917,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 1.782,
+      "step": 259
+    },
+    {
+      "epoch": 0.17591339648173207,
+      "grad_norm": 11.688029779834665,
+      "learning_rate": 5.855855855855856e-06,
+      "loss": 1.8429,
+      "step": 260
+    },
+    {
+      "epoch": 0.17658998646820026,
+      "grad_norm": 9.621847962601386,
+      "learning_rate": 5.8783783783783786e-06,
+      "loss": 1.7889,
+      "step": 261
+    },
+    {
+      "epoch": 0.17726657645466848,
+      "grad_norm": 11.637111117095575,
+      "learning_rate": 5.900900900900901e-06,
+      "loss": 1.7855,
+      "step": 262
+    },
+    {
+      "epoch": 0.17794316644113667,
+      "grad_norm": 15.340550756068295,
+      "learning_rate": 5.923423423423423e-06,
+      "loss": 1.8019,
+      "step": 263
+    },
+    {
+      "epoch": 0.17861975642760486,
+      "grad_norm": 11.88029892034545,
+      "learning_rate": 5.945945945945947e-06,
+      "loss": 1.7269,
+      "step": 264
+    },
+    {
+      "epoch": 0.17929634641407308,
+      "grad_norm": 10.274778570979059,
+      "learning_rate": 5.9684684684684694e-06,
+      "loss": 1.703,
+      "step": 265
+    },
+    {
+      "epoch": 0.17997293640054127,
+      "grad_norm": 13.750962892213895,
+      "learning_rate": 5.990990990990992e-06,
+      "loss": 1.7555,
+      "step": 266
+    },
+    {
+      "epoch": 0.18064952638700948,
+      "grad_norm": 11.485943554362244,
+      "learning_rate": 6.013513513513514e-06,
+      "loss": 1.734,
+      "step": 267
+    },
+    {
+      "epoch": 0.18132611637347767,
+      "grad_norm": 13.961022884592403,
+      "learning_rate": 6.036036036036037e-06,
+      "loss": 1.7312,
+      "step": 268
+    },
+    {
+      "epoch": 0.18200270635994586,
+      "grad_norm": 13.429683042761775,
+      "learning_rate": 6.0585585585585595e-06,
+      "loss": 1.6984,
+      "step": 269
+    },
+    {
+      "epoch": 0.18267929634641408,
+      "grad_norm": 10.691005666701471,
+      "learning_rate": 6.081081081081082e-06,
+      "loss": 1.7554,
+      "step": 270
+    },
+    {
+      "epoch": 0.18335588633288227,
+      "grad_norm": 9.791557551761352,
+      "learning_rate": 6.103603603603604e-06,
+      "loss": 1.7218,
+      "step": 271
+    },
+    {
+      "epoch": 0.18403247631935046,
+      "grad_norm": 12.162695738299966,
+      "learning_rate": 6.126126126126126e-06,
+      "loss": 1.6978,
+      "step": 272
+    },
+    {
+      "epoch": 0.18470906630581868,
+      "grad_norm": 12.52239274552492,
+      "learning_rate": 6.1486486486486495e-06,
+      "loss": 1.7204,
+      "step": 273
+    },
+    {
+      "epoch": 0.18538565629228687,
+      "grad_norm": 11.069520567295438,
+      "learning_rate": 6.171171171171172e-06,
+      "loss": 1.8051,
+      "step": 274
+    },
+    {
+      "epoch": 0.18606224627875506,
+      "grad_norm": 13.646838283946098,
+      "learning_rate": 6.193693693693694e-06,
+      "loss": 1.7197,
+      "step": 275
+    },
+    {
+      "epoch": 0.18673883626522328,
+      "grad_norm": 10.739838933303147,
+      "learning_rate": 6.2162162162162164e-06,
+      "loss": 1.6782,
+      "step": 276
+    },
+    {
+      "epoch": 0.18741542625169147,
+      "grad_norm": 11.402083709510906,
+      "learning_rate": 6.238738738738739e-06,
+      "loss": 1.6366,
+      "step": 277
+    },
+    {
+      "epoch": 0.1880920162381597,
+      "grad_norm": 12.6985139467155,
+      "learning_rate": 6.261261261261262e-06,
+      "loss": 1.7035,
+      "step": 278
+    },
+    {
+      "epoch": 0.18876860622462788,
+      "grad_norm": 11.95572588600681,
+      "learning_rate": 6.283783783783784e-06,
+      "loss": 1.7244,
+      "step": 279
+    },
+    {
+      "epoch": 0.18944519621109607,
+      "grad_norm": 13.151801727488719,
+      "learning_rate": 6.3063063063063065e-06,
+      "loss": 1.7128,
+      "step": 280
+    },
+    {
+      "epoch": 0.19012178619756429,
+      "grad_norm": 9.841504797900614,
+      "learning_rate": 6.328828828828829e-06,
+      "loss": 1.6895,
+      "step": 281
+    },
+    {
+      "epoch": 0.19079837618403248,
+      "grad_norm": 9.787646149578237,
+      "learning_rate": 6.351351351351351e-06,
+      "loss": 1.7138,
+      "step": 282
+    },
+    {
+      "epoch": 0.19147496617050067,
+      "grad_norm": 11.161339025880242,
+      "learning_rate": 6.373873873873875e-06,
+      "loss": 1.6576,
+      "step": 283
+    },
+    {
+      "epoch": 0.19215155615696888,
+      "grad_norm": 9.159618429348464,
+      "learning_rate": 6.396396396396397e-06,
+      "loss": 1.5906,
+      "step": 284
+    },
+    {
+      "epoch": 0.19282814614343707,
+      "grad_norm": 10.811401066646267,
+      "learning_rate": 6.41891891891892e-06,
+      "loss": 1.6524,
+      "step": 285
+    },
+    {
+      "epoch": 0.19350473612990526,
+      "grad_norm": 11.045079449022317,
+      "learning_rate": 6.441441441441442e-06,
+      "loss": 1.6468,
+      "step": 286
+    },
+    {
+      "epoch": 0.19418132611637348,
+      "grad_norm": 9.955353642039189,
+      "learning_rate": 6.463963963963964e-06,
+      "loss": 1.6601,
+      "step": 287
+    },
+    {
+      "epoch": 0.19485791610284167,
+      "grad_norm": 11.353233362482896,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 1.6481,
+      "step": 288
+    },
+    {
+      "epoch": 0.1955345060893099,
+      "grad_norm": 11.4616219039945,
+      "learning_rate": 6.50900900900901e-06,
+      "loss": 1.6293,
+      "step": 289
+    },
+    {
+      "epoch": 0.19621109607577808,
+      "grad_norm": 8.799864129115917,
+      "learning_rate": 6.531531531531532e-06,
+      "loss": 1.548,
+      "step": 290
+    },
+    {
+      "epoch": 0.19688768606224627,
+      "grad_norm": 12.497463518680908,
+      "learning_rate": 6.554054054054054e-06,
+      "loss": 1.6479,
+      "step": 291
+    },
+    {
+      "epoch": 0.1975642760487145,
+      "grad_norm": 11.300656550432288,
+      "learning_rate": 6.5765765765765775e-06,
+      "loss": 1.5232,
+      "step": 292
+    },
+    {
+      "epoch": 0.19824086603518268,
+      "grad_norm": 11.503528750517631,
+      "learning_rate": 6.5990990990991e-06,
+      "loss": 1.6006,
+      "step": 293
+    },
+    {
+      "epoch": 0.19891745602165087,
+      "grad_norm": 7.911223197398728,
+      "learning_rate": 6.621621621621622e-06,
+      "loss": 1.5595,
+      "step": 294
+    },
+    {
+      "epoch": 0.19959404600811909,
+      "grad_norm": 11.536759095047055,
+      "learning_rate": 6.644144144144144e-06,
+      "loss": 1.5718,
+      "step": 295
+    },
+    {
+      "epoch": 0.20027063599458728,
+      "grad_norm": 8.515018062684854,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.5974,
+      "step": 296
+    },
+    {
+      "epoch": 0.2009472259810555,
+      "grad_norm": 11.189846996969566,
+      "learning_rate": 6.689189189189191e-06,
+      "loss": 1.548,
+      "step": 297
+    },
+    {
+      "epoch": 0.20162381596752368,
+      "grad_norm": 8.714980297029278,
+      "learning_rate": 6.711711711711713e-06,
+      "loss": 1.5136,
+      "step": 298
+    },
+    {
+      "epoch": 0.20230040595399187,
+      "grad_norm": 12.00408954819733,
+      "learning_rate": 6.734234234234235e-06,
+      "loss": 1.6347,
+      "step": 299
+    },
+    {
+      "epoch": 0.2029769959404601,
+      "grad_norm": 10.798788556467306,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 1.5315,
+      "step": 300
+    },
+    {
+      "epoch": 0.20365358592692828,
+      "grad_norm": 9.59027269733314,
+      "learning_rate": 6.77927927927928e-06,
+      "loss": 1.4653,
+      "step": 301
+    },
+    {
+      "epoch": 0.20433017591339647,
+      "grad_norm": 9.209203507811885,
+      "learning_rate": 6.801801801801803e-06,
+      "loss": 1.5597,
+      "step": 302
+    },
+    {
+      "epoch": 0.2050067658998647,
+      "grad_norm": 10.137126068903864,
+      "learning_rate": 6.824324324324325e-06,
+      "loss": 1.5163,
+      "step": 303
+    },
+    {
+      "epoch": 0.20568335588633288,
+      "grad_norm": 9.68307545199484,
+      "learning_rate": 6.846846846846848e-06,
+      "loss": 1.5265,
+      "step": 304
+    },
+    {
+      "epoch": 0.20635994587280107,
+      "grad_norm": 8.807889928010955,
+      "learning_rate": 6.86936936936937e-06,
+      "loss": 1.4709,
+      "step": 305
+    },
+    {
+      "epoch": 0.2070365358592693,
+      "grad_norm": 11.257125948904491,
+      "learning_rate": 6.891891891891892e-06,
+      "loss": 1.5649,
+      "step": 306
+    },
+    {
+      "epoch": 0.20771312584573748,
+      "grad_norm": 8.703225933763383,
+      "learning_rate": 6.914414414414415e-06,
+      "loss": 1.4575,
+      "step": 307
+    },
+    {
+      "epoch": 0.2083897158322057,
+      "grad_norm": 9.511676852333661,
+      "learning_rate": 6.936936936936938e-06,
+      "loss": 1.5125,
+      "step": 308
+    },
+    {
+      "epoch": 0.2090663058186739,
+      "grad_norm": 8.873193504576163,
+      "learning_rate": 6.95945945945946e-06,
+      "loss": 1.4193,
+      "step": 309
+    },
+    {
+      "epoch": 0.20974289580514208,
+      "grad_norm": 8.468313508182078,
+      "learning_rate": 6.981981981981982e-06,
+      "loss": 1.494,
+      "step": 310
+    },
+    {
+      "epoch": 0.2104194857916103,
+      "grad_norm": 8.614596249065656,
+      "learning_rate": 7.0045045045045045e-06,
+      "loss": 1.5012,
+      "step": 311
+    },
+    {
+      "epoch": 0.21109607577807848,
+      "grad_norm": 9.01193447897298,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 1.4495,
+      "step": 312
+    },
+    {
+      "epoch": 0.21177266576454667,
+      "grad_norm": 9.198271014590313,
+      "learning_rate": 7.04954954954955e-06,
+      "loss": 1.4238,
+      "step": 313
+    },
+    {
+      "epoch": 0.2124492557510149,
+      "grad_norm": 8.315800768848687,
+      "learning_rate": 7.072072072072072e-06,
+      "loss": 1.4382,
+      "step": 314
+    },
+    {
+      "epoch": 0.21312584573748308,
+      "grad_norm": 8.162579495233,
+      "learning_rate": 7.0945945945945946e-06,
+      "loss": 1.4358,
+      "step": 315
+    },
+    {
+      "epoch": 0.21380243572395127,
+      "grad_norm": 8.382615195195696,
+      "learning_rate": 7.117117117117117e-06,
+      "loss": 1.4738,
+      "step": 316
+    },
+    {
+      "epoch": 0.2144790257104195,
+      "grad_norm": 6.873539297867009,
+      "learning_rate": 7.139639639639641e-06,
+      "loss": 1.4068,
+      "step": 317
+    },
+    {
+      "epoch": 0.21515561569688768,
+      "grad_norm": 8.057956476417067,
+      "learning_rate": 7.162162162162163e-06,
+      "loss": 1.4231,
+      "step": 318
+    },
+    {
+      "epoch": 0.2158322056833559,
+      "grad_norm": 8.760194161062197,
+      "learning_rate": 7.1846846846846855e-06,
+      "loss": 1.432,
+      "step": 319
+    },
+    {
+      "epoch": 0.2165087956698241,
+      "grad_norm": 8.088862479265627,
+      "learning_rate": 7.207207207207208e-06,
+      "loss": 1.317,
+      "step": 320
+    },
+    {
+      "epoch": 0.21718538565629228,
+      "grad_norm": 7.349788195182965,
+      "learning_rate": 7.229729729729731e-06,
+      "loss": 1.3908,
+      "step": 321
+    },
+    {
+      "epoch": 0.2178619756427605,
+      "grad_norm": 7.161188967637041,
+      "learning_rate": 7.252252252252253e-06,
+      "loss": 1.3914,
+      "step": 322
+    },
+    {
+      "epoch": 0.2185385656292287,
+      "grad_norm": 7.435624966331327,
+      "learning_rate": 7.2747747747747755e-06,
+      "loss": 1.4447,
+      "step": 323
+    },
+    {
+      "epoch": 0.21921515561569688,
+      "grad_norm": 6.450117600056855,
+      "learning_rate": 7.297297297297298e-06,
+      "loss": 1.353,
+      "step": 324
+    },
+    {
+      "epoch": 0.2198917456021651,
+      "grad_norm": 9.342882898605941,
+      "learning_rate": 7.31981981981982e-06,
+      "loss": 1.3382,
+      "step": 325
+    },
+    {
+      "epoch": 0.22056833558863329,
+      "grad_norm": 9.279944488980599,
+      "learning_rate": 7.342342342342343e-06,
+      "loss": 1.414,
+      "step": 326
+    },
+    {
+      "epoch": 0.22124492557510148,
+      "grad_norm": 7.099845515132716,
+      "learning_rate": 7.3648648648648655e-06,
+      "loss": 1.3692,
+      "step": 327
+    },
+    {
+      "epoch": 0.2219215155615697,
+      "grad_norm": 9.031425541989027,
+      "learning_rate": 7.387387387387388e-06,
+      "loss": 1.4173,
+      "step": 328
+    },
+    {
+      "epoch": 0.22259810554803788,
+      "grad_norm": 7.423615320613807,
+      "learning_rate": 7.40990990990991e-06,
+      "loss": 1.3511,
+      "step": 329
+    },
+    {
+      "epoch": 0.2232746955345061,
+      "grad_norm": 8.339528431625467,
+      "learning_rate": 7.4324324324324324e-06,
+      "loss": 1.3445,
+      "step": 330
+    },
+    {
+      "epoch": 0.2239512855209743,
+      "grad_norm": 8.991790302489601,
+      "learning_rate": 7.4549549549549564e-06,
+      "loss": 1.3122,
+      "step": 331
+    },
+    {
+      "epoch": 0.22462787550744248,
+      "grad_norm": 6.199790631454467,
+      "learning_rate": 7.477477477477479e-06,
+      "loss": 1.3809,
+      "step": 332
+    },
+    {
+      "epoch": 0.2253044654939107,
+      "grad_norm": 8.099748400756368,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.3352,
+      "step": 333
+    },
+    {
+      "epoch": 0.2259810554803789,
+      "grad_norm": 8.056852169659802,
+      "learning_rate": 7.5225225225225225e-06,
+      "loss": 1.3156,
+      "step": 334
+    },
+    {
+      "epoch": 0.22665764546684708,
+      "grad_norm": 6.686240008496426,
+      "learning_rate": 7.545045045045045e-06,
+      "loss": 1.3499,
+      "step": 335
+    },
+    {
+      "epoch": 0.2273342354533153,
+      "grad_norm": 7.9309282102043035,
+      "learning_rate": 7.567567567567569e-06,
+      "loss": 1.3324,
+      "step": 336
+    },
+    {
+      "epoch": 0.2280108254397835,
+      "grad_norm": 7.77658996644435,
+      "learning_rate": 7.590090090090091e-06,
+      "loss": 1.2565,
+      "step": 337
+    },
+    {
+      "epoch": 0.22868741542625168,
+      "grad_norm": 6.324303404965645,
+      "learning_rate": 7.612612612612613e-06,
+      "loss": 1.3374,
+      "step": 338
+    },
+    {
+      "epoch": 0.2293640054127199,
+      "grad_norm": 8.628015931255625,
+      "learning_rate": 7.635135135135135e-06,
+      "loss": 1.3013,
+      "step": 339
+    },
+    {
+      "epoch": 0.23004059539918809,
+      "grad_norm": 7.683595210549564,
+      "learning_rate": 7.657657657657658e-06,
+      "loss": 1.3019,
+      "step": 340
+    },
+    {
+      "epoch": 0.2307171853856563,
+      "grad_norm": 7.104540289089895,
+      "learning_rate": 7.680180180180181e-06,
+      "loss": 1.2834,
+      "step": 341
+    },
+    {
+      "epoch": 0.2313937753721245,
+      "grad_norm": 10.484418721383305,
+      "learning_rate": 7.702702702702704e-06,
+      "loss": 1.3192,
+      "step": 342
+    },
+    {
+      "epoch": 0.23207036535859268,
+      "grad_norm": 7.529555458611565,
+      "learning_rate": 7.725225225225226e-06,
+      "loss": 1.3345,
+      "step": 343
+    },
+    {
+      "epoch": 0.2327469553450609,
+      "grad_norm": 7.304557513705602,
+      "learning_rate": 7.747747747747749e-06,
+      "loss": 1.2718,
+      "step": 344
+    },
+    {
+      "epoch": 0.2334235453315291,
+      "grad_norm": 7.468281142178066,
+      "learning_rate": 7.77027027027027e-06,
+      "loss": 1.3459,
+      "step": 345
+    },
+    {
+      "epoch": 0.23410013531799728,
+      "grad_norm": 7.308215956657339,
+      "learning_rate": 7.792792792792793e-06,
+      "loss": 1.3549,
+      "step": 346
+    },
+    {
+      "epoch": 0.2347767253044655,
+      "grad_norm": 7.511453954818753,
+      "learning_rate": 7.815315315315317e-06,
+      "loss": 1.2729,
+      "step": 347
+    },
+    {
+      "epoch": 0.2354533152909337,
+      "grad_norm": 6.692314776874314,
+      "learning_rate": 7.837837837837838e-06,
+      "loss": 1.3146,
+      "step": 348
+    },
+    {
+      "epoch": 0.2361299052774019,
+      "grad_norm": 7.555862621620149,
+      "learning_rate": 7.860360360360361e-06,
+      "loss": 1.3208,
+      "step": 349
+    },
+    {
+      "epoch": 0.2368064952638701,
+      "grad_norm": 7.867176009962364,
+      "learning_rate": 7.882882882882884e-06,
+      "loss": 1.2837,
+      "step": 350
+    },
+    {
+      "epoch": 0.2374830852503383,
+      "grad_norm": 6.821961011231683,
+      "learning_rate": 7.905405405405406e-06,
+      "loss": 1.2689,
+      "step": 351
+    },
+    {
+      "epoch": 0.2381596752368065,
+      "grad_norm": 6.829813150156765,
+      "learning_rate": 7.927927927927929e-06,
+      "loss": 1.2085,
+      "step": 352
+    },
+    {
+      "epoch": 0.2388362652232747,
+      "grad_norm": 6.416445487360607,
+      "learning_rate": 7.95045045045045e-06,
+      "loss": 1.3142,
+      "step": 353
+    },
+    {
+      "epoch": 0.2395128552097429,
+      "grad_norm": 6.755985297167506,
+      "learning_rate": 7.972972972972974e-06,
+      "loss": 1.2437,
+      "step": 354
+    },
+    {
+      "epoch": 0.2401894451962111,
+      "grad_norm": 5.63750947182298,
+      "learning_rate": 7.995495495495497e-06,
+      "loss": 1.1827,
+      "step": 355
+    },
+    {
+      "epoch": 0.2408660351826793,
+      "grad_norm": 8.41215274775629,
+      "learning_rate": 8.018018018018018e-06,
+      "loss": 1.3291,
+      "step": 356
+    },
+    {
+      "epoch": 0.24154262516914748,
+      "grad_norm": 5.695299828620484,
+      "learning_rate": 8.040540540540541e-06,
+      "loss": 1.2339,
+      "step": 357
+    },
+    {
+      "epoch": 0.2422192151556157,
+      "grad_norm": 6.303783125664122,
+      "learning_rate": 8.063063063063063e-06,
+      "loss": 1.2393,
+      "step": 358
+    },
+    {
+      "epoch": 0.2428958051420839,
+      "grad_norm": 5.818013793977989,
+      "learning_rate": 8.085585585585586e-06,
+      "loss": 1.264,
+      "step": 359
+    },
+    {
+      "epoch": 0.2435723951285521,
+      "grad_norm": 6.00149757812294,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 1.25,
+      "step": 360
+    },
+    {
+      "epoch": 0.2442489851150203,
+      "grad_norm": 7.294950151985955,
+      "learning_rate": 8.130630630630632e-06,
+      "loss": 1.2442,
+      "step": 361
+    },
+    {
+      "epoch": 0.2449255751014885,
+      "grad_norm": 7.319385571720754,
+      "learning_rate": 8.153153153153154e-06,
+      "loss": 1.2517,
+      "step": 362
+    },
+    {
+      "epoch": 0.2456021650879567,
+      "grad_norm": 6.556333265442866,
+      "learning_rate": 8.175675675675677e-06,
+      "loss": 1.2393,
+      "step": 363
+    },
+    {
+      "epoch": 0.2462787550744249,
+      "grad_norm": 5.537101135836502,
+      "learning_rate": 8.198198198198198e-06,
+      "loss": 1.1824,
+      "step": 364
+    },
+    {
+      "epoch": 0.2469553450608931,
+      "grad_norm": 6.095023950167767,
+      "learning_rate": 8.220720720720721e-06,
+      "loss": 1.2078,
+      "step": 365
+    },
+    {
+      "epoch": 0.2476319350473613,
+      "grad_norm": 6.693002070569304,
+      "learning_rate": 8.243243243243245e-06,
+      "loss": 1.1672,
+      "step": 366
+    },
+    {
+      "epoch": 0.2483085250338295,
+      "grad_norm": 4.778897755508491,
+      "learning_rate": 8.265765765765766e-06,
+      "loss": 1.2018,
+      "step": 367
+    },
+    {
+      "epoch": 0.2489851150202977,
+      "grad_norm": 6.41741003265316,
+      "learning_rate": 8.288288288288289e-06,
+      "loss": 1.2732,
+      "step": 368
+    },
+    {
+      "epoch": 0.2496617050067659,
+      "grad_norm": 6.5817670328837545,
+      "learning_rate": 8.31081081081081e-06,
+      "loss": 1.138,
+      "step": 369
+    },
+    {
+      "epoch": 0.2503382949932341,
+      "grad_norm": 4.797250856682376,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.1449,
+      "step": 370
+    },
+    {
+      "epoch": 0.2510148849797023,
+      "grad_norm": 6.884122726840821,
+      "learning_rate": 8.355855855855857e-06,
+      "loss": 1.2155,
+      "step": 371
+    },
+    {
+      "epoch": 0.2516914749661705,
+      "grad_norm": 6.599119872104339,
+      "learning_rate": 8.378378378378378e-06,
+      "loss": 1.1852,
+      "step": 372
+    },
+    {
+      "epoch": 0.2523680649526387,
+      "grad_norm": 5.369629044197062,
+      "learning_rate": 8.400900900900901e-06,
+      "loss": 1.1606,
+      "step": 373
+    },
+    {
+      "epoch": 0.2530446549391069,
+      "grad_norm": 5.812549305075738,
+      "learning_rate": 8.423423423423423e-06,
+      "loss": 1.2094,
+      "step": 374
+    },
+    {
+      "epoch": 0.25372124492557513,
+      "grad_norm": 5.223599468834814,
+      "learning_rate": 8.445945945945948e-06,
+      "loss": 1.0854,
+      "step": 375
+    },
+    {
+      "epoch": 0.2543978349120433,
+      "grad_norm": 4.905683717210968,
+      "learning_rate": 8.46846846846847e-06,
+      "loss": 1.1488,
+      "step": 376
+    },
+    {
+      "epoch": 0.2550744248985115,
+      "grad_norm": 5.421234863754259,
+      "learning_rate": 8.490990990990992e-06,
+      "loss": 1.167,
+      "step": 377
+    },
+    {
+      "epoch": 0.2557510148849797,
+      "grad_norm": 6.8698555994818635,
+      "learning_rate": 8.513513513513514e-06,
+      "loss": 1.1931,
+      "step": 378
+    },
+    {
+      "epoch": 0.2564276048714479,
+      "grad_norm": 5.214335512005905,
+      "learning_rate": 8.536036036036037e-06,
+      "loss": 1.1036,
+      "step": 379
+    },
+    {
+      "epoch": 0.2571041948579161,
+      "grad_norm": 5.631060836540709,
+      "learning_rate": 8.55855855855856e-06,
+      "loss": 1.1909,
+      "step": 380
+    },
+    {
+      "epoch": 0.2577807848443843,
+      "grad_norm": 4.675672390186224,
+      "learning_rate": 8.581081081081082e-06,
+      "loss": 1.1228,
+      "step": 381
+    },
+    {
+      "epoch": 0.2584573748308525,
+      "grad_norm": 6.334523111748299,
+      "learning_rate": 8.603603603603605e-06,
+      "loss": 1.1194,
+      "step": 382
+    },
+    {
+      "epoch": 0.2591339648173207,
+      "grad_norm": 5.391808799329787,
+      "learning_rate": 8.626126126126126e-06,
+      "loss": 1.2031,
+      "step": 383
+    },
+    {
+      "epoch": 0.2598105548037889,
+      "grad_norm": 4.951535252198407,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 1.1924,
+      "step": 384
+    },
+    {
+      "epoch": 0.2604871447902571,
+      "grad_norm": 5.833225590221079,
+      "learning_rate": 8.671171171171172e-06,
+      "loss": 1.1505,
+      "step": 385
+    },
+    {
+      "epoch": 0.26116373477672533,
+      "grad_norm": 4.990875084577596,
+      "learning_rate": 8.693693693693694e-06,
+      "loss": 1.1134,
+      "step": 386
+    },
+    {
+      "epoch": 0.2618403247631935,
+      "grad_norm": 4.902817282647241,
+      "learning_rate": 8.716216216216217e-06,
+      "loss": 1.1128,
+      "step": 387
+    },
+    {
+      "epoch": 0.2625169147496617,
+      "grad_norm": 4.494411184327583,
+      "learning_rate": 8.738738738738739e-06,
+      "loss": 1.1129,
+      "step": 388
+    },
+    {
+      "epoch": 0.2631935047361299,
+      "grad_norm": 4.888423603959499,
+      "learning_rate": 8.761261261261262e-06,
+      "loss": 1.18,
+      "step": 389
+    },
+    {
+      "epoch": 0.2638700947225981,
+      "grad_norm": 5.832594841589209,
+      "learning_rate": 8.783783783783785e-06,
+      "loss": 1.1547,
+      "step": 390
+    },
+    {
+      "epoch": 0.2645466847090663,
+      "grad_norm": 5.235913407042474,
+      "learning_rate": 8.806306306306306e-06,
+      "loss": 1.1537,
+      "step": 391
+    },
+    {
+      "epoch": 0.2652232746955345,
+      "grad_norm": 4.785604512532903,
+      "learning_rate": 8.82882882882883e-06,
+      "loss": 1.1732,
+      "step": 392
+    },
+    {
+      "epoch": 0.2658998646820027,
+      "grad_norm": 5.313118558349346,
+      "learning_rate": 8.851351351351351e-06,
+      "loss": 1.0713,
+      "step": 393
+    },
+    {
+      "epoch": 0.2665764546684709,
+      "grad_norm": 4.827250818420794,
+      "learning_rate": 8.873873873873876e-06,
+      "loss": 1.0751,
+      "step": 394
+    },
+    {
+      "epoch": 0.2672530446549391,
+      "grad_norm": 4.874301936478751,
+      "learning_rate": 8.896396396396397e-06,
+      "loss": 1.0652,
+      "step": 395
+    },
+    {
+      "epoch": 0.2679296346414073,
+      "grad_norm": 4.8435780244924915,
+      "learning_rate": 8.91891891891892e-06,
+      "loss": 1.0936,
+      "step": 396
+    },
+    {
+      "epoch": 0.26860622462787553,
+      "grad_norm": 4.343161185327432,
+      "learning_rate": 8.941441441441442e-06,
+      "loss": 1.0229,
+      "step": 397
+    },
+    {
+      "epoch": 0.2692828146143437,
+      "grad_norm": 4.913175619810589,
+      "learning_rate": 8.963963963963965e-06,
+      "loss": 1.0129,
+      "step": 398
+    },
+    {
+      "epoch": 0.2699594046008119,
+      "grad_norm": 4.206729558415461,
+      "learning_rate": 8.986486486486488e-06,
+      "loss": 1.07,
+      "step": 399
+    },
+    {
+      "epoch": 0.2706359945872801,
+      "grad_norm": 4.5465248961906175,
+      "learning_rate": 9.00900900900901e-06,
+      "loss": 1.006,
+      "step": 400
+    },
+    {
+      "epoch": 0.2713125845737483,
+      "grad_norm": 4.64094117178924,
+      "learning_rate": 9.031531531531533e-06,
+      "loss": 1.0352,
+      "step": 401
+    },
+    {
+      "epoch": 0.2719891745602165,
+      "grad_norm": 4.941072707337412,
+      "learning_rate": 9.054054054054054e-06,
+      "loss": 1.1037,
+      "step": 402
+    },
+    {
+      "epoch": 0.27266576454668473,
+      "grad_norm": 5.532157984209573,
+      "learning_rate": 9.076576576576577e-06,
+      "loss": 1.0686,
+      "step": 403
+    },
+    {
+      "epoch": 0.2733423545331529,
+      "grad_norm": 4.324221301885782,
+      "learning_rate": 9.0990990990991e-06,
+      "loss": 1.0886,
+      "step": 404
+    },
+    {
+      "epoch": 0.2740189445196211,
+      "grad_norm": 4.672612441909186,
+      "learning_rate": 9.121621621621622e-06,
+      "loss": 1.0678,
+      "step": 405
+    },
+    {
+      "epoch": 0.2746955345060893,
+      "grad_norm": 3.8658548636324337,
+      "learning_rate": 9.144144144144145e-06,
+      "loss": 1.0214,
+      "step": 406
+    },
+    {
+      "epoch": 0.2753721244925575,
+      "grad_norm": 4.774743242339752,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 1.0256,
+      "step": 407
+    },
+    {
+      "epoch": 0.27604871447902574,
+      "grad_norm": 5.3993821745337565,
+      "learning_rate": 9.189189189189191e-06,
+      "loss": 1.0791,
+      "step": 408
+    },
+    {
+      "epoch": 0.2767253044654939,
+      "grad_norm": 5.057580982546297,
+      "learning_rate": 9.211711711711713e-06,
+      "loss": 1.0208,
+      "step": 409
+    },
+    {
+      "epoch": 0.2774018944519621,
+      "grad_norm": 4.924665267405325,
+      "learning_rate": 9.234234234234236e-06,
+      "loss": 1.0023,
+      "step": 410
+    },
+    {
+      "epoch": 0.2780784844384303,
+      "grad_norm": 4.232329566300859,
+      "learning_rate": 9.256756756756757e-06,
+      "loss": 1.0575,
+      "step": 411
+    },
+    {
+      "epoch": 0.2787550744248985,
+      "grad_norm": 5.239366528410246,
+      "learning_rate": 9.27927927927928e-06,
+      "loss": 1.0371,
+      "step": 412
+    },
+    {
+      "epoch": 0.2794316644113667,
+      "grad_norm": 4.783501098068623,
+      "learning_rate": 9.301801801801804e-06,
+      "loss": 1.0027,
+      "step": 413
+    },
+    {
+      "epoch": 0.28010825439783493,
+      "grad_norm": 3.9728448745576896,
+      "learning_rate": 9.324324324324325e-06,
+      "loss": 1.0467,
+      "step": 414
+    },
+    {
+      "epoch": 0.2807848443843031,
+      "grad_norm": 4.681528219035177,
+      "learning_rate": 9.346846846846848e-06,
+      "loss": 1.0132,
+      "step": 415
+    },
+    {
+      "epoch": 0.2814614343707713,
+      "grad_norm": 4.206918370997306,
+      "learning_rate": 9.36936936936937e-06,
+      "loss": 1.0553,
+      "step": 416
+    },
+    {
+      "epoch": 0.2821380243572395,
+      "grad_norm": 4.553415102778668,
+      "learning_rate": 9.391891891891893e-06,
+      "loss": 1.0007,
+      "step": 417
+    },
+    {
+      "epoch": 0.2828146143437077,
+      "grad_norm": 5.0528643415633905,
+      "learning_rate": 9.414414414414416e-06,
+      "loss": 1.0204,
+      "step": 418
+    },
+    {
+      "epoch": 0.28349120433017594,
+      "grad_norm": 4.3331179687879695,
+      "learning_rate": 9.436936936936937e-06,
+      "loss": 1.0374,
+      "step": 419
+    },
+    {
+      "epoch": 0.28416779431664413,
+      "grad_norm": 4.494054237607846,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 0.9342,
+      "step": 420
+    },
+    {
+      "epoch": 0.2848443843031123,
+      "grad_norm": 3.8300564753424964,
+      "learning_rate": 9.481981981981982e-06,
+      "loss": 1.0262,
+      "step": 421
+    },
+    {
+      "epoch": 0.2855209742895805,
+      "grad_norm": 3.9015045170124183,
+      "learning_rate": 9.504504504504505e-06,
+      "loss": 0.9976,
+      "step": 422
+    },
+    {
+      "epoch": 0.2861975642760487,
+      "grad_norm": 4.7139488491269725,
+      "learning_rate": 9.527027027027028e-06,
+      "loss": 0.9667,
+      "step": 423
+    },
+    {
+      "epoch": 0.2868741542625169,
+      "grad_norm": 3.9906815821303794,
+      "learning_rate": 9.54954954954955e-06,
+      "loss": 1.0557,
+      "step": 424
+    },
+    {
+      "epoch": 0.28755074424898514,
+      "grad_norm": 4.883974178298484,
+      "learning_rate": 9.572072072072073e-06,
+      "loss": 1.0379,
+      "step": 425
+    },
+    {
+      "epoch": 0.2882273342354533,
+      "grad_norm": 4.125238209478557,
+      "learning_rate": 9.594594594594594e-06,
+      "loss": 0.948,
+      "step": 426
+    },
+    {
+      "epoch": 0.2889039242219215,
+      "grad_norm": 4.067349144549682,
+      "learning_rate": 9.617117117117117e-06,
+      "loss": 0.9698,
+      "step": 427
+    },
+    {
+      "epoch": 0.2895805142083897,
+      "grad_norm": 4.479942419738399,
+      "learning_rate": 9.63963963963964e-06,
+      "loss": 0.9977,
+      "step": 428
+    },
+    {
+      "epoch": 0.2902571041948579,
+      "grad_norm": 3.849447458341483,
+      "learning_rate": 9.662162162162164e-06,
+      "loss": 0.989,
+      "step": 429
+    },
+    {
+      "epoch": 0.29093369418132614,
+      "grad_norm": 4.042946064561824,
+      "learning_rate": 9.684684684684685e-06,
+      "loss": 0.9692,
+      "step": 430
+    },
+    {
+      "epoch": 0.29161028416779433,
+      "grad_norm": 4.091640101512462,
+      "learning_rate": 9.707207207207208e-06,
+      "loss": 0.9442,
+      "step": 431
+    },
+    {
+      "epoch": 0.2922868741542625,
+      "grad_norm": 4.049936658082999,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 1.0213,
+      "step": 432
+    },
+    {
+      "epoch": 0.2929634641407307,
+      "grad_norm": 3.6882609568257134,
+      "learning_rate": 9.752252252252253e-06,
+      "loss": 0.9687,
+      "step": 433
+    },
+    {
+      "epoch": 0.2936400541271989,
+      "grad_norm": 3.9102149602955025,
+      "learning_rate": 9.774774774774776e-06,
+      "loss": 0.941,
+      "step": 434
+    },
+    {
+      "epoch": 0.2943166441136671,
+      "grad_norm": 4.223777918374561,
+      "learning_rate": 9.797297297297298e-06,
+      "loss": 0.9704,
+      "step": 435
+    },
+    {
+      "epoch": 0.29499323410013534,
+      "grad_norm": 3.8011374914119425,
+      "learning_rate": 9.81981981981982e-06,
+      "loss": 0.9702,
+      "step": 436
+    },
+    {
+      "epoch": 0.2956698240866035,
+      "grad_norm": 3.7303737377420894,
+      "learning_rate": 9.842342342342344e-06,
+      "loss": 0.9378,
+      "step": 437
+    },
+    {
+      "epoch": 0.2963464140730717,
+      "grad_norm": 3.521072752470763,
+      "learning_rate": 9.864864864864865e-06,
+      "loss": 0.9507,
+      "step": 438
+    },
+    {
+      "epoch": 0.2970230040595399,
+      "grad_norm": 4.316004502684595,
+      "learning_rate": 9.887387387387388e-06,
+      "loss": 0.9734,
+      "step": 439
+    },
+    {
+      "epoch": 0.2976995940460081,
+      "grad_norm": 3.8335521195044318,
+      "learning_rate": 9.90990990990991e-06,
+      "loss": 0.9253,
+      "step": 440
+    },
+    {
+      "epoch": 0.29837618403247634,
+      "grad_norm": 3.665705979258958,
+      "learning_rate": 9.932432432432433e-06,
+      "loss": 0.9713,
+      "step": 441
+    },
+    {
+      "epoch": 0.29905277401894453,
+      "grad_norm": 4.334698014272042,
+      "learning_rate": 9.954954954954956e-06,
+      "loss": 0.9083,
+      "step": 442
+    },
+    {
+      "epoch": 0.2997293640054127,
+      "grad_norm": 3.3447072281334513,
+      "learning_rate": 9.97747747747748e-06,
+      "loss": 0.9034,
+      "step": 443
+    },
+    {
+      "epoch": 0.3004059539918809,
+      "grad_norm": 4.1589594059536985,
+      "learning_rate": 1e-05,
+      "loss": 0.9321,
+      "step": 444
+    },
+    {
+      "epoch": 0.3010825439783491,
+      "grad_norm": 4.1209097446031935,
+      "learning_rate": 9.999998450134754e-06,
+      "loss": 0.9187,
+      "step": 445
+    },
+    {
+      "epoch": 0.3017591339648173,
+      "grad_norm": 3.6342562123201354,
+      "learning_rate": 9.999993800539971e-06,
+      "loss": 0.9433,
+      "step": 446
+    },
+    {
+      "epoch": 0.30243572395128554,
+      "grad_norm": 3.8838265923669137,
+      "learning_rate": 9.999986051218538e-06,
+      "loss": 0.9375,
+      "step": 447
+    },
+    {
+      "epoch": 0.30311231393775373,
+      "grad_norm": 3.8982724762364014,
+      "learning_rate": 9.999975202175256e-06,
+      "loss": 0.9194,
+      "step": 448
+    },
+    {
+      "epoch": 0.3037889039242219,
+      "grad_norm": 4.0073538315187225,
+      "learning_rate": 9.999961253416853e-06,
+      "loss": 0.9047,
+      "step": 449
+    },
+    {
+      "epoch": 0.3044654939106901,
+      "grad_norm": 3.83724752407595,
+      "learning_rate": 9.999944204951974e-06,
+      "loss": 0.9214,
+      "step": 450
+    },
+    {
+      "epoch": 0.3051420838971583,
+      "grad_norm": 3.5702265127516117,
+      "learning_rate": 9.999924056791192e-06,
+      "loss": 0.8795,
+      "step": 451
+    },
+    {
+      "epoch": 0.30581867388362655,
+      "grad_norm": 3.492179005663864,
+      "learning_rate": 9.999900808946996e-06,
+      "loss": 0.9131,
+      "step": 452
+    },
+    {
+      "epoch": 0.30649526387009474,
+      "grad_norm": 3.1745948209546837,
+      "learning_rate": 9.999874461433796e-06,
+      "loss": 0.9165,
+      "step": 453
+    },
+    {
+      "epoch": 0.3071718538565629,
+      "grad_norm": 3.552119904729672,
+      "learning_rate": 9.999845014267928e-06,
+      "loss": 0.9205,
+      "step": 454
+    },
+    {
+      "epoch": 0.3078484438430311,
+      "grad_norm": 3.095271394681658,
+      "learning_rate": 9.99981246746765e-06,
+      "loss": 0.8814,
+      "step": 455
+    },
+    {
+      "epoch": 0.3085250338294993,
+      "grad_norm": 3.5904697324098733,
+      "learning_rate": 9.999776821053134e-06,
+      "loss": 0.9115,
+      "step": 456
+    },
+    {
+      "epoch": 0.3092016238159675,
+      "grad_norm": 3.696586268891529,
+      "learning_rate": 9.999738075046483e-06,
+      "loss": 0.9036,
+      "step": 457
+    },
+    {
+      "epoch": 0.30987821380243574,
+      "grad_norm": 3.666514530641993,
+      "learning_rate": 9.999696229471716e-06,
+      "loss": 0.9171,
+      "step": 458
+    },
+    {
+      "epoch": 0.31055480378890393,
+      "grad_norm": 3.5284682583414426,
+      "learning_rate": 9.999651284354774e-06,
+      "loss": 0.8933,
+      "step": 459
+    },
+    {
+      "epoch": 0.3112313937753721,
+      "grad_norm": 3.55289773993609,
+      "learning_rate": 9.999603239723524e-06,
+      "loss": 0.872,
+      "step": 460
+    },
+    {
+      "epoch": 0.3119079837618403,
+      "grad_norm": 3.2368814708165794,
+      "learning_rate": 9.999552095607748e-06,
+      "loss": 0.8605,
+      "step": 461
+    },
+    {
+      "epoch": 0.3125845737483085,
+      "grad_norm": 3.0334660120605452,
+      "learning_rate": 9.999497852039152e-06,
+      "loss": 0.9141,
+      "step": 462
+    },
+    {
+      "epoch": 0.31326116373477675,
+      "grad_norm": 3.314329217403211,
+      "learning_rate": 9.999440509051367e-06,
+      "loss": 0.8968,
+      "step": 463
+    },
+    {
+      "epoch": 0.31393775372124494,
+      "grad_norm": 3.570487557645587,
+      "learning_rate": 9.999380066679943e-06,
+      "loss": 0.9137,
+      "step": 464
+    },
+    {
+      "epoch": 0.31461434370771313,
+      "grad_norm": 3.2744333929814706,
+      "learning_rate": 9.999316524962347e-06,
+      "loss": 0.9098,
+      "step": 465
+    },
+    {
+      "epoch": 0.3152909336941813,
+      "grad_norm": 3.003532560029087,
+      "learning_rate": 9.999249883937971e-06,
+      "loss": 0.8658,
+      "step": 466
+    },
+    {
+      "epoch": 0.3159675236806495,
+      "grad_norm": 3.1147658376753915,
+      "learning_rate": 9.999180143648136e-06,
+      "loss": 0.8807,
+      "step": 467
+    },
+    {
+      "epoch": 0.3166441136671177,
+      "grad_norm": 3.2904201115134244,
+      "learning_rate": 9.999107304136068e-06,
+      "loss": 0.918,
+      "step": 468
+    },
+    {
+      "epoch": 0.31732070365358594,
+      "grad_norm": 3.346893818987595,
+      "learning_rate": 9.999031365446932e-06,
+      "loss": 0.8351,
+      "step": 469
+    },
+    {
+      "epoch": 0.31799729364005414,
+      "grad_norm": 3.088776815713552,
+      "learning_rate": 9.9989523276278e-06,
+      "loss": 0.8829,
+      "step": 470
+    },
+    {
+      "epoch": 0.3186738836265223,
+      "grad_norm": 4.0936798977412625,
+      "learning_rate": 9.998870190727674e-06,
+      "loss": 0.8432,
+      "step": 471
+    },
+    {
+      "epoch": 0.3193504736129905,
+      "grad_norm": 3.1764697217902946,
+      "learning_rate": 9.998784954797474e-06,
+      "loss": 0.8613,
+      "step": 472
+    },
+    {
+      "epoch": 0.3200270635994587,
+      "grad_norm": 3.193647148522581,
+      "learning_rate": 9.99869661989004e-06,
+      "loss": 0.8693,
+      "step": 473
+    },
+    {
+      "epoch": 0.32070365358592695,
+      "grad_norm": 3.4120440580585303,
+      "learning_rate": 9.998605186060138e-06,
+      "loss": 0.8244,
+      "step": 474
+    },
+    {
+      "epoch": 0.32138024357239514,
+      "grad_norm": 3.0258544107304832,
+      "learning_rate": 9.998510653364449e-06,
+      "loss": 0.8517,
+      "step": 475
+    },
+    {
+      "epoch": 0.32205683355886333,
+      "grad_norm": 3.1034878568318263,
+      "learning_rate": 9.998413021861581e-06,
+      "loss": 0.8578,
+      "step": 476
+    },
+    {
+      "epoch": 0.3227334235453315,
+      "grad_norm": 3.250075725402367,
+      "learning_rate": 9.998312291612056e-06,
+      "loss": 0.877,
+      "step": 477
+    },
+    {
+      "epoch": 0.3234100135317997,
+      "grad_norm": 2.7733424457408034,
+      "learning_rate": 9.998208462678328e-06,
+      "loss": 0.8353,
+      "step": 478
+    },
+    {
+      "epoch": 0.32408660351826796,
+      "grad_norm": 3.0189955639643955,
+      "learning_rate": 9.998101535124758e-06,
+      "loss": 0.8852,
+      "step": 479
+    },
+    {
+      "epoch": 0.32476319350473615,
+      "grad_norm": 2.8949415050560456,
+      "learning_rate": 9.99799150901764e-06,
+      "loss": 0.8907,
+      "step": 480
+    },
+    {
+      "epoch": 0.32543978349120434,
+      "grad_norm": 3.0534310639406885,
+      "learning_rate": 9.997878384425183e-06,
+      "loss": 0.8456,
+      "step": 481
+    },
+    {
+      "epoch": 0.3261163734776725,
+      "grad_norm": 2.855996354930219,
+      "learning_rate": 9.997762161417517e-06,
+      "loss": 0.8338,
+      "step": 482
+    },
+    {
+      "epoch": 0.3267929634641407,
+      "grad_norm": 2.890677997136448,
+      "learning_rate": 9.997642840066696e-06,
+      "loss": 0.8582,
+      "step": 483
+    },
+    {
+      "epoch": 0.3274695534506089,
+      "grad_norm": 2.8955609614653155,
+      "learning_rate": 9.997520420446694e-06,
+      "loss": 0.8533,
+      "step": 484
+    },
+    {
+      "epoch": 0.32814614343707715,
+      "grad_norm": 3.094942114313454,
+      "learning_rate": 9.9973949026334e-06,
+      "loss": 0.824,
+      "step": 485
+    },
+    {
+      "epoch": 0.32882273342354534,
+      "grad_norm": 3.063619152732945,
+      "learning_rate": 9.99726628670463e-06,
+      "loss": 0.8555,
+      "step": 486
+    },
+    {
+      "epoch": 0.32949932341001353,
+      "grad_norm": 2.819488565448912,
+      "learning_rate": 9.997134572740122e-06,
+      "loss": 0.8464,
+      "step": 487
+    },
+    {
+      "epoch": 0.3301759133964817,
+      "grad_norm": 2.5548684405877973,
+      "learning_rate": 9.996999760821529e-06,
+      "loss": 0.7959,
+      "step": 488
+    },
+    {
+      "epoch": 0.3308525033829499,
+      "grad_norm": 2.8786002140608313,
+      "learning_rate": 9.996861851032426e-06,
+      "loss": 0.8573,
+      "step": 489
+    },
+    {
+      "epoch": 0.33152909336941816,
+      "grad_norm": 2.807515168968411,
+      "learning_rate": 9.996720843458312e-06,
+      "loss": 0.8447,
+      "step": 490
+    },
+    {
+      "epoch": 0.33220568335588635,
+      "grad_norm": 2.937361095247826,
+      "learning_rate": 9.996576738186602e-06,
+      "loss": 0.8275,
+      "step": 491
+    },
+    {
+      "epoch": 0.33288227334235454,
+      "grad_norm": 2.7581361587891218,
+      "learning_rate": 9.996429535306638e-06,
+      "loss": 0.8625,
+      "step": 492
+    },
+    {
+      "epoch": 0.33355886332882273,
+      "grad_norm": 3.060203502125078,
+      "learning_rate": 9.996279234909672e-06,
+      "loss": 0.8504,
+      "step": 493
+    },
+    {
+      "epoch": 0.3342354533152909,
+      "grad_norm": 2.9680097813210553,
+      "learning_rate": 9.996125837088883e-06,
+      "loss": 0.846,
+      "step": 494
+    },
+    {
+      "epoch": 0.3349120433017591,
+      "grad_norm": 2.6959945304869635,
+      "learning_rate": 9.995969341939373e-06,
+      "loss": 0.8065,
+      "step": 495
+    },
+    {
+      "epoch": 0.33558863328822736,
+      "grad_norm": 3.177115967659662,
+      "learning_rate": 9.995809749558159e-06,
+      "loss": 0.8275,
+      "step": 496
+    },
+    {
+      "epoch": 0.33626522327469555,
+      "grad_norm": 2.944430539362135,
+      "learning_rate": 9.995647060044178e-06,
+      "loss": 0.8373,
+      "step": 497
+    },
+    {
+      "epoch": 0.33694181326116374,
+      "grad_norm": 2.9290586110244305,
+      "learning_rate": 9.995481273498291e-06,
+      "loss": 0.7995,
+      "step": 498
+    },
+    {
+      "epoch": 0.3376184032476319,
+      "grad_norm": 3.206371405004454,
+      "learning_rate": 9.995312390023275e-06,
+      "loss": 0.8409,
+      "step": 499
+    },
+    {
+      "epoch": 0.3382949932341001,
+      "grad_norm": 2.7223624769282857,
+      "learning_rate": 9.995140409723831e-06,
+      "loss": 0.8337,
+      "step": 500
+    },
+    {
+      "epoch": 0.33897158322056836,
+      "grad_norm": 2.7354940352519708,
+      "learning_rate": 9.994965332706574e-06,
+      "loss": 0.8128,
+      "step": 501
+    },
+    {
+      "epoch": 0.33964817320703655,
+      "grad_norm": 2.742152403931936,
+      "learning_rate": 9.994787159080046e-06,
+      "loss": 0.8372,
+      "step": 502
+    },
+    {
+      "epoch": 0.34032476319350474,
+      "grad_norm": 2.407710662991995,
+      "learning_rate": 9.994605888954701e-06,
+      "loss": 0.7752,
+      "step": 503
+    },
+    {
+      "epoch": 0.34100135317997293,
+      "grad_norm": 2.648539541200312,
+      "learning_rate": 9.99442152244292e-06,
+      "loss": 0.8351,
+      "step": 504
+    },
+    {
+      "epoch": 0.3416779431664411,
+      "grad_norm": 2.7987393621847256,
+      "learning_rate": 9.994234059658998e-06,
+      "loss": 0.8253,
+      "step": 505
+    },
+    {
+      "epoch": 0.3423545331529093,
+      "grad_norm": 2.6611235079805544,
+      "learning_rate": 9.994043500719155e-06,
+      "loss": 0.7732,
+      "step": 506
+    },
+    {
+      "epoch": 0.34303112313937756,
+      "grad_norm": 2.824414736639999,
+      "learning_rate": 9.993849845741525e-06,
+      "loss": 0.8232,
+      "step": 507
+    },
+    {
+      "epoch": 0.34370771312584575,
+      "grad_norm": 3.278643791346545,
+      "learning_rate": 9.993653094846162e-06,
+      "loss": 0.833,
+      "step": 508
+    },
+    {
+      "epoch": 0.34438430311231394,
+      "grad_norm": 2.6755857541883263,
+      "learning_rate": 9.993453248155044e-06,
+      "loss": 0.8138,
+      "step": 509
+    },
+    {
+      "epoch": 0.34506089309878213,
+      "grad_norm": 3.534895888050554,
+      "learning_rate": 9.993250305792067e-06,
+      "loss": 0.8327,
+      "step": 510
+    },
+    {
+      "epoch": 0.3457374830852503,
+      "grad_norm": 2.544304824354249,
+      "learning_rate": 9.993044267883039e-06,
+      "loss": 0.7876,
+      "step": 511
+    },
+    {
+      "epoch": 0.34641407307171856,
+      "grad_norm": 2.9700680528467163,
+      "learning_rate": 9.992835134555694e-06,
+      "loss": 0.8245,
+      "step": 512
+    },
+    {
+      "epoch": 0.34709066305818675,
+      "grad_norm": 2.5626117802153,
+      "learning_rate": 9.992622905939686e-06,
+      "loss": 0.8026,
+      "step": 513
+    },
+    {
+      "epoch": 0.34776725304465494,
+      "grad_norm": 2.687982257062625,
+      "learning_rate": 9.992407582166582e-06,
+      "loss": 0.817,
+      "step": 514
+    },
+    {
+      "epoch": 0.34844384303112313,
+      "grad_norm": 2.682185575359109,
+      "learning_rate": 9.992189163369873e-06,
+      "loss": 0.8024,
+      "step": 515
+    },
+    {
+      "epoch": 0.3491204330175913,
+      "grad_norm": 2.621193747220266,
+      "learning_rate": 9.991967649684967e-06,
+      "loss": 0.7859,
+      "step": 516
+    },
+    {
+      "epoch": 0.3497970230040595,
+      "grad_norm": 2.6783740060221803,
+      "learning_rate": 9.99174304124919e-06,
+      "loss": 0.7966,
+      "step": 517
+    },
+    {
+      "epoch": 0.35047361299052776,
+      "grad_norm": 2.638660466868078,
+      "learning_rate": 9.991515338201787e-06,
+      "loss": 0.8155,
+      "step": 518
+    },
+    {
+      "epoch": 0.35115020297699595,
+      "grad_norm": 2.6796352812176827,
+      "learning_rate": 9.991284540683922e-06,
+      "loss": 0.8187,
+      "step": 519
+    },
+    {
+      "epoch": 0.35182679296346414,
+      "grad_norm": 2.4636478517253098,
+      "learning_rate": 9.991050648838676e-06,
+      "loss": 0.79,
+      "step": 520
+    },
+    {
+      "epoch": 0.35250338294993233,
+      "grad_norm": 2.5132501947626134,
+      "learning_rate": 9.990813662811052e-06,
+      "loss": 0.8107,
+      "step": 521
+    },
+    {
+      "epoch": 0.3531799729364005,
+      "grad_norm": 2.3102149207906657,
+      "learning_rate": 9.990573582747965e-06,
+      "loss": 0.7221,
+      "step": 522
+    },
+    {
+      "epoch": 0.35385656292286877,
+      "grad_norm": 2.8388086625182076,
+      "learning_rate": 9.990330408798255e-06,
+      "loss": 0.8176,
+      "step": 523
+    },
+    {
+      "epoch": 0.35453315290933696,
+      "grad_norm": 2.547424702471631,
+      "learning_rate": 9.990084141112674e-06,
+      "loss": 0.7757,
+      "step": 524
+    },
+    {
+      "epoch": 0.35520974289580515,
+      "grad_norm": 2.459038671782621,
+      "learning_rate": 9.989834779843895e-06,
+      "loss": 0.7571,
+      "step": 525
+    },
+    {
+      "epoch": 0.35588633288227334,
+      "grad_norm": 2.343072915349538,
+      "learning_rate": 9.989582325146511e-06,
+      "loss": 0.7588,
+      "step": 526
+    },
+    {
+      "epoch": 0.3565629228687415,
+      "grad_norm": 2.6318457433995643,
+      "learning_rate": 9.98932677717703e-06,
+      "loss": 0.7863,
+      "step": 527
+    },
+    {
+      "epoch": 0.3572395128552097,
+      "grad_norm": 2.841829010674145,
+      "learning_rate": 9.989068136093873e-06,
+      "loss": 0.8111,
+      "step": 528
+    },
+    {
+      "epoch": 0.35791610284167796,
+      "grad_norm": 2.798257035023864,
+      "learning_rate": 9.98880640205739e-06,
+      "loss": 0.8038,
+      "step": 529
+    },
+    {
+      "epoch": 0.35859269282814615,
+      "grad_norm": 2.7565105246706367,
+      "learning_rate": 9.988541575229837e-06,
+      "loss": 0.7761,
+      "step": 530
+    },
+    {
+      "epoch": 0.35926928281461434,
+      "grad_norm": 2.6356624492568357,
+      "learning_rate": 9.988273655775398e-06,
+      "loss": 0.8054,
+      "step": 531
+    },
+    {
+      "epoch": 0.35994587280108253,
+      "grad_norm": 2.6569910293091574,
+      "learning_rate": 9.988002643860162e-06,
+      "loss": 0.7642,
+      "step": 532
+    },
+    {
+      "epoch": 0.3606224627875507,
+      "grad_norm": 2.779551959212093,
+      "learning_rate": 9.987728539652145e-06,
+      "loss": 0.7719,
+      "step": 533
+    },
+    {
+      "epoch": 0.36129905277401897,
+      "grad_norm": 2.821073101690452,
+      "learning_rate": 9.98745134332128e-06,
+      "loss": 0.723,
+      "step": 534
+    },
+    {
+      "epoch": 0.36197564276048716,
+      "grad_norm": 2.665438624418873,
+      "learning_rate": 9.987171055039409e-06,
+      "loss": 0.8325,
+      "step": 535
+    },
+    {
+      "epoch": 0.36265223274695535,
+      "grad_norm": 2.602468827659332,
+      "learning_rate": 9.986887674980297e-06,
+      "loss": 0.7768,
+      "step": 536
+    },
+    {
+      "epoch": 0.36332882273342354,
+      "grad_norm": 2.5104464519095866,
+      "learning_rate": 9.986601203319623e-06,
+      "loss": 0.735,
+      "step": 537
+    },
+    {
+      "epoch": 0.36400541271989173,
+      "grad_norm": 2.4650931692682723,
+      "learning_rate": 9.986311640234988e-06,
+      "loss": 0.7359,
+      "step": 538
+    },
+    {
+      "epoch": 0.3646820027063599,
+      "grad_norm": 2.266836893273909,
+      "learning_rate": 9.986018985905901e-06,
+      "loss": 0.7646,
+      "step": 539
+    },
+    {
+      "epoch": 0.36535859269282817,
+      "grad_norm": 2.3159717821800627,
+      "learning_rate": 9.985723240513795e-06,
+      "loss": 0.7479,
+      "step": 540
+    },
+    {
+      "epoch": 0.36603518267929636,
+      "grad_norm": 2.3267215054412,
+      "learning_rate": 9.985424404242015e-06,
+      "loss": 0.7429,
+      "step": 541
+    },
+    {
+      "epoch": 0.36671177266576455,
+      "grad_norm": 2.445693725429499,
+      "learning_rate": 9.985122477275824e-06,
+      "loss": 0.7349,
+      "step": 542
+    },
+    {
+      "epoch": 0.36738836265223274,
+      "grad_norm": 2.4198599886280623,
+      "learning_rate": 9.9848174598024e-06,
+      "loss": 0.7323,
+      "step": 543
+    },
+    {
+      "epoch": 0.3680649526387009,
+      "grad_norm": 2.2812273337003917,
+      "learning_rate": 9.984509352010839e-06,
+      "loss": 0.6918,
+      "step": 544
+    },
+    {
+      "epoch": 0.36874154262516917,
+      "grad_norm": 2.2654499896491487,
+      "learning_rate": 9.984198154092147e-06,
+      "loss": 0.7169,
+      "step": 545
+    },
+    {
+      "epoch": 0.36941813261163736,
+      "grad_norm": 2.2119471183099235,
+      "learning_rate": 9.983883866239253e-06,
+      "loss": 0.7525,
+      "step": 546
+    },
+    {
+      "epoch": 0.37009472259810555,
+      "grad_norm": 2.239396967454031,
+      "learning_rate": 9.983566488647e-06,
+      "loss": 0.7633,
+      "step": 547
+    },
+    {
+      "epoch": 0.37077131258457374,
+      "grad_norm": 2.3671195766363646,
+      "learning_rate": 9.98324602151214e-06,
+      "loss": 0.7287,
+      "step": 548
+    },
+    {
+      "epoch": 0.37144790257104193,
+      "grad_norm": 2.3516713935587545,
+      "learning_rate": 9.98292246503335e-06,
+      "loss": 0.7845,
+      "step": 549
+    },
+    {
+      "epoch": 0.3721244925575101,
+      "grad_norm": 2.5614066335000887,
+      "learning_rate": 9.982595819411216e-06,
+      "loss": 0.772,
+      "step": 550
+    },
+    {
+      "epoch": 0.37280108254397837,
+      "grad_norm": 2.268730726420046,
+      "learning_rate": 9.98226608484824e-06,
+      "loss": 0.7265,
+      "step": 551
+    },
+    {
+      "epoch": 0.37347767253044656,
+      "grad_norm": 2.4673866865308485,
+      "learning_rate": 9.981933261548841e-06,
+      "loss": 0.7357,
+      "step": 552
+    },
+    {
+      "epoch": 0.37415426251691475,
+      "grad_norm": 2.296525880694683,
+      "learning_rate": 9.981597349719351e-06,
+      "loss": 0.7124,
+      "step": 553
+    },
+    {
+      "epoch": 0.37483085250338294,
+      "grad_norm": 2.265625643278614,
+      "learning_rate": 9.981258349568018e-06,
+      "loss": 0.7373,
+      "step": 554
+    },
+    {
+      "epoch": 0.37550744248985113,
+      "grad_norm": 2.2696703683486157,
+      "learning_rate": 9.980916261305002e-06,
+      "loss": 0.7247,
+      "step": 555
+    },
+    {
+      "epoch": 0.3761840324763194,
+      "grad_norm": 2.177261495035988,
+      "learning_rate": 9.980571085142381e-06,
+      "loss": 0.747,
+      "step": 556
+    },
+    {
+      "epoch": 0.37686062246278756,
+      "grad_norm": 2.362176160288204,
+      "learning_rate": 9.980222821294143e-06,
+      "loss": 0.7246,
+      "step": 557
+    },
+    {
+      "epoch": 0.37753721244925575,
+      "grad_norm": 2.1022685100770646,
+      "learning_rate": 9.979871469976197e-06,
+      "loss": 0.6982,
+      "step": 558
+    },
+    {
+      "epoch": 0.37821380243572394,
+      "grad_norm": 2.236688754096613,
+      "learning_rate": 9.979517031406357e-06,
+      "loss": 0.7417,
+      "step": 559
+    },
+    {
+      "epoch": 0.37889039242219213,
+      "grad_norm": 2.1466742469590834,
+      "learning_rate": 9.97915950580436e-06,
+      "loss": 0.7377,
+      "step": 560
+    },
+    {
+      "epoch": 0.3795669824086603,
+      "grad_norm": 2.182650476403489,
+      "learning_rate": 9.97879889339185e-06,
+      "loss": 0.7209,
+      "step": 561
+    },
+    {
+      "epoch": 0.38024357239512857,
+      "grad_norm": 2.1526694838589755,
+      "learning_rate": 9.97843519439239e-06,
+      "loss": 0.7522,
+      "step": 562
+    },
+    {
+      "epoch": 0.38092016238159676,
+      "grad_norm": 2.1371520397539943,
+      "learning_rate": 9.978068409031449e-06,
+      "loss": 0.7189,
+      "step": 563
+    },
+    {
+      "epoch": 0.38159675236806495,
+      "grad_norm": 2.102484057128259,
+      "learning_rate": 9.97769853753642e-06,
+      "loss": 0.7035,
+      "step": 564
+    },
+    {
+      "epoch": 0.38227334235453314,
+      "grad_norm": 2.1862049962798777,
+      "learning_rate": 9.977325580136598e-06,
+      "loss": 0.7131,
+      "step": 565
+    },
+    {
+      "epoch": 0.38294993234100133,
+      "grad_norm": 2.366449337575754,
+      "learning_rate": 9.9769495370632e-06,
+      "loss": 0.7405,
+      "step": 566
+    },
+    {
+      "epoch": 0.3836265223274696,
+      "grad_norm": 2.088294738417784,
+      "learning_rate": 9.97657040854935e-06,
+      "loss": 0.6992,
+      "step": 567
+    },
+    {
+      "epoch": 0.38430311231393777,
+      "grad_norm": 2.309847902881155,
+      "learning_rate": 9.976188194830092e-06,
+      "loss": 0.7349,
+      "step": 568
+    },
+    {
+      "epoch": 0.38497970230040596,
+      "grad_norm": 2.1445079732726624,
+      "learning_rate": 9.975802896142373e-06,
+      "loss": 0.7408,
+      "step": 569
+    },
+    {
+      "epoch": 0.38565629228687415,
+      "grad_norm": 2.239660655657111,
+      "learning_rate": 9.975414512725058e-06,
+      "loss": 0.7203,
+      "step": 570
+    },
+    {
+      "epoch": 0.38633288227334234,
+      "grad_norm": 2.1721580645053273,
+      "learning_rate": 9.975023044818925e-06,
+      "loss": 0.7365,
+      "step": 571
+    },
+    {
+      "epoch": 0.3870094722598105,
+      "grad_norm": 2.1101165945964455,
+      "learning_rate": 9.974628492666664e-06,
+      "loss": 0.6948,
+      "step": 572
+    },
+    {
+      "epoch": 0.3876860622462788,
+      "grad_norm": 2.412564136594261,
+      "learning_rate": 9.974230856512874e-06,
+      "loss": 0.7622,
+      "step": 573
+    },
+    {
+      "epoch": 0.38836265223274696,
+      "grad_norm": 2.2856492277962097,
+      "learning_rate": 9.973830136604068e-06,
+      "loss": 0.728,
+      "step": 574
+    },
+    {
+      "epoch": 0.38903924221921515,
+      "grad_norm": 2.180935064812689,
+      "learning_rate": 9.973426333188673e-06,
+      "loss": 0.7116,
+      "step": 575
+    },
+    {
+      "epoch": 0.38971583220568334,
+      "grad_norm": 2.276771558917313,
+      "learning_rate": 9.973019446517023e-06,
+      "loss": 0.7363,
+      "step": 576
+    },
+    {
+      "epoch": 0.39039242219215153,
+      "grad_norm": 2.0804479250377907,
+      "learning_rate": 9.972609476841368e-06,
+      "loss": 0.729,
+      "step": 577
+    },
+    {
+      "epoch": 0.3910690121786198,
+      "grad_norm": 2.293959982563185,
+      "learning_rate": 9.972196424415865e-06,
+      "loss": 0.7168,
+      "step": 578
+    },
+    {
+      "epoch": 0.39174560216508797,
+      "grad_norm": 2.019436821032936,
+      "learning_rate": 9.971780289496585e-06,
+      "loss": 0.715,
+      "step": 579
+    },
+    {
+      "epoch": 0.39242219215155616,
+      "grad_norm": 2.2784237326655625,
+      "learning_rate": 9.971361072341509e-06,
+      "loss": 0.7036,
+      "step": 580
+    },
+    {
+      "epoch": 0.39309878213802435,
+      "grad_norm": 2.059377297470316,
+      "learning_rate": 9.97093877321053e-06,
+      "loss": 0.721,
+      "step": 581
+    },
+    {
+      "epoch": 0.39377537212449254,
+      "grad_norm": 2.139362525141615,
+      "learning_rate": 9.970513392365449e-06,
+      "loss": 0.715,
+      "step": 582
+    },
+    {
+      "epoch": 0.3944519621109608,
+      "grad_norm": 2.0176705974240003,
+      "learning_rate": 9.970084930069982e-06,
+      "loss": 0.7114,
+      "step": 583
+    },
+    {
+      "epoch": 0.395128552097429,
+      "grad_norm": 2.0092807332993137,
+      "learning_rate": 9.969653386589749e-06,
+      "loss": 0.6947,
+      "step": 584
+    },
+    {
+      "epoch": 0.39580514208389717,
+      "grad_norm": 2.1129018989942683,
+      "learning_rate": 9.969218762192286e-06,
+      "loss": 0.7647,
+      "step": 585
+    },
+    {
+      "epoch": 0.39648173207036536,
+      "grad_norm": 2.1316606162902563,
+      "learning_rate": 9.968781057147036e-06,
+      "loss": 0.7023,
+      "step": 586
+    },
+    {
+      "epoch": 0.39715832205683355,
+      "grad_norm": 2.232107359256458,
+      "learning_rate": 9.968340271725352e-06,
+      "loss": 0.7147,
+      "step": 587
+    },
+    {
+      "epoch": 0.39783491204330174,
+      "grad_norm": 2.0234712969002477,
+      "learning_rate": 9.967896406200498e-06,
+      "loss": 0.7091,
+      "step": 588
+    },
+    {
+      "epoch": 0.39851150202977,
+      "grad_norm": 2.141612227613064,
+      "learning_rate": 9.967449460847648e-06,
+      "loss": 0.7107,
+      "step": 589
+    },
+    {
+      "epoch": 0.39918809201623817,
+      "grad_norm": 2.0552325338452735,
+      "learning_rate": 9.966999435943882e-06,
+      "loss": 0.682,
+      "step": 590
+    },
+    {
+      "epoch": 0.39986468200270636,
+      "grad_norm": 2.265846296289829,
+      "learning_rate": 9.966546331768192e-06,
+      "loss": 0.7215,
+      "step": 591
+    },
+    {
+      "epoch": 0.40054127198917455,
+      "grad_norm": 2.2436402593988833,
+      "learning_rate": 9.966090148601477e-06,
+      "loss": 0.6817,
+      "step": 592
+    },
+    {
+      "epoch": 0.40121786197564274,
+      "grad_norm": 2.179910678350655,
+      "learning_rate": 9.965630886726548e-06,
+      "loss": 0.7052,
+      "step": 593
+    },
+    {
+      "epoch": 0.401894451962111,
+      "grad_norm": 2.48734340383341,
+      "learning_rate": 9.965168546428122e-06,
+      "loss": 0.714,
+      "step": 594
+    },
+    {
+      "epoch": 0.4025710419485792,
+      "grad_norm": 2.163089894373313,
+      "learning_rate": 9.964703127992822e-06,
+      "loss": 0.69,
+      "step": 595
+    },
+    {
+      "epoch": 0.40324763193504737,
+      "grad_norm": 2.4952934205498045,
+      "learning_rate": 9.964234631709188e-06,
+      "loss": 0.7057,
+      "step": 596
+    },
+    {
+      "epoch": 0.40392422192151556,
+      "grad_norm": 2.515865519983522,
+      "learning_rate": 9.963763057867658e-06,
+      "loss": 0.7149,
+      "step": 597
+    },
+    {
+      "epoch": 0.40460081190798375,
+      "grad_norm": 2.0215440093353227,
+      "learning_rate": 9.963288406760584e-06,
+      "loss": 0.6926,
+      "step": 598
+    },
+    {
+      "epoch": 0.40527740189445194,
+      "grad_norm": 2.1935710081830577,
+      "learning_rate": 9.962810678682223e-06,
+      "loss": 0.7007,
+      "step": 599
+    },
+    {
+      "epoch": 0.4059539918809202,
+      "grad_norm": 2.027744782273937,
+      "learning_rate": 9.962329873928743e-06,
+      "loss": 0.7042,
+      "step": 600
+    },
+    {
+      "epoch": 0.4066305818673884,
+      "grad_norm": 2.034332358536131,
+      "learning_rate": 9.961845992798213e-06,
+      "loss": 0.7381,
+      "step": 601
+    },
+    {
+      "epoch": 0.40730717185385656,
+      "grad_norm": 2.0084286006292515,
+      "learning_rate": 9.961359035590619e-06,
+      "loss": 0.6834,
+      "step": 602
+    },
+    {
+      "epoch": 0.40798376184032475,
+      "grad_norm": 2.362532652474117,
+      "learning_rate": 9.960869002607843e-06,
+      "loss": 0.7133,
+      "step": 603
+    },
+    {
+      "epoch": 0.40866035182679294,
+      "grad_norm": 2.100407719983761,
+      "learning_rate": 9.960375894153682e-06,
+      "loss": 0.6868,
+      "step": 604
+    },
+    {
+      "epoch": 0.4093369418132612,
+      "grad_norm": 2.1368565789833265,
+      "learning_rate": 9.959879710533835e-06,
+      "loss": 0.6934,
+      "step": 605
+    },
+    {
+      "epoch": 0.4100135317997294,
+      "grad_norm": 2.0566631156288375,
+      "learning_rate": 9.959380452055909e-06,
+      "loss": 0.6786,
+      "step": 606
+    },
+    {
+      "epoch": 0.41069012178619757,
+      "grad_norm": 1.9593553918937765,
+      "learning_rate": 9.958878119029419e-06,
+      "loss": 0.7015,
+      "step": 607
+    },
+    {
+      "epoch": 0.41136671177266576,
+      "grad_norm": 2.018785216368903,
+      "learning_rate": 9.958372711765785e-06,
+      "loss": 0.6794,
+      "step": 608
+    },
+    {
+      "epoch": 0.41204330175913395,
+      "grad_norm": 1.9075054250954835,
+      "learning_rate": 9.95786423057833e-06,
+      "loss": 0.6691,
+      "step": 609
+    },
+    {
+      "epoch": 0.41271989174560214,
+      "grad_norm": 2.0871748298241766,
+      "learning_rate": 9.957352675782283e-06,
+      "loss": 0.6762,
+      "step": 610
+    },
+    {
+      "epoch": 0.4133964817320704,
+      "grad_norm": 1.9188715246488623,
+      "learning_rate": 9.956838047694785e-06,
+      "loss": 0.6627,
+      "step": 611
+    },
+    {
+      "epoch": 0.4140730717185386,
+      "grad_norm": 1.9847524523392805,
+      "learning_rate": 9.956320346634877e-06,
+      "loss": 0.6884,
+      "step": 612
+    },
+    {
+      "epoch": 0.41474966170500677,
+      "grad_norm": 2.043064857643346,
+      "learning_rate": 9.955799572923503e-06,
+      "loss": 0.6711,
+      "step": 613
+    },
+    {
+      "epoch": 0.41542625169147496,
+      "grad_norm": 2.0325247485054296,
+      "learning_rate": 9.955275726883517e-06,
+      "loss": 0.6838,
+      "step": 614
+    },
+    {
+      "epoch": 0.41610284167794315,
+      "grad_norm": 1.8907423347407817,
+      "learning_rate": 9.954748808839675e-06,
+      "loss": 0.6816,
+      "step": 615
+    },
+    {
+      "epoch": 0.4167794316644114,
+      "grad_norm": 2.1406953226662515,
+      "learning_rate": 9.954218819118636e-06,
+      "loss": 0.6949,
+      "step": 616
+    },
+    {
+      "epoch": 0.4174560216508796,
+      "grad_norm": 2.001786047271728,
+      "learning_rate": 9.953685758048968e-06,
+      "loss": 0.6815,
+      "step": 617
+    },
+    {
+      "epoch": 0.4181326116373478,
+      "grad_norm": 1.929934647977257,
+      "learning_rate": 9.953149625961136e-06,
+      "loss": 0.6852,
+      "step": 618
+    },
+    {
+      "epoch": 0.41880920162381596,
+      "grad_norm": 2.0589536480408883,
+      "learning_rate": 9.952610423187516e-06,
+      "loss": 0.6902,
+      "step": 619
+    },
+    {
+      "epoch": 0.41948579161028415,
+      "grad_norm": 1.9996210703499524,
+      "learning_rate": 9.952068150062386e-06,
+      "loss": 0.6707,
+      "step": 620
+    },
+    {
+      "epoch": 0.42016238159675234,
+      "grad_norm": 2.0134316880601206,
+      "learning_rate": 9.951522806921922e-06,
+      "loss": 0.683,
+      "step": 621
+    },
+    {
+      "epoch": 0.4208389715832206,
+      "grad_norm": 1.86633078703543,
+      "learning_rate": 9.95097439410421e-06,
+      "loss": 0.641,
+      "step": 622
+    },
+    {
+      "epoch": 0.4215155615696888,
+      "grad_norm": 2.2129596018316136,
+      "learning_rate": 9.950422911949238e-06,
+      "loss": 0.6932,
+      "step": 623
+    },
+    {
+      "epoch": 0.42219215155615697,
+      "grad_norm": 1.944116798384578,
+      "learning_rate": 9.949868360798893e-06,
+      "loss": 0.7012,
+      "step": 624
+    },
+    {
+      "epoch": 0.42286874154262516,
+      "grad_norm": 1.9308190608146374,
+      "learning_rate": 9.949310740996964e-06,
+      "loss": 0.6716,
+      "step": 625
+    },
+    {
+      "epoch": 0.42354533152909335,
+      "grad_norm": 1.782688098228263,
+      "learning_rate": 9.94875005288915e-06,
+      "loss": 0.674,
+      "step": 626
+    },
+    {
+      "epoch": 0.4242219215155616,
+      "grad_norm": 2.1653710361638945,
+      "learning_rate": 9.948186296823048e-06,
+      "loss": 0.6705,
+      "step": 627
+    },
+    {
+      "epoch": 0.4248985115020298,
+      "grad_norm": 1.9491818637213443,
+      "learning_rate": 9.947619473148152e-06,
+      "loss": 0.6792,
+      "step": 628
+    },
+    {
+      "epoch": 0.425575101488498,
+      "grad_norm": 1.995741143847466,
+      "learning_rate": 9.947049582215862e-06,
+      "loss": 0.7007,
+      "step": 629
+    },
+    {
+      "epoch": 0.42625169147496617,
+      "grad_norm": 1.8127064040211225,
+      "learning_rate": 9.946476624379485e-06,
+      "loss": 0.6555,
+      "step": 630
+    },
+    {
+      "epoch": 0.42692828146143436,
+      "grad_norm": 2.002480955111533,
+      "learning_rate": 9.945900599994219e-06,
+      "loss": 0.6943,
+      "step": 631
+    },
+    {
+      "epoch": 0.42760487144790255,
+      "grad_norm": 2.0318946970414564,
+      "learning_rate": 9.94532150941717e-06,
+      "loss": 0.6629,
+      "step": 632
+    },
+    {
+      "epoch": 0.4282814614343708,
+      "grad_norm": 1.9877226730218924,
+      "learning_rate": 9.944739353007344e-06,
+      "loss": 0.6803,
+      "step": 633
+    },
+    {
+      "epoch": 0.428958051420839,
+      "grad_norm": 1.809109992463502,
+      "learning_rate": 9.944154131125643e-06,
+      "loss": 0.6559,
+      "step": 634
+    },
+    {
+      "epoch": 0.42963464140730717,
+      "grad_norm": 1.997159047981803,
+      "learning_rate": 9.943565844134877e-06,
+      "loss": 0.6867,
+      "step": 635
+    },
+    {
+      "epoch": 0.43031123139377536,
+      "grad_norm": 1.810602576207129,
+      "learning_rate": 9.942974492399751e-06,
+      "loss": 0.6702,
+      "step": 636
+    },
+    {
+      "epoch": 0.43098782138024355,
+      "grad_norm": 2.0451887760889322,
+      "learning_rate": 9.94238007628687e-06,
+      "loss": 0.6728,
+      "step": 637
+    },
+    {
+      "epoch": 0.4316644113667118,
+      "grad_norm": 1.9773424679166982,
+      "learning_rate": 9.94178259616474e-06,
+      "loss": 0.6554,
+      "step": 638
+    },
+    {
+      "epoch": 0.43234100135318,
+      "grad_norm": 1.983323900570349,
+      "learning_rate": 9.941182052403768e-06,
+      "loss": 0.6753,
+      "step": 639
+    },
+    {
+      "epoch": 0.4330175913396482,
+      "grad_norm": 1.9549786355670833,
+      "learning_rate": 9.940578445376259e-06,
+      "loss": 0.6905,
+      "step": 640
+    },
+    {
+      "epoch": 0.43369418132611637,
+      "grad_norm": 1.997179151236007,
+      "learning_rate": 9.939971775456416e-06,
+      "loss": 0.6847,
+      "step": 641
+    },
+    {
+      "epoch": 0.43437077131258456,
+      "grad_norm": 1.8413656351114143,
+      "learning_rate": 9.93936204302034e-06,
+      "loss": 0.6888,
+      "step": 642
+    },
+    {
+      "epoch": 0.43504736129905275,
+      "grad_norm": 1.8738449595961248,
+      "learning_rate": 9.938749248446033e-06,
+      "loss": 0.6413,
+      "step": 643
+    },
+    {
+      "epoch": 0.435723951285521,
+      "grad_norm": 1.9154050527684428,
+      "learning_rate": 9.938133392113399e-06,
+      "loss": 0.6786,
+      "step": 644
+    },
+    {
+      "epoch": 0.4364005412719892,
+      "grad_norm": 1.9692984990498263,
+      "learning_rate": 9.937514474404229e-06,
+      "loss": 0.652,
+      "step": 645
+    },
+    {
+      "epoch": 0.4370771312584574,
+      "grad_norm": 1.8363218558877794,
+      "learning_rate": 9.936892495702222e-06,
+      "loss": 0.6318,
+      "step": 646
+    },
+    {
+      "epoch": 0.43775372124492556,
+      "grad_norm": 1.8058194378723227,
+      "learning_rate": 9.936267456392971e-06,
+      "loss": 0.6534,
+      "step": 647
+    },
+    {
+      "epoch": 0.43843031123139375,
+      "grad_norm": 1.9353376187771927,
+      "learning_rate": 9.935639356863966e-06,
+      "loss": 0.6969,
+      "step": 648
+    },
+    {
+      "epoch": 0.439106901217862,
+      "grad_norm": 1.8956154103192093,
+      "learning_rate": 9.935008197504596e-06,
+      "loss": 0.6521,
+      "step": 649
+    },
+    {
+      "epoch": 0.4397834912043302,
+      "grad_norm": 1.7924097574849651,
+      "learning_rate": 9.934373978706147e-06,
+      "loss": 0.604,
+      "step": 650
+    },
+    {
+      "epoch": 0.4404600811907984,
+      "grad_norm": 1.9110970217222203,
+      "learning_rate": 9.933736700861798e-06,
+      "loss": 0.6808,
+      "step": 651
+    },
+    {
+      "epoch": 0.44113667117726657,
+      "grad_norm": 1.7458422006902716,
+      "learning_rate": 9.933096364366625e-06,
+      "loss": 0.6512,
+      "step": 652
+    },
+    {
+      "epoch": 0.44181326116373476,
+      "grad_norm": 1.7984905753531888,
+      "learning_rate": 9.932452969617607e-06,
+      "loss": 0.6278,
+      "step": 653
+    },
+    {
+      "epoch": 0.44248985115020295,
+      "grad_norm": 1.8244577525823302,
+      "learning_rate": 9.931806517013612e-06,
+      "loss": 0.643,
+      "step": 654
+    },
+    {
+      "epoch": 0.4431664411366712,
+      "grad_norm": 1.8462614021324941,
+      "learning_rate": 9.931157006955406e-06,
+      "loss": 0.6822,
+      "step": 655
+    },
+    {
+      "epoch": 0.4438430311231394,
+      "grad_norm": 1.8237762592102016,
+      "learning_rate": 9.93050443984565e-06,
+      "loss": 0.6549,
+      "step": 656
+    },
+    {
+      "epoch": 0.4445196211096076,
+      "grad_norm": 2.0264274302521432,
+      "learning_rate": 9.929848816088898e-06,
+      "loss": 0.6764,
+      "step": 657
+    },
+    {
+      "epoch": 0.44519621109607577,
+      "grad_norm": 1.9028989377044923,
+      "learning_rate": 9.929190136091604e-06,
+      "loss": 0.6619,
+      "step": 658
+    },
+    {
+      "epoch": 0.44587280108254396,
+      "grad_norm": 1.9349232488713752,
+      "learning_rate": 9.928528400262116e-06,
+      "loss": 0.6592,
+      "step": 659
+    },
+    {
+      "epoch": 0.4465493910690122,
+      "grad_norm": 1.8586228478591957,
+      "learning_rate": 9.92786360901067e-06,
+      "loss": 0.6867,
+      "step": 660
+    },
+    {
+      "epoch": 0.4472259810554804,
+      "grad_norm": 1.7870259169553473,
+      "learning_rate": 9.927195762749405e-06,
+      "loss": 0.6487,
+      "step": 661
+    },
+    {
+      "epoch": 0.4479025710419486,
+      "grad_norm": 1.8126776620578224,
+      "learning_rate": 9.926524861892346e-06,
+      "loss": 0.6567,
+      "step": 662
+    },
+    {
+      "epoch": 0.4485791610284168,
+      "grad_norm": 1.8737868617314668,
+      "learning_rate": 9.925850906855419e-06,
+      "loss": 0.6614,
+      "step": 663
+    },
+    {
+      "epoch": 0.44925575101488496,
+      "grad_norm": 1.7713697769223755,
+      "learning_rate": 9.925173898056436e-06,
+      "loss": 0.6278,
+      "step": 664
+    },
+    {
+      "epoch": 0.44993234100135315,
+      "grad_norm": 1.9092961289094104,
+      "learning_rate": 9.924493835915108e-06,
+      "loss": 0.6668,
+      "step": 665
+    },
+    {
+      "epoch": 0.4506089309878214,
+      "grad_norm": 1.7496215142400025,
+      "learning_rate": 9.923810720853038e-06,
+      "loss": 0.6373,
+      "step": 666
+    },
+    {
+      "epoch": 0.4512855209742896,
+      "grad_norm": 1.883308707797926,
+      "learning_rate": 9.923124553293718e-06,
+      "loss": 0.6623,
+      "step": 667
+    },
+    {
+      "epoch": 0.4519621109607578,
+      "grad_norm": 1.8381070698311122,
+      "learning_rate": 9.922435333662537e-06,
+      "loss": 0.6577,
+      "step": 668
+    },
+    {
+      "epoch": 0.45263870094722597,
+      "grad_norm": 1.8134254950348916,
+      "learning_rate": 9.921743062386773e-06,
+      "loss": 0.6799,
+      "step": 669
+    },
+    {
+      "epoch": 0.45331529093369416,
+      "grad_norm": 1.8177625611343793,
+      "learning_rate": 9.921047739895596e-06,
+      "loss": 0.6206,
+      "step": 670
+    },
+    {
+      "epoch": 0.4539918809201624,
+      "grad_norm": 1.916375693161388,
+      "learning_rate": 9.92034936662007e-06,
+      "loss": 0.6466,
+      "step": 671
+    },
+    {
+      "epoch": 0.4546684709066306,
+      "grad_norm": 1.8666278331742292,
+      "learning_rate": 9.91964794299315e-06,
+      "loss": 0.6472,
+      "step": 672
+    },
+    {
+      "epoch": 0.4553450608930988,
+      "grad_norm": 1.841159739003564,
+      "learning_rate": 9.918943469449676e-06,
+      "loss": 0.6738,
+      "step": 673
+    },
+    {
+      "epoch": 0.456021650879567,
+      "grad_norm": 1.931039343691473,
+      "learning_rate": 9.918235946426389e-06,
+      "loss": 0.6464,
+      "step": 674
+    },
+    {
+      "epoch": 0.45669824086603517,
+      "grad_norm": 1.6913206369911358,
+      "learning_rate": 9.917525374361913e-06,
+      "loss": 0.6387,
+      "step": 675
+    },
+    {
+      "epoch": 0.45737483085250336,
+      "grad_norm": 1.9636634964730848,
+      "learning_rate": 9.916811753696764e-06,
+      "loss": 0.6377,
+      "step": 676
+    },
+    {
+      "epoch": 0.4580514208389716,
+      "grad_norm": 1.7872712718739638,
+      "learning_rate": 9.916095084873348e-06,
+      "loss": 0.646,
+      "step": 677
+    },
+    {
+      "epoch": 0.4587280108254398,
+      "grad_norm": 1.6642630676043548,
+      "learning_rate": 9.915375368335962e-06,
+      "loss": 0.5887,
+      "step": 678
+    },
+    {
+      "epoch": 0.459404600811908,
+      "grad_norm": 1.810919197529432,
+      "learning_rate": 9.91465260453079e-06,
+      "loss": 0.6268,
+      "step": 679
+    },
+    {
+      "epoch": 0.46008119079837617,
+      "grad_norm": 1.7866371130807348,
+      "learning_rate": 9.913926793905909e-06,
+      "loss": 0.6155,
+      "step": 680
+    },
+    {
+      "epoch": 0.46075778078484436,
+      "grad_norm": 1.738555445668848,
+      "learning_rate": 9.91319793691128e-06,
+      "loss": 0.6722,
+      "step": 681
+    },
+    {
+      "epoch": 0.4614343707713126,
+      "grad_norm": 1.8542733755959186,
+      "learning_rate": 9.912466033998758e-06,
+      "loss": 0.6604,
+      "step": 682
+    },
+    {
+      "epoch": 0.4621109607577808,
+      "grad_norm": 1.6243959213835935,
+      "learning_rate": 9.91173108562208e-06,
+      "loss": 0.6177,
+      "step": 683
+    },
+    {
+      "epoch": 0.462787550744249,
+      "grad_norm": 1.7088849631880303,
+      "learning_rate": 9.910993092236878e-06,
+      "loss": 0.6321,
+      "step": 684
+    },
+    {
+      "epoch": 0.4634641407307172,
+      "grad_norm": 1.744776998203165,
+      "learning_rate": 9.910252054300664e-06,
+      "loss": 0.6234,
+      "step": 685
+    },
+    {
+      "epoch": 0.46414073071718537,
+      "grad_norm": 1.713082320412285,
+      "learning_rate": 9.909507972272845e-06,
+      "loss": 0.6204,
+      "step": 686
+    },
+    {
+      "epoch": 0.4648173207036536,
+      "grad_norm": 1.7906364998078443,
+      "learning_rate": 9.90876084661471e-06,
+      "loss": 0.6279,
+      "step": 687
+    },
+    {
+      "epoch": 0.4654939106901218,
+      "grad_norm": 1.7059545171919712,
+      "learning_rate": 9.908010677789437e-06,
+      "loss": 0.6217,
+      "step": 688
+    },
+    {
+      "epoch": 0.46617050067659,
+      "grad_norm": 1.7506382698310816,
+      "learning_rate": 9.90725746626209e-06,
+      "loss": 0.6492,
+      "step": 689
+    },
+    {
+      "epoch": 0.4668470906630582,
+      "grad_norm": 1.7765082721148218,
+      "learning_rate": 9.90650121249962e-06,
+      "loss": 0.6602,
+      "step": 690
+    },
+    {
+      "epoch": 0.4675236806495264,
+      "grad_norm": 1.6752513183319067,
+      "learning_rate": 9.905741916970863e-06,
+      "loss": 0.6338,
+      "step": 691
+    },
+    {
+      "epoch": 0.46820027063599456,
+      "grad_norm": 1.7186093366528268,
+      "learning_rate": 9.904979580146544e-06,
+      "loss": 0.5968,
+      "step": 692
+    },
+    {
+      "epoch": 0.4688768606224628,
+      "grad_norm": 1.665313992607149,
+      "learning_rate": 9.904214202499266e-06,
+      "loss": 0.6499,
+      "step": 693
+    },
+    {
+      "epoch": 0.469553450608931,
+      "grad_norm": 1.716744851695661,
+      "learning_rate": 9.903445784503525e-06,
+      "loss": 0.6388,
+      "step": 694
+    },
+    {
+      "epoch": 0.4702300405953992,
+      "grad_norm": 1.7245973304097408,
+      "learning_rate": 9.902674326635698e-06,
+      "loss": 0.6164,
+      "step": 695
+    },
+    {
+      "epoch": 0.4709066305818674,
+      "grad_norm": 1.7252557740173053,
+      "learning_rate": 9.901899829374048e-06,
+      "loss": 0.6271,
+      "step": 696
+    },
+    {
+      "epoch": 0.47158322056833557,
+      "grad_norm": 1.7400065832286413,
+      "learning_rate": 9.90112229319872e-06,
+      "loss": 0.6115,
+      "step": 697
+    },
+    {
+      "epoch": 0.4722598105548038,
+      "grad_norm": 1.7879186578579442,
+      "learning_rate": 9.900341718591746e-06,
+      "loss": 0.6612,
+      "step": 698
+    },
+    {
+      "epoch": 0.472936400541272,
+      "grad_norm": 1.7582531777040444,
+      "learning_rate": 9.899558106037039e-06,
+      "loss": 0.6318,
+      "step": 699
+    },
+    {
+      "epoch": 0.4736129905277402,
+      "grad_norm": 1.7056978103836629,
+      "learning_rate": 9.898771456020397e-06,
+      "loss": 0.6496,
+      "step": 700
+    },
+    {
+      "epoch": 0.4742895805142084,
+      "grad_norm": 1.8130704404644866,
+      "learning_rate": 9.897981769029504e-06,
+      "loss": 0.6539,
+      "step": 701
+    },
+    {
+      "epoch": 0.4749661705006766,
+      "grad_norm": 1.6519286721992499,
+      "learning_rate": 9.897189045553917e-06,
+      "loss": 0.6403,
+      "step": 702
+    },
+    {
+      "epoch": 0.47564276048714477,
+      "grad_norm": 1.662796737313033,
+      "learning_rate": 9.896393286085085e-06,
+      "loss": 0.6284,
+      "step": 703
+    },
+    {
+      "epoch": 0.476319350473613,
+      "grad_norm": 1.6513492760188824,
+      "learning_rate": 9.895594491116336e-06,
+      "loss": 0.6105,
+      "step": 704
+    },
+    {
+      "epoch": 0.4769959404600812,
+      "grad_norm": 1.815000983504719,
+      "learning_rate": 9.89479266114288e-06,
+      "loss": 0.6721,
+      "step": 705
+    },
+    {
+      "epoch": 0.4776725304465494,
+      "grad_norm": 1.7292582267683765,
+      "learning_rate": 9.893987796661809e-06,
+      "loss": 0.6248,
+      "step": 706
+    },
+    {
+      "epoch": 0.4783491204330176,
+      "grad_norm": 1.7289472357290376,
+      "learning_rate": 9.893179898172095e-06,
+      "loss": 0.6493,
+      "step": 707
+    },
+    {
+      "epoch": 0.4790257104194858,
+      "grad_norm": 1.6593555289396826,
+      "learning_rate": 9.89236896617459e-06,
+      "loss": 0.5745,
+      "step": 708
+    },
+    {
+      "epoch": 0.479702300405954,
+      "grad_norm": 1.672847157894918,
+      "learning_rate": 9.891555001172032e-06,
+      "loss": 0.6026,
+      "step": 709
+    },
+    {
+      "epoch": 0.4803788903924222,
+      "grad_norm": 1.7502390676959785,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 0.6003,
+      "step": 710
+    },
+    {
+      "epoch": 0.4810554803788904,
+      "grad_norm": 1.7317795497035786,
+      "learning_rate": 9.88991797417208e-06,
+      "loss": 0.6252,
+      "step": 711
+    },
+    {
+      "epoch": 0.4817320703653586,
+      "grad_norm": 1.785190533669979,
+      "learning_rate": 9.889094913189561e-06,
+      "loss": 0.6374,
+      "step": 712
+    },
+    {
+      "epoch": 0.4824086603518268,
+      "grad_norm": 1.578833584389062,
+      "learning_rate": 9.888268821231721e-06,
+      "loss": 0.6063,
+      "step": 713
+    },
+    {
+      "epoch": 0.48308525033829497,
+      "grad_norm": 1.7596384787332495,
+      "learning_rate": 9.887439698810694e-06,
+      "loss": 0.62,
+      "step": 714
+    },
+    {
+      "epoch": 0.4837618403247632,
+      "grad_norm": 1.594461218310731,
+      "learning_rate": 9.886607546440492e-06,
+      "loss": 0.6045,
+      "step": 715
+    },
+    {
+      "epoch": 0.4844384303112314,
+      "grad_norm": 1.7861022819494534,
+      "learning_rate": 9.885772364637002e-06,
+      "loss": 0.6406,
+      "step": 716
+    },
+    {
+      "epoch": 0.4851150202976996,
+      "grad_norm": 1.7637826966498402,
+      "learning_rate": 9.884934153917998e-06,
+      "loss": 0.6699,
+      "step": 717
+    },
+    {
+      "epoch": 0.4857916102841678,
+      "grad_norm": 1.7715147918492709,
+      "learning_rate": 9.884092914803119e-06,
+      "loss": 0.6445,
+      "step": 718
+    },
+    {
+      "epoch": 0.486468200270636,
+      "grad_norm": 1.7337695186422941,
+      "learning_rate": 9.88324864781389e-06,
+      "loss": 0.6273,
+      "step": 719
+    },
+    {
+      "epoch": 0.4871447902571042,
+      "grad_norm": 1.5963191484454953,
+      "learning_rate": 9.882401353473711e-06,
+      "loss": 0.5803,
+      "step": 720
+    },
+    {
+      "epoch": 0.4878213802435724,
+      "grad_norm": 1.6713741090198078,
+      "learning_rate": 9.881551032307859e-06,
+      "loss": 0.6024,
+      "step": 721
+    },
+    {
+      "epoch": 0.4884979702300406,
+      "grad_norm": 1.6048808576870026,
+      "learning_rate": 9.880697684843487e-06,
+      "loss": 0.6145,
+      "step": 722
+    },
+    {
+      "epoch": 0.4891745602165088,
+      "grad_norm": 1.7435957538894575,
+      "learning_rate": 9.879841311609625e-06,
+      "loss": 0.6097,
+      "step": 723
+    },
+    {
+      "epoch": 0.489851150202977,
+      "grad_norm": 1.6588461692725849,
+      "learning_rate": 9.878981913137178e-06,
+      "loss": 0.6155,
+      "step": 724
+    },
+    {
+      "epoch": 0.49052774018944517,
+      "grad_norm": 1.7396401883773205,
+      "learning_rate": 9.878119489958929e-06,
+      "loss": 0.6211,
+      "step": 725
+    },
+    {
+      "epoch": 0.4912043301759134,
+      "grad_norm": 1.6723586481609591,
+      "learning_rate": 9.877254042609529e-06,
+      "loss": 0.6384,
+      "step": 726
+    },
+    {
+      "epoch": 0.4918809201623816,
+      "grad_norm": 1.5772435277903296,
+      "learning_rate": 9.87638557162551e-06,
+      "loss": 0.6014,
+      "step": 727
+    },
+    {
+      "epoch": 0.4925575101488498,
+      "grad_norm": 1.6462833868011817,
+      "learning_rate": 9.875514077545282e-06,
+      "loss": 0.6496,
+      "step": 728
+    },
+    {
+      "epoch": 0.493234100135318,
+      "grad_norm": 1.7827549905726272,
+      "learning_rate": 9.874639560909118e-06,
+      "loss": 0.6522,
+      "step": 729
+    },
+    {
+      "epoch": 0.4939106901217862,
+      "grad_norm": 1.623572952971577,
+      "learning_rate": 9.873762022259177e-06,
+      "loss": 0.6044,
+      "step": 730
+    },
+    {
+      "epoch": 0.4945872801082544,
+      "grad_norm": 1.6716564125292326,
+      "learning_rate": 9.87288146213948e-06,
+      "loss": 0.5997,
+      "step": 731
+    },
+    {
+      "epoch": 0.4952638700947226,
+      "grad_norm": 1.7137680916001972,
+      "learning_rate": 9.87199788109593e-06,
+      "loss": 0.6069,
+      "step": 732
+    },
+    {
+      "epoch": 0.4959404600811908,
+      "grad_norm": 1.7164683410507124,
+      "learning_rate": 9.8711112796763e-06,
+      "loss": 0.6033,
+      "step": 733
+    },
+    {
+      "epoch": 0.496617050067659,
+      "grad_norm": 1.7395270346998304,
+      "learning_rate": 9.870221658430233e-06,
+      "loss": 0.5927,
+      "step": 734
+    },
+    {
+      "epoch": 0.4972936400541272,
+      "grad_norm": 1.6223821863202548,
+      "learning_rate": 9.869329017909248e-06,
+      "loss": 0.6003,
+      "step": 735
+    },
+    {
+      "epoch": 0.4979702300405954,
+      "grad_norm": 1.6582083441683446,
+      "learning_rate": 9.868433358666734e-06,
+      "loss": 0.5864,
+      "step": 736
+    },
+    {
+      "epoch": 0.4986468200270636,
+      "grad_norm": 1.6235788830079854,
+      "learning_rate": 9.86753468125795e-06,
+      "loss": 0.6234,
+      "step": 737
+    },
+    {
+      "epoch": 0.4993234100135318,
+      "grad_norm": 1.5575952668537478,
+      "learning_rate": 9.86663298624003e-06,
+      "loss": 0.6021,
+      "step": 738
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.659423383816892,
+      "learning_rate": 9.865728274171972e-06,
+      "loss": 0.6102,
+      "step": 739
+    },
+    {
+      "epoch": 0.5006765899864682,
+      "grad_norm": 1.6923854605985684,
+      "learning_rate": 9.864820545614656e-06,
+      "loss": 0.6446,
+      "step": 740
+    },
+    {
+      "epoch": 0.5013531799729364,
+      "grad_norm": 1.6846452394519977,
+      "learning_rate": 9.863909801130816e-06,
+      "loss": 0.6482,
+      "step": 741
+    },
+    {
+      "epoch": 0.5020297699594046,
+      "grad_norm": 1.7552123660153136,
+      "learning_rate": 9.862996041285071e-06,
+      "loss": 0.5958,
+      "step": 742
+    },
+    {
+      "epoch": 0.5027063599458728,
+      "grad_norm": 1.6085553057005082,
+      "learning_rate": 9.862079266643899e-06,
+      "loss": 0.6171,
+      "step": 743
+    },
+    {
+      "epoch": 0.503382949932341,
+      "grad_norm": 1.6659215114844248,
+      "learning_rate": 9.861159477775653e-06,
+      "loss": 0.611,
+      "step": 744
+    },
+    {
+      "epoch": 0.5040595399188093,
+      "grad_norm": 1.622234077476183,
+      "learning_rate": 9.860236675250553e-06,
+      "loss": 0.629,
+      "step": 745
+    },
+    {
+      "epoch": 0.5047361299052774,
+      "grad_norm": 1.6778622263084553,
+      "learning_rate": 9.859310859640685e-06,
+      "loss": 0.6113,
+      "step": 746
+    },
+    {
+      "epoch": 0.5054127198917456,
+      "grad_norm": 1.6063455871758163,
+      "learning_rate": 9.858382031520005e-06,
+      "loss": 0.6181,
+      "step": 747
+    },
+    {
+      "epoch": 0.5060893098782138,
+      "grad_norm": 1.6017126433116808,
+      "learning_rate": 9.857450191464337e-06,
+      "loss": 0.604,
+      "step": 748
+    },
+    {
+      "epoch": 0.506765899864682,
+      "grad_norm": 1.5800409303674505,
+      "learning_rate": 9.856515340051374e-06,
+      "loss": 0.621,
+      "step": 749
+    },
+    {
+      "epoch": 0.5074424898511503,
+      "grad_norm": 1.6335282822148747,
+      "learning_rate": 9.855577477860669e-06,
+      "loss": 0.6128,
+      "step": 750
+    },
+    {
+      "epoch": 0.5081190798376184,
+      "grad_norm": 1.6471217728117251,
+      "learning_rate": 9.854636605473647e-06,
+      "loss": 0.6264,
+      "step": 751
+    },
+    {
+      "epoch": 0.5087956698240866,
+      "grad_norm": 1.6708409892034344,
+      "learning_rate": 9.8536927234736e-06,
+      "loss": 0.5891,
+      "step": 752
+    },
+    {
+      "epoch": 0.5094722598105548,
+      "grad_norm": 1.5678996027099266,
+      "learning_rate": 9.852745832445684e-06,
+      "loss": 0.621,
+      "step": 753
+    },
+    {
+      "epoch": 0.510148849797023,
+      "grad_norm": 1.5662685910655971,
+      "learning_rate": 9.851795932976919e-06,
+      "loss": 0.5957,
+      "step": 754
+    },
+    {
+      "epoch": 0.5108254397834912,
+      "grad_norm": 1.541521445093749,
+      "learning_rate": 9.850843025656194e-06,
+      "loss": 0.6006,
+      "step": 755
+    },
+    {
+      "epoch": 0.5115020297699594,
+      "grad_norm": 1.548693267118839,
+      "learning_rate": 9.849887111074256e-06,
+      "loss": 0.5999,
+      "step": 756
+    },
+    {
+      "epoch": 0.5121786197564276,
+      "grad_norm": 1.6330795062621721,
+      "learning_rate": 9.848928189823724e-06,
+      "loss": 0.6405,
+      "step": 757
+    },
+    {
+      "epoch": 0.5128552097428958,
+      "grad_norm": 1.6165681102295442,
+      "learning_rate": 9.847966262499073e-06,
+      "loss": 0.6105,
+      "step": 758
+    },
+    {
+      "epoch": 0.513531799729364,
+      "grad_norm": 1.6337780848379253,
+      "learning_rate": 9.847001329696653e-06,
+      "loss": 0.6194,
+      "step": 759
+    },
+    {
+      "epoch": 0.5142083897158322,
+      "grad_norm": 1.598034839418694,
+      "learning_rate": 9.846033392014665e-06,
+      "loss": 0.6161,
+      "step": 760
+    },
+    {
+      "epoch": 0.5148849797023004,
+      "grad_norm": 1.5342729581057934,
+      "learning_rate": 9.84506245005318e-06,
+      "loss": 0.6009,
+      "step": 761
+    },
+    {
+      "epoch": 0.5155615696887687,
+      "grad_norm": 1.5924867997084955,
+      "learning_rate": 9.84408850441413e-06,
+      "loss": 0.6206,
+      "step": 762
+    },
+    {
+      "epoch": 0.5162381596752368,
+      "grad_norm": 1.6130256943080854,
+      "learning_rate": 9.843111555701307e-06,
+      "loss": 0.5903,
+      "step": 763
+    },
+    {
+      "epoch": 0.516914749661705,
+      "grad_norm": 1.692415463857787,
+      "learning_rate": 9.84213160452037e-06,
+      "loss": 0.619,
+      "step": 764
+    },
+    {
+      "epoch": 0.5175913396481732,
+      "grad_norm": 1.6261092756599715,
+      "learning_rate": 9.841148651478833e-06,
+      "loss": 0.6073,
+      "step": 765
+    },
+    {
+      "epoch": 0.5182679296346414,
+      "grad_norm": 1.6455088298521052,
+      "learning_rate": 9.840162697186075e-06,
+      "loss": 0.6099,
+      "step": 766
+    },
+    {
+      "epoch": 0.5189445196211097,
+      "grad_norm": 1.5521392563452239,
+      "learning_rate": 9.839173742253334e-06,
+      "loss": 0.6111,
+      "step": 767
+    },
+    {
+      "epoch": 0.5196211096075778,
+      "grad_norm": 1.5765567484232255,
+      "learning_rate": 9.838181787293707e-06,
+      "loss": 0.5873,
+      "step": 768
+    },
+    {
+      "epoch": 0.520297699594046,
+      "grad_norm": 1.580053176799126,
+      "learning_rate": 9.837186832922157e-06,
+      "loss": 0.6019,
+      "step": 769
+    },
+    {
+      "epoch": 0.5209742895805142,
+      "grad_norm": 1.4773449572850088,
+      "learning_rate": 9.8361888797555e-06,
+      "loss": 0.5992,
+      "step": 770
+    },
+    {
+      "epoch": 0.5216508795669824,
+      "grad_norm": 1.5134890082335568,
+      "learning_rate": 9.835187928412412e-06,
+      "loss": 0.5826,
+      "step": 771
+    },
+    {
+      "epoch": 0.5223274695534507,
+      "grad_norm": 1.5839800415867358,
+      "learning_rate": 9.834183979513427e-06,
+      "loss": 0.6098,
+      "step": 772
+    },
+    {
+      "epoch": 0.5230040595399188,
+      "grad_norm": 1.560504675974206,
+      "learning_rate": 9.833177033680945e-06,
+      "loss": 0.5776,
+      "step": 773
+    },
+    {
+      "epoch": 0.523680649526387,
+      "grad_norm": 1.5092576948094416,
+      "learning_rate": 9.832167091539215e-06,
+      "loss": 0.5801,
+      "step": 774
+    },
+    {
+      "epoch": 0.5243572395128552,
+      "grad_norm": 1.5258055048481598,
+      "learning_rate": 9.831154153714344e-06,
+      "loss": 0.6124,
+      "step": 775
+    },
+    {
+      "epoch": 0.5250338294993234,
+      "grad_norm": 1.6141761010516198,
+      "learning_rate": 9.830138220834305e-06,
+      "loss": 0.5627,
+      "step": 776
+    },
+    {
+      "epoch": 0.5257104194857916,
+      "grad_norm": 1.4899611207648424,
+      "learning_rate": 9.829119293528916e-06,
+      "loss": 0.5995,
+      "step": 777
+    },
+    {
+      "epoch": 0.5263870094722598,
+      "grad_norm": 1.5297959832164443,
+      "learning_rate": 9.82809737242986e-06,
+      "loss": 0.5994,
+      "step": 778
+    },
+    {
+      "epoch": 0.527063599458728,
+      "grad_norm": 1.5009548259970094,
+      "learning_rate": 9.827072458170673e-06,
+      "loss": 0.5871,
+      "step": 779
+    },
+    {
+      "epoch": 0.5277401894451962,
+      "grad_norm": 1.534081377423664,
+      "learning_rate": 9.826044551386743e-06,
+      "loss": 0.6127,
+      "step": 780
+    },
+    {
+      "epoch": 0.5284167794316644,
+      "grad_norm": 1.5835054937186908,
+      "learning_rate": 9.825013652715323e-06,
+      "loss": 0.5705,
+      "step": 781
+    },
+    {
+      "epoch": 0.5290933694181326,
+      "grad_norm": 1.6197476891525862,
+      "learning_rate": 9.82397976279551e-06,
+      "loss": 0.6249,
+      "step": 782
+    },
+    {
+      "epoch": 0.5297699594046008,
+      "grad_norm": 1.5118940285786684,
+      "learning_rate": 9.822942882268261e-06,
+      "loss": 0.5644,
+      "step": 783
+    },
+    {
+      "epoch": 0.530446549391069,
+      "grad_norm": 1.61150903255849,
+      "learning_rate": 9.821903011776385e-06,
+      "loss": 0.5976,
+      "step": 784
+    },
+    {
+      "epoch": 0.5311231393775372,
+      "grad_norm": 1.5878566064030444,
+      "learning_rate": 9.820860151964548e-06,
+      "loss": 0.5916,
+      "step": 785
+    },
+    {
+      "epoch": 0.5317997293640054,
+      "grad_norm": 1.5508462981554838,
+      "learning_rate": 9.819814303479268e-06,
+      "loss": 0.6005,
+      "step": 786
+    },
+    {
+      "epoch": 0.5324763193504736,
+      "grad_norm": 1.5243635027775746,
+      "learning_rate": 9.818765466968909e-06,
+      "loss": 0.5954,
+      "step": 787
+    },
+    {
+      "epoch": 0.5331529093369418,
+      "grad_norm": 1.5074445872918962,
+      "learning_rate": 9.8177136430837e-06,
+      "loss": 0.5652,
+      "step": 788
+    },
+    {
+      "epoch": 0.5338294993234101,
+      "grad_norm": 1.5398916190523528,
+      "learning_rate": 9.816658832475709e-06,
+      "loss": 0.6261,
+      "step": 789
+    },
+    {
+      "epoch": 0.5345060893098782,
+      "grad_norm": 1.4913814473732248,
+      "learning_rate": 9.815601035798866e-06,
+      "loss": 0.5544,
+      "step": 790
+    },
+    {
+      "epoch": 0.5351826792963464,
+      "grad_norm": 1.512191575421141,
+      "learning_rate": 9.814540253708945e-06,
+      "loss": 0.6049,
+      "step": 791
+    },
+    {
+      "epoch": 0.5358592692828146,
+      "grad_norm": 1.506943852426354,
+      "learning_rate": 9.813476486863575e-06,
+      "loss": 0.5747,
+      "step": 792
+    },
+    {
+      "epoch": 0.5365358592692828,
+      "grad_norm": 1.5289902512576314,
+      "learning_rate": 9.812409735922236e-06,
+      "loss": 0.5927,
+      "step": 793
+    },
+    {
+      "epoch": 0.5372124492557511,
+      "grad_norm": 1.4688894059588264,
+      "learning_rate": 9.811340001546252e-06,
+      "loss": 0.5626,
+      "step": 794
+    },
+    {
+      "epoch": 0.5378890392422192,
+      "grad_norm": 1.5502677365309896,
+      "learning_rate": 9.810267284398805e-06,
+      "loss": 0.5775,
+      "step": 795
+    },
+    {
+      "epoch": 0.5385656292286874,
+      "grad_norm": 1.504260579122203,
+      "learning_rate": 9.80919158514492e-06,
+      "loss": 0.5816,
+      "step": 796
+    },
+    {
+      "epoch": 0.5392422192151556,
+      "grad_norm": 1.5332683654835957,
+      "learning_rate": 9.80811290445147e-06,
+      "loss": 0.61,
+      "step": 797
+    },
+    {
+      "epoch": 0.5399188092016238,
+      "grad_norm": 1.5829335228456114,
+      "learning_rate": 9.807031242987182e-06,
+      "loss": 0.5739,
+      "step": 798
+    },
+    {
+      "epoch": 0.540595399188092,
+      "grad_norm": 1.477030825458606,
+      "learning_rate": 9.805946601422628e-06,
+      "loss": 0.5709,
+      "step": 799
+    },
+    {
+      "epoch": 0.5412719891745602,
+      "grad_norm": 1.5310442073122814,
+      "learning_rate": 9.804858980430225e-06,
+      "loss": 0.5774,
+      "step": 800
+    },
+    {
+      "epoch": 0.5419485791610285,
+      "grad_norm": 1.6198458755827938,
+      "learning_rate": 9.803768380684242e-06,
+      "loss": 0.5983,
+      "step": 801
+    },
+    {
+      "epoch": 0.5426251691474966,
+      "grad_norm": 1.5722213284156308,
+      "learning_rate": 9.80267480286079e-06,
+      "loss": 0.592,
+      "step": 802
+    },
+    {
+      "epoch": 0.5433017591339648,
+      "grad_norm": 1.5478352782171632,
+      "learning_rate": 9.801578247637828e-06,
+      "loss": 0.5861,
+      "step": 803
+    },
+    {
+      "epoch": 0.543978349120433,
+      "grad_norm": 1.5819907116760954,
+      "learning_rate": 9.800478715695165e-06,
+      "loss": 0.6087,
+      "step": 804
+    },
+    {
+      "epoch": 0.5446549391069012,
+      "grad_norm": 1.524512354441419,
+      "learning_rate": 9.799376207714446e-06,
+      "loss": 0.5879,
+      "step": 805
+    },
+    {
+      "epoch": 0.5453315290933695,
+      "grad_norm": 1.517953472171127,
+      "learning_rate": 9.79827072437917e-06,
+      "loss": 0.5882,
+      "step": 806
+    },
+    {
+      "epoch": 0.5460081190798376,
+      "grad_norm": 1.6697483870541,
+      "learning_rate": 9.797162266374677e-06,
+      "loss": 0.6129,
+      "step": 807
+    },
+    {
+      "epoch": 0.5466847090663058,
+      "grad_norm": 1.5406927455004877,
+      "learning_rate": 9.79605083438815e-06,
+      "loss": 0.6111,
+      "step": 808
+    },
+    {
+      "epoch": 0.547361299052774,
+      "grad_norm": 1.573139563138134,
+      "learning_rate": 9.794936429108617e-06,
+      "loss": 0.5508,
+      "step": 809
+    },
+    {
+      "epoch": 0.5480378890392422,
+      "grad_norm": 1.5569949399993268,
+      "learning_rate": 9.79381905122695e-06,
+      "loss": 0.5967,
+      "step": 810
+    },
+    {
+      "epoch": 0.5487144790257105,
+      "grad_norm": 1.440985317133278,
+      "learning_rate": 9.792698701435863e-06,
+      "loss": 0.5741,
+      "step": 811
+    },
+    {
+      "epoch": 0.5493910690121786,
+      "grad_norm": 1.4814660344152049,
+      "learning_rate": 9.791575380429911e-06,
+      "loss": 0.5865,
+      "step": 812
+    },
+    {
+      "epoch": 0.5500676589986468,
+      "grad_norm": 1.4855959485099581,
+      "learning_rate": 9.790449088905496e-06,
+      "loss": 0.5847,
+      "step": 813
+    },
+    {
+      "epoch": 0.550744248985115,
+      "grad_norm": 1.3877950490527795,
+      "learning_rate": 9.789319827560854e-06,
+      "loss": 0.5636,
+      "step": 814
+    },
+    {
+      "epoch": 0.5514208389715832,
+      "grad_norm": 1.4871104111545084,
+      "learning_rate": 9.78818759709607e-06,
+      "loss": 0.5699,
+      "step": 815
+    },
+    {
+      "epoch": 0.5520974289580515,
+      "grad_norm": 1.5520974904698874,
+      "learning_rate": 9.787052398213062e-06,
+      "loss": 0.6043,
+      "step": 816
+    },
+    {
+      "epoch": 0.5527740189445196,
+      "grad_norm": 1.51941049125711,
+      "learning_rate": 9.785914231615595e-06,
+      "loss": 0.5894,
+      "step": 817
+    },
+    {
+      "epoch": 0.5534506089309879,
+      "grad_norm": 1.51253479768057,
+      "learning_rate": 9.784773098009269e-06,
+      "loss": 0.5941,
+      "step": 818
+    },
+    {
+      "epoch": 0.554127198917456,
+      "grad_norm": 1.4985468635005212,
+      "learning_rate": 9.783628998101525e-06,
+      "loss": 0.6045,
+      "step": 819
+    },
+    {
+      "epoch": 0.5548037889039242,
+      "grad_norm": 1.7399730564579428,
+      "learning_rate": 9.782481932601643e-06,
+      "loss": 0.6035,
+      "step": 820
+    },
+    {
+      "epoch": 0.5554803788903924,
+      "grad_norm": 1.5148424096689828,
+      "learning_rate": 9.781331902220748e-06,
+      "loss": 0.5818,
+      "step": 821
+    },
+    {
+      "epoch": 0.5561569688768606,
+      "grad_norm": 1.5734432459925136,
+      "learning_rate": 9.780178907671788e-06,
+      "loss": 0.598,
+      "step": 822
+    },
+    {
+      "epoch": 0.5568335588633289,
+      "grad_norm": 1.4341192532764733,
+      "learning_rate": 9.779022949669565e-06,
+      "loss": 0.5712,
+      "step": 823
+    },
+    {
+      "epoch": 0.557510148849797,
+      "grad_norm": 1.4828360153292617,
+      "learning_rate": 9.777864028930705e-06,
+      "loss": 0.5762,
+      "step": 824
+    },
+    {
+      "epoch": 0.5581867388362652,
+      "grad_norm": 1.5272581217393904,
+      "learning_rate": 9.776702146173678e-06,
+      "loss": 0.5902,
+      "step": 825
+    },
+    {
+      "epoch": 0.5588633288227334,
+      "grad_norm": 1.474462218909566,
+      "learning_rate": 9.775537302118791e-06,
+      "loss": 0.5937,
+      "step": 826
+    },
+    {
+      "epoch": 0.5595399188092016,
+      "grad_norm": 1.4711865855163142,
+      "learning_rate": 9.77436949748818e-06,
+      "loss": 0.5624,
+      "step": 827
+    },
+    {
+      "epoch": 0.5602165087956699,
+      "grad_norm": 1.5541161951656597,
+      "learning_rate": 9.773198733005827e-06,
+      "loss": 0.5636,
+      "step": 828
+    },
+    {
+      "epoch": 0.560893098782138,
+      "grad_norm": 1.4528736826698192,
+      "learning_rate": 9.772025009397538e-06,
+      "loss": 0.5708,
+      "step": 829
+    },
+    {
+      "epoch": 0.5615696887686062,
+      "grad_norm": 1.5348342474317382,
+      "learning_rate": 9.770848327390961e-06,
+      "loss": 0.578,
+      "step": 830
+    },
+    {
+      "epoch": 0.5622462787550744,
+      "grad_norm": 1.5639549858597588,
+      "learning_rate": 9.769668687715572e-06,
+      "loss": 0.5973,
+      "step": 831
+    },
+    {
+      "epoch": 0.5629228687415426,
+      "grad_norm": 1.4919566493475664,
+      "learning_rate": 9.76848609110269e-06,
+      "loss": 0.5771,
+      "step": 832
+    },
+    {
+      "epoch": 0.5635994587280109,
+      "grad_norm": 1.590278449722994,
+      "learning_rate": 9.767300538285454e-06,
+      "loss": 0.5713,
+      "step": 833
+    },
+    {
+      "epoch": 0.564276048714479,
+      "grad_norm": 1.5159888392403542,
+      "learning_rate": 9.766112029998847e-06,
+      "loss": 0.5746,
+      "step": 834
+    },
+    {
+      "epoch": 0.5649526387009473,
+      "grad_norm": 1.4087331775743346,
+      "learning_rate": 9.76492056697968e-06,
+      "loss": 0.5656,
+      "step": 835
+    },
+    {
+      "epoch": 0.5656292286874154,
+      "grad_norm": 1.5401107560462461,
+      "learning_rate": 9.763726149966596e-06,
+      "loss": 0.5855,
+      "step": 836
+    },
+    {
+      "epoch": 0.5663058186738836,
+      "grad_norm": 1.4859624312720217,
+      "learning_rate": 9.762528779700067e-06,
+      "loss": 0.5793,
+      "step": 837
+    },
+    {
+      "epoch": 0.5669824086603519,
+      "grad_norm": 1.5000618258437695,
+      "learning_rate": 9.7613284569224e-06,
+      "loss": 0.5952,
+      "step": 838
+    },
+    {
+      "epoch": 0.56765899864682,
+      "grad_norm": 1.4847226268347697,
+      "learning_rate": 9.760125182377732e-06,
+      "loss": 0.5599,
+      "step": 839
+    },
+    {
+      "epoch": 0.5683355886332883,
+      "grad_norm": 1.3974150433692198,
+      "learning_rate": 9.758918956812024e-06,
+      "loss": 0.5836,
+      "step": 840
+    },
+    {
+      "epoch": 0.5690121786197564,
+      "grad_norm": 1.5439055222320957,
+      "learning_rate": 9.757709780973074e-06,
+      "loss": 0.5825,
+      "step": 841
+    },
+    {
+      "epoch": 0.5696887686062246,
+      "grad_norm": 1.4336760360463616,
+      "learning_rate": 9.756497655610503e-06,
+      "loss": 0.5572,
+      "step": 842
+    },
+    {
+      "epoch": 0.5703653585926928,
+      "grad_norm": 1.4289167361745316,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.5614,
+      "step": 843
+    },
+    {
+      "epoch": 0.571041948579161,
+      "grad_norm": 1.4397286373393339,
+      "learning_rate": 9.754064559322147e-06,
+      "loss": 0.5759,
+      "step": 844
+    },
+    {
+      "epoch": 0.5717185385656293,
+      "grad_norm": 1.3921029382982602,
+      "learning_rate": 9.752843589904746e-06,
+      "loss": 0.5647,
+      "step": 845
+    },
+    {
+      "epoch": 0.5723951285520974,
+      "grad_norm": 1.4569036063966363,
+      "learning_rate": 9.751619673980503e-06,
+      "loss": 0.5993,
+      "step": 846
+    },
+    {
+      "epoch": 0.5730717185385656,
+      "grad_norm": 1.4955773165520974,
+      "learning_rate": 9.75039281230818e-06,
+      "loss": 0.5747,
+      "step": 847
+    },
+    {
+      "epoch": 0.5737483085250338,
+      "grad_norm": 1.43011602275059,
+      "learning_rate": 9.749163005648362e-06,
+      "loss": 0.556,
+      "step": 848
+    },
+    {
+      "epoch": 0.574424898511502,
+      "grad_norm": 1.4987929090940766,
+      "learning_rate": 9.747930254763467e-06,
+      "loss": 0.601,
+      "step": 849
+    },
+    {
+      "epoch": 0.5751014884979703,
+      "grad_norm": 1.5221805690391956,
+      "learning_rate": 9.746694560417731e-06,
+      "loss": 0.5792,
+      "step": 850
+    },
+    {
+      "epoch": 0.5757780784844384,
+      "grad_norm": 1.4409909021292568,
+      "learning_rate": 9.745455923377218e-06,
+      "loss": 0.5686,
+      "step": 851
+    },
+    {
+      "epoch": 0.5764546684709067,
+      "grad_norm": 1.5331733122946738,
+      "learning_rate": 9.74421434440982e-06,
+      "loss": 0.5877,
+      "step": 852
+    },
+    {
+      "epoch": 0.5771312584573748,
+      "grad_norm": 1.7728555669909092,
+      "learning_rate": 9.742969824285244e-06,
+      "loss": 0.5796,
+      "step": 853
+    },
+    {
+      "epoch": 0.577807848443843,
+      "grad_norm": 1.5527178949346045,
+      "learning_rate": 9.741722363775029e-06,
+      "loss": 0.5775,
+      "step": 854
+    },
+    {
+      "epoch": 0.5784844384303113,
+      "grad_norm": 1.4144723729143567,
+      "learning_rate": 9.74047196365253e-06,
+      "loss": 0.5894,
+      "step": 855
+    },
+    {
+      "epoch": 0.5791610284167794,
+      "grad_norm": 1.4405172045715295,
+      "learning_rate": 9.73921862469293e-06,
+      "loss": 0.56,
+      "step": 856
+    },
+    {
+      "epoch": 0.5798376184032477,
+      "grad_norm": 1.4593763462458083,
+      "learning_rate": 9.737962347673232e-06,
+      "loss": 0.5653,
+      "step": 857
+    },
+    {
+      "epoch": 0.5805142083897158,
+      "grad_norm": 1.5231015715647205,
+      "learning_rate": 9.736703133372259e-06,
+      "loss": 0.6081,
+      "step": 858
+    },
+    {
+      "epoch": 0.581190798376184,
+      "grad_norm": 1.3640730713147275,
+      "learning_rate": 9.735440982570656e-06,
+      "loss": 0.5729,
+      "step": 859
+    },
+    {
+      "epoch": 0.5818673883626523,
+      "grad_norm": 1.4477405879003105,
+      "learning_rate": 9.734175896050889e-06,
+      "loss": 0.5443,
+      "step": 860
+    },
+    {
+      "epoch": 0.5825439783491204,
+      "grad_norm": 1.3710307808039208,
+      "learning_rate": 9.732907874597241e-06,
+      "loss": 0.542,
+      "step": 861
+    },
+    {
+      "epoch": 0.5832205683355887,
+      "grad_norm": 1.3824132249109022,
+      "learning_rate": 9.731636918995821e-06,
+      "loss": 0.574,
+      "step": 862
+    },
+    {
+      "epoch": 0.5838971583220568,
+      "grad_norm": 1.4496950380801756,
+      "learning_rate": 9.730363030034551e-06,
+      "loss": 0.5632,
+      "step": 863
+    },
+    {
+      "epoch": 0.584573748308525,
+      "grad_norm": 1.389239972724504,
+      "learning_rate": 9.729086208503174e-06,
+      "loss": 0.5623,
+      "step": 864
+    },
+    {
+      "epoch": 0.5852503382949933,
+      "grad_norm": 1.491143461609495,
+      "learning_rate": 9.72780645519325e-06,
+      "loss": 0.5456,
+      "step": 865
+    },
+    {
+      "epoch": 0.5859269282814614,
+      "grad_norm": 1.5239062721142338,
+      "learning_rate": 9.726523770898157e-06,
+      "loss": 0.5616,
+      "step": 866
+    },
+    {
+      "epoch": 0.5866035182679297,
+      "grad_norm": 1.5155111551809417,
+      "learning_rate": 9.725238156413089e-06,
+      "loss": 0.5865,
+      "step": 867
+    },
+    {
+      "epoch": 0.5872801082543978,
+      "grad_norm": 1.477315881988796,
+      "learning_rate": 9.72394961253506e-06,
+      "loss": 0.5578,
+      "step": 868
+    },
+    {
+      "epoch": 0.587956698240866,
+      "grad_norm": 1.3431613060276728,
+      "learning_rate": 9.722658140062898e-06,
+      "loss": 0.555,
+      "step": 869
+    },
+    {
+      "epoch": 0.5886332882273342,
+      "grad_norm": 1.472403673794486,
+      "learning_rate": 9.721363739797243e-06,
+      "loss": 0.5638,
+      "step": 870
+    },
+    {
+      "epoch": 0.5893098782138024,
+      "grad_norm": 1.5119770565727169,
+      "learning_rate": 9.720066412540554e-06,
+      "loss": 0.5509,
+      "step": 871
+    },
+    {
+      "epoch": 0.5899864682002707,
+      "grad_norm": 1.4581998018729656,
+      "learning_rate": 9.718766159097109e-06,
+      "loss": 0.5741,
+      "step": 872
+    },
+    {
+      "epoch": 0.5906630581867388,
+      "grad_norm": 1.4778398991082073,
+      "learning_rate": 9.717462980272989e-06,
+      "loss": 0.5838,
+      "step": 873
+    },
+    {
+      "epoch": 0.591339648173207,
+      "grad_norm": 1.450744346748018,
+      "learning_rate": 9.716156876876096e-06,
+      "loss": 0.5735,
+      "step": 874
+    },
+    {
+      "epoch": 0.5920162381596752,
+      "grad_norm": 1.4330043370067966,
+      "learning_rate": 9.714847849716149e-06,
+      "loss": 0.5486,
+      "step": 875
+    },
+    {
+      "epoch": 0.5926928281461434,
+      "grad_norm": 1.3295272980967532,
+      "learning_rate": 9.713535899604667e-06,
+      "loss": 0.5562,
+      "step": 876
+    },
+    {
+      "epoch": 0.5933694181326117,
+      "grad_norm": 1.5018536264245457,
+      "learning_rate": 9.71222102735499e-06,
+      "loss": 0.5559,
+      "step": 877
+    },
+    {
+      "epoch": 0.5940460081190798,
+      "grad_norm": 1.5885433466214243,
+      "learning_rate": 9.710903233782273e-06,
+      "loss": 0.59,
+      "step": 878
+    },
+    {
+      "epoch": 0.5947225981055481,
+      "grad_norm": 1.371744266648417,
+      "learning_rate": 9.70958251970347e-06,
+      "loss": 0.5463,
+      "step": 879
+    },
+    {
+      "epoch": 0.5953991880920162,
+      "grad_norm": 1.3863121051011533,
+      "learning_rate": 9.708258885937359e-06,
+      "loss": 0.5481,
+      "step": 880
+    },
+    {
+      "epoch": 0.5960757780784844,
+      "grad_norm": 1.4983021715077849,
+      "learning_rate": 9.706932333304518e-06,
+      "loss": 0.599,
+      "step": 881
+    },
+    {
+      "epoch": 0.5967523680649527,
+      "grad_norm": 1.4007593996650822,
+      "learning_rate": 9.705602862627335e-06,
+      "loss": 0.557,
+      "step": 882
+    },
+    {
+      "epoch": 0.5974289580514208,
+      "grad_norm": 1.4820775219149764,
+      "learning_rate": 9.704270474730018e-06,
+      "loss": 0.5603,
+      "step": 883
+    },
+    {
+      "epoch": 0.5981055480378891,
+      "grad_norm": 1.5332245521088,
+      "learning_rate": 9.70293517043857e-06,
+      "loss": 0.5657,
+      "step": 884
+    },
+    {
+      "epoch": 0.5987821380243572,
+      "grad_norm": 1.5419686111048978,
+      "learning_rate": 9.701596950580807e-06,
+      "loss": 0.5884,
+      "step": 885
+    },
+    {
+      "epoch": 0.5994587280108254,
+      "grad_norm": 1.4434909655666306,
+      "learning_rate": 9.700255815986357e-06,
+      "loss": 0.5599,
+      "step": 886
+    },
+    {
+      "epoch": 0.6001353179972937,
+      "grad_norm": 1.3979104123355959,
+      "learning_rate": 9.69891176748665e-06,
+      "loss": 0.5444,
+      "step": 887
+    },
+    {
+      "epoch": 0.6008119079837618,
+      "grad_norm": 1.331308664936877,
+      "learning_rate": 9.697564805914922e-06,
+      "loss": 0.5499,
+      "step": 888
+    },
+    {
+      "epoch": 0.6014884979702301,
+      "grad_norm": 1.327499797705502,
+      "learning_rate": 9.696214932106218e-06,
+      "loss": 0.5408,
+      "step": 889
+    },
+    {
+      "epoch": 0.6021650879566982,
+      "grad_norm": 1.3379638644164007,
+      "learning_rate": 9.694862146897385e-06,
+      "loss": 0.5497,
+      "step": 890
+    },
+    {
+      "epoch": 0.6028416779431665,
+      "grad_norm": 1.3288675507428929,
+      "learning_rate": 9.693506451127082e-06,
+      "loss": 0.5512,
+      "step": 891
+    },
+    {
+      "epoch": 0.6035182679296346,
+      "grad_norm": 1.3719618397137918,
+      "learning_rate": 9.692147845635761e-06,
+      "loss": 0.5609,
+      "step": 892
+    },
+    {
+      "epoch": 0.6041948579161028,
+      "grad_norm": 1.309436398801699,
+      "learning_rate": 9.690786331265687e-06,
+      "loss": 0.5488,
+      "step": 893
+    },
+    {
+      "epoch": 0.6048714479025711,
+      "grad_norm": 1.3943922122388013,
+      "learning_rate": 9.689421908860928e-06,
+      "loss": 0.5471,
+      "step": 894
+    },
+    {
+      "epoch": 0.6055480378890392,
+      "grad_norm": 1.4847830573388547,
+      "learning_rate": 9.688054579267347e-06,
+      "loss": 0.5655,
+      "step": 895
+    },
+    {
+      "epoch": 0.6062246278755075,
+      "grad_norm": 1.4118053169125588,
+      "learning_rate": 9.68668434333262e-06,
+      "loss": 0.5689,
+      "step": 896
+    },
+    {
+      "epoch": 0.6069012178619756,
+      "grad_norm": 1.4584980750152714,
+      "learning_rate": 9.685311201906216e-06,
+      "loss": 0.5756,
+      "step": 897
+    },
+    {
+      "epoch": 0.6075778078484438,
+      "grad_norm": 1.425428426125189,
+      "learning_rate": 9.683935155839408e-06,
+      "loss": 0.5874,
+      "step": 898
+    },
+    {
+      "epoch": 0.6082543978349121,
+      "grad_norm": 1.3970744593822986,
+      "learning_rate": 9.682556205985274e-06,
+      "loss": 0.5661,
+      "step": 899
+    },
+    {
+      "epoch": 0.6089309878213802,
+      "grad_norm": 1.3837680826838403,
+      "learning_rate": 9.681174353198687e-06,
+      "loss": 0.5714,
+      "step": 900
+    },
+    {
+      "epoch": 0.6096075778078485,
+      "grad_norm": 1.441948220204842,
+      "learning_rate": 9.67978959833632e-06,
+      "loss": 0.5478,
+      "step": 901
+    },
+    {
+      "epoch": 0.6102841677943166,
+      "grad_norm": 1.3741120155021025,
+      "learning_rate": 9.678401942256648e-06,
+      "loss": 0.5837,
+      "step": 902
+    },
+    {
+      "epoch": 0.6109607577807848,
+      "grad_norm": 1.3963868245289623,
+      "learning_rate": 9.67701138581994e-06,
+      "loss": 0.5608,
+      "step": 903
+    },
+    {
+      "epoch": 0.6116373477672531,
+      "grad_norm": 1.384844476497834,
+      "learning_rate": 9.675617929888271e-06,
+      "loss": 0.5368,
+      "step": 904
+    },
+    {
+      "epoch": 0.6123139377537212,
+      "grad_norm": 1.3278573496251747,
+      "learning_rate": 9.674221575325503e-06,
+      "loss": 0.54,
+      "step": 905
+    },
+    {
+      "epoch": 0.6129905277401895,
+      "grad_norm": 1.3923228339398561,
+      "learning_rate": 9.672822322997305e-06,
+      "loss": 0.561,
+      "step": 906
+    },
+    {
+      "epoch": 0.6136671177266576,
+      "grad_norm": 1.395088698787219,
+      "learning_rate": 9.671420173771135e-06,
+      "loss": 0.5593,
+      "step": 907
+    },
+    {
+      "epoch": 0.6143437077131259,
+      "grad_norm": 1.4409163976038764,
+      "learning_rate": 9.670015128516253e-06,
+      "loss": 0.5521,
+      "step": 908
+    },
+    {
+      "epoch": 0.6150202976995941,
+      "grad_norm": 1.3785741338135242,
+      "learning_rate": 9.668607188103708e-06,
+      "loss": 0.5645,
+      "step": 909
+    },
+    {
+      "epoch": 0.6156968876860622,
+      "grad_norm": 1.4063255874161504,
+      "learning_rate": 9.667196353406352e-06,
+      "loss": 0.5514,
+      "step": 910
+    },
+    {
+      "epoch": 0.6163734776725305,
+      "grad_norm": 1.3357192560767774,
+      "learning_rate": 9.665782625298821e-06,
+      "loss": 0.5383,
+      "step": 911
+    },
+    {
+      "epoch": 0.6170500676589986,
+      "grad_norm": 1.3539482074197706,
+      "learning_rate": 9.664366004657553e-06,
+      "loss": 0.5604,
+      "step": 912
+    },
+    {
+      "epoch": 0.6177266576454669,
+      "grad_norm": 1.310691687673011,
+      "learning_rate": 9.662946492360777e-06,
+      "loss": 0.5303,
+      "step": 913
+    },
+    {
+      "epoch": 0.618403247631935,
+      "grad_norm": 1.2800569076417998,
+      "learning_rate": 9.66152408928851e-06,
+      "loss": 0.556,
+      "step": 914
+    },
+    {
+      "epoch": 0.6190798376184032,
+      "grad_norm": 1.3973489965953194,
+      "learning_rate": 9.66009879632257e-06,
+      "loss": 0.5725,
+      "step": 915
+    },
+    {
+      "epoch": 0.6197564276048715,
+      "grad_norm": 1.451780064487979,
+      "learning_rate": 9.65867061434656e-06,
+      "loss": 0.5617,
+      "step": 916
+    },
+    {
+      "epoch": 0.6204330175913396,
+      "grad_norm": 1.35933524582482,
+      "learning_rate": 9.657239544245877e-06,
+      "loss": 0.5573,
+      "step": 917
+    },
+    {
+      "epoch": 0.6211096075778079,
+      "grad_norm": 1.4232172017425695,
+      "learning_rate": 9.655805586907705e-06,
+      "loss": 0.5466,
+      "step": 918
+    },
+    {
+      "epoch": 0.621786197564276,
+      "grad_norm": 1.4312081095893574,
+      "learning_rate": 9.654368743221022e-06,
+      "loss": 0.5625,
+      "step": 919
+    },
+    {
+      "epoch": 0.6224627875507442,
+      "grad_norm": 1.3912552358253953,
+      "learning_rate": 9.652929014076593e-06,
+      "loss": 0.5683,
+      "step": 920
+    },
+    {
+      "epoch": 0.6231393775372125,
+      "grad_norm": 1.3702305761206166,
+      "learning_rate": 9.651486400366972e-06,
+      "loss": 0.5302,
+      "step": 921
+    },
+    {
+      "epoch": 0.6238159675236806,
+      "grad_norm": 1.3785136595357277,
+      "learning_rate": 9.650040902986504e-06,
+      "loss": 0.5267,
+      "step": 922
+    },
+    {
+      "epoch": 0.6244925575101489,
+      "grad_norm": 1.42489356713615,
+      "learning_rate": 9.648592522831316e-06,
+      "loss": 0.5678,
+      "step": 923
+    },
+    {
+      "epoch": 0.625169147496617,
+      "grad_norm": 1.458705178998173,
+      "learning_rate": 9.64714126079933e-06,
+      "loss": 0.5903,
+      "step": 924
+    },
+    {
+      "epoch": 0.6258457374830853,
+      "grad_norm": 1.2919714844532744,
+      "learning_rate": 9.645687117790246e-06,
+      "loss": 0.539,
+      "step": 925
+    },
+    {
+      "epoch": 0.6265223274695535,
+      "grad_norm": 1.3159779958295432,
+      "learning_rate": 9.644230094705555e-06,
+      "loss": 0.5482,
+      "step": 926
+    },
+    {
+      "epoch": 0.6271989174560216,
+      "grad_norm": 1.372558886610807,
+      "learning_rate": 9.642770192448537e-06,
+      "loss": 0.5513,
+      "step": 927
+    },
+    {
+      "epoch": 0.6278755074424899,
+      "grad_norm": 1.4623448167001696,
+      "learning_rate": 9.641307411924246e-06,
+      "loss": 0.553,
+      "step": 928
+    },
+    {
+      "epoch": 0.628552097428958,
+      "grad_norm": 1.2857193407365461,
+      "learning_rate": 9.639841754039534e-06,
+      "loss": 0.5448,
+      "step": 929
+    },
+    {
+      "epoch": 0.6292286874154263,
+      "grad_norm": 1.3979647765995682,
+      "learning_rate": 9.638373219703023e-06,
+      "loss": 0.5409,
+      "step": 930
+    },
+    {
+      "epoch": 0.6299052774018945,
+      "grad_norm": 1.4450876574726212,
+      "learning_rate": 9.63690180982513e-06,
+      "loss": 0.5499,
+      "step": 931
+    },
+    {
+      "epoch": 0.6305818673883626,
+      "grad_norm": 1.387580596637908,
+      "learning_rate": 9.635427525318048e-06,
+      "loss": 0.5444,
+      "step": 932
+    },
+    {
+      "epoch": 0.6312584573748309,
+      "grad_norm": 1.3897869799870526,
+      "learning_rate": 9.633950367095758e-06,
+      "loss": 0.5409,
+      "step": 933
+    },
+    {
+      "epoch": 0.631935047361299,
+      "grad_norm": 1.3799414280334457,
+      "learning_rate": 9.632470336074009e-06,
+      "loss": 0.5615,
+      "step": 934
+    },
+    {
+      "epoch": 0.6326116373477673,
+      "grad_norm": 1.397109410439002,
+      "learning_rate": 9.63098743317035e-06,
+      "loss": 0.5588,
+      "step": 935
+    },
+    {
+      "epoch": 0.6332882273342354,
+      "grad_norm": 1.3877108753591307,
+      "learning_rate": 9.629501659304096e-06,
+      "loss": 0.5332,
+      "step": 936
+    },
+    {
+      "epoch": 0.6339648173207036,
+      "grad_norm": 1.425214924037294,
+      "learning_rate": 9.628013015396347e-06,
+      "loss": 0.5406,
+      "step": 937
+    },
+    {
+      "epoch": 0.6346414073071719,
+      "grad_norm": 1.427636608291554,
+      "learning_rate": 9.626521502369984e-06,
+      "loss": 0.5588,
+      "step": 938
+    },
+    {
+      "epoch": 0.63531799729364,
+      "grad_norm": 1.3998173217547065,
+      "learning_rate": 9.625027121149665e-06,
+      "loss": 0.5755,
+      "step": 939
+    },
+    {
+      "epoch": 0.6359945872801083,
+      "grad_norm": 1.3913272168959687,
+      "learning_rate": 9.623529872661821e-06,
+      "loss": 0.5911,
+      "step": 940
+    },
+    {
+      "epoch": 0.6366711772665764,
+      "grad_norm": 1.4119241403062532,
+      "learning_rate": 9.62202975783467e-06,
+      "loss": 0.5698,
+      "step": 941
+    },
+    {
+      "epoch": 0.6373477672530447,
+      "grad_norm": 1.369482238930773,
+      "learning_rate": 9.620526777598202e-06,
+      "loss": 0.5501,
+      "step": 942
+    },
+    {
+      "epoch": 0.6380243572395129,
+      "grad_norm": 1.4337966859514184,
+      "learning_rate": 9.619020932884182e-06,
+      "loss": 0.5809,
+      "step": 943
+    },
+    {
+      "epoch": 0.638700947225981,
+      "grad_norm": 1.3801684958426206,
+      "learning_rate": 9.617512224626153e-06,
+      "loss": 0.538,
+      "step": 944
+    },
+    {
+      "epoch": 0.6393775372124493,
+      "grad_norm": 1.3568642938213509,
+      "learning_rate": 9.616000653759435e-06,
+      "loss": 0.5234,
+      "step": 945
+    },
+    {
+      "epoch": 0.6400541271989174,
+      "grad_norm": 1.4120466838855552,
+      "learning_rate": 9.614486221221115e-06,
+      "loss": 0.5589,
+      "step": 946
+    },
+    {
+      "epoch": 0.6407307171853857,
+      "grad_norm": 1.4200432234755074,
+      "learning_rate": 9.612968927950066e-06,
+      "loss": 0.5589,
+      "step": 947
+    },
+    {
+      "epoch": 0.6414073071718539,
+      "grad_norm": 1.514542623903677,
+      "learning_rate": 9.611448774886925e-06,
+      "loss": 0.5467,
+      "step": 948
+    },
+    {
+      "epoch": 0.642083897158322,
+      "grad_norm": 1.3394445600719525,
+      "learning_rate": 9.609925762974103e-06,
+      "loss": 0.5461,
+      "step": 949
+    },
+    {
+      "epoch": 0.6427604871447903,
+      "grad_norm": 1.3558253169872096,
+      "learning_rate": 9.60839989315579e-06,
+      "loss": 0.5464,
+      "step": 950
+    },
+    {
+      "epoch": 0.6434370771312584,
+      "grad_norm": 1.3355916041692906,
+      "learning_rate": 9.606871166377939e-06,
+      "loss": 0.5473,
+      "step": 951
+    },
+    {
+      "epoch": 0.6441136671177267,
+      "grad_norm": 1.4277624564048805,
+      "learning_rate": 9.60533958358828e-06,
+      "loss": 0.5627,
+      "step": 952
+    },
+    {
+      "epoch": 0.6447902571041949,
+      "grad_norm": 1.39810856974054,
+      "learning_rate": 9.603805145736311e-06,
+      "loss": 0.5506,
+      "step": 953
+    },
+    {
+      "epoch": 0.645466847090663,
+      "grad_norm": 1.399253302116368,
+      "learning_rate": 9.602267853773301e-06,
+      "loss": 0.525,
+      "step": 954
+    },
+    {
+      "epoch": 0.6461434370771313,
+      "grad_norm": 1.3599740675606824,
+      "learning_rate": 9.60072770865229e-06,
+      "loss": 0.5466,
+      "step": 955
+    },
+    {
+      "epoch": 0.6468200270635994,
+      "grad_norm": 1.3238748547686656,
+      "learning_rate": 9.599184711328082e-06,
+      "loss": 0.5251,
+      "step": 956
+    },
+    {
+      "epoch": 0.6474966170500677,
+      "grad_norm": 1.405404092108196,
+      "learning_rate": 9.597638862757255e-06,
+      "loss": 0.5839,
+      "step": 957
+    },
+    {
+      "epoch": 0.6481732070365359,
+      "grad_norm": 1.3278904637408784,
+      "learning_rate": 9.596090163898148e-06,
+      "loss": 0.5473,
+      "step": 958
+    },
+    {
+      "epoch": 0.648849797023004,
+      "grad_norm": 1.3142690515046698,
+      "learning_rate": 9.594538615710875e-06,
+      "loss": 0.5397,
+      "step": 959
+    },
+    {
+      "epoch": 0.6495263870094723,
+      "grad_norm": 1.2226279962051259,
+      "learning_rate": 9.59298421915731e-06,
+      "loss": 0.5249,
+      "step": 960
+    },
+    {
+      "epoch": 0.6502029769959404,
+      "grad_norm": 1.3097653165372385,
+      "learning_rate": 9.591426975201093e-06,
+      "loss": 0.5431,
+      "step": 961
+    },
+    {
+      "epoch": 0.6508795669824087,
+      "grad_norm": 1.3418537512883573,
+      "learning_rate": 9.589866884807637e-06,
+      "loss": 0.5573,
+      "step": 962
+    },
+    {
+      "epoch": 0.6515561569688768,
+      "grad_norm": 1.3365246025089783,
+      "learning_rate": 9.588303948944109e-06,
+      "loss": 0.5154,
+      "step": 963
+    },
+    {
+      "epoch": 0.652232746955345,
+      "grad_norm": 1.3271238141898198,
+      "learning_rate": 9.586738168579446e-06,
+      "loss": 0.5466,
+      "step": 964
+    },
+    {
+      "epoch": 0.6529093369418133,
+      "grad_norm": 1.3178585046613578,
+      "learning_rate": 9.58516954468435e-06,
+      "loss": 0.5526,
+      "step": 965
+    },
+    {
+      "epoch": 0.6535859269282814,
+      "grad_norm": 1.3989072692037787,
+      "learning_rate": 9.58359807823128e-06,
+      "loss": 0.5663,
+      "step": 966
+    },
+    {
+      "epoch": 0.6542625169147497,
+      "grad_norm": 1.3555037515851058,
+      "learning_rate": 9.582023770194462e-06,
+      "loss": 0.5547,
+      "step": 967
+    },
+    {
+      "epoch": 0.6549391069012178,
+      "grad_norm": 1.3813861872852315,
+      "learning_rate": 9.580446621549883e-06,
+      "loss": 0.5479,
+      "step": 968
+    },
+    {
+      "epoch": 0.6556156968876861,
+      "grad_norm": 1.424745070176629,
+      "learning_rate": 9.578866633275289e-06,
+      "loss": 0.568,
+      "step": 969
+    },
+    {
+      "epoch": 0.6562922868741543,
+      "grad_norm": 1.2674000115717234,
+      "learning_rate": 9.577283806350186e-06,
+      "loss": 0.4873,
+      "step": 970
+    },
+    {
+      "epoch": 0.6569688768606224,
+      "grad_norm": 1.4144882152699472,
+      "learning_rate": 9.575698141755844e-06,
+      "loss": 0.5732,
+      "step": 971
+    },
+    {
+      "epoch": 0.6576454668470907,
+      "grad_norm": 1.3286139699632946,
+      "learning_rate": 9.57410964047529e-06,
+      "loss": 0.549,
+      "step": 972
+    },
+    {
+      "epoch": 0.6583220568335588,
+      "grad_norm": 1.369980768822487,
+      "learning_rate": 9.572518303493305e-06,
+      "loss": 0.5637,
+      "step": 973
+    },
+    {
+      "epoch": 0.6589986468200271,
+      "grad_norm": 1.3191259196641867,
+      "learning_rate": 9.570924131796437e-06,
+      "loss": 0.5219,
+      "step": 974
+    },
+    {
+      "epoch": 0.6596752368064953,
+      "grad_norm": 1.2876748949527599,
+      "learning_rate": 9.569327126372985e-06,
+      "loss": 0.5313,
+      "step": 975
+    },
+    {
+      "epoch": 0.6603518267929634,
+      "grad_norm": 1.2517769114590152,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.5229,
+      "step": 976
+    },
+    {
+      "epoch": 0.6610284167794317,
+      "grad_norm": 1.3386273910558741,
+      "learning_rate": 9.566124618308312e-06,
+      "loss": 0.5554,
+      "step": 977
+    },
+    {
+      "epoch": 0.6617050067658998,
+      "grad_norm": 1.2835140736196629,
+      "learning_rate": 9.564519117652473e-06,
+      "loss": 0.5691,
+      "step": 978
+    },
+    {
+      "epoch": 0.6623815967523681,
+      "grad_norm": 1.3637938650411934,
+      "learning_rate": 9.562910787240814e-06,
+      "loss": 0.5489,
+      "step": 979
+    },
+    {
+      "epoch": 0.6630581867388363,
+      "grad_norm": 1.349687301178395,
+      "learning_rate": 9.56129962807041e-06,
+      "loss": 0.5521,
+      "step": 980
+    },
+    {
+      "epoch": 0.6637347767253045,
+      "grad_norm": 1.4169245371858592,
+      "learning_rate": 9.559685641140098e-06,
+      "loss": 0.5587,
+      "step": 981
+    },
+    {
+      "epoch": 0.6644113667117727,
+      "grad_norm": 1.3985311547678738,
+      "learning_rate": 9.55806882745046e-06,
+      "loss": 0.5285,
+      "step": 982
+    },
+    {
+      "epoch": 0.6650879566982408,
+      "grad_norm": 1.307960497977671,
+      "learning_rate": 9.556449188003831e-06,
+      "loss": 0.5028,
+      "step": 983
+    },
+    {
+      "epoch": 0.6657645466847091,
+      "grad_norm": 1.501170859949113,
+      "learning_rate": 9.554826723804304e-06,
+      "loss": 0.5557,
+      "step": 984
+    },
+    {
+      "epoch": 0.6664411366711772,
+      "grad_norm": 1.5050837366352103,
+      "learning_rate": 9.553201435857718e-06,
+      "loss": 0.5449,
+      "step": 985
+    },
+    {
+      "epoch": 0.6671177266576455,
+      "grad_norm": 1.279971240070052,
+      "learning_rate": 9.551573325171662e-06,
+      "loss": 0.52,
+      "step": 986
+    },
+    {
+      "epoch": 0.6677943166441137,
+      "grad_norm": 1.3580033928572774,
+      "learning_rate": 9.54994239275548e-06,
+      "loss": 0.5281,
+      "step": 987
+    },
+    {
+      "epoch": 0.6684709066305818,
+      "grad_norm": 1.4122733978886406,
+      "learning_rate": 9.54830863962026e-06,
+      "loss": 0.5377,
+      "step": 988
+    },
+    {
+      "epoch": 0.6691474966170501,
+      "grad_norm": 1.318103326767308,
+      "learning_rate": 9.546672066778842e-06,
+      "loss": 0.5491,
+      "step": 989
+    },
+    {
+      "epoch": 0.6698240866035182,
+      "grad_norm": 1.367910485968864,
+      "learning_rate": 9.545032675245814e-06,
+      "loss": 0.5509,
+      "step": 990
+    },
+    {
+      "epoch": 0.6705006765899865,
+      "grad_norm": 1.3568884466439552,
+      "learning_rate": 9.543390466037507e-06,
+      "loss": 0.5568,
+      "step": 991
+    },
+    {
+      "epoch": 0.6711772665764547,
+      "grad_norm": 1.2525351435097483,
+      "learning_rate": 9.541745440172006e-06,
+      "loss": 0.5121,
+      "step": 992
+    },
+    {
+      "epoch": 0.6718538565629228,
+      "grad_norm": 1.2936482783093453,
+      "learning_rate": 9.540097598669135e-06,
+      "loss": 0.5144,
+      "step": 993
+    },
+    {
+      "epoch": 0.6725304465493911,
+      "grad_norm": 1.36600537631378,
+      "learning_rate": 9.538446942550468e-06,
+      "loss": 0.5528,
+      "step": 994
+    },
+    {
+      "epoch": 0.6732070365358592,
+      "grad_norm": 1.2729433973145832,
+      "learning_rate": 9.536793472839325e-06,
+      "loss": 0.5239,
+      "step": 995
+    },
+    {
+      "epoch": 0.6738836265223275,
+      "grad_norm": 1.327473038211668,
+      "learning_rate": 9.535137190560765e-06,
+      "loss": 0.5554,
+      "step": 996
+    },
+    {
+      "epoch": 0.6745602165087957,
+      "grad_norm": 1.3515073959577413,
+      "learning_rate": 9.533478096741597e-06,
+      "loss": 0.5361,
+      "step": 997
+    },
+    {
+      "epoch": 0.6752368064952639,
+      "grad_norm": 1.2826188441164121,
+      "learning_rate": 9.531816192410366e-06,
+      "loss": 0.5459,
+      "step": 998
+    },
+    {
+      "epoch": 0.6759133964817321,
+      "grad_norm": 1.300107320008121,
+      "learning_rate": 9.530151478597366e-06,
+      "loss": 0.5608,
+      "step": 999
+    },
+    {
+      "epoch": 0.6765899864682002,
+      "grad_norm": 1.3356839606812958,
+      "learning_rate": 9.528483956334628e-06,
+      "loss": 0.5147,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6772665764546685,
+      "grad_norm": 1.3120279365134704,
+      "learning_rate": 9.526813626655929e-06,
+      "loss": 0.536,
+      "step": 1001
+    },
+    {
+      "epoch": 0.6779431664411367,
+      "grad_norm": 1.3297459079562277,
+      "learning_rate": 9.525140490596778e-06,
+      "loss": 0.5389,
+      "step": 1002
+    },
+    {
+      "epoch": 0.6786197564276049,
+      "grad_norm": 1.263353060560232,
+      "learning_rate": 9.523464549194434e-06,
+      "loss": 0.5222,
+      "step": 1003
+    },
+    {
+      "epoch": 0.6792963464140731,
+      "grad_norm": 1.3766618870033855,
+      "learning_rate": 9.521785803487888e-06,
+      "loss": 0.527,
+      "step": 1004
+    },
+    {
+      "epoch": 0.6799729364005412,
+      "grad_norm": 1.3777397427193374,
+      "learning_rate": 9.520104254517873e-06,
+      "loss": 0.5463,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6806495263870095,
+      "grad_norm": 1.2408275412363299,
+      "learning_rate": 9.518419903326859e-06,
+      "loss": 0.5164,
+      "step": 1006
+    },
+    {
+      "epoch": 0.6813261163734776,
+      "grad_norm": 1.3468025610994963,
+      "learning_rate": 9.51673275095905e-06,
+      "loss": 0.5432,
+      "step": 1007
+    },
+    {
+      "epoch": 0.6820027063599459,
+      "grad_norm": 1.363867425812368,
+      "learning_rate": 9.515042798460393e-06,
+      "loss": 0.5657,
+      "step": 1008
+    },
+    {
+      "epoch": 0.6826792963464141,
+      "grad_norm": 1.2738870118029653,
+      "learning_rate": 9.513350046878565e-06,
+      "loss": 0.5526,
+      "step": 1009
+    },
+    {
+      "epoch": 0.6833558863328822,
+      "grad_norm": 1.3541456471191247,
+      "learning_rate": 9.511654497262984e-06,
+      "loss": 0.5484,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6840324763193505,
+      "grad_norm": 1.2885587079894851,
+      "learning_rate": 9.509956150664796e-06,
+      "loss": 0.5126,
+      "step": 1011
+    },
+    {
+      "epoch": 0.6847090663058186,
+      "grad_norm": 1.3438732729106644,
+      "learning_rate": 9.508255008136885e-06,
+      "loss": 0.5218,
+      "step": 1012
+    },
+    {
+      "epoch": 0.6853856562922869,
+      "grad_norm": 1.3795448163943937,
+      "learning_rate": 9.506551070733869e-06,
+      "loss": 0.5489,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6860622462787551,
+      "grad_norm": 1.285620830621772,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.5345,
+      "step": 1014
+    },
+    {
+      "epoch": 0.6867388362652233,
+      "grad_norm": 1.2756329996389932,
+      "learning_rate": 9.50313481552965e-06,
+      "loss": 0.5126,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6874154262516915,
+      "grad_norm": 1.2432220267937537,
+      "learning_rate": 9.501422499846338e-06,
+      "loss": 0.524,
+      "step": 1016
+    },
+    {
+      "epoch": 0.6880920162381596,
+      "grad_norm": 1.257677476756879,
+      "learning_rate": 9.49970739352371e-06,
+      "loss": 0.534,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6887686062246279,
+      "grad_norm": 1.3273447461283017,
+      "learning_rate": 9.497989497625036e-06,
+      "loss": 0.5406,
+      "step": 1018
+    },
+    {
+      "epoch": 0.6894451962110961,
+      "grad_norm": 1.260547433668687,
+      "learning_rate": 9.49626881321532e-06,
+      "loss": 0.5173,
+      "step": 1019
+    },
+    {
+      "epoch": 0.6901217861975643,
+      "grad_norm": 1.2886542014829554,
+      "learning_rate": 9.494545341361291e-06,
+      "loss": 0.5143,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6907983761840325,
+      "grad_norm": 1.343253740959597,
+      "learning_rate": 9.492819083131412e-06,
+      "loss": 0.5618,
+      "step": 1021
+    },
+    {
+      "epoch": 0.6914749661705006,
+      "grad_norm": 1.2555010974446998,
+      "learning_rate": 9.491090039595869e-06,
+      "loss": 0.5319,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6921515561569689,
+      "grad_norm": 1.319505203780607,
+      "learning_rate": 9.489358211826577e-06,
+      "loss": 0.5375,
+      "step": 1023
+    },
+    {
+      "epoch": 0.6928281461434371,
+      "grad_norm": 1.1996956522628355,
+      "learning_rate": 9.487623600897172e-06,
+      "loss": 0.4947,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6935047361299053,
+      "grad_norm": 1.3127047147597104,
+      "learning_rate": 9.485886207883022e-06,
+      "loss": 0.528,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6941813261163735,
+      "grad_norm": 1.3191073278457845,
+      "learning_rate": 9.484146033861216e-06,
+      "loss": 0.5259,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6948579161028416,
+      "grad_norm": 1.254310556929913,
+      "learning_rate": 9.482403079910571e-06,
+      "loss": 0.5119,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6955345060893099,
+      "grad_norm": 1.2337606715988945,
+      "learning_rate": 9.480657347111621e-06,
+      "loss": 0.503,
+      "step": 1028
+    },
+    {
+      "epoch": 0.696211096075778,
+      "grad_norm": 1.3540905295740424,
+      "learning_rate": 9.478908836546629e-06,
+      "loss": 0.5292,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6968876860622463,
+      "grad_norm": 1.4171859855033513,
+      "learning_rate": 9.477157549299574e-06,
+      "loss": 0.5524,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6975642760487145,
+      "grad_norm": 1.393922847605659,
+      "learning_rate": 9.475403486456162e-06,
+      "loss": 0.5356,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6982408660351827,
+      "grad_norm": 1.2893129326178892,
+      "learning_rate": 9.473646649103819e-06,
+      "loss": 0.5303,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6989174560216509,
+      "grad_norm": 1.395325205307781,
+      "learning_rate": 9.471887038331686e-06,
+      "loss": 0.5336,
+      "step": 1033
+    },
+    {
+      "epoch": 0.699594046008119,
+      "grad_norm": 1.2208612224676296,
+      "learning_rate": 9.470124655230627e-06,
+      "loss": 0.5163,
+      "step": 1034
+    },
+    {
+      "epoch": 0.7002706359945873,
+      "grad_norm": 1.2460030453002091,
+      "learning_rate": 9.468359500893227e-06,
+      "loss": 0.5118,
+      "step": 1035
+    },
+    {
+      "epoch": 0.7009472259810555,
+      "grad_norm": 1.3306247260952895,
+      "learning_rate": 9.466591576413785e-06,
+      "loss": 0.5362,
+      "step": 1036
+    },
+    {
+      "epoch": 0.7016238159675237,
+      "grad_norm": 1.2585485402143146,
+      "learning_rate": 9.464820882888319e-06,
+      "loss": 0.5258,
+      "step": 1037
+    },
+    {
+      "epoch": 0.7023004059539919,
+      "grad_norm": 1.288794719662461,
+      "learning_rate": 9.463047421414564e-06,
+      "loss": 0.5386,
+      "step": 1038
+    },
+    {
+      "epoch": 0.70297699594046,
+      "grad_norm": 1.2475012289505003,
+      "learning_rate": 9.461271193091971e-06,
+      "loss": 0.5256,
+      "step": 1039
+    },
+    {
+      "epoch": 0.7036535859269283,
+      "grad_norm": 1.2087947388839142,
+      "learning_rate": 9.459492199021705e-06,
+      "loss": 0.5451,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7043301759133965,
+      "grad_norm": 1.3104079751387059,
+      "learning_rate": 9.457710440306645e-06,
+      "loss": 0.5265,
+      "step": 1041
+    },
+    {
+      "epoch": 0.7050067658998647,
+      "grad_norm": 1.3380393908643815,
+      "learning_rate": 9.455925918051388e-06,
+      "loss": 0.5369,
+      "step": 1042
+    },
+    {
+      "epoch": 0.7056833558863329,
+      "grad_norm": 1.2511546699253946,
+      "learning_rate": 9.454138633362241e-06,
+      "loss": 0.5297,
+      "step": 1043
+    },
+    {
+      "epoch": 0.706359945872801,
+      "grad_norm": 1.324148288096631,
+      "learning_rate": 9.452348587347224e-06,
+      "loss": 0.5317,
+      "step": 1044
+    },
+    {
+      "epoch": 0.7070365358592693,
+      "grad_norm": 1.2494905498440645,
+      "learning_rate": 9.450555781116068e-06,
+      "loss": 0.5247,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7077131258457375,
+      "grad_norm": 1.2657195358378996,
+      "learning_rate": 9.448760215780218e-06,
+      "loss": 0.5301,
+      "step": 1046
+    },
+    {
+      "epoch": 0.7083897158322057,
+      "grad_norm": 1.252536940258962,
+      "learning_rate": 9.446961892452824e-06,
+      "loss": 0.517,
+      "step": 1047
+    },
+    {
+      "epoch": 0.7090663058186739,
+      "grad_norm": 1.2036919668954769,
+      "learning_rate": 9.445160812248754e-06,
+      "loss": 0.4953,
+      "step": 1048
+    },
+    {
+      "epoch": 0.709742895805142,
+      "grad_norm": 1.1974599304320643,
+      "learning_rate": 9.44335697628458e-06,
+      "loss": 0.5222,
+      "step": 1049
+    },
+    {
+      "epoch": 0.7104194857916103,
+      "grad_norm": 1.2640687097542616,
+      "learning_rate": 9.44155038567858e-06,
+      "loss": 0.5276,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7110960757780784,
+      "grad_norm": 1.26580964036148,
+      "learning_rate": 9.439741041550745e-06,
+      "loss": 0.5436,
+      "step": 1051
+    },
+    {
+      "epoch": 0.7117726657645467,
+      "grad_norm": 1.3520832943905736,
+      "learning_rate": 9.437928945022772e-06,
+      "loss": 0.5569,
+      "step": 1052
+    },
+    {
+      "epoch": 0.7124492557510149,
+      "grad_norm": 1.2471314764984125,
+      "learning_rate": 9.43611409721806e-06,
+      "loss": 0.5244,
+      "step": 1053
+    },
+    {
+      "epoch": 0.713125845737483,
+      "grad_norm": 1.274199489244252,
+      "learning_rate": 9.434296499261719e-06,
+      "loss": 0.5553,
+      "step": 1054
+    },
+    {
+      "epoch": 0.7138024357239513,
+      "grad_norm": 1.2611714641375649,
+      "learning_rate": 9.432476152280562e-06,
+      "loss": 0.5495,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7144790257104194,
+      "grad_norm": 1.2933516856333653,
+      "learning_rate": 9.430653057403105e-06,
+      "loss": 0.5214,
+      "step": 1056
+    },
+    {
+      "epoch": 0.7151556156968877,
+      "grad_norm": 1.2639444832070281,
+      "learning_rate": 9.428827215759569e-06,
+      "loss": 0.5409,
+      "step": 1057
+    },
+    {
+      "epoch": 0.7158322056833559,
+      "grad_norm": 1.2755101715166999,
+      "learning_rate": 9.426998628481876e-06,
+      "loss": 0.5045,
+      "step": 1058
+    },
+    {
+      "epoch": 0.7165087956698241,
+      "grad_norm": 1.2809343276489422,
+      "learning_rate": 9.425167296703655e-06,
+      "loss": 0.5021,
+      "step": 1059
+    },
+    {
+      "epoch": 0.7171853856562923,
+      "grad_norm": 1.270706769695537,
+      "learning_rate": 9.42333322156023e-06,
+      "loss": 0.5096,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7178619756427604,
+      "grad_norm": 1.2258189002444924,
+      "learning_rate": 9.42149640418863e-06,
+      "loss": 0.5203,
+      "step": 1061
+    },
+    {
+      "epoch": 0.7185385656292287,
+      "grad_norm": 1.2091660343486292,
+      "learning_rate": 9.419656845727582e-06,
+      "loss": 0.5109,
+      "step": 1062
+    },
+    {
+      "epoch": 0.7192151556156969,
+      "grad_norm": 1.2393047302711406,
+      "learning_rate": 9.417814547317513e-06,
+      "loss": 0.528,
+      "step": 1063
+    },
+    {
+      "epoch": 0.7198917456021651,
+      "grad_norm": 1.3022882402386513,
+      "learning_rate": 9.415969510100549e-06,
+      "loss": 0.5511,
+      "step": 1064
+    },
+    {
+      "epoch": 0.7205683355886333,
+      "grad_norm": 1.2218640795866045,
+      "learning_rate": 9.414121735220513e-06,
+      "loss": 0.5154,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7212449255751014,
+      "grad_norm": 1.2191141863190391,
+      "learning_rate": 9.412271223822929e-06,
+      "loss": 0.5215,
+      "step": 1066
+    },
+    {
+      "epoch": 0.7219215155615697,
+      "grad_norm": 1.2037212415605165,
+      "learning_rate": 9.41041797705501e-06,
+      "loss": 0.5179,
+      "step": 1067
+    },
+    {
+      "epoch": 0.7225981055480379,
+      "grad_norm": 1.289576805052528,
+      "learning_rate": 9.408561996065672e-06,
+      "loss": 0.5528,
+      "step": 1068
+    },
+    {
+      "epoch": 0.7232746955345061,
+      "grad_norm": 1.3073734499243528,
+      "learning_rate": 9.406703282005523e-06,
+      "loss": 0.529,
+      "step": 1069
+    },
+    {
+      "epoch": 0.7239512855209743,
+      "grad_norm": 1.3050448266339914,
+      "learning_rate": 9.404841836026863e-06,
+      "loss": 0.5329,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7246278755074425,
+      "grad_norm": 1.2290680927440425,
+      "learning_rate": 9.40297765928369e-06,
+      "loss": 0.5367,
+      "step": 1071
+    },
+    {
+      "epoch": 0.7253044654939107,
+      "grad_norm": 1.2240668245069777,
+      "learning_rate": 9.401110752931694e-06,
+      "loss": 0.531,
+      "step": 1072
+    },
+    {
+      "epoch": 0.725981055480379,
+      "grad_norm": 1.3473562899592888,
+      "learning_rate": 9.399241118128255e-06,
+      "loss": 0.531,
+      "step": 1073
+    },
+    {
+      "epoch": 0.7266576454668471,
+      "grad_norm": 1.2581370172967685,
+      "learning_rate": 9.397368756032445e-06,
+      "loss": 0.5321,
+      "step": 1074
+    },
+    {
+      "epoch": 0.7273342354533153,
+      "grad_norm": 1.2573947330764252,
+      "learning_rate": 9.395493667805032e-06,
+      "loss": 0.5193,
+      "step": 1075
+    },
+    {
+      "epoch": 0.7280108254397835,
+      "grad_norm": 1.2241008215655933,
+      "learning_rate": 9.393615854608461e-06,
+      "loss": 0.5122,
+      "step": 1076
+    },
+    {
+      "epoch": 0.7286874154262517,
+      "grad_norm": 1.1777671303740376,
+      "learning_rate": 9.391735317606885e-06,
+      "loss": 0.5104,
+      "step": 1077
+    },
+    {
+      "epoch": 0.7293640054127198,
+      "grad_norm": 1.197878175202554,
+      "learning_rate": 9.389852057966129e-06,
+      "loss": 0.5157,
+      "step": 1078
+    },
+    {
+      "epoch": 0.7300405953991881,
+      "grad_norm": 1.237944117266921,
+      "learning_rate": 9.387966076853714e-06,
+      "loss": 0.5188,
+      "step": 1079
+    },
+    {
+      "epoch": 0.7307171853856563,
+      "grad_norm": 1.2016964462933304,
+      "learning_rate": 9.386077375438848e-06,
+      "loss": 0.5008,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7313937753721245,
+      "grad_norm": 1.3030730797360548,
+      "learning_rate": 9.384185954892423e-06,
+      "loss": 0.5141,
+      "step": 1081
+    },
+    {
+      "epoch": 0.7320703653585927,
+      "grad_norm": 1.204426766247516,
+      "learning_rate": 9.382291816387018e-06,
+      "loss": 0.5177,
+      "step": 1082
+    },
+    {
+      "epoch": 0.7327469553450608,
+      "grad_norm": 1.239964128818739,
+      "learning_rate": 9.380394961096895e-06,
+      "loss": 0.5486,
+      "step": 1083
+    },
+    {
+      "epoch": 0.7334235453315291,
+      "grad_norm": 1.2519067594437676,
+      "learning_rate": 9.378495390198005e-06,
+      "loss": 0.5271,
+      "step": 1084
+    },
+    {
+      "epoch": 0.7341001353179973,
+      "grad_norm": 1.2018233212299891,
+      "learning_rate": 9.376593104867976e-06,
+      "loss": 0.5207,
+      "step": 1085
+    },
+    {
+      "epoch": 0.7347767253044655,
+      "grad_norm": 1.1511603822022256,
+      "learning_rate": 9.374688106286127e-06,
+      "loss": 0.4949,
+      "step": 1086
+    },
+    {
+      "epoch": 0.7354533152909337,
+      "grad_norm": 1.2957006130750042,
+      "learning_rate": 9.372780395633451e-06,
+      "loss": 0.5345,
+      "step": 1087
+    },
+    {
+      "epoch": 0.7361299052774019,
+      "grad_norm": 1.2723770737474842,
+      "learning_rate": 9.370869974092628e-06,
+      "loss": 0.5281,
+      "step": 1088
+    },
+    {
+      "epoch": 0.7368064952638701,
+      "grad_norm": 1.241815277191254,
+      "learning_rate": 9.368956842848014e-06,
+      "loss": 0.543,
+      "step": 1089
+    },
+    {
+      "epoch": 0.7374830852503383,
+      "grad_norm": 1.2032263815000845,
+      "learning_rate": 9.36704100308565e-06,
+      "loss": 0.492,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7381596752368065,
+      "grad_norm": 1.19514936869428,
+      "learning_rate": 9.36512245599325e-06,
+      "loss": 0.5199,
+      "step": 1091
+    },
+    {
+      "epoch": 0.7388362652232747,
+      "grad_norm": 1.2230982484528294,
+      "learning_rate": 9.363201202760212e-06,
+      "loss": 0.5248,
+      "step": 1092
+    },
+    {
+      "epoch": 0.7395128552097429,
+      "grad_norm": 1.2752199312029666,
+      "learning_rate": 9.36127724457761e-06,
+      "loss": 0.523,
+      "step": 1093
+    },
+    {
+      "epoch": 0.7401894451962111,
+      "grad_norm": 1.21153928005643,
+      "learning_rate": 9.359350582638193e-06,
+      "loss": 0.515,
+      "step": 1094
+    },
+    {
+      "epoch": 0.7408660351826793,
+      "grad_norm": 1.2424880078858398,
+      "learning_rate": 9.357421218136387e-06,
+      "loss": 0.5287,
+      "step": 1095
+    },
+    {
+      "epoch": 0.7415426251691475,
+      "grad_norm": 1.283084430351766,
+      "learning_rate": 9.355489152268296e-06,
+      "loss": 0.5345,
+      "step": 1096
+    },
+    {
+      "epoch": 0.7422192151556157,
+      "grad_norm": 1.16681955496085,
+      "learning_rate": 9.353554386231697e-06,
+      "loss": 0.4996,
+      "step": 1097
+    },
+    {
+      "epoch": 0.7428958051420839,
+      "grad_norm": 1.249942106329068,
+      "learning_rate": 9.351616921226036e-06,
+      "loss": 0.5522,
+      "step": 1098
+    },
+    {
+      "epoch": 0.7435723951285521,
+      "grad_norm": 1.245397417814634,
+      "learning_rate": 9.349676758452441e-06,
+      "loss": 0.5419,
+      "step": 1099
+    },
+    {
+      "epoch": 0.7442489851150202,
+      "grad_norm": 1.2050839062440089,
+      "learning_rate": 9.347733899113709e-06,
+      "loss": 0.5016,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7449255751014885,
+      "grad_norm": 1.2192149937877752,
+      "learning_rate": 9.345788344414306e-06,
+      "loss": 0.523,
+      "step": 1101
+    },
+    {
+      "epoch": 0.7456021650879567,
+      "grad_norm": 1.1860068074567807,
+      "learning_rate": 9.343840095560373e-06,
+      "loss": 0.5115,
+      "step": 1102
+    },
+    {
+      "epoch": 0.7462787550744249,
+      "grad_norm": 1.2506071981429465,
+      "learning_rate": 9.341889153759715e-06,
+      "loss": 0.5048,
+      "step": 1103
+    },
+    {
+      "epoch": 0.7469553450608931,
+      "grad_norm": 1.2416686227007272,
+      "learning_rate": 9.339935520221816e-06,
+      "loss": 0.5216,
+      "step": 1104
+    },
+    {
+      "epoch": 0.7476319350473613,
+      "grad_norm": 1.1918928798340078,
+      "learning_rate": 9.33797919615782e-06,
+      "loss": 0.5276,
+      "step": 1105
+    },
+    {
+      "epoch": 0.7483085250338295,
+      "grad_norm": 1.2608880099009294,
+      "learning_rate": 9.336020182780545e-06,
+      "loss": 0.5002,
+      "step": 1106
+    },
+    {
+      "epoch": 0.7489851150202977,
+      "grad_norm": 1.220186200328264,
+      "learning_rate": 9.33405848130447e-06,
+      "loss": 0.5308,
+      "step": 1107
+    },
+    {
+      "epoch": 0.7496617050067659,
+      "grad_norm": 1.2292827308408865,
+      "learning_rate": 9.332094092945749e-06,
+      "loss": 0.501,
+      "step": 1108
+    },
+    {
+      "epoch": 0.7503382949932341,
+      "grad_norm": 1.1640792670480113,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.4993,
+      "step": 1109
+    },
+    {
+      "epoch": 0.7510148849797023,
+      "grad_norm": 1.1442652537974987,
+      "learning_rate": 9.328157260453286e-06,
+      "loss": 0.5014,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7516914749661705,
+      "grad_norm": 1.2270186722583278,
+      "learning_rate": 9.326184818760167e-06,
+      "loss": 0.5132,
+      "step": 1111
+    },
+    {
+      "epoch": 0.7523680649526387,
+      "grad_norm": 1.1923161453599922,
+      "learning_rate": 9.324209695065644e-06,
+      "loss": 0.5212,
+      "step": 1112
+    },
+    {
+      "epoch": 0.7530446549391069,
+      "grad_norm": 1.2391194563334127,
+      "learning_rate": 9.322231890594193e-06,
+      "loss": 0.5451,
+      "step": 1113
+    },
+    {
+      "epoch": 0.7537212449255751,
+      "grad_norm": 1.2344078457049639,
+      "learning_rate": 9.32025140657194e-06,
+      "loss": 0.5251,
+      "step": 1114
+    },
+    {
+      "epoch": 0.7543978349120433,
+      "grad_norm": 1.2519285005335228,
+      "learning_rate": 9.318268244226681e-06,
+      "loss": 0.5056,
+      "step": 1115
+    },
+    {
+      "epoch": 0.7550744248985115,
+      "grad_norm": 1.242774477788238,
+      "learning_rate": 9.31628240478787e-06,
+      "loss": 0.5277,
+      "step": 1116
+    },
+    {
+      "epoch": 0.7557510148849798,
+      "grad_norm": 1.282012041641155,
+      "learning_rate": 9.31429388948662e-06,
+      "loss": 0.5167,
+      "step": 1117
+    },
+    {
+      "epoch": 0.7564276048714479,
+      "grad_norm": 1.184970210583831,
+      "learning_rate": 9.312302699555701e-06,
+      "loss": 0.5201,
+      "step": 1118
+    },
+    {
+      "epoch": 0.7571041948579161,
+      "grad_norm": 1.219400714159224,
+      "learning_rate": 9.310308836229548e-06,
+      "loss": 0.4936,
+      "step": 1119
+    },
+    {
+      "epoch": 0.7577807848443843,
+      "grad_norm": 1.1382113207645481,
+      "learning_rate": 9.308312300744247e-06,
+      "loss": 0.4951,
+      "step": 1120
+    },
+    {
+      "epoch": 0.7584573748308525,
+      "grad_norm": 1.167666202797914,
+      "learning_rate": 9.306313094337539e-06,
+      "loss": 0.4867,
+      "step": 1121
+    },
+    {
+      "epoch": 0.7591339648173207,
+      "grad_norm": 1.2387704994315876,
+      "learning_rate": 9.304311218248828e-06,
+      "loss": 0.4901,
+      "step": 1122
+    },
+    {
+      "epoch": 0.7598105548037889,
+      "grad_norm": 1.1316940775149187,
+      "learning_rate": 9.30230667371917e-06,
+      "loss": 0.4774,
+      "step": 1123
+    },
+    {
+      "epoch": 0.7604871447902571,
+      "grad_norm": 1.2058566340142174,
+      "learning_rate": 9.30029946199127e-06,
+      "loss": 0.52,
+      "step": 1124
+    },
+    {
+      "epoch": 0.7611637347767253,
+      "grad_norm": 1.243961618423197,
+      "learning_rate": 9.298289584309496e-06,
+      "loss": 0.5346,
+      "step": 1125
+    },
+    {
+      "epoch": 0.7618403247631935,
+      "grad_norm": 1.193442295510897,
+      "learning_rate": 9.29627704191986e-06,
+      "loss": 0.5065,
+      "step": 1126
+    },
+    {
+      "epoch": 0.7625169147496617,
+      "grad_norm": 1.2407973475588872,
+      "learning_rate": 9.294261836070034e-06,
+      "loss": 0.524,
+      "step": 1127
+    },
+    {
+      "epoch": 0.7631935047361299,
+      "grad_norm": 1.195632349448027,
+      "learning_rate": 9.292243968009332e-06,
+      "loss": 0.4739,
+      "step": 1128
+    },
+    {
+      "epoch": 0.7638700947225981,
+      "grad_norm": 1.1911527376522375,
+      "learning_rate": 9.290223438988726e-06,
+      "loss": 0.5221,
+      "step": 1129
+    },
+    {
+      "epoch": 0.7645466847090663,
+      "grad_norm": 1.129506736804369,
+      "learning_rate": 9.288200250260836e-06,
+      "loss": 0.4977,
+      "step": 1130
+    },
+    {
+      "epoch": 0.7652232746955345,
+      "grad_norm": 1.3476559038046902,
+      "learning_rate": 9.286174403079928e-06,
+      "loss": 0.5335,
+      "step": 1131
+    },
+    {
+      "epoch": 0.7658998646820027,
+      "grad_norm": 1.2656781723573305,
+      "learning_rate": 9.284145898701921e-06,
+      "loss": 0.5298,
+      "step": 1132
+    },
+    {
+      "epoch": 0.7665764546684709,
+      "grad_norm": 1.2362237441645285,
+      "learning_rate": 9.282114738384375e-06,
+      "loss": 0.5098,
+      "step": 1133
+    },
+    {
+      "epoch": 0.7672530446549392,
+      "grad_norm": 1.2091333429953386,
+      "learning_rate": 9.280080923386501e-06,
+      "loss": 0.5041,
+      "step": 1134
+    },
+    {
+      "epoch": 0.7679296346414073,
+      "grad_norm": 1.1647454812252227,
+      "learning_rate": 9.278044454969157e-06,
+      "loss": 0.4838,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7686062246278755,
+      "grad_norm": 1.2327622765574568,
+      "learning_rate": 9.27600533439484e-06,
+      "loss": 0.4997,
+      "step": 1136
+    },
+    {
+      "epoch": 0.7692828146143437,
+      "grad_norm": 1.205122408855429,
+      "learning_rate": 9.273963562927695e-06,
+      "loss": 0.5062,
+      "step": 1137
+    },
+    {
+      "epoch": 0.7699594046008119,
+      "grad_norm": 1.2455164972733845,
+      "learning_rate": 9.271919141833514e-06,
+      "loss": 0.5077,
+      "step": 1138
+    },
+    {
+      "epoch": 0.7706359945872802,
+      "grad_norm": 1.2212308301938604,
+      "learning_rate": 9.269872072379725e-06,
+      "loss": 0.4984,
+      "step": 1139
+    },
+    {
+      "epoch": 0.7713125845737483,
+      "grad_norm": 1.2614884328893134,
+      "learning_rate": 9.267822355835402e-06,
+      "loss": 0.5473,
+      "step": 1140
+    },
+    {
+      "epoch": 0.7719891745602165,
+      "grad_norm": 1.245191993873177,
+      "learning_rate": 9.265769993471258e-06,
+      "loss": 0.5277,
+      "step": 1141
+    },
+    {
+      "epoch": 0.7726657645466847,
+      "grad_norm": 1.184237449224601,
+      "learning_rate": 9.263714986559647e-06,
+      "loss": 0.5228,
+      "step": 1142
+    },
+    {
+      "epoch": 0.7733423545331529,
+      "grad_norm": 1.2988147104890695,
+      "learning_rate": 9.261657336374561e-06,
+      "loss": 0.5164,
+      "step": 1143
+    },
+    {
+      "epoch": 0.774018944519621,
+      "grad_norm": 1.258595652871956,
+      "learning_rate": 9.259597044191635e-06,
+      "loss": 0.5378,
+      "step": 1144
+    },
+    {
+      "epoch": 0.7746955345060893,
+      "grad_norm": 1.1255557115022607,
+      "learning_rate": 9.25753411128814e-06,
+      "loss": 0.4567,
+      "step": 1145
+    },
+    {
+      "epoch": 0.7753721244925575,
+      "grad_norm": 1.2610790509117351,
+      "learning_rate": 9.25546853894298e-06,
+      "loss": 0.5242,
+      "step": 1146
+    },
+    {
+      "epoch": 0.7760487144790257,
+      "grad_norm": 1.2589644527665236,
+      "learning_rate": 9.253400328436699e-06,
+      "loss": 0.5219,
+      "step": 1147
+    },
+    {
+      "epoch": 0.7767253044654939,
+      "grad_norm": 1.184495299442687,
+      "learning_rate": 9.251329481051476e-06,
+      "loss": 0.5092,
+      "step": 1148
+    },
+    {
+      "epoch": 0.7774018944519621,
+      "grad_norm": 1.1782965516351767,
+      "learning_rate": 9.249255998071127e-06,
+      "loss": 0.5006,
+      "step": 1149
+    },
+    {
+      "epoch": 0.7780784844384303,
+      "grad_norm": 1.1678224212638866,
+      "learning_rate": 9.247179880781099e-06,
+      "loss": 0.5077,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7787550744248986,
+      "grad_norm": 1.172336312307818,
+      "learning_rate": 9.24510113046847e-06,
+      "loss": 0.4821,
+      "step": 1151
+    },
+    {
+      "epoch": 0.7794316644113667,
+      "grad_norm": 1.2326813343485412,
+      "learning_rate": 9.243019748421956e-06,
+      "loss": 0.5238,
+      "step": 1152
+    },
+    {
+      "epoch": 0.7801082543978349,
+      "grad_norm": 1.253336584382025,
+      "learning_rate": 9.2409357359319e-06,
+      "loss": 0.5201,
+      "step": 1153
+    },
+    {
+      "epoch": 0.7807848443843031,
+      "grad_norm": 1.2715452410014232,
+      "learning_rate": 9.238849094290279e-06,
+      "loss": 0.5173,
+      "step": 1154
+    },
+    {
+      "epoch": 0.7814614343707713,
+      "grad_norm": 1.2043169935061349,
+      "learning_rate": 9.236759824790698e-06,
+      "loss": 0.4971,
+      "step": 1155
+    },
+    {
+      "epoch": 0.7821380243572396,
+      "grad_norm": 1.265740963751011,
+      "learning_rate": 9.234667928728392e-06,
+      "loss": 0.5015,
+      "step": 1156
+    },
+    {
+      "epoch": 0.7828146143437077,
+      "grad_norm": 1.1540576910730072,
+      "learning_rate": 9.23257340740022e-06,
+      "loss": 0.5,
+      "step": 1157
+    },
+    {
+      "epoch": 0.7834912043301759,
+      "grad_norm": 1.2321409258359763,
+      "learning_rate": 9.230476262104678e-06,
+      "loss": 0.5327,
+      "step": 1158
+    },
+    {
+      "epoch": 0.7841677943166441,
+      "grad_norm": 1.2198752858863542,
+      "learning_rate": 9.22837649414188e-06,
+      "loss": 0.5162,
+      "step": 1159
+    },
+    {
+      "epoch": 0.7848443843031123,
+      "grad_norm": 1.2086741821442035,
+      "learning_rate": 9.226274104813567e-06,
+      "loss": 0.5266,
+      "step": 1160
+    },
+    {
+      "epoch": 0.7855209742895806,
+      "grad_norm": 1.2269901761735669,
+      "learning_rate": 9.22416909542311e-06,
+      "loss": 0.5064,
+      "step": 1161
+    },
+    {
+      "epoch": 0.7861975642760487,
+      "grad_norm": 1.1272107653531314,
+      "learning_rate": 9.222061467275503e-06,
+      "loss": 0.5056,
+      "step": 1162
+    },
+    {
+      "epoch": 0.786874154262517,
+      "grad_norm": 1.1682273139005261,
+      "learning_rate": 9.219951221677356e-06,
+      "loss": 0.5057,
+      "step": 1163
+    },
+    {
+      "epoch": 0.7875507442489851,
+      "grad_norm": 1.2047454363953354,
+      "learning_rate": 9.217838359936914e-06,
+      "loss": 0.4973,
+      "step": 1164
+    },
+    {
+      "epoch": 0.7882273342354533,
+      "grad_norm": 1.1996132345158188,
+      "learning_rate": 9.215722883364033e-06,
+      "loss": 0.5061,
+      "step": 1165
+    },
+    {
+      "epoch": 0.7889039242219216,
+      "grad_norm": 1.2175204839574072,
+      "learning_rate": 9.213604793270196e-06,
+      "loss": 0.5087,
+      "step": 1166
+    },
+    {
+      "epoch": 0.7895805142083897,
+      "grad_norm": 1.163682009112604,
+      "learning_rate": 9.211484090968505e-06,
+      "loss": 0.496,
+      "step": 1167
+    },
+    {
+      "epoch": 0.790257104194858,
+      "grad_norm": 1.1551956368039216,
+      "learning_rate": 9.20936077777368e-06,
+      "loss": 0.5006,
+      "step": 1168
+    },
+    {
+      "epoch": 0.7909336941813261,
+      "grad_norm": 1.286480056932424,
+      "learning_rate": 9.207234855002062e-06,
+      "loss": 0.5427,
+      "step": 1169
+    },
+    {
+      "epoch": 0.7916102841677943,
+      "grad_norm": 1.2416731588400851,
+      "learning_rate": 9.205106323971607e-06,
+      "loss": 0.5128,
+      "step": 1170
+    },
+    {
+      "epoch": 0.7922868741542625,
+      "grad_norm": 1.2035277303836733,
+      "learning_rate": 9.202975186001892e-06,
+      "loss": 0.5028,
+      "step": 1171
+    },
+    {
+      "epoch": 0.7929634641407307,
+      "grad_norm": 1.172801805095853,
+      "learning_rate": 9.200841442414106e-06,
+      "loss": 0.5045,
+      "step": 1172
+    },
+    {
+      "epoch": 0.793640054127199,
+      "grad_norm": 1.2246920957554206,
+      "learning_rate": 9.198705094531053e-06,
+      "loss": 0.5215,
+      "step": 1173
+    },
+    {
+      "epoch": 0.7943166441136671,
+      "grad_norm": 1.2210194718468,
+      "learning_rate": 9.196566143677157e-06,
+      "loss": 0.5161,
+      "step": 1174
+    },
+    {
+      "epoch": 0.7949932341001353,
+      "grad_norm": 1.2207221685901417,
+      "learning_rate": 9.19442459117845e-06,
+      "loss": 0.5358,
+      "step": 1175
+    },
+    {
+      "epoch": 0.7956698240866035,
+      "grad_norm": 1.1836268293335501,
+      "learning_rate": 9.192280438362581e-06,
+      "loss": 0.4876,
+      "step": 1176
+    },
+    {
+      "epoch": 0.7963464140730717,
+      "grad_norm": 1.166241983031272,
+      "learning_rate": 9.190133686558809e-06,
+      "loss": 0.4922,
+      "step": 1177
+    },
+    {
+      "epoch": 0.79702300405954,
+      "grad_norm": 1.1556650614887145,
+      "learning_rate": 9.187984337098002e-06,
+      "loss": 0.5073,
+      "step": 1178
+    },
+    {
+      "epoch": 0.7976995940460081,
+      "grad_norm": 1.1921213306616103,
+      "learning_rate": 9.185832391312644e-06,
+      "loss": 0.5091,
+      "step": 1179
+    },
+    {
+      "epoch": 0.7983761840324763,
+      "grad_norm": 1.214325909552027,
+      "learning_rate": 9.183677850536823e-06,
+      "loss": 0.5153,
+      "step": 1180
+    },
+    {
+      "epoch": 0.7990527740189445,
+      "grad_norm": 1.1979689868096728,
+      "learning_rate": 9.181520716106238e-06,
+      "loss": 0.5019,
+      "step": 1181
+    },
+    {
+      "epoch": 0.7997293640054127,
+      "grad_norm": 1.1835623771156139,
+      "learning_rate": 9.179360989358199e-06,
+      "loss": 0.5011,
+      "step": 1182
+    },
+    {
+      "epoch": 0.800405953991881,
+      "grad_norm": 1.1896726896387138,
+      "learning_rate": 9.177198671631616e-06,
+      "loss": 0.5043,
+      "step": 1183
+    },
+    {
+      "epoch": 0.8010825439783491,
+      "grad_norm": 1.1453404765067188,
+      "learning_rate": 9.175033764267013e-06,
+      "loss": 0.4804,
+      "step": 1184
+    },
+    {
+      "epoch": 0.8017591339648173,
+      "grad_norm": 1.1653256478294909,
+      "learning_rate": 9.172866268606514e-06,
+      "loss": 0.5065,
+      "step": 1185
+    },
+    {
+      "epoch": 0.8024357239512855,
+      "grad_norm": 1.1459358079448239,
+      "learning_rate": 9.17069618599385e-06,
+      "loss": 0.5052,
+      "step": 1186
+    },
+    {
+      "epoch": 0.8031123139377537,
+      "grad_norm": 1.141288346093606,
+      "learning_rate": 9.168523517774356e-06,
+      "loss": 0.5004,
+      "step": 1187
+    },
+    {
+      "epoch": 0.803788903924222,
+      "grad_norm": 1.2081823619592962,
+      "learning_rate": 9.166348265294968e-06,
+      "loss": 0.5044,
+      "step": 1188
+    },
+    {
+      "epoch": 0.8044654939106901,
+      "grad_norm": 1.161845597498268,
+      "learning_rate": 9.164170429904224e-06,
+      "loss": 0.5178,
+      "step": 1189
+    },
+    {
+      "epoch": 0.8051420838971584,
+      "grad_norm": 1.1575450572593282,
+      "learning_rate": 9.16199001295227e-06,
+      "loss": 0.5085,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8058186738836265,
+      "grad_norm": 1.201488524435077,
+      "learning_rate": 9.15980701579084e-06,
+      "loss": 0.5055,
+      "step": 1191
+    },
+    {
+      "epoch": 0.8064952638700947,
+      "grad_norm": 1.1885579926096035,
+      "learning_rate": 9.157621439773278e-06,
+      "loss": 0.4957,
+      "step": 1192
+    },
+    {
+      "epoch": 0.8071718538565629,
+      "grad_norm": 1.121439214618471,
+      "learning_rate": 9.155433286254524e-06,
+      "loss": 0.4931,
+      "step": 1193
+    },
+    {
+      "epoch": 0.8078484438430311,
+      "grad_norm": 1.2044542210279365,
+      "learning_rate": 9.153242556591115e-06,
+      "loss": 0.504,
+      "step": 1194
+    },
+    {
+      "epoch": 0.8085250338294994,
+      "grad_norm": 1.1627187962465653,
+      "learning_rate": 9.151049252141185e-06,
+      "loss": 0.4976,
+      "step": 1195
+    },
+    {
+      "epoch": 0.8092016238159675,
+      "grad_norm": 1.136773615776934,
+      "learning_rate": 9.148853374264463e-06,
+      "loss": 0.4992,
+      "step": 1196
+    },
+    {
+      "epoch": 0.8098782138024357,
+      "grad_norm": 1.154489603413787,
+      "learning_rate": 9.146654924322277e-06,
+      "loss": 0.5052,
+      "step": 1197
+    },
+    {
+      "epoch": 0.8105548037889039,
+      "grad_norm": 1.158757882340262,
+      "learning_rate": 9.144453903677546e-06,
+      "loss": 0.4983,
+      "step": 1198
+    },
+    {
+      "epoch": 0.8112313937753721,
+      "grad_norm": 1.170001361279437,
+      "learning_rate": 9.142250313694785e-06,
+      "loss": 0.5098,
+      "step": 1199
+    },
+    {
+      "epoch": 0.8119079837618404,
+      "grad_norm": 1.198423599885229,
+      "learning_rate": 9.140044155740102e-06,
+      "loss": 0.4989,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8125845737483085,
+      "grad_norm": 1.2762899051831884,
+      "learning_rate": 9.137835431181192e-06,
+      "loss": 0.5235,
+      "step": 1201
+    },
+    {
+      "epoch": 0.8132611637347767,
+      "grad_norm": 1.1672888453026706,
+      "learning_rate": 9.13562414138735e-06,
+      "loss": 0.4845,
+      "step": 1202
+    },
+    {
+      "epoch": 0.8139377537212449,
+      "grad_norm": 1.1916703372146515,
+      "learning_rate": 9.133410287729454e-06,
+      "loss": 0.5103,
+      "step": 1203
+    },
+    {
+      "epoch": 0.8146143437077131,
+      "grad_norm": 1.1707987920004868,
+      "learning_rate": 9.131193871579975e-06,
+      "loss": 0.5033,
+      "step": 1204
+    },
+    {
+      "epoch": 0.8152909336941814,
+      "grad_norm": 1.167866774362817,
+      "learning_rate": 9.12897489431297e-06,
+      "loss": 0.5102,
+      "step": 1205
+    },
+    {
+      "epoch": 0.8159675236806495,
+      "grad_norm": 1.1591757897777635,
+      "learning_rate": 9.126753357304088e-06,
+      "loss": 0.5005,
+      "step": 1206
+    },
+    {
+      "epoch": 0.8166441136671178,
+      "grad_norm": 1.134558974635681,
+      "learning_rate": 9.12452926193056e-06,
+      "loss": 0.4934,
+      "step": 1207
+    },
+    {
+      "epoch": 0.8173207036535859,
+      "grad_norm": 1.176088029742004,
+      "learning_rate": 9.122302609571204e-06,
+      "loss": 0.5012,
+      "step": 1208
+    },
+    {
+      "epoch": 0.8179972936400541,
+      "grad_norm": 1.1784767687786055,
+      "learning_rate": 9.120073401606427e-06,
+      "loss": 0.5136,
+      "step": 1209
+    },
+    {
+      "epoch": 0.8186738836265224,
+      "grad_norm": 1.1933785692052794,
+      "learning_rate": 9.117841639418218e-06,
+      "loss": 0.506,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8193504736129905,
+      "grad_norm": 1.1658587414583894,
+      "learning_rate": 9.115607324390146e-06,
+      "loss": 0.4713,
+      "step": 1211
+    },
+    {
+      "epoch": 0.8200270635994588,
+      "grad_norm": 1.2206670194557647,
+      "learning_rate": 9.11337045790737e-06,
+      "loss": 0.5055,
+      "step": 1212
+    },
+    {
+      "epoch": 0.8207036535859269,
+      "grad_norm": 1.1596600395068082,
+      "learning_rate": 9.111131041356624e-06,
+      "loss": 0.5046,
+      "step": 1213
+    },
+    {
+      "epoch": 0.8213802435723951,
+      "grad_norm": 1.2417732230382863,
+      "learning_rate": 9.108889076126226e-06,
+      "loss": 0.5111,
+      "step": 1214
+    },
+    {
+      "epoch": 0.8220568335588633,
+      "grad_norm": 1.1095547390157674,
+      "learning_rate": 9.106644563606076e-06,
+      "loss": 0.5052,
+      "step": 1215
+    },
+    {
+      "epoch": 0.8227334235453315,
+      "grad_norm": 1.1437208716968439,
+      "learning_rate": 9.104397505187645e-06,
+      "loss": 0.4926,
+      "step": 1216
+    },
+    {
+      "epoch": 0.8234100135317998,
+      "grad_norm": 1.1504515072112609,
+      "learning_rate": 9.102147902263994e-06,
+      "loss": 0.478,
+      "step": 1217
+    },
+    {
+      "epoch": 0.8240866035182679,
+      "grad_norm": 1.1443597297700818,
+      "learning_rate": 9.099895756229754e-06,
+      "loss": 0.5202,
+      "step": 1218
+    },
+    {
+      "epoch": 0.8247631935047361,
+      "grad_norm": 1.0841569991671136,
+      "learning_rate": 9.097641068481133e-06,
+      "loss": 0.4873,
+      "step": 1219
+    },
+    {
+      "epoch": 0.8254397834912043,
+      "grad_norm": 1.184769872454869,
+      "learning_rate": 9.095383840415915e-06,
+      "loss": 0.5065,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8261163734776725,
+      "grad_norm": 1.2340019807532674,
+      "learning_rate": 9.093124073433464e-06,
+      "loss": 0.5045,
+      "step": 1221
+    },
+    {
+      "epoch": 0.8267929634641408,
+      "grad_norm": 1.1484478756692174,
+      "learning_rate": 9.090861768934708e-06,
+      "loss": 0.4832,
+      "step": 1222
+    },
+    {
+      "epoch": 0.8274695534506089,
+      "grad_norm": 1.215347927222981,
+      "learning_rate": 9.088596928322158e-06,
+      "loss": 0.5044,
+      "step": 1223
+    },
+    {
+      "epoch": 0.8281461434370772,
+      "grad_norm": 1.2136813200888974,
+      "learning_rate": 9.08632955299989e-06,
+      "loss": 0.5075,
+      "step": 1224
+    },
+    {
+      "epoch": 0.8288227334235453,
+      "grad_norm": 1.21118000221307,
+      "learning_rate": 9.084059644373558e-06,
+      "loss": 0.502,
+      "step": 1225
+    },
+    {
+      "epoch": 0.8294993234100135,
+      "grad_norm": 1.1992669669850768,
+      "learning_rate": 9.08178720385038e-06,
+      "loss": 0.4914,
+      "step": 1226
+    },
+    {
+      "epoch": 0.8301759133964818,
+      "grad_norm": 1.159224784891141,
+      "learning_rate": 9.07951223283915e-06,
+      "loss": 0.4941,
+      "step": 1227
+    },
+    {
+      "epoch": 0.8308525033829499,
+      "grad_norm": 1.152140367381401,
+      "learning_rate": 9.077234732750223e-06,
+      "loss": 0.4862,
+      "step": 1228
+    },
+    {
+      "epoch": 0.8315290933694182,
+      "grad_norm": 1.2178795785849492,
+      "learning_rate": 9.074954704995532e-06,
+      "loss": 0.5277,
+      "step": 1229
+    },
+    {
+      "epoch": 0.8322056833558863,
+      "grad_norm": 1.178803162593111,
+      "learning_rate": 9.072672150988563e-06,
+      "loss": 0.5019,
+      "step": 1230
+    },
+    {
+      "epoch": 0.8328822733423545,
+      "grad_norm": 1.084268178799295,
+      "learning_rate": 9.070387072144386e-06,
+      "loss": 0.4917,
+      "step": 1231
+    },
+    {
+      "epoch": 0.8335588633288228,
+      "grad_norm": 1.2088539096551216,
+      "learning_rate": 9.06809946987962e-06,
+      "loss": 0.5342,
+      "step": 1232
+    },
+    {
+      "epoch": 0.8342354533152909,
+      "grad_norm": 1.1143419493443054,
+      "learning_rate": 9.065809345612458e-06,
+      "loss": 0.4885,
+      "step": 1233
+    },
+    {
+      "epoch": 0.8349120433017592,
+      "grad_norm": 1.1393488661081579,
+      "learning_rate": 9.06351670076265e-06,
+      "loss": 0.4705,
+      "step": 1234
+    },
+    {
+      "epoch": 0.8355886332882273,
+      "grad_norm": 1.150681280569992,
+      "learning_rate": 9.061221536751517e-06,
+      "loss": 0.498,
+      "step": 1235
+    },
+    {
+      "epoch": 0.8362652232746955,
+      "grad_norm": 1.1133905701950872,
+      "learning_rate": 9.058923855001935e-06,
+      "loss": 0.4846,
+      "step": 1236
+    },
+    {
+      "epoch": 0.8369418132611637,
+      "grad_norm": 1.1451489378732804,
+      "learning_rate": 9.056623656938344e-06,
+      "loss": 0.4815,
+      "step": 1237
+    },
+    {
+      "epoch": 0.8376184032476319,
+      "grad_norm": 1.163503758049376,
+      "learning_rate": 9.05432094398674e-06,
+      "loss": 0.4907,
+      "step": 1238
+    },
+    {
+      "epoch": 0.8382949932341002,
+      "grad_norm": 1.231452882592344,
+      "learning_rate": 9.052015717574683e-06,
+      "loss": 0.5093,
+      "step": 1239
+    },
+    {
+      "epoch": 0.8389715832205683,
+      "grad_norm": 1.1863568491997516,
+      "learning_rate": 9.049707979131288e-06,
+      "loss": 0.5195,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8396481732070366,
+      "grad_norm": 1.1659036571243266,
+      "learning_rate": 9.04739773008723e-06,
+      "loss": 0.4938,
+      "step": 1241
+    },
+    {
+      "epoch": 0.8403247631935047,
+      "grad_norm": 1.1394606038594781,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.4942,
+      "step": 1242
+    },
+    {
+      "epoch": 0.8410013531799729,
+      "grad_norm": 1.1688212688465167,
+      "learning_rate": 9.042769705927597e-06,
+      "loss": 0.5065,
+      "step": 1243
+    },
+    {
+      "epoch": 0.8416779431664412,
+      "grad_norm": 1.1400435992590443,
+      "learning_rate": 9.040451933681148e-06,
+      "loss": 0.5017,
+      "step": 1244
+    },
+    {
+      "epoch": 0.8423545331529093,
+      "grad_norm": 1.1393080053253095,
+      "learning_rate": 9.038131656572284e-06,
+      "loss": 0.4949,
+      "step": 1245
+    },
+    {
+      "epoch": 0.8430311231393776,
+      "grad_norm": 1.1392590264755564,
+      "learning_rate": 9.035808876039451e-06,
+      "loss": 0.5007,
+      "step": 1246
+    },
+    {
+      "epoch": 0.8437077131258457,
+      "grad_norm": 1.08930459707302,
+      "learning_rate": 9.033483593522652e-06,
+      "loss": 0.4833,
+      "step": 1247
+    },
+    {
+      "epoch": 0.8443843031123139,
+      "grad_norm": 1.1060110312753122,
+      "learning_rate": 9.03115581046343e-06,
+      "loss": 0.4859,
+      "step": 1248
+    },
+    {
+      "epoch": 0.8450608930987822,
+      "grad_norm": 1.1427215978050838,
+      "learning_rate": 9.028825528304892e-06,
+      "loss": 0.4952,
+      "step": 1249
+    },
+    {
+      "epoch": 0.8457374830852503,
+      "grad_norm": 1.1074913254567567,
+      "learning_rate": 9.026492748491683e-06,
+      "loss": 0.4836,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8464140730717186,
+      "grad_norm": 1.1303369847573572,
+      "learning_rate": 9.02415747247e-06,
+      "loss": 0.4955,
+      "step": 1251
+    },
+    {
+      "epoch": 0.8470906630581867,
+      "grad_norm": 1.1404785046476127,
+      "learning_rate": 9.02181970168759e-06,
+      "loss": 0.4996,
+      "step": 1252
+    },
+    {
+      "epoch": 0.847767253044655,
+      "grad_norm": 1.157184953961382,
+      "learning_rate": 9.019479437593748e-06,
+      "loss": 0.5111,
+      "step": 1253
+    },
+    {
+      "epoch": 0.8484438430311232,
+      "grad_norm": 1.1346714000625626,
+      "learning_rate": 9.017136681639307e-06,
+      "loss": 0.4995,
+      "step": 1254
+    },
+    {
+      "epoch": 0.8491204330175913,
+      "grad_norm": 1.1129518663288842,
+      "learning_rate": 9.014791435276651e-06,
+      "loss": 0.4679,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8497970230040596,
+      "grad_norm": 1.1332449083138452,
+      "learning_rate": 9.012443699959706e-06,
+      "loss": 0.4596,
+      "step": 1256
+    },
+    {
+      "epoch": 0.8504736129905277,
+      "grad_norm": 1.1928224747810408,
+      "learning_rate": 9.010093477143942e-06,
+      "loss": 0.4868,
+      "step": 1257
+    },
+    {
+      "epoch": 0.851150202976996,
+      "grad_norm": 1.1330133012179797,
+      "learning_rate": 9.007740768286369e-06,
+      "loss": 0.4745,
+      "step": 1258
+    },
+    {
+      "epoch": 0.8518267929634641,
+      "grad_norm": 1.1022108256784344,
+      "learning_rate": 9.005385574845543e-06,
+      "loss": 0.4702,
+      "step": 1259
+    },
+    {
+      "epoch": 0.8525033829499323,
+      "grad_norm": 1.1663024420978705,
+      "learning_rate": 9.003027898281551e-06,
+      "loss": 0.5006,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8531799729364006,
+      "grad_norm": 1.1614830762105077,
+      "learning_rate": 9.000667740056033e-06,
+      "loss": 0.4788,
+      "step": 1261
+    },
+    {
+      "epoch": 0.8538565629228687,
+      "grad_norm": 1.1599210618285292,
+      "learning_rate": 8.998305101632155e-06,
+      "loss": 0.5145,
+      "step": 1262
+    },
+    {
+      "epoch": 0.854533152909337,
+      "grad_norm": 1.1997132116001858,
+      "learning_rate": 8.995939984474624e-06,
+      "loss": 0.522,
+      "step": 1263
+    },
+    {
+      "epoch": 0.8552097428958051,
+      "grad_norm": 1.1647077190927393,
+      "learning_rate": 8.99357239004969e-06,
+      "loss": 0.5069,
+      "step": 1264
+    },
+    {
+      "epoch": 0.8558863328822733,
+      "grad_norm": 1.1597263193249239,
+      "learning_rate": 8.991202319825131e-06,
+      "loss": 0.5082,
+      "step": 1265
+    },
+    {
+      "epoch": 0.8565629228687416,
+      "grad_norm": 1.1767426661006315,
+      "learning_rate": 8.988829775270265e-06,
+      "loss": 0.4927,
+      "step": 1266
+    },
+    {
+      "epoch": 0.8572395128552097,
+      "grad_norm": 1.1537489991545355,
+      "learning_rate": 8.986454757855938e-06,
+      "loss": 0.5006,
+      "step": 1267
+    },
+    {
+      "epoch": 0.857916102841678,
+      "grad_norm": 1.2096771024003516,
+      "learning_rate": 8.984077269054535e-06,
+      "loss": 0.5222,
+      "step": 1268
+    },
+    {
+      "epoch": 0.8585926928281461,
+      "grad_norm": 1.2327653148038258,
+      "learning_rate": 8.981697310339972e-06,
+      "loss": 0.4996,
+      "step": 1269
+    },
+    {
+      "epoch": 0.8592692828146143,
+      "grad_norm": 1.152318472987713,
+      "learning_rate": 8.979314883187694e-06,
+      "loss": 0.5223,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8599458728010826,
+      "grad_norm": 1.1380987339073596,
+      "learning_rate": 8.976929989074677e-06,
+      "loss": 0.4876,
+      "step": 1271
+    },
+    {
+      "epoch": 0.8606224627875507,
+      "grad_norm": 1.1091524025592998,
+      "learning_rate": 8.974542629479426e-06,
+      "loss": 0.5002,
+      "step": 1272
+    },
+    {
+      "epoch": 0.861299052774019,
+      "grad_norm": 1.213865438308819,
+      "learning_rate": 8.972152805881978e-06,
+      "loss": 0.5077,
+      "step": 1273
+    },
+    {
+      "epoch": 0.8619756427604871,
+      "grad_norm": 1.240209755681695,
+      "learning_rate": 8.969760519763891e-06,
+      "loss": 0.5141,
+      "step": 1274
+    },
+    {
+      "epoch": 0.8626522327469553,
+      "grad_norm": 1.0695291806021396,
+      "learning_rate": 8.967365772608258e-06,
+      "loss": 0.4657,
+      "step": 1275
+    },
+    {
+      "epoch": 0.8633288227334236,
+      "grad_norm": 1.0946797927801621,
+      "learning_rate": 8.96496856589969e-06,
+      "loss": 0.4994,
+      "step": 1276
+    },
+    {
+      "epoch": 0.8640054127198917,
+      "grad_norm": 1.1415766090022756,
+      "learning_rate": 8.962568901124326e-06,
+      "loss": 0.5136,
+      "step": 1277
+    },
+    {
+      "epoch": 0.86468200270636,
+      "grad_norm": 1.1298423136870914,
+      "learning_rate": 8.96016677976983e-06,
+      "loss": 0.4915,
+      "step": 1278
+    },
+    {
+      "epoch": 0.8653585926928281,
+      "grad_norm": 1.1243889547151007,
+      "learning_rate": 8.957762203325389e-06,
+      "loss": 0.4997,
+      "step": 1279
+    },
+    {
+      "epoch": 0.8660351826792964,
+      "grad_norm": 1.0795482679115407,
+      "learning_rate": 8.955355173281709e-06,
+      "loss": 0.495,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8667117726657646,
+      "grad_norm": 1.140820850967581,
+      "learning_rate": 8.952945691131016e-06,
+      "loss": 0.4894,
+      "step": 1281
+    },
+    {
+      "epoch": 0.8673883626522327,
+      "grad_norm": 1.151613276845393,
+      "learning_rate": 8.950533758367063e-06,
+      "loss": 0.5205,
+      "step": 1282
+    },
+    {
+      "epoch": 0.868064952638701,
+      "grad_norm": 1.1590996796684934,
+      "learning_rate": 8.948119376485119e-06,
+      "loss": 0.4832,
+      "step": 1283
+    },
+    {
+      "epoch": 0.8687415426251691,
+      "grad_norm": 1.110183660912312,
+      "learning_rate": 8.94570254698197e-06,
+      "loss": 0.4669,
+      "step": 1284
+    },
+    {
+      "epoch": 0.8694181326116374,
+      "grad_norm": 1.1112549714756523,
+      "learning_rate": 8.943283271355915e-06,
+      "loss": 0.5022,
+      "step": 1285
+    },
+    {
+      "epoch": 0.8700947225981055,
+      "grad_norm": 1.1418881471068483,
+      "learning_rate": 8.940861551106784e-06,
+      "loss": 0.4926,
+      "step": 1286
+    },
+    {
+      "epoch": 0.8707713125845737,
+      "grad_norm": 1.1118737025422658,
+      "learning_rate": 8.938437387735903e-06,
+      "loss": 0.4622,
+      "step": 1287
+    },
+    {
+      "epoch": 0.871447902571042,
+      "grad_norm": 1.1881042509588093,
+      "learning_rate": 8.93601078274613e-06,
+      "loss": 0.5061,
+      "step": 1288
+    },
+    {
+      "epoch": 0.8721244925575101,
+      "grad_norm": 1.1247767210782038,
+      "learning_rate": 8.933581737641824e-06,
+      "loss": 0.4938,
+      "step": 1289
+    },
+    {
+      "epoch": 0.8728010825439784,
+      "grad_norm": 1.1390569526887053,
+      "learning_rate": 8.931150253928866e-06,
+      "loss": 0.5012,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8734776725304465,
+      "grad_norm": 1.1035813141538635,
+      "learning_rate": 8.928716333114643e-06,
+      "loss": 0.4891,
+      "step": 1291
+    },
+    {
+      "epoch": 0.8741542625169147,
+      "grad_norm": 1.056265978387651,
+      "learning_rate": 8.926279976708056e-06,
+      "loss": 0.4493,
+      "step": 1292
+    },
+    {
+      "epoch": 0.874830852503383,
+      "grad_norm": 1.1386457355437434,
+      "learning_rate": 8.923841186219512e-06,
+      "loss": 0.5046,
+      "step": 1293
+    },
+    {
+      "epoch": 0.8755074424898511,
+      "grad_norm": 1.111246334626937,
+      "learning_rate": 8.921399963160934e-06,
+      "loss": 0.492,
+      "step": 1294
+    },
+    {
+      "epoch": 0.8761840324763194,
+      "grad_norm": 1.1413217971002307,
+      "learning_rate": 8.918956309045743e-06,
+      "loss": 0.4949,
+      "step": 1295
+    },
+    {
+      "epoch": 0.8768606224627875,
+      "grad_norm": 1.0499552388509685,
+      "learning_rate": 8.916510225388878e-06,
+      "loss": 0.4783,
+      "step": 1296
+    },
+    {
+      "epoch": 0.8775372124492558,
+      "grad_norm": 1.0885209774631313,
+      "learning_rate": 8.914061713706776e-06,
+      "loss": 0.494,
+      "step": 1297
+    },
+    {
+      "epoch": 0.878213802435724,
+      "grad_norm": 1.1554655320262635,
+      "learning_rate": 8.911610775517383e-06,
+      "loss": 0.5158,
+      "step": 1298
+    },
+    {
+      "epoch": 0.8788903924221921,
+      "grad_norm": 1.1058022445695312,
+      "learning_rate": 8.90915741234015e-06,
+      "loss": 0.5058,
+      "step": 1299
+    },
+    {
+      "epoch": 0.8795669824086604,
+      "grad_norm": 1.093566761068326,
+      "learning_rate": 8.906701625696028e-06,
+      "loss": 0.4625,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8802435723951285,
+      "grad_norm": 1.1322104456009245,
+      "learning_rate": 8.904243417107473e-06,
+      "loss": 0.5016,
+      "step": 1301
+    },
+    {
+      "epoch": 0.8809201623815968,
+      "grad_norm": 1.1148533155687137,
+      "learning_rate": 8.901782788098442e-06,
+      "loss": 0.4651,
+      "step": 1302
+    },
+    {
+      "epoch": 0.881596752368065,
+      "grad_norm": 1.1803033305243866,
+      "learning_rate": 8.899319740194391e-06,
+      "loss": 0.4861,
+      "step": 1303
+    },
+    {
+      "epoch": 0.8822733423545331,
+      "grad_norm": 1.1627810496841564,
+      "learning_rate": 8.89685427492228e-06,
+      "loss": 0.517,
+      "step": 1304
+    },
+    {
+      "epoch": 0.8829499323410014,
+      "grad_norm": 1.1271570410543166,
+      "learning_rate": 8.894386393810563e-06,
+      "loss": 0.4859,
+      "step": 1305
+    },
+    {
+      "epoch": 0.8836265223274695,
+      "grad_norm": 1.111793161435199,
+      "learning_rate": 8.891916098389193e-06,
+      "loss": 0.4763,
+      "step": 1306
+    },
+    {
+      "epoch": 0.8843031123139378,
+      "grad_norm": 1.1034277151511187,
+      "learning_rate": 8.889443390189618e-06,
+      "loss": 0.49,
+      "step": 1307
+    },
+    {
+      "epoch": 0.8849797023004059,
+      "grad_norm": 1.1140873596427165,
+      "learning_rate": 8.886968270744789e-06,
+      "loss": 0.4851,
+      "step": 1308
+    },
+    {
+      "epoch": 0.8856562922868741,
+      "grad_norm": 1.0885624245963936,
+      "learning_rate": 8.88449074158914e-06,
+      "loss": 0.4827,
+      "step": 1309
+    },
+    {
+      "epoch": 0.8863328822733424,
+      "grad_norm": 1.143391920316101,
+      "learning_rate": 8.882010804258612e-06,
+      "loss": 0.5025,
+      "step": 1310
+    },
+    {
+      "epoch": 0.8870094722598105,
+      "grad_norm": 1.171688364607078,
+      "learning_rate": 8.879528460290628e-06,
+      "loss": 0.5019,
+      "step": 1311
+    },
+    {
+      "epoch": 0.8876860622462788,
+      "grad_norm": 1.0896444499215772,
+      "learning_rate": 8.877043711224109e-06,
+      "loss": 0.4932,
+      "step": 1312
+    },
+    {
+      "epoch": 0.8883626522327469,
+      "grad_norm": 1.1438246799736955,
+      "learning_rate": 8.874556558599465e-06,
+      "loss": 0.4767,
+      "step": 1313
+    },
+    {
+      "epoch": 0.8890392422192152,
+      "grad_norm": 1.173811621971971,
+      "learning_rate": 8.872067003958597e-06,
+      "loss": 0.4927,
+      "step": 1314
+    },
+    {
+      "epoch": 0.8897158322056834,
+      "grad_norm": 1.118986786328927,
+      "learning_rate": 8.869575048844896e-06,
+      "loss": 0.4889,
+      "step": 1315
+    },
+    {
+      "epoch": 0.8903924221921515,
+      "grad_norm": 1.145105965592563,
+      "learning_rate": 8.867080694803238e-06,
+      "loss": 0.5006,
+      "step": 1316
+    },
+    {
+      "epoch": 0.8910690121786198,
+      "grad_norm": 1.1690976704954363,
+      "learning_rate": 8.864583943379987e-06,
+      "loss": 0.48,
+      "step": 1317
+    },
+    {
+      "epoch": 0.8917456021650879,
+      "grad_norm": 1.1877369269813445,
+      "learning_rate": 8.862084796122998e-06,
+      "loss": 0.4963,
+      "step": 1318
+    },
+    {
+      "epoch": 0.8924221921515562,
+      "grad_norm": 1.1109714662658936,
+      "learning_rate": 8.859583254581604e-06,
+      "loss": 0.4803,
+      "step": 1319
+    },
+    {
+      "epoch": 0.8930987821380244,
+      "grad_norm": 1.0543606010594926,
+      "learning_rate": 8.85707932030663e-06,
+      "loss": 0.4698,
+      "step": 1320
+    },
+    {
+      "epoch": 0.8937753721244925,
+      "grad_norm": 1.0918568552170842,
+      "learning_rate": 8.854572994850376e-06,
+      "loss": 0.4821,
+      "step": 1321
+    },
+    {
+      "epoch": 0.8944519621109608,
+      "grad_norm": 1.1171309020142948,
+      "learning_rate": 8.85206427976663e-06,
+      "loss": 0.487,
+      "step": 1322
+    },
+    {
+      "epoch": 0.8951285520974289,
+      "grad_norm": 1.1055727328989207,
+      "learning_rate": 8.849553176610661e-06,
+      "loss": 0.4795,
+      "step": 1323
+    },
+    {
+      "epoch": 0.8958051420838972,
+      "grad_norm": 1.124376201088204,
+      "learning_rate": 8.847039686939218e-06,
+      "loss": 0.5016,
+      "step": 1324
+    },
+    {
+      "epoch": 0.8964817320703654,
+      "grad_norm": 1.164546623553131,
+      "learning_rate": 8.844523812310527e-06,
+      "loss": 0.5052,
+      "step": 1325
+    },
+    {
+      "epoch": 0.8971583220568335,
+      "grad_norm": 1.1225569509504587,
+      "learning_rate": 8.842005554284296e-06,
+      "loss": 0.5065,
+      "step": 1326
+    },
+    {
+      "epoch": 0.8978349120433018,
+      "grad_norm": 1.0816518230791254,
+      "learning_rate": 8.83948491442171e-06,
+      "loss": 0.4812,
+      "step": 1327
+    },
+    {
+      "epoch": 0.8985115020297699,
+      "grad_norm": 1.062718268046039,
+      "learning_rate": 8.836961894285428e-06,
+      "loss": 0.4758,
+      "step": 1328
+    },
+    {
+      "epoch": 0.8991880920162382,
+      "grad_norm": 1.062149555585684,
+      "learning_rate": 8.834436495439588e-06,
+      "loss": 0.4848,
+      "step": 1329
+    },
+    {
+      "epoch": 0.8998646820027063,
+      "grad_norm": 1.133815006969586,
+      "learning_rate": 8.8319087194498e-06,
+      "loss": 0.4837,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9005412719891746,
+      "grad_norm": 1.120722107309994,
+      "learning_rate": 8.829378567883152e-06,
+      "loss": 0.5073,
+      "step": 1331
+    },
+    {
+      "epoch": 0.9012178619756428,
+      "grad_norm": 1.115072165839227,
+      "learning_rate": 8.826846042308195e-06,
+      "loss": 0.5087,
+      "step": 1332
+    },
+    {
+      "epoch": 0.9018944519621109,
+      "grad_norm": 1.0648196052598222,
+      "learning_rate": 8.824311144294966e-06,
+      "loss": 0.4637,
+      "step": 1333
+    },
+    {
+      "epoch": 0.9025710419485792,
+      "grad_norm": 1.1892533528735048,
+      "learning_rate": 8.82177387541496e-06,
+      "loss": 0.5189,
+      "step": 1334
+    },
+    {
+      "epoch": 0.9032476319350473,
+      "grad_norm": 1.109942217898596,
+      "learning_rate": 8.819234237241148e-06,
+      "loss": 0.4783,
+      "step": 1335
+    },
+    {
+      "epoch": 0.9039242219215156,
+      "grad_norm": 1.0853856487309683,
+      "learning_rate": 8.816692231347972e-06,
+      "loss": 0.4759,
+      "step": 1336
+    },
+    {
+      "epoch": 0.9046008119079838,
+      "grad_norm": 1.0652639470040577,
+      "learning_rate": 8.814147859311333e-06,
+      "loss": 0.4791,
+      "step": 1337
+    },
+    {
+      "epoch": 0.9052774018944519,
+      "grad_norm": 1.0929673161367177,
+      "learning_rate": 8.81160112270861e-06,
+      "loss": 0.4887,
+      "step": 1338
+    },
+    {
+      "epoch": 0.9059539918809202,
+      "grad_norm": 1.1495286340763267,
+      "learning_rate": 8.809052023118638e-06,
+      "loss": 0.5051,
+      "step": 1339
+    },
+    {
+      "epoch": 0.9066305818673883,
+      "grad_norm": 1.078036842137762,
+      "learning_rate": 8.806500562121724e-06,
+      "loss": 0.4695,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9073071718538566,
+      "grad_norm": 1.0903934108755975,
+      "learning_rate": 8.803946741299635e-06,
+      "loss": 0.4497,
+      "step": 1341
+    },
+    {
+      "epoch": 0.9079837618403248,
+      "grad_norm": 1.0983398134962221,
+      "learning_rate": 8.801390562235603e-06,
+      "loss": 0.4805,
+      "step": 1342
+    },
+    {
+      "epoch": 0.908660351826793,
+      "grad_norm": 1.0688244745964672,
+      "learning_rate": 8.79883202651432e-06,
+      "loss": 0.4575,
+      "step": 1343
+    },
+    {
+      "epoch": 0.9093369418132612,
+      "grad_norm": 1.1078644727186073,
+      "learning_rate": 8.796271135721944e-06,
+      "loss": 0.4946,
+      "step": 1344
+    },
+    {
+      "epoch": 0.9100135317997293,
+      "grad_norm": 1.1421186491807882,
+      "learning_rate": 8.793707891446086e-06,
+      "loss": 0.4847,
+      "step": 1345
+    },
+    {
+      "epoch": 0.9106901217861976,
+      "grad_norm": 1.093260920106085,
+      "learning_rate": 8.791142295275819e-06,
+      "loss": 0.4751,
+      "step": 1346
+    },
+    {
+      "epoch": 0.9113667117726658,
+      "grad_norm": 1.1136598978673493,
+      "learning_rate": 8.788574348801676e-06,
+      "loss": 0.4806,
+      "step": 1347
+    },
+    {
+      "epoch": 0.912043301759134,
+      "grad_norm": 1.102167214310862,
+      "learning_rate": 8.786004053615642e-06,
+      "loss": 0.4764,
+      "step": 1348
+    },
+    {
+      "epoch": 0.9127198917456022,
+      "grad_norm": 1.1059664458026084,
+      "learning_rate": 8.783431411311165e-06,
+      "loss": 0.4894,
+      "step": 1349
+    },
+    {
+      "epoch": 0.9133964817320703,
+      "grad_norm": 1.0578620170298016,
+      "learning_rate": 8.780856423483145e-06,
+      "loss": 0.4827,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9140730717185386,
+      "grad_norm": 1.0957760371681264,
+      "learning_rate": 8.778279091727933e-06,
+      "loss": 0.503,
+      "step": 1351
+    },
+    {
+      "epoch": 0.9147496617050067,
+      "grad_norm": 1.1096754377279134,
+      "learning_rate": 8.775699417643337e-06,
+      "loss": 0.4719,
+      "step": 1352
+    },
+    {
+      "epoch": 0.915426251691475,
+      "grad_norm": 1.0989769966462268,
+      "learning_rate": 8.773117402828618e-06,
+      "loss": 0.4766,
+      "step": 1353
+    },
+    {
+      "epoch": 0.9161028416779432,
+      "grad_norm": 1.1020082145652916,
+      "learning_rate": 8.770533048884483e-06,
+      "loss": 0.4726,
+      "step": 1354
+    },
+    {
+      "epoch": 0.9167794316644113,
+      "grad_norm": 1.1072270802152941,
+      "learning_rate": 8.767946357413091e-06,
+      "loss": 0.4843,
+      "step": 1355
+    },
+    {
+      "epoch": 0.9174560216508796,
+      "grad_norm": 1.156295673676666,
+      "learning_rate": 8.765357330018056e-06,
+      "loss": 0.4952,
+      "step": 1356
+    },
+    {
+      "epoch": 0.9181326116373477,
+      "grad_norm": 1.1270161829253635,
+      "learning_rate": 8.76276596830443e-06,
+      "loss": 0.5159,
+      "step": 1357
+    },
+    {
+      "epoch": 0.918809201623816,
+      "grad_norm": 1.0891541127007403,
+      "learning_rate": 8.760172273878723e-06,
+      "loss": 0.4905,
+      "step": 1358
+    },
+    {
+      "epoch": 0.9194857916102842,
+      "grad_norm": 1.1225614804261512,
+      "learning_rate": 8.757576248348883e-06,
+      "loss": 0.4899,
+      "step": 1359
+    },
+    {
+      "epoch": 0.9201623815967523,
+      "grad_norm": 1.1093518880030158,
+      "learning_rate": 8.754977893324305e-06,
+      "loss": 0.4836,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9208389715832206,
+      "grad_norm": 1.122660197398548,
+      "learning_rate": 8.75237721041583e-06,
+      "loss": 0.4885,
+      "step": 1361
+    },
+    {
+      "epoch": 0.9215155615696887,
+      "grad_norm": 1.0844909528987519,
+      "learning_rate": 8.74977420123574e-06,
+      "loss": 0.4754,
+      "step": 1362
+    },
+    {
+      "epoch": 0.922192151556157,
+      "grad_norm": 1.070954967599324,
+      "learning_rate": 8.747168867397765e-06,
+      "loss": 0.4771,
+      "step": 1363
+    },
+    {
+      "epoch": 0.9228687415426252,
+      "grad_norm": 1.1557809804366883,
+      "learning_rate": 8.744561210517067e-06,
+      "loss": 0.4649,
+      "step": 1364
+    },
+    {
+      "epoch": 0.9235453315290933,
+      "grad_norm": 1.1020724249035476,
+      "learning_rate": 8.741951232210254e-06,
+      "loss": 0.4759,
+      "step": 1365
+    },
+    {
+      "epoch": 0.9242219215155616,
+      "grad_norm": 1.0679150863324274,
+      "learning_rate": 8.73933893409537e-06,
+      "loss": 0.4782,
+      "step": 1366
+    },
+    {
+      "epoch": 0.9248985115020297,
+      "grad_norm": 1.0974332171975851,
+      "learning_rate": 8.736724317791903e-06,
+      "loss": 0.4697,
+      "step": 1367
+    },
+    {
+      "epoch": 0.925575101488498,
+      "grad_norm": 1.1369134886294892,
+      "learning_rate": 8.734107384920771e-06,
+      "loss": 0.4654,
+      "step": 1368
+    },
+    {
+      "epoch": 0.9262516914749662,
+      "grad_norm": 1.1723999917840484,
+      "learning_rate": 8.731488137104332e-06,
+      "loss": 0.5069,
+      "step": 1369
+    },
+    {
+      "epoch": 0.9269282814614344,
+      "grad_norm": 1.1442159695171845,
+      "learning_rate": 8.728866575966379e-06,
+      "loss": 0.4697,
+      "step": 1370
+    },
+    {
+      "epoch": 0.9276048714479026,
+      "grad_norm": 1.1039418864188117,
+      "learning_rate": 8.726242703132139e-06,
+      "loss": 0.4749,
+      "step": 1371
+    },
+    {
+      "epoch": 0.9282814614343707,
+      "grad_norm": 1.158271427116123,
+      "learning_rate": 8.72361652022827e-06,
+      "loss": 0.4635,
+      "step": 1372
+    },
+    {
+      "epoch": 0.928958051420839,
+      "grad_norm": 1.1324175018730116,
+      "learning_rate": 8.720988028882867e-06,
+      "loss": 0.4826,
+      "step": 1373
+    },
+    {
+      "epoch": 0.9296346414073072,
+      "grad_norm": 1.0592613871370138,
+      "learning_rate": 8.71835723072545e-06,
+      "loss": 0.4733,
+      "step": 1374
+    },
+    {
+      "epoch": 0.9303112313937754,
+      "grad_norm": 1.108042678700999,
+      "learning_rate": 8.715724127386971e-06,
+      "loss": 0.4845,
+      "step": 1375
+    },
+    {
+      "epoch": 0.9309878213802436,
+      "grad_norm": 1.0490795210120185,
+      "learning_rate": 8.713088720499817e-06,
+      "loss": 0.4674,
+      "step": 1376
+    },
+    {
+      "epoch": 0.9316644113667117,
+      "grad_norm": 1.079599177008303,
+      "learning_rate": 8.710451011697794e-06,
+      "loss": 0.4581,
+      "step": 1377
+    },
+    {
+      "epoch": 0.93234100135318,
+      "grad_norm": 1.101341041886667,
+      "learning_rate": 8.70781100261614e-06,
+      "loss": 0.4712,
+      "step": 1378
+    },
+    {
+      "epoch": 0.9330175913396481,
+      "grad_norm": 1.128688170954129,
+      "learning_rate": 8.705168694891522e-06,
+      "loss": 0.503,
+      "step": 1379
+    },
+    {
+      "epoch": 0.9336941813261164,
+      "grad_norm": 1.1146512655692067,
+      "learning_rate": 8.702524090162023e-06,
+      "loss": 0.458,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9343707713125846,
+      "grad_norm": 1.0875249983285529,
+      "learning_rate": 8.699877190067158e-06,
+      "loss": 0.4644,
+      "step": 1381
+    },
+    {
+      "epoch": 0.9350473612990527,
+      "grad_norm": 1.0827841064440842,
+      "learning_rate": 8.697227996247861e-06,
+      "loss": 0.4785,
+      "step": 1382
+    },
+    {
+      "epoch": 0.935723951285521,
+      "grad_norm": 1.1930711262650655,
+      "learning_rate": 8.694576510346493e-06,
+      "loss": 0.5006,
+      "step": 1383
+    },
+    {
+      "epoch": 0.9364005412719891,
+      "grad_norm": 1.1121646400437968,
+      "learning_rate": 8.691922734006828e-06,
+      "loss": 0.4747,
+      "step": 1384
+    },
+    {
+      "epoch": 0.9370771312584574,
+      "grad_norm": 1.1020820481150944,
+      "learning_rate": 8.689266668874067e-06,
+      "loss": 0.4793,
+      "step": 1385
+    },
+    {
+      "epoch": 0.9377537212449256,
+      "grad_norm": 1.092052589638481,
+      "learning_rate": 8.686608316594826e-06,
+      "loss": 0.4748,
+      "step": 1386
+    },
+    {
+      "epoch": 0.9384303112313938,
+      "grad_norm": 1.1178469084348743,
+      "learning_rate": 8.683947678817139e-06,
+      "loss": 0.4734,
+      "step": 1387
+    },
+    {
+      "epoch": 0.939106901217862,
+      "grad_norm": 1.0851127636579434,
+      "learning_rate": 8.681284757190462e-06,
+      "loss": 0.4793,
+      "step": 1388
+    },
+    {
+      "epoch": 0.9397834912043301,
+      "grad_norm": 1.1681335670461341,
+      "learning_rate": 8.67861955336566e-06,
+      "loss": 0.5047,
+      "step": 1389
+    },
+    {
+      "epoch": 0.9404600811907984,
+      "grad_norm": 1.1228454390584235,
+      "learning_rate": 8.675952068995014e-06,
+      "loss": 0.4761,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9411366711772666,
+      "grad_norm": 1.155355536164632,
+      "learning_rate": 8.673282305732225e-06,
+      "loss": 0.4931,
+      "step": 1391
+    },
+    {
+      "epoch": 0.9418132611637348,
+      "grad_norm": 1.0975254096590343,
+      "learning_rate": 8.670610265232398e-06,
+      "loss": 0.4952,
+      "step": 1392
+    },
+    {
+      "epoch": 0.942489851150203,
+      "grad_norm": 1.1263807002918433,
+      "learning_rate": 8.667935949152057e-06,
+      "loss": 0.4973,
+      "step": 1393
+    },
+    {
+      "epoch": 0.9431664411366711,
+      "grad_norm": 1.0717855553357298,
+      "learning_rate": 8.665259359149132e-06,
+      "loss": 0.4604,
+      "step": 1394
+    },
+    {
+      "epoch": 0.9438430311231394,
+      "grad_norm": 1.115443505803302,
+      "learning_rate": 8.662580496882967e-06,
+      "loss": 0.4638,
+      "step": 1395
+    },
+    {
+      "epoch": 0.9445196211096076,
+      "grad_norm": 1.0613925003688098,
+      "learning_rate": 8.659899364014309e-06,
+      "loss": 0.4661,
+      "step": 1396
+    },
+    {
+      "epoch": 0.9451962110960758,
+      "grad_norm": 1.1160444526814355,
+      "learning_rate": 8.657215962205318e-06,
+      "loss": 0.4826,
+      "step": 1397
+    },
+    {
+      "epoch": 0.945872801082544,
+      "grad_norm": 1.0982289880424816,
+      "learning_rate": 8.654530293119558e-06,
+      "loss": 0.4869,
+      "step": 1398
+    },
+    {
+      "epoch": 0.9465493910690121,
+      "grad_norm": 1.042399225804521,
+      "learning_rate": 8.651842358421999e-06,
+      "loss": 0.4491,
+      "step": 1399
+    },
+    {
+      "epoch": 0.9472259810554804,
+      "grad_norm": 1.025045296939848,
+      "learning_rate": 8.649152159779015e-06,
+      "loss": 0.4716,
+      "step": 1400
+    },
+    {
+      "epoch": 0.9479025710419485,
+      "grad_norm": 1.0188020827441104,
+      "learning_rate": 8.646459698858386e-06,
+      "loss": 0.4498,
+      "step": 1401
+    },
+    {
+      "epoch": 0.9485791610284168,
+      "grad_norm": 1.131377324119661,
+      "learning_rate": 8.64376497732929e-06,
+      "loss": 0.4726,
+      "step": 1402
+    },
+    {
+      "epoch": 0.949255751014885,
+      "grad_norm": 1.1033023675409355,
+      "learning_rate": 8.64106799686231e-06,
+      "loss": 0.4882,
+      "step": 1403
+    },
+    {
+      "epoch": 0.9499323410013532,
+      "grad_norm": 1.1570690756002653,
+      "learning_rate": 8.638368759129433e-06,
+      "loss": 0.4711,
+      "step": 1404
+    },
+    {
+      "epoch": 0.9506089309878214,
+      "grad_norm": 1.0936792629313752,
+      "learning_rate": 8.635667265804034e-06,
+      "loss": 0.4602,
+      "step": 1405
+    },
+    {
+      "epoch": 0.9512855209742895,
+      "grad_norm": 1.1097972552291377,
+      "learning_rate": 8.632963518560894e-06,
+      "loss": 0.4769,
+      "step": 1406
+    },
+    {
+      "epoch": 0.9519621109607578,
+      "grad_norm": 1.0600437578483592,
+      "learning_rate": 8.630257519076196e-06,
+      "loss": 0.4764,
+      "step": 1407
+    },
+    {
+      "epoch": 0.952638700947226,
+      "grad_norm": 1.0810009270230665,
+      "learning_rate": 8.627549269027509e-06,
+      "loss": 0.4563,
+      "step": 1408
+    },
+    {
+      "epoch": 0.9533152909336942,
+      "grad_norm": 1.1378313133212332,
+      "learning_rate": 8.624838770093805e-06,
+      "loss": 0.5015,
+      "step": 1409
+    },
+    {
+      "epoch": 0.9539918809201624,
+      "grad_norm": 1.0730409166924595,
+      "learning_rate": 8.622126023955446e-06,
+      "loss": 0.486,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9546684709066305,
+      "grad_norm": 1.0924453900556088,
+      "learning_rate": 8.619411032294187e-06,
+      "loss": 0.4815,
+      "step": 1411
+    },
+    {
+      "epoch": 0.9553450608930988,
+      "grad_norm": 1.1196772622569686,
+      "learning_rate": 8.616693796793178e-06,
+      "loss": 0.4908,
+      "step": 1412
+    },
+    {
+      "epoch": 0.956021650879567,
+      "grad_norm": 1.1279912000536065,
+      "learning_rate": 8.613974319136959e-06,
+      "loss": 0.4697,
+      "step": 1413
+    },
+    {
+      "epoch": 0.9566982408660352,
+      "grad_norm": 1.0686371294128816,
+      "learning_rate": 8.611252601011457e-06,
+      "loss": 0.4557,
+      "step": 1414
+    },
+    {
+      "epoch": 0.9573748308525034,
+      "grad_norm": 1.060571430752584,
+      "learning_rate": 8.608528644103994e-06,
+      "loss": 0.4773,
+      "step": 1415
+    },
+    {
+      "epoch": 0.9580514208389715,
+      "grad_norm": 1.0943973067230608,
+      "learning_rate": 8.605802450103276e-06,
+      "loss": 0.4692,
+      "step": 1416
+    },
+    {
+      "epoch": 0.9587280108254398,
+      "grad_norm": 1.1135642740683926,
+      "learning_rate": 8.603074020699393e-06,
+      "loss": 0.4632,
+      "step": 1417
+    },
+    {
+      "epoch": 0.959404600811908,
+      "grad_norm": 1.1564508268609643,
+      "learning_rate": 8.600343357583826e-06,
+      "loss": 0.4832,
+      "step": 1418
+    },
+    {
+      "epoch": 0.9600811907983762,
+      "grad_norm": 1.188240940834957,
+      "learning_rate": 8.597610462449441e-06,
+      "loss": 0.5119,
+      "step": 1419
+    },
+    {
+      "epoch": 0.9607577807848444,
+      "grad_norm": 1.0838446802817607,
+      "learning_rate": 8.594875336990482e-06,
+      "loss": 0.4628,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9614343707713126,
+      "grad_norm": 1.1083838323042579,
+      "learning_rate": 8.592137982902585e-06,
+      "loss": 0.4823,
+      "step": 1421
+    },
+    {
+      "epoch": 0.9621109607577808,
+      "grad_norm": 1.0957894962472958,
+      "learning_rate": 8.589398401882755e-06,
+      "loss": 0.4904,
+      "step": 1422
+    },
+    {
+      "epoch": 0.9627875507442489,
+      "grad_norm": 1.0759682815221387,
+      "learning_rate": 8.586656595629387e-06,
+      "loss": 0.4579,
+      "step": 1423
+    },
+    {
+      "epoch": 0.9634641407307172,
+      "grad_norm": 1.101268240904258,
+      "learning_rate": 8.583912565842258e-06,
+      "loss": 0.4619,
+      "step": 1424
+    },
+    {
+      "epoch": 0.9641407307171854,
+      "grad_norm": 1.056032844943916,
+      "learning_rate": 8.581166314222512e-06,
+      "loss": 0.4782,
+      "step": 1425
+    },
+    {
+      "epoch": 0.9648173207036536,
+      "grad_norm": 1.0247273710377593,
+      "learning_rate": 8.57841784247268e-06,
+      "loss": 0.4733,
+      "step": 1426
+    },
+    {
+      "epoch": 0.9654939106901218,
+      "grad_norm": 1.0633616342864958,
+      "learning_rate": 8.575667152296666e-06,
+      "loss": 0.4964,
+      "step": 1427
+    },
+    {
+      "epoch": 0.9661705006765899,
+      "grad_norm": 1.09250706289183,
+      "learning_rate": 8.572914245399748e-06,
+      "loss": 0.4868,
+      "step": 1428
+    },
+    {
+      "epoch": 0.9668470906630582,
+      "grad_norm": 1.0532665047266043,
+      "learning_rate": 8.570159123488584e-06,
+      "loss": 0.4591,
+      "step": 1429
+    },
+    {
+      "epoch": 0.9675236806495264,
+      "grad_norm": 1.1648035844376532,
+      "learning_rate": 8.567401788271195e-06,
+      "loss": 0.4792,
+      "step": 1430
+    },
+    {
+      "epoch": 0.9682002706359946,
+      "grad_norm": 1.0526720189732204,
+      "learning_rate": 8.564642241456986e-06,
+      "loss": 0.4745,
+      "step": 1431
+    },
+    {
+      "epoch": 0.9688768606224628,
+      "grad_norm": 1.1939719091186145,
+      "learning_rate": 8.561880484756726e-06,
+      "loss": 0.4979,
+      "step": 1432
+    },
+    {
+      "epoch": 0.969553450608931,
+      "grad_norm": 1.0643289776647997,
+      "learning_rate": 8.559116519882551e-06,
+      "loss": 0.4606,
+      "step": 1433
+    },
+    {
+      "epoch": 0.9702300405953992,
+      "grad_norm": 1.0954360300513486,
+      "learning_rate": 8.556350348547978e-06,
+      "loss": 0.4672,
+      "step": 1434
+    },
+    {
+      "epoch": 0.9709066305818674,
+      "grad_norm": 1.1244464441728843,
+      "learning_rate": 8.553581972467875e-06,
+      "loss": 0.481,
+      "step": 1435
+    },
+    {
+      "epoch": 0.9715832205683356,
+      "grad_norm": 1.131798533312523,
+      "learning_rate": 8.550811393358494e-06,
+      "loss": 0.4831,
+      "step": 1436
+    },
+    {
+      "epoch": 0.9722598105548038,
+      "grad_norm": 1.1107322276914728,
+      "learning_rate": 8.54803861293744e-06,
+      "loss": 0.4627,
+      "step": 1437
+    },
+    {
+      "epoch": 0.972936400541272,
+      "grad_norm": 1.0968367892460305,
+      "learning_rate": 8.545263632923687e-06,
+      "loss": 0.4716,
+      "step": 1438
+    },
+    {
+      "epoch": 0.9736129905277402,
+      "grad_norm": 1.0702423575289433,
+      "learning_rate": 8.542486455037578e-06,
+      "loss": 0.4892,
+      "step": 1439
+    },
+    {
+      "epoch": 0.9742895805142084,
+      "grad_norm": 1.0901673460271821,
+      "learning_rate": 8.539707081000808e-06,
+      "loss": 0.4802,
+      "step": 1440
+    },
+    {
+      "epoch": 0.9749661705006766,
+      "grad_norm": 1.0775967217988527,
+      "learning_rate": 8.536925512536441e-06,
+      "loss": 0.4852,
+      "step": 1441
+    },
+    {
+      "epoch": 0.9756427604871448,
+      "grad_norm": 1.0804318364012613,
+      "learning_rate": 8.534141751368901e-06,
+      "loss": 0.4754,
+      "step": 1442
+    },
+    {
+      "epoch": 0.976319350473613,
+      "grad_norm": 1.0948792749230274,
+      "learning_rate": 8.531355799223968e-06,
+      "loss": 0.475,
+      "step": 1443
+    },
+    {
+      "epoch": 0.9769959404600812,
+      "grad_norm": 1.040576660022014,
+      "learning_rate": 8.528567657828785e-06,
+      "loss": 0.4667,
+      "step": 1444
+    },
+    {
+      "epoch": 0.9776725304465493,
+      "grad_norm": 1.168361912067637,
+      "learning_rate": 8.525777328911846e-06,
+      "loss": 0.4813,
+      "step": 1445
+    },
+    {
+      "epoch": 0.9783491204330176,
+      "grad_norm": 1.0415737120357251,
+      "learning_rate": 8.522984814203006e-06,
+      "loss": 0.4697,
+      "step": 1446
+    },
+    {
+      "epoch": 0.9790257104194858,
+      "grad_norm": 1.135579034154808,
+      "learning_rate": 8.520190115433473e-06,
+      "loss": 0.4964,
+      "step": 1447
+    },
+    {
+      "epoch": 0.979702300405954,
+      "grad_norm": 1.0571980952930058,
+      "learning_rate": 8.517393234335812e-06,
+      "loss": 0.4839,
+      "step": 1448
+    },
+    {
+      "epoch": 0.9803788903924222,
+      "grad_norm": 1.0516782289808286,
+      "learning_rate": 8.514594172643934e-06,
+      "loss": 0.4715,
+      "step": 1449
+    },
+    {
+      "epoch": 0.9810554803788903,
+      "grad_norm": 1.0840929954449103,
+      "learning_rate": 8.51179293209311e-06,
+      "loss": 0.4601,
+      "step": 1450
+    },
+    {
+      "epoch": 0.9817320703653586,
+      "grad_norm": 1.0817488259851176,
+      "learning_rate": 8.508989514419959e-06,
+      "loss": 0.4858,
+      "step": 1451
+    },
+    {
+      "epoch": 0.9824086603518268,
+      "grad_norm": 1.04091882722342,
+      "learning_rate": 8.506183921362443e-06,
+      "loss": 0.4398,
+      "step": 1452
+    },
+    {
+      "epoch": 0.983085250338295,
+      "grad_norm": 1.0565015043834196,
+      "learning_rate": 8.503376154659886e-06,
+      "loss": 0.4723,
+      "step": 1453
+    },
+    {
+      "epoch": 0.9837618403247632,
+      "grad_norm": 1.0955380238432584,
+      "learning_rate": 8.500566216052948e-06,
+      "loss": 0.4826,
+      "step": 1454
+    },
+    {
+      "epoch": 0.9844384303112313,
+      "grad_norm": 1.0444130204805198,
+      "learning_rate": 8.497754107283637e-06,
+      "loss": 0.4566,
+      "step": 1455
+    },
+    {
+      "epoch": 0.9851150202976996,
+      "grad_norm": 0.9885016010515429,
+      "learning_rate": 8.494939830095315e-06,
+      "loss": 0.4475,
+      "step": 1456
+    },
+    {
+      "epoch": 0.9857916102841678,
+      "grad_norm": 1.0177590872550946,
+      "learning_rate": 8.492123386232678e-06,
+      "loss": 0.4618,
+      "step": 1457
+    },
+    {
+      "epoch": 0.986468200270636,
+      "grad_norm": 1.060331320050249,
+      "learning_rate": 8.489304777441772e-06,
+      "loss": 0.4804,
+      "step": 1458
+    },
+    {
+      "epoch": 0.9871447902571042,
+      "grad_norm": 1.0130478712982525,
+      "learning_rate": 8.486484005469977e-06,
+      "loss": 0.4645,
+      "step": 1459
+    },
+    {
+      "epoch": 0.9878213802435724,
+      "grad_norm": 1.090740661889863,
+      "learning_rate": 8.483661072066027e-06,
+      "loss": 0.483,
+      "step": 1460
+    },
+    {
+      "epoch": 0.9884979702300406,
+      "grad_norm": 1.084950732414929,
+      "learning_rate": 8.480835978979983e-06,
+      "loss": 0.4557,
+      "step": 1461
+    },
+    {
+      "epoch": 0.9891745602165088,
+      "grad_norm": 1.0397139752573001,
+      "learning_rate": 8.478008727963253e-06,
+      "loss": 0.4669,
+      "step": 1462
+    },
+    {
+      "epoch": 0.989851150202977,
+      "grad_norm": 1.027183170924242,
+      "learning_rate": 8.475179320768581e-06,
+      "loss": 0.4683,
+      "step": 1463
+    },
+    {
+      "epoch": 0.9905277401894452,
+      "grad_norm": 1.065718624494268,
+      "learning_rate": 8.472347759150044e-06,
+      "loss": 0.4697,
+      "step": 1464
+    },
+    {
+      "epoch": 0.9912043301759134,
+      "grad_norm": 1.048211630683376,
+      "learning_rate": 8.46951404486306e-06,
+      "loss": 0.4424,
+      "step": 1465
+    },
+    {
+      "epoch": 0.9918809201623816,
+      "grad_norm": 1.0681969687831656,
+      "learning_rate": 8.466678179664378e-06,
+      "loss": 0.4583,
+      "step": 1466
+    },
+    {
+      "epoch": 0.9925575101488497,
+      "grad_norm": 1.0772139404680132,
+      "learning_rate": 8.463840165312083e-06,
+      "loss": 0.4539,
+      "step": 1467
+    },
+    {
+      "epoch": 0.993234100135318,
+      "grad_norm": 1.1274787922980964,
+      "learning_rate": 8.461000003565588e-06,
+      "loss": 0.478,
+      "step": 1468
+    },
+    {
+      "epoch": 0.9939106901217862,
+      "grad_norm": 1.1519269128455025,
+      "learning_rate": 8.458157696185643e-06,
+      "loss": 0.4796,
+      "step": 1469
+    },
+    {
+      "epoch": 0.9945872801082544,
+      "grad_norm": 1.0985704356091963,
+      "learning_rate": 8.455313244934324e-06,
+      "loss": 0.4945,
+      "step": 1470
+    },
+    {
+      "epoch": 0.9952638700947226,
+      "grad_norm": 1.130466681997261,
+      "learning_rate": 8.452466651575039e-06,
+      "loss": 0.4701,
+      "step": 1471
+    },
+    {
+      "epoch": 0.9959404600811907,
+      "grad_norm": 1.0723777920342406,
+      "learning_rate": 8.44961791787252e-06,
+      "loss": 0.4781,
+      "step": 1472
+    },
+    {
+      "epoch": 0.996617050067659,
+      "grad_norm": 1.0129052107576901,
+      "learning_rate": 8.446767045592829e-06,
+      "loss": 0.4472,
+      "step": 1473
+    },
+    {
+      "epoch": 0.9972936400541272,
+      "grad_norm": 1.0822449262813412,
+      "learning_rate": 8.443914036503356e-06,
+      "loss": 0.5021,
+      "step": 1474
+    },
+    {
+      "epoch": 0.9979702300405954,
+      "grad_norm": 1.058103427210262,
+      "learning_rate": 8.44105889237281e-06,
+      "loss": 0.493,
+      "step": 1475
+    },
+    {
+      "epoch": 0.9986468200270636,
+      "grad_norm": 1.021156031374969,
+      "learning_rate": 8.438201614971227e-06,
+      "loss": 0.4592,
+      "step": 1476
+    },
+    {
+      "epoch": 0.9993234100135318,
+      "grad_norm": 1.0687638217219952,
+      "learning_rate": 8.435342206069965e-06,
+      "loss": 0.4808,
+      "step": 1477
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0693278482698754,
+      "learning_rate": 8.432480667441703e-06,
+      "loss": 0.4824,
+      "step": 1478
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.47269096970558167,
+      "eval_runtime": 442.6041,
+      "eval_samples_per_second": 22.492,
+      "eval_steps_per_second": 0.705,
+      "step": 1478
+    },
+    {
+      "epoch": 1.0006765899864682,
+      "grad_norm": 1.0631144044441243,
+      "learning_rate": 8.429617000860441e-06,
+      "loss": 0.447,
+      "step": 1479
+    },
+    {
+      "epoch": 1.0013531799729365,
+      "grad_norm": 1.0134654167802628,
+      "learning_rate": 8.4267512081015e-06,
+      "loss": 0.4393,
+      "step": 1480
+    },
+    {
+      "epoch": 1.0020297699594045,
+      "grad_norm": 1.025784764238537,
+      "learning_rate": 8.423883290941514e-06,
+      "loss": 0.4477,
+      "step": 1481
+    },
+    {
+      "epoch": 1.0027063599458728,
+      "grad_norm": 1.015539247825642,
+      "learning_rate": 8.421013251158437e-06,
+      "loss": 0.4295,
+      "step": 1482
+    },
+    {
+      "epoch": 1.003382949932341,
+      "grad_norm": 1.1028614850365017,
+      "learning_rate": 8.418141090531543e-06,
+      "loss": 0.4453,
+      "step": 1483
+    },
+    {
+      "epoch": 1.0040595399188093,
+      "grad_norm": 0.9976452983820293,
+      "learning_rate": 8.415266810841412e-06,
+      "loss": 0.4188,
+      "step": 1484
+    },
+    {
+      "epoch": 1.0047361299052775,
+      "grad_norm": 1.095388091254139,
+      "learning_rate": 8.412390413869944e-06,
+      "loss": 0.4539,
+      "step": 1485
+    },
+    {
+      "epoch": 1.0054127198917455,
+      "grad_norm": 1.0429180460374765,
+      "learning_rate": 8.409511901400351e-06,
+      "loss": 0.4235,
+      "step": 1486
+    },
+    {
+      "epoch": 1.0060893098782138,
+      "grad_norm": 1.0659771378200944,
+      "learning_rate": 8.406631275217156e-06,
+      "loss": 0.4343,
+      "step": 1487
+    },
+    {
+      "epoch": 1.006765899864682,
+      "grad_norm": 1.0523513744241333,
+      "learning_rate": 8.40374853710619e-06,
+      "loss": 0.4487,
+      "step": 1488
+    },
+    {
+      "epoch": 1.0074424898511503,
+      "grad_norm": 1.1118925477216597,
+      "learning_rate": 8.400863688854598e-06,
+      "loss": 0.4507,
+      "step": 1489
+    },
+    {
+      "epoch": 1.0081190798376185,
+      "grad_norm": 1.0249195627264522,
+      "learning_rate": 8.397976732250827e-06,
+      "loss": 0.4186,
+      "step": 1490
+    },
+    {
+      "epoch": 1.0087956698240865,
+      "grad_norm": 1.091226462243931,
+      "learning_rate": 8.395087669084638e-06,
+      "loss": 0.4297,
+      "step": 1491
+    },
+    {
+      "epoch": 1.0094722598105548,
+      "grad_norm": 1.1595732717937002,
+      "learning_rate": 8.392196501147092e-06,
+      "loss": 0.4315,
+      "step": 1492
+    },
+    {
+      "epoch": 1.010148849797023,
+      "grad_norm": 1.102426306487242,
+      "learning_rate": 8.389303230230556e-06,
+      "loss": 0.4451,
+      "step": 1493
+    },
+    {
+      "epoch": 1.0108254397834913,
+      "grad_norm": 1.1335299423993308,
+      "learning_rate": 8.386407858128707e-06,
+      "loss": 0.4573,
+      "step": 1494
+    },
+    {
+      "epoch": 1.0115020297699595,
+      "grad_norm": 1.047802436754095,
+      "learning_rate": 8.383510386636516e-06,
+      "loss": 0.4198,
+      "step": 1495
+    },
+    {
+      "epoch": 1.0121786197564275,
+      "grad_norm": 1.0871744256064688,
+      "learning_rate": 8.380610817550256e-06,
+      "loss": 0.4385,
+      "step": 1496
+    },
+    {
+      "epoch": 1.0128552097428958,
+      "grad_norm": 1.1213217903098696,
+      "learning_rate": 8.377709152667513e-06,
+      "loss": 0.4608,
+      "step": 1497
+    },
+    {
+      "epoch": 1.013531799729364,
+      "grad_norm": 1.0568178302832327,
+      "learning_rate": 8.374805393787154e-06,
+      "loss": 0.4295,
+      "step": 1498
+    },
+    {
+      "epoch": 1.0142083897158323,
+      "grad_norm": 1.0866289032102656,
+      "learning_rate": 8.371899542709355e-06,
+      "loss": 0.4371,
+      "step": 1499
+    },
+    {
+      "epoch": 1.0148849797023005,
+      "grad_norm": 1.0691387752221155,
+      "learning_rate": 8.36899160123559e-06,
+      "loss": 0.4357,
+      "step": 1500
+    },
+    {
+      "epoch": 1.0155615696887685,
+      "grad_norm": 1.056073550598995,
+      "learning_rate": 8.366081571168625e-06,
+      "loss": 0.4327,
+      "step": 1501
+    },
+    {
+      "epoch": 1.0162381596752368,
+      "grad_norm": 1.1018468294030312,
+      "learning_rate": 8.363169454312518e-06,
+      "loss": 0.4437,
+      "step": 1502
+    },
+    {
+      "epoch": 1.016914749661705,
+      "grad_norm": 1.142094501778782,
+      "learning_rate": 8.36025525247263e-06,
+      "loss": 0.4566,
+      "step": 1503
+    },
+    {
+      "epoch": 1.0175913396481733,
+      "grad_norm": 1.1384720029539852,
+      "learning_rate": 8.357338967455605e-06,
+      "loss": 0.4558,
+      "step": 1504
+    },
+    {
+      "epoch": 1.0182679296346413,
+      "grad_norm": 1.0774145244239763,
+      "learning_rate": 8.354420601069384e-06,
+      "loss": 0.416,
+      "step": 1505
+    },
+    {
+      "epoch": 1.0189445196211095,
+      "grad_norm": 1.0638346943367993,
+      "learning_rate": 8.3515001551232e-06,
+      "loss": 0.4091,
+      "step": 1506
+    },
+    {
+      "epoch": 1.0196211096075778,
+      "grad_norm": 1.0742898883746654,
+      "learning_rate": 8.348577631427565e-06,
+      "loss": 0.4243,
+      "step": 1507
+    },
+    {
+      "epoch": 1.020297699594046,
+      "grad_norm": 1.1129992247357157,
+      "learning_rate": 8.345653031794292e-06,
+      "loss": 0.4484,
+      "step": 1508
+    },
+    {
+      "epoch": 1.0209742895805143,
+      "grad_norm": 1.1382557195899554,
+      "learning_rate": 8.342726358036473e-06,
+      "loss": 0.4537,
+      "step": 1509
+    },
+    {
+      "epoch": 1.0216508795669823,
+      "grad_norm": 1.1258317090624272,
+      "learning_rate": 8.339797611968488e-06,
+      "loss": 0.4212,
+      "step": 1510
+    },
+    {
+      "epoch": 1.0223274695534506,
+      "grad_norm": 1.0670788201964598,
+      "learning_rate": 8.336866795406003e-06,
+      "loss": 0.4386,
+      "step": 1511
+    },
+    {
+      "epoch": 1.0230040595399188,
+      "grad_norm": 1.080078835351708,
+      "learning_rate": 8.333933910165964e-06,
+      "loss": 0.4304,
+      "step": 1512
+    },
+    {
+      "epoch": 1.023680649526387,
+      "grad_norm": 1.0720296166600187,
+      "learning_rate": 8.3309989580666e-06,
+      "loss": 0.4505,
+      "step": 1513
+    },
+    {
+      "epoch": 1.0243572395128553,
+      "grad_norm": 1.086828424479624,
+      "learning_rate": 8.32806194092743e-06,
+      "loss": 0.4438,
+      "step": 1514
+    },
+    {
+      "epoch": 1.0250338294993233,
+      "grad_norm": 1.0861309314852288,
+      "learning_rate": 8.325122860569241e-06,
+      "loss": 0.4399,
+      "step": 1515
+    },
+    {
+      "epoch": 1.0257104194857916,
+      "grad_norm": 1.0419786852269919,
+      "learning_rate": 8.322181718814107e-06,
+      "loss": 0.4522,
+      "step": 1516
+    },
+    {
+      "epoch": 1.0263870094722598,
+      "grad_norm": 1.1276730398979562,
+      "learning_rate": 8.319238517485376e-06,
+      "loss": 0.4376,
+      "step": 1517
+    },
+    {
+      "epoch": 1.027063599458728,
+      "grad_norm": 1.0543197707149141,
+      "learning_rate": 8.316293258407673e-06,
+      "loss": 0.4276,
+      "step": 1518
+    },
+    {
+      "epoch": 1.0277401894451963,
+      "grad_norm": 1.0992416942236956,
+      "learning_rate": 8.313345943406903e-06,
+      "loss": 0.45,
+      "step": 1519
+    },
+    {
+      "epoch": 1.0284167794316643,
+      "grad_norm": 1.04032592481762,
+      "learning_rate": 8.310396574310239e-06,
+      "loss": 0.4142,
+      "step": 1520
+    },
+    {
+      "epoch": 1.0290933694181326,
+      "grad_norm": 1.0538815128264054,
+      "learning_rate": 8.307445152946133e-06,
+      "loss": 0.4455,
+      "step": 1521
+    },
+    {
+      "epoch": 1.0297699594046008,
+      "grad_norm": 1.0511836445262603,
+      "learning_rate": 8.304491681144306e-06,
+      "loss": 0.4416,
+      "step": 1522
+    },
+    {
+      "epoch": 1.030446549391069,
+      "grad_norm": 1.0377368559914297,
+      "learning_rate": 8.301536160735752e-06,
+      "loss": 0.419,
+      "step": 1523
+    },
+    {
+      "epoch": 1.0311231393775373,
+      "grad_norm": 1.0648575048510316,
+      "learning_rate": 8.298578593552737e-06,
+      "loss": 0.4403,
+      "step": 1524
+    },
+    {
+      "epoch": 1.0317997293640053,
+      "grad_norm": 1.0027013773600582,
+      "learning_rate": 8.295618981428788e-06,
+      "loss": 0.4161,
+      "step": 1525
+    },
+    {
+      "epoch": 1.0324763193504736,
+      "grad_norm": 1.100756799313297,
+      "learning_rate": 8.292657326198707e-06,
+      "loss": 0.4471,
+      "step": 1526
+    },
+    {
+      "epoch": 1.0331529093369418,
+      "grad_norm": 1.0709669894813856,
+      "learning_rate": 8.289693629698564e-06,
+      "loss": 0.4206,
+      "step": 1527
+    },
+    {
+      "epoch": 1.03382949932341,
+      "grad_norm": 1.1721190145944764,
+      "learning_rate": 8.286727893765687e-06,
+      "loss": 0.463,
+      "step": 1528
+    },
+    {
+      "epoch": 1.0345060893098783,
+      "grad_norm": 1.0595164921651812,
+      "learning_rate": 8.283760120238672e-06,
+      "loss": 0.4199,
+      "step": 1529
+    },
+    {
+      "epoch": 1.0351826792963463,
+      "grad_norm": 1.1084052432329463,
+      "learning_rate": 8.280790310957382e-06,
+      "loss": 0.436,
+      "step": 1530
+    },
+    {
+      "epoch": 1.0358592692828146,
+      "grad_norm": 1.1029448130094104,
+      "learning_rate": 8.277818467762937e-06,
+      "loss": 0.4217,
+      "step": 1531
+    },
+    {
+      "epoch": 1.0365358592692828,
+      "grad_norm": 1.0702473614609793,
+      "learning_rate": 8.27484459249772e-06,
+      "loss": 0.4369,
+      "step": 1532
+    },
+    {
+      "epoch": 1.037212449255751,
+      "grad_norm": 1.0715283201679295,
+      "learning_rate": 8.271868687005371e-06,
+      "loss": 0.4499,
+      "step": 1533
+    },
+    {
+      "epoch": 1.0378890392422193,
+      "grad_norm": 1.0883918479678936,
+      "learning_rate": 8.268890753130794e-06,
+      "loss": 0.4409,
+      "step": 1534
+    },
+    {
+      "epoch": 1.0385656292286873,
+      "grad_norm": 1.150533123005341,
+      "learning_rate": 8.265910792720147e-06,
+      "loss": 0.443,
+      "step": 1535
+    },
+    {
+      "epoch": 1.0392422192151556,
+      "grad_norm": 1.050674093677084,
+      "learning_rate": 8.262928807620843e-06,
+      "loss": 0.4245,
+      "step": 1536
+    },
+    {
+      "epoch": 1.0399188092016238,
+      "grad_norm": 1.0584633423392775,
+      "learning_rate": 8.259944799681555e-06,
+      "loss": 0.458,
+      "step": 1537
+    },
+    {
+      "epoch": 1.040595399188092,
+      "grad_norm": 1.0235428089559888,
+      "learning_rate": 8.256958770752203e-06,
+      "loss": 0.4144,
+      "step": 1538
+    },
+    {
+      "epoch": 1.0412719891745603,
+      "grad_norm": 1.129290621568861,
+      "learning_rate": 8.253970722683968e-06,
+      "loss": 0.4257,
+      "step": 1539
+    },
+    {
+      "epoch": 1.0419485791610283,
+      "grad_norm": 1.105278970568012,
+      "learning_rate": 8.250980657329278e-06,
+      "loss": 0.4439,
+      "step": 1540
+    },
+    {
+      "epoch": 1.0426251691474966,
+      "grad_norm": 1.10926981882421,
+      "learning_rate": 8.24798857654181e-06,
+      "loss": 0.4383,
+      "step": 1541
+    },
+    {
+      "epoch": 1.0433017591339648,
+      "grad_norm": 1.081545700249644,
+      "learning_rate": 8.244994482176495e-06,
+      "loss": 0.419,
+      "step": 1542
+    },
+    {
+      "epoch": 1.043978349120433,
+      "grad_norm": 1.0848326455416037,
+      "learning_rate": 8.241998376089508e-06,
+      "loss": 0.4323,
+      "step": 1543
+    },
+    {
+      "epoch": 1.044654939106901,
+      "grad_norm": 1.0956179683369176,
+      "learning_rate": 8.239000260138277e-06,
+      "loss": 0.4503,
+      "step": 1544
+    },
+    {
+      "epoch": 1.0453315290933693,
+      "grad_norm": 1.117039820391544,
+      "learning_rate": 8.236000136181468e-06,
+      "loss": 0.4533,
+      "step": 1545
+    },
+    {
+      "epoch": 1.0460081190798376,
+      "grad_norm": 1.0843051570447386,
+      "learning_rate": 8.232998006078998e-06,
+      "loss": 0.4451,
+      "step": 1546
+    },
+    {
+      "epoch": 1.0466847090663058,
+      "grad_norm": 1.0446116875656069,
+      "learning_rate": 8.229993871692028e-06,
+      "loss": 0.4231,
+      "step": 1547
+    },
+    {
+      "epoch": 1.047361299052774,
+      "grad_norm": 1.0394361472364462,
+      "learning_rate": 8.226987734882956e-06,
+      "loss": 0.4231,
+      "step": 1548
+    },
+    {
+      "epoch": 1.048037889039242,
+      "grad_norm": 1.107479284428652,
+      "learning_rate": 8.223979597515425e-06,
+      "loss": 0.4157,
+      "step": 1549
+    },
+    {
+      "epoch": 1.0487144790257104,
+      "grad_norm": 1.079236083196405,
+      "learning_rate": 8.220969461454322e-06,
+      "loss": 0.4323,
+      "step": 1550
+    },
+    {
+      "epoch": 1.0493910690121786,
+      "grad_norm": 1.1201728023363162,
+      "learning_rate": 8.217957328565765e-06,
+      "loss": 0.4481,
+      "step": 1551
+    },
+    {
+      "epoch": 1.0500676589986468,
+      "grad_norm": 1.1327604344703373,
+      "learning_rate": 8.214943200717114e-06,
+      "loss": 0.4624,
+      "step": 1552
+    },
+    {
+      "epoch": 1.050744248985115,
+      "grad_norm": 1.079497183316191,
+      "learning_rate": 8.211927079776969e-06,
+      "loss": 0.4398,
+      "step": 1553
+    },
+    {
+      "epoch": 1.0514208389715831,
+      "grad_norm": 1.0822644057515787,
+      "learning_rate": 8.208908967615159e-06,
+      "loss": 0.4321,
+      "step": 1554
+    },
+    {
+      "epoch": 1.0520974289580514,
+      "grad_norm": 1.0670379255597637,
+      "learning_rate": 8.205888866102753e-06,
+      "loss": 0.4209,
+      "step": 1555
+    },
+    {
+      "epoch": 1.0527740189445196,
+      "grad_norm": 1.0584843192491298,
+      "learning_rate": 8.202866777112049e-06,
+      "loss": 0.4274,
+      "step": 1556
+    },
+    {
+      "epoch": 1.0534506089309879,
+      "grad_norm": 1.0825157020208016,
+      "learning_rate": 8.199842702516584e-06,
+      "loss": 0.4231,
+      "step": 1557
+    },
+    {
+      "epoch": 1.054127198917456,
+      "grad_norm": 1.181058072108386,
+      "learning_rate": 8.196816644191116e-06,
+      "loss": 0.4543,
+      "step": 1558
+    },
+    {
+      "epoch": 1.0548037889039241,
+      "grad_norm": 1.1465126958375453,
+      "learning_rate": 8.193788604011639e-06,
+      "loss": 0.4317,
+      "step": 1559
+    },
+    {
+      "epoch": 1.0554803788903924,
+      "grad_norm": 1.1798399915490474,
+      "learning_rate": 8.190758583855379e-06,
+      "loss": 0.4373,
+      "step": 1560
+    },
+    {
+      "epoch": 1.0561569688768606,
+      "grad_norm": 1.0958346515864203,
+      "learning_rate": 8.187726585600779e-06,
+      "loss": 0.4418,
+      "step": 1561
+    },
+    {
+      "epoch": 1.0568335588633289,
+      "grad_norm": 1.0854684147062201,
+      "learning_rate": 8.18469261112752e-06,
+      "loss": 0.4438,
+      "step": 1562
+    },
+    {
+      "epoch": 1.057510148849797,
+      "grad_norm": 1.120287507520833,
+      "learning_rate": 8.181656662316498e-06,
+      "loss": 0.4565,
+      "step": 1563
+    },
+    {
+      "epoch": 1.0581867388362651,
+      "grad_norm": 1.1343778766597947,
+      "learning_rate": 8.178618741049841e-06,
+      "loss": 0.4498,
+      "step": 1564
+    },
+    {
+      "epoch": 1.0588633288227334,
+      "grad_norm": 1.1136835433012175,
+      "learning_rate": 8.175578849210894e-06,
+      "loss": 0.4388,
+      "step": 1565
+    },
+    {
+      "epoch": 1.0595399188092016,
+      "grad_norm": 1.0857050880221115,
+      "learning_rate": 8.172536988684227e-06,
+      "loss": 0.4312,
+      "step": 1566
+    },
+    {
+      "epoch": 1.0602165087956699,
+      "grad_norm": 1.1483632473808032,
+      "learning_rate": 8.169493161355632e-06,
+      "loss": 0.4456,
+      "step": 1567
+    },
+    {
+      "epoch": 1.060893098782138,
+      "grad_norm": 1.0666774990800618,
+      "learning_rate": 8.166447369112115e-06,
+      "loss": 0.4272,
+      "step": 1568
+    },
+    {
+      "epoch": 1.0615696887686061,
+      "grad_norm": 1.1191435336157294,
+      "learning_rate": 8.163399613841903e-06,
+      "loss": 0.4575,
+      "step": 1569
+    },
+    {
+      "epoch": 1.0622462787550744,
+      "grad_norm": 1.1086911588514294,
+      "learning_rate": 8.160349897434441e-06,
+      "loss": 0.4386,
+      "step": 1570
+    },
+    {
+      "epoch": 1.0629228687415426,
+      "grad_norm": 1.0592884396850444,
+      "learning_rate": 8.157298221780388e-06,
+      "loss": 0.4127,
+      "step": 1571
+    },
+    {
+      "epoch": 1.0635994587280109,
+      "grad_norm": 1.0673531117047905,
+      "learning_rate": 8.15424458877162e-06,
+      "loss": 0.4336,
+      "step": 1572
+    },
+    {
+      "epoch": 1.0642760487144791,
+      "grad_norm": 1.0447219327894612,
+      "learning_rate": 8.151189000301223e-06,
+      "loss": 0.4187,
+      "step": 1573
+    },
+    {
+      "epoch": 1.0649526387009471,
+      "grad_norm": 1.1356290836158105,
+      "learning_rate": 8.148131458263499e-06,
+      "loss": 0.4403,
+      "step": 1574
+    },
+    {
+      "epoch": 1.0656292286874154,
+      "grad_norm": 1.067719985999696,
+      "learning_rate": 8.145071964553956e-06,
+      "loss": 0.4495,
+      "step": 1575
+    },
+    {
+      "epoch": 1.0663058186738836,
+      "grad_norm": 1.0700587064794442,
+      "learning_rate": 8.142010521069319e-06,
+      "loss": 0.4354,
+      "step": 1576
+    },
+    {
+      "epoch": 1.0669824086603519,
+      "grad_norm": 1.066775298182691,
+      "learning_rate": 8.138947129707517e-06,
+      "loss": 0.4459,
+      "step": 1577
+    },
+    {
+      "epoch": 1.0676589986468201,
+      "grad_norm": 1.015095176896773,
+      "learning_rate": 8.135881792367686e-06,
+      "loss": 0.4104,
+      "step": 1578
+    },
+    {
+      "epoch": 1.0683355886332881,
+      "grad_norm": 1.093006565458466,
+      "learning_rate": 8.132814510950172e-06,
+      "loss": 0.443,
+      "step": 1579
+    },
+    {
+      "epoch": 1.0690121786197564,
+      "grad_norm": 1.1394810941904618,
+      "learning_rate": 8.129745287356521e-06,
+      "loss": 0.4419,
+      "step": 1580
+    },
+    {
+      "epoch": 1.0696887686062246,
+      "grad_norm": 1.0757647619739268,
+      "learning_rate": 8.12667412348949e-06,
+      "loss": 0.4099,
+      "step": 1581
+    },
+    {
+      "epoch": 1.0703653585926929,
+      "grad_norm": 1.0635248887059734,
+      "learning_rate": 8.12360102125303e-06,
+      "loss": 0.4362,
+      "step": 1582
+    },
+    {
+      "epoch": 1.0710419485791611,
+      "grad_norm": 1.1229702673160813,
+      "learning_rate": 8.120525982552304e-06,
+      "loss": 0.4478,
+      "step": 1583
+    },
+    {
+      "epoch": 1.0717185385656292,
+      "grad_norm": 1.1139017704916843,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.4457,
+      "step": 1584
+    },
+    {
+      "epoch": 1.0723951285520974,
+      "grad_norm": 1.1654561080780221,
+      "learning_rate": 8.11437010338468e-06,
+      "loss": 0.4449,
+      "step": 1585
+    },
+    {
+      "epoch": 1.0730717185385656,
+      "grad_norm": 1.0648716386807038,
+      "learning_rate": 8.111289266734095e-06,
+      "loss": 0.4176,
+      "step": 1586
+    },
+    {
+      "epoch": 1.073748308525034,
+      "grad_norm": 1.0903946735801888,
+      "learning_rate": 8.108206501251868e-06,
+      "loss": 0.4388,
+      "step": 1587
+    },
+    {
+      "epoch": 1.0744248985115021,
+      "grad_norm": 1.1089102923249174,
+      "learning_rate": 8.105121808849143e-06,
+      "loss": 0.4386,
+      "step": 1588
+    },
+    {
+      "epoch": 1.0751014884979702,
+      "grad_norm": 1.10952490226382,
+      "learning_rate": 8.102035191438268e-06,
+      "loss": 0.4499,
+      "step": 1589
+    },
+    {
+      "epoch": 1.0757780784844384,
+      "grad_norm": 1.0363507803276348,
+      "learning_rate": 8.098946650932776e-06,
+      "loss": 0.4223,
+      "step": 1590
+    },
+    {
+      "epoch": 1.0764546684709067,
+      "grad_norm": 1.1276298629414534,
+      "learning_rate": 8.095856189247396e-06,
+      "loss": 0.464,
+      "step": 1591
+    },
+    {
+      "epoch": 1.077131258457375,
+      "grad_norm": 1.0915285211783439,
+      "learning_rate": 8.092763808298048e-06,
+      "loss": 0.4268,
+      "step": 1592
+    },
+    {
+      "epoch": 1.0778078484438431,
+      "grad_norm": 1.0929302102253826,
+      "learning_rate": 8.089669510001843e-06,
+      "loss": 0.4348,
+      "step": 1593
+    },
+    {
+      "epoch": 1.0784844384303112,
+      "grad_norm": 1.0418638573485512,
+      "learning_rate": 8.086573296277078e-06,
+      "loss": 0.4188,
+      "step": 1594
+    },
+    {
+      "epoch": 1.0791610284167794,
+      "grad_norm": 1.1262701903447359,
+      "learning_rate": 8.083475169043237e-06,
+      "loss": 0.442,
+      "step": 1595
+    },
+    {
+      "epoch": 1.0798376184032477,
+      "grad_norm": 1.0278587665611612,
+      "learning_rate": 8.080375130220995e-06,
+      "loss": 0.4231,
+      "step": 1596
+    },
+    {
+      "epoch": 1.080514208389716,
+      "grad_norm": 1.0563790251020893,
+      "learning_rate": 8.077273181732207e-06,
+      "loss": 0.4177,
+      "step": 1597
+    },
+    {
+      "epoch": 1.0811907983761841,
+      "grad_norm": 1.0920356215333442,
+      "learning_rate": 8.074169325499915e-06,
+      "loss": 0.4391,
+      "step": 1598
+    },
+    {
+      "epoch": 1.0818673883626522,
+      "grad_norm": 1.1080050801969525,
+      "learning_rate": 8.071063563448341e-06,
+      "loss": 0.4494,
+      "step": 1599
+    },
+    {
+      "epoch": 1.0825439783491204,
+      "grad_norm": 1.0302226769676113,
+      "learning_rate": 8.06795589750289e-06,
+      "loss": 0.4275,
+      "step": 1600
+    },
+    {
+      "epoch": 1.0832205683355887,
+      "grad_norm": 1.0325206481334062,
+      "learning_rate": 8.06484632959015e-06,
+      "loss": 0.4297,
+      "step": 1601
+    },
+    {
+      "epoch": 1.083897158322057,
+      "grad_norm": 1.0812778227744115,
+      "learning_rate": 8.061734861637883e-06,
+      "loss": 0.431,
+      "step": 1602
+    },
+    {
+      "epoch": 1.084573748308525,
+      "grad_norm": 1.080243386973949,
+      "learning_rate": 8.058621495575032e-06,
+      "loss": 0.4419,
+      "step": 1603
+    },
+    {
+      "epoch": 1.0852503382949932,
+      "grad_norm": 1.0736255387871239,
+      "learning_rate": 8.055506233331718e-06,
+      "loss": 0.4333,
+      "step": 1604
+    },
+    {
+      "epoch": 1.0859269282814614,
+      "grad_norm": 1.0677306821071324,
+      "learning_rate": 8.052389076839233e-06,
+      "loss": 0.4346,
+      "step": 1605
+    },
+    {
+      "epoch": 1.0866035182679297,
+      "grad_norm": 1.1076916294256132,
+      "learning_rate": 8.049270028030045e-06,
+      "loss": 0.4269,
+      "step": 1606
+    },
+    {
+      "epoch": 1.087280108254398,
+      "grad_norm": 1.0213960167654488,
+      "learning_rate": 8.046149088837803e-06,
+      "loss": 0.4278,
+      "step": 1607
+    },
+    {
+      "epoch": 1.087956698240866,
+      "grad_norm": 1.0634181798533768,
+      "learning_rate": 8.043026261197312e-06,
+      "loss": 0.4283,
+      "step": 1608
+    },
+    {
+      "epoch": 1.0886332882273342,
+      "grad_norm": 1.1108831071175451,
+      "learning_rate": 8.039901547044564e-06,
+      "loss": 0.4376,
+      "step": 1609
+    },
+    {
+      "epoch": 1.0893098782138024,
+      "grad_norm": 1.1452809988683992,
+      "learning_rate": 8.03677494831671e-06,
+      "loss": 0.4391,
+      "step": 1610
+    },
+    {
+      "epoch": 1.0899864682002707,
+      "grad_norm": 1.0419528450291402,
+      "learning_rate": 8.033646466952072e-06,
+      "loss": 0.4256,
+      "step": 1611
+    },
+    {
+      "epoch": 1.090663058186739,
+      "grad_norm": 1.072886295704959,
+      "learning_rate": 8.03051610489014e-06,
+      "loss": 0.4272,
+      "step": 1612
+    },
+    {
+      "epoch": 1.091339648173207,
+      "grad_norm": 1.0659495437851878,
+      "learning_rate": 8.027383864071573e-06,
+      "loss": 0.4238,
+      "step": 1613
+    },
+    {
+      "epoch": 1.0920162381596752,
+      "grad_norm": 1.037692414853626,
+      "learning_rate": 8.024249746438189e-06,
+      "loss": 0.4231,
+      "step": 1614
+    },
+    {
+      "epoch": 1.0926928281461434,
+      "grad_norm": 1.054437815665456,
+      "learning_rate": 8.021113753932972e-06,
+      "loss": 0.4244,
+      "step": 1615
+    },
+    {
+      "epoch": 1.0933694181326117,
+      "grad_norm": 1.0311388550159604,
+      "learning_rate": 8.017975888500067e-06,
+      "loss": 0.4125,
+      "step": 1616
+    },
+    {
+      "epoch": 1.09404600811908,
+      "grad_norm": 1.082570047344452,
+      "learning_rate": 8.014836152084784e-06,
+      "loss": 0.4128,
+      "step": 1617
+    },
+    {
+      "epoch": 1.094722598105548,
+      "grad_norm": 1.1164160306228927,
+      "learning_rate": 8.01169454663359e-06,
+      "loss": 0.4283,
+      "step": 1618
+    },
+    {
+      "epoch": 1.0953991880920162,
+      "grad_norm": 1.1089682584435943,
+      "learning_rate": 8.008551074094108e-06,
+      "loss": 0.4249,
+      "step": 1619
+    },
+    {
+      "epoch": 1.0960757780784844,
+      "grad_norm": 1.0509464150438623,
+      "learning_rate": 8.005405736415127e-06,
+      "loss": 0.4338,
+      "step": 1620
+    },
+    {
+      "epoch": 1.0967523680649527,
+      "grad_norm": 1.1040363318896815,
+      "learning_rate": 8.00225853554658e-06,
+      "loss": 0.4366,
+      "step": 1621
+    },
+    {
+      "epoch": 1.097428958051421,
+      "grad_norm": 1.0837023230855898,
+      "learning_rate": 7.99910947343957e-06,
+      "loss": 0.4384,
+      "step": 1622
+    },
+    {
+      "epoch": 1.098105548037889,
+      "grad_norm": 0.9985884908899905,
+      "learning_rate": 7.995958552046338e-06,
+      "loss": 0.4087,
+      "step": 1623
+    },
+    {
+      "epoch": 1.0987821380243572,
+      "grad_norm": 1.1238500609596405,
+      "learning_rate": 7.99280577332029e-06,
+      "loss": 0.4446,
+      "step": 1624
+    },
+    {
+      "epoch": 1.0994587280108254,
+      "grad_norm": 1.0806190421892934,
+      "learning_rate": 7.989651139215979e-06,
+      "loss": 0.4263,
+      "step": 1625
+    },
+    {
+      "epoch": 1.1001353179972937,
+      "grad_norm": 1.1036194550943939,
+      "learning_rate": 7.986494651689104e-06,
+      "loss": 0.4327,
+      "step": 1626
+    },
+    {
+      "epoch": 1.100811907983762,
+      "grad_norm": 1.0420743085626456,
+      "learning_rate": 7.983336312696521e-06,
+      "loss": 0.401,
+      "step": 1627
+    },
+    {
+      "epoch": 1.10148849797023,
+      "grad_norm": 1.1034746489027385,
+      "learning_rate": 7.980176124196231e-06,
+      "loss": 0.4186,
+      "step": 1628
+    },
+    {
+      "epoch": 1.1021650879566982,
+      "grad_norm": 1.0759985686089186,
+      "learning_rate": 7.977014088147375e-06,
+      "loss": 0.4426,
+      "step": 1629
+    },
+    {
+      "epoch": 1.1028416779431665,
+      "grad_norm": 1.0744308311896629,
+      "learning_rate": 7.973850206510251e-06,
+      "loss": 0.4208,
+      "step": 1630
+    },
+    {
+      "epoch": 1.1035182679296347,
+      "grad_norm": 1.0771285237222417,
+      "learning_rate": 7.970684481246291e-06,
+      "loss": 0.4208,
+      "step": 1631
+    },
+    {
+      "epoch": 1.104194857916103,
+      "grad_norm": 1.0803456413753414,
+      "learning_rate": 7.967516914318075e-06,
+      "loss": 0.4373,
+      "step": 1632
+    },
+    {
+      "epoch": 1.104871447902571,
+      "grad_norm": 1.0294107599945008,
+      "learning_rate": 7.964347507689325e-06,
+      "loss": 0.4179,
+      "step": 1633
+    },
+    {
+      "epoch": 1.1055480378890392,
+      "grad_norm": 1.0942476140701136,
+      "learning_rate": 7.961176263324902e-06,
+      "loss": 0.4409,
+      "step": 1634
+    },
+    {
+      "epoch": 1.1062246278755075,
+      "grad_norm": 1.0550016369795785,
+      "learning_rate": 7.958003183190804e-06,
+      "loss": 0.4229,
+      "step": 1635
+    },
+    {
+      "epoch": 1.1069012178619757,
+      "grad_norm": 1.0228007409500721,
+      "learning_rate": 7.954828269254173e-06,
+      "loss": 0.4125,
+      "step": 1636
+    },
+    {
+      "epoch": 1.1075778078484437,
+      "grad_norm": 1.0810395794512198,
+      "learning_rate": 7.951651523483283e-06,
+      "loss": 0.4216,
+      "step": 1637
+    },
+    {
+      "epoch": 1.108254397834912,
+      "grad_norm": 1.1131415243707072,
+      "learning_rate": 7.948472947847546e-06,
+      "loss": 0.4459,
+      "step": 1638
+    },
+    {
+      "epoch": 1.1089309878213802,
+      "grad_norm": 1.1130044089257685,
+      "learning_rate": 7.945292544317505e-06,
+      "loss": 0.4293,
+      "step": 1639
+    },
+    {
+      "epoch": 1.1096075778078485,
+      "grad_norm": 1.0714274606208691,
+      "learning_rate": 7.942110314864842e-06,
+      "loss": 0.4462,
+      "step": 1640
+    },
+    {
+      "epoch": 1.1102841677943167,
+      "grad_norm": 1.0671923969322772,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.4472,
+      "step": 1641
+    },
+    {
+      "epoch": 1.1109607577807847,
+      "grad_norm": 1.0769530076724063,
+      "learning_rate": 7.93574038608402e-06,
+      "loss": 0.4363,
+      "step": 1642
+    },
+    {
+      "epoch": 1.111637347767253,
+      "grad_norm": 1.024809503611195,
+      "learning_rate": 7.932552690704871e-06,
+      "loss": 0.4316,
+      "step": 1643
+    },
+    {
+      "epoch": 1.1123139377537212,
+      "grad_norm": 1.103139310008525,
+      "learning_rate": 7.929363177301124e-06,
+      "loss": 0.4363,
+      "step": 1644
+    },
+    {
+      "epoch": 1.1129905277401895,
+      "grad_norm": 1.0840748778510945,
+      "learning_rate": 7.926171847850101e-06,
+      "loss": 0.4403,
+      "step": 1645
+    },
+    {
+      "epoch": 1.1136671177266577,
+      "grad_norm": 1.1234584719469616,
+      "learning_rate": 7.922978704330257e-06,
+      "loss": 0.4454,
+      "step": 1646
+    },
+    {
+      "epoch": 1.1143437077131257,
+      "grad_norm": 1.047915762428863,
+      "learning_rate": 7.919783748721169e-06,
+      "loss": 0.4319,
+      "step": 1647
+    },
+    {
+      "epoch": 1.115020297699594,
+      "grad_norm": 1.047247514056654,
+      "learning_rate": 7.916586983003534e-06,
+      "loss": 0.4294,
+      "step": 1648
+    },
+    {
+      "epoch": 1.1156968876860622,
+      "grad_norm": 1.017429778766552,
+      "learning_rate": 7.913388409159175e-06,
+      "loss": 0.401,
+      "step": 1649
+    },
+    {
+      "epoch": 1.1163734776725305,
+      "grad_norm": 1.0535665838044446,
+      "learning_rate": 7.910188029171039e-06,
+      "loss": 0.4375,
+      "step": 1650
+    },
+    {
+      "epoch": 1.1170500676589987,
+      "grad_norm": 1.0442179948495736,
+      "learning_rate": 7.906985845023187e-06,
+      "loss": 0.4111,
+      "step": 1651
+    },
+    {
+      "epoch": 1.1177266576454667,
+      "grad_norm": 1.132898799560178,
+      "learning_rate": 7.903781858700799e-06,
+      "loss": 0.449,
+      "step": 1652
+    },
+    {
+      "epoch": 1.118403247631935,
+      "grad_norm": 1.0525393249078858,
+      "learning_rate": 7.900576072190177e-06,
+      "loss": 0.4283,
+      "step": 1653
+    },
+    {
+      "epoch": 1.1190798376184032,
+      "grad_norm": 1.0962682310224106,
+      "learning_rate": 7.897368487478733e-06,
+      "loss": 0.435,
+      "step": 1654
+    },
+    {
+      "epoch": 1.1197564276048715,
+      "grad_norm": 1.1492955729743703,
+      "learning_rate": 7.894159106554997e-06,
+      "loss": 0.4382,
+      "step": 1655
+    },
+    {
+      "epoch": 1.1204330175913397,
+      "grad_norm": 1.1188180147367437,
+      "learning_rate": 7.890947931408614e-06,
+      "loss": 0.4313,
+      "step": 1656
+    },
+    {
+      "epoch": 1.1211096075778078,
+      "grad_norm": 1.0602995977916774,
+      "learning_rate": 7.887734964030337e-06,
+      "loss": 0.4404,
+      "step": 1657
+    },
+    {
+      "epoch": 1.121786197564276,
+      "grad_norm": 1.1525937637165338,
+      "learning_rate": 7.884520206412036e-06,
+      "loss": 0.4456,
+      "step": 1658
+    },
+    {
+      "epoch": 1.1224627875507442,
+      "grad_norm": 1.092653723544954,
+      "learning_rate": 7.881303660546684e-06,
+      "loss": 0.4438,
+      "step": 1659
+    },
+    {
+      "epoch": 1.1231393775372125,
+      "grad_norm": 1.100506828048419,
+      "learning_rate": 7.87808532842837e-06,
+      "loss": 0.4472,
+      "step": 1660
+    },
+    {
+      "epoch": 1.1238159675236807,
+      "grad_norm": 1.1236745282818088,
+      "learning_rate": 7.87486521205228e-06,
+      "loss": 0.4485,
+      "step": 1661
+    },
+    {
+      "epoch": 1.1244925575101488,
+      "grad_norm": 1.0868924336816534,
+      "learning_rate": 7.871643313414718e-06,
+      "loss": 0.4328,
+      "step": 1662
+    },
+    {
+      "epoch": 1.125169147496617,
+      "grad_norm": 1.1118360878728375,
+      "learning_rate": 7.868419634513087e-06,
+      "loss": 0.4251,
+      "step": 1663
+    },
+    {
+      "epoch": 1.1258457374830853,
+      "grad_norm": 1.0614445418034166,
+      "learning_rate": 7.865194177345894e-06,
+      "loss": 0.4133,
+      "step": 1664
+    },
+    {
+      "epoch": 1.1265223274695535,
+      "grad_norm": 1.0529800234272884,
+      "learning_rate": 7.861966943912746e-06,
+      "loss": 0.4228,
+      "step": 1665
+    },
+    {
+      "epoch": 1.1271989174560217,
+      "grad_norm": 1.1584257350576903,
+      "learning_rate": 7.858737936214355e-06,
+      "loss": 0.4319,
+      "step": 1666
+    },
+    {
+      "epoch": 1.1278755074424898,
+      "grad_norm": 1.1349258108372389,
+      "learning_rate": 7.855507156252536e-06,
+      "loss": 0.4516,
+      "step": 1667
+    },
+    {
+      "epoch": 1.128552097428958,
+      "grad_norm": 1.0868559683076846,
+      "learning_rate": 7.852274606030191e-06,
+      "loss": 0.4331,
+      "step": 1668
+    },
+    {
+      "epoch": 1.1292286874154263,
+      "grad_norm": 1.0401928949814874,
+      "learning_rate": 7.849040287551331e-06,
+      "loss": 0.4358,
+      "step": 1669
+    },
+    {
+      "epoch": 1.1299052774018945,
+      "grad_norm": 1.0452570529625524,
+      "learning_rate": 7.84580420282106e-06,
+      "loss": 0.4344,
+      "step": 1670
+    },
+    {
+      "epoch": 1.1305818673883627,
+      "grad_norm": 1.117649207760153,
+      "learning_rate": 7.842566353845575e-06,
+      "loss": 0.4471,
+      "step": 1671
+    },
+    {
+      "epoch": 1.1312584573748308,
+      "grad_norm": 1.0851161080602902,
+      "learning_rate": 7.839326742632168e-06,
+      "loss": 0.425,
+      "step": 1672
+    },
+    {
+      "epoch": 1.131935047361299,
+      "grad_norm": 1.0466829112444798,
+      "learning_rate": 7.836085371189221e-06,
+      "loss": 0.4281,
+      "step": 1673
+    },
+    {
+      "epoch": 1.1326116373477673,
+      "grad_norm": 1.0475345778625593,
+      "learning_rate": 7.832842241526212e-06,
+      "loss": 0.4167,
+      "step": 1674
+    },
+    {
+      "epoch": 1.1332882273342355,
+      "grad_norm": 1.014363698097369,
+      "learning_rate": 7.829597355653707e-06,
+      "loss": 0.4098,
+      "step": 1675
+    },
+    {
+      "epoch": 1.1339648173207038,
+      "grad_norm": 1.0372480643891817,
+      "learning_rate": 7.82635071558336e-06,
+      "loss": 0.4379,
+      "step": 1676
+    },
+    {
+      "epoch": 1.1346414073071718,
+      "grad_norm": 1.0327954851408154,
+      "learning_rate": 7.82310232332791e-06,
+      "loss": 0.4272,
+      "step": 1677
+    },
+    {
+      "epoch": 1.13531799729364,
+      "grad_norm": 1.0835803359927758,
+      "learning_rate": 7.81985218090119e-06,
+      "loss": 0.4432,
+      "step": 1678
+    },
+    {
+      "epoch": 1.1359945872801083,
+      "grad_norm": 1.1131813840327025,
+      "learning_rate": 7.81660029031811e-06,
+      "loss": 0.444,
+      "step": 1679
+    },
+    {
+      "epoch": 1.1366711772665765,
+      "grad_norm": 1.080588584432741,
+      "learning_rate": 7.813346653594667e-06,
+      "loss": 0.4344,
+      "step": 1680
+    },
+    {
+      "epoch": 1.1373477672530448,
+      "grad_norm": 1.0744618414363065,
+      "learning_rate": 7.810091272747943e-06,
+      "loss": 0.4042,
+      "step": 1681
+    },
+    {
+      "epoch": 1.1380243572395128,
+      "grad_norm": 1.0936668071244344,
+      "learning_rate": 7.806834149796094e-06,
+      "loss": 0.4375,
+      "step": 1682
+    },
+    {
+      "epoch": 1.138700947225981,
+      "grad_norm": 1.1156391174164546,
+      "learning_rate": 7.803575286758365e-06,
+      "loss": 0.4416,
+      "step": 1683
+    },
+    {
+      "epoch": 1.1393775372124493,
+      "grad_norm": 1.0789725899731237,
+      "learning_rate": 7.800314685655072e-06,
+      "loss": 0.4342,
+      "step": 1684
+    },
+    {
+      "epoch": 1.1400541271989175,
+      "grad_norm": 1.1220753499974567,
+      "learning_rate": 7.797052348507614e-06,
+      "loss": 0.4468,
+      "step": 1685
+    },
+    {
+      "epoch": 1.1407307171853858,
+      "grad_norm": 1.0665818324168077,
+      "learning_rate": 7.793788277338464e-06,
+      "loss": 0.4237,
+      "step": 1686
+    },
+    {
+      "epoch": 1.1414073071718538,
+      "grad_norm": 1.0722274248439019,
+      "learning_rate": 7.790522474171171e-06,
+      "loss": 0.4257,
+      "step": 1687
+    },
+    {
+      "epoch": 1.142083897158322,
+      "grad_norm": 1.0955116714359816,
+      "learning_rate": 7.787254941030353e-06,
+      "loss": 0.428,
+      "step": 1688
+    },
+    {
+      "epoch": 1.1427604871447903,
+      "grad_norm": 1.0726365796467348,
+      "learning_rate": 7.78398567994171e-06,
+      "loss": 0.4148,
+      "step": 1689
+    },
+    {
+      "epoch": 1.1434370771312585,
+      "grad_norm": 1.0856929960125399,
+      "learning_rate": 7.780714692932002e-06,
+      "loss": 0.4265,
+      "step": 1690
+    },
+    {
+      "epoch": 1.1441136671177268,
+      "grad_norm": 1.0473114825099195,
+      "learning_rate": 7.777441982029072e-06,
+      "loss": 0.4115,
+      "step": 1691
+    },
+    {
+      "epoch": 1.1447902571041948,
+      "grad_norm": 1.0510596721044594,
+      "learning_rate": 7.774167549261817e-06,
+      "loss": 0.4145,
+      "step": 1692
+    },
+    {
+      "epoch": 1.145466847090663,
+      "grad_norm": 1.0577858768604804,
+      "learning_rate": 7.770891396660212e-06,
+      "loss": 0.4363,
+      "step": 1693
+    },
+    {
+      "epoch": 1.1461434370771313,
+      "grad_norm": 1.1009706681569988,
+      "learning_rate": 7.767613526255296e-06,
+      "loss": 0.4451,
+      "step": 1694
+    },
+    {
+      "epoch": 1.1468200270635995,
+      "grad_norm": 1.0041036698381118,
+      "learning_rate": 7.764333940079169e-06,
+      "loss": 0.4029,
+      "step": 1695
+    },
+    {
+      "epoch": 1.1474966170500678,
+      "grad_norm": 1.0887492064401576,
+      "learning_rate": 7.761052640165e-06,
+      "loss": 0.4282,
+      "step": 1696
+    },
+    {
+      "epoch": 1.1481732070365358,
+      "grad_norm": 1.1042918742864336,
+      "learning_rate": 7.757769628547018e-06,
+      "loss": 0.4415,
+      "step": 1697
+    },
+    {
+      "epoch": 1.148849797023004,
+      "grad_norm": 1.0795478153043707,
+      "learning_rate": 7.754484907260513e-06,
+      "loss": 0.4363,
+      "step": 1698
+    },
+    {
+      "epoch": 1.1495263870094723,
+      "grad_norm": 1.1211007669102866,
+      "learning_rate": 7.751198478341836e-06,
+      "loss": 0.4494,
+      "step": 1699
+    },
+    {
+      "epoch": 1.1502029769959405,
+      "grad_norm": 1.0461161449887515,
+      "learning_rate": 7.747910343828391e-06,
+      "loss": 0.4185,
+      "step": 1700
+    },
+    {
+      "epoch": 1.1508795669824086,
+      "grad_norm": 1.0250573855164917,
+      "learning_rate": 7.744620505758652e-06,
+      "loss": 0.4149,
+      "step": 1701
+    },
+    {
+      "epoch": 1.1515561569688768,
+      "grad_norm": 1.1020192972492517,
+      "learning_rate": 7.741328966172134e-06,
+      "loss": 0.4289,
+      "step": 1702
+    },
+    {
+      "epoch": 1.152232746955345,
+      "grad_norm": 1.078114451975737,
+      "learning_rate": 7.738035727109418e-06,
+      "loss": 0.4157,
+      "step": 1703
+    },
+    {
+      "epoch": 1.1529093369418133,
+      "grad_norm": 1.0370301746075743,
+      "learning_rate": 7.734740790612137e-06,
+      "loss": 0.4434,
+      "step": 1704
+    },
+    {
+      "epoch": 1.1535859269282815,
+      "grad_norm": 1.0787767586201404,
+      "learning_rate": 7.731444158722967e-06,
+      "loss": 0.4315,
+      "step": 1705
+    },
+    {
+      "epoch": 1.1542625169147496,
+      "grad_norm": 1.0932127316445852,
+      "learning_rate": 7.728145833485647e-06,
+      "loss": 0.4349,
+      "step": 1706
+    },
+    {
+      "epoch": 1.1549391069012178,
+      "grad_norm": 1.035041805932224,
+      "learning_rate": 7.724845816944962e-06,
+      "loss": 0.438,
+      "step": 1707
+    },
+    {
+      "epoch": 1.155615696887686,
+      "grad_norm": 1.0324469297545122,
+      "learning_rate": 7.72154411114674e-06,
+      "loss": 0.4159,
+      "step": 1708
+    },
+    {
+      "epoch": 1.1562922868741543,
+      "grad_norm": 1.0235591812233056,
+      "learning_rate": 7.718240718137863e-06,
+      "loss": 0.3985,
+      "step": 1709
+    },
+    {
+      "epoch": 1.1569688768606226,
+      "grad_norm": 1.109226175562451,
+      "learning_rate": 7.714935639966257e-06,
+      "loss": 0.4203,
+      "step": 1710
+    },
+    {
+      "epoch": 1.1576454668470906,
+      "grad_norm": 1.1047819932685186,
+      "learning_rate": 7.711628878680892e-06,
+      "loss": 0.4381,
+      "step": 1711
+    },
+    {
+      "epoch": 1.1583220568335588,
+      "grad_norm": 1.0899695902742073,
+      "learning_rate": 7.708320436331782e-06,
+      "loss": 0.4391,
+      "step": 1712
+    },
+    {
+      "epoch": 1.158998646820027,
+      "grad_norm": 1.1171269964445691,
+      "learning_rate": 7.705010314969983e-06,
+      "loss": 0.449,
+      "step": 1713
+    },
+    {
+      "epoch": 1.1596752368064953,
+      "grad_norm": 1.134971538790631,
+      "learning_rate": 7.70169851664759e-06,
+      "loss": 0.4369,
+      "step": 1714
+    },
+    {
+      "epoch": 1.1603518267929636,
+      "grad_norm": 1.0516597377957497,
+      "learning_rate": 7.698385043417741e-06,
+      "loss": 0.4009,
+      "step": 1715
+    },
+    {
+      "epoch": 1.1610284167794316,
+      "grad_norm": 1.0468780025418476,
+      "learning_rate": 7.695069897334613e-06,
+      "loss": 0.4137,
+      "step": 1716
+    },
+    {
+      "epoch": 1.1617050067658998,
+      "grad_norm": 1.1094964357843518,
+      "learning_rate": 7.691753080453413e-06,
+      "loss": 0.4387,
+      "step": 1717
+    },
+    {
+      "epoch": 1.162381596752368,
+      "grad_norm": 1.1545923437236598,
+      "learning_rate": 7.688434594830392e-06,
+      "loss": 0.4404,
+      "step": 1718
+    },
+    {
+      "epoch": 1.1630581867388363,
+      "grad_norm": 1.0608018567651694,
+      "learning_rate": 7.685114442522831e-06,
+      "loss": 0.416,
+      "step": 1719
+    },
+    {
+      "epoch": 1.1637347767253043,
+      "grad_norm": 1.0968318468852076,
+      "learning_rate": 7.681792625589046e-06,
+      "loss": 0.4431,
+      "step": 1720
+    },
+    {
+      "epoch": 1.1644113667117726,
+      "grad_norm": 1.0488681736636545,
+      "learning_rate": 7.678469146088385e-06,
+      "loss": 0.4313,
+      "step": 1721
+    },
+    {
+      "epoch": 1.1650879566982408,
+      "grad_norm": 1.0400668779503623,
+      "learning_rate": 7.675144006081225e-06,
+      "loss": 0.4082,
+      "step": 1722
+    },
+    {
+      "epoch": 1.165764546684709,
+      "grad_norm": 1.0632258969178827,
+      "learning_rate": 7.671817207628973e-06,
+      "loss": 0.4234,
+      "step": 1723
+    },
+    {
+      "epoch": 1.1664411366711773,
+      "grad_norm": 1.030818486913185,
+      "learning_rate": 7.668488752794067e-06,
+      "loss": 0.4307,
+      "step": 1724
+    },
+    {
+      "epoch": 1.1671177266576453,
+      "grad_norm": 1.098843164214365,
+      "learning_rate": 7.66515864363997e-06,
+      "loss": 0.4415,
+      "step": 1725
+    },
+    {
+      "epoch": 1.1677943166441136,
+      "grad_norm": 1.0680730548706054,
+      "learning_rate": 7.661826882231165e-06,
+      "loss": 0.4418,
+      "step": 1726
+    },
+    {
+      "epoch": 1.1684709066305818,
+      "grad_norm": 1.0834139813120975,
+      "learning_rate": 7.658493470633173e-06,
+      "loss": 0.4426,
+      "step": 1727
+    },
+    {
+      "epoch": 1.16914749661705,
+      "grad_norm": 1.122949182521549,
+      "learning_rate": 7.65515841091252e-06,
+      "loss": 0.4591,
+      "step": 1728
+    },
+    {
+      "epoch": 1.1698240866035183,
+      "grad_norm": 1.0777417035253167,
+      "learning_rate": 7.651821705136771e-06,
+      "loss": 0.4362,
+      "step": 1729
+    },
+    {
+      "epoch": 1.1705006765899864,
+      "grad_norm": 1.0344411056612368,
+      "learning_rate": 7.648483355374496e-06,
+      "loss": 0.4196,
+      "step": 1730
+    },
+    {
+      "epoch": 1.1711772665764546,
+      "grad_norm": 1.0400708813650934,
+      "learning_rate": 7.645143363695302e-06,
+      "loss": 0.4296,
+      "step": 1731
+    },
+    {
+      "epoch": 1.1718538565629228,
+      "grad_norm": 1.0462815157853997,
+      "learning_rate": 7.641801732169796e-06,
+      "loss": 0.43,
+      "step": 1732
+    },
+    {
+      "epoch": 1.172530446549391,
+      "grad_norm": 1.082777131755936,
+      "learning_rate": 7.63845846286961e-06,
+      "loss": 0.4317,
+      "step": 1733
+    },
+    {
+      "epoch": 1.1732070365358593,
+      "grad_norm": 1.090354407656327,
+      "learning_rate": 7.635113557867395e-06,
+      "loss": 0.4321,
+      "step": 1734
+    },
+    {
+      "epoch": 1.1738836265223274,
+      "grad_norm": 1.0788999438827593,
+      "learning_rate": 7.63176701923681e-06,
+      "loss": 0.4276,
+      "step": 1735
+    },
+    {
+      "epoch": 1.1745602165087956,
+      "grad_norm": 1.0793553420262387,
+      "learning_rate": 7.628418849052523e-06,
+      "loss": 0.4232,
+      "step": 1736
+    },
+    {
+      "epoch": 1.1752368064952639,
+      "grad_norm": 1.057948289081961,
+      "learning_rate": 7.625069049390228e-06,
+      "loss": 0.4259,
+      "step": 1737
+    },
+    {
+      "epoch": 1.175913396481732,
+      "grad_norm": 1.0683544191354293,
+      "learning_rate": 7.621717622326617e-06,
+      "loss": 0.4332,
+      "step": 1738
+    },
+    {
+      "epoch": 1.1765899864682003,
+      "grad_norm": 1.11566521100598,
+      "learning_rate": 7.61836456993939e-06,
+      "loss": 0.4405,
+      "step": 1739
+    },
+    {
+      "epoch": 1.1772665764546684,
+      "grad_norm": 1.0636654982281053,
+      "learning_rate": 7.615009894307263e-06,
+      "loss": 0.426,
+      "step": 1740
+    },
+    {
+      "epoch": 1.1779431664411366,
+      "grad_norm": 1.0773563958574097,
+      "learning_rate": 7.611653597509954e-06,
+      "loss": 0.4278,
+      "step": 1741
+    },
+    {
+      "epoch": 1.1786197564276049,
+      "grad_norm": 1.0135105759166276,
+      "learning_rate": 7.608295681628185e-06,
+      "loss": 0.4164,
+      "step": 1742
+    },
+    {
+      "epoch": 1.179296346414073,
+      "grad_norm": 1.0820504998658946,
+      "learning_rate": 7.604936148743682e-06,
+      "loss": 0.4202,
+      "step": 1743
+    },
+    {
+      "epoch": 1.1799729364005414,
+      "grad_norm": 1.1241255054691106,
+      "learning_rate": 7.6015750009391776e-06,
+      "loss": 0.4572,
+      "step": 1744
+    },
+    {
+      "epoch": 1.1806495263870094,
+      "grad_norm": 1.0775197137939894,
+      "learning_rate": 7.5982122402983986e-06,
+      "loss": 0.4487,
+      "step": 1745
+    },
+    {
+      "epoch": 1.1813261163734776,
+      "grad_norm": 1.0262329998702866,
+      "learning_rate": 7.594847868906076e-06,
+      "loss": 0.4138,
+      "step": 1746
+    },
+    {
+      "epoch": 1.1820027063599459,
+      "grad_norm": 1.0803460081167837,
+      "learning_rate": 7.5914818888479406e-06,
+      "loss": 0.4315,
+      "step": 1747
+    },
+    {
+      "epoch": 1.182679296346414,
+      "grad_norm": 1.0771088692402047,
+      "learning_rate": 7.588114302210719e-06,
+      "loss": 0.4364,
+      "step": 1748
+    },
+    {
+      "epoch": 1.1833558863328824,
+      "grad_norm": 1.0496325532095967,
+      "learning_rate": 7.584745111082128e-06,
+      "loss": 0.4338,
+      "step": 1749
+    },
+    {
+      "epoch": 1.1840324763193504,
+      "grad_norm": 1.0852773892249585,
+      "learning_rate": 7.5813743175508914e-06,
+      "loss": 0.4351,
+      "step": 1750
+    },
+    {
+      "epoch": 1.1847090663058186,
+      "grad_norm": 1.0352845427311212,
+      "learning_rate": 7.578001923706715e-06,
+      "loss": 0.4362,
+      "step": 1751
+    },
+    {
+      "epoch": 1.1853856562922869,
+      "grad_norm": 1.0373837306819627,
+      "learning_rate": 7.574627931640304e-06,
+      "loss": 0.4294,
+      "step": 1752
+    },
+    {
+      "epoch": 1.1860622462787551,
+      "grad_norm": 1.0572118401090786,
+      "learning_rate": 7.571252343443349e-06,
+      "loss": 0.4432,
+      "step": 1753
+    },
+    {
+      "epoch": 1.1867388362652234,
+      "grad_norm": 1.0452270042217684,
+      "learning_rate": 7.5678751612085344e-06,
+      "loss": 0.422,
+      "step": 1754
+    },
+    {
+      "epoch": 1.1874154262516914,
+      "grad_norm": 1.0276305064576137,
+      "learning_rate": 7.564496387029532e-06,
+      "loss": 0.4243,
+      "step": 1755
+    },
+    {
+      "epoch": 1.1880920162381596,
+      "grad_norm": 1.0927408405400587,
+      "learning_rate": 7.5611160230009975e-06,
+      "loss": 0.4397,
+      "step": 1756
+    },
+    {
+      "epoch": 1.1887686062246279,
+      "grad_norm": 1.0048031575387393,
+      "learning_rate": 7.557734071218575e-06,
+      "loss": 0.3995,
+      "step": 1757
+    },
+    {
+      "epoch": 1.1894451962110961,
+      "grad_norm": 1.0367309076105888,
+      "learning_rate": 7.5543505337788934e-06,
+      "loss": 0.4099,
+      "step": 1758
+    },
+    {
+      "epoch": 1.1901217861975644,
+      "grad_norm": 1.055713401251462,
+      "learning_rate": 7.550965412779563e-06,
+      "loss": 0.4281,
+      "step": 1759
+    },
+    {
+      "epoch": 1.1907983761840324,
+      "grad_norm": 1.0926686257988263,
+      "learning_rate": 7.547578710319174e-06,
+      "loss": 0.4178,
+      "step": 1760
+    },
+    {
+      "epoch": 1.1914749661705006,
+      "grad_norm": 1.0797406046782434,
+      "learning_rate": 7.544190428497304e-06,
+      "loss": 0.4279,
+      "step": 1761
+    },
+    {
+      "epoch": 1.1921515561569689,
+      "grad_norm": 1.140108523007309,
+      "learning_rate": 7.540800569414501e-06,
+      "loss": 0.4614,
+      "step": 1762
+    },
+    {
+      "epoch": 1.1928281461434371,
+      "grad_norm": 1.049646539641545,
+      "learning_rate": 7.537409135172298e-06,
+      "loss": 0.4231,
+      "step": 1763
+    },
+    {
+      "epoch": 1.1935047361299054,
+      "grad_norm": 1.0769851264218875,
+      "learning_rate": 7.5340161278732e-06,
+      "loss": 0.4276,
+      "step": 1764
+    },
+    {
+      "epoch": 1.1941813261163734,
+      "grad_norm": 1.001464436924969,
+      "learning_rate": 7.530621549620689e-06,
+      "loss": 0.3953,
+      "step": 1765
+    },
+    {
+      "epoch": 1.1948579161028416,
+      "grad_norm": 1.097639622865649,
+      "learning_rate": 7.527225402519218e-06,
+      "loss": 0.4405,
+      "step": 1766
+    },
+    {
+      "epoch": 1.19553450608931,
+      "grad_norm": 1.0327683397057092,
+      "learning_rate": 7.52382768867422e-06,
+      "loss": 0.4259,
+      "step": 1767
+    },
+    {
+      "epoch": 1.1962110960757781,
+      "grad_norm": 1.127561600104448,
+      "learning_rate": 7.52042841019209e-06,
+      "loss": 0.4506,
+      "step": 1768
+    },
+    {
+      "epoch": 1.1968876860622464,
+      "grad_norm": 1.0511337706371238,
+      "learning_rate": 7.5170275691802e-06,
+      "loss": 0.4239,
+      "step": 1769
+    },
+    {
+      "epoch": 1.1975642760487144,
+      "grad_norm": 1.0096823700913338,
+      "learning_rate": 7.5136251677468856e-06,
+      "loss": 0.417,
+      "step": 1770
+    },
+    {
+      "epoch": 1.1982408660351827,
+      "grad_norm": 1.0442975619891155,
+      "learning_rate": 7.510221208001457e-06,
+      "loss": 0.4328,
+      "step": 1771
+    },
+    {
+      "epoch": 1.198917456021651,
+      "grad_norm": 1.056310375902745,
+      "learning_rate": 7.50681569205418e-06,
+      "loss": 0.4343,
+      "step": 1772
+    },
+    {
+      "epoch": 1.1995940460081191,
+      "grad_norm": 1.0485435366935498,
+      "learning_rate": 7.5034086220162945e-06,
+      "loss": 0.4219,
+      "step": 1773
+    },
+    {
+      "epoch": 1.2002706359945874,
+      "grad_norm": 1.1091019027460352,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.4204,
+      "step": 1774
+    },
+    {
+      "epoch": 1.2009472259810554,
+      "grad_norm": 1.1208881780578195,
+      "learning_rate": 7.496589828118458e-06,
+      "loss": 0.4501,
+      "step": 1775
+    },
+    {
+      "epoch": 1.2016238159675237,
+      "grad_norm": 1.036187467489681,
+      "learning_rate": 7.4931781084857915e-06,
+      "loss": 0.4152,
+      "step": 1776
+    },
+    {
+      "epoch": 1.202300405953992,
+      "grad_norm": 1.0362678918135637,
+      "learning_rate": 7.489764843217082e-06,
+      "loss": 0.4356,
+      "step": 1777
+    },
+    {
+      "epoch": 1.2029769959404601,
+      "grad_norm": 1.0492880981253974,
+      "learning_rate": 7.4863500344283715e-06,
+      "loss": 0.419,
+      "step": 1778
+    },
+    {
+      "epoch": 1.2036535859269284,
+      "grad_norm": 1.0804969954100565,
+      "learning_rate": 7.482933684236654e-06,
+      "loss": 0.4426,
+      "step": 1779
+    },
+    {
+      "epoch": 1.2043301759133964,
+      "grad_norm": 1.0420846359657001,
+      "learning_rate": 7.4795157947598864e-06,
+      "loss": 0.4394,
+      "step": 1780
+    },
+    {
+      "epoch": 1.2050067658998647,
+      "grad_norm": 1.0496467723396619,
+      "learning_rate": 7.476096368116974e-06,
+      "loss": 0.4212,
+      "step": 1781
+    },
+    {
+      "epoch": 1.205683355886333,
+      "grad_norm": 1.073609812824759,
+      "learning_rate": 7.4726754064277775e-06,
+      "loss": 0.4211,
+      "step": 1782
+    },
+    {
+      "epoch": 1.2063599458728012,
+      "grad_norm": 1.147810099634884,
+      "learning_rate": 7.469252911813107e-06,
+      "loss": 0.4456,
+      "step": 1783
+    },
+    {
+      "epoch": 1.2070365358592694,
+      "grad_norm": 1.0339857863508586,
+      "learning_rate": 7.465828886394729e-06,
+      "loss": 0.4146,
+      "step": 1784
+    },
+    {
+      "epoch": 1.2077131258457374,
+      "grad_norm": 1.05538374872532,
+      "learning_rate": 7.462403332295351e-06,
+      "loss": 0.4249,
+      "step": 1785
+    },
+    {
+      "epoch": 1.2083897158322057,
+      "grad_norm": 1.0270527698895682,
+      "learning_rate": 7.458976251638632e-06,
+      "loss": 0.4046,
+      "step": 1786
+    },
+    {
+      "epoch": 1.209066305818674,
+      "grad_norm": 1.0359624154039528,
+      "learning_rate": 7.455547646549179e-06,
+      "loss": 0.4247,
+      "step": 1787
+    },
+    {
+      "epoch": 1.2097428958051422,
+      "grad_norm": 1.0643263297091912,
+      "learning_rate": 7.452117519152542e-06,
+      "loss": 0.4119,
+      "step": 1788
+    },
+    {
+      "epoch": 1.2104194857916104,
+      "grad_norm": 1.0496257822970312,
+      "learning_rate": 7.448685871575213e-06,
+      "loss": 0.4312,
+      "step": 1789
+    },
+    {
+      "epoch": 1.2110960757780784,
+      "grad_norm": 1.0405282605598096,
+      "learning_rate": 7.445252705944632e-06,
+      "loss": 0.4171,
+      "step": 1790
+    },
+    {
+      "epoch": 1.2117726657645467,
+      "grad_norm": 1.0884311545690581,
+      "learning_rate": 7.441818024389173e-06,
+      "loss": 0.4194,
+      "step": 1791
+    },
+    {
+      "epoch": 1.212449255751015,
+      "grad_norm": 1.0476343136961743,
+      "learning_rate": 7.438381829038157e-06,
+      "loss": 0.4167,
+      "step": 1792
+    },
+    {
+      "epoch": 1.2131258457374832,
+      "grad_norm": 1.0717160931695264,
+      "learning_rate": 7.434944122021837e-06,
+      "loss": 0.4238,
+      "step": 1793
+    },
+    {
+      "epoch": 1.2138024357239512,
+      "grad_norm": 1.1084054339649736,
+      "learning_rate": 7.431504905471407e-06,
+      "loss": 0.4451,
+      "step": 1794
+    },
+    {
+      "epoch": 1.2144790257104194,
+      "grad_norm": 1.0729490077307373,
+      "learning_rate": 7.428064181518997e-06,
+      "loss": 0.4462,
+      "step": 1795
+    },
+    {
+      "epoch": 1.2151556156968877,
+      "grad_norm": 0.9860342582386903,
+      "learning_rate": 7.424621952297668e-06,
+      "loss": 0.4068,
+      "step": 1796
+    },
+    {
+      "epoch": 1.215832205683356,
+      "grad_norm": 1.0684842057776966,
+      "learning_rate": 7.4211782199414204e-06,
+      "loss": 0.4214,
+      "step": 1797
+    },
+    {
+      "epoch": 1.2165087956698242,
+      "grad_norm": 1.0366597281113505,
+      "learning_rate": 7.417732986585179e-06,
+      "loss": 0.4157,
+      "step": 1798
+    },
+    {
+      "epoch": 1.2171853856562922,
+      "grad_norm": 1.0329970051374666,
+      "learning_rate": 7.414286254364804e-06,
+      "loss": 0.4174,
+      "step": 1799
+    },
+    {
+      "epoch": 1.2178619756427604,
+      "grad_norm": 1.0586898711198538,
+      "learning_rate": 7.410838025417083e-06,
+      "loss": 0.4304,
+      "step": 1800
+    },
+    {
+      "epoch": 1.2185385656292287,
+      "grad_norm": 1.0280475621624243,
+      "learning_rate": 7.407388301879735e-06,
+      "loss": 0.4366,
+      "step": 1801
+    },
+    {
+      "epoch": 1.219215155615697,
+      "grad_norm": 1.0674000428420767,
+      "learning_rate": 7.403937085891397e-06,
+      "loss": 0.4259,
+      "step": 1802
+    },
+    {
+      "epoch": 1.2198917456021652,
+      "grad_norm": 1.0588603505389673,
+      "learning_rate": 7.400484379591644e-06,
+      "loss": 0.4378,
+      "step": 1803
+    },
+    {
+      "epoch": 1.2205683355886332,
+      "grad_norm": 1.0690086552832514,
+      "learning_rate": 7.397030185120962e-06,
+      "loss": 0.4307,
+      "step": 1804
+    },
+    {
+      "epoch": 1.2212449255751014,
+      "grad_norm": 1.0446017776878338,
+      "learning_rate": 7.393574504620767e-06,
+      "loss": 0.4356,
+      "step": 1805
+    },
+    {
+      "epoch": 1.2219215155615697,
+      "grad_norm": 1.0879163576888138,
+      "learning_rate": 7.390117340233396e-06,
+      "loss": 0.4295,
+      "step": 1806
+    },
+    {
+      "epoch": 1.222598105548038,
+      "grad_norm": 1.064772020945927,
+      "learning_rate": 7.386658694102103e-06,
+      "loss": 0.4466,
+      "step": 1807
+    },
+    {
+      "epoch": 1.2232746955345062,
+      "grad_norm": 1.0774821775397263,
+      "learning_rate": 7.383198568371064e-06,
+      "loss": 0.439,
+      "step": 1808
+    },
+    {
+      "epoch": 1.2239512855209742,
+      "grad_norm": 1.0307840381048303,
+      "learning_rate": 7.379736965185369e-06,
+      "loss": 0.4259,
+      "step": 1809
+    },
+    {
+      "epoch": 1.2246278755074425,
+      "grad_norm": 1.0893265648975483,
+      "learning_rate": 7.376273886691024e-06,
+      "loss": 0.4326,
+      "step": 1810
+    },
+    {
+      "epoch": 1.2253044654939107,
+      "grad_norm": 1.0586550592993633,
+      "learning_rate": 7.372809335034955e-06,
+      "loss": 0.4247,
+      "step": 1811
+    },
+    {
+      "epoch": 1.225981055480379,
+      "grad_norm": 0.9868054895328412,
+      "learning_rate": 7.369343312364994e-06,
+      "loss": 0.4075,
+      "step": 1812
+    },
+    {
+      "epoch": 1.226657645466847,
+      "grad_norm": 1.0251115832629527,
+      "learning_rate": 7.365875820829889e-06,
+      "loss": 0.4123,
+      "step": 1813
+    },
+    {
+      "epoch": 1.2273342354533152,
+      "grad_norm": 1.0494949178161048,
+      "learning_rate": 7.362406862579299e-06,
+      "loss": 0.4309,
+      "step": 1814
+    },
+    {
+      "epoch": 1.2280108254397835,
+      "grad_norm": 1.0830705838191654,
+      "learning_rate": 7.358936439763789e-06,
+      "loss": 0.4304,
+      "step": 1815
+    },
+    {
+      "epoch": 1.2286874154262517,
+      "grad_norm": 1.0482339849082631,
+      "learning_rate": 7.355464554534837e-06,
+      "loss": 0.4159,
+      "step": 1816
+    },
+    {
+      "epoch": 1.22936400541272,
+      "grad_norm": 1.0525305578933541,
+      "learning_rate": 7.351991209044822e-06,
+      "loss": 0.4203,
+      "step": 1817
+    },
+    {
+      "epoch": 1.230040595399188,
+      "grad_norm": 1.1684816170354542,
+      "learning_rate": 7.348516405447031e-06,
+      "loss": 0.4507,
+      "step": 1818
+    },
+    {
+      "epoch": 1.2307171853856562,
+      "grad_norm": 1.0527470120746363,
+      "learning_rate": 7.345040145895656e-06,
+      "loss": 0.4132,
+      "step": 1819
+    },
+    {
+      "epoch": 1.2313937753721245,
+      "grad_norm": 1.011165560475297,
+      "learning_rate": 7.341562432545793e-06,
+      "loss": 0.4105,
+      "step": 1820
+    },
+    {
+      "epoch": 1.2320703653585927,
+      "grad_norm": 1.0646474186484702,
+      "learning_rate": 7.338083267553433e-06,
+      "loss": 0.4406,
+      "step": 1821
+    },
+    {
+      "epoch": 1.232746955345061,
+      "grad_norm": 1.0489389597724974,
+      "learning_rate": 7.334602653075471e-06,
+      "loss": 0.4213,
+      "step": 1822
+    },
+    {
+      "epoch": 1.233423545331529,
+      "grad_norm": 1.0561877797503911,
+      "learning_rate": 7.331120591269701e-06,
+      "loss": 0.4302,
+      "step": 1823
+    },
+    {
+      "epoch": 1.2341001353179972,
+      "grad_norm": 1.0340380529270852,
+      "learning_rate": 7.327637084294818e-06,
+      "loss": 0.4167,
+      "step": 1824
+    },
+    {
+      "epoch": 1.2347767253044655,
+      "grad_norm": 1.006360752229811,
+      "learning_rate": 7.324152134310401e-06,
+      "loss": 0.4014,
+      "step": 1825
+    },
+    {
+      "epoch": 1.2354533152909337,
+      "grad_norm": 1.0498725723777145,
+      "learning_rate": 7.3206657434769354e-06,
+      "loss": 0.4218,
+      "step": 1826
+    },
+    {
+      "epoch": 1.236129905277402,
+      "grad_norm": 1.1197102651242337,
+      "learning_rate": 7.317177913955795e-06,
+      "loss": 0.4561,
+      "step": 1827
+    },
+    {
+      "epoch": 1.23680649526387,
+      "grad_norm": 1.047845814323881,
+      "learning_rate": 7.313688647909245e-06,
+      "loss": 0.4319,
+      "step": 1828
+    },
+    {
+      "epoch": 1.2374830852503382,
+      "grad_norm": 1.0133934913396878,
+      "learning_rate": 7.310197947500446e-06,
+      "loss": 0.3936,
+      "step": 1829
+    },
+    {
+      "epoch": 1.2381596752368065,
+      "grad_norm": 1.0610646071502368,
+      "learning_rate": 7.30670581489344e-06,
+      "loss": 0.4312,
+      "step": 1830
+    },
+    {
+      "epoch": 1.2388362652232747,
+      "grad_norm": 1.0550038089887133,
+      "learning_rate": 7.303212252253163e-06,
+      "loss": 0.4241,
+      "step": 1831
+    },
+    {
+      "epoch": 1.239512855209743,
+      "grad_norm": 1.0078787601100263,
+      "learning_rate": 7.2997172617454335e-06,
+      "loss": 0.4201,
+      "step": 1832
+    },
+    {
+      "epoch": 1.240189445196211,
+      "grad_norm": 1.040439043416705,
+      "learning_rate": 7.29622084553696e-06,
+      "loss": 0.4112,
+      "step": 1833
+    },
+    {
+      "epoch": 1.2408660351826792,
+      "grad_norm": 1.0851214055242746,
+      "learning_rate": 7.29272300579533e-06,
+      "loss": 0.434,
+      "step": 1834
+    },
+    {
+      "epoch": 1.2415426251691475,
+      "grad_norm": 1.0205069422957933,
+      "learning_rate": 7.289223744689018e-06,
+      "loss": 0.4144,
+      "step": 1835
+    },
+    {
+      "epoch": 1.2422192151556157,
+      "grad_norm": 1.040925371820256,
+      "learning_rate": 7.285723064387373e-06,
+      "loss": 0.4371,
+      "step": 1836
+    },
+    {
+      "epoch": 1.242895805142084,
+      "grad_norm": 1.0483087504178972,
+      "learning_rate": 7.282220967060634e-06,
+      "loss": 0.4329,
+      "step": 1837
+    },
+    {
+      "epoch": 1.243572395128552,
+      "grad_norm": 1.0723707821472495,
+      "learning_rate": 7.278717454879907e-06,
+      "loss": 0.4402,
+      "step": 1838
+    },
+    {
+      "epoch": 1.2442489851150202,
+      "grad_norm": 1.0259857378494874,
+      "learning_rate": 7.2752125300171835e-06,
+      "loss": 0.4124,
+      "step": 1839
+    },
+    {
+      "epoch": 1.2449255751014885,
+      "grad_norm": 1.056909604175791,
+      "learning_rate": 7.271706194645327e-06,
+      "loss": 0.4258,
+      "step": 1840
+    },
+    {
+      "epoch": 1.2456021650879567,
+      "grad_norm": 1.0202072520898602,
+      "learning_rate": 7.26819845093808e-06,
+      "loss": 0.4073,
+      "step": 1841
+    },
+    {
+      "epoch": 1.246278755074425,
+      "grad_norm": 1.033461438488984,
+      "learning_rate": 7.264689301070048e-06,
+      "loss": 0.4291,
+      "step": 1842
+    },
+    {
+      "epoch": 1.246955345060893,
+      "grad_norm": 1.0372665916116917,
+      "learning_rate": 7.2611787472167194e-06,
+      "loss": 0.4124,
+      "step": 1843
+    },
+    {
+      "epoch": 1.2476319350473613,
+      "grad_norm": 1.0272576712115402,
+      "learning_rate": 7.257666791554448e-06,
+      "loss": 0.4129,
+      "step": 1844
+    },
+    {
+      "epoch": 1.2483085250338295,
+      "grad_norm": 1.0549832243545578,
+      "learning_rate": 7.254153436260456e-06,
+      "loss": 0.4024,
+      "step": 1845
+    },
+    {
+      "epoch": 1.2489851150202977,
+      "grad_norm": 1.0410383317623635,
+      "learning_rate": 7.250638683512833e-06,
+      "loss": 0.426,
+      "step": 1846
+    },
+    {
+      "epoch": 1.249661705006766,
+      "grad_norm": 1.0669834282377504,
+      "learning_rate": 7.247122535490539e-06,
+      "loss": 0.4314,
+      "step": 1847
+    },
+    {
+      "epoch": 1.250338294993234,
+      "grad_norm": 1.0692057782691242,
+      "learning_rate": 7.2436049943733955e-06,
+      "loss": 0.431,
+      "step": 1848
+    },
+    {
+      "epoch": 1.2510148849797023,
+      "grad_norm": 1.064160593240285,
+      "learning_rate": 7.240086062342087e-06,
+      "loss": 0.4137,
+      "step": 1849
+    },
+    {
+      "epoch": 1.2516914749661705,
+      "grad_norm": 1.0382964704263993,
+      "learning_rate": 7.236565741578163e-06,
+      "loss": 0.4306,
+      "step": 1850
+    },
+    {
+      "epoch": 1.2523680649526387,
+      "grad_norm": 1.046704426082285,
+      "learning_rate": 7.233044034264034e-06,
+      "loss": 0.4324,
+      "step": 1851
+    },
+    {
+      "epoch": 1.253044654939107,
+      "grad_norm": 1.0918029011354768,
+      "learning_rate": 7.229520942582965e-06,
+      "loss": 0.4241,
+      "step": 1852
+    },
+    {
+      "epoch": 1.253721244925575,
+      "grad_norm": 1.0507343263088116,
+      "learning_rate": 7.2259964687190855e-06,
+      "loss": 0.4131,
+      "step": 1853
+    },
+    {
+      "epoch": 1.2543978349120433,
+      "grad_norm": 1.014695676771801,
+      "learning_rate": 7.22247061485738e-06,
+      "loss": 0.4224,
+      "step": 1854
+    },
+    {
+      "epoch": 1.2550744248985115,
+      "grad_norm": 1.027518620809915,
+      "learning_rate": 7.218943383183684e-06,
+      "loss": 0.4279,
+      "step": 1855
+    },
+    {
+      "epoch": 1.2557510148849798,
+      "grad_norm": 1.052531461411292,
+      "learning_rate": 7.215414775884695e-06,
+      "loss": 0.4331,
+      "step": 1856
+    },
+    {
+      "epoch": 1.256427604871448,
+      "grad_norm": 1.0567183807459564,
+      "learning_rate": 7.211884795147958e-06,
+      "loss": 0.4187,
+      "step": 1857
+    },
+    {
+      "epoch": 1.257104194857916,
+      "grad_norm": 0.9995176408166365,
+      "learning_rate": 7.208353443161871e-06,
+      "loss": 0.4206,
+      "step": 1858
+    },
+    {
+      "epoch": 1.2577807848443843,
+      "grad_norm": 1.0076902801485583,
+      "learning_rate": 7.204820722115681e-06,
+      "loss": 0.4188,
+      "step": 1859
+    },
+    {
+      "epoch": 1.2584573748308525,
+      "grad_norm": 1.0663423805198837,
+      "learning_rate": 7.201286634199484e-06,
+      "loss": 0.4268,
+      "step": 1860
+    },
+    {
+      "epoch": 1.2591339648173208,
+      "grad_norm": 1.0446200390665148,
+      "learning_rate": 7.197751181604228e-06,
+      "loss": 0.4322,
+      "step": 1861
+    },
+    {
+      "epoch": 1.259810554803789,
+      "grad_norm": 1.0921478581993227,
+      "learning_rate": 7.194214366521699e-06,
+      "loss": 0.4276,
+      "step": 1862
+    },
+    {
+      "epoch": 1.260487144790257,
+      "grad_norm": 1.0589696751775382,
+      "learning_rate": 7.190676191144532e-06,
+      "loss": 0.4037,
+      "step": 1863
+    },
+    {
+      "epoch": 1.2611637347767253,
+      "grad_norm": 1.0290308964919637,
+      "learning_rate": 7.187136657666208e-06,
+      "loss": 0.417,
+      "step": 1864
+    },
+    {
+      "epoch": 1.2618403247631935,
+      "grad_norm": 1.0470068438719795,
+      "learning_rate": 7.183595768281044e-06,
+      "loss": 0.4339,
+      "step": 1865
+    },
+    {
+      "epoch": 1.2625169147496618,
+      "grad_norm": 0.98901399837732,
+      "learning_rate": 7.180053525184202e-06,
+      "loss": 0.4198,
+      "step": 1866
+    },
+    {
+      "epoch": 1.26319350473613,
+      "grad_norm": 1.052338787970324,
+      "learning_rate": 7.176509930571682e-06,
+      "loss": 0.4084,
+      "step": 1867
+    },
+    {
+      "epoch": 1.263870094722598,
+      "grad_norm": 1.0619778656699166,
+      "learning_rate": 7.172964986640319e-06,
+      "loss": 0.4376,
+      "step": 1868
+    },
+    {
+      "epoch": 1.2645466847090663,
+      "grad_norm": 0.9921952317723164,
+      "learning_rate": 7.169418695587791e-06,
+      "loss": 0.4077,
+      "step": 1869
+    },
+    {
+      "epoch": 1.2652232746955345,
+      "grad_norm": 1.0602673593791327,
+      "learning_rate": 7.165871059612604e-06,
+      "loss": 0.4224,
+      "step": 1870
+    },
+    {
+      "epoch": 1.2658998646820028,
+      "grad_norm": 1.040367719037345,
+      "learning_rate": 7.162322080914106e-06,
+      "loss": 0.4229,
+      "step": 1871
+    },
+    {
+      "epoch": 1.266576454668471,
+      "grad_norm": 1.0672355169561336,
+      "learning_rate": 7.158771761692464e-06,
+      "loss": 0.4291,
+      "step": 1872
+    },
+    {
+      "epoch": 1.267253044654939,
+      "grad_norm": 1.0416062326184927,
+      "learning_rate": 7.155220104148694e-06,
+      "loss": 0.4217,
+      "step": 1873
+    },
+    {
+      "epoch": 1.2679296346414073,
+      "grad_norm": 1.036889202967758,
+      "learning_rate": 7.151667110484626e-06,
+      "loss": 0.4232,
+      "step": 1874
+    },
+    {
+      "epoch": 1.2686062246278755,
+      "grad_norm": 1.0191265847071933,
+      "learning_rate": 7.148112782902927e-06,
+      "loss": 0.4103,
+      "step": 1875
+    },
+    {
+      "epoch": 1.2692828146143438,
+      "grad_norm": 1.039796595773164,
+      "learning_rate": 7.144557123607087e-06,
+      "loss": 0.4279,
+      "step": 1876
+    },
+    {
+      "epoch": 1.269959404600812,
+      "grad_norm": 1.0191522117178033,
+      "learning_rate": 7.141000134801426e-06,
+      "loss": 0.4181,
+      "step": 1877
+    },
+    {
+      "epoch": 1.27063599458728,
+      "grad_norm": 1.0407018633602043,
+      "learning_rate": 7.137441818691081e-06,
+      "loss": 0.4352,
+      "step": 1878
+    },
+    {
+      "epoch": 1.2713125845737483,
+      "grad_norm": 1.0361978369363967,
+      "learning_rate": 7.133882177482019e-06,
+      "loss": 0.4121,
+      "step": 1879
+    },
+    {
+      "epoch": 1.2719891745602165,
+      "grad_norm": 1.05544765232057,
+      "learning_rate": 7.130321213381025e-06,
+      "loss": 0.4224,
+      "step": 1880
+    },
+    {
+      "epoch": 1.2726657645466848,
+      "grad_norm": 1.0429317514953815,
+      "learning_rate": 7.1267589285957075e-06,
+      "loss": 0.4173,
+      "step": 1881
+    },
+    {
+      "epoch": 1.273342354533153,
+      "grad_norm": 1.081130599728119,
+      "learning_rate": 7.123195325334486e-06,
+      "loss": 0.416,
+      "step": 1882
+    },
+    {
+      "epoch": 1.274018944519621,
+      "grad_norm": 1.0590471273414273,
+      "learning_rate": 7.119630405806607e-06,
+      "loss": 0.4196,
+      "step": 1883
+    },
+    {
+      "epoch": 1.2746955345060893,
+      "grad_norm": 1.027365914525238,
+      "learning_rate": 7.1160641722221255e-06,
+      "loss": 0.4017,
+      "step": 1884
+    },
+    {
+      "epoch": 1.2753721244925575,
+      "grad_norm": 1.0036207664540902,
+      "learning_rate": 7.112496626791915e-06,
+      "loss": 0.4074,
+      "step": 1885
+    },
+    {
+      "epoch": 1.2760487144790258,
+      "grad_norm": 1.102698366148571,
+      "learning_rate": 7.108927771727661e-06,
+      "loss": 0.4271,
+      "step": 1886
+    },
+    {
+      "epoch": 1.276725304465494,
+      "grad_norm": 1.0286285818381042,
+      "learning_rate": 7.105357609241863e-06,
+      "loss": 0.4157,
+      "step": 1887
+    },
+    {
+      "epoch": 1.277401894451962,
+      "grad_norm": 1.0715158955133592,
+      "learning_rate": 7.101786141547829e-06,
+      "loss": 0.4138,
+      "step": 1888
+    },
+    {
+      "epoch": 1.2780784844384303,
+      "grad_norm": 1.0377719757570174,
+      "learning_rate": 7.098213370859673e-06,
+      "loss": 0.4187,
+      "step": 1889
+    },
+    {
+      "epoch": 1.2787550744248986,
+      "grad_norm": 1.0309289124148049,
+      "learning_rate": 7.094639299392324e-06,
+      "loss": 0.4242,
+      "step": 1890
+    },
+    {
+      "epoch": 1.2794316644113666,
+      "grad_norm": 1.0905463430934894,
+      "learning_rate": 7.0910639293615125e-06,
+      "loss": 0.4291,
+      "step": 1891
+    },
+    {
+      "epoch": 1.280108254397835,
+      "grad_norm": 1.061943255649957,
+      "learning_rate": 7.087487262983776e-06,
+      "loss": 0.4369,
+      "step": 1892
+    },
+    {
+      "epoch": 1.280784844384303,
+      "grad_norm": 1.005229810215619,
+      "learning_rate": 7.083909302476453e-06,
+      "loss": 0.4151,
+      "step": 1893
+    },
+    {
+      "epoch": 1.2814614343707713,
+      "grad_norm": 1.0222024289145817,
+      "learning_rate": 7.080330050057687e-06,
+      "loss": 0.4155,
+      "step": 1894
+    },
+    {
+      "epoch": 1.2821380243572396,
+      "grad_norm": 1.0359849066352849,
+      "learning_rate": 7.076749507946422e-06,
+      "loss": 0.4209,
+      "step": 1895
+    },
+    {
+      "epoch": 1.2828146143437076,
+      "grad_norm": 0.9993712113084391,
+      "learning_rate": 7.0731676783624015e-06,
+      "loss": 0.4053,
+      "step": 1896
+    },
+    {
+      "epoch": 1.283491204330176,
+      "grad_norm": 1.0754978857289135,
+      "learning_rate": 7.069584563526166e-06,
+      "loss": 0.4193,
+      "step": 1897
+    },
+    {
+      "epoch": 1.284167794316644,
+      "grad_norm": 1.0682295147426433,
+      "learning_rate": 7.066000165659054e-06,
+      "loss": 0.4218,
+      "step": 1898
+    },
+    {
+      "epoch": 1.2848443843031123,
+      "grad_norm": 1.137024056549916,
+      "learning_rate": 7.062414486983197e-06,
+      "loss": 0.422,
+      "step": 1899
+    },
+    {
+      "epoch": 1.2855209742895806,
+      "grad_norm": 0.9893047815127538,
+      "learning_rate": 7.058827529721526e-06,
+      "loss": 0.407,
+      "step": 1900
+    },
+    {
+      "epoch": 1.2861975642760486,
+      "grad_norm": 1.093759495759805,
+      "learning_rate": 7.055239296097758e-06,
+      "loss": 0.4206,
+      "step": 1901
+    },
+    {
+      "epoch": 1.2868741542625168,
+      "grad_norm": 1.0389243522476432,
+      "learning_rate": 7.051649788336405e-06,
+      "loss": 0.4351,
+      "step": 1902
+    },
+    {
+      "epoch": 1.287550744248985,
+      "grad_norm": 1.031624389267756,
+      "learning_rate": 7.048059008662772e-06,
+      "loss": 0.4292,
+      "step": 1903
+    },
+    {
+      "epoch": 1.2882273342354533,
+      "grad_norm": 1.0127401227745771,
+      "learning_rate": 7.044466959302945e-06,
+      "loss": 0.4075,
+      "step": 1904
+    },
+    {
+      "epoch": 1.2889039242219216,
+      "grad_norm": 1.094989203921133,
+      "learning_rate": 7.040873642483801e-06,
+      "loss": 0.4302,
+      "step": 1905
+    },
+    {
+      "epoch": 1.2895805142083896,
+      "grad_norm": 1.0731525895979097,
+      "learning_rate": 7.037279060433004e-06,
+      "loss": 0.4233,
+      "step": 1906
+    },
+    {
+      "epoch": 1.2902571041948578,
+      "grad_norm": 1.092679228067818,
+      "learning_rate": 7.033683215379002e-06,
+      "loss": 0.4248,
+      "step": 1907
+    },
+    {
+      "epoch": 1.290933694181326,
+      "grad_norm": 1.0633464006107094,
+      "learning_rate": 7.030086109551023e-06,
+      "loss": 0.4207,
+      "step": 1908
+    },
+    {
+      "epoch": 1.2916102841677943,
+      "grad_norm": 1.0655938041658124,
+      "learning_rate": 7.02648774517908e-06,
+      "loss": 0.4223,
+      "step": 1909
+    },
+    {
+      "epoch": 1.2922868741542626,
+      "grad_norm": 1.002809005330905,
+      "learning_rate": 7.022888124493964e-06,
+      "loss": 0.4162,
+      "step": 1910
+    },
+    {
+      "epoch": 1.2929634641407306,
+      "grad_norm": 0.9667935377624202,
+      "learning_rate": 7.019287249727248e-06,
+      "loss": 0.3918,
+      "step": 1911
+    },
+    {
+      "epoch": 1.2936400541271988,
+      "grad_norm": 1.011605643629267,
+      "learning_rate": 7.015685123111276e-06,
+      "loss": 0.4224,
+      "step": 1912
+    },
+    {
+      "epoch": 1.294316644113667,
+      "grad_norm": 1.0589780284119585,
+      "learning_rate": 7.012081746879178e-06,
+      "loss": 0.4371,
+      "step": 1913
+    },
+    {
+      "epoch": 1.2949932341001353,
+      "grad_norm": 1.0241759976541929,
+      "learning_rate": 7.008477123264849e-06,
+      "loss": 0.4193,
+      "step": 1914
+    },
+    {
+      "epoch": 1.2956698240866036,
+      "grad_norm": 1.0397577085592005,
+      "learning_rate": 7.004871254502962e-06,
+      "loss": 0.4232,
+      "step": 1915
+    },
+    {
+      "epoch": 1.2963464140730716,
+      "grad_norm": 1.0363618531551906,
+      "learning_rate": 7.001264142828961e-06,
+      "loss": 0.4176,
+      "step": 1916
+    },
+    {
+      "epoch": 1.2970230040595399,
+      "grad_norm": 1.0314756599160502,
+      "learning_rate": 6.997655790479062e-06,
+      "loss": 0.44,
+      "step": 1917
+    },
+    {
+      "epoch": 1.297699594046008,
+      "grad_norm": 1.3284490053441678,
+      "learning_rate": 6.9940461996902495e-06,
+      "loss": 0.4167,
+      "step": 1918
+    },
+    {
+      "epoch": 1.2983761840324763,
+      "grad_norm": 1.0562678507252015,
+      "learning_rate": 6.990435372700273e-06,
+      "loss": 0.4316,
+      "step": 1919
+    },
+    {
+      "epoch": 1.2990527740189446,
+      "grad_norm": 1.042199906868355,
+      "learning_rate": 6.986823311747652e-06,
+      "loss": 0.4307,
+      "step": 1920
+    },
+    {
+      "epoch": 1.2997293640054126,
+      "grad_norm": 0.9947084159324686,
+      "learning_rate": 6.983210019071671e-06,
+      "loss": 0.4116,
+      "step": 1921
+    },
+    {
+      "epoch": 1.3004059539918809,
+      "grad_norm": 1.0684694784407172,
+      "learning_rate": 6.979595496912374e-06,
+      "loss": 0.4314,
+      "step": 1922
+    },
+    {
+      "epoch": 1.301082543978349,
+      "grad_norm": 1.0900857885218591,
+      "learning_rate": 6.97597974751057e-06,
+      "loss": 0.4322,
+      "step": 1923
+    },
+    {
+      "epoch": 1.3017591339648173,
+      "grad_norm": 1.0342858420227146,
+      "learning_rate": 6.972362773107832e-06,
+      "loss": 0.4217,
+      "step": 1924
+    },
+    {
+      "epoch": 1.3024357239512856,
+      "grad_norm": 1.0308843397983698,
+      "learning_rate": 6.968744575946484e-06,
+      "loss": 0.4242,
+      "step": 1925
+    },
+    {
+      "epoch": 1.3031123139377536,
+      "grad_norm": 1.0443067751322757,
+      "learning_rate": 6.965125158269619e-06,
+      "loss": 0.4258,
+      "step": 1926
+    },
+    {
+      "epoch": 1.3037889039242219,
+      "grad_norm": 1.0145026800131371,
+      "learning_rate": 6.961504522321077e-06,
+      "loss": 0.4162,
+      "step": 1927
+    },
+    {
+      "epoch": 1.30446549391069,
+      "grad_norm": 1.0363141094777273,
+      "learning_rate": 6.957882670345458e-06,
+      "loss": 0.4012,
+      "step": 1928
+    },
+    {
+      "epoch": 1.3051420838971584,
+      "grad_norm": 1.0071780713629512,
+      "learning_rate": 6.954259604588114e-06,
+      "loss": 0.4048,
+      "step": 1929
+    },
+    {
+      "epoch": 1.3058186738836266,
+      "grad_norm": 1.0243348488897355,
+      "learning_rate": 6.950635327295154e-06,
+      "loss": 0.4206,
+      "step": 1930
+    },
+    {
+      "epoch": 1.3064952638700946,
+      "grad_norm": 1.0591372180408505,
+      "learning_rate": 6.94700984071343e-06,
+      "loss": 0.4257,
+      "step": 1931
+    },
+    {
+      "epoch": 1.3071718538565629,
+      "grad_norm": 1.0298745842766885,
+      "learning_rate": 6.943383147090552e-06,
+      "loss": 0.4098,
+      "step": 1932
+    },
+    {
+      "epoch": 1.3078484438430311,
+      "grad_norm": 1.032539916463029,
+      "learning_rate": 6.939755248674872e-06,
+      "loss": 0.4161,
+      "step": 1933
+    },
+    {
+      "epoch": 1.3085250338294994,
+      "grad_norm": 1.025676896237079,
+      "learning_rate": 6.936126147715494e-06,
+      "loss": 0.4305,
+      "step": 1934
+    },
+    {
+      "epoch": 1.3092016238159676,
+      "grad_norm": 1.038434492928235,
+      "learning_rate": 6.932495846462262e-06,
+      "loss": 0.4011,
+      "step": 1935
+    },
+    {
+      "epoch": 1.3098782138024356,
+      "grad_norm": 1.0325297415759267,
+      "learning_rate": 6.928864347165769e-06,
+      "loss": 0.4156,
+      "step": 1936
+    },
+    {
+      "epoch": 1.3105548037889039,
+      "grad_norm": 1.045956006199918,
+      "learning_rate": 6.925231652077349e-06,
+      "loss": 0.4222,
+      "step": 1937
+    },
+    {
+      "epoch": 1.3112313937753721,
+      "grad_norm": 1.00841192315478,
+      "learning_rate": 6.921597763449075e-06,
+      "loss": 0.4145,
+      "step": 1938
+    },
+    {
+      "epoch": 1.3119079837618404,
+      "grad_norm": 1.0257453293856218,
+      "learning_rate": 6.917962683533765e-06,
+      "loss": 0.4056,
+      "step": 1939
+    },
+    {
+      "epoch": 1.3125845737483086,
+      "grad_norm": 1.0965061832624963,
+      "learning_rate": 6.914326414584971e-06,
+      "loss": 0.4286,
+      "step": 1940
+    },
+    {
+      "epoch": 1.3132611637347766,
+      "grad_norm": 1.1030288462358044,
+      "learning_rate": 6.9106889588569845e-06,
+      "loss": 0.4273,
+      "step": 1941
+    },
+    {
+      "epoch": 1.3139377537212449,
+      "grad_norm": 1.0400203867263067,
+      "learning_rate": 6.907050318604831e-06,
+      "loss": 0.436,
+      "step": 1942
+    },
+    {
+      "epoch": 1.3146143437077131,
+      "grad_norm": 1.0485655153728435,
+      "learning_rate": 6.903410496084272e-06,
+      "loss": 0.4172,
+      "step": 1943
+    },
+    {
+      "epoch": 1.3152909336941814,
+      "grad_norm": 1.03732143615595,
+      "learning_rate": 6.8997694935518e-06,
+      "loss": 0.4174,
+      "step": 1944
+    },
+    {
+      "epoch": 1.3159675236806496,
+      "grad_norm": 1.0543855252616179,
+      "learning_rate": 6.896127313264643e-06,
+      "loss": 0.4362,
+      "step": 1945
+    },
+    {
+      "epoch": 1.3166441136671176,
+      "grad_norm": 1.0945303569147031,
+      "learning_rate": 6.892483957480754e-06,
+      "loss": 0.441,
+      "step": 1946
+    },
+    {
+      "epoch": 1.317320703653586,
+      "grad_norm": 1.0308613405126186,
+      "learning_rate": 6.888839428458819e-06,
+      "loss": 0.4259,
+      "step": 1947
+    },
+    {
+      "epoch": 1.3179972936400541,
+      "grad_norm": 1.0053459570484582,
+      "learning_rate": 6.885193728458247e-06,
+      "loss": 0.42,
+      "step": 1948
+    },
+    {
+      "epoch": 1.3186738836265224,
+      "grad_norm": 1.032777221954674,
+      "learning_rate": 6.8815468597391785e-06,
+      "loss": 0.4127,
+      "step": 1949
+    },
+    {
+      "epoch": 1.3193504736129906,
+      "grad_norm": 1.0933368075310617,
+      "learning_rate": 6.877898824562472e-06,
+      "loss": 0.4295,
+      "step": 1950
+    },
+    {
+      "epoch": 1.3200270635994586,
+      "grad_norm": 1.039978254256892,
+      "learning_rate": 6.8742496251897185e-06,
+      "loss": 0.4408,
+      "step": 1951
+    },
+    {
+      "epoch": 1.320703653585927,
+      "grad_norm": 1.0321175085385472,
+      "learning_rate": 6.8705992638832185e-06,
+      "loss": 0.4315,
+      "step": 1952
+    },
+    {
+      "epoch": 1.3213802435723951,
+      "grad_norm": 1.0411565722494915,
+      "learning_rate": 6.8669477429060026e-06,
+      "loss": 0.4295,
+      "step": 1953
+    },
+    {
+      "epoch": 1.3220568335588634,
+      "grad_norm": 1.0913366053147617,
+      "learning_rate": 6.863295064521816e-06,
+      "loss": 0.4329,
+      "step": 1954
+    },
+    {
+      "epoch": 1.3227334235453316,
+      "grad_norm": 0.9443365970442889,
+      "learning_rate": 6.859641230995123e-06,
+      "loss": 0.3914,
+      "step": 1955
+    },
+    {
+      "epoch": 1.3234100135317997,
+      "grad_norm": 1.039597800151262,
+      "learning_rate": 6.855986244591104e-06,
+      "loss": 0.4338,
+      "step": 1956
+    },
+    {
+      "epoch": 1.324086603518268,
+      "grad_norm": 1.0747903357364792,
+      "learning_rate": 6.852330107575653e-06,
+      "loss": 0.4458,
+      "step": 1957
+    },
+    {
+      "epoch": 1.3247631935047361,
+      "grad_norm": 1.004428701653835,
+      "learning_rate": 6.848672822215378e-06,
+      "loss": 0.4238,
+      "step": 1958
+    },
+    {
+      "epoch": 1.3254397834912044,
+      "grad_norm": 1.0512313876126667,
+      "learning_rate": 6.845014390777595e-06,
+      "loss": 0.4378,
+      "step": 1959
+    },
+    {
+      "epoch": 1.3261163734776726,
+      "grad_norm": 1.0011327011312865,
+      "learning_rate": 6.841354815530341e-06,
+      "loss": 0.4019,
+      "step": 1960
+    },
+    {
+      "epoch": 1.3267929634641407,
+      "grad_norm": 0.9870214563568758,
+      "learning_rate": 6.8376940987423526e-06,
+      "loss": 0.4074,
+      "step": 1961
+    },
+    {
+      "epoch": 1.327469553450609,
+      "grad_norm": 1.0191409671026699,
+      "learning_rate": 6.834032242683075e-06,
+      "loss": 0.4155,
+      "step": 1962
+    },
+    {
+      "epoch": 1.3281461434370772,
+      "grad_norm": 1.0503029826899775,
+      "learning_rate": 6.830369249622663e-06,
+      "loss": 0.418,
+      "step": 1963
+    },
+    {
+      "epoch": 1.3288227334235454,
+      "grad_norm": 1.0715770347345146,
+      "learning_rate": 6.8267051218319766e-06,
+      "loss": 0.4395,
+      "step": 1964
+    },
+    {
+      "epoch": 1.3294993234100136,
+      "grad_norm": 1.029799173753757,
+      "learning_rate": 6.823039861582574e-06,
+      "loss": 0.4058,
+      "step": 1965
+    },
+    {
+      "epoch": 1.3301759133964817,
+      "grad_norm": 1.039409884365538,
+      "learning_rate": 6.819373471146722e-06,
+      "loss": 0.4178,
+      "step": 1966
+    },
+    {
+      "epoch": 1.33085250338295,
+      "grad_norm": 1.1011530152736475,
+      "learning_rate": 6.815705952797383e-06,
+      "loss": 0.4235,
+      "step": 1967
+    },
+    {
+      "epoch": 1.3315290933694182,
+      "grad_norm": 1.0272643488408197,
+      "learning_rate": 6.8120373088082215e-06,
+      "loss": 0.4036,
+      "step": 1968
+    },
+    {
+      "epoch": 1.3322056833558864,
+      "grad_norm": 1.0734396966690711,
+      "learning_rate": 6.808367541453599e-06,
+      "loss": 0.4242,
+      "step": 1969
+    },
+    {
+      "epoch": 1.3328822733423547,
+      "grad_norm": 1.0440675115802303,
+      "learning_rate": 6.804696653008574e-06,
+      "loss": 0.4169,
+      "step": 1970
+    },
+    {
+      "epoch": 1.3335588633288227,
+      "grad_norm": 1.0527102213457342,
+      "learning_rate": 6.801024645748899e-06,
+      "loss": 0.419,
+      "step": 1971
+    },
+    {
+      "epoch": 1.334235453315291,
+      "grad_norm": 1.0403169623255601,
+      "learning_rate": 6.797351521951021e-06,
+      "loss": 0.4256,
+      "step": 1972
+    },
+    {
+      "epoch": 1.3349120433017592,
+      "grad_norm": 1.0058598851333849,
+      "learning_rate": 6.793677283892077e-06,
+      "loss": 0.4106,
+      "step": 1973
+    },
+    {
+      "epoch": 1.3355886332882274,
+      "grad_norm": 0.9864499034412528,
+      "learning_rate": 6.7900019338499005e-06,
+      "loss": 0.4105,
+      "step": 1974
+    },
+    {
+      "epoch": 1.3362652232746957,
+      "grad_norm": 1.0851498101661252,
+      "learning_rate": 6.786325474103006e-06,
+      "loss": 0.4278,
+      "step": 1975
+    },
+    {
+      "epoch": 1.3369418132611637,
+      "grad_norm": 1.0427232147668426,
+      "learning_rate": 6.782647906930602e-06,
+      "loss": 0.4123,
+      "step": 1976
+    },
+    {
+      "epoch": 1.337618403247632,
+      "grad_norm": 1.008879779853957,
+      "learning_rate": 6.778969234612583e-06,
+      "loss": 0.3981,
+      "step": 1977
+    },
+    {
+      "epoch": 1.3382949932341002,
+      "grad_norm": 1.0195763958236552,
+      "learning_rate": 6.775289459429526e-06,
+      "loss": 0.4169,
+      "step": 1978
+    },
+    {
+      "epoch": 1.3389715832205684,
+      "grad_norm": 1.030043921102332,
+      "learning_rate": 6.771608583662694e-06,
+      "loss": 0.4169,
+      "step": 1979
+    },
+    {
+      "epoch": 1.3396481732070367,
+      "grad_norm": 1.04060394170562,
+      "learning_rate": 6.767926609594032e-06,
+      "loss": 0.4166,
+      "step": 1980
+    },
+    {
+      "epoch": 1.3403247631935047,
+      "grad_norm": 1.0596763439742476,
+      "learning_rate": 6.764243539506166e-06,
+      "loss": 0.4275,
+      "step": 1981
+    },
+    {
+      "epoch": 1.341001353179973,
+      "grad_norm": 1.0121725599137021,
+      "learning_rate": 6.760559375682398e-06,
+      "loss": 0.427,
+      "step": 1982
+    },
+    {
+      "epoch": 1.3416779431664412,
+      "grad_norm": 1.0288082176213642,
+      "learning_rate": 6.7568741204067145e-06,
+      "loss": 0.4204,
+      "step": 1983
+    },
+    {
+      "epoch": 1.3423545331529092,
+      "grad_norm": 1.0499843130824649,
+      "learning_rate": 6.753187775963773e-06,
+      "loss": 0.4253,
+      "step": 1984
+    },
+    {
+      "epoch": 1.3430311231393777,
+      "grad_norm": 1.0174389797105612,
+      "learning_rate": 6.749500344638908e-06,
+      "loss": 0.4063,
+      "step": 1985
+    },
+    {
+      "epoch": 1.3437077131258457,
+      "grad_norm": 1.0812254539748511,
+      "learning_rate": 6.74581182871813e-06,
+      "loss": 0.451,
+      "step": 1986
+    },
+    {
+      "epoch": 1.344384303112314,
+      "grad_norm": 1.0487451208277823,
+      "learning_rate": 6.7421222304881194e-06,
+      "loss": 0.439,
+      "step": 1987
+    },
+    {
+      "epoch": 1.3450608930987822,
+      "grad_norm": 1.0063093729936436,
+      "learning_rate": 6.738431552236228e-06,
+      "loss": 0.396,
+      "step": 1988
+    },
+    {
+      "epoch": 1.3457374830852502,
+      "grad_norm": 1.0503976905174983,
+      "learning_rate": 6.734739796250477e-06,
+      "loss": 0.4331,
+      "step": 1989
+    },
+    {
+      "epoch": 1.3464140730717187,
+      "grad_norm": 1.097386145909681,
+      "learning_rate": 6.731046964819555e-06,
+      "loss": 0.4431,
+      "step": 1990
+    },
+    {
+      "epoch": 1.3470906630581867,
+      "grad_norm": 1.0106047494897548,
+      "learning_rate": 6.727353060232822e-06,
+      "loss": 0.4049,
+      "step": 1991
+    },
+    {
+      "epoch": 1.347767253044655,
+      "grad_norm": 1.029135946008643,
+      "learning_rate": 6.723658084780297e-06,
+      "loss": 0.4341,
+      "step": 1992
+    },
+    {
+      "epoch": 1.3484438430311232,
+      "grad_norm": 0.9863741857138194,
+      "learning_rate": 6.719962040752665e-06,
+      "loss": 0.4093,
+      "step": 1993
+    },
+    {
+      "epoch": 1.3491204330175912,
+      "grad_norm": 1.044214163248404,
+      "learning_rate": 6.716264930441279e-06,
+      "loss": 0.4127,
+      "step": 1994
+    },
+    {
+      "epoch": 1.3497970230040595,
+      "grad_norm": 0.9859722455064147,
+      "learning_rate": 6.712566756138142e-06,
+      "loss": 0.3925,
+      "step": 1995
+    },
+    {
+      "epoch": 1.3504736129905277,
+      "grad_norm": 1.0171853848379828,
+      "learning_rate": 6.708867520135924e-06,
+      "loss": 0.4066,
+      "step": 1996
+    },
+    {
+      "epoch": 1.351150202976996,
+      "grad_norm": 1.136459883768459,
+      "learning_rate": 6.705167224727956e-06,
+      "loss": 0.439,
+      "step": 1997
+    },
+    {
+      "epoch": 1.3518267929634642,
+      "grad_norm": 1.0696783687374793,
+      "learning_rate": 6.701465872208216e-06,
+      "loss": 0.4246,
+      "step": 1998
+    },
+    {
+      "epoch": 1.3525033829499322,
+      "grad_norm": 1.0319016617433867,
+      "learning_rate": 6.697763464871346e-06,
+      "loss": 0.4214,
+      "step": 1999
+    },
+    {
+      "epoch": 1.3531799729364005,
+      "grad_norm": 1.0524230751022488,
+      "learning_rate": 6.694060005012642e-06,
+      "loss": 0.4408,
+      "step": 2000
+    },
+    {
+      "epoch": 1.3538565629228687,
+      "grad_norm": 1.0122123524630355,
+      "learning_rate": 6.690355494928043e-06,
+      "loss": 0.416,
+      "step": 2001
+    },
+    {
+      "epoch": 1.354533152909337,
+      "grad_norm": 1.0570408088248036,
+      "learning_rate": 6.686649936914151e-06,
+      "loss": 0.422,
+      "step": 2002
+    },
+    {
+      "epoch": 1.3552097428958052,
+      "grad_norm": 0.973013215879346,
+      "learning_rate": 6.682943333268208e-06,
+      "loss": 0.3978,
+      "step": 2003
+    },
+    {
+      "epoch": 1.3558863328822732,
+      "grad_norm": 1.0308636855480167,
+      "learning_rate": 6.6792356862881144e-06,
+      "loss": 0.4413,
+      "step": 2004
+    },
+    {
+      "epoch": 1.3565629228687415,
+      "grad_norm": 1.0410125908595642,
+      "learning_rate": 6.675526998272405e-06,
+      "loss": 0.4312,
+      "step": 2005
+    },
+    {
+      "epoch": 1.3572395128552097,
+      "grad_norm": 1.0297870259383395,
+      "learning_rate": 6.671817271520269e-06,
+      "loss": 0.4096,
+      "step": 2006
+    },
+    {
+      "epoch": 1.357916102841678,
+      "grad_norm": 1.0118324274652348,
+      "learning_rate": 6.668106508331539e-06,
+      "loss": 0.4151,
+      "step": 2007
+    },
+    {
+      "epoch": 1.3585926928281462,
+      "grad_norm": 0.987166650559704,
+      "learning_rate": 6.664394711006684e-06,
+      "loss": 0.4148,
+      "step": 2008
+    },
+    {
+      "epoch": 1.3592692828146142,
+      "grad_norm": 0.9710453550802868,
+      "learning_rate": 6.660681881846822e-06,
+      "loss": 0.3888,
+      "step": 2009
+    },
+    {
+      "epoch": 1.3599458728010825,
+      "grad_norm": 0.9792754179223846,
+      "learning_rate": 6.656968023153706e-06,
+      "loss": 0.4149,
+      "step": 2010
+    },
+    {
+      "epoch": 1.3606224627875507,
+      "grad_norm": 1.0259782184965631,
+      "learning_rate": 6.653253137229727e-06,
+      "loss": 0.3937,
+      "step": 2011
+    },
+    {
+      "epoch": 1.361299052774019,
+      "grad_norm": 1.0024310585481508,
+      "learning_rate": 6.6495372263779145e-06,
+      "loss": 0.4013,
+      "step": 2012
+    },
+    {
+      "epoch": 1.3619756427604872,
+      "grad_norm": 1.0481091780226433,
+      "learning_rate": 6.6458202929019345e-06,
+      "loss": 0.4241,
+      "step": 2013
+    },
+    {
+      "epoch": 1.3626522327469552,
+      "grad_norm": 1.0117337707229543,
+      "learning_rate": 6.6421023391060845e-06,
+      "loss": 0.4019,
+      "step": 2014
+    },
+    {
+      "epoch": 1.3633288227334235,
+      "grad_norm": 1.0499358074064378,
+      "learning_rate": 6.6383833672952945e-06,
+      "loss": 0.4141,
+      "step": 2015
+    },
+    {
+      "epoch": 1.3640054127198917,
+      "grad_norm": 1.0432854536830993,
+      "learning_rate": 6.634663379775126e-06,
+      "loss": 0.4162,
+      "step": 2016
+    },
+    {
+      "epoch": 1.36468200270636,
+      "grad_norm": 1.0471834865487055,
+      "learning_rate": 6.630942378851774e-06,
+      "loss": 0.4214,
+      "step": 2017
+    },
+    {
+      "epoch": 1.3653585926928282,
+      "grad_norm": 1.0215669999280084,
+      "learning_rate": 6.627220366832056e-06,
+      "loss": 0.4304,
+      "step": 2018
+    },
+    {
+      "epoch": 1.3660351826792962,
+      "grad_norm": 1.0546946843995326,
+      "learning_rate": 6.6234973460234184e-06,
+      "loss": 0.4275,
+      "step": 2019
+    },
+    {
+      "epoch": 1.3667117726657645,
+      "grad_norm": 1.011735647399513,
+      "learning_rate": 6.619773318733934e-06,
+      "loss": 0.4216,
+      "step": 2020
+    },
+    {
+      "epoch": 1.3673883626522327,
+      "grad_norm": 0.9877857356075881,
+      "learning_rate": 6.616048287272301e-06,
+      "loss": 0.4086,
+      "step": 2021
+    },
+    {
+      "epoch": 1.368064952638701,
+      "grad_norm": 0.9817834055860715,
+      "learning_rate": 6.612322253947836e-06,
+      "loss": 0.3926,
+      "step": 2022
+    },
+    {
+      "epoch": 1.3687415426251692,
+      "grad_norm": 1.0050196414953598,
+      "learning_rate": 6.608595221070478e-06,
+      "loss": 0.4195,
+      "step": 2023
+    },
+    {
+      "epoch": 1.3694181326116373,
+      "grad_norm": 0.9867981999368625,
+      "learning_rate": 6.60486719095079e-06,
+      "loss": 0.4187,
+      "step": 2024
+    },
+    {
+      "epoch": 1.3700947225981055,
+      "grad_norm": 1.0113170727426537,
+      "learning_rate": 6.601138165899945e-06,
+      "loss": 0.4138,
+      "step": 2025
+    },
+    {
+      "epoch": 1.3707713125845737,
+      "grad_norm": 1.0229877977928143,
+      "learning_rate": 6.597408148229742e-06,
+      "loss": 0.427,
+      "step": 2026
+    },
+    {
+      "epoch": 1.371447902571042,
+      "grad_norm": 1.0297031727093853,
+      "learning_rate": 6.5936771402525875e-06,
+      "loss": 0.4225,
+      "step": 2027
+    },
+    {
+      "epoch": 1.3721244925575102,
+      "grad_norm": 1.0542911396541623,
+      "learning_rate": 6.589945144281508e-06,
+      "loss": 0.4312,
+      "step": 2028
+    },
+    {
+      "epoch": 1.3728010825439783,
+      "grad_norm": 1.0548128227567142,
+      "learning_rate": 6.586212162630137e-06,
+      "loss": 0.4139,
+      "step": 2029
+    },
+    {
+      "epoch": 1.3734776725304465,
+      "grad_norm": 1.040140995830836,
+      "learning_rate": 6.582478197612725e-06,
+      "loss": 0.4271,
+      "step": 2030
+    },
+    {
+      "epoch": 1.3741542625169147,
+      "grad_norm": 1.0320816500314305,
+      "learning_rate": 6.578743251544128e-06,
+      "loss": 0.4332,
+      "step": 2031
+    },
+    {
+      "epoch": 1.374830852503383,
+      "grad_norm": 1.0135099972477304,
+      "learning_rate": 6.57500732673981e-06,
+      "loss": 0.414,
+      "step": 2032
+    },
+    {
+      "epoch": 1.3755074424898512,
+      "grad_norm": 1.0028503153709398,
+      "learning_rate": 6.571270425515843e-06,
+      "loss": 0.4115,
+      "step": 2033
+    },
+    {
+      "epoch": 1.3761840324763193,
+      "grad_norm": 1.0493627329600907,
+      "learning_rate": 6.567532550188908e-06,
+      "loss": 0.4242,
+      "step": 2034
+    },
+    {
+      "epoch": 1.3768606224627875,
+      "grad_norm": 1.0858798690797127,
+      "learning_rate": 6.56379370307628e-06,
+      "loss": 0.4403,
+      "step": 2035
+    },
+    {
+      "epoch": 1.3775372124492558,
+      "grad_norm": 1.0232116076183813,
+      "learning_rate": 6.560053886495847e-06,
+      "loss": 0.3924,
+      "step": 2036
+    },
+    {
+      "epoch": 1.378213802435724,
+      "grad_norm": 0.940722554882363,
+      "learning_rate": 6.556313102766094e-06,
+      "loss": 0.3973,
+      "step": 2037
+    },
+    {
+      "epoch": 1.3788903924221922,
+      "grad_norm": 0.9745813657743075,
+      "learning_rate": 6.552571354206104e-06,
+      "loss": 0.4095,
+      "step": 2038
+    },
+    {
+      "epoch": 1.3795669824086603,
+      "grad_norm": 1.0675000533588468,
+      "learning_rate": 6.548828643135559e-06,
+      "loss": 0.422,
+      "step": 2039
+    },
+    {
+      "epoch": 1.3802435723951285,
+      "grad_norm": 1.0638864029128485,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.4185,
+      "step": 2040
+    },
+    {
+      "epoch": 1.3809201623815968,
+      "grad_norm": 1.060855984837904,
+      "learning_rate": 6.541340342744517e-06,
+      "loss": 0.4373,
+      "step": 2041
+    },
+    {
+      "epoch": 1.381596752368065,
+      "grad_norm": 0.966572894794586,
+      "learning_rate": 6.537594758066362e-06,
+      "loss": 0.3806,
+      "step": 2042
+    },
+    {
+      "epoch": 1.3822733423545333,
+      "grad_norm": 1.0389503133408153,
+      "learning_rate": 6.533848220162336e-06,
+      "loss": 0.4157,
+      "step": 2043
+    },
+    {
+      "epoch": 1.3829499323410013,
+      "grad_norm": 1.0053055566917941,
+      "learning_rate": 6.530100731355089e-06,
+      "loss": 0.4088,
+      "step": 2044
+    },
+    {
+      "epoch": 1.3836265223274695,
+      "grad_norm": 1.0341325677032138,
+      "learning_rate": 6.5263522939678626e-06,
+      "loss": 0.4213,
+      "step": 2045
+    },
+    {
+      "epoch": 1.3843031123139378,
+      "grad_norm": 1.0285726317859492,
+      "learning_rate": 6.5226029103244846e-06,
+      "loss": 0.4221,
+      "step": 2046
+    },
+    {
+      "epoch": 1.384979702300406,
+      "grad_norm": 1.071233932449683,
+      "learning_rate": 6.518852582749373e-06,
+      "loss": 0.4259,
+      "step": 2047
+    },
+    {
+      "epoch": 1.3856562922868743,
+      "grad_norm": 1.0102528084469493,
+      "learning_rate": 6.515101313567529e-06,
+      "loss": 0.4098,
+      "step": 2048
+    },
+    {
+      "epoch": 1.3863328822733423,
+      "grad_norm": 0.9843297015291131,
+      "learning_rate": 6.511349105104534e-06,
+      "loss": 0.4036,
+      "step": 2049
+    },
+    {
+      "epoch": 1.3870094722598105,
+      "grad_norm": 1.0064849296435698,
+      "learning_rate": 6.507595959686558e-06,
+      "loss": 0.4153,
+      "step": 2050
+    },
+    {
+      "epoch": 1.3876860622462788,
+      "grad_norm": 1.0630195299277263,
+      "learning_rate": 6.503841879640349e-06,
+      "loss": 0.4071,
+      "step": 2051
+    },
+    {
+      "epoch": 1.388362652232747,
+      "grad_norm": 1.032856512191563,
+      "learning_rate": 6.500086867293231e-06,
+      "loss": 0.4201,
+      "step": 2052
+    },
+    {
+      "epoch": 1.3890392422192153,
+      "grad_norm": 1.0589845753660267,
+      "learning_rate": 6.496330924973112e-06,
+      "loss": 0.4328,
+      "step": 2053
+    },
+    {
+      "epoch": 1.3897158322056833,
+      "grad_norm": 1.0496781629534784,
+      "learning_rate": 6.492574055008474e-06,
+      "loss": 0.4292,
+      "step": 2054
+    },
+    {
+      "epoch": 1.3903924221921515,
+      "grad_norm": 1.0373155937489198,
+      "learning_rate": 6.488816259728372e-06,
+      "loss": 0.396,
+      "step": 2055
+    },
+    {
+      "epoch": 1.3910690121786198,
+      "grad_norm": 1.0082194646039746,
+      "learning_rate": 6.4850575414624385e-06,
+      "loss": 0.408,
+      "step": 2056
+    },
+    {
+      "epoch": 1.391745602165088,
+      "grad_norm": 1.0112868305447484,
+      "learning_rate": 6.481297902540875e-06,
+      "loss": 0.4075,
+      "step": 2057
+    },
+    {
+      "epoch": 1.3924221921515563,
+      "grad_norm": 0.9777946393058407,
+      "learning_rate": 6.477537345294455e-06,
+      "loss": 0.4149,
+      "step": 2058
+    },
+    {
+      "epoch": 1.3930987821380243,
+      "grad_norm": 1.0429932967518827,
+      "learning_rate": 6.473775872054522e-06,
+      "loss": 0.4277,
+      "step": 2059
+    },
+    {
+      "epoch": 1.3937753721244925,
+      "grad_norm": 1.0203817930711212,
+      "learning_rate": 6.4700134851529864e-06,
+      "loss": 0.4085,
+      "step": 2060
+    },
+    {
+      "epoch": 1.3944519621109608,
+      "grad_norm": 0.9904306968624756,
+      "learning_rate": 6.466250186922325e-06,
+      "loss": 0.4212,
+      "step": 2061
+    },
+    {
+      "epoch": 1.395128552097429,
+      "grad_norm": 1.0431618615213187,
+      "learning_rate": 6.46248597969558e-06,
+      "loss": 0.4472,
+      "step": 2062
+    },
+    {
+      "epoch": 1.3958051420838973,
+      "grad_norm": 1.0176240279001862,
+      "learning_rate": 6.458720865806356e-06,
+      "loss": 0.4092,
+      "step": 2063
+    },
+    {
+      "epoch": 1.3964817320703653,
+      "grad_norm": 1.039550103362902,
+      "learning_rate": 6.454954847588824e-06,
+      "loss": 0.4401,
+      "step": 2064
+    },
+    {
+      "epoch": 1.3971583220568335,
+      "grad_norm": 1.0041541238833982,
+      "learning_rate": 6.4511879273777065e-06,
+      "loss": 0.4055,
+      "step": 2065
+    },
+    {
+      "epoch": 1.3978349120433018,
+      "grad_norm": 1.0145696013341563,
+      "learning_rate": 6.447420107508297e-06,
+      "loss": 0.4194,
+      "step": 2066
+    },
+    {
+      "epoch": 1.39851150202977,
+      "grad_norm": 0.986030628475444,
+      "learning_rate": 6.443651390316438e-06,
+      "loss": 0.4078,
+      "step": 2067
+    },
+    {
+      "epoch": 1.3991880920162383,
+      "grad_norm": 1.0336096121951381,
+      "learning_rate": 6.439881778138531e-06,
+      "loss": 0.4244,
+      "step": 2068
+    },
+    {
+      "epoch": 1.3998646820027063,
+      "grad_norm": 0.9870058474041503,
+      "learning_rate": 6.436111273311533e-06,
+      "loss": 0.406,
+      "step": 2069
+    },
+    {
+      "epoch": 1.4005412719891746,
+      "grad_norm": 1.0670008226310135,
+      "learning_rate": 6.4323398781729525e-06,
+      "loss": 0.4317,
+      "step": 2070
+    },
+    {
+      "epoch": 1.4012178619756428,
+      "grad_norm": 0.9814484086507456,
+      "learning_rate": 6.428567595060853e-06,
+      "loss": 0.3929,
+      "step": 2071
+    },
+    {
+      "epoch": 1.401894451962111,
+      "grad_norm": 1.0117139748524482,
+      "learning_rate": 6.424794426313845e-06,
+      "loss": 0.4199,
+      "step": 2072
+    },
+    {
+      "epoch": 1.4025710419485793,
+      "grad_norm": 1.0161994901612237,
+      "learning_rate": 6.42102037427109e-06,
+      "loss": 0.4111,
+      "step": 2073
+    },
+    {
+      "epoch": 1.4032476319350473,
+      "grad_norm": 1.0304585972871476,
+      "learning_rate": 6.417245441272299e-06,
+      "loss": 0.4104,
+      "step": 2074
+    },
+    {
+      "epoch": 1.4039242219215156,
+      "grad_norm": 1.0314546611378963,
+      "learning_rate": 6.413469629657724e-06,
+      "loss": 0.4224,
+      "step": 2075
+    },
+    {
+      "epoch": 1.4046008119079838,
+      "grad_norm": 1.0168116234374232,
+      "learning_rate": 6.409692941768166e-06,
+      "loss": 0.4173,
+      "step": 2076
+    },
+    {
+      "epoch": 1.4052774018944518,
+      "grad_norm": 1.0033202876518077,
+      "learning_rate": 6.405915379944967e-06,
+      "loss": 0.3835,
+      "step": 2077
+    },
+    {
+      "epoch": 1.4059539918809203,
+      "grad_norm": 1.035970558795831,
+      "learning_rate": 6.402136946530014e-06,
+      "loss": 0.4165,
+      "step": 2078
+    },
+    {
+      "epoch": 1.4066305818673883,
+      "grad_norm": 1.0186604980458462,
+      "learning_rate": 6.398357643865731e-06,
+      "loss": 0.399,
+      "step": 2079
+    },
+    {
+      "epoch": 1.4073071718538566,
+      "grad_norm": 1.0214763885041875,
+      "learning_rate": 6.394577474295081e-06,
+      "loss": 0.4055,
+      "step": 2080
+    },
+    {
+      "epoch": 1.4079837618403248,
+      "grad_norm": 1.021359429184758,
+      "learning_rate": 6.390796440161566e-06,
+      "loss": 0.4162,
+      "step": 2081
+    },
+    {
+      "epoch": 1.4086603518267928,
+      "grad_norm": 1.025956336559153,
+      "learning_rate": 6.387014543809224e-06,
+      "loss": 0.4123,
+      "step": 2082
+    },
+    {
+      "epoch": 1.4093369418132613,
+      "grad_norm": 0.9847594412419571,
+      "learning_rate": 6.383231787582625e-06,
+      "loss": 0.4145,
+      "step": 2083
+    },
+    {
+      "epoch": 1.4100135317997293,
+      "grad_norm": 1.0159414826010698,
+      "learning_rate": 6.3794481738268765e-06,
+      "loss": 0.4151,
+      "step": 2084
+    },
+    {
+      "epoch": 1.4106901217861976,
+      "grad_norm": 0.9925542054758073,
+      "learning_rate": 6.375663704887614e-06,
+      "loss": 0.4114,
+      "step": 2085
+    },
+    {
+      "epoch": 1.4113667117726658,
+      "grad_norm": 0.9996896795037346,
+      "learning_rate": 6.371878383111002e-06,
+      "loss": 0.4165,
+      "step": 2086
+    },
+    {
+      "epoch": 1.4120433017591338,
+      "grad_norm": 1.004250347530124,
+      "learning_rate": 6.368092210843739e-06,
+      "loss": 0.4146,
+      "step": 2087
+    },
+    {
+      "epoch": 1.412719891745602,
+      "grad_norm": 0.9784966367414293,
+      "learning_rate": 6.364305190433049e-06,
+      "loss": 0.3999,
+      "step": 2088
+    },
+    {
+      "epoch": 1.4133964817320703,
+      "grad_norm": 1.0620579280721916,
+      "learning_rate": 6.360517324226676e-06,
+      "loss": 0.404,
+      "step": 2089
+    },
+    {
+      "epoch": 1.4140730717185386,
+      "grad_norm": 1.0224848940212419,
+      "learning_rate": 6.3567286145728944e-06,
+      "loss": 0.4223,
+      "step": 2090
+    },
+    {
+      "epoch": 1.4147496617050068,
+      "grad_norm": 1.0376396714818348,
+      "learning_rate": 6.3529390638205036e-06,
+      "loss": 0.4346,
+      "step": 2091
+    },
+    {
+      "epoch": 1.4154262516914748,
+      "grad_norm": 1.059024918556977,
+      "learning_rate": 6.349148674318816e-06,
+      "loss": 0.4127,
+      "step": 2092
+    },
+    {
+      "epoch": 1.416102841677943,
+      "grad_norm": 1.007584639816123,
+      "learning_rate": 6.34535744841767e-06,
+      "loss": 0.3986,
+      "step": 2093
+    },
+    {
+      "epoch": 1.4167794316644113,
+      "grad_norm": 0.9533261973399811,
+      "learning_rate": 6.341565388467425e-06,
+      "loss": 0.3835,
+      "step": 2094
+    },
+    {
+      "epoch": 1.4174560216508796,
+      "grad_norm": 1.0029995368131168,
+      "learning_rate": 6.3377724968189494e-06,
+      "loss": 0.397,
+      "step": 2095
+    },
+    {
+      "epoch": 1.4181326116373478,
+      "grad_norm": 1.043716743652025,
+      "learning_rate": 6.3339787758236316e-06,
+      "loss": 0.4079,
+      "step": 2096
+    },
+    {
+      "epoch": 1.4188092016238159,
+      "grad_norm": 1.0160555984925257,
+      "learning_rate": 6.330184227833376e-06,
+      "loss": 0.4151,
+      "step": 2097
+    },
+    {
+      "epoch": 1.419485791610284,
+      "grad_norm": 1.0329018979428226,
+      "learning_rate": 6.326388855200598e-06,
+      "loss": 0.4101,
+      "step": 2098
+    },
+    {
+      "epoch": 1.4201623815967523,
+      "grad_norm": 0.9657804031861481,
+      "learning_rate": 6.322592660278223e-06,
+      "loss": 0.4059,
+      "step": 2099
+    },
+    {
+      "epoch": 1.4208389715832206,
+      "grad_norm": 1.0086465820944381,
+      "learning_rate": 6.3187956454196885e-06,
+      "loss": 0.3947,
+      "step": 2100
+    },
+    {
+      "epoch": 1.4215155615696888,
+      "grad_norm": 1.052454309985347,
+      "learning_rate": 6.314997812978938e-06,
+      "loss": 0.432,
+      "step": 2101
+    },
+    {
+      "epoch": 1.4221921515561569,
+      "grad_norm": 1.0484186597494867,
+      "learning_rate": 6.311199165310422e-06,
+      "loss": 0.4273,
+      "step": 2102
+    },
+    {
+      "epoch": 1.422868741542625,
+      "grad_norm": 0.9937917427499688,
+      "learning_rate": 6.3073997047691e-06,
+      "loss": 0.4187,
+      "step": 2103
+    },
+    {
+      "epoch": 1.4235453315290933,
+      "grad_norm": 1.0071429853112834,
+      "learning_rate": 6.30359943371043e-06,
+      "loss": 0.408,
+      "step": 2104
+    },
+    {
+      "epoch": 1.4242219215155616,
+      "grad_norm": 0.9953451190094351,
+      "learning_rate": 6.299798354490376e-06,
+      "loss": 0.4115,
+      "step": 2105
+    },
+    {
+      "epoch": 1.4248985115020298,
+      "grad_norm": 0.9670336945990794,
+      "learning_rate": 6.295996469465404e-06,
+      "loss": 0.3935,
+      "step": 2106
+    },
+    {
+      "epoch": 1.4255751014884979,
+      "grad_norm": 0.9823171512966387,
+      "learning_rate": 6.292193780992475e-06,
+      "loss": 0.3895,
+      "step": 2107
+    },
+    {
+      "epoch": 1.426251691474966,
+      "grad_norm": 1.0153987229933188,
+      "learning_rate": 6.288390291429054e-06,
+      "loss": 0.3911,
+      "step": 2108
+    },
+    {
+      "epoch": 1.4269282814614344,
+      "grad_norm": 1.035840492614243,
+      "learning_rate": 6.284586003133096e-06,
+      "loss": 0.4104,
+      "step": 2109
+    },
+    {
+      "epoch": 1.4276048714479026,
+      "grad_norm": 1.0085552883909579,
+      "learning_rate": 6.280780918463057e-06,
+      "loss": 0.4161,
+      "step": 2110
+    },
+    {
+      "epoch": 1.4282814614343708,
+      "grad_norm": 0.9956474044663148,
+      "learning_rate": 6.276975039777885e-06,
+      "loss": 0.3851,
+      "step": 2111
+    },
+    {
+      "epoch": 1.4289580514208389,
+      "grad_norm": 1.0342229865985904,
+      "learning_rate": 6.2731683694370185e-06,
+      "loss": 0.4256,
+      "step": 2112
+    },
+    {
+      "epoch": 1.4296346414073071,
+      "grad_norm": 1.0160702632344145,
+      "learning_rate": 6.269360909800386e-06,
+      "loss": 0.4168,
+      "step": 2113
+    },
+    {
+      "epoch": 1.4303112313937754,
+      "grad_norm": 0.9608203581897241,
+      "learning_rate": 6.265552663228411e-06,
+      "loss": 0.3755,
+      "step": 2114
+    },
+    {
+      "epoch": 1.4309878213802436,
+      "grad_norm": 1.0125062904723066,
+      "learning_rate": 6.261743632081998e-06,
+      "loss": 0.4154,
+      "step": 2115
+    },
+    {
+      "epoch": 1.4316644113667119,
+      "grad_norm": 0.9670174920797765,
+      "learning_rate": 6.257933818722544e-06,
+      "loss": 0.3824,
+      "step": 2116
+    },
+    {
+      "epoch": 1.4323410013531799,
+      "grad_norm": 1.0672483252699188,
+      "learning_rate": 6.254123225511924e-06,
+      "loss": 0.4192,
+      "step": 2117
+    },
+    {
+      "epoch": 1.4330175913396481,
+      "grad_norm": 1.0042193693383927,
+      "learning_rate": 6.250311854812504e-06,
+      "loss": 0.4081,
+      "step": 2118
+    },
+    {
+      "epoch": 1.4336941813261164,
+      "grad_norm": 1.0040947846388422,
+      "learning_rate": 6.246499708987127e-06,
+      "loss": 0.4172,
+      "step": 2119
+    },
+    {
+      "epoch": 1.4343707713125846,
+      "grad_norm": 1.0531594718451485,
+      "learning_rate": 6.242686790399117e-06,
+      "loss": 0.41,
+      "step": 2120
+    },
+    {
+      "epoch": 1.4350473612990529,
+      "grad_norm": 1.035884758642596,
+      "learning_rate": 6.238873101412282e-06,
+      "loss": 0.4059,
+      "step": 2121
+    },
+    {
+      "epoch": 1.4357239512855209,
+      "grad_norm": 1.0148740218596994,
+      "learning_rate": 6.2350586443908965e-06,
+      "loss": 0.4029,
+      "step": 2122
+    },
+    {
+      "epoch": 1.4364005412719891,
+      "grad_norm": 0.9839871761940735,
+      "learning_rate": 6.231243421699725e-06,
+      "loss": 0.4098,
+      "step": 2123
+    },
+    {
+      "epoch": 1.4370771312584574,
+      "grad_norm": 0.9675074879941716,
+      "learning_rate": 6.227427435703997e-06,
+      "loss": 0.4171,
+      "step": 2124
+    },
+    {
+      "epoch": 1.4377537212449256,
+      "grad_norm": 1.011359475891578,
+      "learning_rate": 6.223610688769418e-06,
+      "loss": 0.4101,
+      "step": 2125
+    },
+    {
+      "epoch": 1.4384303112313939,
+      "grad_norm": 0.9841656612063509,
+      "learning_rate": 6.219793183262165e-06,
+      "loss": 0.3997,
+      "step": 2126
+    },
+    {
+      "epoch": 1.439106901217862,
+      "grad_norm": 1.0397277570030254,
+      "learning_rate": 6.215974921548888e-06,
+      "loss": 0.4202,
+      "step": 2127
+    },
+    {
+      "epoch": 1.4397834912043301,
+      "grad_norm": 0.9854370592211071,
+      "learning_rate": 6.2121559059966995e-06,
+      "loss": 0.4138,
+      "step": 2128
+    },
+    {
+      "epoch": 1.4404600811907984,
+      "grad_norm": 1.0094625634095216,
+      "learning_rate": 6.2083361389731874e-06,
+      "loss": 0.3963,
+      "step": 2129
+    },
+    {
+      "epoch": 1.4411366711772666,
+      "grad_norm": 1.0825200264477999,
+      "learning_rate": 6.204515622846399e-06,
+      "loss": 0.4205,
+      "step": 2130
+    },
+    {
+      "epoch": 1.4418132611637349,
+      "grad_norm": 0.9902815540778717,
+      "learning_rate": 6.200694359984849e-06,
+      "loss": 0.3992,
+      "step": 2131
+    },
+    {
+      "epoch": 1.442489851150203,
+      "grad_norm": 1.0698463297234384,
+      "learning_rate": 6.1968723527575155e-06,
+      "loss": 0.4092,
+      "step": 2132
+    },
+    {
+      "epoch": 1.4431664411366711,
+      "grad_norm": 1.0377798757102696,
+      "learning_rate": 6.193049603533835e-06,
+      "loss": 0.4216,
+      "step": 2133
+    },
+    {
+      "epoch": 1.4438430311231394,
+      "grad_norm": 1.0075304616902754,
+      "learning_rate": 6.189226114683708e-06,
+      "loss": 0.4117,
+      "step": 2134
+    },
+    {
+      "epoch": 1.4445196211096076,
+      "grad_norm": 0.9797038194536343,
+      "learning_rate": 6.185401888577488e-06,
+      "loss": 0.3993,
+      "step": 2135
+    },
+    {
+      "epoch": 1.4451962110960759,
+      "grad_norm": 0.9999885464307031,
+      "learning_rate": 6.181576927585993e-06,
+      "loss": 0.403,
+      "step": 2136
+    },
+    {
+      "epoch": 1.445872801082544,
+      "grad_norm": 1.0130171281322573,
+      "learning_rate": 6.177751234080491e-06,
+      "loss": 0.4189,
+      "step": 2137
+    },
+    {
+      "epoch": 1.4465493910690121,
+      "grad_norm": 1.068396234897941,
+      "learning_rate": 6.173924810432705e-06,
+      "loss": 0.4254,
+      "step": 2138
+    },
+    {
+      "epoch": 1.4472259810554804,
+      "grad_norm": 1.030850482764561,
+      "learning_rate": 6.170097659014812e-06,
+      "loss": 0.4076,
+      "step": 2139
+    },
+    {
+      "epoch": 1.4479025710419486,
+      "grad_norm": 1.0792286942112153,
+      "learning_rate": 6.166269782199441e-06,
+      "loss": 0.4173,
+      "step": 2140
+    },
+    {
+      "epoch": 1.4485791610284169,
+      "grad_norm": 1.0148254137711399,
+      "learning_rate": 6.162441182359667e-06,
+      "loss": 0.408,
+      "step": 2141
+    },
+    {
+      "epoch": 1.449255751014885,
+      "grad_norm": 1.0562316910922944,
+      "learning_rate": 6.158611861869018e-06,
+      "loss": 0.4346,
+      "step": 2142
+    },
+    {
+      "epoch": 1.4499323410013532,
+      "grad_norm": 1.0185731414358785,
+      "learning_rate": 6.154781823101463e-06,
+      "loss": 0.4117,
+      "step": 2143
+    },
+    {
+      "epoch": 1.4506089309878214,
+      "grad_norm": 1.0430597801857664,
+      "learning_rate": 6.150951068431424e-06,
+      "loss": 0.4493,
+      "step": 2144
+    },
+    {
+      "epoch": 1.4512855209742896,
+      "grad_norm": 1.084614047825812,
+      "learning_rate": 6.147119600233758e-06,
+      "loss": 0.4253,
+      "step": 2145
+    },
+    {
+      "epoch": 1.451962110960758,
+      "grad_norm": 1.0144796122155888,
+      "learning_rate": 6.143287420883772e-06,
+      "loss": 0.4051,
+      "step": 2146
+    },
+    {
+      "epoch": 1.452638700947226,
+      "grad_norm": 1.0024725767130016,
+      "learning_rate": 6.1394545327572086e-06,
+      "loss": 0.421,
+      "step": 2147
+    },
+    {
+      "epoch": 1.4533152909336942,
+      "grad_norm": 1.0032027162996657,
+      "learning_rate": 6.135620938230254e-06,
+      "loss": 0.4253,
+      "step": 2148
+    },
+    {
+      "epoch": 1.4539918809201624,
+      "grad_norm": 1.0336166024199123,
+      "learning_rate": 6.131786639679527e-06,
+      "loss": 0.4356,
+      "step": 2149
+    },
+    {
+      "epoch": 1.4546684709066307,
+      "grad_norm": 0.9555994545271854,
+      "learning_rate": 6.127951639482088e-06,
+      "loss": 0.4114,
+      "step": 2150
+    },
+    {
+      "epoch": 1.455345060893099,
+      "grad_norm": 1.0132960523516634,
+      "learning_rate": 6.1241159400154306e-06,
+      "loss": 0.4209,
+      "step": 2151
+    },
+    {
+      "epoch": 1.456021650879567,
+      "grad_norm": 1.0214122607983387,
+      "learning_rate": 6.12027954365748e-06,
+      "loss": 0.4061,
+      "step": 2152
+    },
+    {
+      "epoch": 1.4566982408660352,
+      "grad_norm": 1.0266615457272228,
+      "learning_rate": 6.116442452786599e-06,
+      "loss": 0.4309,
+      "step": 2153
+    },
+    {
+      "epoch": 1.4573748308525034,
+      "grad_norm": 1.011781840536705,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.3876,
+      "step": 2154
+    },
+    {
+      "epoch": 1.4580514208389717,
+      "grad_norm": 1.0360269028968327,
+      "learning_rate": 6.108766197021623e-06,
+      "loss": 0.4088,
+      "step": 2155
+    },
+    {
+      "epoch": 1.45872801082544,
+      "grad_norm": 0.9855891213130734,
+      "learning_rate": 6.104927036886392e-06,
+      "loss": 0.3855,
+      "step": 2156
+    },
+    {
+      "epoch": 1.459404600811908,
+      "grad_norm": 1.0539991329465321,
+      "learning_rate": 6.101087191755958e-06,
+      "loss": 0.4295,
+      "step": 2157
+    },
+    {
+      "epoch": 1.4600811907983762,
+      "grad_norm": 0.9684348580901866,
+      "learning_rate": 6.097246664010813e-06,
+      "loss": 0.3954,
+      "step": 2158
+    },
+    {
+      "epoch": 1.4607577807848444,
+      "grad_norm": 0.9983027694439996,
+      "learning_rate": 6.09340545603188e-06,
+      "loss": 0.4062,
+      "step": 2159
+    },
+    {
+      "epoch": 1.4614343707713127,
+      "grad_norm": 1.0069651354132814,
+      "learning_rate": 6.0895635702004985e-06,
+      "loss": 0.4052,
+      "step": 2160
+    },
+    {
+      "epoch": 1.462110960757781,
+      "grad_norm": 0.9881751226110508,
+      "learning_rate": 6.085721008898434e-06,
+      "loss": 0.409,
+      "step": 2161
+    },
+    {
+      "epoch": 1.462787550744249,
+      "grad_norm": 0.9881654970084861,
+      "learning_rate": 6.081877774507864e-06,
+      "loss": 0.4092,
+      "step": 2162
+    },
+    {
+      "epoch": 1.4634641407307172,
+      "grad_norm": 1.045557510597242,
+      "learning_rate": 6.078033869411389e-06,
+      "loss": 0.4258,
+      "step": 2163
+    },
+    {
+      "epoch": 1.4641407307171854,
+      "grad_norm": 0.976336614296186,
+      "learning_rate": 6.0741892959920205e-06,
+      "loss": 0.3912,
+      "step": 2164
+    },
+    {
+      "epoch": 1.4648173207036537,
+      "grad_norm": 1.000923515659377,
+      "learning_rate": 6.070344056633189e-06,
+      "loss": 0.402,
+      "step": 2165
+    },
+    {
+      "epoch": 1.465493910690122,
+      "grad_norm": 1.0262794468116898,
+      "learning_rate": 6.066498153718735e-06,
+      "loss": 0.4124,
+      "step": 2166
+    },
+    {
+      "epoch": 1.46617050067659,
+      "grad_norm": 0.9873685270438126,
+      "learning_rate": 6.062651589632911e-06,
+      "loss": 0.4041,
+      "step": 2167
+    },
+    {
+      "epoch": 1.4668470906630582,
+      "grad_norm": 0.9844673104985412,
+      "learning_rate": 6.05880436676038e-06,
+      "loss": 0.4114,
+      "step": 2168
+    },
+    {
+      "epoch": 1.4675236806495264,
+      "grad_norm": 0.9890288008350204,
+      "learning_rate": 6.054956487486212e-06,
+      "loss": 0.3869,
+      "step": 2169
+    },
+    {
+      "epoch": 1.4682002706359945,
+      "grad_norm": 1.0490782938832361,
+      "learning_rate": 6.0511079541958825e-06,
+      "loss": 0.4363,
+      "step": 2170
+    },
+    {
+      "epoch": 1.468876860622463,
+      "grad_norm": 1.0553022937329775,
+      "learning_rate": 6.04725876927528e-06,
+      "loss": 0.4331,
+      "step": 2171
+    },
+    {
+      "epoch": 1.469553450608931,
+      "grad_norm": 1.0339233337705305,
+      "learning_rate": 6.043408935110688e-06,
+      "loss": 0.4149,
+      "step": 2172
+    },
+    {
+      "epoch": 1.4702300405953992,
+      "grad_norm": 1.0117611069346526,
+      "learning_rate": 6.039558454088796e-06,
+      "loss": 0.4212,
+      "step": 2173
+    },
+    {
+      "epoch": 1.4709066305818674,
+      "grad_norm": 1.0203444894179419,
+      "learning_rate": 6.035707328596698e-06,
+      "loss": 0.4112,
+      "step": 2174
+    },
+    {
+      "epoch": 1.4715832205683355,
+      "grad_norm": 0.9874046098207819,
+      "learning_rate": 6.0318555610218796e-06,
+      "loss": 0.4023,
+      "step": 2175
+    },
+    {
+      "epoch": 1.472259810554804,
+      "grad_norm": 1.0204399045804808,
+      "learning_rate": 6.0280031537522335e-06,
+      "loss": 0.4057,
+      "step": 2176
+    },
+    {
+      "epoch": 1.472936400541272,
+      "grad_norm": 1.0045546043147813,
+      "learning_rate": 6.02415010917604e-06,
+      "loss": 0.4088,
+      "step": 2177
+    },
+    {
+      "epoch": 1.4736129905277402,
+      "grad_norm": 0.9858879316465735,
+      "learning_rate": 6.020296429681985e-06,
+      "loss": 0.3895,
+      "step": 2178
+    },
+    {
+      "epoch": 1.4742895805142084,
+      "grad_norm": 1.0639995087095757,
+      "learning_rate": 6.016442117659135e-06,
+      "loss": 0.4374,
+      "step": 2179
+    },
+    {
+      "epoch": 1.4749661705006765,
+      "grad_norm": 1.030599144176719,
+      "learning_rate": 6.0125871754969614e-06,
+      "loss": 0.4104,
+      "step": 2180
+    },
+    {
+      "epoch": 1.4756427604871447,
+      "grad_norm": 1.0499180208111727,
+      "learning_rate": 6.0087316055853175e-06,
+      "loss": 0.4295,
+      "step": 2181
+    },
+    {
+      "epoch": 1.476319350473613,
+      "grad_norm": 1.0024112069476467,
+      "learning_rate": 6.00487541031445e-06,
+      "loss": 0.4093,
+      "step": 2182
+    },
+    {
+      "epoch": 1.4769959404600812,
+      "grad_norm": 1.0045344325247096,
+      "learning_rate": 6.001018592074991e-06,
+      "loss": 0.3969,
+      "step": 2183
+    },
+    {
+      "epoch": 1.4776725304465494,
+      "grad_norm": 1.0484222432402965,
+      "learning_rate": 5.997161153257963e-06,
+      "loss": 0.4168,
+      "step": 2184
+    },
+    {
+      "epoch": 1.4783491204330175,
+      "grad_norm": 0.9747271322636657,
+      "learning_rate": 5.9933030962547656e-06,
+      "loss": 0.3912,
+      "step": 2185
+    },
+    {
+      "epoch": 1.4790257104194857,
+      "grad_norm": 0.9710713217965347,
+      "learning_rate": 5.989444423457189e-06,
+      "loss": 0.4005,
+      "step": 2186
+    },
+    {
+      "epoch": 1.479702300405954,
+      "grad_norm": 1.0312150264264268,
+      "learning_rate": 5.985585137257401e-06,
+      "loss": 0.4015,
+      "step": 2187
+    },
+    {
+      "epoch": 1.4803788903924222,
+      "grad_norm": 0.9633740731424717,
+      "learning_rate": 5.981725240047954e-06,
+      "loss": 0.3908,
+      "step": 2188
+    },
+    {
+      "epoch": 1.4810554803788905,
+      "grad_norm": 0.9531478331831216,
+      "learning_rate": 5.977864734221773e-06,
+      "loss": 0.3845,
+      "step": 2189
+    },
+    {
+      "epoch": 1.4817320703653585,
+      "grad_norm": 1.0021867670006501,
+      "learning_rate": 5.974003622172167e-06,
+      "loss": 0.4056,
+      "step": 2190
+    },
+    {
+      "epoch": 1.4824086603518267,
+      "grad_norm": 0.971336601001469,
+      "learning_rate": 5.9701419062928125e-06,
+      "loss": 0.3924,
+      "step": 2191
+    },
+    {
+      "epoch": 1.483085250338295,
+      "grad_norm": 0.9917086008158201,
+      "learning_rate": 5.9662795889777666e-06,
+      "loss": 0.4317,
+      "step": 2192
+    },
+    {
+      "epoch": 1.4837618403247632,
+      "grad_norm": 1.0056804736424774,
+      "learning_rate": 5.962416672621461e-06,
+      "loss": 0.4243,
+      "step": 2193
+    },
+    {
+      "epoch": 1.4844384303112315,
+      "grad_norm": 1.0147683891138681,
+      "learning_rate": 5.958553159618693e-06,
+      "loss": 0.4154,
+      "step": 2194
+    },
+    {
+      "epoch": 1.4851150202976995,
+      "grad_norm": 1.0097165966172468,
+      "learning_rate": 5.954689052364633e-06,
+      "loss": 0.404,
+      "step": 2195
+    },
+    {
+      "epoch": 1.4857916102841677,
+      "grad_norm": 1.0387173804317473,
+      "learning_rate": 5.950824353254818e-06,
+      "loss": 0.4005,
+      "step": 2196
+    },
+    {
+      "epoch": 1.486468200270636,
+      "grad_norm": 1.043193783611487,
+      "learning_rate": 5.946959064685156e-06,
+      "loss": 0.413,
+      "step": 2197
+    },
+    {
+      "epoch": 1.4871447902571042,
+      "grad_norm": 1.0233606891645106,
+      "learning_rate": 5.943093189051916e-06,
+      "loss": 0.4177,
+      "step": 2198
+    },
+    {
+      "epoch": 1.4878213802435725,
+      "grad_norm": 1.0477094318719762,
+      "learning_rate": 5.939226728751733e-06,
+      "loss": 0.4386,
+      "step": 2199
+    },
+    {
+      "epoch": 1.4884979702300405,
+      "grad_norm": 0.983423855482106,
+      "learning_rate": 5.9353596861816e-06,
+      "loss": 0.3966,
+      "step": 2200
+    },
+    {
+      "epoch": 1.4891745602165087,
+      "grad_norm": 0.9964883532351336,
+      "learning_rate": 5.931492063738882e-06,
+      "loss": 0.4394,
+      "step": 2201
+    },
+    {
+      "epoch": 1.489851150202977,
+      "grad_norm": 0.978593472732646,
+      "learning_rate": 5.92762386382129e-06,
+      "loss": 0.4026,
+      "step": 2202
+    },
+    {
+      "epoch": 1.4905277401894452,
+      "grad_norm": 1.0112570921479211,
+      "learning_rate": 5.9237550888269045e-06,
+      "loss": 0.408,
+      "step": 2203
+    },
+    {
+      "epoch": 1.4912043301759135,
+      "grad_norm": 0.9935296823955936,
+      "learning_rate": 5.919885741154155e-06,
+      "loss": 0.4171,
+      "step": 2204
+    },
+    {
+      "epoch": 1.4918809201623815,
+      "grad_norm": 1.033162613222971,
+      "learning_rate": 5.916015823201827e-06,
+      "loss": 0.4321,
+      "step": 2205
+    },
+    {
+      "epoch": 1.4925575101488497,
+      "grad_norm": 0.9962948137010584,
+      "learning_rate": 5.912145337369064e-06,
+      "loss": 0.3992,
+      "step": 2206
+    },
+    {
+      "epoch": 1.493234100135318,
+      "grad_norm": 0.9590877277172909,
+      "learning_rate": 5.908274286055358e-06,
+      "loss": 0.3957,
+      "step": 2207
+    },
+    {
+      "epoch": 1.4939106901217862,
+      "grad_norm": 0.995820507945258,
+      "learning_rate": 5.904402671660551e-06,
+      "loss": 0.3919,
+      "step": 2208
+    },
+    {
+      "epoch": 1.4945872801082545,
+      "grad_norm": 1.0116404554341,
+      "learning_rate": 5.900530496584834e-06,
+      "loss": 0.41,
+      "step": 2209
+    },
+    {
+      "epoch": 1.4952638700947225,
+      "grad_norm": 1.070836313357937,
+      "learning_rate": 5.8966577632287506e-06,
+      "loss": 0.4464,
+      "step": 2210
+    },
+    {
+      "epoch": 1.4959404600811907,
+      "grad_norm": 1.0142241367544296,
+      "learning_rate": 5.892784473993184e-06,
+      "loss": 0.4204,
+      "step": 2211
+    },
+    {
+      "epoch": 1.496617050067659,
+      "grad_norm": 0.9766416633150202,
+      "learning_rate": 5.888910631279366e-06,
+      "loss": 0.4072,
+      "step": 2212
+    },
+    {
+      "epoch": 1.4972936400541272,
+      "grad_norm": 1.0323348362597207,
+      "learning_rate": 5.885036237488868e-06,
+      "loss": 0.4228,
+      "step": 2213
+    },
+    {
+      "epoch": 1.4979702300405955,
+      "grad_norm": 0.9965720514178534,
+      "learning_rate": 5.88116129502361e-06,
+      "loss": 0.424,
+      "step": 2214
+    },
+    {
+      "epoch": 1.4986468200270635,
+      "grad_norm": 0.97259341698905,
+      "learning_rate": 5.8772858062858414e-06,
+      "loss": 0.406,
+      "step": 2215
+    },
+    {
+      "epoch": 1.4993234100135318,
+      "grad_norm": 0.9522869853187673,
+      "learning_rate": 5.873409773678163e-06,
+      "loss": 0.3924,
+      "step": 2216
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.009839991700367,
+      "learning_rate": 5.869533199603498e-06,
+      "loss": 0.4014,
+      "step": 2217
+    },
+    {
+      "epoch": 1.5006765899864682,
+      "grad_norm": 0.9636034088064258,
+      "learning_rate": 5.8656560864651225e-06,
+      "loss": 0.3903,
+      "step": 2218
+    },
+    {
+      "epoch": 1.5013531799729365,
+      "grad_norm": 1.0144118187504318,
+      "learning_rate": 5.861778436666631e-06,
+      "loss": 0.425,
+      "step": 2219
+    },
+    {
+      "epoch": 1.5020297699594045,
+      "grad_norm": 0.9916576777743628,
+      "learning_rate": 5.857900252611959e-06,
+      "loss": 0.4061,
+      "step": 2220
+    },
+    {
+      "epoch": 1.5027063599458728,
+      "grad_norm": 0.9706289075843019,
+      "learning_rate": 5.854021536705373e-06,
+      "loss": 0.3867,
+      "step": 2221
+    },
+    {
+      "epoch": 1.503382949932341,
+      "grad_norm": 0.971966039458518,
+      "learning_rate": 5.8501422913514665e-06,
+      "loss": 0.3951,
+      "step": 2222
+    },
+    {
+      "epoch": 1.5040595399188093,
+      "grad_norm": 0.985326801428694,
+      "learning_rate": 5.846262518955163e-06,
+      "loss": 0.4032,
+      "step": 2223
+    },
+    {
+      "epoch": 1.5047361299052775,
+      "grad_norm": 0.9551996737468442,
+      "learning_rate": 5.842382221921711e-06,
+      "loss": 0.3962,
+      "step": 2224
+    },
+    {
+      "epoch": 1.5054127198917455,
+      "grad_norm": 0.9561646383114321,
+      "learning_rate": 5.838501402656688e-06,
+      "loss": 0.3946,
+      "step": 2225
+    },
+    {
+      "epoch": 1.5060893098782138,
+      "grad_norm": 1.0144061776251505,
+      "learning_rate": 5.83462006356599e-06,
+      "loss": 0.4134,
+      "step": 2226
+    },
+    {
+      "epoch": 1.506765899864682,
+      "grad_norm": 1.005883555201239,
+      "learning_rate": 5.830738207055841e-06,
+      "loss": 0.421,
+      "step": 2227
+    },
+    {
+      "epoch": 1.5074424898511503,
+      "grad_norm": 1.0269198057962325,
+      "learning_rate": 5.8268558355327795e-06,
+      "loss": 0.4198,
+      "step": 2228
+    },
+    {
+      "epoch": 1.5081190798376185,
+      "grad_norm": 1.0346316593454372,
+      "learning_rate": 5.82297295140367e-06,
+      "loss": 0.4206,
+      "step": 2229
+    },
+    {
+      "epoch": 1.5087956698240865,
+      "grad_norm": 1.000171251123499,
+      "learning_rate": 5.819089557075689e-06,
+      "loss": 0.4125,
+      "step": 2230
+    },
+    {
+      "epoch": 1.5094722598105548,
+      "grad_norm": 0.9753200371400269,
+      "learning_rate": 5.815205654956333e-06,
+      "loss": 0.401,
+      "step": 2231
+    },
+    {
+      "epoch": 1.510148849797023,
+      "grad_norm": 1.0139267001233923,
+      "learning_rate": 5.811321247453409e-06,
+      "loss": 0.4191,
+      "step": 2232
+    },
+    {
+      "epoch": 1.510825439783491,
+      "grad_norm": 1.0389399982637497,
+      "learning_rate": 5.807436336975045e-06,
+      "loss": 0.4456,
+      "step": 2233
+    },
+    {
+      "epoch": 1.5115020297699595,
+      "grad_norm": 0.9659952802787669,
+      "learning_rate": 5.803550925929673e-06,
+      "loss": 0.3862,
+      "step": 2234
+    },
+    {
+      "epoch": 1.5121786197564275,
+      "grad_norm": 1.0134302993890432,
+      "learning_rate": 5.799665016726039e-06,
+      "loss": 0.4042,
+      "step": 2235
+    },
+    {
+      "epoch": 1.5128552097428958,
+      "grad_norm": 0.976441311394339,
+      "learning_rate": 5.795778611773197e-06,
+      "loss": 0.4035,
+      "step": 2236
+    },
+    {
+      "epoch": 1.513531799729364,
+      "grad_norm": 1.0074819417472705,
+      "learning_rate": 5.791891713480509e-06,
+      "loss": 0.4151,
+      "step": 2237
+    },
+    {
+      "epoch": 1.514208389715832,
+      "grad_norm": 0.9959722043595142,
+      "learning_rate": 5.788004324257643e-06,
+      "loss": 0.4117,
+      "step": 2238
+    },
+    {
+      "epoch": 1.5148849797023005,
+      "grad_norm": 1.01420060485514,
+      "learning_rate": 5.784116446514571e-06,
+      "loss": 0.4131,
+      "step": 2239
+    },
+    {
+      "epoch": 1.5155615696887685,
+      "grad_norm": 0.9829139365052306,
+      "learning_rate": 5.780228082661564e-06,
+      "loss": 0.4041,
+      "step": 2240
+    },
+    {
+      "epoch": 1.5162381596752368,
+      "grad_norm": 1.0053876242969089,
+      "learning_rate": 5.776339235109203e-06,
+      "loss": 0.422,
+      "step": 2241
+    },
+    {
+      "epoch": 1.516914749661705,
+      "grad_norm": 1.013534384685476,
+      "learning_rate": 5.772449906268362e-06,
+      "loss": 0.4136,
+      "step": 2242
+    },
+    {
+      "epoch": 1.517591339648173,
+      "grad_norm": 1.0148785727587901,
+      "learning_rate": 5.768560098550213e-06,
+      "loss": 0.4031,
+      "step": 2243
+    },
+    {
+      "epoch": 1.5182679296346415,
+      "grad_norm": 1.0474190558815468,
+      "learning_rate": 5.764669814366231e-06,
+      "loss": 0.4234,
+      "step": 2244
+    },
+    {
+      "epoch": 1.5189445196211095,
+      "grad_norm": 1.0018100450636278,
+      "learning_rate": 5.760779056128178e-06,
+      "loss": 0.4046,
+      "step": 2245
+    },
+    {
+      "epoch": 1.5196211096075778,
+      "grad_norm": 1.0116566315373592,
+      "learning_rate": 5.756887826248118e-06,
+      "loss": 0.4063,
+      "step": 2246
+    },
+    {
+      "epoch": 1.520297699594046,
+      "grad_norm": 0.9556860738614604,
+      "learning_rate": 5.752996127138404e-06,
+      "loss": 0.382,
+      "step": 2247
+    },
+    {
+      "epoch": 1.520974289580514,
+      "grad_norm": 1.0173269939376488,
+      "learning_rate": 5.749103961211679e-06,
+      "loss": 0.4043,
+      "step": 2248
+    },
+    {
+      "epoch": 1.5216508795669825,
+      "grad_norm": 0.9942473878490836,
+      "learning_rate": 5.745211330880872e-06,
+      "loss": 0.4094,
+      "step": 2249
+    },
+    {
+      "epoch": 1.5223274695534506,
+      "grad_norm": 0.9556359663419925,
+      "learning_rate": 5.74131823855921e-06,
+      "loss": 0.3828,
+      "step": 2250
+    },
+    {
+      "epoch": 1.5230040595399188,
+      "grad_norm": 1.0265181287659098,
+      "learning_rate": 5.737424686660198e-06,
+      "loss": 0.4019,
+      "step": 2251
+    },
+    {
+      "epoch": 1.523680649526387,
+      "grad_norm": 0.9726488479080468,
+      "learning_rate": 5.733530677597627e-06,
+      "loss": 0.3878,
+      "step": 2252
+    },
+    {
+      "epoch": 1.524357239512855,
+      "grad_norm": 1.0499864864766668,
+      "learning_rate": 5.729636213785574e-06,
+      "loss": 0.423,
+      "step": 2253
+    },
+    {
+      "epoch": 1.5250338294993235,
+      "grad_norm": 1.0068708560616513,
+      "learning_rate": 5.725741297638399e-06,
+      "loss": 0.4278,
+      "step": 2254
+    },
+    {
+      "epoch": 1.5257104194857916,
+      "grad_norm": 1.034615105276963,
+      "learning_rate": 5.721845931570734e-06,
+      "loss": 0.4387,
+      "step": 2255
+    },
+    {
+      "epoch": 1.5263870094722598,
+      "grad_norm": 1.0290506331977352,
+      "learning_rate": 5.717950117997502e-06,
+      "loss": 0.433,
+      "step": 2256
+    },
+    {
+      "epoch": 1.527063599458728,
+      "grad_norm": 0.9811439512559603,
+      "learning_rate": 5.714053859333893e-06,
+      "loss": 0.4013,
+      "step": 2257
+    },
+    {
+      "epoch": 1.527740189445196,
+      "grad_norm": 0.9810755960910839,
+      "learning_rate": 5.710157157995382e-06,
+      "loss": 0.4209,
+      "step": 2258
+    },
+    {
+      "epoch": 1.5284167794316645,
+      "grad_norm": 1.0003383439623459,
+      "learning_rate": 5.70626001639771e-06,
+      "loss": 0.4003,
+      "step": 2259
+    },
+    {
+      "epoch": 1.5290933694181326,
+      "grad_norm": 1.0134598468978735,
+      "learning_rate": 5.702362436956895e-06,
+      "loss": 0.4277,
+      "step": 2260
+    },
+    {
+      "epoch": 1.5297699594046008,
+      "grad_norm": 0.9768564362949052,
+      "learning_rate": 5.6984644220892295e-06,
+      "loss": 0.3834,
+      "step": 2261
+    },
+    {
+      "epoch": 1.530446549391069,
+      "grad_norm": 1.0157926527820418,
+      "learning_rate": 5.694565974211267e-06,
+      "loss": 0.4137,
+      "step": 2262
+    },
+    {
+      "epoch": 1.531123139377537,
+      "grad_norm": 1.0021830220183225,
+      "learning_rate": 5.69066709573984e-06,
+      "loss": 0.4025,
+      "step": 2263
+    },
+    {
+      "epoch": 1.5317997293640055,
+      "grad_norm": 0.9950282307580286,
+      "learning_rate": 5.686767789092041e-06,
+      "loss": 0.4057,
+      "step": 2264
+    },
+    {
+      "epoch": 1.5324763193504736,
+      "grad_norm": 0.9449335611367016,
+      "learning_rate": 5.6828680566852314e-06,
+      "loss": 0.3674,
+      "step": 2265
+    },
+    {
+      "epoch": 1.5331529093369418,
+      "grad_norm": 1.0421155842733352,
+      "learning_rate": 5.678967900937032e-06,
+      "loss": 0.4137,
+      "step": 2266
+    },
+    {
+      "epoch": 1.53382949932341,
+      "grad_norm": 0.9408262218022814,
+      "learning_rate": 5.675067324265332e-06,
+      "loss": 0.3802,
+      "step": 2267
+    },
+    {
+      "epoch": 1.534506089309878,
+      "grad_norm": 0.996290281295342,
+      "learning_rate": 5.671166329088278e-06,
+      "loss": 0.4076,
+      "step": 2268
+    },
+    {
+      "epoch": 1.5351826792963466,
+      "grad_norm": 0.9779660773938046,
+      "learning_rate": 5.667264917824277e-06,
+      "loss": 0.3816,
+      "step": 2269
+    },
+    {
+      "epoch": 1.5358592692828146,
+      "grad_norm": 0.991178142562463,
+      "learning_rate": 5.663363092891991e-06,
+      "loss": 0.4073,
+      "step": 2270
+    },
+    {
+      "epoch": 1.5365358592692828,
+      "grad_norm": 0.9668804347055321,
+      "learning_rate": 5.659460856710346e-06,
+      "loss": 0.3883,
+      "step": 2271
+    },
+    {
+      "epoch": 1.537212449255751,
+      "grad_norm": 0.9683349351541414,
+      "learning_rate": 5.655558211698513e-06,
+      "loss": 0.3926,
+      "step": 2272
+    },
+    {
+      "epoch": 1.537889039242219,
+      "grad_norm": 1.017749415000904,
+      "learning_rate": 5.651655160275925e-06,
+      "loss": 0.3997,
+      "step": 2273
+    },
+    {
+      "epoch": 1.5385656292286876,
+      "grad_norm": 1.0059778627270217,
+      "learning_rate": 5.647751704862263e-06,
+      "loss": 0.4133,
+      "step": 2274
+    },
+    {
+      "epoch": 1.5392422192151556,
+      "grad_norm": 0.9583458928957137,
+      "learning_rate": 5.643847847877458e-06,
+      "loss": 0.3857,
+      "step": 2275
+    },
+    {
+      "epoch": 1.5399188092016238,
+      "grad_norm": 0.9918021848342551,
+      "learning_rate": 5.639943591741691e-06,
+      "loss": 0.4095,
+      "step": 2276
+    },
+    {
+      "epoch": 1.540595399188092,
+      "grad_norm": 1.039030873789592,
+      "learning_rate": 5.636038938875391e-06,
+      "loss": 0.4143,
+      "step": 2277
+    },
+    {
+      "epoch": 1.54127198917456,
+      "grad_norm": 0.9961728205433525,
+      "learning_rate": 5.632133891699232e-06,
+      "loss": 0.4002,
+      "step": 2278
+    },
+    {
+      "epoch": 1.5419485791610286,
+      "grad_norm": 1.0206552817038608,
+      "learning_rate": 5.628228452634132e-06,
+      "loss": 0.4271,
+      "step": 2279
+    },
+    {
+      "epoch": 1.5426251691474966,
+      "grad_norm": 0.9854776972135874,
+      "learning_rate": 5.624322624101255e-06,
+      "loss": 0.4144,
+      "step": 2280
+    },
+    {
+      "epoch": 1.5433017591339648,
+      "grad_norm": 0.9846204437039179,
+      "learning_rate": 5.620416408522002e-06,
+      "loss": 0.3846,
+      "step": 2281
+    },
+    {
+      "epoch": 1.543978349120433,
+      "grad_norm": 0.9734074897864438,
+      "learning_rate": 5.616509808318017e-06,
+      "loss": 0.3956,
+      "step": 2282
+    },
+    {
+      "epoch": 1.544654939106901,
+      "grad_norm": 0.9593990971851657,
+      "learning_rate": 5.612602825911179e-06,
+      "loss": 0.3977,
+      "step": 2283
+    },
+    {
+      "epoch": 1.5453315290933696,
+      "grad_norm": 1.011829251442061,
+      "learning_rate": 5.608695463723614e-06,
+      "loss": 0.3878,
+      "step": 2284
+    },
+    {
+      "epoch": 1.5460081190798376,
+      "grad_norm": 0.9979774715964842,
+      "learning_rate": 5.604787724177666e-06,
+      "loss": 0.4093,
+      "step": 2285
+    },
+    {
+      "epoch": 1.5466847090663058,
+      "grad_norm": 0.9995402562129577,
+      "learning_rate": 5.600879609695929e-06,
+      "loss": 0.4019,
+      "step": 2286
+    },
+    {
+      "epoch": 1.547361299052774,
+      "grad_norm": 0.9885585941891722,
+      "learning_rate": 5.596971122701221e-06,
+      "loss": 0.3953,
+      "step": 2287
+    },
+    {
+      "epoch": 1.548037889039242,
+      "grad_norm": 1.0157358178499796,
+      "learning_rate": 5.593062265616598e-06,
+      "loss": 0.4108,
+      "step": 2288
+    },
+    {
+      "epoch": 1.5487144790257106,
+      "grad_norm": 1.0101678433679804,
+      "learning_rate": 5.589153040865333e-06,
+      "loss": 0.3982,
+      "step": 2289
+    },
+    {
+      "epoch": 1.5493910690121786,
+      "grad_norm": 0.9783707287552068,
+      "learning_rate": 5.585243450870941e-06,
+      "loss": 0.3984,
+      "step": 2290
+    },
+    {
+      "epoch": 1.5500676589986468,
+      "grad_norm": 1.0180500062920153,
+      "learning_rate": 5.581333498057153e-06,
+      "loss": 0.4109,
+      "step": 2291
+    },
+    {
+      "epoch": 1.550744248985115,
+      "grad_norm": 0.9728140985192564,
+      "learning_rate": 5.577423184847932e-06,
+      "loss": 0.3989,
+      "step": 2292
+    },
+    {
+      "epoch": 1.5514208389715831,
+      "grad_norm": 1.0024004656794963,
+      "learning_rate": 5.573512513667459e-06,
+      "loss": 0.4061,
+      "step": 2293
+    },
+    {
+      "epoch": 1.5520974289580516,
+      "grad_norm": 1.0317757167651191,
+      "learning_rate": 5.56960148694014e-06,
+      "loss": 0.4211,
+      "step": 2294
+    },
+    {
+      "epoch": 1.5527740189445196,
+      "grad_norm": 1.041495961186371,
+      "learning_rate": 5.565690107090603e-06,
+      "loss": 0.4235,
+      "step": 2295
+    },
+    {
+      "epoch": 1.5534506089309879,
+      "grad_norm": 0.9670002460128947,
+      "learning_rate": 5.5617783765436894e-06,
+      "loss": 0.4011,
+      "step": 2296
+    },
+    {
+      "epoch": 1.554127198917456,
+      "grad_norm": 1.0031794362545174,
+      "learning_rate": 5.557866297724462e-06,
+      "loss": 0.4095,
+      "step": 2297
+    },
+    {
+      "epoch": 1.5548037889039241,
+      "grad_norm": 1.035520247341661,
+      "learning_rate": 5.553953873058201e-06,
+      "loss": 0.4095,
+      "step": 2298
+    },
+    {
+      "epoch": 1.5554803788903924,
+      "grad_norm": 1.028372688486636,
+      "learning_rate": 5.550041104970398e-06,
+      "loss": 0.4029,
+      "step": 2299
+    },
+    {
+      "epoch": 1.5561569688768606,
+      "grad_norm": 1.0468296714090302,
+      "learning_rate": 5.5461279958867556e-06,
+      "loss": 0.4129,
+      "step": 2300
+    },
+    {
+      "epoch": 1.5568335588633289,
+      "grad_norm": 0.9934287194979324,
+      "learning_rate": 5.542214548233195e-06,
+      "loss": 0.4019,
+      "step": 2301
+    },
+    {
+      "epoch": 1.557510148849797,
+      "grad_norm": 0.9697645355800891,
+      "learning_rate": 5.538300764435838e-06,
+      "loss": 0.3932,
+      "step": 2302
+    },
+    {
+      "epoch": 1.5581867388362651,
+      "grad_norm": 0.960511373094424,
+      "learning_rate": 5.534386646921023e-06,
+      "loss": 0.3925,
+      "step": 2303
+    },
+    {
+      "epoch": 1.5588633288227334,
+      "grad_norm": 0.9698412020165607,
+      "learning_rate": 5.530472198115291e-06,
+      "loss": 0.4018,
+      "step": 2304
+    },
+    {
+      "epoch": 1.5595399188092016,
+      "grad_norm": 1.0075412021710797,
+      "learning_rate": 5.52655742044539e-06,
+      "loss": 0.4014,
+      "step": 2305
+    },
+    {
+      "epoch": 1.5602165087956699,
+      "grad_norm": 0.9531646833014361,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 0.3907,
+      "step": 2306
+    },
+    {
+      "epoch": 1.560893098782138,
+      "grad_norm": 0.975687940759243,
+      "learning_rate": 5.518726888221082e-06,
+      "loss": 0.4209,
+      "step": 2307
+    },
+    {
+      "epoch": 1.5615696887686061,
+      "grad_norm": 0.9912853978664564,
+      "learning_rate": 5.514811138521186e-06,
+      "loss": 0.4105,
+      "step": 2308
+    },
+    {
+      "epoch": 1.5622462787550744,
+      "grad_norm": 1.0186549000050666,
+      "learning_rate": 5.510895069666132e-06,
+      "loss": 0.4078,
+      "step": 2309
+    },
+    {
+      "epoch": 1.5629228687415426,
+      "grad_norm": 0.958316811831813,
+      "learning_rate": 5.506978684083672e-06,
+      "loss": 0.3905,
+      "step": 2310
+    },
+    {
+      "epoch": 1.5635994587280109,
+      "grad_norm": 0.9774153785639785,
+      "learning_rate": 5.503061984201755e-06,
+      "loss": 0.399,
+      "step": 2311
+    },
+    {
+      "epoch": 1.5642760487144791,
+      "grad_norm": 0.955794405103259,
+      "learning_rate": 5.499144972448525e-06,
+      "loss": 0.3931,
+      "step": 2312
+    },
+    {
+      "epoch": 1.5649526387009471,
+      "grad_norm": 0.9774132951159862,
+      "learning_rate": 5.495227651252315e-06,
+      "loss": 0.3832,
+      "step": 2313
+    },
+    {
+      "epoch": 1.5656292286874154,
+      "grad_norm": 0.9987220192751699,
+      "learning_rate": 5.4913100230416536e-06,
+      "loss": 0.4018,
+      "step": 2314
+    },
+    {
+      "epoch": 1.5663058186738836,
+      "grad_norm": 1.0118246470256507,
+      "learning_rate": 5.48739209024526e-06,
+      "loss": 0.3949,
+      "step": 2315
+    },
+    {
+      "epoch": 1.5669824086603519,
+      "grad_norm": 1.0111098350208763,
+      "learning_rate": 5.483473855292043e-06,
+      "loss": 0.4113,
+      "step": 2316
+    },
+    {
+      "epoch": 1.5676589986468201,
+      "grad_norm": 1.011694614551445,
+      "learning_rate": 5.479555320611094e-06,
+      "loss": 0.4105,
+      "step": 2317
+    },
+    {
+      "epoch": 1.5683355886332881,
+      "grad_norm": 0.9944630703558041,
+      "learning_rate": 5.475636488631697e-06,
+      "loss": 0.4006,
+      "step": 2318
+    },
+    {
+      "epoch": 1.5690121786197564,
+      "grad_norm": 0.9691193619216525,
+      "learning_rate": 5.471717361783312e-06,
+      "loss": 0.3984,
+      "step": 2319
+    },
+    {
+      "epoch": 1.5696887686062246,
+      "grad_norm": 0.9681395025541236,
+      "learning_rate": 5.46779794249559e-06,
+      "loss": 0.402,
+      "step": 2320
+    },
+    {
+      "epoch": 1.5703653585926927,
+      "grad_norm": 1.0003940204844282,
+      "learning_rate": 5.463878233198358e-06,
+      "loss": 0.4187,
+      "step": 2321
+    },
+    {
+      "epoch": 1.5710419485791611,
+      "grad_norm": 0.9839529228017467,
+      "learning_rate": 5.459958236321625e-06,
+      "loss": 0.4055,
+      "step": 2322
+    },
+    {
+      "epoch": 1.5717185385656292,
+      "grad_norm": 0.9743159591644858,
+      "learning_rate": 5.4560379542955766e-06,
+      "loss": 0.4036,
+      "step": 2323
+    },
+    {
+      "epoch": 1.5723951285520974,
+      "grad_norm": 0.9310310633257152,
+      "learning_rate": 5.45211738955058e-06,
+      "loss": 0.3816,
+      "step": 2324
+    },
+    {
+      "epoch": 1.5730717185385656,
+      "grad_norm": 1.0129131101388067,
+      "learning_rate": 5.448196544517168e-06,
+      "loss": 0.4107,
+      "step": 2325
+    },
+    {
+      "epoch": 1.5737483085250337,
+      "grad_norm": 0.9800142033118875,
+      "learning_rate": 5.444275421626058e-06,
+      "loss": 0.3877,
+      "step": 2326
+    },
+    {
+      "epoch": 1.5744248985115021,
+      "grad_norm": 1.0316613940545272,
+      "learning_rate": 5.440354023308134e-06,
+      "loss": 0.4144,
+      "step": 2327
+    },
+    {
+      "epoch": 1.5751014884979702,
+      "grad_norm": 0.9644863429417769,
+      "learning_rate": 5.436432351994452e-06,
+      "loss": 0.3902,
+      "step": 2328
+    },
+    {
+      "epoch": 1.5757780784844384,
+      "grad_norm": 0.9648545223376626,
+      "learning_rate": 5.4325104101162345e-06,
+      "loss": 0.3985,
+      "step": 2329
+    },
+    {
+      "epoch": 1.5764546684709067,
+      "grad_norm": 1.003378886490529,
+      "learning_rate": 5.428588200104875e-06,
+      "loss": 0.4017,
+      "step": 2330
+    },
+    {
+      "epoch": 1.5771312584573747,
+      "grad_norm": 0.9680578123630208,
+      "learning_rate": 5.4246657243919345e-06,
+      "loss": 0.4041,
+      "step": 2331
+    },
+    {
+      "epoch": 1.5778078484438431,
+      "grad_norm": 1.0128271350654896,
+      "learning_rate": 5.420742985409132e-06,
+      "loss": 0.4245,
+      "step": 2332
+    },
+    {
+      "epoch": 1.5784844384303112,
+      "grad_norm": 0.9864696145754461,
+      "learning_rate": 5.41681998558836e-06,
+      "loss": 0.3989,
+      "step": 2333
+    },
+    {
+      "epoch": 1.5791610284167794,
+      "grad_norm": 1.0153830884177073,
+      "learning_rate": 5.412896727361663e-06,
+      "loss": 0.4159,
+      "step": 2334
+    },
+    {
+      "epoch": 1.5798376184032477,
+      "grad_norm": 1.0047043718136448,
+      "learning_rate": 5.408973213161251e-06,
+      "loss": 0.4168,
+      "step": 2335
+    },
+    {
+      "epoch": 1.5805142083897157,
+      "grad_norm": 0.9560154987677202,
+      "learning_rate": 5.405049445419488e-06,
+      "loss": 0.3977,
+      "step": 2336
+    },
+    {
+      "epoch": 1.5811907983761841,
+      "grad_norm": 1.0028043875909254,
+      "learning_rate": 5.401125426568904e-06,
+      "loss": 0.4032,
+      "step": 2337
+    },
+    {
+      "epoch": 1.5818673883626522,
+      "grad_norm": 1.0007929350169738,
+      "learning_rate": 5.397201159042176e-06,
+      "loss": 0.4088,
+      "step": 2338
+    },
+    {
+      "epoch": 1.5825439783491204,
+      "grad_norm": 0.9727404241204132,
+      "learning_rate": 5.393276645272139e-06,
+      "loss": 0.4074,
+      "step": 2339
+    },
+    {
+      "epoch": 1.5832205683355887,
+      "grad_norm": 1.0159029139278153,
+      "learning_rate": 5.3893518876917795e-06,
+      "loss": 0.4188,
+      "step": 2340
+    },
+    {
+      "epoch": 1.5838971583220567,
+      "grad_norm": 0.9950006384350888,
+      "learning_rate": 5.385426888734237e-06,
+      "loss": 0.4278,
+      "step": 2341
+    },
+    {
+      "epoch": 1.5845737483085252,
+      "grad_norm": 0.9691195495194651,
+      "learning_rate": 5.381501650832798e-06,
+      "loss": 0.4054,
+      "step": 2342
+    },
+    {
+      "epoch": 1.5852503382949932,
+      "grad_norm": 0.9892340362703014,
+      "learning_rate": 5.377576176420899e-06,
+      "loss": 0.3977,
+      "step": 2343
+    },
+    {
+      "epoch": 1.5859269282814614,
+      "grad_norm": 0.9905294500933729,
+      "learning_rate": 5.373650467932122e-06,
+      "loss": 0.4087,
+      "step": 2344
+    },
+    {
+      "epoch": 1.5866035182679297,
+      "grad_norm": 0.9884003949708602,
+      "learning_rate": 5.3697245278001956e-06,
+      "loss": 0.3924,
+      "step": 2345
+    },
+    {
+      "epoch": 1.5872801082543977,
+      "grad_norm": 1.0255615096231483,
+      "learning_rate": 5.365798358458989e-06,
+      "loss": 0.4116,
+      "step": 2346
+    },
+    {
+      "epoch": 1.5879566982408662,
+      "grad_norm": 0.9813091714942518,
+      "learning_rate": 5.361871962342519e-06,
+      "loss": 0.3929,
+      "step": 2347
+    },
+    {
+      "epoch": 1.5886332882273342,
+      "grad_norm": 0.9838680036315577,
+      "learning_rate": 5.357945341884936e-06,
+      "loss": 0.3949,
+      "step": 2348
+    },
+    {
+      "epoch": 1.5893098782138024,
+      "grad_norm": 0.9736474740941714,
+      "learning_rate": 5.354018499520536e-06,
+      "loss": 0.4057,
+      "step": 2349
+    },
+    {
+      "epoch": 1.5899864682002707,
+      "grad_norm": 0.9952129739574762,
+      "learning_rate": 5.350091437683746e-06,
+      "loss": 0.3984,
+      "step": 2350
+    },
+    {
+      "epoch": 1.5906630581867387,
+      "grad_norm": 1.0008006940252432,
+      "learning_rate": 5.346164158809136e-06,
+      "loss": 0.4107,
+      "step": 2351
+    },
+    {
+      "epoch": 1.5913396481732072,
+      "grad_norm": 0.9824612424974478,
+      "learning_rate": 5.342236665331407e-06,
+      "loss": 0.3954,
+      "step": 2352
+    },
+    {
+      "epoch": 1.5920162381596752,
+      "grad_norm": 0.9938453964896405,
+      "learning_rate": 5.338308959685391e-06,
+      "loss": 0.3972,
+      "step": 2353
+    },
+    {
+      "epoch": 1.5926928281461434,
+      "grad_norm": 1.0569164944507126,
+      "learning_rate": 5.334381044306057e-06,
+      "loss": 0.4054,
+      "step": 2354
+    },
+    {
+      "epoch": 1.5933694181326117,
+      "grad_norm": 0.990150980524875,
+      "learning_rate": 5.3304529216284974e-06,
+      "loss": 0.405,
+      "step": 2355
+    },
+    {
+      "epoch": 1.5940460081190797,
+      "grad_norm": 0.9685797819320492,
+      "learning_rate": 5.32652459408794e-06,
+      "loss": 0.3947,
+      "step": 2356
+    },
+    {
+      "epoch": 1.5947225981055482,
+      "grad_norm": 1.0024465568043461,
+      "learning_rate": 5.322596064119731e-06,
+      "loss": 0.4055,
+      "step": 2357
+    },
+    {
+      "epoch": 1.5953991880920162,
+      "grad_norm": 0.9920328432585775,
+      "learning_rate": 5.318667334159354e-06,
+      "loss": 0.4213,
+      "step": 2358
+    },
+    {
+      "epoch": 1.5960757780784844,
+      "grad_norm": 0.9765680870242228,
+      "learning_rate": 5.314738406642405e-06,
+      "loss": 0.3984,
+      "step": 2359
+    },
+    {
+      "epoch": 1.5967523680649527,
+      "grad_norm": 1.005283725203942,
+      "learning_rate": 5.310809284004608e-06,
+      "loss": 0.416,
+      "step": 2360
+    },
+    {
+      "epoch": 1.5974289580514207,
+      "grad_norm": 0.961858610443524,
+      "learning_rate": 5.306879968681808e-06,
+      "loss": 0.3816,
+      "step": 2361
+    },
+    {
+      "epoch": 1.5981055480378892,
+      "grad_norm": 0.9989041031269195,
+      "learning_rate": 5.30295046310997e-06,
+      "loss": 0.4014,
+      "step": 2362
+    },
+    {
+      "epoch": 1.5987821380243572,
+      "grad_norm": 0.9950318811336293,
+      "learning_rate": 5.299020769725172e-06,
+      "loss": 0.4052,
+      "step": 2363
+    },
+    {
+      "epoch": 1.5994587280108254,
+      "grad_norm": 0.9948212172466734,
+      "learning_rate": 5.2950908909636144e-06,
+      "loss": 0.3939,
+      "step": 2364
+    },
+    {
+      "epoch": 1.6001353179972937,
+      "grad_norm": 1.107328121830043,
+      "learning_rate": 5.2911608292616116e-06,
+      "loss": 0.4115,
+      "step": 2365
+    },
+    {
+      "epoch": 1.6008119079837617,
+      "grad_norm": 1.0294051869885203,
+      "learning_rate": 5.2872305870555874e-06,
+      "loss": 0.4012,
+      "step": 2366
+    },
+    {
+      "epoch": 1.6014884979702302,
+      "grad_norm": 0.9501992577493624,
+      "learning_rate": 5.2833001667820815e-06,
+      "loss": 0.4,
+      "step": 2367
+    },
+    {
+      "epoch": 1.6021650879566982,
+      "grad_norm": 0.9790989327468724,
+      "learning_rate": 5.279369570877742e-06,
+      "loss": 0.4052,
+      "step": 2368
+    },
+    {
+      "epoch": 1.6028416779431665,
+      "grad_norm": 1.0294988926251059,
+      "learning_rate": 5.275438801779328e-06,
+      "loss": 0.4221,
+      "step": 2369
+    },
+    {
+      "epoch": 1.6035182679296347,
+      "grad_norm": 0.9899006835399469,
+      "learning_rate": 5.271507861923701e-06,
+      "loss": 0.3871,
+      "step": 2370
+    },
+    {
+      "epoch": 1.6041948579161027,
+      "grad_norm": 0.9615138135448351,
+      "learning_rate": 5.267576753747839e-06,
+      "loss": 0.3903,
+      "step": 2371
+    },
+    {
+      "epoch": 1.6048714479025712,
+      "grad_norm": 1.0248485881243827,
+      "learning_rate": 5.263645479688807e-06,
+      "loss": 0.4108,
+      "step": 2372
+    },
+    {
+      "epoch": 1.6055480378890392,
+      "grad_norm": 1.0032178822921418,
+      "learning_rate": 5.2597140421837915e-06,
+      "loss": 0.3958,
+      "step": 2373
+    },
+    {
+      "epoch": 1.6062246278755075,
+      "grad_norm": 0.9727493284366812,
+      "learning_rate": 5.255782443670068e-06,
+      "loss": 0.3955,
+      "step": 2374
+    },
+    {
+      "epoch": 1.6069012178619757,
+      "grad_norm": 0.9653040391428355,
+      "learning_rate": 5.251850686585015e-06,
+      "loss": 0.3929,
+      "step": 2375
+    },
+    {
+      "epoch": 1.6075778078484437,
+      "grad_norm": 0.9697107515103446,
+      "learning_rate": 5.247918773366112e-06,
+      "loss": 0.3951,
+      "step": 2376
+    },
+    {
+      "epoch": 1.6082543978349122,
+      "grad_norm": 0.9399237390205499,
+      "learning_rate": 5.243986706450933e-06,
+      "loss": 0.3958,
+      "step": 2377
+    },
+    {
+      "epoch": 1.6089309878213802,
+      "grad_norm": 0.9319844417225434,
+      "learning_rate": 5.240054488277148e-06,
+      "loss": 0.3835,
+      "step": 2378
+    },
+    {
+      "epoch": 1.6096075778078485,
+      "grad_norm": 0.9830045208853183,
+      "learning_rate": 5.2361221212825175e-06,
+      "loss": 0.3961,
+      "step": 2379
+    },
+    {
+      "epoch": 1.6102841677943167,
+      "grad_norm": 0.9819929312852093,
+      "learning_rate": 5.2321896079048994e-06,
+      "loss": 0.3924,
+      "step": 2380
+    },
+    {
+      "epoch": 1.6109607577807847,
+      "grad_norm": 1.0094043944216367,
+      "learning_rate": 5.2282569505822414e-06,
+      "loss": 0.4044,
+      "step": 2381
+    },
+    {
+      "epoch": 1.6116373477672532,
+      "grad_norm": 1.02076523814372,
+      "learning_rate": 5.224324151752575e-06,
+      "loss": 0.4013,
+      "step": 2382
+    },
+    {
+      "epoch": 1.6123139377537212,
+      "grad_norm": 0.954164908647174,
+      "learning_rate": 5.220391213854028e-06,
+      "loss": 0.3812,
+      "step": 2383
+    },
+    {
+      "epoch": 1.6129905277401895,
+      "grad_norm": 0.986288816316368,
+      "learning_rate": 5.216458139324806e-06,
+      "loss": 0.3974,
+      "step": 2384
+    },
+    {
+      "epoch": 1.6136671177266577,
+      "grad_norm": 0.9596211458199466,
+      "learning_rate": 5.212524930603205e-06,
+      "loss": 0.3899,
+      "step": 2385
+    },
+    {
+      "epoch": 1.6143437077131257,
+      "grad_norm": 0.9468358695130175,
+      "learning_rate": 5.208591590127603e-06,
+      "loss": 0.3809,
+      "step": 2386
+    },
+    {
+      "epoch": 1.6150202976995942,
+      "grad_norm": 0.9676314826767896,
+      "learning_rate": 5.2046581203364585e-06,
+      "loss": 0.4009,
+      "step": 2387
+    },
+    {
+      "epoch": 1.6156968876860622,
+      "grad_norm": 0.9503631507105473,
+      "learning_rate": 5.200724523668311e-06,
+      "loss": 0.3966,
+      "step": 2388
+    },
+    {
+      "epoch": 1.6163734776725305,
+      "grad_norm": 0.995098593676996,
+      "learning_rate": 5.196790802561776e-06,
+      "loss": 0.4008,
+      "step": 2389
+    },
+    {
+      "epoch": 1.6170500676589987,
+      "grad_norm": 1.017968085182166,
+      "learning_rate": 5.192856959455552e-06,
+      "loss": 0.3968,
+      "step": 2390
+    },
+    {
+      "epoch": 1.6177266576454667,
+      "grad_norm": 0.9789732775986919,
+      "learning_rate": 5.188922996788409e-06,
+      "loss": 0.4187,
+      "step": 2391
+    },
+    {
+      "epoch": 1.618403247631935,
+      "grad_norm": 1.0108573019123042,
+      "learning_rate": 5.184988916999191e-06,
+      "loss": 0.4223,
+      "step": 2392
+    },
+    {
+      "epoch": 1.6190798376184032,
+      "grad_norm": 1.0106158369845477,
+      "learning_rate": 5.181054722526815e-06,
+      "loss": 0.4037,
+      "step": 2393
+    },
+    {
+      "epoch": 1.6197564276048715,
+      "grad_norm": 0.990290753929674,
+      "learning_rate": 5.177120415810271e-06,
+      "loss": 0.4034,
+      "step": 2394
+    },
+    {
+      "epoch": 1.6204330175913397,
+      "grad_norm": 1.0226528890275217,
+      "learning_rate": 5.173185999288615e-06,
+      "loss": 0.3997,
+      "step": 2395
+    },
+    {
+      "epoch": 1.6211096075778078,
+      "grad_norm": 0.9778349078234524,
+      "learning_rate": 5.1692514754009744e-06,
+      "loss": 0.3897,
+      "step": 2396
+    },
+    {
+      "epoch": 1.621786197564276,
+      "grad_norm": 1.007186403602729,
+      "learning_rate": 5.165316846586541e-06,
+      "loss": 0.4009,
+      "step": 2397
+    },
+    {
+      "epoch": 1.6224627875507442,
+      "grad_norm": 1.0023724528046685,
+      "learning_rate": 5.161382115284576e-06,
+      "loss": 0.3971,
+      "step": 2398
+    },
+    {
+      "epoch": 1.6231393775372125,
+      "grad_norm": 1.017661408717099,
+      "learning_rate": 5.1574472839343956e-06,
+      "loss": 0.4205,
+      "step": 2399
+    },
+    {
+      "epoch": 1.6238159675236807,
+      "grad_norm": 0.9487143093152202,
+      "learning_rate": 5.153512354975388e-06,
+      "loss": 0.3901,
+      "step": 2400
+    },
+    {
+      "epoch": 1.6244925575101488,
+      "grad_norm": 1.0116142038234044,
+      "learning_rate": 5.1495773308469935e-06,
+      "loss": 0.4142,
+      "step": 2401
+    },
+    {
+      "epoch": 1.625169147496617,
+      "grad_norm": 0.989352812414496,
+      "learning_rate": 5.145642213988716e-06,
+      "loss": 0.4017,
+      "step": 2402
+    },
+    {
+      "epoch": 1.6258457374830853,
+      "grad_norm": 1.0344613071561655,
+      "learning_rate": 5.1417070068401165e-06,
+      "loss": 0.4245,
+      "step": 2403
+    },
+    {
+      "epoch": 1.6265223274695535,
+      "grad_norm": 0.9608071894768904,
+      "learning_rate": 5.137771711840811e-06,
+      "loss": 0.4013,
+      "step": 2404
+    },
+    {
+      "epoch": 1.6271989174560217,
+      "grad_norm": 0.9400399139794705,
+      "learning_rate": 5.133836331430469e-06,
+      "loss": 0.3793,
+      "step": 2405
+    },
+    {
+      "epoch": 1.6278755074424898,
+      "grad_norm": 0.9709653980951414,
+      "learning_rate": 5.129900868048817e-06,
+      "loss": 0.3963,
+      "step": 2406
+    },
+    {
+      "epoch": 1.628552097428958,
+      "grad_norm": 0.9429233603810708,
+      "learning_rate": 5.1259653241356275e-06,
+      "loss": 0.3926,
+      "step": 2407
+    },
+    {
+      "epoch": 1.6292286874154263,
+      "grad_norm": 1.0217327433525651,
+      "learning_rate": 5.1220297021307275e-06,
+      "loss": 0.4026,
+      "step": 2408
+    },
+    {
+      "epoch": 1.6299052774018945,
+      "grad_norm": 1.0123619398831514,
+      "learning_rate": 5.11809400447399e-06,
+      "loss": 0.4031,
+      "step": 2409
+    },
+    {
+      "epoch": 1.6305818673883627,
+      "grad_norm": 1.0188862677375232,
+      "learning_rate": 5.114158233605334e-06,
+      "loss": 0.4104,
+      "step": 2410
+    },
+    {
+      "epoch": 1.6312584573748308,
+      "grad_norm": 1.0241358372537814,
+      "learning_rate": 5.110222391964728e-06,
+      "loss": 0.3973,
+      "step": 2411
+    },
+    {
+      "epoch": 1.631935047361299,
+      "grad_norm": 0.9663345385656488,
+      "learning_rate": 5.106286481992179e-06,
+      "loss": 0.3815,
+      "step": 2412
+    },
+    {
+      "epoch": 1.6326116373477673,
+      "grad_norm": 1.0022872276250057,
+      "learning_rate": 5.1023505061277405e-06,
+      "loss": 0.393,
+      "step": 2413
+    },
+    {
+      "epoch": 1.6332882273342353,
+      "grad_norm": 1.0059115627431798,
+      "learning_rate": 5.098414466811504e-06,
+      "loss": 0.418,
+      "step": 2414
+    },
+    {
+      "epoch": 1.6339648173207038,
+      "grad_norm": 0.9583990161145314,
+      "learning_rate": 5.094478366483604e-06,
+      "loss": 0.3999,
+      "step": 2415
+    },
+    {
+      "epoch": 1.6346414073071718,
+      "grad_norm": 0.9622835480998754,
+      "learning_rate": 5.090542207584207e-06,
+      "loss": 0.3834,
+      "step": 2416
+    },
+    {
+      "epoch": 1.63531799729364,
+      "grad_norm": 1.0064874711712921,
+      "learning_rate": 5.086605992553524e-06,
+      "loss": 0.4022,
+      "step": 2417
+    },
+    {
+      "epoch": 1.6359945872801083,
+      "grad_norm": 0.9908528598407368,
+      "learning_rate": 5.082669723831793e-06,
+      "loss": 0.3993,
+      "step": 2418
+    },
+    {
+      "epoch": 1.6366711772665763,
+      "grad_norm": 0.9615276986995829,
+      "learning_rate": 5.07873340385929e-06,
+      "loss": 0.3902,
+      "step": 2419
+    },
+    {
+      "epoch": 1.6373477672530448,
+      "grad_norm": 1.0495752425472613,
+      "learning_rate": 5.074797035076319e-06,
+      "loss": 0.4242,
+      "step": 2420
+    },
+    {
+      "epoch": 1.6380243572395128,
+      "grad_norm": 0.9585523749831784,
+      "learning_rate": 5.070860619923218e-06,
+      "loss": 0.3969,
+      "step": 2421
+    },
+    {
+      "epoch": 1.638700947225981,
+      "grad_norm": 1.015554940825568,
+      "learning_rate": 5.066924160840353e-06,
+      "loss": 0.4233,
+      "step": 2422
+    },
+    {
+      "epoch": 1.6393775372124493,
+      "grad_norm": 0.9467678571613855,
+      "learning_rate": 5.062987660268114e-06,
+      "loss": 0.388,
+      "step": 2423
+    },
+    {
+      "epoch": 1.6400541271989173,
+      "grad_norm": 1.000460880805651,
+      "learning_rate": 5.059051120646924e-06,
+      "loss": 0.4028,
+      "step": 2424
+    },
+    {
+      "epoch": 1.6407307171853858,
+      "grad_norm": 0.9797102196071796,
+      "learning_rate": 5.055114544417219e-06,
+      "loss": 0.4048,
+      "step": 2425
+    },
+    {
+      "epoch": 1.6414073071718538,
+      "grad_norm": 0.9702410011455744,
+      "learning_rate": 5.051177934019468e-06,
+      "loss": 0.3915,
+      "step": 2426
+    },
+    {
+      "epoch": 1.642083897158322,
+      "grad_norm": 0.9370772551216627,
+      "learning_rate": 5.047241291894156e-06,
+      "loss": 0.3731,
+      "step": 2427
+    },
+    {
+      "epoch": 1.6427604871447903,
+      "grad_norm": 1.026925123148177,
+      "learning_rate": 5.043304620481791e-06,
+      "loss": 0.407,
+      "step": 2428
+    },
+    {
+      "epoch": 1.6434370771312583,
+      "grad_norm": 0.9992844526014942,
+      "learning_rate": 5.039367922222894e-06,
+      "loss": 0.3979,
+      "step": 2429
+    },
+    {
+      "epoch": 1.6441136671177268,
+      "grad_norm": 1.0054139793882098,
+      "learning_rate": 5.035431199558008e-06,
+      "loss": 0.4016,
+      "step": 2430
+    },
+    {
+      "epoch": 1.6447902571041948,
+      "grad_norm": 0.9497452369957721,
+      "learning_rate": 5.031494454927688e-06,
+      "loss": 0.3975,
+      "step": 2431
+    },
+    {
+      "epoch": 1.645466847090663,
+      "grad_norm": 0.9291845580578972,
+      "learning_rate": 5.027557690772503e-06,
+      "loss": 0.3897,
+      "step": 2432
+    },
+    {
+      "epoch": 1.6461434370771313,
+      "grad_norm": 0.9440350987672868,
+      "learning_rate": 5.0236209095330344e-06,
+      "loss": 0.391,
+      "step": 2433
+    },
+    {
+      "epoch": 1.6468200270635993,
+      "grad_norm": 1.0391267060279628,
+      "learning_rate": 5.019684113649877e-06,
+      "loss": 0.4143,
+      "step": 2434
+    },
+    {
+      "epoch": 1.6474966170500678,
+      "grad_norm": 0.9543893791124238,
+      "learning_rate": 5.0157473055636285e-06,
+      "loss": 0.3975,
+      "step": 2435
+    },
+    {
+      "epoch": 1.6481732070365358,
+      "grad_norm": 1.0037319738186286,
+      "learning_rate": 5.011810487714901e-06,
+      "loss": 0.4081,
+      "step": 2436
+    },
+    {
+      "epoch": 1.648849797023004,
+      "grad_norm": 0.9959185049663246,
+      "learning_rate": 5.007873662544306e-06,
+      "loss": 0.3842,
+      "step": 2437
+    },
+    {
+      "epoch": 1.6495263870094723,
+      "grad_norm": 1.0285901605597476,
+      "learning_rate": 5.003936832492465e-06,
+      "loss": 0.4217,
+      "step": 2438
+    },
+    {
+      "epoch": 1.6502029769959403,
+      "grad_norm": 0.996720165050114,
+      "learning_rate": 5e-06,
+      "loss": 0.4069,
+      "step": 2439
+    },
+    {
+      "epoch": 1.6508795669824088,
+      "grad_norm": 1.0165325946286154,
+      "learning_rate": 4.9960631675075364e-06,
+      "loss": 0.4122,
+      "step": 2440
+    },
+    {
+      "epoch": 1.6515561569688768,
+      "grad_norm": 0.9838467350512484,
+      "learning_rate": 4.9921263374556946e-06,
+      "loss": 0.4112,
+      "step": 2441
+    },
+    {
+      "epoch": 1.652232746955345,
+      "grad_norm": 0.9358784159452869,
+      "learning_rate": 4.988189512285101e-06,
+      "loss": 0.3737,
+      "step": 2442
+    },
+    {
+      "epoch": 1.6529093369418133,
+      "grad_norm": 1.0590403923273386,
+      "learning_rate": 4.984252694436373e-06,
+      "loss": 0.4196,
+      "step": 2443
+    },
+    {
+      "epoch": 1.6535859269282813,
+      "grad_norm": 0.9860837940120097,
+      "learning_rate": 4.980315886350125e-06,
+      "loss": 0.4086,
+      "step": 2444
+    },
+    {
+      "epoch": 1.6542625169147498,
+      "grad_norm": 0.9410651720083746,
+      "learning_rate": 4.976379090466966e-06,
+      "loss": 0.3785,
+      "step": 2445
+    },
+    {
+      "epoch": 1.6549391069012178,
+      "grad_norm": 1.0073897065366546,
+      "learning_rate": 4.972442309227498e-06,
+      "loss": 0.4233,
+      "step": 2446
+    },
+    {
+      "epoch": 1.655615696887686,
+      "grad_norm": 0.9760170203175808,
+      "learning_rate": 4.968505545072314e-06,
+      "loss": 0.3953,
+      "step": 2447
+    },
+    {
+      "epoch": 1.6562922868741543,
+      "grad_norm": 0.9929780657948141,
+      "learning_rate": 4.964568800441993e-06,
+      "loss": 0.401,
+      "step": 2448
+    },
+    {
+      "epoch": 1.6569688768606223,
+      "grad_norm": 1.0440276577316752,
+      "learning_rate": 4.960632077777107e-06,
+      "loss": 0.4195,
+      "step": 2449
+    },
+    {
+      "epoch": 1.6576454668470908,
+      "grad_norm": 0.9456506570153204,
+      "learning_rate": 4.956695379518211e-06,
+      "loss": 0.3831,
+      "step": 2450
+    },
+    {
+      "epoch": 1.6583220568335588,
+      "grad_norm": 0.9965709761859097,
+      "learning_rate": 4.952758708105845e-06,
+      "loss": 0.4035,
+      "step": 2451
+    },
+    {
+      "epoch": 1.658998646820027,
+      "grad_norm": 1.0007676453145316,
+      "learning_rate": 4.948822065980533e-06,
+      "loss": 0.4133,
+      "step": 2452
+    },
+    {
+      "epoch": 1.6596752368064953,
+      "grad_norm": 1.018873735482474,
+      "learning_rate": 4.944885455582783e-06,
+      "loss": 0.4055,
+      "step": 2453
+    },
+    {
+      "epoch": 1.6603518267929633,
+      "grad_norm": 0.9989665958176519,
+      "learning_rate": 4.940948879353078e-06,
+      "loss": 0.4063,
+      "step": 2454
+    },
+    {
+      "epoch": 1.6610284167794318,
+      "grad_norm": 0.9575429430528978,
+      "learning_rate": 4.937012339731886e-06,
+      "loss": 0.3895,
+      "step": 2455
+    },
+    {
+      "epoch": 1.6617050067658998,
+      "grad_norm": 1.0094319061016028,
+      "learning_rate": 4.933075839159649e-06,
+      "loss": 0.414,
+      "step": 2456
+    },
+    {
+      "epoch": 1.662381596752368,
+      "grad_norm": 1.0013037553216197,
+      "learning_rate": 4.929139380076784e-06,
+      "loss": 0.4009,
+      "step": 2457
+    },
+    {
+      "epoch": 1.6630581867388363,
+      "grad_norm": 0.9701707451088951,
+      "learning_rate": 4.9252029649236835e-06,
+      "loss": 0.3904,
+      "step": 2458
+    },
+    {
+      "epoch": 1.6637347767253043,
+      "grad_norm": 1.1201475247092365,
+      "learning_rate": 4.921266596140712e-06,
+      "loss": 0.427,
+      "step": 2459
+    },
+    {
+      "epoch": 1.6644113667117728,
+      "grad_norm": 0.9840473255127528,
+      "learning_rate": 4.917330276168208e-06,
+      "loss": 0.4007,
+      "step": 2460
+    },
+    {
+      "epoch": 1.6650879566982408,
+      "grad_norm": 0.9613880804638744,
+      "learning_rate": 4.913394007446477e-06,
+      "loss": 0.4058,
+      "step": 2461
+    },
+    {
+      "epoch": 1.665764546684709,
+      "grad_norm": 0.9918636978939089,
+      "learning_rate": 4.909457792415793e-06,
+      "loss": 0.3991,
+      "step": 2462
+    },
+    {
+      "epoch": 1.6664411366711773,
+      "grad_norm": 0.9647445437940695,
+      "learning_rate": 4.905521633516399e-06,
+      "loss": 0.3828,
+      "step": 2463
+    },
+    {
+      "epoch": 1.6671177266576453,
+      "grad_norm": 1.052504476643506,
+      "learning_rate": 4.9015855331884984e-06,
+      "loss": 0.419,
+      "step": 2464
+    },
+    {
+      "epoch": 1.6677943166441138,
+      "grad_norm": 0.9525407604860707,
+      "learning_rate": 4.897649493872262e-06,
+      "loss": 0.3746,
+      "step": 2465
+    },
+    {
+      "epoch": 1.6684709066305818,
+      "grad_norm": 0.9277772490225074,
+      "learning_rate": 4.8937135180078236e-06,
+      "loss": 0.3884,
+      "step": 2466
+    },
+    {
+      "epoch": 1.66914749661705,
+      "grad_norm": 1.0117669002250895,
+      "learning_rate": 4.889777608035273e-06,
+      "loss": 0.4063,
+      "step": 2467
+    },
+    {
+      "epoch": 1.6698240866035183,
+      "grad_norm": 1.055597674608897,
+      "learning_rate": 4.8858417663946665e-06,
+      "loss": 0.4061,
+      "step": 2468
+    },
+    {
+      "epoch": 1.6705006765899864,
+      "grad_norm": 0.9758827935957363,
+      "learning_rate": 4.8819059955260105e-06,
+      "loss": 0.4125,
+      "step": 2469
+    },
+    {
+      "epoch": 1.6711772665764548,
+      "grad_norm": 1.0205766549635598,
+      "learning_rate": 4.877970297869273e-06,
+      "loss": 0.4306,
+      "step": 2470
+    },
+    {
+      "epoch": 1.6718538565629228,
+      "grad_norm": 0.9600860616578318,
+      "learning_rate": 4.874034675864373e-06,
+      "loss": 0.397,
+      "step": 2471
+    },
+    {
+      "epoch": 1.672530446549391,
+      "grad_norm": 0.974992711207263,
+      "learning_rate": 4.870099131951185e-06,
+      "loss": 0.3937,
+      "step": 2472
+    },
+    {
+      "epoch": 1.6732070365358593,
+      "grad_norm": 0.9862395000213485,
+      "learning_rate": 4.866163668569531e-06,
+      "loss": 0.415,
+      "step": 2473
+    },
+    {
+      "epoch": 1.6738836265223274,
+      "grad_norm": 0.9325508078781863,
+      "learning_rate": 4.862228288159191e-06,
+      "loss": 0.383,
+      "step": 2474
+    },
+    {
+      "epoch": 1.6745602165087958,
+      "grad_norm": 0.9730870854691235,
+      "learning_rate": 4.858292993159884e-06,
+      "loss": 0.4132,
+      "step": 2475
+    },
+    {
+      "epoch": 1.6752368064952639,
+      "grad_norm": 0.9273973957731927,
+      "learning_rate": 4.854357786011286e-06,
+      "loss": 0.3898,
+      "step": 2476
+    },
+    {
+      "epoch": 1.675913396481732,
+      "grad_norm": 0.9323393999201217,
+      "learning_rate": 4.850422669153009e-06,
+      "loss": 0.3856,
+      "step": 2477
+    },
+    {
+      "epoch": 1.6765899864682003,
+      "grad_norm": 0.9978979968034617,
+      "learning_rate": 4.846487645024614e-06,
+      "loss": 0.4136,
+      "step": 2478
+    },
+    {
+      "epoch": 1.6772665764546684,
+      "grad_norm": 0.9919189258668775,
+      "learning_rate": 4.842552716065605e-06,
+      "loss": 0.393,
+      "step": 2479
+    },
+    {
+      "epoch": 1.6779431664411368,
+      "grad_norm": 0.9860228586672177,
+      "learning_rate": 4.838617884715425e-06,
+      "loss": 0.3926,
+      "step": 2480
+    },
+    {
+      "epoch": 1.6786197564276049,
+      "grad_norm": 0.9497105596467433,
+      "learning_rate": 4.8346831534134595e-06,
+      "loss": 0.4007,
+      "step": 2481
+    },
+    {
+      "epoch": 1.679296346414073,
+      "grad_norm": 0.946837631080875,
+      "learning_rate": 4.830748524599026e-06,
+      "loss": 0.3793,
+      "step": 2482
+    },
+    {
+      "epoch": 1.6799729364005414,
+      "grad_norm": 0.9792036312094987,
+      "learning_rate": 4.826814000711388e-06,
+      "loss": 0.4066,
+      "step": 2483
+    },
+    {
+      "epoch": 1.6806495263870094,
+      "grad_norm": 0.9987975919217557,
+      "learning_rate": 4.822879584189732e-06,
+      "loss": 0.4023,
+      "step": 2484
+    },
+    {
+      "epoch": 1.6813261163734776,
+      "grad_norm": 0.974885705723936,
+      "learning_rate": 4.818945277473187e-06,
+      "loss": 0.3855,
+      "step": 2485
+    },
+    {
+      "epoch": 1.6820027063599459,
+      "grad_norm": 0.9920840388676263,
+      "learning_rate": 4.81501108300081e-06,
+      "loss": 0.4026,
+      "step": 2486
+    },
+    {
+      "epoch": 1.682679296346414,
+      "grad_norm": 0.9793117271477235,
+      "learning_rate": 4.811077003211592e-06,
+      "loss": 0.4056,
+      "step": 2487
+    },
+    {
+      "epoch": 1.6833558863328824,
+      "grad_norm": 0.9590240533571854,
+      "learning_rate": 4.807143040544448e-06,
+      "loss": 0.3866,
+      "step": 2488
+    },
+    {
+      "epoch": 1.6840324763193504,
+      "grad_norm": 0.9938264882104233,
+      "learning_rate": 4.803209197438224e-06,
+      "loss": 0.3956,
+      "step": 2489
+    },
+    {
+      "epoch": 1.6847090663058186,
+      "grad_norm": 0.9751500959166967,
+      "learning_rate": 4.799275476331692e-06,
+      "loss": 0.3935,
+      "step": 2490
+    },
+    {
+      "epoch": 1.6853856562922869,
+      "grad_norm": 0.9848046221330152,
+      "learning_rate": 4.795341879663543e-06,
+      "loss": 0.3991,
+      "step": 2491
+    },
+    {
+      "epoch": 1.6860622462787551,
+      "grad_norm": 0.9968903361134329,
+      "learning_rate": 4.791408409872398e-06,
+      "loss": 0.3877,
+      "step": 2492
+    },
+    {
+      "epoch": 1.6867388362652234,
+      "grad_norm": 0.9559870333735009,
+      "learning_rate": 4.787475069396796e-06,
+      "loss": 0.3815,
+      "step": 2493
+    },
+    {
+      "epoch": 1.6874154262516914,
+      "grad_norm": 1.0082635190519689,
+      "learning_rate": 4.783541860675195e-06,
+      "loss": 0.4023,
+      "step": 2494
+    },
+    {
+      "epoch": 1.6880920162381596,
+      "grad_norm": 0.9340656171911037,
+      "learning_rate": 4.779608786145974e-06,
+      "loss": 0.372,
+      "step": 2495
+    },
+    {
+      "epoch": 1.6887686062246279,
+      "grad_norm": 1.035107897292333,
+      "learning_rate": 4.775675848247427e-06,
+      "loss": 0.3865,
+      "step": 2496
+    },
+    {
+      "epoch": 1.6894451962110961,
+      "grad_norm": 0.9419311321149763,
+      "learning_rate": 4.771743049417761e-06,
+      "loss": 0.3835,
+      "step": 2497
+    },
+    {
+      "epoch": 1.6901217861975644,
+      "grad_norm": 0.9655304719839376,
+      "learning_rate": 4.767810392095102e-06,
+      "loss": 0.3812,
+      "step": 2498
+    },
+    {
+      "epoch": 1.6907983761840324,
+      "grad_norm": 1.0089002997344731,
+      "learning_rate": 4.763877878717484e-06,
+      "loss": 0.3771,
+      "step": 2499
+    },
+    {
+      "epoch": 1.6914749661705006,
+      "grad_norm": 1.0046405398410736,
+      "learning_rate": 4.759945511722854e-06,
+      "loss": 0.4076,
+      "step": 2500
+    },
+    {
+      "epoch": 1.6921515561569689,
+      "grad_norm": 1.0425561925430737,
+      "learning_rate": 4.756013293549067e-06,
+      "loss": 0.4041,
+      "step": 2501
+    },
+    {
+      "epoch": 1.6928281461434371,
+      "grad_norm": 1.0285633291185599,
+      "learning_rate": 4.752081226633888e-06,
+      "loss": 0.4041,
+      "step": 2502
+    },
+    {
+      "epoch": 1.6935047361299054,
+      "grad_norm": 0.9941045938081207,
+      "learning_rate": 4.748149313414987e-06,
+      "loss": 0.3897,
+      "step": 2503
+    },
+    {
+      "epoch": 1.6941813261163734,
+      "grad_norm": 0.9794331946229003,
+      "learning_rate": 4.744217556329935e-06,
+      "loss": 0.4013,
+      "step": 2504
+    },
+    {
+      "epoch": 1.6948579161028416,
+      "grad_norm": 0.9781149286705279,
+      "learning_rate": 4.740285957816211e-06,
+      "loss": 0.3984,
+      "step": 2505
+    },
+    {
+      "epoch": 1.69553450608931,
+      "grad_norm": 0.9883140881255817,
+      "learning_rate": 4.736354520311194e-06,
+      "loss": 0.4042,
+      "step": 2506
+    },
+    {
+      "epoch": 1.696211096075778,
+      "grad_norm": 0.9809521054755324,
+      "learning_rate": 4.732423246252164e-06,
+      "loss": 0.4019,
+      "step": 2507
+    },
+    {
+      "epoch": 1.6968876860622464,
+      "grad_norm": 0.9582374094988552,
+      "learning_rate": 4.728492138076299e-06,
+      "loss": 0.3904,
+      "step": 2508
+    },
+    {
+      "epoch": 1.6975642760487144,
+      "grad_norm": 1.0099402802386952,
+      "learning_rate": 4.724561198220672e-06,
+      "loss": 0.4152,
+      "step": 2509
+    },
+    {
+      "epoch": 1.6982408660351827,
+      "grad_norm": 0.9576961148741598,
+      "learning_rate": 4.7206304291222585e-06,
+      "loss": 0.412,
+      "step": 2510
+    },
+    {
+      "epoch": 1.698917456021651,
+      "grad_norm": 0.9957008712591706,
+      "learning_rate": 4.71669983321792e-06,
+      "loss": 0.3992,
+      "step": 2511
+    },
+    {
+      "epoch": 1.699594046008119,
+      "grad_norm": 0.9975015024573402,
+      "learning_rate": 4.712769412944413e-06,
+      "loss": 0.4093,
+      "step": 2512
+    },
+    {
+      "epoch": 1.7002706359945874,
+      "grad_norm": 1.0042184608801534,
+      "learning_rate": 4.70883917073839e-06,
+      "loss": 0.4211,
+      "step": 2513
+    },
+    {
+      "epoch": 1.7009472259810554,
+      "grad_norm": 0.9747676471139658,
+      "learning_rate": 4.704909109036387e-06,
+      "loss": 0.4074,
+      "step": 2514
+    },
+    {
+      "epoch": 1.7016238159675237,
+      "grad_norm": 1.0168200233189133,
+      "learning_rate": 4.700979230274829e-06,
+      "loss": 0.4067,
+      "step": 2515
+    },
+    {
+      "epoch": 1.702300405953992,
+      "grad_norm": 0.9894428805887173,
+      "learning_rate": 4.697049536890033e-06,
+      "loss": 0.3858,
+      "step": 2516
+    },
+    {
+      "epoch": 1.70297699594046,
+      "grad_norm": 0.963903563044763,
+      "learning_rate": 4.693120031318194e-06,
+      "loss": 0.4,
+      "step": 2517
+    },
+    {
+      "epoch": 1.7036535859269284,
+      "grad_norm": 0.9568426588036553,
+      "learning_rate": 4.6891907159953935e-06,
+      "loss": 0.4018,
+      "step": 2518
+    },
+    {
+      "epoch": 1.7043301759133964,
+      "grad_norm": 0.9947049613394754,
+      "learning_rate": 4.685261593357598e-06,
+      "loss": 0.406,
+      "step": 2519
+    },
+    {
+      "epoch": 1.7050067658998647,
+      "grad_norm": 1.0081997102706153,
+      "learning_rate": 4.681332665840647e-06,
+      "loss": 0.409,
+      "step": 2520
+    },
+    {
+      "epoch": 1.705683355886333,
+      "grad_norm": 0.949419793529757,
+      "learning_rate": 4.677403935880269e-06,
+      "loss": 0.3956,
+      "step": 2521
+    },
+    {
+      "epoch": 1.706359945872801,
+      "grad_norm": 1.0266615324466901,
+      "learning_rate": 4.673475405912061e-06,
+      "loss": 0.3994,
+      "step": 2522
+    },
+    {
+      "epoch": 1.7070365358592694,
+      "grad_norm": 0.9190420839204616,
+      "learning_rate": 4.669547078371503e-06,
+      "loss": 0.3902,
+      "step": 2523
+    },
+    {
+      "epoch": 1.7077131258457374,
+      "grad_norm": 0.9818924505177589,
+      "learning_rate": 4.6656189556939446e-06,
+      "loss": 0.3862,
+      "step": 2524
+    },
+    {
+      "epoch": 1.7083897158322057,
+      "grad_norm": 0.9706183912843495,
+      "learning_rate": 4.6616910403146095e-06,
+      "loss": 0.3762,
+      "step": 2525
+    },
+    {
+      "epoch": 1.709066305818674,
+      "grad_norm": 0.9735002002315447,
+      "learning_rate": 4.657763334668594e-06,
+      "loss": 0.3921,
+      "step": 2526
+    },
+    {
+      "epoch": 1.709742895805142,
+      "grad_norm": 0.9538228455638207,
+      "learning_rate": 4.653835841190865e-06,
+      "loss": 0.3955,
+      "step": 2527
+    },
+    {
+      "epoch": 1.7104194857916104,
+      "grad_norm": 0.9702689262931259,
+      "learning_rate": 4.649908562316255e-06,
+      "loss": 0.4019,
+      "step": 2528
+    },
+    {
+      "epoch": 1.7110960757780784,
+      "grad_norm": 0.9772188973398663,
+      "learning_rate": 4.645981500479466e-06,
+      "loss": 0.3902,
+      "step": 2529
+    },
+    {
+      "epoch": 1.7117726657645467,
+      "grad_norm": 0.9759152384890787,
+      "learning_rate": 4.6420546581150665e-06,
+      "loss": 0.4073,
+      "step": 2530
+    },
+    {
+      "epoch": 1.712449255751015,
+      "grad_norm": 0.9932995343832387,
+      "learning_rate": 4.6381280376574836e-06,
+      "loss": 0.4033,
+      "step": 2531
+    },
+    {
+      "epoch": 1.713125845737483,
+      "grad_norm": 1.0010008985759689,
+      "learning_rate": 4.634201641541013e-06,
+      "loss": 0.4068,
+      "step": 2532
+    },
+    {
+      "epoch": 1.7138024357239514,
+      "grad_norm": 0.961355633082657,
+      "learning_rate": 4.630275472199805e-06,
+      "loss": 0.3874,
+      "step": 2533
+    },
+    {
+      "epoch": 1.7144790257104194,
+      "grad_norm": 0.9217728068074276,
+      "learning_rate": 4.626349532067879e-06,
+      "loss": 0.3934,
+      "step": 2534
+    },
+    {
+      "epoch": 1.7151556156968877,
+      "grad_norm": 0.976942632753318,
+      "learning_rate": 4.622423823579102e-06,
+      "loss": 0.3887,
+      "step": 2535
+    },
+    {
+      "epoch": 1.715832205683356,
+      "grad_norm": 1.015293084944598,
+      "learning_rate": 4.618498349167204e-06,
+      "loss": 0.4073,
+      "step": 2536
+    },
+    {
+      "epoch": 1.716508795669824,
+      "grad_norm": 0.9932608942785961,
+      "learning_rate": 4.6145731112657644e-06,
+      "loss": 0.411,
+      "step": 2537
+    },
+    {
+      "epoch": 1.7171853856562924,
+      "grad_norm": 0.9641347696128801,
+      "learning_rate": 4.610648112308221e-06,
+      "loss": 0.4009,
+      "step": 2538
+    },
+    {
+      "epoch": 1.7178619756427604,
+      "grad_norm": 0.9671623185500506,
+      "learning_rate": 4.6067233547278614e-06,
+      "loss": 0.3955,
+      "step": 2539
+    },
+    {
+      "epoch": 1.7185385656292287,
+      "grad_norm": 0.9321717137763164,
+      "learning_rate": 4.602798840957825e-06,
+      "loss": 0.3892,
+      "step": 2540
+    },
+    {
+      "epoch": 1.719215155615697,
+      "grad_norm": 0.9263826797640947,
+      "learning_rate": 4.598874573431097e-06,
+      "loss": 0.3874,
+      "step": 2541
+    },
+    {
+      "epoch": 1.719891745602165,
+      "grad_norm": 0.970645935808104,
+      "learning_rate": 4.594950554580512e-06,
+      "loss": 0.4122,
+      "step": 2542
+    },
+    {
+      "epoch": 1.7205683355886334,
+      "grad_norm": 0.9428103978257253,
+      "learning_rate": 4.5910267868387525e-06,
+      "loss": 0.3962,
+      "step": 2543
+    },
+    {
+      "epoch": 1.7212449255751014,
+      "grad_norm": 0.9054889673155659,
+      "learning_rate": 4.587103272638339e-06,
+      "loss": 0.3775,
+      "step": 2544
+    },
+    {
+      "epoch": 1.7219215155615697,
+      "grad_norm": 0.9552481048147414,
+      "learning_rate": 4.583180014411642e-06,
+      "loss": 0.3932,
+      "step": 2545
+    },
+    {
+      "epoch": 1.722598105548038,
+      "grad_norm": 0.9731394076224001,
+      "learning_rate": 4.579257014590869e-06,
+      "loss": 0.3829,
+      "step": 2546
+    },
+    {
+      "epoch": 1.723274695534506,
+      "grad_norm": 0.9975616552805409,
+      "learning_rate": 4.575334275608067e-06,
+      "loss": 0.4022,
+      "step": 2547
+    },
+    {
+      "epoch": 1.7239512855209744,
+      "grad_norm": 0.9729225738080103,
+      "learning_rate": 4.571411799895126e-06,
+      "loss": 0.3901,
+      "step": 2548
+    },
+    {
+      "epoch": 1.7246278755074425,
+      "grad_norm": 0.9797995715333138,
+      "learning_rate": 4.567489589883766e-06,
+      "loss": 0.4007,
+      "step": 2549
+    },
+    {
+      "epoch": 1.7253044654939107,
+      "grad_norm": 0.9844888604879379,
+      "learning_rate": 4.563567648005551e-06,
+      "loss": 0.3951,
+      "step": 2550
+    },
+    {
+      "epoch": 1.725981055480379,
+      "grad_norm": 0.9689950884325625,
+      "learning_rate": 4.559645976691868e-06,
+      "loss": 0.369,
+      "step": 2551
+    },
+    {
+      "epoch": 1.726657645466847,
+      "grad_norm": 0.9984196359505926,
+      "learning_rate": 4.5557245783739425e-06,
+      "loss": 0.4027,
+      "step": 2552
+    },
+    {
+      "epoch": 1.7273342354533154,
+      "grad_norm": 0.9805503226918398,
+      "learning_rate": 4.551803455482833e-06,
+      "loss": 0.4208,
+      "step": 2553
+    },
+    {
+      "epoch": 1.7280108254397835,
+      "grad_norm": 1.0077826768104767,
+      "learning_rate": 4.5478826104494225e-06,
+      "loss": 0.4092,
+      "step": 2554
+    },
+    {
+      "epoch": 1.7286874154262517,
+      "grad_norm": 0.9510535599740801,
+      "learning_rate": 4.543962045704424e-06,
+      "loss": 0.3783,
+      "step": 2555
+    },
+    {
+      "epoch": 1.72936400541272,
+      "grad_norm": 0.9902279326458776,
+      "learning_rate": 4.540041763678377e-06,
+      "loss": 0.409,
+      "step": 2556
+    },
+    {
+      "epoch": 1.730040595399188,
+      "grad_norm": 0.9746722082380599,
+      "learning_rate": 4.536121766801645e-06,
+      "loss": 0.3987,
+      "step": 2557
+    },
+    {
+      "epoch": 1.7307171853856564,
+      "grad_norm": 0.9550626556260181,
+      "learning_rate": 4.532202057504412e-06,
+      "loss": 0.4047,
+      "step": 2558
+    },
+    {
+      "epoch": 1.7313937753721245,
+      "grad_norm": 0.9969625364693467,
+      "learning_rate": 4.528282638216689e-06,
+      "loss": 0.4134,
+      "step": 2559
+    },
+    {
+      "epoch": 1.7320703653585927,
+      "grad_norm": 0.9720163552146841,
+      "learning_rate": 4.524363511368304e-06,
+      "loss": 0.3854,
+      "step": 2560
+    },
+    {
+      "epoch": 1.732746955345061,
+      "grad_norm": 0.999014438323783,
+      "learning_rate": 4.520444679388906e-06,
+      "loss": 0.3974,
+      "step": 2561
+    },
+    {
+      "epoch": 1.733423545331529,
+      "grad_norm": 0.953774151841673,
+      "learning_rate": 4.516526144707957e-06,
+      "loss": 0.3794,
+      "step": 2562
+    },
+    {
+      "epoch": 1.7341001353179974,
+      "grad_norm": 0.9454426187396396,
+      "learning_rate": 4.512607909754741e-06,
+      "loss": 0.3867,
+      "step": 2563
+    },
+    {
+      "epoch": 1.7347767253044655,
+      "grad_norm": 0.9992784563132288,
+      "learning_rate": 4.508689976958348e-06,
+      "loss": 0.4067,
+      "step": 2564
+    },
+    {
+      "epoch": 1.7354533152909337,
+      "grad_norm": 0.958222621957998,
+      "learning_rate": 4.504772348747687e-06,
+      "loss": 0.3937,
+      "step": 2565
+    },
+    {
+      "epoch": 1.736129905277402,
+      "grad_norm": 1.0209496378822407,
+      "learning_rate": 4.500855027551477e-06,
+      "loss": 0.4102,
+      "step": 2566
+    },
+    {
+      "epoch": 1.73680649526387,
+      "grad_norm": 0.9895744674110816,
+      "learning_rate": 4.496938015798246e-06,
+      "loss": 0.3941,
+      "step": 2567
+    },
+    {
+      "epoch": 1.7374830852503385,
+      "grad_norm": 0.946289958972408,
+      "learning_rate": 4.493021315916328e-06,
+      "loss": 0.3786,
+      "step": 2568
+    },
+    {
+      "epoch": 1.7381596752368065,
+      "grad_norm": 0.9676733990694013,
+      "learning_rate": 4.48910493033387e-06,
+      "loss": 0.3799,
+      "step": 2569
+    },
+    {
+      "epoch": 1.7388362652232747,
+      "grad_norm": 0.904426391775157,
+      "learning_rate": 4.485188861478817e-06,
+      "loss": 0.3675,
+      "step": 2570
+    },
+    {
+      "epoch": 1.739512855209743,
+      "grad_norm": 0.9186497123107233,
+      "learning_rate": 4.481273111778919e-06,
+      "loss": 0.3873,
+      "step": 2571
+    },
+    {
+      "epoch": 1.740189445196211,
+      "grad_norm": 0.9458210866205528,
+      "learning_rate": 4.477357683661734e-06,
+      "loss": 0.4012,
+      "step": 2572
+    },
+    {
+      "epoch": 1.7408660351826795,
+      "grad_norm": 0.9594112673318085,
+      "learning_rate": 4.473442579554612e-06,
+      "loss": 0.3978,
+      "step": 2573
+    },
+    {
+      "epoch": 1.7415426251691475,
+      "grad_norm": 1.001722672912692,
+      "learning_rate": 4.46952780188471e-06,
+      "loss": 0.4078,
+      "step": 2574
+    },
+    {
+      "epoch": 1.7422192151556157,
+      "grad_norm": 0.9942070997910358,
+      "learning_rate": 4.465613353078978e-06,
+      "loss": 0.4187,
+      "step": 2575
+    },
+    {
+      "epoch": 1.742895805142084,
+      "grad_norm": 0.9257521647079752,
+      "learning_rate": 4.461699235564164e-06,
+      "loss": 0.3893,
+      "step": 2576
+    },
+    {
+      "epoch": 1.743572395128552,
+      "grad_norm": 0.9908152539375658,
+      "learning_rate": 4.457785451766808e-06,
+      "loss": 0.4008,
+      "step": 2577
+    },
+    {
+      "epoch": 1.7442489851150202,
+      "grad_norm": 1.0562975524693239,
+      "learning_rate": 4.453872004113247e-06,
+      "loss": 0.3936,
+      "step": 2578
+    },
+    {
+      "epoch": 1.7449255751014885,
+      "grad_norm": 0.9301518609938227,
+      "learning_rate": 4.449958895029604e-06,
+      "loss": 0.3727,
+      "step": 2579
+    },
+    {
+      "epoch": 1.7456021650879567,
+      "grad_norm": 0.9508606126679935,
+      "learning_rate": 4.446046126941801e-06,
+      "loss": 0.3863,
+      "step": 2580
+    },
+    {
+      "epoch": 1.746278755074425,
+      "grad_norm": 0.9848562022486078,
+      "learning_rate": 4.442133702275539e-06,
+      "loss": 0.4021,
+      "step": 2581
+    },
+    {
+      "epoch": 1.746955345060893,
+      "grad_norm": 0.9995111740140712,
+      "learning_rate": 4.438221623456312e-06,
+      "loss": 0.3918,
+      "step": 2582
+    },
+    {
+      "epoch": 1.7476319350473613,
+      "grad_norm": 1.0064645844396989,
+      "learning_rate": 4.4343098929094e-06,
+      "loss": 0.4127,
+      "step": 2583
+    },
+    {
+      "epoch": 1.7483085250338295,
+      "grad_norm": 0.9872433886775531,
+      "learning_rate": 4.4303985130598615e-06,
+      "loss": 0.3964,
+      "step": 2584
+    },
+    {
+      "epoch": 1.7489851150202977,
+      "grad_norm": 0.9802756173637774,
+      "learning_rate": 4.426487486332544e-06,
+      "loss": 0.4066,
+      "step": 2585
+    },
+    {
+      "epoch": 1.749661705006766,
+      "grad_norm": 0.9557454752543781,
+      "learning_rate": 4.42257681515207e-06,
+      "loss": 0.3957,
+      "step": 2586
+    },
+    {
+      "epoch": 1.750338294993234,
+      "grad_norm": 0.9189226504260087,
+      "learning_rate": 4.4186665019428485e-06,
+      "loss": 0.3829,
+      "step": 2587
+    },
+    {
+      "epoch": 1.7510148849797023,
+      "grad_norm": 0.9879276621095745,
+      "learning_rate": 4.41475654912906e-06,
+      "loss": 0.4164,
+      "step": 2588
+    },
+    {
+      "epoch": 1.7516914749661705,
+      "grad_norm": 0.9525601118399755,
+      "learning_rate": 4.410846959134667e-06,
+      "loss": 0.3764,
+      "step": 2589
+    },
+    {
+      "epoch": 1.7523680649526387,
+      "grad_norm": 1.0173347974012619,
+      "learning_rate": 4.406937734383405e-06,
+      "loss": 0.4016,
+      "step": 2590
+    },
+    {
+      "epoch": 1.753044654939107,
+      "grad_norm": 0.9863231506254914,
+      "learning_rate": 4.4030288772987795e-06,
+      "loss": 0.3904,
+      "step": 2591
+    },
+    {
+      "epoch": 1.753721244925575,
+      "grad_norm": 0.9614479440117635,
+      "learning_rate": 4.399120390304072e-06,
+      "loss": 0.3959,
+      "step": 2592
+    },
+    {
+      "epoch": 1.7543978349120433,
+      "grad_norm": 0.9263515045538261,
+      "learning_rate": 4.395212275822336e-06,
+      "loss": 0.3921,
+      "step": 2593
+    },
+    {
+      "epoch": 1.7550744248985115,
+      "grad_norm": 0.9319742115582491,
+      "learning_rate": 4.391304536276389e-06,
+      "loss": 0.3662,
+      "step": 2594
+    },
+    {
+      "epoch": 1.7557510148849798,
+      "grad_norm": 0.9971421392739953,
+      "learning_rate": 4.3873971740888205e-06,
+      "loss": 0.39,
+      "step": 2595
+    },
+    {
+      "epoch": 1.756427604871448,
+      "grad_norm": 0.9510498106724605,
+      "learning_rate": 4.383490191681985e-06,
+      "loss": 0.3749,
+      "step": 2596
+    },
+    {
+      "epoch": 1.757104194857916,
+      "grad_norm": 1.0002185766225264,
+      "learning_rate": 4.379583591477999e-06,
+      "loss": 0.4079,
+      "step": 2597
+    },
+    {
+      "epoch": 1.7577807848443843,
+      "grad_norm": 1.0015549185137866,
+      "learning_rate": 4.375677375898746e-06,
+      "loss": 0.4016,
+      "step": 2598
+    },
+    {
+      "epoch": 1.7584573748308525,
+      "grad_norm": 1.0650451438844017,
+      "learning_rate": 4.371771547365869e-06,
+      "loss": 0.4166,
+      "step": 2599
+    },
+    {
+      "epoch": 1.7591339648173205,
+      "grad_norm": 1.026298144196455,
+      "learning_rate": 4.367866108300769e-06,
+      "loss": 0.4091,
+      "step": 2600
+    },
+    {
+      "epoch": 1.759810554803789,
+      "grad_norm": 0.9603534848580563,
+      "learning_rate": 4.3639610611246106e-06,
+      "loss": 0.3858,
+      "step": 2601
+    },
+    {
+      "epoch": 1.760487144790257,
+      "grad_norm": 0.949781912855817,
+      "learning_rate": 4.36005640825831e-06,
+      "loss": 0.3845,
+      "step": 2602
+    },
+    {
+      "epoch": 1.7611637347767253,
+      "grad_norm": 0.9796734545820398,
+      "learning_rate": 4.3561521521225445e-06,
+      "loss": 0.388,
+      "step": 2603
+    },
+    {
+      "epoch": 1.7618403247631935,
+      "grad_norm": 0.966494254091051,
+      "learning_rate": 4.352248295137739e-06,
+      "loss": 0.4071,
+      "step": 2604
+    },
+    {
+      "epoch": 1.7625169147496615,
+      "grad_norm": 0.9506188669527146,
+      "learning_rate": 4.348344839724076e-06,
+      "loss": 0.4017,
+      "step": 2605
+    },
+    {
+      "epoch": 1.76319350473613,
+      "grad_norm": 0.94332338237239,
+      "learning_rate": 4.3444417883014885e-06,
+      "loss": 0.3781,
+      "step": 2606
+    },
+    {
+      "epoch": 1.763870094722598,
+      "grad_norm": 0.9593292089830917,
+      "learning_rate": 4.340539143289655e-06,
+      "loss": 0.3844,
+      "step": 2607
+    },
+    {
+      "epoch": 1.7645466847090663,
+      "grad_norm": 0.9654001859757949,
+      "learning_rate": 4.33663690710801e-06,
+      "loss": 0.3958,
+      "step": 2608
+    },
+    {
+      "epoch": 1.7652232746955345,
+      "grad_norm": 0.9567085243098365,
+      "learning_rate": 4.332735082175724e-06,
+      "loss": 0.3908,
+      "step": 2609
+    },
+    {
+      "epoch": 1.7658998646820026,
+      "grad_norm": 1.0020124880525525,
+      "learning_rate": 4.3288336709117246e-06,
+      "loss": 0.4193,
+      "step": 2610
+    },
+    {
+      "epoch": 1.766576454668471,
+      "grad_norm": 0.9887964853614057,
+      "learning_rate": 4.32493267573467e-06,
+      "loss": 0.391,
+      "step": 2611
+    },
+    {
+      "epoch": 1.767253044654939,
+      "grad_norm": 0.985615480192359,
+      "learning_rate": 4.3210320990629696e-06,
+      "loss": 0.3904,
+      "step": 2612
+    },
+    {
+      "epoch": 1.7679296346414073,
+      "grad_norm": 0.9480404220784492,
+      "learning_rate": 4.31713194331477e-06,
+      "loss": 0.3927,
+      "step": 2613
+    },
+    {
+      "epoch": 1.7686062246278755,
+      "grad_norm": 0.9404850542944785,
+      "learning_rate": 4.313232210907959e-06,
+      "loss": 0.3786,
+      "step": 2614
+    },
+    {
+      "epoch": 1.7692828146143436,
+      "grad_norm": 0.9429814884624229,
+      "learning_rate": 4.30933290426016e-06,
+      "loss": 0.378,
+      "step": 2615
+    },
+    {
+      "epoch": 1.769959404600812,
+      "grad_norm": 0.9777368455535926,
+      "learning_rate": 4.305434025788735e-06,
+      "loss": 0.4096,
+      "step": 2616
+    },
+    {
+      "epoch": 1.77063599458728,
+      "grad_norm": 1.0115823585875652,
+      "learning_rate": 4.301535577910774e-06,
+      "loss": 0.4199,
+      "step": 2617
+    },
+    {
+      "epoch": 1.7713125845737483,
+      "grad_norm": 0.9776174108039173,
+      "learning_rate": 4.297637563043106e-06,
+      "loss": 0.4014,
+      "step": 2618
+    },
+    {
+      "epoch": 1.7719891745602165,
+      "grad_norm": 0.9286579482031964,
+      "learning_rate": 4.293739983602292e-06,
+      "loss": 0.3726,
+      "step": 2619
+    },
+    {
+      "epoch": 1.7726657645466846,
+      "grad_norm": 0.9537553847089697,
+      "learning_rate": 4.28984284200462e-06,
+      "loss": 0.3854,
+      "step": 2620
+    },
+    {
+      "epoch": 1.773342354533153,
+      "grad_norm": 0.9546262623135069,
+      "learning_rate": 4.285946140666107e-06,
+      "loss": 0.4035,
+      "step": 2621
+    },
+    {
+      "epoch": 1.774018944519621,
+      "grad_norm": 1.0004135073921654,
+      "learning_rate": 4.282049882002499e-06,
+      "loss": 0.3859,
+      "step": 2622
+    },
+    {
+      "epoch": 1.7746955345060893,
+      "grad_norm": 1.0216652599147376,
+      "learning_rate": 4.278154068429268e-06,
+      "loss": 0.4145,
+      "step": 2623
+    },
+    {
+      "epoch": 1.7753721244925575,
+      "grad_norm": 0.9577692175618054,
+      "learning_rate": 4.274258702361604e-06,
+      "loss": 0.3813,
+      "step": 2624
+    },
+    {
+      "epoch": 1.7760487144790256,
+      "grad_norm": 0.9802380852400819,
+      "learning_rate": 4.270363786214427e-06,
+      "loss": 0.4077,
+      "step": 2625
+    },
+    {
+      "epoch": 1.776725304465494,
+      "grad_norm": 0.9665671248211338,
+      "learning_rate": 4.266469322402374e-06,
+      "loss": 0.3959,
+      "step": 2626
+    },
+    {
+      "epoch": 1.777401894451962,
+      "grad_norm": 0.9309997043303668,
+      "learning_rate": 4.2625753133398036e-06,
+      "loss": 0.3792,
+      "step": 2627
+    },
+    {
+      "epoch": 1.7780784844384303,
+      "grad_norm": 0.9657798512939555,
+      "learning_rate": 4.25868176144079e-06,
+      "loss": 0.3952,
+      "step": 2628
+    },
+    {
+      "epoch": 1.7787550744248986,
+      "grad_norm": 0.9700576818111204,
+      "learning_rate": 4.254788669119127e-06,
+      "loss": 0.3886,
+      "step": 2629
+    },
+    {
+      "epoch": 1.7794316644113666,
+      "grad_norm": 1.03212159402125,
+      "learning_rate": 4.250896038788324e-06,
+      "loss": 0.4208,
+      "step": 2630
+    },
+    {
+      "epoch": 1.780108254397835,
+      "grad_norm": 0.9616286968417425,
+      "learning_rate": 4.247003872861598e-06,
+      "loss": 0.3997,
+      "step": 2631
+    },
+    {
+      "epoch": 1.780784844384303,
+      "grad_norm": 0.9258081970944867,
+      "learning_rate": 4.2431121737518824e-06,
+      "loss": 0.3705,
+      "step": 2632
+    },
+    {
+      "epoch": 1.7814614343707713,
+      "grad_norm": 0.9676995922370372,
+      "learning_rate": 4.239220943871823e-06,
+      "loss": 0.3885,
+      "step": 2633
+    },
+    {
+      "epoch": 1.7821380243572396,
+      "grad_norm": 0.9676763528747013,
+      "learning_rate": 4.23533018563377e-06,
+      "loss": 0.3937,
+      "step": 2634
+    },
+    {
+      "epoch": 1.7828146143437076,
+      "grad_norm": 0.9807034187447221,
+      "learning_rate": 4.231439901449788e-06,
+      "loss": 0.3983,
+      "step": 2635
+    },
+    {
+      "epoch": 1.783491204330176,
+      "grad_norm": 0.948920942161716,
+      "learning_rate": 4.227550093731641e-06,
+      "loss": 0.3936,
+      "step": 2636
+    },
+    {
+      "epoch": 1.784167794316644,
+      "grad_norm": 0.9647084658527924,
+      "learning_rate": 4.223660764890799e-06,
+      "loss": 0.3724,
+      "step": 2637
+    },
+    {
+      "epoch": 1.7848443843031123,
+      "grad_norm": 0.9664239389181164,
+      "learning_rate": 4.2197719173384374e-06,
+      "loss": 0.3955,
+      "step": 2638
+    },
+    {
+      "epoch": 1.7855209742895806,
+      "grad_norm": 0.9278347339777552,
+      "learning_rate": 4.215883553485431e-06,
+      "loss": 0.392,
+      "step": 2639
+    },
+    {
+      "epoch": 1.7861975642760486,
+      "grad_norm": 1.0060101602339717,
+      "learning_rate": 4.211995675742358e-06,
+      "loss": 0.4117,
+      "step": 2640
+    },
+    {
+      "epoch": 1.786874154262517,
+      "grad_norm": 1.0135331058750892,
+      "learning_rate": 4.208108286519491e-06,
+      "loss": 0.4075,
+      "step": 2641
+    },
+    {
+      "epoch": 1.787550744248985,
+      "grad_norm": 1.0396604216111962,
+      "learning_rate": 4.204221388226803e-06,
+      "loss": 0.4108,
+      "step": 2642
+    },
+    {
+      "epoch": 1.7882273342354533,
+      "grad_norm": 0.9397392360088378,
+      "learning_rate": 4.2003349832739624e-06,
+      "loss": 0.3936,
+      "step": 2643
+    },
+    {
+      "epoch": 1.7889039242219216,
+      "grad_norm": 0.9390935068668829,
+      "learning_rate": 4.196449074070329e-06,
+      "loss": 0.3768,
+      "step": 2644
+    },
+    {
+      "epoch": 1.7895805142083896,
+      "grad_norm": 0.961629240316789,
+      "learning_rate": 4.1925636630249565e-06,
+      "loss": 0.3868,
+      "step": 2645
+    },
+    {
+      "epoch": 1.790257104194858,
+      "grad_norm": 0.9693878022246628,
+      "learning_rate": 4.1886787525465914e-06,
+      "loss": 0.3923,
+      "step": 2646
+    },
+    {
+      "epoch": 1.790933694181326,
+      "grad_norm": 0.9788227794864665,
+      "learning_rate": 4.184794345043668e-06,
+      "loss": 0.3955,
+      "step": 2647
+    },
+    {
+      "epoch": 1.7916102841677943,
+      "grad_norm": 1.0285734374589028,
+      "learning_rate": 4.180910442924312e-06,
+      "loss": 0.4248,
+      "step": 2648
+    },
+    {
+      "epoch": 1.7922868741542626,
+      "grad_norm": 0.9516476002575738,
+      "learning_rate": 4.17702704859633e-06,
+      "loss": 0.3742,
+      "step": 2649
+    },
+    {
+      "epoch": 1.7929634641407306,
+      "grad_norm": 0.9234015190624797,
+      "learning_rate": 4.173144164467221e-06,
+      "loss": 0.3924,
+      "step": 2650
+    },
+    {
+      "epoch": 1.793640054127199,
+      "grad_norm": 1.006276307711603,
+      "learning_rate": 4.169261792944161e-06,
+      "loss": 0.4126,
+      "step": 2651
+    },
+    {
+      "epoch": 1.794316644113667,
+      "grad_norm": 0.9489648526139915,
+      "learning_rate": 4.165379936434011e-06,
+      "loss": 0.3786,
+      "step": 2652
+    },
+    {
+      "epoch": 1.7949932341001353,
+      "grad_norm": 0.9669686659997468,
+      "learning_rate": 4.161498597343313e-06,
+      "loss": 0.4056,
+      "step": 2653
+    },
+    {
+      "epoch": 1.7956698240866036,
+      "grad_norm": 0.9658512544044849,
+      "learning_rate": 4.15761777807829e-06,
+      "loss": 0.3727,
+      "step": 2654
+    },
+    {
+      "epoch": 1.7963464140730716,
+      "grad_norm": 0.9627151248354098,
+      "learning_rate": 4.153737481044838e-06,
+      "loss": 0.3811,
+      "step": 2655
+    },
+    {
+      "epoch": 1.79702300405954,
+      "grad_norm": 0.9926586492534942,
+      "learning_rate": 4.149857708648536e-06,
+      "loss": 0.4075,
+      "step": 2656
+    },
+    {
+      "epoch": 1.797699594046008,
+      "grad_norm": 0.9776992890715819,
+      "learning_rate": 4.1459784632946295e-06,
+      "loss": 0.3965,
+      "step": 2657
+    },
+    {
+      "epoch": 1.7983761840324763,
+      "grad_norm": 0.9615659408473649,
+      "learning_rate": 4.142099747388042e-06,
+      "loss": 0.4024,
+      "step": 2658
+    },
+    {
+      "epoch": 1.7990527740189446,
+      "grad_norm": 0.98622655383633,
+      "learning_rate": 4.138221563333371e-06,
+      "loss": 0.3889,
+      "step": 2659
+    },
+    {
+      "epoch": 1.7997293640054126,
+      "grad_norm": 0.9563767130211177,
+      "learning_rate": 4.134343913534879e-06,
+      "loss": 0.386,
+      "step": 2660
+    },
+    {
+      "epoch": 1.800405953991881,
+      "grad_norm": 0.9234774059435488,
+      "learning_rate": 4.1304668003965016e-06,
+      "loss": 0.3831,
+      "step": 2661
+    },
+    {
+      "epoch": 1.801082543978349,
+      "grad_norm": 1.0327608879236543,
+      "learning_rate": 4.126590226321838e-06,
+      "loss": 0.4124,
+      "step": 2662
+    },
+    {
+      "epoch": 1.8017591339648173,
+      "grad_norm": 0.9670260073711607,
+      "learning_rate": 4.12271419371416e-06,
+      "loss": 0.3965,
+      "step": 2663
+    },
+    {
+      "epoch": 1.8024357239512856,
+      "grad_norm": 0.9812009881153325,
+      "learning_rate": 4.118838704976392e-06,
+      "loss": 0.3781,
+      "step": 2664
+    },
+    {
+      "epoch": 1.8031123139377536,
+      "grad_norm": 0.9201699169620586,
+      "learning_rate": 4.114963762511134e-06,
+      "loss": 0.3799,
+      "step": 2665
+    },
+    {
+      "epoch": 1.803788903924222,
+      "grad_norm": 0.9953322972521901,
+      "learning_rate": 4.111089368720635e-06,
+      "loss": 0.4127,
+      "step": 2666
+    },
+    {
+      "epoch": 1.80446549391069,
+      "grad_norm": 0.9507164657537404,
+      "learning_rate": 4.107215526006818e-06,
+      "loss": 0.3826,
+      "step": 2667
+    },
+    {
+      "epoch": 1.8051420838971584,
+      "grad_norm": 0.9893140146578959,
+      "learning_rate": 4.10334223677125e-06,
+      "loss": 0.4072,
+      "step": 2668
+    },
+    {
+      "epoch": 1.8058186738836266,
+      "grad_norm": 0.9879521587844674,
+      "learning_rate": 4.099469503415167e-06,
+      "loss": 0.4058,
+      "step": 2669
+    },
+    {
+      "epoch": 1.8064952638700946,
+      "grad_norm": 0.9442173163374029,
+      "learning_rate": 4.0955973283394525e-06,
+      "loss": 0.3842,
+      "step": 2670
+    },
+    {
+      "epoch": 1.8071718538565629,
+      "grad_norm": 1.0009768172003903,
+      "learning_rate": 4.091725713944644e-06,
+      "loss": 0.4018,
+      "step": 2671
+    },
+    {
+      "epoch": 1.8078484438430311,
+      "grad_norm": 0.9331511419632416,
+      "learning_rate": 4.087854662630937e-06,
+      "loss": 0.3717,
+      "step": 2672
+    },
+    {
+      "epoch": 1.8085250338294994,
+      "grad_norm": 0.986564902343268,
+      "learning_rate": 4.083984176798175e-06,
+      "loss": 0.3869,
+      "step": 2673
+    },
+    {
+      "epoch": 1.8092016238159676,
+      "grad_norm": 0.9985087496075559,
+      "learning_rate": 4.080114258845846e-06,
+      "loss": 0.4135,
+      "step": 2674
+    },
+    {
+      "epoch": 1.8098782138024356,
+      "grad_norm": 0.9826142874991449,
+      "learning_rate": 4.076244911173097e-06,
+      "loss": 0.4002,
+      "step": 2675
+    },
+    {
+      "epoch": 1.8105548037889039,
+      "grad_norm": 1.0101246168508613,
+      "learning_rate": 4.072376136178712e-06,
+      "loss": 0.4208,
+      "step": 2676
+    },
+    {
+      "epoch": 1.8112313937753721,
+      "grad_norm": 0.982051469432822,
+      "learning_rate": 4.06850793626112e-06,
+      "loss": 0.3806,
+      "step": 2677
+    },
+    {
+      "epoch": 1.8119079837618404,
+      "grad_norm": 1.0006965108244374,
+      "learning_rate": 4.064640313818401e-06,
+      "loss": 0.4249,
+      "step": 2678
+    },
+    {
+      "epoch": 1.8125845737483086,
+      "grad_norm": 1.02113807045358,
+      "learning_rate": 4.06077327124827e-06,
+      "loss": 0.3953,
+      "step": 2679
+    },
+    {
+      "epoch": 1.8132611637347766,
+      "grad_norm": 0.9188533856662646,
+      "learning_rate": 4.056906810948086e-06,
+      "loss": 0.3844,
+      "step": 2680
+    },
+    {
+      "epoch": 1.8139377537212449,
+      "grad_norm": 0.9823063149835631,
+      "learning_rate": 4.053040935314845e-06,
+      "loss": 0.392,
+      "step": 2681
+    },
+    {
+      "epoch": 1.8146143437077131,
+      "grad_norm": 0.9586090895924663,
+      "learning_rate": 4.049175646745182e-06,
+      "loss": 0.4027,
+      "step": 2682
+    },
+    {
+      "epoch": 1.8152909336941814,
+      "grad_norm": 0.9210304149156828,
+      "learning_rate": 4.045310947635369e-06,
+      "loss": 0.3691,
+      "step": 2683
+    },
+    {
+      "epoch": 1.8159675236806496,
+      "grad_norm": 1.004792893723896,
+      "learning_rate": 4.041446840381309e-06,
+      "loss": 0.3948,
+      "step": 2684
+    },
+    {
+      "epoch": 1.8166441136671176,
+      "grad_norm": 0.9640420720464306,
+      "learning_rate": 4.03758332737854e-06,
+      "loss": 0.3885,
+      "step": 2685
+    },
+    {
+      "epoch": 1.817320703653586,
+      "grad_norm": 0.9761056105847574,
+      "learning_rate": 4.033720411022235e-06,
+      "loss": 0.4018,
+      "step": 2686
+    },
+    {
+      "epoch": 1.8179972936400541,
+      "grad_norm": 0.9940739219726901,
+      "learning_rate": 4.02985809370719e-06,
+      "loss": 0.399,
+      "step": 2687
+    },
+    {
+      "epoch": 1.8186738836265224,
+      "grad_norm": 0.9297981262839745,
+      "learning_rate": 4.025996377827836e-06,
+      "loss": 0.3831,
+      "step": 2688
+    },
+    {
+      "epoch": 1.8193504736129906,
+      "grad_norm": 0.9538154269143837,
+      "learning_rate": 4.022135265778226e-06,
+      "loss": 0.379,
+      "step": 2689
+    },
+    {
+      "epoch": 1.8200270635994586,
+      "grad_norm": 0.9263921989936219,
+      "learning_rate": 4.018274759952047e-06,
+      "loss": 0.4042,
+      "step": 2690
+    },
+    {
+      "epoch": 1.820703653585927,
+      "grad_norm": 0.977585672815,
+      "learning_rate": 4.0144148627426e-06,
+      "loss": 0.3972,
+      "step": 2691
+    },
+    {
+      "epoch": 1.8213802435723951,
+      "grad_norm": 0.9605042438747269,
+      "learning_rate": 4.010555576542812e-06,
+      "loss": 0.4013,
+      "step": 2692
+    },
+    {
+      "epoch": 1.8220568335588632,
+      "grad_norm": 0.9462577925101531,
+      "learning_rate": 4.006696903745236e-06,
+      "loss": 0.3864,
+      "step": 2693
+    },
+    {
+      "epoch": 1.8227334235453316,
+      "grad_norm": 0.9587643452720432,
+      "learning_rate": 4.002838846742039e-06,
+      "loss": 0.3914,
+      "step": 2694
+    },
+    {
+      "epoch": 1.8234100135317997,
+      "grad_norm": 0.9498816356715988,
+      "learning_rate": 3.998981407925009e-06,
+      "loss": 0.3879,
+      "step": 2695
+    },
+    {
+      "epoch": 1.824086603518268,
+      "grad_norm": 0.920748654494678,
+      "learning_rate": 3.995124589685552e-06,
+      "loss": 0.3731,
+      "step": 2696
+    },
+    {
+      "epoch": 1.8247631935047361,
+      "grad_norm": 0.9648858745318133,
+      "learning_rate": 3.991268394414685e-06,
+      "loss": 0.3941,
+      "step": 2697
+    },
+    {
+      "epoch": 1.8254397834912042,
+      "grad_norm": 0.9301512330105254,
+      "learning_rate": 3.987412824503041e-06,
+      "loss": 0.3734,
+      "step": 2698
+    },
+    {
+      "epoch": 1.8261163734776726,
+      "grad_norm": 0.9844036475247598,
+      "learning_rate": 3.983557882340866e-06,
+      "loss": 0.3926,
+      "step": 2699
+    },
+    {
+      "epoch": 1.8267929634641407,
+      "grad_norm": 0.9906461597560315,
+      "learning_rate": 3.979703570318017e-06,
+      "loss": 0.4173,
+      "step": 2700
+    },
+    {
+      "epoch": 1.827469553450609,
+      "grad_norm": 0.9505497224356612,
+      "learning_rate": 3.97584989082396e-06,
+      "loss": 0.3916,
+      "step": 2701
+    },
+    {
+      "epoch": 1.8281461434370772,
+      "grad_norm": 0.9359231694932668,
+      "learning_rate": 3.971996846247767e-06,
+      "loss": 0.3939,
+      "step": 2702
+    },
+    {
+      "epoch": 1.8288227334235452,
+      "grad_norm": 0.9399633161213121,
+      "learning_rate": 3.968144438978121e-06,
+      "loss": 0.3742,
+      "step": 2703
+    },
+    {
+      "epoch": 1.8294993234100136,
+      "grad_norm": 0.9549451444962479,
+      "learning_rate": 3.964292671403303e-06,
+      "loss": 0.4075,
+      "step": 2704
+    },
+    {
+      "epoch": 1.8301759133964817,
+      "grad_norm": 0.9618288042986526,
+      "learning_rate": 3.960441545911205e-06,
+      "loss": 0.3884,
+      "step": 2705
+    },
+    {
+      "epoch": 1.83085250338295,
+      "grad_norm": 0.9383453073080402,
+      "learning_rate": 3.956591064889313e-06,
+      "loss": 0.3832,
+      "step": 2706
+    },
+    {
+      "epoch": 1.8315290933694182,
+      "grad_norm": 0.9367145772648151,
+      "learning_rate": 3.952741230724721e-06,
+      "loss": 0.3816,
+      "step": 2707
+    },
+    {
+      "epoch": 1.8322056833558862,
+      "grad_norm": 0.9654028112278012,
+      "learning_rate": 3.948892045804117e-06,
+      "loss": 0.3713,
+      "step": 2708
+    },
+    {
+      "epoch": 1.8328822733423547,
+      "grad_norm": 0.9957421271193522,
+      "learning_rate": 3.94504351251379e-06,
+      "loss": 0.3944,
+      "step": 2709
+    },
+    {
+      "epoch": 1.8335588633288227,
+      "grad_norm": 0.9391522691304137,
+      "learning_rate": 3.9411956332396224e-06,
+      "loss": 0.3771,
+      "step": 2710
+    },
+    {
+      "epoch": 1.834235453315291,
+      "grad_norm": 0.9736197204389111,
+      "learning_rate": 3.937348410367091e-06,
+      "loss": 0.3816,
+      "step": 2711
+    },
+    {
+      "epoch": 1.8349120433017592,
+      "grad_norm": 1.0525202114093986,
+      "learning_rate": 3.9335018462812664e-06,
+      "loss": 0.4316,
+      "step": 2712
+    },
+    {
+      "epoch": 1.8355886332882272,
+      "grad_norm": 0.9756580087226838,
+      "learning_rate": 3.929655943366812e-06,
+      "loss": 0.3857,
+      "step": 2713
+    },
+    {
+      "epoch": 1.8362652232746957,
+      "grad_norm": 0.9336004222503018,
+      "learning_rate": 3.92581070400798e-06,
+      "loss": 0.3816,
+      "step": 2714
+    },
+    {
+      "epoch": 1.8369418132611637,
+      "grad_norm": 0.9880252843245524,
+      "learning_rate": 3.921966130588612e-06,
+      "loss": 0.3949,
+      "step": 2715
+    },
+    {
+      "epoch": 1.837618403247632,
+      "grad_norm": 0.950446373911663,
+      "learning_rate": 3.918122225492139e-06,
+      "loss": 0.3997,
+      "step": 2716
+    },
+    {
+      "epoch": 1.8382949932341002,
+      "grad_norm": 0.9621395005451292,
+      "learning_rate": 3.914278991101568e-06,
+      "loss": 0.393,
+      "step": 2717
+    },
+    {
+      "epoch": 1.8389715832205682,
+      "grad_norm": 0.9824936681078674,
+      "learning_rate": 3.910436429799503e-06,
+      "loss": 0.3995,
+      "step": 2718
+    },
+    {
+      "epoch": 1.8396481732070367,
+      "grad_norm": 0.9946873820346164,
+      "learning_rate": 3.906594543968122e-06,
+      "loss": 0.3998,
+      "step": 2719
+    },
+    {
+      "epoch": 1.8403247631935047,
+      "grad_norm": 0.9509829767988345,
+      "learning_rate": 3.902753335989188e-06,
+      "loss": 0.389,
+      "step": 2720
+    },
+    {
+      "epoch": 1.841001353179973,
+      "grad_norm": 0.9461778708114417,
+      "learning_rate": 3.898912808244043e-06,
+      "loss": 0.3846,
+      "step": 2721
+    },
+    {
+      "epoch": 1.8416779431664412,
+      "grad_norm": 0.9334507205472801,
+      "learning_rate": 3.895072963113607e-06,
+      "loss": 0.3867,
+      "step": 2722
+    },
+    {
+      "epoch": 1.8423545331529092,
+      "grad_norm": 0.9432282296290785,
+      "learning_rate": 3.89123380297838e-06,
+      "loss": 0.3915,
+      "step": 2723
+    },
+    {
+      "epoch": 1.8430311231393777,
+      "grad_norm": 0.9606854252233376,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.3812,
+      "step": 2724
+    },
+    {
+      "epoch": 1.8437077131258457,
+      "grad_norm": 0.9703721196340805,
+      "learning_rate": 3.883557547213404e-06,
+      "loss": 0.3897,
+      "step": 2725
+    },
+    {
+      "epoch": 1.844384303112314,
+      "grad_norm": 0.9642040443018232,
+      "learning_rate": 3.8797204563425215e-06,
+      "loss": 0.3937,
+      "step": 2726
+    },
+    {
+      "epoch": 1.8450608930987822,
+      "grad_norm": 0.9882315760969734,
+      "learning_rate": 3.875884059984571e-06,
+      "loss": 0.3948,
+      "step": 2727
+    },
+    {
+      "epoch": 1.8457374830852502,
+      "grad_norm": 0.9661486897118964,
+      "learning_rate": 3.872048360517914e-06,
+      "loss": 0.3937,
+      "step": 2728
+    },
+    {
+      "epoch": 1.8464140730717187,
+      "grad_norm": 0.9408857728057767,
+      "learning_rate": 3.868213360320474e-06,
+      "loss": 0.3589,
+      "step": 2729
+    },
+    {
+      "epoch": 1.8470906630581867,
+      "grad_norm": 0.9498115251523825,
+      "learning_rate": 3.864379061769749e-06,
+      "loss": 0.3772,
+      "step": 2730
+    },
+    {
+      "epoch": 1.847767253044655,
+      "grad_norm": 0.9391010920454833,
+      "learning_rate": 3.860545467242793e-06,
+      "loss": 0.3814,
+      "step": 2731
+    },
+    {
+      "epoch": 1.8484438430311232,
+      "grad_norm": 0.9602026878812513,
+      "learning_rate": 3.856712579116229e-06,
+      "loss": 0.3806,
+      "step": 2732
+    },
+    {
+      "epoch": 1.8491204330175912,
+      "grad_norm": 1.017718591836699,
+      "learning_rate": 3.852880399766243e-06,
+      "loss": 0.4049,
+      "step": 2733
+    },
+    {
+      "epoch": 1.8497970230040597,
+      "grad_norm": 0.9905362353007553,
+      "learning_rate": 3.8490489315685764e-06,
+      "loss": 0.3944,
+      "step": 2734
+    },
+    {
+      "epoch": 1.8504736129905277,
+      "grad_norm": 0.990721430207882,
+      "learning_rate": 3.845218176898537e-06,
+      "loss": 0.4101,
+      "step": 2735
+    },
+    {
+      "epoch": 1.851150202976996,
+      "grad_norm": 0.9000359867998304,
+      "learning_rate": 3.8413881381309845e-06,
+      "loss": 0.3726,
+      "step": 2736
+    },
+    {
+      "epoch": 1.8518267929634642,
+      "grad_norm": 0.9844985020788298,
+      "learning_rate": 3.837558817640334e-06,
+      "loss": 0.3995,
+      "step": 2737
+    },
+    {
+      "epoch": 1.8525033829499322,
+      "grad_norm": 0.896020116602737,
+      "learning_rate": 3.8337302178005605e-06,
+      "loss": 0.3756,
+      "step": 2738
+    },
+    {
+      "epoch": 1.8531799729364007,
+      "grad_norm": 0.9453510728604818,
+      "learning_rate": 3.829902340985189e-06,
+      "loss": 0.3845,
+      "step": 2739
+    },
+    {
+      "epoch": 1.8538565629228687,
+      "grad_norm": 1.0149339322894193,
+      "learning_rate": 3.826075189567296e-06,
+      "loss": 0.3954,
+      "step": 2740
+    },
+    {
+      "epoch": 1.854533152909337,
+      "grad_norm": 0.9369259685457049,
+      "learning_rate": 3.82224876591951e-06,
+      "loss": 0.3915,
+      "step": 2741
+    },
+    {
+      "epoch": 1.8552097428958052,
+      "grad_norm": 0.9924464907482013,
+      "learning_rate": 3.818423072414007e-06,
+      "loss": 0.4011,
+      "step": 2742
+    },
+    {
+      "epoch": 1.8558863328822732,
+      "grad_norm": 0.9844163643599041,
+      "learning_rate": 3.8145981114225135e-06,
+      "loss": 0.3947,
+      "step": 2743
+    },
+    {
+      "epoch": 1.8565629228687417,
+      "grad_norm": 0.990797930854464,
+      "learning_rate": 3.8107738853162953e-06,
+      "loss": 0.3878,
+      "step": 2744
+    },
+    {
+      "epoch": 1.8572395128552097,
+      "grad_norm": 0.9848806184867611,
+      "learning_rate": 3.8069503964661656e-06,
+      "loss": 0.3891,
+      "step": 2745
+    },
+    {
+      "epoch": 1.857916102841678,
+      "grad_norm": 0.9747597649510048,
+      "learning_rate": 3.803127647242486e-06,
+      "loss": 0.3919,
+      "step": 2746
+    },
+    {
+      "epoch": 1.8585926928281462,
+      "grad_norm": 0.9035277076684101,
+      "learning_rate": 3.7993056400151516e-06,
+      "loss": 0.3629,
+      "step": 2747
+    },
+    {
+      "epoch": 1.8592692828146142,
+      "grad_norm": 0.9818462370713226,
+      "learning_rate": 3.795484377153601e-06,
+      "loss": 0.3951,
+      "step": 2748
+    },
+    {
+      "epoch": 1.8599458728010827,
+      "grad_norm": 0.9477913833635639,
+      "learning_rate": 3.791663861026814e-06,
+      "loss": 0.3945,
+      "step": 2749
+    },
+    {
+      "epoch": 1.8606224627875507,
+      "grad_norm": 0.9145189752205615,
+      "learning_rate": 3.787844094003302e-06,
+      "loss": 0.3741,
+      "step": 2750
+    },
+    {
+      "epoch": 1.861299052774019,
+      "grad_norm": 0.9236608243752011,
+      "learning_rate": 3.7840250784511147e-06,
+      "loss": 0.387,
+      "step": 2751
+    },
+    {
+      "epoch": 1.8619756427604872,
+      "grad_norm": 0.9938022598784646,
+      "learning_rate": 3.780206816737837e-06,
+      "loss": 0.4062,
+      "step": 2752
+    },
+    {
+      "epoch": 1.8626522327469552,
+      "grad_norm": 0.9461125538937857,
+      "learning_rate": 3.776389311230584e-06,
+      "loss": 0.3759,
+      "step": 2753
+    },
+    {
+      "epoch": 1.8633288227334237,
+      "grad_norm": 0.9787522934889943,
+      "learning_rate": 3.7725725642960047e-06,
+      "loss": 0.3842,
+      "step": 2754
+    },
+    {
+      "epoch": 1.8640054127198917,
+      "grad_norm": 0.9653692368265662,
+      "learning_rate": 3.7687565783002754e-06,
+      "loss": 0.3947,
+      "step": 2755
+    },
+    {
+      "epoch": 1.86468200270636,
+      "grad_norm": 1.0032055690109953,
+      "learning_rate": 3.7649413556091047e-06,
+      "loss": 0.3941,
+      "step": 2756
+    },
+    {
+      "epoch": 1.8653585926928282,
+      "grad_norm": 0.980333510274981,
+      "learning_rate": 3.7611268985877213e-06,
+      "loss": 0.3968,
+      "step": 2757
+    },
+    {
+      "epoch": 1.8660351826792962,
+      "grad_norm": 0.9941678857876616,
+      "learning_rate": 3.7573132096008843e-06,
+      "loss": 0.3887,
+      "step": 2758
+    },
+    {
+      "epoch": 1.8667117726657647,
+      "grad_norm": 0.9909540797694224,
+      "learning_rate": 3.753500291012874e-06,
+      "loss": 0.3804,
+      "step": 2759
+    },
+    {
+      "epoch": 1.8673883626522327,
+      "grad_norm": 0.9381334042245338,
+      "learning_rate": 3.749688145187497e-06,
+      "loss": 0.3924,
+      "step": 2760
+    },
+    {
+      "epoch": 1.868064952638701,
+      "grad_norm": 1.0047585975417583,
+      "learning_rate": 3.7458767744880763e-06,
+      "loss": 0.3909,
+      "step": 2761
+    },
+    {
+      "epoch": 1.8687415426251692,
+      "grad_norm": 0.9883657107026129,
+      "learning_rate": 3.7420661812774577e-06,
+      "loss": 0.3899,
+      "step": 2762
+    },
+    {
+      "epoch": 1.8694181326116373,
+      "grad_norm": 1.0175792405139148,
+      "learning_rate": 3.738256367918004e-06,
+      "loss": 0.3928,
+      "step": 2763
+    },
+    {
+      "epoch": 1.8700947225981055,
+      "grad_norm": 0.9411438913245966,
+      "learning_rate": 3.734447336771591e-06,
+      "loss": 0.3717,
+      "step": 2764
+    },
+    {
+      "epoch": 1.8707713125845737,
+      "grad_norm": 0.9939326799564947,
+      "learning_rate": 3.730639090199616e-06,
+      "loss": 0.3985,
+      "step": 2765
+    },
+    {
+      "epoch": 1.871447902571042,
+      "grad_norm": 1.0061801509678485,
+      "learning_rate": 3.7268316305629836e-06,
+      "loss": 0.3993,
+      "step": 2766
+    },
+    {
+      "epoch": 1.8721244925575102,
+      "grad_norm": 0.9492586830061073,
+      "learning_rate": 3.7230249602221163e-06,
+      "loss": 0.3708,
+      "step": 2767
+    },
+    {
+      "epoch": 1.8728010825439783,
+      "grad_norm": 0.9653781636902337,
+      "learning_rate": 3.719219081536942e-06,
+      "loss": 0.3783,
+      "step": 2768
+    },
+    {
+      "epoch": 1.8734776725304465,
+      "grad_norm": 0.9620951517242323,
+      "learning_rate": 3.7154139968669043e-06,
+      "loss": 0.4085,
+      "step": 2769
+    },
+    {
+      "epoch": 1.8741542625169147,
+      "grad_norm": 0.9590883410512174,
+      "learning_rate": 3.711609708570948e-06,
+      "loss": 0.3918,
+      "step": 2770
+    },
+    {
+      "epoch": 1.874830852503383,
+      "grad_norm": 0.9876900472948228,
+      "learning_rate": 3.7078062190075264e-06,
+      "loss": 0.4055,
+      "step": 2771
+    },
+    {
+      "epoch": 1.8755074424898512,
+      "grad_norm": 0.961043824549599,
+      "learning_rate": 3.704003530534597e-06,
+      "loss": 0.3919,
+      "step": 2772
+    },
+    {
+      "epoch": 1.8761840324763193,
+      "grad_norm": 0.9505863236459713,
+      "learning_rate": 3.7002016455096247e-06,
+      "loss": 0.3778,
+      "step": 2773
+    },
+    {
+      "epoch": 1.8768606224627875,
+      "grad_norm": 0.974330024617366,
+      "learning_rate": 3.696400566289571e-06,
+      "loss": 0.3891,
+      "step": 2774
+    },
+    {
+      "epoch": 1.8775372124492558,
+      "grad_norm": 0.9519792071742638,
+      "learning_rate": 3.6926002952309015e-06,
+      "loss": 0.3826,
+      "step": 2775
+    },
+    {
+      "epoch": 1.878213802435724,
+      "grad_norm": 0.9790413464789157,
+      "learning_rate": 3.6888008346895797e-06,
+      "loss": 0.3984,
+      "step": 2776
+    },
+    {
+      "epoch": 1.8788903924221922,
+      "grad_norm": 0.9446688806491784,
+      "learning_rate": 3.685002187021064e-06,
+      "loss": 0.3953,
+      "step": 2777
+    },
+    {
+      "epoch": 1.8795669824086603,
+      "grad_norm": 0.9755760917432024,
+      "learning_rate": 3.681204354580313e-06,
+      "loss": 0.4098,
+      "step": 2778
+    },
+    {
+      "epoch": 1.8802435723951285,
+      "grad_norm": 0.9818430663351374,
+      "learning_rate": 3.6774073397217786e-06,
+      "loss": 0.4019,
+      "step": 2779
+    },
+    {
+      "epoch": 1.8809201623815968,
+      "grad_norm": 0.9552511405776277,
+      "learning_rate": 3.6736111447994026e-06,
+      "loss": 0.4035,
+      "step": 2780
+    },
+    {
+      "epoch": 1.881596752368065,
+      "grad_norm": 0.9552607028054969,
+      "learning_rate": 3.669815772166625e-06,
+      "loss": 0.3849,
+      "step": 2781
+    },
+    {
+      "epoch": 1.8822733423545333,
+      "grad_norm": 0.9536352716432847,
+      "learning_rate": 3.6660212241763692e-06,
+      "loss": 0.3911,
+      "step": 2782
+    },
+    {
+      "epoch": 1.8829499323410013,
+      "grad_norm": 0.9921563560458461,
+      "learning_rate": 3.662227503181054e-06,
+      "loss": 0.4034,
+      "step": 2783
+    },
+    {
+      "epoch": 1.8836265223274695,
+      "grad_norm": 1.0185812571841373,
+      "learning_rate": 3.658434611532578e-06,
+      "loss": 0.3915,
+      "step": 2784
+    },
+    {
+      "epoch": 1.8843031123139378,
+      "grad_norm": 1.0241495689199167,
+      "learning_rate": 3.65464255158233e-06,
+      "loss": 0.4069,
+      "step": 2785
+    },
+    {
+      "epoch": 1.8849797023004058,
+      "grad_norm": 1.0183859549395882,
+      "learning_rate": 3.6508513256811856e-06,
+      "loss": 0.4062,
+      "step": 2786
+    },
+    {
+      "epoch": 1.8856562922868743,
+      "grad_norm": 0.982689482340641,
+      "learning_rate": 3.6470609361794972e-06,
+      "loss": 0.3803,
+      "step": 2787
+    },
+    {
+      "epoch": 1.8863328822733423,
+      "grad_norm": 0.9794251537636517,
+      "learning_rate": 3.643271385427105e-06,
+      "loss": 0.3851,
+      "step": 2788
+    },
+    {
+      "epoch": 1.8870094722598105,
+      "grad_norm": 0.9682502978651679,
+      "learning_rate": 3.639482675773324e-06,
+      "loss": 0.4056,
+      "step": 2789
+    },
+    {
+      "epoch": 1.8876860622462788,
+      "grad_norm": 0.937772382058869,
+      "learning_rate": 3.635694809566954e-06,
+      "loss": 0.3946,
+      "step": 2790
+    },
+    {
+      "epoch": 1.8883626522327468,
+      "grad_norm": 0.9495086893245787,
+      "learning_rate": 3.6319077891562616e-06,
+      "loss": 0.4001,
+      "step": 2791
+    },
+    {
+      "epoch": 1.8890392422192153,
+      "grad_norm": 1.0283747007480388,
+      "learning_rate": 3.6281216168889993e-06,
+      "loss": 0.412,
+      "step": 2792
+    },
+    {
+      "epoch": 1.8897158322056833,
+      "grad_norm": 0.9069792284601774,
+      "learning_rate": 3.624336295112388e-06,
+      "loss": 0.3674,
+      "step": 2793
+    },
+    {
+      "epoch": 1.8903924221921515,
+      "grad_norm": 0.9522981535658404,
+      "learning_rate": 3.6205518261731247e-06,
+      "loss": 0.387,
+      "step": 2794
+    },
+    {
+      "epoch": 1.8910690121786198,
+      "grad_norm": 0.9582503735982182,
+      "learning_rate": 3.616768212417375e-06,
+      "loss": 0.3877,
+      "step": 2795
+    },
+    {
+      "epoch": 1.8917456021650878,
+      "grad_norm": 0.957293919962844,
+      "learning_rate": 3.6129854561907786e-06,
+      "loss": 0.3827,
+      "step": 2796
+    },
+    {
+      "epoch": 1.8924221921515563,
+      "grad_norm": 0.946894126538796,
+      "learning_rate": 3.6092035598384356e-06,
+      "loss": 0.3799,
+      "step": 2797
+    },
+    {
+      "epoch": 1.8930987821380243,
+      "grad_norm": 0.9447382438618128,
+      "learning_rate": 3.6054225257049204e-06,
+      "loss": 0.3708,
+      "step": 2798
+    },
+    {
+      "epoch": 1.8937753721244925,
+      "grad_norm": 0.9500877877986935,
+      "learning_rate": 3.6016423561342707e-06,
+      "loss": 0.3773,
+      "step": 2799
+    },
+    {
+      "epoch": 1.8944519621109608,
+      "grad_norm": 0.9641975593944571,
+      "learning_rate": 3.5978630534699873e-06,
+      "loss": 0.3795,
+      "step": 2800
+    },
+    {
+      "epoch": 1.8951285520974288,
+      "grad_norm": 0.9453746819066781,
+      "learning_rate": 3.5940846200550327e-06,
+      "loss": 0.3957,
+      "step": 2801
+    },
+    {
+      "epoch": 1.8958051420838973,
+      "grad_norm": 0.9508531731708235,
+      "learning_rate": 3.5903070582318356e-06,
+      "loss": 0.3838,
+      "step": 2802
+    },
+    {
+      "epoch": 1.8964817320703653,
+      "grad_norm": 0.9608120145748116,
+      "learning_rate": 3.5865303703422794e-06,
+      "loss": 0.39,
+      "step": 2803
+    },
+    {
+      "epoch": 1.8971583220568335,
+      "grad_norm": 0.9692517729007015,
+      "learning_rate": 3.5827545587277033e-06,
+      "loss": 0.3946,
+      "step": 2804
+    },
+    {
+      "epoch": 1.8978349120433018,
+      "grad_norm": 0.9670414702909218,
+      "learning_rate": 3.5789796257289117e-06,
+      "loss": 0.3763,
+      "step": 2805
+    },
+    {
+      "epoch": 1.8985115020297698,
+      "grad_norm": 0.9630253511275844,
+      "learning_rate": 3.5752055736861567e-06,
+      "loss": 0.391,
+      "step": 2806
+    },
+    {
+      "epoch": 1.8991880920162383,
+      "grad_norm": 0.9201773183133836,
+      "learning_rate": 3.571432404939149e-06,
+      "loss": 0.368,
+      "step": 2807
+    },
+    {
+      "epoch": 1.8998646820027063,
+      "grad_norm": 1.0000228594068894,
+      "learning_rate": 3.567660121827048e-06,
+      "loss": 0.3884,
+      "step": 2808
+    },
+    {
+      "epoch": 1.9005412719891746,
+      "grad_norm": 1.001379977382614,
+      "learning_rate": 3.5638887266884682e-06,
+      "loss": 0.3985,
+      "step": 2809
+    },
+    {
+      "epoch": 1.9012178619756428,
+      "grad_norm": 1.0050413652915795,
+      "learning_rate": 3.5601182218614706e-06,
+      "loss": 0.4064,
+      "step": 2810
+    },
+    {
+      "epoch": 1.9018944519621108,
+      "grad_norm": 0.942357973561415,
+      "learning_rate": 3.5563486096835643e-06,
+      "loss": 0.3868,
+      "step": 2811
+    },
+    {
+      "epoch": 1.9025710419485793,
+      "grad_norm": 0.9452019830664797,
+      "learning_rate": 3.552579892491704e-06,
+      "loss": 0.3906,
+      "step": 2812
+    },
+    {
+      "epoch": 1.9032476319350473,
+      "grad_norm": 0.9744015353774492,
+      "learning_rate": 3.548812072622294e-06,
+      "loss": 0.381,
+      "step": 2813
+    },
+    {
+      "epoch": 1.9039242219215156,
+      "grad_norm": 0.971122503753358,
+      "learning_rate": 3.545045152411178e-06,
+      "loss": 0.3837,
+      "step": 2814
+    },
+    {
+      "epoch": 1.9046008119079838,
+      "grad_norm": 0.9527853306453381,
+      "learning_rate": 3.5412791341936446e-06,
+      "loss": 0.3846,
+      "step": 2815
+    },
+    {
+      "epoch": 1.9052774018944518,
+      "grad_norm": 0.9159435943064962,
+      "learning_rate": 3.5375140203044233e-06,
+      "loss": 0.3792,
+      "step": 2816
+    },
+    {
+      "epoch": 1.9059539918809203,
+      "grad_norm": 0.9578571255864694,
+      "learning_rate": 3.533749813077677e-06,
+      "loss": 0.3731,
+      "step": 2817
+    },
+    {
+      "epoch": 1.9066305818673883,
+      "grad_norm": 0.9605758862853653,
+      "learning_rate": 3.5299865148470157e-06,
+      "loss": 0.3922,
+      "step": 2818
+    },
+    {
+      "epoch": 1.9073071718538566,
+      "grad_norm": 1.0324707495534615,
+      "learning_rate": 3.526224127945479e-06,
+      "loss": 0.4137,
+      "step": 2819
+    },
+    {
+      "epoch": 1.9079837618403248,
+      "grad_norm": 0.9866128094125127,
+      "learning_rate": 3.5224626547055463e-06,
+      "loss": 0.387,
+      "step": 2820
+    },
+    {
+      "epoch": 1.9086603518267928,
+      "grad_norm": 0.9579418544230219,
+      "learning_rate": 3.518702097459126e-06,
+      "loss": 0.3788,
+      "step": 2821
+    },
+    {
+      "epoch": 1.9093369418132613,
+      "grad_norm": 1.0191306852982904,
+      "learning_rate": 3.5149424585375623e-06,
+      "loss": 0.4013,
+      "step": 2822
+    },
+    {
+      "epoch": 1.9100135317997293,
+      "grad_norm": 0.973425893268513,
+      "learning_rate": 3.5111837402716297e-06,
+      "loss": 0.3799,
+      "step": 2823
+    },
+    {
+      "epoch": 1.9106901217861976,
+      "grad_norm": 0.945748236263133,
+      "learning_rate": 3.507425944991529e-06,
+      "loss": 0.3794,
+      "step": 2824
+    },
+    {
+      "epoch": 1.9113667117726658,
+      "grad_norm": 0.9875665785501894,
+      "learning_rate": 3.5036690750268897e-06,
+      "loss": 0.3863,
+      "step": 2825
+    },
+    {
+      "epoch": 1.9120433017591338,
+      "grad_norm": 0.9574025280037418,
+      "learning_rate": 3.499913132706771e-06,
+      "loss": 0.3903,
+      "step": 2826
+    },
+    {
+      "epoch": 1.9127198917456023,
+      "grad_norm": 0.9934220298931912,
+      "learning_rate": 3.496158120359653e-06,
+      "loss": 0.374,
+      "step": 2827
+    },
+    {
+      "epoch": 1.9133964817320703,
+      "grad_norm": 0.9842609953846655,
+      "learning_rate": 3.492404040313443e-06,
+      "loss": 0.3896,
+      "step": 2828
+    },
+    {
+      "epoch": 1.9140730717185386,
+      "grad_norm": 0.9350775444620787,
+      "learning_rate": 3.4886508948954656e-06,
+      "loss": 0.3931,
+      "step": 2829
+    },
+    {
+      "epoch": 1.9147496617050068,
+      "grad_norm": 0.9529282501576434,
+      "learning_rate": 3.484898686432473e-06,
+      "loss": 0.378,
+      "step": 2830
+    },
+    {
+      "epoch": 1.9154262516914748,
+      "grad_norm": 0.9054300867056541,
+      "learning_rate": 3.4811474172506277e-06,
+      "loss": 0.3606,
+      "step": 2831
+    },
+    {
+      "epoch": 1.9161028416779433,
+      "grad_norm": 0.9730209052798792,
+      "learning_rate": 3.4773970896755167e-06,
+      "loss": 0.402,
+      "step": 2832
+    },
+    {
+      "epoch": 1.9167794316644113,
+      "grad_norm": 0.9711823319274516,
+      "learning_rate": 3.4736477060321387e-06,
+      "loss": 0.397,
+      "step": 2833
+    },
+    {
+      "epoch": 1.9174560216508796,
+      "grad_norm": 0.9724124107493505,
+      "learning_rate": 3.469899268644913e-06,
+      "loss": 0.3985,
+      "step": 2834
+    },
+    {
+      "epoch": 1.9181326116373478,
+      "grad_norm": 0.9626689911589725,
+      "learning_rate": 3.466151779837665e-06,
+      "loss": 0.3871,
+      "step": 2835
+    },
+    {
+      "epoch": 1.9188092016238159,
+      "grad_norm": 0.9288249770414444,
+      "learning_rate": 3.4624052419336395e-06,
+      "loss": 0.3974,
+      "step": 2836
+    },
+    {
+      "epoch": 1.9194857916102843,
+      "grad_norm": 0.9648098540021439,
+      "learning_rate": 3.458659657255486e-06,
+      "loss": 0.3914,
+      "step": 2837
+    },
+    {
+      "epoch": 1.9201623815967523,
+      "grad_norm": 1.0116173148389211,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.4055,
+      "step": 2838
+    },
+    {
+      "epoch": 1.9208389715832206,
+      "grad_norm": 0.9648934805064656,
+      "learning_rate": 3.4511713568644432e-06,
+      "loss": 0.4077,
+      "step": 2839
+    },
+    {
+      "epoch": 1.9215155615696888,
+      "grad_norm": 0.9505351658348619,
+      "learning_rate": 3.4474286457938976e-06,
+      "loss": 0.3908,
+      "step": 2840
+    },
+    {
+      "epoch": 1.9221921515561569,
+      "grad_norm": 0.9178824448310808,
+      "learning_rate": 3.4436868972339073e-06,
+      "loss": 0.378,
+      "step": 2841
+    },
+    {
+      "epoch": 1.9228687415426253,
+      "grad_norm": 0.934040816987798,
+      "learning_rate": 3.4399461135041525e-06,
+      "loss": 0.3667,
+      "step": 2842
+    },
+    {
+      "epoch": 1.9235453315290933,
+      "grad_norm": 0.9379242534960136,
+      "learning_rate": 3.4362062969237227e-06,
+      "loss": 0.3941,
+      "step": 2843
+    },
+    {
+      "epoch": 1.9242219215155616,
+      "grad_norm": 0.9514297065230649,
+      "learning_rate": 3.4324674498110956e-06,
+      "loss": 0.3855,
+      "step": 2844
+    },
+    {
+      "epoch": 1.9248985115020298,
+      "grad_norm": 0.9103341771896453,
+      "learning_rate": 3.4287295744841588e-06,
+      "loss": 0.374,
+      "step": 2845
+    },
+    {
+      "epoch": 1.9255751014884979,
+      "grad_norm": 0.9761131062101809,
+      "learning_rate": 3.4249926732601914e-06,
+      "loss": 0.3738,
+      "step": 2846
+    },
+    {
+      "epoch": 1.9262516914749663,
+      "grad_norm": 0.9696485524449143,
+      "learning_rate": 3.4212567484558735e-06,
+      "loss": 0.3919,
+      "step": 2847
+    },
+    {
+      "epoch": 1.9269282814614344,
+      "grad_norm": 0.9538678927713922,
+      "learning_rate": 3.4175218023872753e-06,
+      "loss": 0.3911,
+      "step": 2848
+    },
+    {
+      "epoch": 1.9276048714479026,
+      "grad_norm": 0.9389077060372707,
+      "learning_rate": 3.413787837369863e-06,
+      "loss": 0.3895,
+      "step": 2849
+    },
+    {
+      "epoch": 1.9282814614343708,
+      "grad_norm": 0.9972066459954332,
+      "learning_rate": 3.4100548557184944e-06,
+      "loss": 0.4015,
+      "step": 2850
+    },
+    {
+      "epoch": 1.9289580514208389,
+      "grad_norm": 0.9216920339702025,
+      "learning_rate": 3.4063228597474133e-06,
+      "loss": 0.3687,
+      "step": 2851
+    },
+    {
+      "epoch": 1.9296346414073073,
+      "grad_norm": 0.9738682835054276,
+      "learning_rate": 3.40259185177026e-06,
+      "loss": 0.3874,
+      "step": 2852
+    },
+    {
+      "epoch": 1.9303112313937754,
+      "grad_norm": 0.9608396408178609,
+      "learning_rate": 3.3988618341000566e-06,
+      "loss": 0.3761,
+      "step": 2853
+    },
+    {
+      "epoch": 1.9309878213802436,
+      "grad_norm": 1.010039770864873,
+      "learning_rate": 3.395132809049212e-06,
+      "loss": 0.3998,
+      "step": 2854
+    },
+    {
+      "epoch": 1.9316644113667119,
+      "grad_norm": 0.9708243237418063,
+      "learning_rate": 3.391404778929523e-06,
+      "loss": 0.3907,
+      "step": 2855
+    },
+    {
+      "epoch": 1.9323410013531799,
+      "grad_norm": 0.9047366472542193,
+      "learning_rate": 3.3876777460521647e-06,
+      "loss": 0.3585,
+      "step": 2856
+    },
+    {
+      "epoch": 1.9330175913396481,
+      "grad_norm": 0.9734212954300306,
+      "learning_rate": 3.383951712727701e-06,
+      "loss": 0.3929,
+      "step": 2857
+    },
+    {
+      "epoch": 1.9336941813261164,
+      "grad_norm": 0.9324070495018218,
+      "learning_rate": 3.3802266812660674e-06,
+      "loss": 0.3788,
+      "step": 2858
+    },
+    {
+      "epoch": 1.9343707713125846,
+      "grad_norm": 0.947309948772134,
+      "learning_rate": 3.3765026539765832e-06,
+      "loss": 0.3821,
+      "step": 2859
+    },
+    {
+      "epoch": 1.9350473612990529,
+      "grad_norm": 0.9021359226034089,
+      "learning_rate": 3.372779633167946e-06,
+      "loss": 0.3718,
+      "step": 2860
+    },
+    {
+      "epoch": 1.9357239512855209,
+      "grad_norm": 0.9935978264460849,
+      "learning_rate": 3.369057621148227e-06,
+      "loss": 0.3821,
+      "step": 2861
+    },
+    {
+      "epoch": 1.9364005412719891,
+      "grad_norm": 0.9164059370888998,
+      "learning_rate": 3.3653366202248738e-06,
+      "loss": 0.3814,
+      "step": 2862
+    },
+    {
+      "epoch": 1.9370771312584574,
+      "grad_norm": 0.9674431295211486,
+      "learning_rate": 3.3616166327047084e-06,
+      "loss": 0.3935,
+      "step": 2863
+    },
+    {
+      "epoch": 1.9377537212449256,
+      "grad_norm": 0.915352582882395,
+      "learning_rate": 3.3578976608939184e-06,
+      "loss": 0.3819,
+      "step": 2864
+    },
+    {
+      "epoch": 1.9384303112313939,
+      "grad_norm": 0.9567493191761282,
+      "learning_rate": 3.3541797070980663e-06,
+      "loss": 0.3864,
+      "step": 2865
+    },
+    {
+      "epoch": 1.939106901217862,
+      "grad_norm": 0.91384633516482,
+      "learning_rate": 3.3504627736220863e-06,
+      "loss": 0.3624,
+      "step": 2866
+    },
+    {
+      "epoch": 1.9397834912043301,
+      "grad_norm": 0.931640281843564,
+      "learning_rate": 3.3467468627702736e-06,
+      "loss": 0.3748,
+      "step": 2867
+    },
+    {
+      "epoch": 1.9404600811907984,
+      "grad_norm": 0.9584167372969533,
+      "learning_rate": 3.3430319768462956e-06,
+      "loss": 0.3855,
+      "step": 2868
+    },
+    {
+      "epoch": 1.9411366711772666,
+      "grad_norm": 0.9405848069357464,
+      "learning_rate": 3.3393181181531785e-06,
+      "loss": 0.3758,
+      "step": 2869
+    },
+    {
+      "epoch": 1.9418132611637349,
+      "grad_norm": 1.0128862798860447,
+      "learning_rate": 3.3356052889933177e-06,
+      "loss": 0.3944,
+      "step": 2870
+    },
+    {
+      "epoch": 1.942489851150203,
+      "grad_norm": 0.9654630010379878,
+      "learning_rate": 3.331893491668464e-06,
+      "loss": 0.3848,
+      "step": 2871
+    },
+    {
+      "epoch": 1.9431664411366711,
+      "grad_norm": 0.9471093982546952,
+      "learning_rate": 3.3281827284797317e-06,
+      "loss": 0.401,
+      "step": 2872
+    },
+    {
+      "epoch": 1.9438430311231394,
+      "grad_norm": 0.9104591170027402,
+      "learning_rate": 3.3244730017275974e-06,
+      "loss": 0.3756,
+      "step": 2873
+    },
+    {
+      "epoch": 1.9445196211096076,
+      "grad_norm": 0.9585672602525391,
+      "learning_rate": 3.3207643137118872e-06,
+      "loss": 0.3787,
+      "step": 2874
+    },
+    {
+      "epoch": 1.9451962110960759,
+      "grad_norm": 0.9051463480050137,
+      "learning_rate": 3.3170566667317917e-06,
+      "loss": 0.3672,
+      "step": 2875
+    },
+    {
+      "epoch": 1.945872801082544,
+      "grad_norm": 0.9602883465334724,
+      "learning_rate": 3.3133500630858507e-06,
+      "loss": 0.3868,
+      "step": 2876
+    },
+    {
+      "epoch": 1.9465493910690121,
+      "grad_norm": 0.8848662794491753,
+      "learning_rate": 3.309644505071959e-06,
+      "loss": 0.3656,
+      "step": 2877
+    },
+    {
+      "epoch": 1.9472259810554804,
+      "grad_norm": 0.8776402634641406,
+      "learning_rate": 3.3059399949873605e-06,
+      "loss": 0.3613,
+      "step": 2878
+    },
+    {
+      "epoch": 1.9479025710419484,
+      "grad_norm": 0.9278448142055523,
+      "learning_rate": 3.3022365351286545e-06,
+      "loss": 0.3688,
+      "step": 2879
+    },
+    {
+      "epoch": 1.9485791610284169,
+      "grad_norm": 0.9318795655942258,
+      "learning_rate": 3.298534127791785e-06,
+      "loss": 0.3845,
+      "step": 2880
+    },
+    {
+      "epoch": 1.949255751014885,
+      "grad_norm": 0.9702668985938906,
+      "learning_rate": 3.2948327752720464e-06,
+      "loss": 0.3986,
+      "step": 2881
+    },
+    {
+      "epoch": 1.9499323410013532,
+      "grad_norm": 0.9421255263990418,
+      "learning_rate": 3.2911324798640764e-06,
+      "loss": 0.396,
+      "step": 2882
+    },
+    {
+      "epoch": 1.9506089309878214,
+      "grad_norm": 0.9636564263109836,
+      "learning_rate": 3.2874332438618607e-06,
+      "loss": 0.3935,
+      "step": 2883
+    },
+    {
+      "epoch": 1.9512855209742894,
+      "grad_norm": 0.9357187148746272,
+      "learning_rate": 3.2837350695587237e-06,
+      "loss": 0.3829,
+      "step": 2884
+    },
+    {
+      "epoch": 1.951962110960758,
+      "grad_norm": 0.9276320040386208,
+      "learning_rate": 3.280037959247336e-06,
+      "loss": 0.3847,
+      "step": 2885
+    },
+    {
+      "epoch": 1.952638700947226,
+      "grad_norm": 0.9247564963846984,
+      "learning_rate": 3.276341915219704e-06,
+      "loss": 0.3819,
+      "step": 2886
+    },
+    {
+      "epoch": 1.9533152909336942,
+      "grad_norm": 0.9518951118579614,
+      "learning_rate": 3.2726469397671797e-06,
+      "loss": 0.3871,
+      "step": 2887
+    },
+    {
+      "epoch": 1.9539918809201624,
+      "grad_norm": 0.9502773764867274,
+      "learning_rate": 3.268953035180445e-06,
+      "loss": 0.3835,
+      "step": 2888
+    },
+    {
+      "epoch": 1.9546684709066304,
+      "grad_norm": 0.9853605160786062,
+      "learning_rate": 3.2652602037495247e-06,
+      "loss": 0.388,
+      "step": 2889
+    },
+    {
+      "epoch": 1.955345060893099,
+      "grad_norm": 0.9808710233742232,
+      "learning_rate": 3.261568447763775e-06,
+      "loss": 0.4045,
+      "step": 2890
+    },
+    {
+      "epoch": 1.956021650879567,
+      "grad_norm": 0.9350573077338281,
+      "learning_rate": 3.2578777695118822e-06,
+      "loss": 0.3916,
+      "step": 2891
+    },
+    {
+      "epoch": 1.9566982408660352,
+      "grad_norm": 0.9990549083597429,
+      "learning_rate": 3.254188171281871e-06,
+      "loss": 0.3927,
+      "step": 2892
+    },
+    {
+      "epoch": 1.9573748308525034,
+      "grad_norm": 0.9300835015869638,
+      "learning_rate": 3.2504996553610924e-06,
+      "loss": 0.3829,
+      "step": 2893
+    },
+    {
+      "epoch": 1.9580514208389714,
+      "grad_norm": 0.9337900156826742,
+      "learning_rate": 3.2468122240362287e-06,
+      "loss": 0.3834,
+      "step": 2894
+    },
+    {
+      "epoch": 1.95872801082544,
+      "grad_norm": 0.9748744128707281,
+      "learning_rate": 3.2431258795932863e-06,
+      "loss": 0.3924,
+      "step": 2895
+    },
+    {
+      "epoch": 1.959404600811908,
+      "grad_norm": 0.9355614717781006,
+      "learning_rate": 3.2394406243176025e-06,
+      "loss": 0.3784,
+      "step": 2896
+    },
+    {
+      "epoch": 1.9600811907983762,
+      "grad_norm": 0.9587164673655073,
+      "learning_rate": 3.2357564604938363e-06,
+      "loss": 0.3854,
+      "step": 2897
+    },
+    {
+      "epoch": 1.9607577807848444,
+      "grad_norm": 0.8825622603252513,
+      "learning_rate": 3.232073390405969e-06,
+      "loss": 0.361,
+      "step": 2898
+    },
+    {
+      "epoch": 1.9614343707713124,
+      "grad_norm": 0.9285993721908563,
+      "learning_rate": 3.2283914163373064e-06,
+      "loss": 0.3789,
+      "step": 2899
+    },
+    {
+      "epoch": 1.962110960757781,
+      "grad_norm": 0.9564243877471528,
+      "learning_rate": 3.224710540570475e-06,
+      "loss": 0.3867,
+      "step": 2900
+    },
+    {
+      "epoch": 1.962787550744249,
+      "grad_norm": 1.019897269612813,
+      "learning_rate": 3.2210307653874175e-06,
+      "loss": 0.4011,
+      "step": 2901
+    },
+    {
+      "epoch": 1.9634641407307172,
+      "grad_norm": 0.9139846983316788,
+      "learning_rate": 3.2173520930693987e-06,
+      "loss": 0.3662,
+      "step": 2902
+    },
+    {
+      "epoch": 1.9641407307171854,
+      "grad_norm": 0.920949944630482,
+      "learning_rate": 3.2136745258969965e-06,
+      "loss": 0.3789,
+      "step": 2903
+    },
+    {
+      "epoch": 1.9648173207036534,
+      "grad_norm": 0.9650036235115824,
+      "learning_rate": 3.2099980661501016e-06,
+      "loss": 0.4025,
+      "step": 2904
+    },
+    {
+      "epoch": 1.965493910690122,
+      "grad_norm": 0.9547398265775282,
+      "learning_rate": 3.2063227161079234e-06,
+      "loss": 0.3871,
+      "step": 2905
+    },
+    {
+      "epoch": 1.96617050067659,
+      "grad_norm": 0.9612879639781636,
+      "learning_rate": 3.202648478048981e-06,
+      "loss": 0.3795,
+      "step": 2906
+    },
+    {
+      "epoch": 1.9668470906630582,
+      "grad_norm": 0.9477117701441659,
+      "learning_rate": 3.1989753542511016e-06,
+      "loss": 0.3732,
+      "step": 2907
+    },
+    {
+      "epoch": 1.9675236806495264,
+      "grad_norm": 0.9761982483445683,
+      "learning_rate": 3.1953033469914273e-06,
+      "loss": 0.3894,
+      "step": 2908
+    },
+    {
+      "epoch": 1.9682002706359945,
+      "grad_norm": 0.9215681550973517,
+      "learning_rate": 3.191632458546401e-06,
+      "loss": 0.3722,
+      "step": 2909
+    },
+    {
+      "epoch": 1.968876860622463,
+      "grad_norm": 0.9412949533312766,
+      "learning_rate": 3.1879626911917806e-06,
+      "loss": 0.3901,
+      "step": 2910
+    },
+    {
+      "epoch": 1.969553450608931,
+      "grad_norm": 0.9691348369623856,
+      "learning_rate": 3.1842940472026194e-06,
+      "loss": 0.392,
+      "step": 2911
+    },
+    {
+      "epoch": 1.9702300405953992,
+      "grad_norm": 0.9758249542217725,
+      "learning_rate": 3.18062652885328e-06,
+      "loss": 0.3757,
+      "step": 2912
+    },
+    {
+      "epoch": 1.9709066305818674,
+      "grad_norm": 0.9751745215266456,
+      "learning_rate": 3.1769601384174274e-06,
+      "loss": 0.3986,
+      "step": 2913
+    },
+    {
+      "epoch": 1.9715832205683355,
+      "grad_norm": 0.9087155948453963,
+      "learning_rate": 3.173294878168025e-06,
+      "loss": 0.3661,
+      "step": 2914
+    },
+    {
+      "epoch": 1.972259810554804,
+      "grad_norm": 0.9523954700522769,
+      "learning_rate": 3.169630750377337e-06,
+      "loss": 0.3713,
+      "step": 2915
+    },
+    {
+      "epoch": 1.972936400541272,
+      "grad_norm": 0.9720553313694448,
+      "learning_rate": 3.165967757316925e-06,
+      "loss": 0.388,
+      "step": 2916
+    },
+    {
+      "epoch": 1.9736129905277402,
+      "grad_norm": 0.9590252306028271,
+      "learning_rate": 3.16230590125765e-06,
+      "loss": 0.3943,
+      "step": 2917
+    },
+    {
+      "epoch": 1.9742895805142084,
+      "grad_norm": 0.9783369003095155,
+      "learning_rate": 3.1586451844696596e-06,
+      "loss": 0.3923,
+      "step": 2918
+    },
+    {
+      "epoch": 1.9749661705006765,
+      "grad_norm": 0.9707047467522676,
+      "learning_rate": 3.154985609222405e-06,
+      "loss": 0.3897,
+      "step": 2919
+    },
+    {
+      "epoch": 1.975642760487145,
+      "grad_norm": 0.9834293043937559,
+      "learning_rate": 3.1513271777846244e-06,
+      "loss": 0.3967,
+      "step": 2920
+    },
+    {
+      "epoch": 1.976319350473613,
+      "grad_norm": 0.9513142124211352,
+      "learning_rate": 3.1476698924243487e-06,
+      "loss": 0.3982,
+      "step": 2921
+    },
+    {
+      "epoch": 1.9769959404600812,
+      "grad_norm": 0.9412110601695037,
+      "learning_rate": 3.1440137554088957e-06,
+      "loss": 0.387,
+      "step": 2922
+    },
+    {
+      "epoch": 1.9776725304465494,
+      "grad_norm": 0.9159394025749934,
+      "learning_rate": 3.1403587690048775e-06,
+      "loss": 0.3786,
+      "step": 2923
+    },
+    {
+      "epoch": 1.9783491204330175,
+      "grad_norm": 0.9395174913262756,
+      "learning_rate": 3.1367049354781854e-06,
+      "loss": 0.3946,
+      "step": 2924
+    },
+    {
+      "epoch": 1.979025710419486,
+      "grad_norm": 0.9383214231914938,
+      "learning_rate": 3.1330522570939987e-06,
+      "loss": 0.3842,
+      "step": 2925
+    },
+    {
+      "epoch": 1.979702300405954,
+      "grad_norm": 0.9686656631817997,
+      "learning_rate": 3.129400736116783e-06,
+      "loss": 0.3975,
+      "step": 2926
+    },
+    {
+      "epoch": 1.9803788903924222,
+      "grad_norm": 0.9671352879487586,
+      "learning_rate": 3.125750374810283e-06,
+      "loss": 0.3767,
+      "step": 2927
+    },
+    {
+      "epoch": 1.9810554803788905,
+      "grad_norm": 0.9242821609719624,
+      "learning_rate": 3.1221011754375275e-06,
+      "loss": 0.383,
+      "step": 2928
+    },
+    {
+      "epoch": 1.9817320703653585,
+      "grad_norm": 0.959463781802831,
+      "learning_rate": 3.118453140260823e-06,
+      "loss": 0.3792,
+      "step": 2929
+    },
+    {
+      "epoch": 1.982408660351827,
+      "grad_norm": 0.9515821291385462,
+      "learning_rate": 3.1148062715417553e-06,
+      "loss": 0.3903,
+      "step": 2930
+    },
+    {
+      "epoch": 1.983085250338295,
+      "grad_norm": 0.9327741393852783,
+      "learning_rate": 3.111160571541183e-06,
+      "loss": 0.3777,
+      "step": 2931
+    },
+    {
+      "epoch": 1.9837618403247632,
+      "grad_norm": 0.8922039273045852,
+      "learning_rate": 3.107516042519248e-06,
+      "loss": 0.3535,
+      "step": 2932
+    },
+    {
+      "epoch": 1.9844384303112315,
+      "grad_norm": 0.966783229592581,
+      "learning_rate": 3.1038726867353587e-06,
+      "loss": 0.3761,
+      "step": 2933
+    },
+    {
+      "epoch": 1.9851150202976995,
+      "grad_norm": 0.9620160626649537,
+      "learning_rate": 3.1002305064482006e-06,
+      "loss": 0.3754,
+      "step": 2934
+    },
+    {
+      "epoch": 1.985791610284168,
+      "grad_norm": 0.9700042274732317,
+      "learning_rate": 3.096589503915729e-06,
+      "loss": 0.3779,
+      "step": 2935
+    },
+    {
+      "epoch": 1.986468200270636,
+      "grad_norm": 0.9842938956826476,
+      "learning_rate": 3.09294968139517e-06,
+      "loss": 0.3882,
+      "step": 2936
+    },
+    {
+      "epoch": 1.9871447902571042,
+      "grad_norm": 0.9773708083210262,
+      "learning_rate": 3.089311041143017e-06,
+      "loss": 0.3811,
+      "step": 2937
+    },
+    {
+      "epoch": 1.9878213802435725,
+      "grad_norm": 0.9484892252396426,
+      "learning_rate": 3.085673585415031e-06,
+      "loss": 0.3884,
+      "step": 2938
+    },
+    {
+      "epoch": 1.9884979702300405,
+      "grad_norm": 0.9783067505673053,
+      "learning_rate": 3.082037316466236e-06,
+      "loss": 0.3886,
+      "step": 2939
+    },
+    {
+      "epoch": 1.989174560216509,
+      "grad_norm": 0.9345842184278825,
+      "learning_rate": 3.078402236550926e-06,
+      "loss": 0.3762,
+      "step": 2940
+    },
+    {
+      "epoch": 1.989851150202977,
+      "grad_norm": 0.9368630543795368,
+      "learning_rate": 3.074768347922652e-06,
+      "loss": 0.3741,
+      "step": 2941
+    },
+    {
+      "epoch": 1.9905277401894452,
+      "grad_norm": 0.9335098843050984,
+      "learning_rate": 3.0711356528342316e-06,
+      "loss": 0.3712,
+      "step": 2942
+    },
+    {
+      "epoch": 1.9912043301759135,
+      "grad_norm": 0.9786673322188582,
+      "learning_rate": 3.06750415353774e-06,
+      "loss": 0.3958,
+      "step": 2943
+    },
+    {
+      "epoch": 1.9918809201623815,
+      "grad_norm": 0.9166916609198876,
+      "learning_rate": 3.063873852284508e-06,
+      "loss": 0.3788,
+      "step": 2944
+    },
+    {
+      "epoch": 1.9925575101488497,
+      "grad_norm": 0.9407536988683556,
+      "learning_rate": 3.0602447513251287e-06,
+      "loss": 0.389,
+      "step": 2945
+    },
+    {
+      "epoch": 1.993234100135318,
+      "grad_norm": 0.9169991498718381,
+      "learning_rate": 3.0566168529094485e-06,
+      "loss": 0.3808,
+      "step": 2946
+    },
+    {
+      "epoch": 1.9939106901217862,
+      "grad_norm": 0.925987917772624,
+      "learning_rate": 3.0529901592865705e-06,
+      "loss": 0.3706,
+      "step": 2947
+    },
+    {
+      "epoch": 1.9945872801082545,
+      "grad_norm": 0.9076401766987651,
+      "learning_rate": 3.0493646727048463e-06,
+      "loss": 0.3646,
+      "step": 2948
+    },
+    {
+      "epoch": 1.9952638700947225,
+      "grad_norm": 0.9495977118673681,
+      "learning_rate": 3.045740395411886e-06,
+      "loss": 0.3805,
+      "step": 2949
+    },
+    {
+      "epoch": 1.9959404600811907,
+      "grad_norm": 1.0197104166964828,
+      "learning_rate": 3.042117329654544e-06,
+      "loss": 0.3983,
+      "step": 2950
+    },
+    {
+      "epoch": 1.996617050067659,
+      "grad_norm": 0.9726730375147471,
+      "learning_rate": 3.0384954776789255e-06,
+      "loss": 0.3996,
+      "step": 2951
+    },
+    {
+      "epoch": 1.9972936400541272,
+      "grad_norm": 0.9158795363181512,
+      "learning_rate": 3.0348748417303826e-06,
+      "loss": 0.3733,
+      "step": 2952
+    },
+    {
+      "epoch": 1.9979702300405955,
+      "grad_norm": 0.9505212869269329,
+      "learning_rate": 3.0312554240535166e-06,
+      "loss": 0.3865,
+      "step": 2953
+    },
+    {
+      "epoch": 1.9986468200270635,
+      "grad_norm": 0.9862754977243918,
+      "learning_rate": 3.0276372268921694e-06,
+      "loss": 0.3877,
+      "step": 2954
+    },
+    {
+      "epoch": 1.9993234100135318,
+      "grad_norm": 1.0263494141291425,
+      "learning_rate": 3.0240202524894304e-06,
+      "loss": 0.3886,
+      "step": 2955
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8873739972552815,
+      "learning_rate": 3.0204045030876267e-06,
+      "loss": 0.3638,
+      "step": 2956
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.40424153208732605,
+      "eval_runtime": 431.6998,
+      "eval_samples_per_second": 23.06,
+      "eval_steps_per_second": 0.723,
+      "step": 2956
+    },
+    {
+      "epoch": 2.000676589986468,
+      "grad_norm": 0.9038130925946041,
+      "learning_rate": 3.016789980928331e-06,
+      "loss": 0.3329,
+      "step": 2957
+    },
+    {
+      "epoch": 2.0013531799729365,
+      "grad_norm": 0.8816872368815492,
+      "learning_rate": 3.013176688252349e-06,
+      "loss": 0.3185,
+      "step": 2958
+    },
+    {
+      "epoch": 2.0020297699594045,
+      "grad_norm": 0.9200538669123807,
+      "learning_rate": 3.009564627299728e-06,
+      "loss": 0.3411,
+      "step": 2959
+    },
+    {
+      "epoch": 2.002706359945873,
+      "grad_norm": 0.9047575100079928,
+      "learning_rate": 3.005953800309752e-06,
+      "loss": 0.3248,
+      "step": 2960
+    },
+    {
+      "epoch": 2.003382949932341,
+      "grad_norm": 0.9004158664370083,
+      "learning_rate": 3.0023442095209386e-06,
+      "loss": 0.3231,
+      "step": 2961
+    },
+    {
+      "epoch": 2.004059539918809,
+      "grad_norm": 0.9776540657457075,
+      "learning_rate": 2.9987358571710394e-06,
+      "loss": 0.3456,
+      "step": 2962
+    },
+    {
+      "epoch": 2.0047361299052775,
+      "grad_norm": 0.9376307261051864,
+      "learning_rate": 2.9951287454970405e-06,
+      "loss": 0.3278,
+      "step": 2963
+    },
+    {
+      "epoch": 2.0054127198917455,
+      "grad_norm": 1.0297889212940174,
+      "learning_rate": 2.991522876735154e-06,
+      "loss": 0.3374,
+      "step": 2964
+    },
+    {
+      "epoch": 2.006089309878214,
+      "grad_norm": 1.0007457037767937,
+      "learning_rate": 2.987918253120824e-06,
+      "loss": 0.3315,
+      "step": 2965
+    },
+    {
+      "epoch": 2.006765899864682,
+      "grad_norm": 0.9459797377430096,
+      "learning_rate": 2.984314876888725e-06,
+      "loss": 0.3109,
+      "step": 2966
+    },
+    {
+      "epoch": 2.00744248985115,
+      "grad_norm": 1.0540213518299297,
+      "learning_rate": 2.980712750272754e-06,
+      "loss": 0.3395,
+      "step": 2967
+    },
+    {
+      "epoch": 2.0081190798376185,
+      "grad_norm": 1.0139479525332433,
+      "learning_rate": 2.9771118755060368e-06,
+      "loss": 0.3369,
+      "step": 2968
+    },
+    {
+      "epoch": 2.0087956698240865,
+      "grad_norm": 0.9501841209981149,
+      "learning_rate": 2.9735122548209204e-06,
+      "loss": 0.3225,
+      "step": 2969
+    },
+    {
+      "epoch": 2.009472259810555,
+      "grad_norm": 0.9104085672574963,
+      "learning_rate": 2.96991389044898e-06,
+      "loss": 0.3173,
+      "step": 2970
+    },
+    {
+      "epoch": 2.010148849797023,
+      "grad_norm": 0.9795858637459921,
+      "learning_rate": 2.966316784621e-06,
+      "loss": 0.3297,
+      "step": 2971
+    },
+    {
+      "epoch": 2.010825439783491,
+      "grad_norm": 0.9498513959342357,
+      "learning_rate": 2.9627209395669978e-06,
+      "loss": 0.327,
+      "step": 2972
+    },
+    {
+      "epoch": 2.0115020297699595,
+      "grad_norm": 0.9486617766874432,
+      "learning_rate": 2.9591263575162e-06,
+      "loss": 0.3171,
+      "step": 2973
+    },
+    {
+      "epoch": 2.0121786197564275,
+      "grad_norm": 0.9425205170580246,
+      "learning_rate": 2.9555330406970568e-06,
+      "loss": 0.3262,
+      "step": 2974
+    },
+    {
+      "epoch": 2.012855209742896,
+      "grad_norm": 0.9553225339558391,
+      "learning_rate": 2.9519409913372286e-06,
+      "loss": 0.3257,
+      "step": 2975
+    },
+    {
+      "epoch": 2.013531799729364,
+      "grad_norm": 1.0181726236530662,
+      "learning_rate": 2.9483502116635943e-06,
+      "loss": 0.3312,
+      "step": 2976
+    },
+    {
+      "epoch": 2.014208389715832,
+      "grad_norm": 0.9212436349490788,
+      "learning_rate": 2.9447607039022443e-06,
+      "loss": 0.3088,
+      "step": 2977
+    },
+    {
+      "epoch": 2.0148849797023005,
+      "grad_norm": 0.9826074890506608,
+      "learning_rate": 2.9411724702784762e-06,
+      "loss": 0.3315,
+      "step": 2978
+    },
+    {
+      "epoch": 2.0155615696887685,
+      "grad_norm": 0.952169968543756,
+      "learning_rate": 2.9375855130168046e-06,
+      "loss": 0.3155,
+      "step": 2979
+    },
+    {
+      "epoch": 2.016238159675237,
+      "grad_norm": 0.986823165347708,
+      "learning_rate": 2.9339998343409484e-06,
+      "loss": 0.3278,
+      "step": 2980
+    },
+    {
+      "epoch": 2.016914749661705,
+      "grad_norm": 0.9663696510862468,
+      "learning_rate": 2.9304154364738358e-06,
+      "loss": 0.3151,
+      "step": 2981
+    },
+    {
+      "epoch": 2.017591339648173,
+      "grad_norm": 1.0038324147866482,
+      "learning_rate": 2.9268323216375997e-06,
+      "loss": 0.3411,
+      "step": 2982
+    },
+    {
+      "epoch": 2.0182679296346415,
+      "grad_norm": 0.9538096600942009,
+      "learning_rate": 2.92325049205358e-06,
+      "loss": 0.3286,
+      "step": 2983
+    },
+    {
+      "epoch": 2.0189445196211095,
+      "grad_norm": 1.0710026705897027,
+      "learning_rate": 2.9196699499423143e-06,
+      "loss": 0.3306,
+      "step": 2984
+    },
+    {
+      "epoch": 2.019621109607578,
+      "grad_norm": 0.9978318187088097,
+      "learning_rate": 2.9160906975235493e-06,
+      "loss": 0.3161,
+      "step": 2985
+    },
+    {
+      "epoch": 2.020297699594046,
+      "grad_norm": 1.0036118000027836,
+      "learning_rate": 2.9125127370162253e-06,
+      "loss": 0.3357,
+      "step": 2986
+    },
+    {
+      "epoch": 2.020974289580514,
+      "grad_norm": 0.9728050363976315,
+      "learning_rate": 2.908936070638487e-06,
+      "loss": 0.3199,
+      "step": 2987
+    },
+    {
+      "epoch": 2.0216508795669825,
+      "grad_norm": 0.9872888844627106,
+      "learning_rate": 2.9053607006076766e-06,
+      "loss": 0.3295,
+      "step": 2988
+    },
+    {
+      "epoch": 2.0223274695534506,
+      "grad_norm": 0.9950866097902137,
+      "learning_rate": 2.9017866291403275e-06,
+      "loss": 0.3154,
+      "step": 2989
+    },
+    {
+      "epoch": 2.023004059539919,
+      "grad_norm": 0.9994860688314967,
+      "learning_rate": 2.8982138584521734e-06,
+      "loss": 0.3144,
+      "step": 2990
+    },
+    {
+      "epoch": 2.023680649526387,
+      "grad_norm": 0.9921616855581498,
+      "learning_rate": 2.8946423907581377e-06,
+      "loss": 0.3295,
+      "step": 2991
+    },
+    {
+      "epoch": 2.024357239512855,
+      "grad_norm": 1.0299146759340774,
+      "learning_rate": 2.8910722282723404e-06,
+      "loss": 0.3275,
+      "step": 2992
+    },
+    {
+      "epoch": 2.0250338294993235,
+      "grad_norm": 1.029696683812635,
+      "learning_rate": 2.8875033732080865e-06,
+      "loss": 0.3433,
+      "step": 2993
+    },
+    {
+      "epoch": 2.0257104194857916,
+      "grad_norm": 1.0043041380089484,
+      "learning_rate": 2.8839358277778758e-06,
+      "loss": 0.3259,
+      "step": 2994
+    },
+    {
+      "epoch": 2.02638700947226,
+      "grad_norm": 1.0166006226414344,
+      "learning_rate": 2.8803695941933933e-06,
+      "loss": 0.3407,
+      "step": 2995
+    },
+    {
+      "epoch": 2.027063599458728,
+      "grad_norm": 1.0217692062128187,
+      "learning_rate": 2.876804674665515e-06,
+      "loss": 0.3376,
+      "step": 2996
+    },
+    {
+      "epoch": 2.027740189445196,
+      "grad_norm": 0.9666269732092508,
+      "learning_rate": 2.873241071404296e-06,
+      "loss": 0.3162,
+      "step": 2997
+    },
+    {
+      "epoch": 2.0284167794316645,
+      "grad_norm": 1.0139647308942759,
+      "learning_rate": 2.869678786618976e-06,
+      "loss": 0.3352,
+      "step": 2998
+    },
+    {
+      "epoch": 2.0290933694181326,
+      "grad_norm": 0.9972898048555547,
+      "learning_rate": 2.866117822517982e-06,
+      "loss": 0.3262,
+      "step": 2999
+    },
+    {
+      "epoch": 2.029769959404601,
+      "grad_norm": 0.9967207575777589,
+      "learning_rate": 2.86255818130892e-06,
+      "loss": 0.3294,
+      "step": 3000
+    },
+    {
+      "epoch": 2.030446549391069,
+      "grad_norm": 1.0790267719312128,
+      "learning_rate": 2.8589998651985775e-06,
+      "loss": 0.3399,
+      "step": 3001
+    },
+    {
+      "epoch": 2.031123139377537,
+      "grad_norm": 1.0059416399173047,
+      "learning_rate": 2.855442876392914e-06,
+      "loss": 0.3351,
+      "step": 3002
+    },
+    {
+      "epoch": 2.0317997293640055,
+      "grad_norm": 1.0123183878980389,
+      "learning_rate": 2.8518872170970758e-06,
+      "loss": 0.3146,
+      "step": 3003
+    },
+    {
+      "epoch": 2.0324763193504736,
+      "grad_norm": 1.0662257057660793,
+      "learning_rate": 2.848332889515375e-06,
+      "loss": 0.3343,
+      "step": 3004
+    },
+    {
+      "epoch": 2.033152909336942,
+      "grad_norm": 1.0578278376530537,
+      "learning_rate": 2.8447798958513082e-06,
+      "loss": 0.3338,
+      "step": 3005
+    },
+    {
+      "epoch": 2.03382949932341,
+      "grad_norm": 1.0526404579909534,
+      "learning_rate": 2.8412282383075362e-06,
+      "loss": 0.3319,
+      "step": 3006
+    },
+    {
+      "epoch": 2.034506089309878,
+      "grad_norm": 1.0372705102399975,
+      "learning_rate": 2.837677919085896e-06,
+      "loss": 0.3273,
+      "step": 3007
+    },
+    {
+      "epoch": 2.0351826792963466,
+      "grad_norm": 0.983954642138402,
+      "learning_rate": 2.8341289403873952e-06,
+      "loss": 0.3102,
+      "step": 3008
+    },
+    {
+      "epoch": 2.0358592692828146,
+      "grad_norm": 1.0214465256758594,
+      "learning_rate": 2.83058130441221e-06,
+      "loss": 0.3255,
+      "step": 3009
+    },
+    {
+      "epoch": 2.0365358592692826,
+      "grad_norm": 1.0569444044633176,
+      "learning_rate": 2.8270350133596824e-06,
+      "loss": 0.3398,
+      "step": 3010
+    },
+    {
+      "epoch": 2.037212449255751,
+      "grad_norm": 0.9854356222554426,
+      "learning_rate": 2.82349006942832e-06,
+      "loss": 0.3214,
+      "step": 3011
+    },
+    {
+      "epoch": 2.037889039242219,
+      "grad_norm": 1.0208905020757164,
+      "learning_rate": 2.8199464748157983e-06,
+      "loss": 0.328,
+      "step": 3012
+    },
+    {
+      "epoch": 2.0385656292286876,
+      "grad_norm": 0.9829155508559776,
+      "learning_rate": 2.816404231718958e-06,
+      "loss": 0.3116,
+      "step": 3013
+    },
+    {
+      "epoch": 2.0392422192151556,
+      "grad_norm": 1.0189846361053592,
+      "learning_rate": 2.8128633423337932e-06,
+      "loss": 0.3219,
+      "step": 3014
+    },
+    {
+      "epoch": 2.0399188092016236,
+      "grad_norm": 0.9756658654015941,
+      "learning_rate": 2.8093238088554676e-06,
+      "loss": 0.3164,
+      "step": 3015
+    },
+    {
+      "epoch": 2.040595399188092,
+      "grad_norm": 1.0397077430063677,
+      "learning_rate": 2.8057856334783006e-06,
+      "loss": 0.3316,
+      "step": 3016
+    },
+    {
+      "epoch": 2.04127198917456,
+      "grad_norm": 1.0399646537929785,
+      "learning_rate": 2.802248818395773e-06,
+      "loss": 0.3311,
+      "step": 3017
+    },
+    {
+      "epoch": 2.0419485791610286,
+      "grad_norm": 1.0206729775037218,
+      "learning_rate": 2.7987133658005174e-06,
+      "loss": 0.3213,
+      "step": 3018
+    },
+    {
+      "epoch": 2.0426251691474966,
+      "grad_norm": 1.063125117670267,
+      "learning_rate": 2.795179277884321e-06,
+      "loss": 0.3318,
+      "step": 3019
+    },
+    {
+      "epoch": 2.0433017591339646,
+      "grad_norm": 1.0226033067382583,
+      "learning_rate": 2.79164655683813e-06,
+      "loss": 0.32,
+      "step": 3020
+    },
+    {
+      "epoch": 2.043978349120433,
+      "grad_norm": 1.028570262460247,
+      "learning_rate": 2.788115204852042e-06,
+      "loss": 0.3211,
+      "step": 3021
+    },
+    {
+      "epoch": 2.044654939106901,
+      "grad_norm": 1.007828137941707,
+      "learning_rate": 2.7845852241153063e-06,
+      "loss": 0.3212,
+      "step": 3022
+    },
+    {
+      "epoch": 2.0453315290933696,
+      "grad_norm": 1.057705082850331,
+      "learning_rate": 2.781056616816319e-06,
+      "loss": 0.328,
+      "step": 3023
+    },
+    {
+      "epoch": 2.0460081190798376,
+      "grad_norm": 1.0025355670518719,
+      "learning_rate": 2.7775293851426233e-06,
+      "loss": 0.3166,
+      "step": 3024
+    },
+    {
+      "epoch": 2.0466847090663056,
+      "grad_norm": 1.0481239008378884,
+      "learning_rate": 2.7740035312809153e-06,
+      "loss": 0.3192,
+      "step": 3025
+    },
+    {
+      "epoch": 2.047361299052774,
+      "grad_norm": 1.0411246643951915,
+      "learning_rate": 2.7704790574170372e-06,
+      "loss": 0.3219,
+      "step": 3026
+    },
+    {
+      "epoch": 2.048037889039242,
+      "grad_norm": 1.02782978110486,
+      "learning_rate": 2.766955965735968e-06,
+      "loss": 0.3224,
+      "step": 3027
+    },
+    {
+      "epoch": 2.0487144790257106,
+      "grad_norm": 1.053754933522437,
+      "learning_rate": 2.7634342584218364e-06,
+      "loss": 0.3221,
+      "step": 3028
+    },
+    {
+      "epoch": 2.0493910690121786,
+      "grad_norm": 1.072464684504966,
+      "learning_rate": 2.759913937657912e-06,
+      "loss": 0.3417,
+      "step": 3029
+    },
+    {
+      "epoch": 2.0500676589986466,
+      "grad_norm": 1.0565733183676842,
+      "learning_rate": 2.7563950056266053e-06,
+      "loss": 0.3328,
+      "step": 3030
+    },
+    {
+      "epoch": 2.050744248985115,
+      "grad_norm": 1.0473899303853187,
+      "learning_rate": 2.752877464509463e-06,
+      "loss": 0.3203,
+      "step": 3031
+    },
+    {
+      "epoch": 2.051420838971583,
+      "grad_norm": 1.0455928615028476,
+      "learning_rate": 2.7493613164871678e-06,
+      "loss": 0.3341,
+      "step": 3032
+    },
+    {
+      "epoch": 2.0520974289580516,
+      "grad_norm": 1.0529692345026416,
+      "learning_rate": 2.745846563739546e-06,
+      "loss": 0.3256,
+      "step": 3033
+    },
+    {
+      "epoch": 2.0527740189445196,
+      "grad_norm": 0.991111229617875,
+      "learning_rate": 2.7423332084455543e-06,
+      "loss": 0.3151,
+      "step": 3034
+    },
+    {
+      "epoch": 2.0534506089309876,
+      "grad_norm": 1.0354432823377364,
+      "learning_rate": 2.7388212527832814e-06,
+      "loss": 0.3343,
+      "step": 3035
+    },
+    {
+      "epoch": 2.054127198917456,
+      "grad_norm": 1.0505842582985414,
+      "learning_rate": 2.7353106989299528e-06,
+      "loss": 0.3274,
+      "step": 3036
+    },
+    {
+      "epoch": 2.054803788903924,
+      "grad_norm": 1.0331790888807864,
+      "learning_rate": 2.731801549061923e-06,
+      "loss": 0.3185,
+      "step": 3037
+    },
+    {
+      "epoch": 2.0554803788903926,
+      "grad_norm": 1.0306134117878722,
+      "learning_rate": 2.7282938053546727e-06,
+      "loss": 0.3262,
+      "step": 3038
+    },
+    {
+      "epoch": 2.0561569688768606,
+      "grad_norm": 1.0289563017047423,
+      "learning_rate": 2.7247874699828186e-06,
+      "loss": 0.3202,
+      "step": 3039
+    },
+    {
+      "epoch": 2.0568335588633286,
+      "grad_norm": 1.0242389485495078,
+      "learning_rate": 2.7212825451200942e-06,
+      "loss": 0.3264,
+      "step": 3040
+    },
+    {
+      "epoch": 2.057510148849797,
+      "grad_norm": 0.9845036838281035,
+      "learning_rate": 2.7177790329393674e-06,
+      "loss": 0.3104,
+      "step": 3041
+    },
+    {
+      "epoch": 2.058186738836265,
+      "grad_norm": 1.0301246381202998,
+      "learning_rate": 2.7142769356126258e-06,
+      "loss": 0.325,
+      "step": 3042
+    },
+    {
+      "epoch": 2.0588633288227336,
+      "grad_norm": 1.030844098927739,
+      "learning_rate": 2.710776255310984e-06,
+      "loss": 0.3184,
+      "step": 3043
+    },
+    {
+      "epoch": 2.0595399188092016,
+      "grad_norm": 1.0694238947846768,
+      "learning_rate": 2.7072769942046716e-06,
+      "loss": 0.3288,
+      "step": 3044
+    },
+    {
+      "epoch": 2.0602165087956696,
+      "grad_norm": 1.0592291058504668,
+      "learning_rate": 2.7037791544630414e-06,
+      "loss": 0.3336,
+      "step": 3045
+    },
+    {
+      "epoch": 2.060893098782138,
+      "grad_norm": 1.0218749998590817,
+      "learning_rate": 2.700282738254567e-06,
+      "loss": 0.3309,
+      "step": 3046
+    },
+    {
+      "epoch": 2.061569688768606,
+      "grad_norm": 1.016657007346522,
+      "learning_rate": 2.6967877477468394e-06,
+      "loss": 0.3148,
+      "step": 3047
+    },
+    {
+      "epoch": 2.0622462787550746,
+      "grad_norm": 1.020059008465994,
+      "learning_rate": 2.693294185106562e-06,
+      "loss": 0.3068,
+      "step": 3048
+    },
+    {
+      "epoch": 2.0629228687415426,
+      "grad_norm": 1.055423523804485,
+      "learning_rate": 2.689802052499555e-06,
+      "loss": 0.3365,
+      "step": 3049
+    },
+    {
+      "epoch": 2.0635994587280106,
+      "grad_norm": 1.0444281385591616,
+      "learning_rate": 2.686311352090756e-06,
+      "loss": 0.3245,
+      "step": 3050
+    },
+    {
+      "epoch": 2.064276048714479,
+      "grad_norm": 1.0947021393057752,
+      "learning_rate": 2.682822086044206e-06,
+      "loss": 0.3406,
+      "step": 3051
+    },
+    {
+      "epoch": 2.064952638700947,
+      "grad_norm": 1.0466321232172768,
+      "learning_rate": 2.6793342565230675e-06,
+      "loss": 0.3209,
+      "step": 3052
+    },
+    {
+      "epoch": 2.0656292286874156,
+      "grad_norm": 1.0259706505060404,
+      "learning_rate": 2.6758478656896015e-06,
+      "loss": 0.31,
+      "step": 3053
+    },
+    {
+      "epoch": 2.0663058186738836,
+      "grad_norm": 1.0604318032232756,
+      "learning_rate": 2.6723629157051844e-06,
+      "loss": 0.3257,
+      "step": 3054
+    },
+    {
+      "epoch": 2.0669824086603517,
+      "grad_norm": 1.0293439262553075,
+      "learning_rate": 2.6688794087302993e-06,
+      "loss": 0.3142,
+      "step": 3055
+    },
+    {
+      "epoch": 2.06765899864682,
+      "grad_norm": 1.0161790362445395,
+      "learning_rate": 2.66539734692453e-06,
+      "loss": 0.314,
+      "step": 3056
+    },
+    {
+      "epoch": 2.068335588633288,
+      "grad_norm": 1.0612915633711357,
+      "learning_rate": 2.66191673244657e-06,
+      "loss": 0.32,
+      "step": 3057
+    },
+    {
+      "epoch": 2.0690121786197566,
+      "grad_norm": 1.048609064297326,
+      "learning_rate": 2.658437567454209e-06,
+      "loss": 0.3259,
+      "step": 3058
+    },
+    {
+      "epoch": 2.0696887686062246,
+      "grad_norm": 1.0385520165410107,
+      "learning_rate": 2.6549598541043433e-06,
+      "loss": 0.3186,
+      "step": 3059
+    },
+    {
+      "epoch": 2.0703653585926927,
+      "grad_norm": 0.9993832964664853,
+      "learning_rate": 2.6514835945529706e-06,
+      "loss": 0.3191,
+      "step": 3060
+    },
+    {
+      "epoch": 2.071041948579161,
+      "grad_norm": 1.0049445279801426,
+      "learning_rate": 2.64800879095518e-06,
+      "loss": 0.307,
+      "step": 3061
+    },
+    {
+      "epoch": 2.071718538565629,
+      "grad_norm": 1.0329942836922585,
+      "learning_rate": 2.644535445465164e-06,
+      "loss": 0.311,
+      "step": 3062
+    },
+    {
+      "epoch": 2.0723951285520976,
+      "grad_norm": 1.0773788016239711,
+      "learning_rate": 2.641063560236212e-06,
+      "loss": 0.3234,
+      "step": 3063
+    },
+    {
+      "epoch": 2.0730717185385656,
+      "grad_norm": 1.0514388681976856,
+      "learning_rate": 2.637593137420702e-06,
+      "loss": 0.3132,
+      "step": 3064
+    },
+    {
+      "epoch": 2.0737483085250337,
+      "grad_norm": 1.060199126971947,
+      "learning_rate": 2.6341241791701126e-06,
+      "loss": 0.3237,
+      "step": 3065
+    },
+    {
+      "epoch": 2.074424898511502,
+      "grad_norm": 1.0328324282426826,
+      "learning_rate": 2.6306566876350072e-06,
+      "loss": 0.3056,
+      "step": 3066
+    },
+    {
+      "epoch": 2.07510148849797,
+      "grad_norm": 1.0398636123520217,
+      "learning_rate": 2.627190664965046e-06,
+      "loss": 0.3251,
+      "step": 3067
+    },
+    {
+      "epoch": 2.0757780784844386,
+      "grad_norm": 1.0276240513510593,
+      "learning_rate": 2.623726113308977e-06,
+      "loss": 0.3218,
+      "step": 3068
+    },
+    {
+      "epoch": 2.0764546684709067,
+      "grad_norm": 1.0142210211703218,
+      "learning_rate": 2.6202630348146323e-06,
+      "loss": 0.3082,
+      "step": 3069
+    },
+    {
+      "epoch": 2.0771312584573747,
+      "grad_norm": 1.028746549621249,
+      "learning_rate": 2.616801431628938e-06,
+      "loss": 0.3153,
+      "step": 3070
+    },
+    {
+      "epoch": 2.077807848443843,
+      "grad_norm": 1.0925764452314193,
+      "learning_rate": 2.613341305897898e-06,
+      "loss": 0.3198,
+      "step": 3071
+    },
+    {
+      "epoch": 2.078484438430311,
+      "grad_norm": 1.0249065357540452,
+      "learning_rate": 2.609882659766605e-06,
+      "loss": 0.3219,
+      "step": 3072
+    },
+    {
+      "epoch": 2.0791610284167796,
+      "grad_norm": 1.0530121262543295,
+      "learning_rate": 2.6064254953792344e-06,
+      "loss": 0.3128,
+      "step": 3073
+    },
+    {
+      "epoch": 2.0798376184032477,
+      "grad_norm": 1.0569473958278888,
+      "learning_rate": 2.6029698148790392e-06,
+      "loss": 0.3343,
+      "step": 3074
+    },
+    {
+      "epoch": 2.0805142083897157,
+      "grad_norm": 1.0286464058137517,
+      "learning_rate": 2.5995156204083573e-06,
+      "loss": 0.3124,
+      "step": 3075
+    },
+    {
+      "epoch": 2.081190798376184,
+      "grad_norm": 1.0300719279849928,
+      "learning_rate": 2.5960629141086014e-06,
+      "loss": 0.3094,
+      "step": 3076
+    },
+    {
+      "epoch": 2.081867388362652,
+      "grad_norm": 1.0508843236502377,
+      "learning_rate": 2.5926116981202688e-06,
+      "loss": 0.3163,
+      "step": 3077
+    },
+    {
+      "epoch": 2.0825439783491206,
+      "grad_norm": 1.0466602530117037,
+      "learning_rate": 2.5891619745829184e-06,
+      "loss": 0.3213,
+      "step": 3078
+    },
+    {
+      "epoch": 2.0832205683355887,
+      "grad_norm": 1.0550055994518281,
+      "learning_rate": 2.585713745635197e-06,
+      "loss": 0.3182,
+      "step": 3079
+    },
+    {
+      "epoch": 2.0838971583220567,
+      "grad_norm": 1.0494429080200303,
+      "learning_rate": 2.5822670134148216e-06,
+      "loss": 0.3207,
+      "step": 3080
+    },
+    {
+      "epoch": 2.084573748308525,
+      "grad_norm": 1.07663821750747,
+      "learning_rate": 2.5788217800585812e-06,
+      "loss": 0.3307,
+      "step": 3081
+    },
+    {
+      "epoch": 2.085250338294993,
+      "grad_norm": 1.0633110377892574,
+      "learning_rate": 2.5753780477023314e-06,
+      "loss": 0.3269,
+      "step": 3082
+    },
+    {
+      "epoch": 2.0859269282814616,
+      "grad_norm": 1.0242153916380263,
+      "learning_rate": 2.571935818481005e-06,
+      "loss": 0.3129,
+      "step": 3083
+    },
+    {
+      "epoch": 2.0866035182679297,
+      "grad_norm": 1.0754341891045889,
+      "learning_rate": 2.5684950945285937e-06,
+      "loss": 0.3314,
+      "step": 3084
+    },
+    {
+      "epoch": 2.0872801082543977,
+      "grad_norm": 1.0115318342883304,
+      "learning_rate": 2.5650558779781635e-06,
+      "loss": 0.3218,
+      "step": 3085
+    },
+    {
+      "epoch": 2.087956698240866,
+      "grad_norm": 1.0557808811469453,
+      "learning_rate": 2.5616181709618447e-06,
+      "loss": 0.3083,
+      "step": 3086
+    },
+    {
+      "epoch": 2.088633288227334,
+      "grad_norm": 1.1107005877830372,
+      "learning_rate": 2.558181975610827e-06,
+      "loss": 0.3409,
+      "step": 3087
+    },
+    {
+      "epoch": 2.089309878213802,
+      "grad_norm": 1.0489520532540877,
+      "learning_rate": 2.5547472940553685e-06,
+      "loss": 0.3235,
+      "step": 3088
+    },
+    {
+      "epoch": 2.0899864682002707,
+      "grad_norm": 1.0736035785016134,
+      "learning_rate": 2.551314128424788e-06,
+      "loss": 0.3362,
+      "step": 3089
+    },
+    {
+      "epoch": 2.0906630581867387,
+      "grad_norm": 1.0545894547228865,
+      "learning_rate": 2.5478824808474613e-06,
+      "loss": 0.3155,
+      "step": 3090
+    },
+    {
+      "epoch": 2.091339648173207,
+      "grad_norm": 1.0130201067249343,
+      "learning_rate": 2.5444523534508225e-06,
+      "loss": 0.3,
+      "step": 3091
+    },
+    {
+      "epoch": 2.092016238159675,
+      "grad_norm": 1.0883982795950715,
+      "learning_rate": 2.5410237483613685e-06,
+      "loss": 0.3324,
+      "step": 3092
+    },
+    {
+      "epoch": 2.092692828146143,
+      "grad_norm": 1.035595955465196,
+      "learning_rate": 2.53759666770465e-06,
+      "loss": 0.3082,
+      "step": 3093
+    },
+    {
+      "epoch": 2.0933694181326117,
+      "grad_norm": 1.089704057262123,
+      "learning_rate": 2.5341711136052728e-06,
+      "loss": 0.3301,
+      "step": 3094
+    },
+    {
+      "epoch": 2.0940460081190797,
+      "grad_norm": 1.0588476791242767,
+      "learning_rate": 2.530747088186893e-06,
+      "loss": 0.3172,
+      "step": 3095
+    },
+    {
+      "epoch": 2.094722598105548,
+      "grad_norm": 1.0835982309517147,
+      "learning_rate": 2.527324593572223e-06,
+      "loss": 0.3385,
+      "step": 3096
+    },
+    {
+      "epoch": 2.095399188092016,
+      "grad_norm": 1.0416893482873841,
+      "learning_rate": 2.523903631883028e-06,
+      "loss": 0.3247,
+      "step": 3097
+    },
+    {
+      "epoch": 2.096075778078484,
+      "grad_norm": 1.043808058353486,
+      "learning_rate": 2.520484205240116e-06,
+      "loss": 0.3177,
+      "step": 3098
+    },
+    {
+      "epoch": 2.0967523680649527,
+      "grad_norm": 1.0492203272402414,
+      "learning_rate": 2.517066315763348e-06,
+      "loss": 0.3256,
+      "step": 3099
+    },
+    {
+      "epoch": 2.0974289580514207,
+      "grad_norm": 1.0702990144584885,
+      "learning_rate": 2.5136499655716306e-06,
+      "loss": 0.3219,
+      "step": 3100
+    },
+    {
+      "epoch": 2.098105548037889,
+      "grad_norm": 1.0308577581436713,
+      "learning_rate": 2.5102351567829187e-06,
+      "loss": 0.3161,
+      "step": 3101
+    },
+    {
+      "epoch": 2.098782138024357,
+      "grad_norm": 1.054329882960072,
+      "learning_rate": 2.5068218915142093e-06,
+      "loss": 0.3235,
+      "step": 3102
+    },
+    {
+      "epoch": 2.0994587280108252,
+      "grad_norm": 1.1157230257088258,
+      "learning_rate": 2.503410171881544e-06,
+      "loss": 0.3348,
+      "step": 3103
+    },
+    {
+      "epoch": 2.1001353179972937,
+      "grad_norm": 1.067196763066268,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.3196,
+      "step": 3104
+    },
+    {
+      "epoch": 2.1008119079837617,
+      "grad_norm": 1.0844946213056303,
+      "learning_rate": 2.496591377983706e-06,
+      "loss": 0.3288,
+      "step": 3105
+    },
+    {
+      "epoch": 2.10148849797023,
+      "grad_norm": 1.1023322609921864,
+      "learning_rate": 2.49318430794582e-06,
+      "loss": 0.3162,
+      "step": 3106
+    },
+    {
+      "epoch": 2.102165087956698,
+      "grad_norm": 1.1388866487761733,
+      "learning_rate": 2.4897787919985457e-06,
+      "loss": 0.3422,
+      "step": 3107
+    },
+    {
+      "epoch": 2.1028416779431662,
+      "grad_norm": 1.1183577118989485,
+      "learning_rate": 2.4863748322531144e-06,
+      "loss": 0.3356,
+      "step": 3108
+    },
+    {
+      "epoch": 2.1035182679296347,
+      "grad_norm": 1.0407826855860005,
+      "learning_rate": 2.4829724308198003e-06,
+      "loss": 0.3128,
+      "step": 3109
+    },
+    {
+      "epoch": 2.1041948579161027,
+      "grad_norm": 1.0743978433089592,
+      "learning_rate": 2.4795715898079116e-06,
+      "loss": 0.3233,
+      "step": 3110
+    },
+    {
+      "epoch": 2.104871447902571,
+      "grad_norm": 1.0903473041474492,
+      "learning_rate": 2.476172311325783e-06,
+      "loss": 0.3319,
+      "step": 3111
+    },
+    {
+      "epoch": 2.105548037889039,
+      "grad_norm": 1.024409836189495,
+      "learning_rate": 2.472774597480783e-06,
+      "loss": 0.3026,
+      "step": 3112
+    },
+    {
+      "epoch": 2.1062246278755072,
+      "grad_norm": 1.064844122177811,
+      "learning_rate": 2.4693784503793128e-06,
+      "loss": 0.3237,
+      "step": 3113
+    },
+    {
+      "epoch": 2.1069012178619757,
+      "grad_norm": 1.083568204903442,
+      "learning_rate": 2.4659838721268005e-06,
+      "loss": 0.3303,
+      "step": 3114
+    },
+    {
+      "epoch": 2.1075778078484437,
+      "grad_norm": 1.0366527980044367,
+      "learning_rate": 2.462590864827703e-06,
+      "loss": 0.3202,
+      "step": 3115
+    },
+    {
+      "epoch": 2.108254397834912,
+      "grad_norm": 1.062597365245354,
+      "learning_rate": 2.4591994305854988e-06,
+      "loss": 0.321,
+      "step": 3116
+    },
+    {
+      "epoch": 2.10893098782138,
+      "grad_norm": 1.0989405982152978,
+      "learning_rate": 2.4558095715026975e-06,
+      "loss": 0.3188,
+      "step": 3117
+    },
+    {
+      "epoch": 2.1096075778078482,
+      "grad_norm": 1.0543491430785725,
+      "learning_rate": 2.4524212896808265e-06,
+      "loss": 0.3195,
+      "step": 3118
+    },
+    {
+      "epoch": 2.1102841677943167,
+      "grad_norm": 1.0698713036562926,
+      "learning_rate": 2.4490345872204403e-06,
+      "loss": 0.3211,
+      "step": 3119
+    },
+    {
+      "epoch": 2.1109607577807847,
+      "grad_norm": 1.0829067748911252,
+      "learning_rate": 2.4456494662211082e-06,
+      "loss": 0.3219,
+      "step": 3120
+    },
+    {
+      "epoch": 2.111637347767253,
+      "grad_norm": 1.0624932575492023,
+      "learning_rate": 2.442265928781426e-06,
+      "loss": 0.3208,
+      "step": 3121
+    },
+    {
+      "epoch": 2.1123139377537212,
+      "grad_norm": 1.0750870009935214,
+      "learning_rate": 2.438883976999003e-06,
+      "loss": 0.333,
+      "step": 3122
+    },
+    {
+      "epoch": 2.1129905277401893,
+      "grad_norm": 1.082437516812866,
+      "learning_rate": 2.43550361297047e-06,
+      "loss": 0.325,
+      "step": 3123
+    },
+    {
+      "epoch": 2.1136671177266577,
+      "grad_norm": 1.0730007247696887,
+      "learning_rate": 2.4321248387914677e-06,
+      "loss": 0.3374,
+      "step": 3124
+    },
+    {
+      "epoch": 2.1143437077131257,
+      "grad_norm": 1.0971426609597585,
+      "learning_rate": 2.4287476565566525e-06,
+      "loss": 0.319,
+      "step": 3125
+    },
+    {
+      "epoch": 2.115020297699594,
+      "grad_norm": 1.0933027030657891,
+      "learning_rate": 2.4253720683596976e-06,
+      "loss": 0.3204,
+      "step": 3126
+    },
+    {
+      "epoch": 2.1156968876860622,
+      "grad_norm": 1.092757872470542,
+      "learning_rate": 2.421998076293285e-06,
+      "loss": 0.3206,
+      "step": 3127
+    },
+    {
+      "epoch": 2.1163734776725303,
+      "grad_norm": 1.027898455561416,
+      "learning_rate": 2.4186256824491106e-06,
+      "loss": 0.3283,
+      "step": 3128
+    },
+    {
+      "epoch": 2.1170500676589987,
+      "grad_norm": 1.0309936626773026,
+      "learning_rate": 2.4152548889178722e-06,
+      "loss": 0.3196,
+      "step": 3129
+    },
+    {
+      "epoch": 2.1177266576454667,
+      "grad_norm": 1.035713973438815,
+      "learning_rate": 2.4118856977892846e-06,
+      "loss": 0.308,
+      "step": 3130
+    },
+    {
+      "epoch": 2.118403247631935,
+      "grad_norm": 1.0300287259100047,
+      "learning_rate": 2.4085181111520607e-06,
+      "loss": 0.3093,
+      "step": 3131
+    },
+    {
+      "epoch": 2.1190798376184032,
+      "grad_norm": 1.0390574413392113,
+      "learning_rate": 2.4051521310939258e-06,
+      "loss": 0.307,
+      "step": 3132
+    },
+    {
+      "epoch": 2.1197564276048713,
+      "grad_norm": 1.1116496454790783,
+      "learning_rate": 2.401787759701603e-06,
+      "loss": 0.3269,
+      "step": 3133
+    },
+    {
+      "epoch": 2.1204330175913397,
+      "grad_norm": 1.0872385577378534,
+      "learning_rate": 2.3984249990608237e-06,
+      "loss": 0.3262,
+      "step": 3134
+    },
+    {
+      "epoch": 2.1211096075778078,
+      "grad_norm": 1.0969175153615704,
+      "learning_rate": 2.3950638512563173e-06,
+      "loss": 0.3279,
+      "step": 3135
+    },
+    {
+      "epoch": 2.121786197564276,
+      "grad_norm": 1.0932050847858432,
+      "learning_rate": 2.3917043183718162e-06,
+      "loss": 0.3154,
+      "step": 3136
+    },
+    {
+      "epoch": 2.1224627875507442,
+      "grad_norm": 1.103043564866908,
+      "learning_rate": 2.3883464024900484e-06,
+      "loss": 0.3292,
+      "step": 3137
+    },
+    {
+      "epoch": 2.1231393775372123,
+      "grad_norm": 1.1119240065901193,
+      "learning_rate": 2.3849901056927383e-06,
+      "loss": 0.3213,
+      "step": 3138
+    },
+    {
+      "epoch": 2.1238159675236807,
+      "grad_norm": 1.0420421027785127,
+      "learning_rate": 2.381635430060611e-06,
+      "loss": 0.3075,
+      "step": 3139
+    },
+    {
+      "epoch": 2.1244925575101488,
+      "grad_norm": 1.0646788200773543,
+      "learning_rate": 2.3782823776733866e-06,
+      "loss": 0.3227,
+      "step": 3140
+    },
+    {
+      "epoch": 2.1251691474966172,
+      "grad_norm": 1.056405175632645,
+      "learning_rate": 2.374930950609773e-06,
+      "loss": 0.3161,
+      "step": 3141
+    },
+    {
+      "epoch": 2.1258457374830853,
+      "grad_norm": 1.0414292704194887,
+      "learning_rate": 2.371581150947476e-06,
+      "loss": 0.3174,
+      "step": 3142
+    },
+    {
+      "epoch": 2.1265223274695533,
+      "grad_norm": 1.1142086471166972,
+      "learning_rate": 2.368232980763194e-06,
+      "loss": 0.3307,
+      "step": 3143
+    },
+    {
+      "epoch": 2.1271989174560217,
+      "grad_norm": 1.0212596172221553,
+      "learning_rate": 2.364886442132606e-06,
+      "loss": 0.3211,
+      "step": 3144
+    },
+    {
+      "epoch": 2.1278755074424898,
+      "grad_norm": 1.1024813366872834,
+      "learning_rate": 2.361541537130392e-06,
+      "loss": 0.3304,
+      "step": 3145
+    },
+    {
+      "epoch": 2.1285520974289582,
+      "grad_norm": 1.0885345365994723,
+      "learning_rate": 2.358198267830206e-06,
+      "loss": 0.3317,
+      "step": 3146
+    },
+    {
+      "epoch": 2.1292286874154263,
+      "grad_norm": 1.0637804849114936,
+      "learning_rate": 2.3548566363046993e-06,
+      "loss": 0.3187,
+      "step": 3147
+    },
+    {
+      "epoch": 2.1299052774018943,
+      "grad_norm": 1.056700279491071,
+      "learning_rate": 2.351516644625502e-06,
+      "loss": 0.3236,
+      "step": 3148
+    },
+    {
+      "epoch": 2.1305818673883627,
+      "grad_norm": 1.0990143212098609,
+      "learning_rate": 2.3481782948632317e-06,
+      "loss": 0.3177,
+      "step": 3149
+    },
+    {
+      "epoch": 2.1312584573748308,
+      "grad_norm": 1.046723744246787,
+      "learning_rate": 2.344841589087482e-06,
+      "loss": 0.3091,
+      "step": 3150
+    },
+    {
+      "epoch": 2.1319350473612992,
+      "grad_norm": 1.0384970244552318,
+      "learning_rate": 2.34150652936683e-06,
+      "loss": 0.3044,
+      "step": 3151
+    },
+    {
+      "epoch": 2.1326116373477673,
+      "grad_norm": 1.110070757716026,
+      "learning_rate": 2.3381731177688346e-06,
+      "loss": 0.3253,
+      "step": 3152
+    },
+    {
+      "epoch": 2.1332882273342353,
+      "grad_norm": 1.0876101006040824,
+      "learning_rate": 2.3348413563600324e-06,
+      "loss": 0.3246,
+      "step": 3153
+    },
+    {
+      "epoch": 2.1339648173207038,
+      "grad_norm": 1.118999567282542,
+      "learning_rate": 2.331511247205933e-06,
+      "loss": 0.3267,
+      "step": 3154
+    },
+    {
+      "epoch": 2.134641407307172,
+      "grad_norm": 1.0663102742420723,
+      "learning_rate": 2.3281827923710265e-06,
+      "loss": 0.3203,
+      "step": 3155
+    },
+    {
+      "epoch": 2.1353179972936402,
+      "grad_norm": 1.065598797598251,
+      "learning_rate": 2.324855993918775e-06,
+      "loss": 0.327,
+      "step": 3156
+    },
+    {
+      "epoch": 2.1359945872801083,
+      "grad_norm": 1.0782751262598218,
+      "learning_rate": 2.321530853911616e-06,
+      "loss": 0.3328,
+      "step": 3157
+    },
+    {
+      "epoch": 2.1366711772665763,
+      "grad_norm": 1.071017983244043,
+      "learning_rate": 2.318207374410956e-06,
+      "loss": 0.3194,
+      "step": 3158
+    },
+    {
+      "epoch": 2.1373477672530448,
+      "grad_norm": 1.084835448741031,
+      "learning_rate": 2.3148855574771706e-06,
+      "loss": 0.3166,
+      "step": 3159
+    },
+    {
+      "epoch": 2.138024357239513,
+      "grad_norm": 1.0504154783898043,
+      "learning_rate": 2.3115654051696097e-06,
+      "loss": 0.3186,
+      "step": 3160
+    },
+    {
+      "epoch": 2.1387009472259813,
+      "grad_norm": 1.0803525128154143,
+      "learning_rate": 2.3082469195465893e-06,
+      "loss": 0.3214,
+      "step": 3161
+    },
+    {
+      "epoch": 2.1393775372124493,
+      "grad_norm": 1.1031238529251692,
+      "learning_rate": 2.304930102665389e-06,
+      "loss": 0.316,
+      "step": 3162
+    },
+    {
+      "epoch": 2.1400541271989173,
+      "grad_norm": 1.0611509470802538,
+      "learning_rate": 2.3016149565822608e-06,
+      "loss": 0.3191,
+      "step": 3163
+    },
+    {
+      "epoch": 2.1407307171853858,
+      "grad_norm": 1.1840898020737045,
+      "learning_rate": 2.2983014833524115e-06,
+      "loss": 0.3436,
+      "step": 3164
+    },
+    {
+      "epoch": 2.141407307171854,
+      "grad_norm": 1.100638275146891,
+      "learning_rate": 2.2949896850300186e-06,
+      "loss": 0.3245,
+      "step": 3165
+    },
+    {
+      "epoch": 2.1420838971583223,
+      "grad_norm": 1.0668940636710917,
+      "learning_rate": 2.2916795636682197e-06,
+      "loss": 0.3131,
+      "step": 3166
+    },
+    {
+      "epoch": 2.1427604871447903,
+      "grad_norm": 1.0747765392926185,
+      "learning_rate": 2.288371121319109e-06,
+      "loss": 0.3116,
+      "step": 3167
+    },
+    {
+      "epoch": 2.1434370771312583,
+      "grad_norm": 1.063660703391759,
+      "learning_rate": 2.2850643600337435e-06,
+      "loss": 0.3123,
+      "step": 3168
+    },
+    {
+      "epoch": 2.1441136671177268,
+      "grad_norm": 1.1520263523555183,
+      "learning_rate": 2.281759281862137e-06,
+      "loss": 0.3456,
+      "step": 3169
+    },
+    {
+      "epoch": 2.144790257104195,
+      "grad_norm": 1.0397793529719004,
+      "learning_rate": 2.278455888853262e-06,
+      "loss": 0.3139,
+      "step": 3170
+    },
+    {
+      "epoch": 2.1454668470906633,
+      "grad_norm": 1.1764636653039535,
+      "learning_rate": 2.2751541830550417e-06,
+      "loss": 0.3477,
+      "step": 3171
+    },
+    {
+      "epoch": 2.1461434370771313,
+      "grad_norm": 1.096044087493195,
+      "learning_rate": 2.2718541665143546e-06,
+      "loss": 0.3176,
+      "step": 3172
+    },
+    {
+      "epoch": 2.1468200270635993,
+      "grad_norm": 1.0971650520919818,
+      "learning_rate": 2.2685558412770344e-06,
+      "loss": 0.3204,
+      "step": 3173
+    },
+    {
+      "epoch": 2.147496617050068,
+      "grad_norm": 1.038875194934764,
+      "learning_rate": 2.265259209387867e-06,
+      "loss": 0.3108,
+      "step": 3174
+    },
+    {
+      "epoch": 2.148173207036536,
+      "grad_norm": 1.0673465086445297,
+      "learning_rate": 2.261964272890582e-06,
+      "loss": 0.3119,
+      "step": 3175
+    },
+    {
+      "epoch": 2.1488497970230043,
+      "grad_norm": 1.084616288360263,
+      "learning_rate": 2.258671033827866e-06,
+      "loss": 0.3166,
+      "step": 3176
+    },
+    {
+      "epoch": 2.1495263870094723,
+      "grad_norm": 1.061414923306606,
+      "learning_rate": 2.2553794942413506e-06,
+      "loss": 0.3116,
+      "step": 3177
+    },
+    {
+      "epoch": 2.1502029769959403,
+      "grad_norm": 1.105324830905539,
+      "learning_rate": 2.2520896561716086e-06,
+      "loss": 0.3187,
+      "step": 3178
+    },
+    {
+      "epoch": 2.150879566982409,
+      "grad_norm": 1.0254265156658997,
+      "learning_rate": 2.248801521658167e-06,
+      "loss": 0.303,
+      "step": 3179
+    },
+    {
+      "epoch": 2.151556156968877,
+      "grad_norm": 1.1040647514313622,
+      "learning_rate": 2.245515092739488e-06,
+      "loss": 0.3246,
+      "step": 3180
+    },
+    {
+      "epoch": 2.1522327469553453,
+      "grad_norm": 1.1083311557638789,
+      "learning_rate": 2.242230371452982e-06,
+      "loss": 0.3169,
+      "step": 3181
+    },
+    {
+      "epoch": 2.1529093369418133,
+      "grad_norm": 1.076610648441405,
+      "learning_rate": 2.2389473598349994e-06,
+      "loss": 0.3179,
+      "step": 3182
+    },
+    {
+      "epoch": 2.1535859269282813,
+      "grad_norm": 1.048408774628365,
+      "learning_rate": 2.2356660599208335e-06,
+      "loss": 0.3018,
+      "step": 3183
+    },
+    {
+      "epoch": 2.15426251691475,
+      "grad_norm": 1.1348917295010865,
+      "learning_rate": 2.2323864737447067e-06,
+      "loss": 0.3282,
+      "step": 3184
+    },
+    {
+      "epoch": 2.154939106901218,
+      "grad_norm": 1.1325651005433037,
+      "learning_rate": 2.229108603339789e-06,
+      "loss": 0.3279,
+      "step": 3185
+    },
+    {
+      "epoch": 2.1556156968876863,
+      "grad_norm": 1.0881271183176577,
+      "learning_rate": 2.2258324507381834e-06,
+      "loss": 0.3198,
+      "step": 3186
+    },
+    {
+      "epoch": 2.1562922868741543,
+      "grad_norm": 1.09201320779718,
+      "learning_rate": 2.2225580179709303e-06,
+      "loss": 0.3227,
+      "step": 3187
+    },
+    {
+      "epoch": 2.1569688768606223,
+      "grad_norm": 1.0944716419384257,
+      "learning_rate": 2.219285307067997e-06,
+      "loss": 0.3277,
+      "step": 3188
+    },
+    {
+      "epoch": 2.157645466847091,
+      "grad_norm": 1.0874606281070929,
+      "learning_rate": 2.2160143200582906e-06,
+      "loss": 0.3159,
+      "step": 3189
+    },
+    {
+      "epoch": 2.158322056833559,
+      "grad_norm": 1.1209681196560373,
+      "learning_rate": 2.2127450589696475e-06,
+      "loss": 0.3168,
+      "step": 3190
+    },
+    {
+      "epoch": 2.1589986468200273,
+      "grad_norm": 1.098721684394217,
+      "learning_rate": 2.209477525828831e-06,
+      "loss": 0.3255,
+      "step": 3191
+    },
+    {
+      "epoch": 2.1596752368064953,
+      "grad_norm": 1.1409105232572194,
+      "learning_rate": 2.2062117226615375e-06,
+      "loss": 0.3271,
+      "step": 3192
+    },
+    {
+      "epoch": 2.1603518267929633,
+      "grad_norm": 1.0909547430292323,
+      "learning_rate": 2.202947651492387e-06,
+      "loss": 0.3185,
+      "step": 3193
+    },
+    {
+      "epoch": 2.161028416779432,
+      "grad_norm": 1.026728556296274,
+      "learning_rate": 2.1996853143449285e-06,
+      "loss": 0.3061,
+      "step": 3194
+    },
+    {
+      "epoch": 2.1617050067659,
+      "grad_norm": 1.0788020345952956,
+      "learning_rate": 2.1964247132416373e-06,
+      "loss": 0.3146,
+      "step": 3195
+    },
+    {
+      "epoch": 2.1623815967523683,
+      "grad_norm": 1.0693341232447195,
+      "learning_rate": 2.1931658502039067e-06,
+      "loss": 0.3279,
+      "step": 3196
+    },
+    {
+      "epoch": 2.1630581867388363,
+      "grad_norm": 1.1088375426769643,
+      "learning_rate": 2.1899087272520596e-06,
+      "loss": 0.3271,
+      "step": 3197
+    },
+    {
+      "epoch": 2.1637347767253043,
+      "grad_norm": 1.102380907668985,
+      "learning_rate": 2.186653346405333e-06,
+      "loss": 0.3209,
+      "step": 3198
+    },
+    {
+      "epoch": 2.164411366711773,
+      "grad_norm": 1.0593601678630866,
+      "learning_rate": 2.1833997096818897e-06,
+      "loss": 0.316,
+      "step": 3199
+    },
+    {
+      "epoch": 2.165087956698241,
+      "grad_norm": 1.1126938288072634,
+      "learning_rate": 2.1801478190988107e-06,
+      "loss": 0.3289,
+      "step": 3200
+    },
+    {
+      "epoch": 2.1657645466847093,
+      "grad_norm": 1.0980632775517076,
+      "learning_rate": 2.1768976766720896e-06,
+      "loss": 0.3212,
+      "step": 3201
+    },
+    {
+      "epoch": 2.1664411366711773,
+      "grad_norm": 1.0952565927822706,
+      "learning_rate": 2.1736492844166406e-06,
+      "loss": 0.3109,
+      "step": 3202
+    },
+    {
+      "epoch": 2.1671177266576453,
+      "grad_norm": 1.0878279545464584,
+      "learning_rate": 2.170402644346294e-06,
+      "loss": 0.3221,
+      "step": 3203
+    },
+    {
+      "epoch": 2.167794316644114,
+      "grad_norm": 1.173408318144088,
+      "learning_rate": 2.16715775847379e-06,
+      "loss": 0.3411,
+      "step": 3204
+    },
+    {
+      "epoch": 2.168470906630582,
+      "grad_norm": 1.068753888253633,
+      "learning_rate": 2.163914628810781e-06,
+      "loss": 0.3158,
+      "step": 3205
+    },
+    {
+      "epoch": 2.16914749661705,
+      "grad_norm": 1.1375964881134428,
+      "learning_rate": 2.1606732573678344e-06,
+      "loss": 0.3256,
+      "step": 3206
+    },
+    {
+      "epoch": 2.1698240866035183,
+      "grad_norm": 1.054216379354541,
+      "learning_rate": 2.157433646154426e-06,
+      "loss": 0.3102,
+      "step": 3207
+    },
+    {
+      "epoch": 2.1705006765899864,
+      "grad_norm": 1.0995668409591008,
+      "learning_rate": 2.154195797178941e-06,
+      "loss": 0.318,
+      "step": 3208
+    },
+    {
+      "epoch": 2.171177266576455,
+      "grad_norm": 1.0364530645555354,
+      "learning_rate": 2.1509597124486693e-06,
+      "loss": 0.308,
+      "step": 3209
+    },
+    {
+      "epoch": 2.171853856562923,
+      "grad_norm": 1.0969483375908629,
+      "learning_rate": 2.147725393969811e-06,
+      "loss": 0.3231,
+      "step": 3210
+    },
+    {
+      "epoch": 2.172530446549391,
+      "grad_norm": 1.0786800183544625,
+      "learning_rate": 2.1444928437474667e-06,
+      "loss": 0.3167,
+      "step": 3211
+    },
+    {
+      "epoch": 2.1732070365358593,
+      "grad_norm": 1.0898201172124493,
+      "learning_rate": 2.1412620637856445e-06,
+      "loss": 0.3105,
+      "step": 3212
+    },
+    {
+      "epoch": 2.1738836265223274,
+      "grad_norm": 1.0684289989990967,
+      "learning_rate": 2.138033056087256e-06,
+      "loss": 0.319,
+      "step": 3213
+    },
+    {
+      "epoch": 2.174560216508796,
+      "grad_norm": 1.0753064396583398,
+      "learning_rate": 2.1348058226541072e-06,
+      "loss": 0.3133,
+      "step": 3214
+    },
+    {
+      "epoch": 2.175236806495264,
+      "grad_norm": 1.1097920340549594,
+      "learning_rate": 2.1315803654869125e-06,
+      "loss": 0.3231,
+      "step": 3215
+    },
+    {
+      "epoch": 2.175913396481732,
+      "grad_norm": 1.1131696393830561,
+      "learning_rate": 2.1283566865852824e-06,
+      "loss": 0.329,
+      "step": 3216
+    },
+    {
+      "epoch": 2.1765899864682003,
+      "grad_norm": 1.0978277518042483,
+      "learning_rate": 2.1251347879477217e-06,
+      "loss": 0.329,
+      "step": 3217
+    },
+    {
+      "epoch": 2.1772665764546684,
+      "grad_norm": 1.08525798764089,
+      "learning_rate": 2.1219146715716332e-06,
+      "loss": 0.3227,
+      "step": 3218
+    },
+    {
+      "epoch": 2.177943166441137,
+      "grad_norm": 1.0584478988081187,
+      "learning_rate": 2.1186963394533165e-06,
+      "loss": 0.3088,
+      "step": 3219
+    },
+    {
+      "epoch": 2.178619756427605,
+      "grad_norm": 1.1183388150044842,
+      "learning_rate": 2.1154797935879647e-06,
+      "loss": 0.3311,
+      "step": 3220
+    },
+    {
+      "epoch": 2.179296346414073,
+      "grad_norm": 1.0851888500390343,
+      "learning_rate": 2.112265035969664e-06,
+      "loss": 0.3211,
+      "step": 3221
+    },
+    {
+      "epoch": 2.1799729364005414,
+      "grad_norm": 1.1071261649066217,
+      "learning_rate": 2.1090520685913874e-06,
+      "loss": 0.3198,
+      "step": 3222
+    },
+    {
+      "epoch": 2.1806495263870094,
+      "grad_norm": 1.0996624108543835,
+      "learning_rate": 2.1058408934450055e-06,
+      "loss": 0.324,
+      "step": 3223
+    },
+    {
+      "epoch": 2.181326116373478,
+      "grad_norm": 1.1979147047573786,
+      "learning_rate": 2.102631512521269e-06,
+      "loss": 0.3285,
+      "step": 3224
+    },
+    {
+      "epoch": 2.182002706359946,
+      "grad_norm": 1.0688814010406351,
+      "learning_rate": 2.099423927809826e-06,
+      "loss": 0.3093,
+      "step": 3225
+    },
+    {
+      "epoch": 2.182679296346414,
+      "grad_norm": 1.1007338667357167,
+      "learning_rate": 2.096218141299203e-06,
+      "loss": 0.3132,
+      "step": 3226
+    },
+    {
+      "epoch": 2.1833558863328824,
+      "grad_norm": 1.03037091500764,
+      "learning_rate": 2.0930141549768145e-06,
+      "loss": 0.3025,
+      "step": 3227
+    },
+    {
+      "epoch": 2.1840324763193504,
+      "grad_norm": 1.0833469799514677,
+      "learning_rate": 2.089811970828961e-06,
+      "loss": 0.3168,
+      "step": 3228
+    },
+    {
+      "epoch": 2.184709066305819,
+      "grad_norm": 1.1265826881938268,
+      "learning_rate": 2.086611590840826e-06,
+      "loss": 0.3189,
+      "step": 3229
+    },
+    {
+      "epoch": 2.185385656292287,
+      "grad_norm": 1.0948813622373013,
+      "learning_rate": 2.0834130169964695e-06,
+      "loss": 0.3181,
+      "step": 3230
+    },
+    {
+      "epoch": 2.186062246278755,
+      "grad_norm": 1.0918852153574339,
+      "learning_rate": 2.0802162512788337e-06,
+      "loss": 0.3211,
+      "step": 3231
+    },
+    {
+      "epoch": 2.1867388362652234,
+      "grad_norm": 1.09555208352549,
+      "learning_rate": 2.0770212956697435e-06,
+      "loss": 0.3291,
+      "step": 3232
+    },
+    {
+      "epoch": 2.1874154262516914,
+      "grad_norm": 1.0875727207058634,
+      "learning_rate": 2.073828152149898e-06,
+      "loss": 0.3328,
+      "step": 3233
+    },
+    {
+      "epoch": 2.18809201623816,
+      "grad_norm": 1.1287418139427734,
+      "learning_rate": 2.0706368226988772e-06,
+      "loss": 0.327,
+      "step": 3234
+    },
+    {
+      "epoch": 2.188768606224628,
+      "grad_norm": 1.1205981844748083,
+      "learning_rate": 2.0674473092951286e-06,
+      "loss": 0.3139,
+      "step": 3235
+    },
+    {
+      "epoch": 2.189445196211096,
+      "grad_norm": 1.1117176246248053,
+      "learning_rate": 2.064259613915981e-06,
+      "loss": 0.3301,
+      "step": 3236
+    },
+    {
+      "epoch": 2.1901217861975644,
+      "grad_norm": 1.0804406611551591,
+      "learning_rate": 2.061073738537635e-06,
+      "loss": 0.3145,
+      "step": 3237
+    },
+    {
+      "epoch": 2.1907983761840324,
+      "grad_norm": 1.1115435682461963,
+      "learning_rate": 2.0578896851351606e-06,
+      "loss": 0.3223,
+      "step": 3238
+    },
+    {
+      "epoch": 2.191474966170501,
+      "grad_norm": 1.152146545504736,
+      "learning_rate": 2.0547074556824964e-06,
+      "loss": 0.3362,
+      "step": 3239
+    },
+    {
+      "epoch": 2.192151556156969,
+      "grad_norm": 1.1197461664326775,
+      "learning_rate": 2.0515270521524562e-06,
+      "loss": 0.317,
+      "step": 3240
+    },
+    {
+      "epoch": 2.192828146143437,
+      "grad_norm": 1.1004599108012882,
+      "learning_rate": 2.0483484765167172e-06,
+      "loss": 0.3128,
+      "step": 3241
+    },
+    {
+      "epoch": 2.1935047361299054,
+      "grad_norm": 1.0878516434129133,
+      "learning_rate": 2.0451717307458287e-06,
+      "loss": 0.3088,
+      "step": 3242
+    },
+    {
+      "epoch": 2.1941813261163734,
+      "grad_norm": 1.0415435208065689,
+      "learning_rate": 2.041996816809197e-06,
+      "loss": 0.2941,
+      "step": 3243
+    },
+    {
+      "epoch": 2.194857916102842,
+      "grad_norm": 1.142413372852838,
+      "learning_rate": 2.0388237366751005e-06,
+      "loss": 0.3242,
+      "step": 3244
+    },
+    {
+      "epoch": 2.19553450608931,
+      "grad_norm": 1.1143746222524338,
+      "learning_rate": 2.0356524923106763e-06,
+      "loss": 0.3205,
+      "step": 3245
+    },
+    {
+      "epoch": 2.196211096075778,
+      "grad_norm": 1.1433323908983573,
+      "learning_rate": 2.032483085681927e-06,
+      "loss": 0.3283,
+      "step": 3246
+    },
+    {
+      "epoch": 2.1968876860622464,
+      "grad_norm": 1.1046803882827345,
+      "learning_rate": 2.029315518753711e-06,
+      "loss": 0.3182,
+      "step": 3247
+    },
+    {
+      "epoch": 2.1975642760487144,
+      "grad_norm": 1.0799027923360895,
+      "learning_rate": 2.0261497934897507e-06,
+      "loss": 0.3165,
+      "step": 3248
+    },
+    {
+      "epoch": 2.198240866035183,
+      "grad_norm": 1.0719060596376357,
+      "learning_rate": 2.0229859118526244e-06,
+      "loss": 0.3104,
+      "step": 3249
+    },
+    {
+      "epoch": 2.198917456021651,
+      "grad_norm": 1.0967964847012261,
+      "learning_rate": 2.019823875803771e-06,
+      "loss": 0.3195,
+      "step": 3250
+    },
+    {
+      "epoch": 2.199594046008119,
+      "grad_norm": 1.0935366419862107,
+      "learning_rate": 2.0166636873034807e-06,
+      "loss": 0.3127,
+      "step": 3251
+    },
+    {
+      "epoch": 2.2002706359945874,
+      "grad_norm": 1.0607589844761274,
+      "learning_rate": 2.0135053483108973e-06,
+      "loss": 0.3138,
+      "step": 3252
+    },
+    {
+      "epoch": 2.2009472259810554,
+      "grad_norm": 1.0647223788232456,
+      "learning_rate": 2.0103488607840233e-06,
+      "loss": 0.3202,
+      "step": 3253
+    },
+    {
+      "epoch": 2.201623815967524,
+      "grad_norm": 1.089194003129358,
+      "learning_rate": 2.00719422667971e-06,
+      "loss": 0.3161,
+      "step": 3254
+    },
+    {
+      "epoch": 2.202300405953992,
+      "grad_norm": 1.1041258251832875,
+      "learning_rate": 2.004041447953663e-06,
+      "loss": 0.3181,
+      "step": 3255
+    },
+    {
+      "epoch": 2.20297699594046,
+      "grad_norm": 1.0599321669780764,
+      "learning_rate": 2.0008905265604316e-06,
+      "loss": 0.3136,
+      "step": 3256
+    },
+    {
+      "epoch": 2.2036535859269284,
+      "grad_norm": 1.0851323571905946,
+      "learning_rate": 1.9977414644534206e-06,
+      "loss": 0.3267,
+      "step": 3257
+    },
+    {
+      "epoch": 2.2043301759133964,
+      "grad_norm": 1.0985336027662924,
+      "learning_rate": 1.9945942635848745e-06,
+      "loss": 0.314,
+      "step": 3258
+    },
+    {
+      "epoch": 2.205006765899865,
+      "grad_norm": 1.085049523892315,
+      "learning_rate": 1.9914489259058933e-06,
+      "loss": 0.3175,
+      "step": 3259
+    },
+    {
+      "epoch": 2.205683355886333,
+      "grad_norm": 1.0936457064241567,
+      "learning_rate": 1.9883054533664128e-06,
+      "loss": 0.3112,
+      "step": 3260
+    },
+    {
+      "epoch": 2.206359945872801,
+      "grad_norm": 1.137587536853472,
+      "learning_rate": 1.985163847915217e-06,
+      "loss": 0.3261,
+      "step": 3261
+    },
+    {
+      "epoch": 2.2070365358592694,
+      "grad_norm": 1.1502869126601807,
+      "learning_rate": 1.9820241114999334e-06,
+      "loss": 0.3335,
+      "step": 3262
+    },
+    {
+      "epoch": 2.2077131258457374,
+      "grad_norm": 1.084424883963079,
+      "learning_rate": 1.9788862460670305e-06,
+      "loss": 0.3133,
+      "step": 3263
+    },
+    {
+      "epoch": 2.208389715832206,
+      "grad_norm": 1.0941965681535513,
+      "learning_rate": 1.9757502535618137e-06,
+      "loss": 0.322,
+      "step": 3264
+    },
+    {
+      "epoch": 2.209066305818674,
+      "grad_norm": 1.1036070624995395,
+      "learning_rate": 1.9726161359284283e-06,
+      "loss": 0.3169,
+      "step": 3265
+    },
+    {
+      "epoch": 2.209742895805142,
+      "grad_norm": 1.1342514431296868,
+      "learning_rate": 1.96948389510986e-06,
+      "loss": 0.3324,
+      "step": 3266
+    },
+    {
+      "epoch": 2.2104194857916104,
+      "grad_norm": 1.1379243987804681,
+      "learning_rate": 1.9663535330479305e-06,
+      "loss": 0.3246,
+      "step": 3267
+    },
+    {
+      "epoch": 2.2110960757780784,
+      "grad_norm": 1.08204091272448,
+      "learning_rate": 1.963225051683292e-06,
+      "loss": 0.3155,
+      "step": 3268
+    },
+    {
+      "epoch": 2.2117726657645465,
+      "grad_norm": 1.0590795572638303,
+      "learning_rate": 1.9600984529554366e-06,
+      "loss": 0.3157,
+      "step": 3269
+    },
+    {
+      "epoch": 2.212449255751015,
+      "grad_norm": 1.1260939943685944,
+      "learning_rate": 1.956973738802689e-06,
+      "loss": 0.3262,
+      "step": 3270
+    },
+    {
+      "epoch": 2.213125845737483,
+      "grad_norm": 1.0964961612693531,
+      "learning_rate": 1.953850911162199e-06,
+      "loss": 0.3111,
+      "step": 3271
+    },
+    {
+      "epoch": 2.2138024357239514,
+      "grad_norm": 1.084500344709526,
+      "learning_rate": 1.950729971969955e-06,
+      "loss": 0.3114,
+      "step": 3272
+    },
+    {
+      "epoch": 2.2144790257104194,
+      "grad_norm": 1.0741713297188433,
+      "learning_rate": 1.9476109231607687e-06,
+      "loss": 0.3211,
+      "step": 3273
+    },
+    {
+      "epoch": 2.2151556156968875,
+      "grad_norm": 1.1128583013601934,
+      "learning_rate": 1.9444937666682834e-06,
+      "loss": 0.3282,
+      "step": 3274
+    },
+    {
+      "epoch": 2.215832205683356,
+      "grad_norm": 1.0783496600813218,
+      "learning_rate": 1.941378504424968e-06,
+      "loss": 0.3188,
+      "step": 3275
+    },
+    {
+      "epoch": 2.216508795669824,
+      "grad_norm": 1.142615372525608,
+      "learning_rate": 1.938265138362118e-06,
+      "loss": 0.3415,
+      "step": 3276
+    },
+    {
+      "epoch": 2.2171853856562924,
+      "grad_norm": 1.1176536155289774,
+      "learning_rate": 1.935153670409853e-06,
+      "loss": 0.3242,
+      "step": 3277
+    },
+    {
+      "epoch": 2.2178619756427604,
+      "grad_norm": 1.1162769030229303,
+      "learning_rate": 1.9320441024971113e-06,
+      "loss": 0.3249,
+      "step": 3278
+    },
+    {
+      "epoch": 2.2185385656292285,
+      "grad_norm": 1.1016011369705836,
+      "learning_rate": 1.928936436551661e-06,
+      "loss": 0.3244,
+      "step": 3279
+    },
+    {
+      "epoch": 2.219215155615697,
+      "grad_norm": 1.12092491179393,
+      "learning_rate": 1.925830674500088e-06,
+      "loss": 0.3326,
+      "step": 3280
+    },
+    {
+      "epoch": 2.219891745602165,
+      "grad_norm": 1.0612791964845465,
+      "learning_rate": 1.922726818267795e-06,
+      "loss": 0.3052,
+      "step": 3281
+    },
+    {
+      "epoch": 2.2205683355886334,
+      "grad_norm": 1.1058906838071672,
+      "learning_rate": 1.9196248697790066e-06,
+      "loss": 0.3153,
+      "step": 3282
+    },
+    {
+      "epoch": 2.2212449255751014,
+      "grad_norm": 1.0743562816872771,
+      "learning_rate": 1.916524830956763e-06,
+      "loss": 0.3144,
+      "step": 3283
+    },
+    {
+      "epoch": 2.2219215155615695,
+      "grad_norm": 1.0751463117493012,
+      "learning_rate": 1.913426703722924e-06,
+      "loss": 0.3158,
+      "step": 3284
+    },
+    {
+      "epoch": 2.222598105548038,
+      "grad_norm": 1.1144507027554167,
+      "learning_rate": 1.9103304899981603e-06,
+      "loss": 0.3091,
+      "step": 3285
+    },
+    {
+      "epoch": 2.223274695534506,
+      "grad_norm": 1.0893710367294782,
+      "learning_rate": 1.9072361917019538e-06,
+      "loss": 0.3055,
+      "step": 3286
+    },
+    {
+      "epoch": 2.2239512855209744,
+      "grad_norm": 1.077157428644335,
+      "learning_rate": 1.9041438107526055e-06,
+      "loss": 0.3162,
+      "step": 3287
+    },
+    {
+      "epoch": 2.2246278755074425,
+      "grad_norm": 1.0969452093928271,
+      "learning_rate": 1.901053349067225e-06,
+      "loss": 0.3106,
+      "step": 3288
+    },
+    {
+      "epoch": 2.2253044654939105,
+      "grad_norm": 1.0875299012338826,
+      "learning_rate": 1.8979648085617342e-06,
+      "loss": 0.3163,
+      "step": 3289
+    },
+    {
+      "epoch": 2.225981055480379,
+      "grad_norm": 1.0762528308437525,
+      "learning_rate": 1.894878191150859e-06,
+      "loss": 0.3148,
+      "step": 3290
+    },
+    {
+      "epoch": 2.226657645466847,
+      "grad_norm": 1.1518906662810515,
+      "learning_rate": 1.891793498748134e-06,
+      "loss": 0.3331,
+      "step": 3291
+    },
+    {
+      "epoch": 2.2273342354533154,
+      "grad_norm": 1.1317721373872631,
+      "learning_rate": 1.888710733265905e-06,
+      "loss": 0.3285,
+      "step": 3292
+    },
+    {
+      "epoch": 2.2280108254397835,
+      "grad_norm": 1.0903729072189,
+      "learning_rate": 1.8856298966153214e-06,
+      "loss": 0.3116,
+      "step": 3293
+    },
+    {
+      "epoch": 2.2286874154262515,
+      "grad_norm": 1.0947288957335477,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.3233,
+      "step": 3294
+    },
+    {
+      "epoch": 2.22936400541272,
+      "grad_norm": 1.0993037404763306,
+      "learning_rate": 1.8794740174476966e-06,
+      "loss": 0.316,
+      "step": 3295
+    },
+    {
+      "epoch": 2.230040595399188,
+      "grad_norm": 1.108437016598285,
+      "learning_rate": 1.87639897874697e-06,
+      "loss": 0.3184,
+      "step": 3296
+    },
+    {
+      "epoch": 2.2307171853856564,
+      "grad_norm": 1.114164944749697,
+      "learning_rate": 1.8733258765105129e-06,
+      "loss": 0.3203,
+      "step": 3297
+    },
+    {
+      "epoch": 2.2313937753721245,
+      "grad_norm": 1.1083368432880243,
+      "learning_rate": 1.8702547126434818e-06,
+      "loss": 0.3183,
+      "step": 3298
+    },
+    {
+      "epoch": 2.2320703653585925,
+      "grad_norm": 1.1175177439313144,
+      "learning_rate": 1.8671854890498308e-06,
+      "loss": 0.3245,
+      "step": 3299
+    },
+    {
+      "epoch": 2.232746955345061,
+      "grad_norm": 1.1254683235367513,
+      "learning_rate": 1.864118207632315e-06,
+      "loss": 0.32,
+      "step": 3300
+    },
+    {
+      "epoch": 2.233423545331529,
+      "grad_norm": 1.0877107601148008,
+      "learning_rate": 1.8610528702924851e-06,
+      "loss": 0.3154,
+      "step": 3301
+    },
+    {
+      "epoch": 2.2341001353179974,
+      "grad_norm": 1.1089743598037287,
+      "learning_rate": 1.8579894789306813e-06,
+      "loss": 0.3241,
+      "step": 3302
+    },
+    {
+      "epoch": 2.2347767253044655,
+      "grad_norm": 1.0782980078206152,
+      "learning_rate": 1.8549280354460437e-06,
+      "loss": 0.3098,
+      "step": 3303
+    },
+    {
+      "epoch": 2.2354533152909335,
+      "grad_norm": 1.0864053175720063,
+      "learning_rate": 1.851868541736503e-06,
+      "loss": 0.31,
+      "step": 3304
+    },
+    {
+      "epoch": 2.236129905277402,
+      "grad_norm": 1.1387094568267146,
+      "learning_rate": 1.8488109996987774e-06,
+      "loss": 0.3072,
+      "step": 3305
+    },
+    {
+      "epoch": 2.23680649526387,
+      "grad_norm": 1.0956085211301947,
+      "learning_rate": 1.845755411228382e-06,
+      "loss": 0.3101,
+      "step": 3306
+    },
+    {
+      "epoch": 2.2374830852503385,
+      "grad_norm": 1.1485300394760511,
+      "learning_rate": 1.8427017782196126e-06,
+      "loss": 0.3279,
+      "step": 3307
+    },
+    {
+      "epoch": 2.2381596752368065,
+      "grad_norm": 1.1474997281514587,
+      "learning_rate": 1.8396501025655594e-06,
+      "loss": 0.3234,
+      "step": 3308
+    },
+    {
+      "epoch": 2.2388362652232745,
+      "grad_norm": 1.0913245328103054,
+      "learning_rate": 1.8366003861580966e-06,
+      "loss": 0.3169,
+      "step": 3309
+    },
+    {
+      "epoch": 2.239512855209743,
+      "grad_norm": 1.123623576016734,
+      "learning_rate": 1.8335526308878877e-06,
+      "loss": 0.3242,
+      "step": 3310
+    },
+    {
+      "epoch": 2.240189445196211,
+      "grad_norm": 1.116077851758487,
+      "learning_rate": 1.8305068386443696e-06,
+      "loss": 0.3152,
+      "step": 3311
+    },
+    {
+      "epoch": 2.2408660351826795,
+      "grad_norm": 1.060396643319164,
+      "learning_rate": 1.8274630113157727e-06,
+      "loss": 0.3067,
+      "step": 3312
+    },
+    {
+      "epoch": 2.2415426251691475,
+      "grad_norm": 1.127108287277122,
+      "learning_rate": 1.8244211507891064e-06,
+      "loss": 0.328,
+      "step": 3313
+    },
+    {
+      "epoch": 2.2422192151556155,
+      "grad_norm": 1.0690531696713788,
+      "learning_rate": 1.8213812589501611e-06,
+      "loss": 0.307,
+      "step": 3314
+    },
+    {
+      "epoch": 2.242895805142084,
+      "grad_norm": 1.122336479062686,
+      "learning_rate": 1.818343337683503e-06,
+      "loss": 0.3227,
+      "step": 3315
+    },
+    {
+      "epoch": 2.243572395128552,
+      "grad_norm": 1.1359795459881408,
+      "learning_rate": 1.815307388872481e-06,
+      "loss": 0.3262,
+      "step": 3316
+    },
+    {
+      "epoch": 2.2442489851150205,
+      "grad_norm": 1.0622564695149823,
+      "learning_rate": 1.8122734143992216e-06,
+      "loss": 0.3005,
+      "step": 3317
+    },
+    {
+      "epoch": 2.2449255751014885,
+      "grad_norm": 1.0814269861374335,
+      "learning_rate": 1.8092414161446225e-06,
+      "loss": 0.3107,
+      "step": 3318
+    },
+    {
+      "epoch": 2.2456021650879565,
+      "grad_norm": 1.1459700024946626,
+      "learning_rate": 1.8062113959883616e-06,
+      "loss": 0.3297,
+      "step": 3319
+    },
+    {
+      "epoch": 2.246278755074425,
+      "grad_norm": 1.1131065589915317,
+      "learning_rate": 1.8031833558088858e-06,
+      "loss": 0.3139,
+      "step": 3320
+    },
+    {
+      "epoch": 2.246955345060893,
+      "grad_norm": 1.1666161981810397,
+      "learning_rate": 1.8001572974834169e-06,
+      "loss": 0.3375,
+      "step": 3321
+    },
+    {
+      "epoch": 2.2476319350473615,
+      "grad_norm": 1.113297941437187,
+      "learning_rate": 1.7971332228879518e-06,
+      "loss": 0.3205,
+      "step": 3322
+    },
+    {
+      "epoch": 2.2483085250338295,
+      "grad_norm": 1.102423882613334,
+      "learning_rate": 1.7941111338972484e-06,
+      "loss": 0.3114,
+      "step": 3323
+    },
+    {
+      "epoch": 2.2489851150202975,
+      "grad_norm": 1.0811960853990636,
+      "learning_rate": 1.7910910323848435e-06,
+      "loss": 0.3027,
+      "step": 3324
+    },
+    {
+      "epoch": 2.249661705006766,
+      "grad_norm": 1.0805295002823965,
+      "learning_rate": 1.7880729202230334e-06,
+      "loss": 0.3221,
+      "step": 3325
+    },
+    {
+      "epoch": 2.250338294993234,
+      "grad_norm": 1.092699611162999,
+      "learning_rate": 1.7850567992828865e-06,
+      "loss": 0.3118,
+      "step": 3326
+    },
+    {
+      "epoch": 2.2510148849797025,
+      "grad_norm": 1.12779435947042,
+      "learning_rate": 1.7820426714342375e-06,
+      "loss": 0.3082,
+      "step": 3327
+    },
+    {
+      "epoch": 2.2516914749661705,
+      "grad_norm": 1.1207219680619251,
+      "learning_rate": 1.7790305385456797e-06,
+      "loss": 0.3339,
+      "step": 3328
+    },
+    {
+      "epoch": 2.2523680649526385,
+      "grad_norm": 1.1140847565481993,
+      "learning_rate": 1.7760204024845745e-06,
+      "loss": 0.3261,
+      "step": 3329
+    },
+    {
+      "epoch": 2.253044654939107,
+      "grad_norm": 1.116370655879667,
+      "learning_rate": 1.7730122651170457e-06,
+      "loss": 0.3204,
+      "step": 3330
+    },
+    {
+      "epoch": 2.253721244925575,
+      "grad_norm": 1.1105220215816718,
+      "learning_rate": 1.7700061283079744e-06,
+      "loss": 0.3191,
+      "step": 3331
+    },
+    {
+      "epoch": 2.2543978349120435,
+      "grad_norm": 1.1039621507071047,
+      "learning_rate": 1.7670019939210025e-06,
+      "loss": 0.3206,
+      "step": 3332
+    },
+    {
+      "epoch": 2.2550744248985115,
+      "grad_norm": 1.1903598350264117,
+      "learning_rate": 1.763999863818533e-06,
+      "loss": 0.3247,
+      "step": 3333
+    },
+    {
+      "epoch": 2.2557510148849795,
+      "grad_norm": 1.108812204563875,
+      "learning_rate": 1.760999739861724e-06,
+      "loss": 0.3186,
+      "step": 3334
+    },
+    {
+      "epoch": 2.256427604871448,
+      "grad_norm": 1.1401390865079983,
+      "learning_rate": 1.7580016239104924e-06,
+      "loss": 0.3288,
+      "step": 3335
+    },
+    {
+      "epoch": 2.257104194857916,
+      "grad_norm": 1.1144430665962406,
+      "learning_rate": 1.755005517823506e-06,
+      "loss": 0.3144,
+      "step": 3336
+    },
+    {
+      "epoch": 2.2577807848443845,
+      "grad_norm": 1.0842963872846536,
+      "learning_rate": 1.7520114234581914e-06,
+      "loss": 0.3057,
+      "step": 3337
+    },
+    {
+      "epoch": 2.2584573748308525,
+      "grad_norm": 1.1424795584299616,
+      "learning_rate": 1.7490193426707236e-06,
+      "loss": 0.3242,
+      "step": 3338
+    },
+    {
+      "epoch": 2.2591339648173205,
+      "grad_norm": 1.1355605059209823,
+      "learning_rate": 1.7460292773160315e-06,
+      "loss": 0.326,
+      "step": 3339
+    },
+    {
+      "epoch": 2.259810554803789,
+      "grad_norm": 1.1011906814131458,
+      "learning_rate": 1.7430412292477978e-06,
+      "loss": 0.312,
+      "step": 3340
+    },
+    {
+      "epoch": 2.260487144790257,
+      "grad_norm": 1.138540676717295,
+      "learning_rate": 1.7400552003184463e-06,
+      "loss": 0.3127,
+      "step": 3341
+    },
+    {
+      "epoch": 2.2611637347767255,
+      "grad_norm": 1.1270333385021083,
+      "learning_rate": 1.7370711923791567e-06,
+      "loss": 0.3122,
+      "step": 3342
+    },
+    {
+      "epoch": 2.2618403247631935,
+      "grad_norm": 1.1015075611748257,
+      "learning_rate": 1.7340892072798544e-06,
+      "loss": 0.3164,
+      "step": 3343
+    },
+    {
+      "epoch": 2.2625169147496615,
+      "grad_norm": 1.140531977166998,
+      "learning_rate": 1.7311092468692082e-06,
+      "loss": 0.3227,
+      "step": 3344
+    },
+    {
+      "epoch": 2.26319350473613,
+      "grad_norm": 1.0957238030600263,
+      "learning_rate": 1.7281313129946302e-06,
+      "loss": 0.3045,
+      "step": 3345
+    },
+    {
+      "epoch": 2.263870094722598,
+      "grad_norm": 1.104971454612028,
+      "learning_rate": 1.725155407502282e-06,
+      "loss": 0.3232,
+      "step": 3346
+    },
+    {
+      "epoch": 2.2645466847090665,
+      "grad_norm": 1.1322417425026432,
+      "learning_rate": 1.7221815322370633e-06,
+      "loss": 0.3251,
+      "step": 3347
+    },
+    {
+      "epoch": 2.2652232746955345,
+      "grad_norm": 1.1350080826911368,
+      "learning_rate": 1.7192096890426192e-06,
+      "loss": 0.3247,
+      "step": 3348
+    },
+    {
+      "epoch": 2.2658998646820026,
+      "grad_norm": 1.0942392860611383,
+      "learning_rate": 1.7162398797613284e-06,
+      "loss": 0.3075,
+      "step": 3349
+    },
+    {
+      "epoch": 2.266576454668471,
+      "grad_norm": 1.1402895028747502,
+      "learning_rate": 1.7132721062343156e-06,
+      "loss": 0.3386,
+      "step": 3350
+    },
+    {
+      "epoch": 2.267253044654939,
+      "grad_norm": 1.1146478475025061,
+      "learning_rate": 1.7103063703014372e-06,
+      "loss": 0.3134,
+      "step": 3351
+    },
+    {
+      "epoch": 2.2679296346414075,
+      "grad_norm": 1.0618294213987924,
+      "learning_rate": 1.7073426738012939e-06,
+      "loss": 0.2987,
+      "step": 3352
+    },
+    {
+      "epoch": 2.2686062246278755,
+      "grad_norm": 1.118063298485389,
+      "learning_rate": 1.7043810185712135e-06,
+      "loss": 0.3193,
+      "step": 3353
+    },
+    {
+      "epoch": 2.2692828146143436,
+      "grad_norm": 1.0807042302554075,
+      "learning_rate": 1.7014214064472646e-06,
+      "loss": 0.3087,
+      "step": 3354
+    },
+    {
+      "epoch": 2.269959404600812,
+      "grad_norm": 1.141316907501067,
+      "learning_rate": 1.6984638392642467e-06,
+      "loss": 0.3253,
+      "step": 3355
+    },
+    {
+      "epoch": 2.27063599458728,
+      "grad_norm": 1.1220599605407322,
+      "learning_rate": 1.6955083188556947e-06,
+      "loss": 0.3262,
+      "step": 3356
+    },
+    {
+      "epoch": 2.2713125845737485,
+      "grad_norm": 1.0923070298453927,
+      "learning_rate": 1.6925548470538695e-06,
+      "loss": 0.3087,
+      "step": 3357
+    },
+    {
+      "epoch": 2.2719891745602165,
+      "grad_norm": 1.1350411400521894,
+      "learning_rate": 1.6896034256897626e-06,
+      "loss": 0.3199,
+      "step": 3358
+    },
+    {
+      "epoch": 2.2726657645466846,
+      "grad_norm": 1.0965410078432,
+      "learning_rate": 1.686654056593099e-06,
+      "loss": 0.3166,
+      "step": 3359
+    },
+    {
+      "epoch": 2.273342354533153,
+      "grad_norm": 1.101257652156372,
+      "learning_rate": 1.683706741592327e-06,
+      "loss": 0.3201,
+      "step": 3360
+    },
+    {
+      "epoch": 2.274018944519621,
+      "grad_norm": 1.1066956228417268,
+      "learning_rate": 1.6807614825146258e-06,
+      "loss": 0.3144,
+      "step": 3361
+    },
+    {
+      "epoch": 2.2746955345060895,
+      "grad_norm": 1.1346421941125058,
+      "learning_rate": 1.6778182811858934e-06,
+      "loss": 0.3215,
+      "step": 3362
+    },
+    {
+      "epoch": 2.2753721244925575,
+      "grad_norm": 1.0664768712631463,
+      "learning_rate": 1.6748771394307584e-06,
+      "loss": 0.3003,
+      "step": 3363
+    },
+    {
+      "epoch": 2.2760487144790256,
+      "grad_norm": 1.058773094429667,
+      "learning_rate": 1.671938059072571e-06,
+      "loss": 0.2911,
+      "step": 3364
+    },
+    {
+      "epoch": 2.276725304465494,
+      "grad_norm": 1.1096086763501996,
+      "learning_rate": 1.6690010419334008e-06,
+      "loss": 0.3125,
+      "step": 3365
+    },
+    {
+      "epoch": 2.277401894451962,
+      "grad_norm": 1.0957051269097933,
+      "learning_rate": 1.6660660898340392e-06,
+      "loss": 0.3105,
+      "step": 3366
+    },
+    {
+      "epoch": 2.2780784844384305,
+      "grad_norm": 1.1054903398293534,
+      "learning_rate": 1.6631332045939996e-06,
+      "loss": 0.3222,
+      "step": 3367
+    },
+    {
+      "epoch": 2.2787550744248986,
+      "grad_norm": 1.13905856570075,
+      "learning_rate": 1.6602023880315126e-06,
+      "loss": 0.3189,
+      "step": 3368
+    },
+    {
+      "epoch": 2.2794316644113666,
+      "grad_norm": 1.0891361743846752,
+      "learning_rate": 1.6572736419635288e-06,
+      "loss": 0.3123,
+      "step": 3369
+    },
+    {
+      "epoch": 2.280108254397835,
+      "grad_norm": 1.1544619993742387,
+      "learning_rate": 1.6543469682057105e-06,
+      "loss": 0.3324,
+      "step": 3370
+    },
+    {
+      "epoch": 2.280784844384303,
+      "grad_norm": 1.121760153736156,
+      "learning_rate": 1.651422368572436e-06,
+      "loss": 0.3034,
+      "step": 3371
+    },
+    {
+      "epoch": 2.2814614343707715,
+      "grad_norm": 1.0785468954532595,
+      "learning_rate": 1.648499844876802e-06,
+      "loss": 0.305,
+      "step": 3372
+    },
+    {
+      "epoch": 2.2821380243572396,
+      "grad_norm": 1.1182286766661833,
+      "learning_rate": 1.6455793989306169e-06,
+      "loss": 0.3097,
+      "step": 3373
+    },
+    {
+      "epoch": 2.2828146143437076,
+      "grad_norm": 1.0734889722576963,
+      "learning_rate": 1.642661032544396e-06,
+      "loss": 0.3085,
+      "step": 3374
+    },
+    {
+      "epoch": 2.283491204330176,
+      "grad_norm": 1.108891989678797,
+      "learning_rate": 1.639744747527371e-06,
+      "loss": 0.318,
+      "step": 3375
+    },
+    {
+      "epoch": 2.284167794316644,
+      "grad_norm": 1.1205993282113098,
+      "learning_rate": 1.636830545687481e-06,
+      "loss": 0.3318,
+      "step": 3376
+    },
+    {
+      "epoch": 2.2848443843031125,
+      "grad_norm": 1.0510503977154908,
+      "learning_rate": 1.6339184288313769e-06,
+      "loss": 0.3025,
+      "step": 3377
+    },
+    {
+      "epoch": 2.2855209742895806,
+      "grad_norm": 1.1120851647899683,
+      "learning_rate": 1.631008398764412e-06,
+      "loss": 0.3144,
+      "step": 3378
+    },
+    {
+      "epoch": 2.2861975642760486,
+      "grad_norm": 1.149039813226276,
+      "learning_rate": 1.6281004572906462e-06,
+      "loss": 0.3222,
+      "step": 3379
+    },
+    {
+      "epoch": 2.286874154262517,
+      "grad_norm": 1.0715588590696454,
+      "learning_rate": 1.6251946062128482e-06,
+      "loss": 0.2923,
+      "step": 3380
+    },
+    {
+      "epoch": 2.287550744248985,
+      "grad_norm": 1.1148798759956529,
+      "learning_rate": 1.6222908473324889e-06,
+      "loss": 0.3116,
+      "step": 3381
+    },
+    {
+      "epoch": 2.2882273342354535,
+      "grad_norm": 1.1030688885896307,
+      "learning_rate": 1.6193891824497438e-06,
+      "loss": 0.3274,
+      "step": 3382
+    },
+    {
+      "epoch": 2.2889039242219216,
+      "grad_norm": 1.1351302355210737,
+      "learning_rate": 1.616489613363486e-06,
+      "loss": 0.3161,
+      "step": 3383
+    },
+    {
+      "epoch": 2.2895805142083896,
+      "grad_norm": 1.0873188603335708,
+      "learning_rate": 1.6135921418712959e-06,
+      "loss": 0.3095,
+      "step": 3384
+    },
+    {
+      "epoch": 2.290257104194858,
+      "grad_norm": 1.099302104968843,
+      "learning_rate": 1.6106967697694442e-06,
+      "loss": 0.3178,
+      "step": 3385
+    },
+    {
+      "epoch": 2.290933694181326,
+      "grad_norm": 1.0932057979976018,
+      "learning_rate": 1.6078034988529112e-06,
+      "loss": 0.3082,
+      "step": 3386
+    },
+    {
+      "epoch": 2.2916102841677946,
+      "grad_norm": 1.1199446924872456,
+      "learning_rate": 1.604912330915364e-06,
+      "loss": 0.3204,
+      "step": 3387
+    },
+    {
+      "epoch": 2.2922868741542626,
+      "grad_norm": 1.1172734074807096,
+      "learning_rate": 1.6020232677491732e-06,
+      "loss": 0.3165,
+      "step": 3388
+    },
+    {
+      "epoch": 2.2929634641407306,
+      "grad_norm": 1.1258570181484318,
+      "learning_rate": 1.5991363111454023e-06,
+      "loss": 0.3185,
+      "step": 3389
+    },
+    {
+      "epoch": 2.293640054127199,
+      "grad_norm": 1.1459021280959485,
+      "learning_rate": 1.5962514628938103e-06,
+      "loss": 0.3202,
+      "step": 3390
+    },
+    {
+      "epoch": 2.294316644113667,
+      "grad_norm": 1.0794144926972977,
+      "learning_rate": 1.5933687247828462e-06,
+      "loss": 0.3115,
+      "step": 3391
+    },
+    {
+      "epoch": 2.2949932341001356,
+      "grad_norm": 1.1517526760721986,
+      "learning_rate": 1.59048809859965e-06,
+      "loss": 0.3306,
+      "step": 3392
+    },
+    {
+      "epoch": 2.2956698240866036,
+      "grad_norm": 1.1561322037872452,
+      "learning_rate": 1.5876095861300567e-06,
+      "loss": 0.3189,
+      "step": 3393
+    },
+    {
+      "epoch": 2.2963464140730716,
+      "grad_norm": 1.1427339573349151,
+      "learning_rate": 1.5847331891585888e-06,
+      "loss": 0.3204,
+      "step": 3394
+    },
+    {
+      "epoch": 2.29702300405954,
+      "grad_norm": 1.1220120342409396,
+      "learning_rate": 1.5818589094684594e-06,
+      "loss": 0.3173,
+      "step": 3395
+    },
+    {
+      "epoch": 2.297699594046008,
+      "grad_norm": 1.1166275446306264,
+      "learning_rate": 1.5789867488415633e-06,
+      "loss": 0.3195,
+      "step": 3396
+    },
+    {
+      "epoch": 2.2983761840324766,
+      "grad_norm": 1.127764635480915,
+      "learning_rate": 1.5761167090584885e-06,
+      "loss": 0.3104,
+      "step": 3397
+    },
+    {
+      "epoch": 2.2990527740189446,
+      "grad_norm": 1.10975894657015,
+      "learning_rate": 1.5732487918985017e-06,
+      "loss": 0.3178,
+      "step": 3398
+    },
+    {
+      "epoch": 2.2997293640054126,
+      "grad_norm": 1.0738324905777639,
+      "learning_rate": 1.5703829991395602e-06,
+      "loss": 0.3055,
+      "step": 3399
+    },
+    {
+      "epoch": 2.300405953991881,
+      "grad_norm": 1.137934750987506,
+      "learning_rate": 1.5675193325582983e-06,
+      "loss": 0.3141,
+      "step": 3400
+    },
+    {
+      "epoch": 2.301082543978349,
+      "grad_norm": 1.1329215898229208,
+      "learning_rate": 1.5646577939300362e-06,
+      "loss": 0.3121,
+      "step": 3401
+    },
+    {
+      "epoch": 2.301759133964817,
+      "grad_norm": 1.102345858270791,
+      "learning_rate": 1.5617983850287737e-06,
+      "loss": 0.3084,
+      "step": 3402
+    },
+    {
+      "epoch": 2.3024357239512856,
+      "grad_norm": 1.1021439141083313,
+      "learning_rate": 1.5589411076271916e-06,
+      "loss": 0.3081,
+      "step": 3403
+    },
+    {
+      "epoch": 2.3031123139377536,
+      "grad_norm": 1.1394970532425004,
+      "learning_rate": 1.5560859634966457e-06,
+      "loss": 0.3211,
+      "step": 3404
+    },
+    {
+      "epoch": 2.303788903924222,
+      "grad_norm": 1.129693204042089,
+      "learning_rate": 1.5532329544071712e-06,
+      "loss": 0.3194,
+      "step": 3405
+    },
+    {
+      "epoch": 2.30446549391069,
+      "grad_norm": 1.107859806206584,
+      "learning_rate": 1.5503820821274812e-06,
+      "loss": 0.3205,
+      "step": 3406
+    },
+    {
+      "epoch": 2.305142083897158,
+      "grad_norm": 1.114577905263349,
+      "learning_rate": 1.5475333484249633e-06,
+      "loss": 0.3164,
+      "step": 3407
+    },
+    {
+      "epoch": 2.3058186738836266,
+      "grad_norm": 1.169566832221299,
+      "learning_rate": 1.544686755065677e-06,
+      "loss": 0.3304,
+      "step": 3408
+    },
+    {
+      "epoch": 2.3064952638700946,
+      "grad_norm": 1.1347548120296829,
+      "learning_rate": 1.5418423038143576e-06,
+      "loss": 0.3097,
+      "step": 3409
+    },
+    {
+      "epoch": 2.307171853856563,
+      "grad_norm": 1.1445595073032628,
+      "learning_rate": 1.5389999964344138e-06,
+      "loss": 0.325,
+      "step": 3410
+    },
+    {
+      "epoch": 2.307848443843031,
+      "grad_norm": 1.1650905686187547,
+      "learning_rate": 1.5361598346879193e-06,
+      "loss": 0.3262,
+      "step": 3411
+    },
+    {
+      "epoch": 2.308525033829499,
+      "grad_norm": 1.0740847149600803,
+      "learning_rate": 1.5333218203356243e-06,
+      "loss": 0.2951,
+      "step": 3412
+    },
+    {
+      "epoch": 2.3092016238159676,
+      "grad_norm": 1.1487668739745571,
+      "learning_rate": 1.5304859551369417e-06,
+      "loss": 0.333,
+      "step": 3413
+    },
+    {
+      "epoch": 2.3098782138024356,
+      "grad_norm": 1.0967164133714282,
+      "learning_rate": 1.5276522408499567e-06,
+      "loss": 0.3073,
+      "step": 3414
+    },
+    {
+      "epoch": 2.310554803788904,
+      "grad_norm": 1.1216155389870373,
+      "learning_rate": 1.5248206792314197e-06,
+      "loss": 0.332,
+      "step": 3415
+    },
+    {
+      "epoch": 2.311231393775372,
+      "grad_norm": 1.1108707580735167,
+      "learning_rate": 1.5219912720367474e-06,
+      "loss": 0.3186,
+      "step": 3416
+    },
+    {
+      "epoch": 2.31190798376184,
+      "grad_norm": 1.128033667017073,
+      "learning_rate": 1.5191640210200186e-06,
+      "loss": 0.318,
+      "step": 3417
+    },
+    {
+      "epoch": 2.3125845737483086,
+      "grad_norm": 1.1257536737162772,
+      "learning_rate": 1.5163389279339746e-06,
+      "loss": 0.3179,
+      "step": 3418
+    },
+    {
+      "epoch": 2.3132611637347766,
+      "grad_norm": 1.1114635673002509,
+      "learning_rate": 1.5135159945300232e-06,
+      "loss": 0.3135,
+      "step": 3419
+    },
+    {
+      "epoch": 2.313937753721245,
+      "grad_norm": 1.113924567701934,
+      "learning_rate": 1.5106952225582312e-06,
+      "loss": 0.3161,
+      "step": 3420
+    },
+    {
+      "epoch": 2.314614343707713,
+      "grad_norm": 1.1127315917413891,
+      "learning_rate": 1.5078766137673229e-06,
+      "loss": 0.3275,
+      "step": 3421
+    },
+    {
+      "epoch": 2.315290933694181,
+      "grad_norm": 1.1070729147789264,
+      "learning_rate": 1.5050601699046852e-06,
+      "loss": 0.3138,
+      "step": 3422
+    },
+    {
+      "epoch": 2.3159675236806496,
+      "grad_norm": 1.1263929228177223,
+      "learning_rate": 1.5022458927163618e-06,
+      "loss": 0.3198,
+      "step": 3423
+    },
+    {
+      "epoch": 2.3166441136671176,
+      "grad_norm": 1.086673640477385,
+      "learning_rate": 1.499433783947054e-06,
+      "loss": 0.3238,
+      "step": 3424
+    },
+    {
+      "epoch": 2.317320703653586,
+      "grad_norm": 1.133152214240818,
+      "learning_rate": 1.4966238453401161e-06,
+      "loss": 0.327,
+      "step": 3425
+    },
+    {
+      "epoch": 2.317997293640054,
+      "grad_norm": 1.1056603902423796,
+      "learning_rate": 1.4938160786375571e-06,
+      "loss": 0.3035,
+      "step": 3426
+    },
+    {
+      "epoch": 2.318673883626522,
+      "grad_norm": 1.1143015228059376,
+      "learning_rate": 1.4910104855800429e-06,
+      "loss": 0.3076,
+      "step": 3427
+    },
+    {
+      "epoch": 2.3193504736129906,
+      "grad_norm": 1.1329223685252325,
+      "learning_rate": 1.488207067906891e-06,
+      "loss": 0.3231,
+      "step": 3428
+    },
+    {
+      "epoch": 2.3200270635994586,
+      "grad_norm": 1.081248642990499,
+      "learning_rate": 1.4854058273560667e-06,
+      "loss": 0.3142,
+      "step": 3429
+    },
+    {
+      "epoch": 2.320703653585927,
+      "grad_norm": 1.1368462902893035,
+      "learning_rate": 1.4826067656641912e-06,
+      "loss": 0.3229,
+      "step": 3430
+    },
+    {
+      "epoch": 2.321380243572395,
+      "grad_norm": 1.1569343820591698,
+      "learning_rate": 1.479809884566528e-06,
+      "loss": 0.3266,
+      "step": 3431
+    },
+    {
+      "epoch": 2.322056833558863,
+      "grad_norm": 1.0941622455360218,
+      "learning_rate": 1.477015185796995e-06,
+      "loss": 0.3027,
+      "step": 3432
+    },
+    {
+      "epoch": 2.3227334235453316,
+      "grad_norm": 1.077477715606638,
+      "learning_rate": 1.4742226710881558e-06,
+      "loss": 0.3058,
+      "step": 3433
+    },
+    {
+      "epoch": 2.3234100135317997,
+      "grad_norm": 1.1232028136964634,
+      "learning_rate": 1.4714323421712163e-06,
+      "loss": 0.3106,
+      "step": 3434
+    },
+    {
+      "epoch": 2.324086603518268,
+      "grad_norm": 1.1789796526479444,
+      "learning_rate": 1.4686442007760315e-06,
+      "loss": 0.3343,
+      "step": 3435
+    },
+    {
+      "epoch": 2.324763193504736,
+      "grad_norm": 1.10183200367498,
+      "learning_rate": 1.465858248631099e-06,
+      "loss": 0.3107,
+      "step": 3436
+    },
+    {
+      "epoch": 2.325439783491204,
+      "grad_norm": 1.1268963270372936,
+      "learning_rate": 1.4630744874635611e-06,
+      "loss": 0.3143,
+      "step": 3437
+    },
+    {
+      "epoch": 2.3261163734776726,
+      "grad_norm": 1.1019247955445997,
+      "learning_rate": 1.460292918999195e-06,
+      "loss": 0.3029,
+      "step": 3438
+    },
+    {
+      "epoch": 2.3267929634641407,
+      "grad_norm": 1.1103547542753434,
+      "learning_rate": 1.4575135449624251e-06,
+      "loss": 0.3126,
+      "step": 3439
+    },
+    {
+      "epoch": 2.3274695534506087,
+      "grad_norm": 1.1700727872289949,
+      "learning_rate": 1.4547363670763138e-06,
+      "loss": 0.3133,
+      "step": 3440
+    },
+    {
+      "epoch": 2.328146143437077,
+      "grad_norm": 1.102597220280328,
+      "learning_rate": 1.4519613870625632e-06,
+      "loss": 0.303,
+      "step": 3441
+    },
+    {
+      "epoch": 2.328822733423545,
+      "grad_norm": 1.0846839260550367,
+      "learning_rate": 1.4491886066415084e-06,
+      "loss": 0.3003,
+      "step": 3442
+    },
+    {
+      "epoch": 2.3294993234100136,
+      "grad_norm": 1.158268108344539,
+      "learning_rate": 1.4464180275321255e-06,
+      "loss": 0.3206,
+      "step": 3443
+    },
+    {
+      "epoch": 2.3301759133964817,
+      "grad_norm": 1.1399961003317345,
+      "learning_rate": 1.4436496514520253e-06,
+      "loss": 0.3177,
+      "step": 3444
+    },
+    {
+      "epoch": 2.3308525033829497,
+      "grad_norm": 1.164233915510866,
+      "learning_rate": 1.4408834801174492e-06,
+      "loss": 0.3269,
+      "step": 3445
+    },
+    {
+      "epoch": 2.331529093369418,
+      "grad_norm": 1.1350971318386764,
+      "learning_rate": 1.438119515243277e-06,
+      "loss": 0.3118,
+      "step": 3446
+    },
+    {
+      "epoch": 2.332205683355886,
+      "grad_norm": 1.1451605130743008,
+      "learning_rate": 1.4353577585430152e-06,
+      "loss": 0.3265,
+      "step": 3447
+    },
+    {
+      "epoch": 2.3328822733423547,
+      "grad_norm": 1.1346672692917785,
+      "learning_rate": 1.4325982117288052e-06,
+      "loss": 0.3218,
+      "step": 3448
+    },
+    {
+      "epoch": 2.3335588633288227,
+      "grad_norm": 1.0893708151229857,
+      "learning_rate": 1.4298408765114191e-06,
+      "loss": 0.3106,
+      "step": 3449
+    },
+    {
+      "epoch": 2.3342354533152907,
+      "grad_norm": 1.1403266646337968,
+      "learning_rate": 1.4270857546002548e-06,
+      "loss": 0.3138,
+      "step": 3450
+    },
+    {
+      "epoch": 2.334912043301759,
+      "grad_norm": 1.0884277973793046,
+      "learning_rate": 1.4243328477033369e-06,
+      "loss": 0.3169,
+      "step": 3451
+    },
+    {
+      "epoch": 2.335588633288227,
+      "grad_norm": 1.117831052522907,
+      "learning_rate": 1.4215821575273219e-06,
+      "loss": 0.297,
+      "step": 3452
+    },
+    {
+      "epoch": 2.3362652232746957,
+      "grad_norm": 1.1159003918941182,
+      "learning_rate": 1.4188336857774892e-06,
+      "loss": 0.3152,
+      "step": 3453
+    },
+    {
+      "epoch": 2.3369418132611637,
+      "grad_norm": 1.1312782791706404,
+      "learning_rate": 1.4160874341577447e-06,
+      "loss": 0.3234,
+      "step": 3454
+    },
+    {
+      "epoch": 2.3376184032476317,
+      "grad_norm": 1.111147460466852,
+      "learning_rate": 1.413343404370613e-06,
+      "loss": 0.3147,
+      "step": 3455
+    },
+    {
+      "epoch": 2.3382949932341,
+      "grad_norm": 1.0705415087278132,
+      "learning_rate": 1.410601598117246e-06,
+      "loss": 0.3079,
+      "step": 3456
+    },
+    {
+      "epoch": 2.338971583220568,
+      "grad_norm": 1.1260681822779628,
+      "learning_rate": 1.4078620170974178e-06,
+      "loss": 0.3162,
+      "step": 3457
+    },
+    {
+      "epoch": 2.3396481732070367,
+      "grad_norm": 1.061992178188652,
+      "learning_rate": 1.4051246630095195e-06,
+      "loss": 0.3001,
+      "step": 3458
+    },
+    {
+      "epoch": 2.3403247631935047,
+      "grad_norm": 1.146386396406404,
+      "learning_rate": 1.4023895375505608e-06,
+      "loss": 0.3188,
+      "step": 3459
+    },
+    {
+      "epoch": 2.3410013531799727,
+      "grad_norm": 1.1256372502555996,
+      "learning_rate": 1.3996566424161746e-06,
+      "loss": 0.3267,
+      "step": 3460
+    },
+    {
+      "epoch": 2.341677943166441,
+      "grad_norm": 1.1170601039612857,
+      "learning_rate": 1.396925979300608e-06,
+      "loss": 0.313,
+      "step": 3461
+    },
+    {
+      "epoch": 2.342354533152909,
+      "grad_norm": 1.068555772983513,
+      "learning_rate": 1.3941975498967265e-06,
+      "loss": 0.2924,
+      "step": 3462
+    },
+    {
+      "epoch": 2.3430311231393777,
+      "grad_norm": 1.0854556175917942,
+      "learning_rate": 1.3914713558960064e-06,
+      "loss": 0.3109,
+      "step": 3463
+    },
+    {
+      "epoch": 2.3437077131258457,
+      "grad_norm": 1.140675727614274,
+      "learning_rate": 1.3887473989885441e-06,
+      "loss": 0.3198,
+      "step": 3464
+    },
+    {
+      "epoch": 2.3443843031123137,
+      "grad_norm": 1.1438387264589154,
+      "learning_rate": 1.3860256808630429e-06,
+      "loss": 0.3188,
+      "step": 3465
+    },
+    {
+      "epoch": 2.345060893098782,
+      "grad_norm": 1.0739040090609289,
+      "learning_rate": 1.383306203206823e-06,
+      "loss": 0.3096,
+      "step": 3466
+    },
+    {
+      "epoch": 2.34573748308525,
+      "grad_norm": 1.1026162523417182,
+      "learning_rate": 1.3805889677058148e-06,
+      "loss": 0.3151,
+      "step": 3467
+    },
+    {
+      "epoch": 2.3464140730717187,
+      "grad_norm": 1.143705999203667,
+      "learning_rate": 1.3778739760445552e-06,
+      "loss": 0.3306,
+      "step": 3468
+    },
+    {
+      "epoch": 2.3470906630581867,
+      "grad_norm": 1.0775382767982196,
+      "learning_rate": 1.375161229906195e-06,
+      "loss": 0.3095,
+      "step": 3469
+    },
+    {
+      "epoch": 2.3477672530446547,
+      "grad_norm": 1.171476023297586,
+      "learning_rate": 1.372450730972491e-06,
+      "loss": 0.3255,
+      "step": 3470
+    },
+    {
+      "epoch": 2.348443843031123,
+      "grad_norm": 1.121195342490227,
+      "learning_rate": 1.3697424809238058e-06,
+      "loss": 0.3165,
+      "step": 3471
+    },
+    {
+      "epoch": 2.349120433017591,
+      "grad_norm": 1.1290495914505805,
+      "learning_rate": 1.3670364814391062e-06,
+      "loss": 0.3286,
+      "step": 3472
+    },
+    {
+      "epoch": 2.3497970230040597,
+      "grad_norm": 1.159182727891767,
+      "learning_rate": 1.3643327341959684e-06,
+      "loss": 0.3287,
+      "step": 3473
+    },
+    {
+      "epoch": 2.3504736129905277,
+      "grad_norm": 1.1507321436774485,
+      "learning_rate": 1.361631240870569e-06,
+      "loss": 0.3233,
+      "step": 3474
+    },
+    {
+      "epoch": 2.3511502029769957,
+      "grad_norm": 1.112282443636384,
+      "learning_rate": 1.35893200313769e-06,
+      "loss": 0.3213,
+      "step": 3475
+    },
+    {
+      "epoch": 2.351826792963464,
+      "grad_norm": 1.140080162349322,
+      "learning_rate": 1.3562350226707106e-06,
+      "loss": 0.3233,
+      "step": 3476
+    },
+    {
+      "epoch": 2.352503382949932,
+      "grad_norm": 1.1200162002913803,
+      "learning_rate": 1.3535403011416159e-06,
+      "loss": 0.3194,
+      "step": 3477
+    },
+    {
+      "epoch": 2.3531799729364007,
+      "grad_norm": 1.122094756099197,
+      "learning_rate": 1.3508478402209858e-06,
+      "loss": 0.3222,
+      "step": 3478
+    },
+    {
+      "epoch": 2.3538565629228687,
+      "grad_norm": 1.1236134275566492,
+      "learning_rate": 1.3481576415780035e-06,
+      "loss": 0.3211,
+      "step": 3479
+    },
+    {
+      "epoch": 2.3545331529093367,
+      "grad_norm": 1.0899993630047502,
+      "learning_rate": 1.3454697068804434e-06,
+      "loss": 0.3041,
+      "step": 3480
+    },
+    {
+      "epoch": 2.355209742895805,
+      "grad_norm": 1.1664970485021213,
+      "learning_rate": 1.3427840377946826e-06,
+      "loss": 0.3195,
+      "step": 3481
+    },
+    {
+      "epoch": 2.3558863328822732,
+      "grad_norm": 1.1441128613778109,
+      "learning_rate": 1.3401006359856916e-06,
+      "loss": 0.3263,
+      "step": 3482
+    },
+    {
+      "epoch": 2.3565629228687417,
+      "grad_norm": 1.1061354898928037,
+      "learning_rate": 1.337419503117035e-06,
+      "loss": 0.3041,
+      "step": 3483
+    },
+    {
+      "epoch": 2.3572395128552097,
+      "grad_norm": 1.1073889865134197,
+      "learning_rate": 1.3347406408508695e-06,
+      "loss": 0.3039,
+      "step": 3484
+    },
+    {
+      "epoch": 2.3579161028416777,
+      "grad_norm": 1.128298768612049,
+      "learning_rate": 1.332064050847945e-06,
+      "loss": 0.3204,
+      "step": 3485
+    },
+    {
+      "epoch": 2.358592692828146,
+      "grad_norm": 1.143187580828165,
+      "learning_rate": 1.3293897347676032e-06,
+      "loss": 0.3195,
+      "step": 3486
+    },
+    {
+      "epoch": 2.3592692828146142,
+      "grad_norm": 1.1433164543433265,
+      "learning_rate": 1.3267176942677763e-06,
+      "loss": 0.3168,
+      "step": 3487
+    },
+    {
+      "epoch": 2.3599458728010827,
+      "grad_norm": 1.1095837297397795,
+      "learning_rate": 1.324047931004987e-06,
+      "loss": 0.3213,
+      "step": 3488
+    },
+    {
+      "epoch": 2.3606224627875507,
+      "grad_norm": 1.181348578466833,
+      "learning_rate": 1.321380446634342e-06,
+      "loss": 0.3142,
+      "step": 3489
+    },
+    {
+      "epoch": 2.3612990527740187,
+      "grad_norm": 1.1002417074246151,
+      "learning_rate": 1.31871524280954e-06,
+      "loss": 0.3046,
+      "step": 3490
+    },
+    {
+      "epoch": 2.361975642760487,
+      "grad_norm": 1.1461889181285505,
+      "learning_rate": 1.3160523211828612e-06,
+      "loss": 0.3255,
+      "step": 3491
+    },
+    {
+      "epoch": 2.3626522327469552,
+      "grad_norm": 1.1683634006387886,
+      "learning_rate": 1.313391683405177e-06,
+      "loss": 0.3304,
+      "step": 3492
+    },
+    {
+      "epoch": 2.3633288227334237,
+      "grad_norm": 1.1264272724900595,
+      "learning_rate": 1.310733331125935e-06,
+      "loss": 0.2987,
+      "step": 3493
+    },
+    {
+      "epoch": 2.3640054127198917,
+      "grad_norm": 1.1110478101900172,
+      "learning_rate": 1.3080772659931728e-06,
+      "loss": 0.3115,
+      "step": 3494
+    },
+    {
+      "epoch": 2.3646820027063598,
+      "grad_norm": 1.1120632764431906,
+      "learning_rate": 1.305423489653508e-06,
+      "loss": 0.3215,
+      "step": 3495
+    },
+    {
+      "epoch": 2.365358592692828,
+      "grad_norm": 1.1074600373325922,
+      "learning_rate": 1.3027720037521397e-06,
+      "loss": 0.3067,
+      "step": 3496
+    },
+    {
+      "epoch": 2.3660351826792962,
+      "grad_norm": 1.3468500227329978,
+      "learning_rate": 1.3001228099328445e-06,
+      "loss": 0.3019,
+      "step": 3497
+    },
+    {
+      "epoch": 2.3667117726657647,
+      "grad_norm": 1.1118779652272481,
+      "learning_rate": 1.297475909837979e-06,
+      "loss": 0.3175,
+      "step": 3498
+    },
+    {
+      "epoch": 2.3673883626522327,
+      "grad_norm": 1.1453009657972386,
+      "learning_rate": 1.29483130510848e-06,
+      "loss": 0.3391,
+      "step": 3499
+    },
+    {
+      "epoch": 2.3680649526387008,
+      "grad_norm": 1.1385287275376463,
+      "learning_rate": 1.2921889973838591e-06,
+      "loss": 0.3217,
+      "step": 3500
+    },
+    {
+      "epoch": 2.3687415426251692,
+      "grad_norm": 1.113156750144898,
+      "learning_rate": 1.289548988302207e-06,
+      "loss": 0.3083,
+      "step": 3501
+    },
+    {
+      "epoch": 2.3694181326116373,
+      "grad_norm": 1.1411851415754353,
+      "learning_rate": 1.2869112795001836e-06,
+      "loss": 0.323,
+      "step": 3502
+    },
+    {
+      "epoch": 2.3700947225981057,
+      "grad_norm": 1.1124203263727617,
+      "learning_rate": 1.2842758726130283e-06,
+      "loss": 0.3232,
+      "step": 3503
+    },
+    {
+      "epoch": 2.3707713125845737,
+      "grad_norm": 1.0845956699853068,
+      "learning_rate": 1.281642769274552e-06,
+      "loss": 0.3101,
+      "step": 3504
+    },
+    {
+      "epoch": 2.3714479025710418,
+      "grad_norm": 1.1343768372333758,
+      "learning_rate": 1.2790119711171356e-06,
+      "loss": 0.3124,
+      "step": 3505
+    },
+    {
+      "epoch": 2.3721244925575102,
+      "grad_norm": 1.1211238121509965,
+      "learning_rate": 1.2763834797717312e-06,
+      "loss": 0.311,
+      "step": 3506
+    },
+    {
+      "epoch": 2.3728010825439783,
+      "grad_norm": 1.1325954165849101,
+      "learning_rate": 1.2737572968678624e-06,
+      "loss": 0.3122,
+      "step": 3507
+    },
+    {
+      "epoch": 2.3734776725304467,
+      "grad_norm": 1.1217703491721847,
+      "learning_rate": 1.2711334240336216e-06,
+      "loss": 0.3152,
+      "step": 3508
+    },
+    {
+      "epoch": 2.3741542625169147,
+      "grad_norm": 1.163818266836355,
+      "learning_rate": 1.26851186289567e-06,
+      "loss": 0.325,
+      "step": 3509
+    },
+    {
+      "epoch": 2.3748308525033828,
+      "grad_norm": 1.1343757479060412,
+      "learning_rate": 1.2658926150792321e-06,
+      "loss": 0.3318,
+      "step": 3510
+    },
+    {
+      "epoch": 2.3755074424898512,
+      "grad_norm": 1.0987108606316085,
+      "learning_rate": 1.2632756822081e-06,
+      "loss": 0.3112,
+      "step": 3511
+    },
+    {
+      "epoch": 2.3761840324763193,
+      "grad_norm": 1.1484261921654009,
+      "learning_rate": 1.2606610659046314e-06,
+      "loss": 0.3193,
+      "step": 3512
+    },
+    {
+      "epoch": 2.3768606224627877,
+      "grad_norm": 1.1631813605994743,
+      "learning_rate": 1.2580487677897496e-06,
+      "loss": 0.3256,
+      "step": 3513
+    },
+    {
+      "epoch": 2.3775372124492558,
+      "grad_norm": 1.1514927477839263,
+      "learning_rate": 1.255438789482935e-06,
+      "loss": 0.323,
+      "step": 3514
+    },
+    {
+      "epoch": 2.378213802435724,
+      "grad_norm": 1.1259927778312828,
+      "learning_rate": 1.2528311326022364e-06,
+      "loss": 0.3093,
+      "step": 3515
+    },
+    {
+      "epoch": 2.3788903924221922,
+      "grad_norm": 1.1632793982771636,
+      "learning_rate": 1.250225798764259e-06,
+      "loss": 0.3299,
+      "step": 3516
+    },
+    {
+      "epoch": 2.3795669824086603,
+      "grad_norm": 1.1730467726536382,
+      "learning_rate": 1.2476227895841714e-06,
+      "loss": 0.3149,
+      "step": 3517
+    },
+    {
+      "epoch": 2.3802435723951287,
+      "grad_norm": 1.10142032031152,
+      "learning_rate": 1.2450221066756973e-06,
+      "loss": 0.3234,
+      "step": 3518
+    },
+    {
+      "epoch": 2.3809201623815968,
+      "grad_norm": 1.1095396209156525,
+      "learning_rate": 1.242423751651119e-06,
+      "loss": 0.3169,
+      "step": 3519
+    },
+    {
+      "epoch": 2.381596752368065,
+      "grad_norm": 1.1286624368515654,
+      "learning_rate": 1.2398277261212777e-06,
+      "loss": 0.3287,
+      "step": 3520
+    },
+    {
+      "epoch": 2.3822733423545333,
+      "grad_norm": 1.1408801632686176,
+      "learning_rate": 1.2372340316955694e-06,
+      "loss": 0.3255,
+      "step": 3521
+    },
+    {
+      "epoch": 2.3829499323410013,
+      "grad_norm": 1.1051274743799062,
+      "learning_rate": 1.234642669981946e-06,
+      "loss": 0.3131,
+      "step": 3522
+    },
+    {
+      "epoch": 2.3836265223274697,
+      "grad_norm": 1.124440199110235,
+      "learning_rate": 1.232053642586909e-06,
+      "loss": 0.3162,
+      "step": 3523
+    },
+    {
+      "epoch": 2.3843031123139378,
+      "grad_norm": 1.1538161779736553,
+      "learning_rate": 1.2294669511155193e-06,
+      "loss": 0.324,
+      "step": 3524
+    },
+    {
+      "epoch": 2.384979702300406,
+      "grad_norm": 1.085256280637149,
+      "learning_rate": 1.2268825971713833e-06,
+      "loss": 0.3147,
+      "step": 3525
+    },
+    {
+      "epoch": 2.3856562922868743,
+      "grad_norm": 1.141637889491272,
+      "learning_rate": 1.2243005823566638e-06,
+      "loss": 0.3165,
+      "step": 3526
+    },
+    {
+      "epoch": 2.3863328822733423,
+      "grad_norm": 1.1273807671838103,
+      "learning_rate": 1.2217209082720677e-06,
+      "loss": 0.321,
+      "step": 3527
+    },
+    {
+      "epoch": 2.3870094722598107,
+      "grad_norm": 1.1446425478149072,
+      "learning_rate": 1.2191435765168557e-06,
+      "loss": 0.3085,
+      "step": 3528
+    },
+    {
+      "epoch": 2.3876860622462788,
+      "grad_norm": 1.138276430666935,
+      "learning_rate": 1.2165685886888346e-06,
+      "loss": 0.3232,
+      "step": 3529
+    },
+    {
+      "epoch": 2.388362652232747,
+      "grad_norm": 1.1092810231609838,
+      "learning_rate": 1.2139959463843593e-06,
+      "loss": 0.3014,
+      "step": 3530
+    },
+    {
+      "epoch": 2.3890392422192153,
+      "grad_norm": 1.1060100759802696,
+      "learning_rate": 1.2114256511983274e-06,
+      "loss": 0.3088,
+      "step": 3531
+    },
+    {
+      "epoch": 2.3897158322056833,
+      "grad_norm": 1.1611344190662929,
+      "learning_rate": 1.2088577047241834e-06,
+      "loss": 0.3336,
+      "step": 3532
+    },
+    {
+      "epoch": 2.3903924221921518,
+      "grad_norm": 1.1127045231776416,
+      "learning_rate": 1.2062921085539152e-06,
+      "loss": 0.3165,
+      "step": 3533
+    },
+    {
+      "epoch": 2.39106901217862,
+      "grad_norm": 1.1749429541623873,
+      "learning_rate": 1.2037288642780575e-06,
+      "loss": 0.3349,
+      "step": 3534
+    },
+    {
+      "epoch": 2.391745602165088,
+      "grad_norm": 1.1400185826065576,
+      "learning_rate": 1.2011679734856796e-06,
+      "loss": 0.3231,
+      "step": 3535
+    },
+    {
+      "epoch": 2.3924221921515563,
+      "grad_norm": 1.1196237995790301,
+      "learning_rate": 1.1986094377643976e-06,
+      "loss": 0.3134,
+      "step": 3536
+    },
+    {
+      "epoch": 2.3930987821380243,
+      "grad_norm": 1.1317444371874341,
+      "learning_rate": 1.1960532587003666e-06,
+      "loss": 0.304,
+      "step": 3537
+    },
+    {
+      "epoch": 2.3937753721244928,
+      "grad_norm": 1.200027462821917,
+      "learning_rate": 1.193499437878277e-06,
+      "loss": 0.3248,
+      "step": 3538
+    },
+    {
+      "epoch": 2.394451962110961,
+      "grad_norm": 1.1030282179931525,
+      "learning_rate": 1.1909479768813641e-06,
+      "loss": 0.3128,
+      "step": 3539
+    },
+    {
+      "epoch": 2.395128552097429,
+      "grad_norm": 1.1240595568357,
+      "learning_rate": 1.1883988772913924e-06,
+      "loss": 0.3139,
+      "step": 3540
+    },
+    {
+      "epoch": 2.3958051420838973,
+      "grad_norm": 1.1469755819623062,
+      "learning_rate": 1.1858521406886674e-06,
+      "loss": 0.3224,
+      "step": 3541
+    },
+    {
+      "epoch": 2.3964817320703653,
+      "grad_norm": 1.1327222168099689,
+      "learning_rate": 1.183307768652029e-06,
+      "loss": 0.3146,
+      "step": 3542
+    },
+    {
+      "epoch": 2.3971583220568338,
+      "grad_norm": 1.112350043947951,
+      "learning_rate": 1.180765762758852e-06,
+      "loss": 0.3168,
+      "step": 3543
+    },
+    {
+      "epoch": 2.397834912043302,
+      "grad_norm": 1.1050977029991658,
+      "learning_rate": 1.1782261245850417e-06,
+      "loss": 0.3214,
+      "step": 3544
+    },
+    {
+      "epoch": 2.39851150202977,
+      "grad_norm": 1.120220381893809,
+      "learning_rate": 1.1756888557050356e-06,
+      "loss": 0.3258,
+      "step": 3545
+    },
+    {
+      "epoch": 2.3991880920162383,
+      "grad_norm": 1.0982917353481483,
+      "learning_rate": 1.173153957691805e-06,
+      "loss": 0.3028,
+      "step": 3546
+    },
+    {
+      "epoch": 2.3998646820027063,
+      "grad_norm": 1.1290017099554381,
+      "learning_rate": 1.1706214321168513e-06,
+      "loss": 0.316,
+      "step": 3547
+    },
+    {
+      "epoch": 2.4005412719891748,
+      "grad_norm": 1.1122629928280499,
+      "learning_rate": 1.1680912805502008e-06,
+      "loss": 0.3099,
+      "step": 3548
+    },
+    {
+      "epoch": 2.401217861975643,
+      "grad_norm": 1.1318301757696607,
+      "learning_rate": 1.165563504560413e-06,
+      "loss": 0.3254,
+      "step": 3549
+    },
+    {
+      "epoch": 2.401894451962111,
+      "grad_norm": 1.1000619570162524,
+      "learning_rate": 1.1630381057145735e-06,
+      "loss": 0.2996,
+      "step": 3550
+    },
+    {
+      "epoch": 2.4025710419485793,
+      "grad_norm": 1.1444264595967888,
+      "learning_rate": 1.1605150855782916e-06,
+      "loss": 0.3231,
+      "step": 3551
+    },
+    {
+      "epoch": 2.4032476319350473,
+      "grad_norm": 1.110204315005672,
+      "learning_rate": 1.157994445715706e-06,
+      "loss": 0.3154,
+      "step": 3552
+    },
+    {
+      "epoch": 2.403924221921516,
+      "grad_norm": 1.1053213243824813,
+      "learning_rate": 1.155476187689475e-06,
+      "loss": 0.3133,
+      "step": 3553
+    },
+    {
+      "epoch": 2.404600811907984,
+      "grad_norm": 1.15353184046916,
+      "learning_rate": 1.1529603130607837e-06,
+      "loss": 0.3167,
+      "step": 3554
+    },
+    {
+      "epoch": 2.405277401894452,
+      "grad_norm": 1.1362190763890363,
+      "learning_rate": 1.1504468233893408e-06,
+      "loss": 0.3198,
+      "step": 3555
+    },
+    {
+      "epoch": 2.4059539918809203,
+      "grad_norm": 1.179608172565361,
+      "learning_rate": 1.1479357202333707e-06,
+      "loss": 0.323,
+      "step": 3556
+    },
+    {
+      "epoch": 2.4066305818673883,
+      "grad_norm": 1.087135666523029,
+      "learning_rate": 1.1454270051496264e-06,
+      "loss": 0.3042,
+      "step": 3557
+    },
+    {
+      "epoch": 2.407307171853857,
+      "grad_norm": 1.181982189311583,
+      "learning_rate": 1.1429206796933717e-06,
+      "loss": 0.3278,
+      "step": 3558
+    },
+    {
+      "epoch": 2.407983761840325,
+      "grad_norm": 1.1681944729415519,
+      "learning_rate": 1.1404167454183957e-06,
+      "loss": 0.3277,
+      "step": 3559
+    },
+    {
+      "epoch": 2.408660351826793,
+      "grad_norm": 1.1340177948443406,
+      "learning_rate": 1.137915203877003e-06,
+      "loss": 0.3221,
+      "step": 3560
+    },
+    {
+      "epoch": 2.4093369418132613,
+      "grad_norm": 1.1176795014616392,
+      "learning_rate": 1.1354160566200128e-06,
+      "loss": 0.3026,
+      "step": 3561
+    },
+    {
+      "epoch": 2.4100135317997293,
+      "grad_norm": 1.144195183861141,
+      "learning_rate": 1.132919305196763e-06,
+      "loss": 0.3098,
+      "step": 3562
+    },
+    {
+      "epoch": 2.410690121786198,
+      "grad_norm": 1.1347214615831074,
+      "learning_rate": 1.130424951155104e-06,
+      "loss": 0.3133,
+      "step": 3563
+    },
+    {
+      "epoch": 2.411366711772666,
+      "grad_norm": 1.126091562477791,
+      "learning_rate": 1.1279329960414047e-06,
+      "loss": 0.3092,
+      "step": 3564
+    },
+    {
+      "epoch": 2.412043301759134,
+      "grad_norm": 1.1729908809034137,
+      "learning_rate": 1.1254434414005367e-06,
+      "loss": 0.3202,
+      "step": 3565
+    },
+    {
+      "epoch": 2.4127198917456023,
+      "grad_norm": 1.1438036146000217,
+      "learning_rate": 1.1229562887758927e-06,
+      "loss": 0.3068,
+      "step": 3566
+    },
+    {
+      "epoch": 2.4133964817320703,
+      "grad_norm": 1.161876628504045,
+      "learning_rate": 1.1204715397093735e-06,
+      "loss": 0.3167,
+      "step": 3567
+    },
+    {
+      "epoch": 2.414073071718539,
+      "grad_norm": 1.1352831581419485,
+      "learning_rate": 1.1179891957413908e-06,
+      "loss": 0.3206,
+      "step": 3568
+    },
+    {
+      "epoch": 2.414749661705007,
+      "grad_norm": 1.1379129348287849,
+      "learning_rate": 1.1155092584108606e-06,
+      "loss": 0.329,
+      "step": 3569
+    },
+    {
+      "epoch": 2.415426251691475,
+      "grad_norm": 1.1582938649505607,
+      "learning_rate": 1.113031729255214e-06,
+      "loss": 0.3138,
+      "step": 3570
+    },
+    {
+      "epoch": 2.4161028416779433,
+      "grad_norm": 1.110109841769803,
+      "learning_rate": 1.1105566098103825e-06,
+      "loss": 0.3059,
+      "step": 3571
+    },
+    {
+      "epoch": 2.4167794316644113,
+      "grad_norm": 1.0712589097819936,
+      "learning_rate": 1.1080839016108086e-06,
+      "loss": 0.2975,
+      "step": 3572
+    },
+    {
+      "epoch": 2.41745602165088,
+      "grad_norm": 1.087618111361069,
+      "learning_rate": 1.1056136061894386e-06,
+      "loss": 0.3076,
+      "step": 3573
+    },
+    {
+      "epoch": 2.418132611637348,
+      "grad_norm": 1.095089762234527,
+      "learning_rate": 1.1031457250777206e-06,
+      "loss": 0.3125,
+      "step": 3574
+    },
+    {
+      "epoch": 2.418809201623816,
+      "grad_norm": 1.1571240889674608,
+      "learning_rate": 1.1006802598056081e-06,
+      "loss": 0.316,
+      "step": 3575
+    },
+    {
+      "epoch": 2.4194857916102843,
+      "grad_norm": 1.1342395599392372,
+      "learning_rate": 1.0982172119015594e-06,
+      "loss": 0.3187,
+      "step": 3576
+    },
+    {
+      "epoch": 2.4201623815967523,
+      "grad_norm": 1.1303956719686752,
+      "learning_rate": 1.0957565828925292e-06,
+      "loss": 0.3159,
+      "step": 3577
+    },
+    {
+      "epoch": 2.420838971583221,
+      "grad_norm": 1.1460891396861255,
+      "learning_rate": 1.0932983743039739e-06,
+      "loss": 0.3195,
+      "step": 3578
+    },
+    {
+      "epoch": 2.421515561569689,
+      "grad_norm": 1.150753764974417,
+      "learning_rate": 1.0908425876598512e-06,
+      "loss": 0.3206,
+      "step": 3579
+    },
+    {
+      "epoch": 2.422192151556157,
+      "grad_norm": 1.1620928025993469,
+      "learning_rate": 1.0883892244826173e-06,
+      "loss": 0.3271,
+      "step": 3580
+    },
+    {
+      "epoch": 2.4228687415426253,
+      "grad_norm": 1.1908733356947119,
+      "learning_rate": 1.0859382862932255e-06,
+      "loss": 0.3278,
+      "step": 3581
+    },
+    {
+      "epoch": 2.4235453315290933,
+      "grad_norm": 1.1332923988073482,
+      "learning_rate": 1.0834897746111233e-06,
+      "loss": 0.3172,
+      "step": 3582
+    },
+    {
+      "epoch": 2.424221921515562,
+      "grad_norm": 1.1109533239182614,
+      "learning_rate": 1.0810436909542571e-06,
+      "loss": 0.3126,
+      "step": 3583
+    },
+    {
+      "epoch": 2.42489851150203,
+      "grad_norm": 1.1261381580784227,
+      "learning_rate": 1.0786000368390686e-06,
+      "loss": 0.3127,
+      "step": 3584
+    },
+    {
+      "epoch": 2.425575101488498,
+      "grad_norm": 1.1294567080481843,
+      "learning_rate": 1.0761588137804896e-06,
+      "loss": 0.324,
+      "step": 3585
+    },
+    {
+      "epoch": 2.4262516914749663,
+      "grad_norm": 1.1844730566208894,
+      "learning_rate": 1.0737200232919465e-06,
+      "loss": 0.3321,
+      "step": 3586
+    },
+    {
+      "epoch": 2.4269282814614344,
+      "grad_norm": 1.14565636803996,
+      "learning_rate": 1.0712836668853583e-06,
+      "loss": 0.3156,
+      "step": 3587
+    },
+    {
+      "epoch": 2.4276048714479024,
+      "grad_norm": 1.1524306442941032,
+      "learning_rate": 1.0688497460711345e-06,
+      "loss": 0.333,
+      "step": 3588
+    },
+    {
+      "epoch": 2.428281461434371,
+      "grad_norm": 1.121942649676295,
+      "learning_rate": 1.0664182623581777e-06,
+      "loss": 0.3201,
+      "step": 3589
+    },
+    {
+      "epoch": 2.428958051420839,
+      "grad_norm": 1.1529333785239038,
+      "learning_rate": 1.0639892172538734e-06,
+      "loss": 0.3103,
+      "step": 3590
+    },
+    {
+      "epoch": 2.4296346414073073,
+      "grad_norm": 1.1193930381983288,
+      "learning_rate": 1.0615626122640988e-06,
+      "loss": 0.3104,
+      "step": 3591
+    },
+    {
+      "epoch": 2.4303112313937754,
+      "grad_norm": 1.1230642424158743,
+      "learning_rate": 1.0591384488932188e-06,
+      "loss": 0.3109,
+      "step": 3592
+    },
+    {
+      "epoch": 2.4309878213802434,
+      "grad_norm": 1.1530014973225775,
+      "learning_rate": 1.0567167286440844e-06,
+      "loss": 0.3221,
+      "step": 3593
+    },
+    {
+      "epoch": 2.431664411366712,
+      "grad_norm": 1.0807062861300396,
+      "learning_rate": 1.0542974530180327e-06,
+      "loss": 0.2891,
+      "step": 3594
+    },
+    {
+      "epoch": 2.43234100135318,
+      "grad_norm": 1.1589482864780765,
+      "learning_rate": 1.0518806235148814e-06,
+      "loss": 0.311,
+      "step": 3595
+    },
+    {
+      "epoch": 2.4330175913396483,
+      "grad_norm": 1.1586065424296987,
+      "learning_rate": 1.0494662416329366e-06,
+      "loss": 0.3228,
+      "step": 3596
+    },
+    {
+      "epoch": 2.4336941813261164,
+      "grad_norm": 1.1502240836292057,
+      "learning_rate": 1.0470543088689855e-06,
+      "loss": 0.3094,
+      "step": 3597
+    },
+    {
+      "epoch": 2.4343707713125844,
+      "grad_norm": 1.112309337032747,
+      "learning_rate": 1.044644826718295e-06,
+      "loss": 0.3166,
+      "step": 3598
+    },
+    {
+      "epoch": 2.435047361299053,
+      "grad_norm": 1.086114625093356,
+      "learning_rate": 1.0422377966746133e-06,
+      "loss": 0.3067,
+      "step": 3599
+    },
+    {
+      "epoch": 2.435723951285521,
+      "grad_norm": 1.1052570795412353,
+      "learning_rate": 1.0398332202301708e-06,
+      "loss": 0.3015,
+      "step": 3600
+    },
+    {
+      "epoch": 2.4364005412719894,
+      "grad_norm": 1.1488816748291193,
+      "learning_rate": 1.0374310988756747e-06,
+      "loss": 0.3226,
+      "step": 3601
+    },
+    {
+      "epoch": 2.4370771312584574,
+      "grad_norm": 1.1266345056168994,
+      "learning_rate": 1.0350314341003121e-06,
+      "loss": 0.3096,
+      "step": 3602
+    },
+    {
+      "epoch": 2.4377537212449254,
+      "grad_norm": 1.1517626607813003,
+      "learning_rate": 1.0326342273917432e-06,
+      "loss": 0.3253,
+      "step": 3603
+    },
+    {
+      "epoch": 2.438430311231394,
+      "grad_norm": 1.145166831273789,
+      "learning_rate": 1.0302394802361104e-06,
+      "loss": 0.3154,
+      "step": 3604
+    },
+    {
+      "epoch": 2.439106901217862,
+      "grad_norm": 1.1386424656308904,
+      "learning_rate": 1.0278471941180245e-06,
+      "loss": 0.3291,
+      "step": 3605
+    },
+    {
+      "epoch": 2.4397834912043304,
+      "grad_norm": 1.1491871424478022,
+      "learning_rate": 1.0254573705205751e-06,
+      "loss": 0.3204,
+      "step": 3606
+    },
+    {
+      "epoch": 2.4404600811907984,
+      "grad_norm": 1.1608825064820834,
+      "learning_rate": 1.0230700109253255e-06,
+      "loss": 0.3171,
+      "step": 3607
+    },
+    {
+      "epoch": 2.4411366711772664,
+      "grad_norm": 1.123465447443952,
+      "learning_rate": 1.0206851168123078e-06,
+      "loss": 0.3153,
+      "step": 3608
+    },
+    {
+      "epoch": 2.441813261163735,
+      "grad_norm": 1.135981167360808,
+      "learning_rate": 1.0183026896600284e-06,
+      "loss": 0.3109,
+      "step": 3609
+    },
+    {
+      "epoch": 2.442489851150203,
+      "grad_norm": 1.193026211766432,
+      "learning_rate": 1.0159227309454662e-06,
+      "loss": 0.3332,
+      "step": 3610
+    },
+    {
+      "epoch": 2.4431664411366714,
+      "grad_norm": 1.1194518586529043,
+      "learning_rate": 1.0135452421440645e-06,
+      "loss": 0.3243,
+      "step": 3611
+    },
+    {
+      "epoch": 2.4438430311231394,
+      "grad_norm": 1.1622110630884213,
+      "learning_rate": 1.0111702247297372e-06,
+      "loss": 0.3215,
+      "step": 3612
+    },
+    {
+      "epoch": 2.4445196211096074,
+      "grad_norm": 1.112150959767557,
+      "learning_rate": 1.0087976801748694e-06,
+      "loss": 0.3081,
+      "step": 3613
+    },
+    {
+      "epoch": 2.445196211096076,
+      "grad_norm": 1.1441128793589497,
+      "learning_rate": 1.00642760995031e-06,
+      "loss": 0.3186,
+      "step": 3614
+    },
+    {
+      "epoch": 2.445872801082544,
+      "grad_norm": 1.1545232417516533,
+      "learning_rate": 1.0040600155253766e-06,
+      "loss": 0.306,
+      "step": 3615
+    },
+    {
+      "epoch": 2.4465493910690124,
+      "grad_norm": 1.119520686574386,
+      "learning_rate": 1.0016948983678471e-06,
+      "loss": 0.3152,
+      "step": 3616
+    },
+    {
+      "epoch": 2.4472259810554804,
+      "grad_norm": 1.1020040761457315,
+      "learning_rate": 9.993322599439692e-07,
+      "loss": 0.3116,
+      "step": 3617
+    },
+    {
+      "epoch": 2.4479025710419484,
+      "grad_norm": 1.1309211532895094,
+      "learning_rate": 9.969721017184492e-07,
+      "loss": 0.3124,
+      "step": 3618
+    },
+    {
+      "epoch": 2.448579161028417,
+      "grad_norm": 1.1247046357851842,
+      "learning_rate": 9.946144251544604e-07,
+      "loss": 0.3038,
+      "step": 3619
+    },
+    {
+      "epoch": 2.449255751014885,
+      "grad_norm": 1.1304169341813408,
+      "learning_rate": 9.92259231713632e-07,
+      "loss": 0.3112,
+      "step": 3620
+    },
+    {
+      "epoch": 2.449932341001353,
+      "grad_norm": 1.1656071517608024,
+      "learning_rate": 9.899065228560596e-07,
+      "loss": 0.3155,
+      "step": 3621
+    },
+    {
+      "epoch": 2.4506089309878214,
+      "grad_norm": 1.1295431320008458,
+      "learning_rate": 9.87556300040295e-07,
+      "loss": 0.3236,
+      "step": 3622
+    },
+    {
+      "epoch": 2.4512855209742894,
+      "grad_norm": 1.0970987982295464,
+      "learning_rate": 9.852085647233505e-07,
+      "loss": 0.3032,
+      "step": 3623
+    },
+    {
+      "epoch": 2.451962110960758,
+      "grad_norm": 1.1159102589736656,
+      "learning_rate": 9.82863318360695e-07,
+      "loss": 0.3136,
+      "step": 3624
+    },
+    {
+      "epoch": 2.452638700947226,
+      "grad_norm": 1.1545687563771772,
+      "learning_rate": 9.805205624062535e-07,
+      "loss": 0.3103,
+      "step": 3625
+    },
+    {
+      "epoch": 2.453315290933694,
+      "grad_norm": 1.1772105291245962,
+      "learning_rate": 9.781802983124094e-07,
+      "loss": 0.325,
+      "step": 3626
+    },
+    {
+      "epoch": 2.4539918809201624,
+      "grad_norm": 1.149176716633529,
+      "learning_rate": 9.758425275299998e-07,
+      "loss": 0.3211,
+      "step": 3627
+    },
+    {
+      "epoch": 2.4546684709066304,
+      "grad_norm": 1.2223225308978591,
+      "learning_rate": 9.735072515083193e-07,
+      "loss": 0.3395,
+      "step": 3628
+    },
+    {
+      "epoch": 2.455345060893099,
+      "grad_norm": 1.109822575849989,
+      "learning_rate": 9.711744716951093e-07,
+      "loss": 0.3083,
+      "step": 3629
+    },
+    {
+      "epoch": 2.456021650879567,
+      "grad_norm": 1.1270779951147665,
+      "learning_rate": 9.688441895365708e-07,
+      "loss": 0.3197,
+      "step": 3630
+    },
+    {
+      "epoch": 2.456698240866035,
+      "grad_norm": 1.1497565738756217,
+      "learning_rate": 9.665164064773496e-07,
+      "loss": 0.3256,
+      "step": 3631
+    },
+    {
+      "epoch": 2.4573748308525034,
+      "grad_norm": 1.0809595186273584,
+      "learning_rate": 9.641911239605494e-07,
+      "loss": 0.2968,
+      "step": 3632
+    },
+    {
+      "epoch": 2.4580514208389714,
+      "grad_norm": 1.1326728787965972,
+      "learning_rate": 9.618683434277176e-07,
+      "loss": 0.32,
+      "step": 3633
+    },
+    {
+      "epoch": 2.45872801082544,
+      "grad_norm": 1.11117448199408,
+      "learning_rate": 9.595480663188528e-07,
+      "loss": 0.3083,
+      "step": 3634
+    },
+    {
+      "epoch": 2.459404600811908,
+      "grad_norm": 1.0851801438798105,
+      "learning_rate": 9.572302940724032e-07,
+      "loss": 0.2976,
+      "step": 3635
+    },
+    {
+      "epoch": 2.460081190798376,
+      "grad_norm": 1.1355557016308093,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.3247,
+      "step": 3636
+    },
+    {
+      "epoch": 2.4607577807848444,
+      "grad_norm": 1.1499066283060941,
+      "learning_rate": 9.526022699127718e-07,
+      "loss": 0.3201,
+      "step": 3637
+    },
+    {
+      "epoch": 2.4614343707713124,
+      "grad_norm": 1.1387358327900405,
+      "learning_rate": 9.502920208687133e-07,
+      "loss": 0.3123,
+      "step": 3638
+    },
+    {
+      "epoch": 2.462110960757781,
+      "grad_norm": 1.075294062624437,
+      "learning_rate": 9.479842824253182e-07,
+      "loss": 0.2992,
+      "step": 3639
+    },
+    {
+      "epoch": 2.462787550744249,
+      "grad_norm": 1.1094037335250049,
+      "learning_rate": 9.456790560132617e-07,
+      "loss": 0.3005,
+      "step": 3640
+    },
+    {
+      "epoch": 2.463464140730717,
+      "grad_norm": 1.123900159236025,
+      "learning_rate": 9.433763430616577e-07,
+      "loss": 0.304,
+      "step": 3641
+    },
+    {
+      "epoch": 2.4641407307171854,
+      "grad_norm": 1.189332240219114,
+      "learning_rate": 9.410761449980654e-07,
+      "loss": 0.3239,
+      "step": 3642
+    },
+    {
+      "epoch": 2.4648173207036534,
+      "grad_norm": 1.1245416558411154,
+      "learning_rate": 9.387784632484825e-07,
+      "loss": 0.3155,
+      "step": 3643
+    },
+    {
+      "epoch": 2.465493910690122,
+      "grad_norm": 1.1019308765739761,
+      "learning_rate": 9.364832992373501e-07,
+      "loss": 0.2953,
+      "step": 3644
+    },
+    {
+      "epoch": 2.46617050067659,
+      "grad_norm": 1.1288751678368858,
+      "learning_rate": 9.341906543875451e-07,
+      "loss": 0.3165,
+      "step": 3645
+    },
+    {
+      "epoch": 2.466847090663058,
+      "grad_norm": 1.160564559428215,
+      "learning_rate": 9.319005301203821e-07,
+      "loss": 0.322,
+      "step": 3646
+    },
+    {
+      "epoch": 2.4675236806495264,
+      "grad_norm": 1.1399809893424149,
+      "learning_rate": 9.296129278556155e-07,
+      "loss": 0.3139,
+      "step": 3647
+    },
+    {
+      "epoch": 2.4682002706359945,
+      "grad_norm": 1.15137782801232,
+      "learning_rate": 9.273278490114357e-07,
+      "loss": 0.3216,
+      "step": 3648
+    },
+    {
+      "epoch": 2.468876860622463,
+      "grad_norm": 1.1225234397219517,
+      "learning_rate": 9.250452950044702e-07,
+      "loss": 0.3059,
+      "step": 3649
+    },
+    {
+      "epoch": 2.469553450608931,
+      "grad_norm": 1.115070313746826,
+      "learning_rate": 9.227652672497761e-07,
+      "loss": 0.3012,
+      "step": 3650
+    },
+    {
+      "epoch": 2.470230040595399,
+      "grad_norm": 1.1311278690032023,
+      "learning_rate": 9.204877671608515e-07,
+      "loss": 0.3099,
+      "step": 3651
+    },
+    {
+      "epoch": 2.4709066305818674,
+      "grad_norm": 1.121211882530858,
+      "learning_rate": 9.182127961496196e-07,
+      "loss": 0.3152,
+      "step": 3652
+    },
+    {
+      "epoch": 2.4715832205683355,
+      "grad_norm": 1.2016290053178065,
+      "learning_rate": 9.159403556264435e-07,
+      "loss": 0.324,
+      "step": 3653
+    },
+    {
+      "epoch": 2.472259810554804,
+      "grad_norm": 1.1272716084337469,
+      "learning_rate": 9.136704470001101e-07,
+      "loss": 0.3166,
+      "step": 3654
+    },
+    {
+      "epoch": 2.472936400541272,
+      "grad_norm": 1.0773651646535676,
+      "learning_rate": 9.114030716778433e-07,
+      "loss": 0.2912,
+      "step": 3655
+    },
+    {
+      "epoch": 2.47361299052774,
+      "grad_norm": 1.112831014496653,
+      "learning_rate": 9.091382310652925e-07,
+      "loss": 0.3121,
+      "step": 3656
+    },
+    {
+      "epoch": 2.4742895805142084,
+      "grad_norm": 1.1228219864597098,
+      "learning_rate": 9.068759265665384e-07,
+      "loss": 0.3115,
+      "step": 3657
+    },
+    {
+      "epoch": 2.4749661705006765,
+      "grad_norm": 1.1035873209666962,
+      "learning_rate": 9.046161595840858e-07,
+      "loss": 0.2965,
+      "step": 3658
+    },
+    {
+      "epoch": 2.475642760487145,
+      "grad_norm": 1.1179364137717536,
+      "learning_rate": 9.023589315188686e-07,
+      "loss": 0.3162,
+      "step": 3659
+    },
+    {
+      "epoch": 2.476319350473613,
+      "grad_norm": 1.1285034156113722,
+      "learning_rate": 9.001042437702468e-07,
+      "loss": 0.3164,
+      "step": 3660
+    },
+    {
+      "epoch": 2.476995940460081,
+      "grad_norm": 1.1429412179668048,
+      "learning_rate": 8.978520977360067e-07,
+      "loss": 0.3171,
+      "step": 3661
+    },
+    {
+      "epoch": 2.4776725304465494,
+      "grad_norm": 1.1121955666599372,
+      "learning_rate": 8.956024948123549e-07,
+      "loss": 0.3098,
+      "step": 3662
+    },
+    {
+      "epoch": 2.4783491204330175,
+      "grad_norm": 1.1697239453051984,
+      "learning_rate": 8.933554363939256e-07,
+      "loss": 0.3181,
+      "step": 3663
+    },
+    {
+      "epoch": 2.479025710419486,
+      "grad_norm": 1.1780879883606745,
+      "learning_rate": 8.911109238737748e-07,
+      "loss": 0.321,
+      "step": 3664
+    },
+    {
+      "epoch": 2.479702300405954,
+      "grad_norm": 1.1493066822023357,
+      "learning_rate": 8.888689586433768e-07,
+      "loss": 0.3086,
+      "step": 3665
+    },
+    {
+      "epoch": 2.480378890392422,
+      "grad_norm": 1.1306115430782535,
+      "learning_rate": 8.866295420926319e-07,
+      "loss": 0.3134,
+      "step": 3666
+    },
+    {
+      "epoch": 2.4810554803788905,
+      "grad_norm": 1.0908449161442337,
+      "learning_rate": 8.843926756098548e-07,
+      "loss": 0.3121,
+      "step": 3667
+    },
+    {
+      "epoch": 2.4817320703653585,
+      "grad_norm": 1.123226274940738,
+      "learning_rate": 8.821583605817835e-07,
+      "loss": 0.3117,
+      "step": 3668
+    },
+    {
+      "epoch": 2.482408660351827,
+      "grad_norm": 1.090837760871994,
+      "learning_rate": 8.799265983935734e-07,
+      "loss": 0.2957,
+      "step": 3669
+    },
+    {
+      "epoch": 2.483085250338295,
+      "grad_norm": 1.1591596583877377,
+      "learning_rate": 8.776973904287972e-07,
+      "loss": 0.3272,
+      "step": 3670
+    },
+    {
+      "epoch": 2.483761840324763,
+      "grad_norm": 1.1741658351855955,
+      "learning_rate": 8.754707380694427e-07,
+      "loss": 0.3157,
+      "step": 3671
+    },
+    {
+      "epoch": 2.4844384303112315,
+      "grad_norm": 1.123503135233597,
+      "learning_rate": 8.732466426959135e-07,
+      "loss": 0.3207,
+      "step": 3672
+    },
+    {
+      "epoch": 2.4851150202976995,
+      "grad_norm": 1.1461377510712878,
+      "learning_rate": 8.7102510568703e-07,
+      "loss": 0.3206,
+      "step": 3673
+    },
+    {
+      "epoch": 2.485791610284168,
+      "grad_norm": 1.1353827801624172,
+      "learning_rate": 8.688061284200266e-07,
+      "loss": 0.3147,
+      "step": 3674
+    },
+    {
+      "epoch": 2.486468200270636,
+      "grad_norm": 1.1246975065115774,
+      "learning_rate": 8.665897122705463e-07,
+      "loss": 0.3121,
+      "step": 3675
+    },
+    {
+      "epoch": 2.487144790257104,
+      "grad_norm": 1.1021770453979274,
+      "learning_rate": 8.6437585861265e-07,
+      "loss": 0.3085,
+      "step": 3676
+    },
+    {
+      "epoch": 2.4878213802435725,
+      "grad_norm": 1.1133565559785967,
+      "learning_rate": 8.621645688188085e-07,
+      "loss": 0.3137,
+      "step": 3677
+    },
+    {
+      "epoch": 2.4884979702300405,
+      "grad_norm": 1.1513927941995301,
+      "learning_rate": 8.599558442598998e-07,
+      "loss": 0.3162,
+      "step": 3678
+    },
+    {
+      "epoch": 2.489174560216509,
+      "grad_norm": 1.130175559035253,
+      "learning_rate": 8.577496863052165e-07,
+      "loss": 0.3101,
+      "step": 3679
+    },
+    {
+      "epoch": 2.489851150202977,
+      "grad_norm": 1.1309432653170697,
+      "learning_rate": 8.555460963224549e-07,
+      "loss": 0.3114,
+      "step": 3680
+    },
+    {
+      "epoch": 2.490527740189445,
+      "grad_norm": 1.1297916506726353,
+      "learning_rate": 8.53345075677724e-07,
+      "loss": 0.3112,
+      "step": 3681
+    },
+    {
+      "epoch": 2.4912043301759135,
+      "grad_norm": 1.1036341209194485,
+      "learning_rate": 8.511466257355384e-07,
+      "loss": 0.3048,
+      "step": 3682
+    },
+    {
+      "epoch": 2.4918809201623815,
+      "grad_norm": 1.1431141396707682,
+      "learning_rate": 8.48950747858816e-07,
+      "loss": 0.3288,
+      "step": 3683
+    },
+    {
+      "epoch": 2.49255751014885,
+      "grad_norm": 1.1329866537932307,
+      "learning_rate": 8.46757443408886e-07,
+      "loss": 0.31,
+      "step": 3684
+    },
+    {
+      "epoch": 2.493234100135318,
+      "grad_norm": 1.1080750786849676,
+      "learning_rate": 8.44566713745476e-07,
+      "loss": 0.3125,
+      "step": 3685
+    },
+    {
+      "epoch": 2.493910690121786,
+      "grad_norm": 1.1099112329245437,
+      "learning_rate": 8.42378560226722e-07,
+      "loss": 0.3004,
+      "step": 3686
+    },
+    {
+      "epoch": 2.4945872801082545,
+      "grad_norm": 1.1417708599820553,
+      "learning_rate": 8.401929842091616e-07,
+      "loss": 0.3103,
+      "step": 3687
+    },
+    {
+      "epoch": 2.4952638700947225,
+      "grad_norm": 1.1374713397857121,
+      "learning_rate": 8.380099870477321e-07,
+      "loss": 0.3107,
+      "step": 3688
+    },
+    {
+      "epoch": 2.495940460081191,
+      "grad_norm": 1.1467448136950438,
+      "learning_rate": 8.358295700957753e-07,
+      "loss": 0.3186,
+      "step": 3689
+    },
+    {
+      "epoch": 2.496617050067659,
+      "grad_norm": 1.1531318849857632,
+      "learning_rate": 8.336517347050327e-07,
+      "loss": 0.3219,
+      "step": 3690
+    },
+    {
+      "epoch": 2.497293640054127,
+      "grad_norm": 1.19361679628736,
+      "learning_rate": 8.314764822256465e-07,
+      "loss": 0.3366,
+      "step": 3691
+    },
+    {
+      "epoch": 2.4979702300405955,
+      "grad_norm": 1.1974777341904552,
+      "learning_rate": 8.293038140061516e-07,
+      "loss": 0.332,
+      "step": 3692
+    },
+    {
+      "epoch": 2.4986468200270635,
+      "grad_norm": 1.1244192531633166,
+      "learning_rate": 8.271337313934869e-07,
+      "loss": 0.3193,
+      "step": 3693
+    },
+    {
+      "epoch": 2.499323410013532,
+      "grad_norm": 1.131779643859016,
+      "learning_rate": 8.24966235732988e-07,
+      "loss": 0.3077,
+      "step": 3694
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.1221170366058266,
+      "learning_rate": 8.22801328368385e-07,
+      "loss": 0.3043,
+      "step": 3695
+    },
+    {
+      "epoch": 2.500676589986468,
+      "grad_norm": 1.1434580711747846,
+      "learning_rate": 8.206390106418028e-07,
+      "loss": 0.3093,
+      "step": 3696
+    },
+    {
+      "epoch": 2.5013531799729365,
+      "grad_norm": 1.1769857800034191,
+      "learning_rate": 8.184792838937633e-07,
+      "loss": 0.3234,
+      "step": 3697
+    },
+    {
+      "epoch": 2.5020297699594045,
+      "grad_norm": 1.1173881343935332,
+      "learning_rate": 8.163221494631785e-07,
+      "loss": 0.2997,
+      "step": 3698
+    },
+    {
+      "epoch": 2.502706359945873,
+      "grad_norm": 1.1271209057768203,
+      "learning_rate": 8.141676086873574e-07,
+      "loss": 0.3104,
+      "step": 3699
+    },
+    {
+      "epoch": 2.503382949932341,
+      "grad_norm": 1.08754106551674,
+      "learning_rate": 8.120156629019987e-07,
+      "loss": 0.3039,
+      "step": 3700
+    },
+    {
+      "epoch": 2.504059539918809,
+      "grad_norm": 1.1513866419083132,
+      "learning_rate": 8.098663134411922e-07,
+      "loss": 0.3116,
+      "step": 3701
+    },
+    {
+      "epoch": 2.5047361299052775,
+      "grad_norm": 1.13204408908435,
+      "learning_rate": 8.077195616374184e-07,
+      "loss": 0.3169,
+      "step": 3702
+    },
+    {
+      "epoch": 2.5054127198917455,
+      "grad_norm": 1.096123392467275,
+      "learning_rate": 8.055754088215501e-07,
+      "loss": 0.2988,
+      "step": 3703
+    },
+    {
+      "epoch": 2.506089309878214,
+      "grad_norm": 1.1118220265427365,
+      "learning_rate": 8.03433856322845e-07,
+      "loss": 0.2933,
+      "step": 3704
+    },
+    {
+      "epoch": 2.506765899864682,
+      "grad_norm": 1.174466968642932,
+      "learning_rate": 8.012949054689484e-07,
+      "loss": 0.3217,
+      "step": 3705
+    },
+    {
+      "epoch": 2.50744248985115,
+      "grad_norm": 1.1553983762354532,
+      "learning_rate": 7.991585575858962e-07,
+      "loss": 0.3216,
+      "step": 3706
+    },
+    {
+      "epoch": 2.5081190798376185,
+      "grad_norm": 1.1660414694117864,
+      "learning_rate": 7.970248139981091e-07,
+      "loss": 0.3265,
+      "step": 3707
+    },
+    {
+      "epoch": 2.5087956698240865,
+      "grad_norm": 1.1654863540447142,
+      "learning_rate": 7.948936760283937e-07,
+      "loss": 0.3184,
+      "step": 3708
+    },
+    {
+      "epoch": 2.509472259810555,
+      "grad_norm": 1.1519665292909043,
+      "learning_rate": 7.92765144997939e-07,
+      "loss": 0.3214,
+      "step": 3709
+    },
+    {
+      "epoch": 2.510148849797023,
+      "grad_norm": 1.163850272236243,
+      "learning_rate": 7.906392222263199e-07,
+      "loss": 0.3162,
+      "step": 3710
+    },
+    {
+      "epoch": 2.510825439783491,
+      "grad_norm": 1.111935055834014,
+      "learning_rate": 7.885159090314959e-07,
+      "loss": 0.3056,
+      "step": 3711
+    },
+    {
+      "epoch": 2.5115020297699595,
+      "grad_norm": 1.1567415229092517,
+      "learning_rate": 7.863952067298042e-07,
+      "loss": 0.324,
+      "step": 3712
+    },
+    {
+      "epoch": 2.5121786197564275,
+      "grad_norm": 1.1350957183916657,
+      "learning_rate": 7.842771166359681e-07,
+      "loss": 0.3152,
+      "step": 3713
+    },
+    {
+      "epoch": 2.512855209742896,
+      "grad_norm": 1.1029667819067668,
+      "learning_rate": 7.821616400630866e-07,
+      "loss": 0.3026,
+      "step": 3714
+    },
+    {
+      "epoch": 2.513531799729364,
+      "grad_norm": 1.120691799087682,
+      "learning_rate": 7.80048778322643e-07,
+      "loss": 0.3087,
+      "step": 3715
+    },
+    {
+      "epoch": 2.514208389715832,
+      "grad_norm": 1.1487156573969688,
+      "learning_rate": 7.779385327244987e-07,
+      "loss": 0.3145,
+      "step": 3716
+    },
+    {
+      "epoch": 2.5148849797023005,
+      "grad_norm": 1.1070717734849849,
+      "learning_rate": 7.758309045768908e-07,
+      "loss": 0.3038,
+      "step": 3717
+    },
+    {
+      "epoch": 2.5155615696887685,
+      "grad_norm": 1.1813271601173971,
+      "learning_rate": 7.737258951864341e-07,
+      "loss": 0.3165,
+      "step": 3718
+    },
+    {
+      "epoch": 2.516238159675237,
+      "grad_norm": 1.12974817586669,
+      "learning_rate": 7.716235058581218e-07,
+      "loss": 0.3149,
+      "step": 3719
+    },
+    {
+      "epoch": 2.516914749661705,
+      "grad_norm": 1.1863843608386213,
+      "learning_rate": 7.695237378953224e-07,
+      "loss": 0.3143,
+      "step": 3720
+    },
+    {
+      "epoch": 2.517591339648173,
+      "grad_norm": 1.1392727807156302,
+      "learning_rate": 7.674265925997804e-07,
+      "loss": 0.315,
+      "step": 3721
+    },
+    {
+      "epoch": 2.5182679296346415,
+      "grad_norm": 1.1324406042303652,
+      "learning_rate": 7.653320712716095e-07,
+      "loss": 0.3157,
+      "step": 3722
+    },
+    {
+      "epoch": 2.5189445196211095,
+      "grad_norm": 1.1233974281574273,
+      "learning_rate": 7.632401752093016e-07,
+      "loss": 0.3143,
+      "step": 3723
+    },
+    {
+      "epoch": 2.519621109607578,
+      "grad_norm": 1.144356827851246,
+      "learning_rate": 7.611509057097211e-07,
+      "loss": 0.3247,
+      "step": 3724
+    },
+    {
+      "epoch": 2.520297699594046,
+      "grad_norm": 1.151259634891893,
+      "learning_rate": 7.590642640681012e-07,
+      "loss": 0.3249,
+      "step": 3725
+    },
+    {
+      "epoch": 2.520974289580514,
+      "grad_norm": 1.127245847804875,
+      "learning_rate": 7.569802515780455e-07,
+      "loss": 0.3027,
+      "step": 3726
+    },
+    {
+      "epoch": 2.5216508795669825,
+      "grad_norm": 1.1119815832495008,
+      "learning_rate": 7.548988695315313e-07,
+      "loss": 0.311,
+      "step": 3727
+    },
+    {
+      "epoch": 2.5223274695534506,
+      "grad_norm": 1.2063885674474828,
+      "learning_rate": 7.528201192189028e-07,
+      "loss": 0.3194,
+      "step": 3728
+    },
+    {
+      "epoch": 2.523004059539919,
+      "grad_norm": 1.132910892314815,
+      "learning_rate": 7.507440019288742e-07,
+      "loss": 0.3125,
+      "step": 3729
+    },
+    {
+      "epoch": 2.523680649526387,
+      "grad_norm": 1.1723996182680465,
+      "learning_rate": 7.486705189485243e-07,
+      "loss": 0.3037,
+      "step": 3730
+    },
+    {
+      "epoch": 2.524357239512855,
+      "grad_norm": 1.0692107181283899,
+      "learning_rate": 7.465996715633028e-07,
+      "loss": 0.2866,
+      "step": 3731
+    },
+    {
+      "epoch": 2.5250338294993235,
+      "grad_norm": 1.1460005722338504,
+      "learning_rate": 7.44531461057022e-07,
+      "loss": 0.3004,
+      "step": 3732
+    },
+    {
+      "epoch": 2.5257104194857916,
+      "grad_norm": 1.1763829195906508,
+      "learning_rate": 7.424658887118613e-07,
+      "loss": 0.3217,
+      "step": 3733
+    },
+    {
+      "epoch": 2.52638700947226,
+      "grad_norm": 1.1178643648656326,
+      "learning_rate": 7.404029558083653e-07,
+      "loss": 0.2952,
+      "step": 3734
+    },
+    {
+      "epoch": 2.527063599458728,
+      "grad_norm": 1.1434491053925546,
+      "learning_rate": 7.383426636254392e-07,
+      "loss": 0.3143,
+      "step": 3735
+    },
+    {
+      "epoch": 2.527740189445196,
+      "grad_norm": 1.159002784993734,
+      "learning_rate": 7.362850134403543e-07,
+      "loss": 0.317,
+      "step": 3736
+    },
+    {
+      "epoch": 2.5284167794316645,
+      "grad_norm": 1.084385535923329,
+      "learning_rate": 7.342300065287439e-07,
+      "loss": 0.2959,
+      "step": 3737
+    },
+    {
+      "epoch": 2.5290933694181326,
+      "grad_norm": 1.1871165584955947,
+      "learning_rate": 7.321776441646001e-07,
+      "loss": 0.3242,
+      "step": 3738
+    },
+    {
+      "epoch": 2.529769959404601,
+      "grad_norm": 1.1493828961775037,
+      "learning_rate": 7.301279276202761e-07,
+      "loss": 0.3118,
+      "step": 3739
+    },
+    {
+      "epoch": 2.530446549391069,
+      "grad_norm": 1.1895350574165584,
+      "learning_rate": 7.280808581664866e-07,
+      "loss": 0.3244,
+      "step": 3740
+    },
+    {
+      "epoch": 2.531123139377537,
+      "grad_norm": 1.2015739901056524,
+      "learning_rate": 7.260364370723044e-07,
+      "loss": 0.3339,
+      "step": 3741
+    },
+    {
+      "epoch": 2.5317997293640055,
+      "grad_norm": 1.1760931721153907,
+      "learning_rate": 7.239946656051622e-07,
+      "loss": 0.3198,
+      "step": 3742
+    },
+    {
+      "epoch": 2.5324763193504736,
+      "grad_norm": 1.1699124745410474,
+      "learning_rate": 7.219555450308446e-07,
+      "loss": 0.3088,
+      "step": 3743
+    },
+    {
+      "epoch": 2.533152909336942,
+      "grad_norm": 1.2099178006365958,
+      "learning_rate": 7.199190766135001e-07,
+      "loss": 0.3436,
+      "step": 3744
+    },
+    {
+      "epoch": 2.53382949932341,
+      "grad_norm": 1.1029140570740334,
+      "learning_rate": 7.178852616156262e-07,
+      "loss": 0.2956,
+      "step": 3745
+    },
+    {
+      "epoch": 2.534506089309878,
+      "grad_norm": 1.143069202420577,
+      "learning_rate": 7.158541012980813e-07,
+      "loss": 0.3029,
+      "step": 3746
+    },
+    {
+      "epoch": 2.5351826792963466,
+      "grad_norm": 1.1596156867229297,
+      "learning_rate": 7.138255969200724e-07,
+      "loss": 0.3184,
+      "step": 3747
+    },
+    {
+      "epoch": 2.5358592692828146,
+      "grad_norm": 1.1502706362373234,
+      "learning_rate": 7.117997497391648e-07,
+      "loss": 0.3114,
+      "step": 3748
+    },
+    {
+      "epoch": 2.536535859269283,
+      "grad_norm": 1.1384305842959306,
+      "learning_rate": 7.097765610112745e-07,
+      "loss": 0.3166,
+      "step": 3749
+    },
+    {
+      "epoch": 2.537212449255751,
+      "grad_norm": 1.114087712937595,
+      "learning_rate": 7.077560319906696e-07,
+      "loss": 0.3088,
+      "step": 3750
+    },
+    {
+      "epoch": 2.537889039242219,
+      "grad_norm": 1.1327181809454117,
+      "learning_rate": 7.057381639299693e-07,
+      "loss": 0.3117,
+      "step": 3751
+    },
+    {
+      "epoch": 2.5385656292286876,
+      "grad_norm": 1.1488625539922845,
+      "learning_rate": 7.037229580801414e-07,
+      "loss": 0.3217,
+      "step": 3752
+    },
+    {
+      "epoch": 2.5392422192151556,
+      "grad_norm": 1.1475327189341715,
+      "learning_rate": 7.017104156905058e-07,
+      "loss": 0.3056,
+      "step": 3753
+    },
+    {
+      "epoch": 2.539918809201624,
+      "grad_norm": 1.0878976960789086,
+      "learning_rate": 6.997005380087301e-07,
+      "loss": 0.2933,
+      "step": 3754
+    },
+    {
+      "epoch": 2.540595399188092,
+      "grad_norm": 1.108388259536206,
+      "learning_rate": 6.976933262808322e-07,
+      "loss": 0.2988,
+      "step": 3755
+    },
+    {
+      "epoch": 2.54127198917456,
+      "grad_norm": 1.141026512880203,
+      "learning_rate": 6.95688781751172e-07,
+      "loss": 0.3227,
+      "step": 3756
+    },
+    {
+      "epoch": 2.5419485791610286,
+      "grad_norm": 1.1479001023892021,
+      "learning_rate": 6.936869056624623e-07,
+      "loss": 0.3138,
+      "step": 3757
+    },
+    {
+      "epoch": 2.5426251691474966,
+      "grad_norm": 1.104268764099118,
+      "learning_rate": 6.916876992557553e-07,
+      "loss": 0.3033,
+      "step": 3758
+    },
+    {
+      "epoch": 2.543301759133965,
+      "grad_norm": 1.164407872390624,
+      "learning_rate": 6.896911637704534e-07,
+      "loss": 0.3319,
+      "step": 3759
+    },
+    {
+      "epoch": 2.543978349120433,
+      "grad_norm": 1.1571743573343427,
+      "learning_rate": 6.876973004442988e-07,
+      "loss": 0.3165,
+      "step": 3760
+    },
+    {
+      "epoch": 2.544654939106901,
+      "grad_norm": 1.150170656754291,
+      "learning_rate": 6.85706110513381e-07,
+      "loss": 0.3249,
+      "step": 3761
+    },
+    {
+      "epoch": 2.5453315290933696,
+      "grad_norm": 1.1314854694700565,
+      "learning_rate": 6.837175952121305e-07,
+      "loss": 0.3062,
+      "step": 3762
+    },
+    {
+      "epoch": 2.5460081190798376,
+      "grad_norm": 1.0945707231098523,
+      "learning_rate": 6.8173175577332e-07,
+      "loss": 0.3095,
+      "step": 3763
+    },
+    {
+      "epoch": 2.546684709066306,
+      "grad_norm": 1.1289025306727016,
+      "learning_rate": 6.797485934280618e-07,
+      "loss": 0.3173,
+      "step": 3764
+    },
+    {
+      "epoch": 2.547361299052774,
+      "grad_norm": 1.1344543682752084,
+      "learning_rate": 6.777681094058087e-07,
+      "loss": 0.3152,
+      "step": 3765
+    },
+    {
+      "epoch": 2.548037889039242,
+      "grad_norm": 1.138694327919493,
+      "learning_rate": 6.757903049343556e-07,
+      "loss": 0.3144,
+      "step": 3766
+    },
+    {
+      "epoch": 2.5487144790257106,
+      "grad_norm": 1.1474526070049569,
+      "learning_rate": 6.738151812398353e-07,
+      "loss": 0.3206,
+      "step": 3767
+    },
+    {
+      "epoch": 2.5493910690121786,
+      "grad_norm": 1.1194031349989313,
+      "learning_rate": 6.718427395467165e-07,
+      "loss": 0.305,
+      "step": 3768
+    },
+    {
+      "epoch": 2.550067658998647,
+      "grad_norm": 1.1629177925372662,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.327,
+      "step": 3769
+    },
+    {
+      "epoch": 2.550744248985115,
+      "grad_norm": 1.1246626093657746,
+      "learning_rate": 6.67905907054251e-07,
+      "loss": 0.3126,
+      "step": 3770
+    },
+    {
+      "epoch": 2.551420838971583,
+      "grad_norm": 1.137151973605821,
+      "learning_rate": 6.659415186955298e-07,
+      "loss": 0.3084,
+      "step": 3771
+    },
+    {
+      "epoch": 2.5520974289580516,
+      "grad_norm": 1.1324360590500044,
+      "learning_rate": 6.639798172194567e-07,
+      "loss": 0.3076,
+      "step": 3772
+    },
+    {
+      "epoch": 2.5527740189445196,
+      "grad_norm": 1.1106642609606951,
+      "learning_rate": 6.620208038421805e-07,
+      "loss": 0.3174,
+      "step": 3773
+    },
+    {
+      "epoch": 2.553450608930988,
+      "grad_norm": 1.1533882904443011,
+      "learning_rate": 6.600644797781847e-07,
+      "loss": 0.3199,
+      "step": 3774
+    },
+    {
+      "epoch": 2.554127198917456,
+      "grad_norm": 1.1191526362128408,
+      "learning_rate": 6.581108462402847e-07,
+      "loss": 0.3048,
+      "step": 3775
+    },
+    {
+      "epoch": 2.554803788903924,
+      "grad_norm": 1.1285933784815176,
+      "learning_rate": 6.561599044396288e-07,
+      "loss": 0.325,
+      "step": 3776
+    },
+    {
+      "epoch": 2.555480378890392,
+      "grad_norm": 1.1398680877615874,
+      "learning_rate": 6.542116555856953e-07,
+      "loss": 0.3194,
+      "step": 3777
+    },
+    {
+      "epoch": 2.5561569688768606,
+      "grad_norm": 1.167366445221115,
+      "learning_rate": 6.522661008862918e-07,
+      "loss": 0.3151,
+      "step": 3778
+    },
+    {
+      "epoch": 2.556833558863329,
+      "grad_norm": 1.1125732379208815,
+      "learning_rate": 6.503232415475591e-07,
+      "loss": 0.2988,
+      "step": 3779
+    },
+    {
+      "epoch": 2.557510148849797,
+      "grad_norm": 1.1365819307875291,
+      "learning_rate": 6.483830787739659e-07,
+      "loss": 0.3076,
+      "step": 3780
+    },
+    {
+      "epoch": 2.558186738836265,
+      "grad_norm": 1.1350557624411308,
+      "learning_rate": 6.464456137683061e-07,
+      "loss": 0.3134,
+      "step": 3781
+    },
+    {
+      "epoch": 2.558863328822733,
+      "grad_norm": 1.1886131788738359,
+      "learning_rate": 6.445108477317046e-07,
+      "loss": 0.3217,
+      "step": 3782
+    },
+    {
+      "epoch": 2.5595399188092016,
+      "grad_norm": 1.1070398661169027,
+      "learning_rate": 6.425787818636131e-07,
+      "loss": 0.3029,
+      "step": 3783
+    },
+    {
+      "epoch": 2.56021650879567,
+      "grad_norm": 1.1265990379205022,
+      "learning_rate": 6.406494173618083e-07,
+      "loss": 0.3008,
+      "step": 3784
+    },
+    {
+      "epoch": 2.560893098782138,
+      "grad_norm": 1.1167239802165645,
+      "learning_rate": 6.387227554223918e-07,
+      "loss": 0.3034,
+      "step": 3785
+    },
+    {
+      "epoch": 2.561569688768606,
+      "grad_norm": 1.1593679924147031,
+      "learning_rate": 6.367987972397887e-07,
+      "loss": 0.3212,
+      "step": 3786
+    },
+    {
+      "epoch": 2.562246278755074,
+      "grad_norm": 1.146965141827802,
+      "learning_rate": 6.348775440067507e-07,
+      "loss": 0.3069,
+      "step": 3787
+    },
+    {
+      "epoch": 2.5629228687415426,
+      "grad_norm": 1.1656148548941394,
+      "learning_rate": 6.329589969143518e-07,
+      "loss": 0.3139,
+      "step": 3788
+    },
+    {
+      "epoch": 2.563599458728011,
+      "grad_norm": 1.1930517509647356,
+      "learning_rate": 6.310431571519865e-07,
+      "loss": 0.3369,
+      "step": 3789
+    },
+    {
+      "epoch": 2.564276048714479,
+      "grad_norm": 1.14838658629636,
+      "learning_rate": 6.291300259073724e-07,
+      "loss": 0.317,
+      "step": 3790
+    },
+    {
+      "epoch": 2.564952638700947,
+      "grad_norm": 1.1043942451679183,
+      "learning_rate": 6.27219604366549e-07,
+      "loss": 0.2992,
+      "step": 3791
+    },
+    {
+      "epoch": 2.565629228687415,
+      "grad_norm": 1.1846120369078166,
+      "learning_rate": 6.25311893713873e-07,
+      "loss": 0.3252,
+      "step": 3792
+    },
+    {
+      "epoch": 2.5663058186738836,
+      "grad_norm": 1.1718124573906015,
+      "learning_rate": 6.234068951320243e-07,
+      "loss": 0.3246,
+      "step": 3793
+    },
+    {
+      "epoch": 2.566982408660352,
+      "grad_norm": 1.1901228285723628,
+      "learning_rate": 6.215046098019967e-07,
+      "loss": 0.3274,
+      "step": 3794
+    },
+    {
+      "epoch": 2.56765899864682,
+      "grad_norm": 1.1265266987656684,
+      "learning_rate": 6.196050389031061e-07,
+      "loss": 0.3106,
+      "step": 3795
+    },
+    {
+      "epoch": 2.568335588633288,
+      "grad_norm": 1.0981062349044424,
+      "learning_rate": 6.177081836129833e-07,
+      "loss": 0.2974,
+      "step": 3796
+    },
+    {
+      "epoch": 2.569012178619756,
+      "grad_norm": 1.183624626867721,
+      "learning_rate": 6.158140451075794e-07,
+      "loss": 0.3178,
+      "step": 3797
+    },
+    {
+      "epoch": 2.5696887686062246,
+      "grad_norm": 1.1475745017616223,
+      "learning_rate": 6.139226245611535e-07,
+      "loss": 0.3075,
+      "step": 3798
+    },
+    {
+      "epoch": 2.5703653585926927,
+      "grad_norm": 1.130162558637046,
+      "learning_rate": 6.120339231462862e-07,
+      "loss": 0.3091,
+      "step": 3799
+    },
+    {
+      "epoch": 2.571041948579161,
+      "grad_norm": 1.1482852572092048,
+      "learning_rate": 6.101479420338713e-07,
+      "loss": 0.3118,
+      "step": 3800
+    },
+    {
+      "epoch": 2.571718538565629,
+      "grad_norm": 1.1741857009951864,
+      "learning_rate": 6.082646823931165e-07,
+      "loss": 0.3227,
+      "step": 3801
+    },
+    {
+      "epoch": 2.572395128552097,
+      "grad_norm": 1.16590564942958,
+      "learning_rate": 6.063841453915381e-07,
+      "loss": 0.3129,
+      "step": 3802
+    },
+    {
+      "epoch": 2.5730717185385656,
+      "grad_norm": 1.1183433012969533,
+      "learning_rate": 6.045063321949696e-07,
+      "loss": 0.3038,
+      "step": 3803
+    },
+    {
+      "epoch": 2.5737483085250337,
+      "grad_norm": 1.1945213716863263,
+      "learning_rate": 6.026312439675553e-07,
+      "loss": 0.3264,
+      "step": 3804
+    },
+    {
+      "epoch": 2.574424898511502,
+      "grad_norm": 1.1543056646963448,
+      "learning_rate": 6.007588818717458e-07,
+      "loss": 0.3112,
+      "step": 3805
+    },
+    {
+      "epoch": 2.57510148849797,
+      "grad_norm": 1.1974830719655707,
+      "learning_rate": 5.988892470683072e-07,
+      "loss": 0.3165,
+      "step": 3806
+    },
+    {
+      "epoch": 2.575778078484438,
+      "grad_norm": 1.1449929264797616,
+      "learning_rate": 5.9702234071631e-07,
+      "loss": 0.3054,
+      "step": 3807
+    },
+    {
+      "epoch": 2.5764546684709067,
+      "grad_norm": 1.133367760593302,
+      "learning_rate": 5.951581639731374e-07,
+      "loss": 0.3147,
+      "step": 3808
+    },
+    {
+      "epoch": 2.5771312584573747,
+      "grad_norm": 1.2114262871091466,
+      "learning_rate": 5.932967179944788e-07,
+      "loss": 0.3255,
+      "step": 3809
+    },
+    {
+      "epoch": 2.577807848443843,
+      "grad_norm": 1.1473435560399297,
+      "learning_rate": 5.914380039343281e-07,
+      "loss": 0.3079,
+      "step": 3810
+    },
+    {
+      "epoch": 2.578484438430311,
+      "grad_norm": 1.1333252360860042,
+      "learning_rate": 5.895820229449906e-07,
+      "loss": 0.3134,
+      "step": 3811
+    },
+    {
+      "epoch": 2.579161028416779,
+      "grad_norm": 1.230923209536186,
+      "learning_rate": 5.877287761770717e-07,
+      "loss": 0.3205,
+      "step": 3812
+    },
+    {
+      "epoch": 2.5798376184032477,
+      "grad_norm": 1.1113958757274316,
+      "learning_rate": 5.858782647794864e-07,
+      "loss": 0.3012,
+      "step": 3813
+    },
+    {
+      "epoch": 2.5805142083897157,
+      "grad_norm": 1.1042971238464216,
+      "learning_rate": 5.84030489899452e-07,
+      "loss": 0.299,
+      "step": 3814
+    },
+    {
+      "epoch": 2.581190798376184,
+      "grad_norm": 1.1580458278579813,
+      "learning_rate": 5.821854526824883e-07,
+      "loss": 0.3193,
+      "step": 3815
+    },
+    {
+      "epoch": 2.581867388362652,
+      "grad_norm": 1.134781982194373,
+      "learning_rate": 5.803431542724192e-07,
+      "loss": 0.3042,
+      "step": 3816
+    },
+    {
+      "epoch": 2.58254397834912,
+      "grad_norm": 1.1392046120583328,
+      "learning_rate": 5.785035958113717e-07,
+      "loss": 0.3144,
+      "step": 3817
+    },
+    {
+      "epoch": 2.5832205683355887,
+      "grad_norm": 1.1381504881565074,
+      "learning_rate": 5.766667784397706e-07,
+      "loss": 0.3183,
+      "step": 3818
+    },
+    {
+      "epoch": 2.5838971583220567,
+      "grad_norm": 1.1643788710651002,
+      "learning_rate": 5.748327032963464e-07,
+      "loss": 0.3222,
+      "step": 3819
+    },
+    {
+      "epoch": 2.584573748308525,
+      "grad_norm": 1.1390705272497892,
+      "learning_rate": 5.730013715181238e-07,
+      "loss": 0.317,
+      "step": 3820
+    },
+    {
+      "epoch": 2.585250338294993,
+      "grad_norm": 1.1334983243193666,
+      "learning_rate": 5.711727842404319e-07,
+      "loss": 0.3083,
+      "step": 3821
+    },
+    {
+      "epoch": 2.585926928281461,
+      "grad_norm": 1.1663671443675592,
+      "learning_rate": 5.693469425968962e-07,
+      "loss": 0.3298,
+      "step": 3822
+    },
+    {
+      "epoch": 2.5866035182679297,
+      "grad_norm": 1.1251060279128053,
+      "learning_rate": 5.675238477194389e-07,
+      "loss": 0.3068,
+      "step": 3823
+    },
+    {
+      "epoch": 2.5872801082543977,
+      "grad_norm": 1.1271727377453336,
+      "learning_rate": 5.657035007382822e-07,
+      "loss": 0.3118,
+      "step": 3824
+    },
+    {
+      "epoch": 2.587956698240866,
+      "grad_norm": 1.1354461790796073,
+      "learning_rate": 5.63885902781941e-07,
+      "loss": 0.3118,
+      "step": 3825
+    },
+    {
+      "epoch": 2.588633288227334,
+      "grad_norm": 1.1252400447339435,
+      "learning_rate": 5.620710549772295e-07,
+      "loss": 0.3097,
+      "step": 3826
+    },
+    {
+      "epoch": 2.589309878213802,
+      "grad_norm": 1.1291326986328731,
+      "learning_rate": 5.602589584492563e-07,
+      "loss": 0.3008,
+      "step": 3827
+    },
+    {
+      "epoch": 2.5899864682002707,
+      "grad_norm": 1.1347672955430506,
+      "learning_rate": 5.584496143214213e-07,
+      "loss": 0.3232,
+      "step": 3828
+    },
+    {
+      "epoch": 2.5906630581867387,
+      "grad_norm": 1.1218525800587003,
+      "learning_rate": 5.566430237154219e-07,
+      "loss": 0.3111,
+      "step": 3829
+    },
+    {
+      "epoch": 2.591339648173207,
+      "grad_norm": 1.1301765416375555,
+      "learning_rate": 5.548391877512471e-07,
+      "loss": 0.3009,
+      "step": 3830
+    },
+    {
+      "epoch": 2.592016238159675,
+      "grad_norm": 1.101603795212092,
+      "learning_rate": 5.530381075471775e-07,
+      "loss": 0.2977,
+      "step": 3831
+    },
+    {
+      "epoch": 2.592692828146143,
+      "grad_norm": 1.0931148059644717,
+      "learning_rate": 5.512397842197847e-07,
+      "loss": 0.3018,
+      "step": 3832
+    },
+    {
+      "epoch": 2.5933694181326117,
+      "grad_norm": 1.1639899007014718,
+      "learning_rate": 5.494442188839333e-07,
+      "loss": 0.3194,
+      "step": 3833
+    },
+    {
+      "epoch": 2.5940460081190797,
+      "grad_norm": 1.1339612664014729,
+      "learning_rate": 5.476514126527771e-07,
+      "loss": 0.3029,
+      "step": 3834
+    },
+    {
+      "epoch": 2.594722598105548,
+      "grad_norm": 1.068036944889056,
+      "learning_rate": 5.458613666377599e-07,
+      "loss": 0.2987,
+      "step": 3835
+    },
+    {
+      "epoch": 2.595399188092016,
+      "grad_norm": 1.1269405083440234,
+      "learning_rate": 5.440740819486123e-07,
+      "loss": 0.2916,
+      "step": 3836
+    },
+    {
+      "epoch": 2.596075778078484,
+      "grad_norm": 1.1489369451334235,
+      "learning_rate": 5.422895596933559e-07,
+      "loss": 0.3283,
+      "step": 3837
+    },
+    {
+      "epoch": 2.5967523680649527,
+      "grad_norm": 1.1596398422234107,
+      "learning_rate": 5.405078009782966e-07,
+      "loss": 0.3088,
+      "step": 3838
+    },
+    {
+      "epoch": 2.5974289580514207,
+      "grad_norm": 1.1886613346019737,
+      "learning_rate": 5.387288069080298e-07,
+      "loss": 0.3266,
+      "step": 3839
+    },
+    {
+      "epoch": 2.598105548037889,
+      "grad_norm": 1.1249537768973477,
+      "learning_rate": 5.369525785854368e-07,
+      "loss": 0.3092,
+      "step": 3840
+    },
+    {
+      "epoch": 2.598782138024357,
+      "grad_norm": 1.0910942396425627,
+      "learning_rate": 5.351791171116815e-07,
+      "loss": 0.2969,
+      "step": 3841
+    },
+    {
+      "epoch": 2.5994587280108252,
+      "grad_norm": 1.1308182243039568,
+      "learning_rate": 5.334084235862158e-07,
+      "loss": 0.2982,
+      "step": 3842
+    },
+    {
+      "epoch": 2.6001353179972937,
+      "grad_norm": 1.173235056178167,
+      "learning_rate": 5.316404991067747e-07,
+      "loss": 0.3276,
+      "step": 3843
+    },
+    {
+      "epoch": 2.6008119079837617,
+      "grad_norm": 1.1379719665613248,
+      "learning_rate": 5.29875344769375e-07,
+      "loss": 0.3106,
+      "step": 3844
+    },
+    {
+      "epoch": 2.60148849797023,
+      "grad_norm": 1.1492949648015884,
+      "learning_rate": 5.281129616683167e-07,
+      "loss": 0.3218,
+      "step": 3845
+    },
+    {
+      "epoch": 2.602165087956698,
+      "grad_norm": 1.183433053907029,
+      "learning_rate": 5.263533508961827e-07,
+      "loss": 0.3211,
+      "step": 3846
+    },
+    {
+      "epoch": 2.6028416779431662,
+      "grad_norm": 1.1355125151963343,
+      "learning_rate": 5.24596513543838e-07,
+      "loss": 0.3102,
+      "step": 3847
+    },
+    {
+      "epoch": 2.6035182679296347,
+      "grad_norm": 1.136279113327033,
+      "learning_rate": 5.228424507004265e-07,
+      "loss": 0.3139,
+      "step": 3848
+    },
+    {
+      "epoch": 2.6041948579161027,
+      "grad_norm": 1.117956201996482,
+      "learning_rate": 5.210911634533722e-07,
+      "loss": 0.2964,
+      "step": 3849
+    },
+    {
+      "epoch": 2.604871447902571,
+      "grad_norm": 1.1359805428603762,
+      "learning_rate": 5.193426528883788e-07,
+      "loss": 0.304,
+      "step": 3850
+    },
+    {
+      "epoch": 2.605548037889039,
+      "grad_norm": 1.146090712694019,
+      "learning_rate": 5.175969200894293e-07,
+      "loss": 0.3023,
+      "step": 3851
+    },
+    {
+      "epoch": 2.6062246278755072,
+      "grad_norm": 1.1620580556060454,
+      "learning_rate": 5.15853966138784e-07,
+      "loss": 0.3122,
+      "step": 3852
+    },
+    {
+      "epoch": 2.6069012178619757,
+      "grad_norm": 1.1298787014335747,
+      "learning_rate": 5.141137921169792e-07,
+      "loss": 0.3049,
+      "step": 3853
+    },
+    {
+      "epoch": 2.6075778078484437,
+      "grad_norm": 1.1399805341720364,
+      "learning_rate": 5.123763991028291e-07,
+      "loss": 0.3163,
+      "step": 3854
+    },
+    {
+      "epoch": 2.608254397834912,
+      "grad_norm": 1.1782093257327317,
+      "learning_rate": 5.106417881734244e-07,
+      "loss": 0.3225,
+      "step": 3855
+    },
+    {
+      "epoch": 2.60893098782138,
+      "grad_norm": 1.1557535236522072,
+      "learning_rate": 5.089099604041314e-07,
+      "loss": 0.3154,
+      "step": 3856
+    },
+    {
+      "epoch": 2.6096075778078482,
+      "grad_norm": 1.1067704143694668,
+      "learning_rate": 5.071809168685887e-07,
+      "loss": 0.3077,
+      "step": 3857
+    },
+    {
+      "epoch": 2.6102841677943167,
+      "grad_norm": 1.1208639217652279,
+      "learning_rate": 5.054546586387093e-07,
+      "loss": 0.3121,
+      "step": 3858
+    },
+    {
+      "epoch": 2.6109607577807847,
+      "grad_norm": 1.0972136590623127,
+      "learning_rate": 5.037311867846817e-07,
+      "loss": 0.2928,
+      "step": 3859
+    },
+    {
+      "epoch": 2.611637347767253,
+      "grad_norm": 1.1555705840082688,
+      "learning_rate": 5.020105023749644e-07,
+      "loss": 0.3045,
+      "step": 3860
+    },
+    {
+      "epoch": 2.6123139377537212,
+      "grad_norm": 1.1430388538618768,
+      "learning_rate": 5.002926064762908e-07,
+      "loss": 0.3076,
+      "step": 3861
+    },
+    {
+      "epoch": 2.6129905277401893,
+      "grad_norm": 1.1655175463541332,
+      "learning_rate": 4.985775001536619e-07,
+      "loss": 0.3173,
+      "step": 3862
+    },
+    {
+      "epoch": 2.6136671177266577,
+      "grad_norm": 1.1864998372109552,
+      "learning_rate": 4.968651844703514e-07,
+      "loss": 0.3313,
+      "step": 3863
+    },
+    {
+      "epoch": 2.6143437077131257,
+      "grad_norm": 1.160701416541282,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.3128,
+      "step": 3864
+    },
+    {
+      "epoch": 2.615020297699594,
+      "grad_norm": 1.1200346113836344,
+      "learning_rate": 4.934489292661326e-07,
+      "loss": 0.2989,
+      "step": 3865
+    },
+    {
+      "epoch": 2.6156968876860622,
+      "grad_norm": 1.1502479978378564,
+      "learning_rate": 4.917449918631162e-07,
+      "loss": 0.312,
+      "step": 3866
+    },
+    {
+      "epoch": 2.6163734776725303,
+      "grad_norm": 1.1481287995177303,
+      "learning_rate": 4.900438493352056e-07,
+      "loss": 0.3154,
+      "step": 3867
+    },
+    {
+      "epoch": 2.6170500676589987,
+      "grad_norm": 1.0868128739565612,
+      "learning_rate": 4.883455027370171e-07,
+      "loss": 0.3029,
+      "step": 3868
+    },
+    {
+      "epoch": 2.6177266576454667,
+      "grad_norm": 1.1093881770639489,
+      "learning_rate": 4.866499531214353e-07,
+      "loss": 0.3003,
+      "step": 3869
+    },
+    {
+      "epoch": 2.618403247631935,
+      "grad_norm": 1.1246919172934888,
+      "learning_rate": 4.849572015396081e-07,
+      "loss": 0.3161,
+      "step": 3870
+    },
+    {
+      "epoch": 2.6190798376184032,
+      "grad_norm": 1.0871026094196163,
+      "learning_rate": 4.832672490409513e-07,
+      "loss": 0.3095,
+      "step": 3871
+    },
+    {
+      "epoch": 2.6197564276048713,
+      "grad_norm": 1.152052591915853,
+      "learning_rate": 4.815800966731432e-07,
+      "loss": 0.3144,
+      "step": 3872
+    },
+    {
+      "epoch": 2.6204330175913397,
+      "grad_norm": 1.0746075049924178,
+      "learning_rate": 4.798957454821285e-07,
+      "loss": 0.2986,
+      "step": 3873
+    },
+    {
+      "epoch": 2.6211096075778078,
+      "grad_norm": 1.1605616867814241,
+      "learning_rate": 4.782141965121129e-07,
+      "loss": 0.3233,
+      "step": 3874
+    },
+    {
+      "epoch": 2.621786197564276,
+      "grad_norm": 1.1428749498555164,
+      "learning_rate": 4.7653545080556694e-07,
+      "loss": 0.3146,
+      "step": 3875
+    },
+    {
+      "epoch": 2.6224627875507442,
+      "grad_norm": 1.1420103428517616,
+      "learning_rate": 4.748595094032221e-07,
+      "loss": 0.3054,
+      "step": 3876
+    },
+    {
+      "epoch": 2.6231393775372123,
+      "grad_norm": 1.1193136035906281,
+      "learning_rate": 4.7318637334407335e-07,
+      "loss": 0.3024,
+      "step": 3877
+    },
+    {
+      "epoch": 2.6238159675236807,
+      "grad_norm": 1.1644387829831746,
+      "learning_rate": 4.715160436653732e-07,
+      "loss": 0.3276,
+      "step": 3878
+    },
+    {
+      "epoch": 2.6244925575101488,
+      "grad_norm": 1.1327022065160812,
+      "learning_rate": 4.698485214026349e-07,
+      "loss": 0.3087,
+      "step": 3879
+    },
+    {
+      "epoch": 2.6251691474966172,
+      "grad_norm": 1.1565690136087279,
+      "learning_rate": 4.6818380758963445e-07,
+      "loss": 0.3138,
+      "step": 3880
+    },
+    {
+      "epoch": 2.6258457374830853,
+      "grad_norm": 1.1361096516042588,
+      "learning_rate": 4.6652190325840396e-07,
+      "loss": 0.3013,
+      "step": 3881
+    },
+    {
+      "epoch": 2.6265223274695533,
+      "grad_norm": 1.0813213385441323,
+      "learning_rate": 4.6486280943923547e-07,
+      "loss": 0.3028,
+      "step": 3882
+    },
+    {
+      "epoch": 2.6271989174560217,
+      "grad_norm": 1.1901263997014517,
+      "learning_rate": 4.632065271606756e-07,
+      "loss": 0.3175,
+      "step": 3883
+    },
+    {
+      "epoch": 2.6278755074424898,
+      "grad_norm": 1.2003580096600932,
+      "learning_rate": 4.615530574495325e-07,
+      "loss": 0.3314,
+      "step": 3884
+    },
+    {
+      "epoch": 2.6285520974289582,
+      "grad_norm": 1.121270228678812,
+      "learning_rate": 4.5990240133086617e-07,
+      "loss": 0.3007,
+      "step": 3885
+    },
+    {
+      "epoch": 2.6292286874154263,
+      "grad_norm": 1.14675010293949,
+      "learning_rate": 4.582545598279964e-07,
+      "loss": 0.3079,
+      "step": 3886
+    },
+    {
+      "epoch": 2.6299052774018943,
+      "grad_norm": 1.162026647892871,
+      "learning_rate": 4.566095339624943e-07,
+      "loss": 0.3221,
+      "step": 3887
+    },
+    {
+      "epoch": 2.6305818673883627,
+      "grad_norm": 1.1839009765579254,
+      "learning_rate": 4.549673247541875e-07,
+      "loss": 0.3141,
+      "step": 3888
+    },
+    {
+      "epoch": 2.6312584573748308,
+      "grad_norm": 1.1185597241812426,
+      "learning_rate": 4.533279332211582e-07,
+      "loss": 0.3049,
+      "step": 3889
+    },
+    {
+      "epoch": 2.6319350473612992,
+      "grad_norm": 1.087754701605978,
+      "learning_rate": 4.516913603797407e-07,
+      "loss": 0.2988,
+      "step": 3890
+    },
+    {
+      "epoch": 2.6326116373477673,
+      "grad_norm": 1.138088046546554,
+      "learning_rate": 4.5005760724452173e-07,
+      "loss": 0.3194,
+      "step": 3891
+    },
+    {
+      "epoch": 2.6332882273342353,
+      "grad_norm": 1.1498375096555644,
+      "learning_rate": 4.484266748283389e-07,
+      "loss": 0.3211,
+      "step": 3892
+    },
+    {
+      "epoch": 2.6339648173207038,
+      "grad_norm": 1.161275366585453,
+      "learning_rate": 4.4679856414228394e-07,
+      "loss": 0.3146,
+      "step": 3893
+    },
+    {
+      "epoch": 2.634641407307172,
+      "grad_norm": 1.149623698273307,
+      "learning_rate": 4.4517327619569784e-07,
+      "loss": 0.3179,
+      "step": 3894
+    },
+    {
+      "epoch": 2.6353179972936402,
+      "grad_norm": 1.199573699525809,
+      "learning_rate": 4.435508119961701e-07,
+      "loss": 0.3232,
+      "step": 3895
+    },
+    {
+      "epoch": 2.6359945872801083,
+      "grad_norm": 1.1401399431685741,
+      "learning_rate": 4.4193117254954174e-07,
+      "loss": 0.3112,
+      "step": 3896
+    },
+    {
+      "epoch": 2.6366711772665763,
+      "grad_norm": 1.1588197843666872,
+      "learning_rate": 4.403143588599029e-07,
+      "loss": 0.3052,
+      "step": 3897
+    },
+    {
+      "epoch": 2.6373477672530448,
+      "grad_norm": 1.187777768827832,
+      "learning_rate": 4.387003719295896e-07,
+      "loss": 0.3195,
+      "step": 3898
+    },
+    {
+      "epoch": 2.638024357239513,
+      "grad_norm": 1.1078991330180872,
+      "learning_rate": 4.37089212759188e-07,
+      "loss": 0.3121,
+      "step": 3899
+    },
+    {
+      "epoch": 2.6387009472259813,
+      "grad_norm": 1.1890306336468357,
+      "learning_rate": 4.3548088234752814e-07,
+      "loss": 0.3202,
+      "step": 3900
+    },
+    {
+      "epoch": 2.6393775372124493,
+      "grad_norm": 1.1449029963949626,
+      "learning_rate": 4.3387538169168905e-07,
+      "loss": 0.3093,
+      "step": 3901
+    },
+    {
+      "epoch": 2.6400541271989173,
+      "grad_norm": 1.1606366928401124,
+      "learning_rate": 4.322727117869951e-07,
+      "loss": 0.3254,
+      "step": 3902
+    },
+    {
+      "epoch": 2.6407307171853858,
+      "grad_norm": 1.1356345185048866,
+      "learning_rate": 4.3067287362701606e-07,
+      "loss": 0.3151,
+      "step": 3903
+    },
+    {
+      "epoch": 2.641407307171854,
+      "grad_norm": 1.1381160631442553,
+      "learning_rate": 4.2907586820356337e-07,
+      "loss": 0.318,
+      "step": 3904
+    },
+    {
+      "epoch": 2.6420838971583223,
+      "grad_norm": 1.156409825974388,
+      "learning_rate": 4.2748169650669524e-07,
+      "loss": 0.3287,
+      "step": 3905
+    },
+    {
+      "epoch": 2.6427604871447903,
+      "grad_norm": 1.14925788312353,
+      "learning_rate": 4.258903595247116e-07,
+      "loss": 0.3123,
+      "step": 3906
+    },
+    {
+      "epoch": 2.6434370771312583,
+      "grad_norm": 1.1401115761497753,
+      "learning_rate": 4.2430185824415717e-07,
+      "loss": 0.3193,
+      "step": 3907
+    },
+    {
+      "epoch": 2.6441136671177268,
+      "grad_norm": 1.1116498482252897,
+      "learning_rate": 4.2271619364981474e-07,
+      "loss": 0.2982,
+      "step": 3908
+    },
+    {
+      "epoch": 2.644790257104195,
+      "grad_norm": 1.115715506840046,
+      "learning_rate": 4.211333667247125e-07,
+      "loss": 0.2977,
+      "step": 3909
+    },
+    {
+      "epoch": 2.6454668470906633,
+      "grad_norm": 1.0934925130360404,
+      "learning_rate": 4.195533784501177e-07,
+      "loss": 0.2971,
+      "step": 3910
+    },
+    {
+      "epoch": 2.6461434370771313,
+      "grad_norm": 1.1315352641263112,
+      "learning_rate": 4.179762298055384e-07,
+      "loss": 0.3123,
+      "step": 3911
+    },
+    {
+      "epoch": 2.6468200270635993,
+      "grad_norm": 1.1835451385787614,
+      "learning_rate": 4.164019217687215e-07,
+      "loss": 0.3256,
+      "step": 3912
+    },
+    {
+      "epoch": 2.647496617050068,
+      "grad_norm": 1.1684264719194448,
+      "learning_rate": 4.1483045531565183e-07,
+      "loss": 0.3154,
+      "step": 3913
+    },
+    {
+      "epoch": 2.648173207036536,
+      "grad_norm": 1.1456838214951601,
+      "learning_rate": 4.132618314205544e-07,
+      "loss": 0.3158,
+      "step": 3914
+    },
+    {
+      "epoch": 2.6488497970230043,
+      "grad_norm": 1.19382123557591,
+      "learning_rate": 4.1169605105589315e-07,
+      "loss": 0.3137,
+      "step": 3915
+    },
+    {
+      "epoch": 2.6495263870094723,
+      "grad_norm": 1.1475399291479902,
+      "learning_rate": 4.101331151923649e-07,
+      "loss": 0.3097,
+      "step": 3916
+    },
+    {
+      "epoch": 2.6502029769959403,
+      "grad_norm": 1.1420180852653452,
+      "learning_rate": 4.085730247989078e-07,
+      "loss": 0.3109,
+      "step": 3917
+    },
+    {
+      "epoch": 2.650879566982409,
+      "grad_norm": 1.0929143586978611,
+      "learning_rate": 4.070157808426928e-07,
+      "loss": 0.2943,
+      "step": 3918
+    },
+    {
+      "epoch": 2.651556156968877,
+      "grad_norm": 1.1198853637818,
+      "learning_rate": 4.0546138428912694e-07,
+      "loss": 0.3071,
+      "step": 3919
+    },
+    {
+      "epoch": 2.6522327469553453,
+      "grad_norm": 1.1754271856360132,
+      "learning_rate": 4.039098361018534e-07,
+      "loss": 0.3237,
+      "step": 3920
+    },
+    {
+      "epoch": 2.6529093369418133,
+      "grad_norm": 1.1545185588249969,
+      "learning_rate": 4.0236113724274716e-07,
+      "loss": 0.312,
+      "step": 3921
+    },
+    {
+      "epoch": 2.6535859269282813,
+      "grad_norm": 1.1252462422848621,
+      "learning_rate": 4.0081528867191854e-07,
+      "loss": 0.3031,
+      "step": 3922
+    },
+    {
+      "epoch": 2.65426251691475,
+      "grad_norm": 1.1312672578179293,
+      "learning_rate": 3.992722913477104e-07,
+      "loss": 0.3124,
+      "step": 3923
+    },
+    {
+      "epoch": 2.654939106901218,
+      "grad_norm": 1.1309409352072899,
+      "learning_rate": 3.9773214622669974e-07,
+      "loss": 0.3067,
+      "step": 3924
+    },
+    {
+      "epoch": 2.6556156968876863,
+      "grad_norm": 1.158834269778159,
+      "learning_rate": 3.9619485426369007e-07,
+      "loss": 0.3164,
+      "step": 3925
+    },
+    {
+      "epoch": 2.6562922868741543,
+      "grad_norm": 1.1211089267543728,
+      "learning_rate": 3.9466041641172126e-07,
+      "loss": 0.3017,
+      "step": 3926
+    },
+    {
+      "epoch": 2.6569688768606223,
+      "grad_norm": 1.1781346466914768,
+      "learning_rate": 3.9312883362206177e-07,
+      "loss": 0.3172,
+      "step": 3927
+    },
+    {
+      "epoch": 2.657645466847091,
+      "grad_norm": 1.1579686733853118,
+      "learning_rate": 3.916001068442116e-07,
+      "loss": 0.3061,
+      "step": 3928
+    },
+    {
+      "epoch": 2.658322056833559,
+      "grad_norm": 1.158083934820583,
+      "learning_rate": 3.90074237025897e-07,
+      "loss": 0.3276,
+      "step": 3929
+    },
+    {
+      "epoch": 2.6589986468200273,
+      "grad_norm": 1.2224644770566764,
+      "learning_rate": 3.885512251130763e-07,
+      "loss": 0.3301,
+      "step": 3930
+    },
+    {
+      "epoch": 2.6596752368064953,
+      "grad_norm": 1.154118992241007,
+      "learning_rate": 3.870310720499354e-07,
+      "loss": 0.3133,
+      "step": 3931
+    },
+    {
+      "epoch": 2.6603518267929633,
+      "grad_norm": 1.1285314214553759,
+      "learning_rate": 3.8551377877888487e-07,
+      "loss": 0.2973,
+      "step": 3932
+    },
+    {
+      "epoch": 2.661028416779432,
+      "grad_norm": 1.1350986384555009,
+      "learning_rate": 3.839993462405678e-07,
+      "loss": 0.3025,
+      "step": 3933
+    },
+    {
+      "epoch": 2.6617050067659,
+      "grad_norm": 1.1691896415739214,
+      "learning_rate": 3.8248777537384763e-07,
+      "loss": 0.3194,
+      "step": 3934
+    },
+    {
+      "epoch": 2.6623815967523683,
+      "grad_norm": 1.1226685200207052,
+      "learning_rate": 3.8097906711581864e-07,
+      "loss": 0.3002,
+      "step": 3935
+    },
+    {
+      "epoch": 2.6630581867388363,
+      "grad_norm": 1.0962648684689034,
+      "learning_rate": 3.794732224017994e-07,
+      "loss": 0.3012,
+      "step": 3936
+    },
+    {
+      "epoch": 2.6637347767253043,
+      "grad_norm": 1.1664936167721887,
+      "learning_rate": 3.7797024216533143e-07,
+      "loss": 0.3177,
+      "step": 3937
+    },
+    {
+      "epoch": 2.664411366711773,
+      "grad_norm": 1.1141677206329326,
+      "learning_rate": 3.764701273381799e-07,
+      "loss": 0.3,
+      "step": 3938
+    },
+    {
+      "epoch": 2.665087956698241,
+      "grad_norm": 1.1920885348497052,
+      "learning_rate": 3.7497287885033763e-07,
+      "loss": 0.3149,
+      "step": 3939
+    },
+    {
+      "epoch": 2.6657645466847093,
+      "grad_norm": 1.1332018548170577,
+      "learning_rate": 3.734784976300165e-07,
+      "loss": 0.3087,
+      "step": 3940
+    },
+    {
+      "epoch": 2.6664411366711773,
+      "grad_norm": 1.1736281956057233,
+      "learning_rate": 3.719869846036539e-07,
+      "loss": 0.3083,
+      "step": 3941
+    },
+    {
+      "epoch": 2.6671177266576453,
+      "grad_norm": 1.1631910018685614,
+      "learning_rate": 3.7049834069590507e-07,
+      "loss": 0.3161,
+      "step": 3942
+    },
+    {
+      "epoch": 2.667794316644114,
+      "grad_norm": 1.1391266383655867,
+      "learning_rate": 3.6901256682965123e-07,
+      "loss": 0.3114,
+      "step": 3943
+    },
+    {
+      "epoch": 2.668470906630582,
+      "grad_norm": 1.1383146039668341,
+      "learning_rate": 3.675296639259912e-07,
+      "loss": 0.3033,
+      "step": 3944
+    },
+    {
+      "epoch": 2.6691474966170503,
+      "grad_norm": 1.1379065654288862,
+      "learning_rate": 3.6604963290424453e-07,
+      "loss": 0.3279,
+      "step": 3945
+    },
+    {
+      "epoch": 2.6698240866035183,
+      "grad_norm": 1.112887933058892,
+      "learning_rate": 3.6457247468195233e-07,
+      "loss": 0.2973,
+      "step": 3946
+    },
+    {
+      "epoch": 2.6705006765899864,
+      "grad_norm": 1.0936236241153474,
+      "learning_rate": 3.6309819017487034e-07,
+      "loss": 0.2962,
+      "step": 3947
+    },
+    {
+      "epoch": 2.671177266576455,
+      "grad_norm": 1.1282324362379799,
+      "learning_rate": 3.6162678029697696e-07,
+      "loss": 0.3013,
+      "step": 3948
+    },
+    {
+      "epoch": 2.671853856562923,
+      "grad_norm": 1.1428927851246673,
+      "learning_rate": 3.60158245960468e-07,
+      "loss": 0.3158,
+      "step": 3949
+    },
+    {
+      "epoch": 2.6725304465493913,
+      "grad_norm": 1.0813084587548822,
+      "learning_rate": 3.5869258807575414e-07,
+      "loss": 0.295,
+      "step": 3950
+    },
+    {
+      "epoch": 2.6732070365358593,
+      "grad_norm": 1.166242524004221,
+      "learning_rate": 3.572298075514652e-07,
+      "loss": 0.305,
+      "step": 3951
+    },
+    {
+      "epoch": 2.6738836265223274,
+      "grad_norm": 1.1212011377351583,
+      "learning_rate": 3.557699052944447e-07,
+      "loss": 0.3134,
+      "step": 3952
+    },
+    {
+      "epoch": 2.674560216508796,
+      "grad_norm": 1.083384538000415,
+      "learning_rate": 3.5431288220975466e-07,
+      "loss": 0.2957,
+      "step": 3953
+    },
+    {
+      "epoch": 2.675236806495264,
+      "grad_norm": 1.180540152010333,
+      "learning_rate": 3.528587392006716e-07,
+      "loss": 0.3195,
+      "step": 3954
+    },
+    {
+      "epoch": 2.6759133964817323,
+      "grad_norm": 1.1121805932210829,
+      "learning_rate": 3.5140747716868375e-07,
+      "loss": 0.2995,
+      "step": 3955
+    },
+    {
+      "epoch": 2.6765899864682003,
+      "grad_norm": 1.1358071974123964,
+      "learning_rate": 3.499590970134964e-07,
+      "loss": 0.3166,
+      "step": 3956
+    },
+    {
+      "epoch": 2.6772665764546684,
+      "grad_norm": 1.146525898007564,
+      "learning_rate": 3.48513599633028e-07,
+      "loss": 0.3237,
+      "step": 3957
+    },
+    {
+      "epoch": 2.677943166441137,
+      "grad_norm": 1.154294420325088,
+      "learning_rate": 3.470709859234084e-07,
+      "loss": 0.3193,
+      "step": 3958
+    },
+    {
+      "epoch": 2.678619756427605,
+      "grad_norm": 1.1522985369452075,
+      "learning_rate": 3.4563125677897936e-07,
+      "loss": 0.3048,
+      "step": 3959
+    },
+    {
+      "epoch": 2.6792963464140733,
+      "grad_norm": 1.111198778038217,
+      "learning_rate": 3.4419441309229587e-07,
+      "loss": 0.306,
+      "step": 3960
+    },
+    {
+      "epoch": 2.6799729364005414,
+      "grad_norm": 1.1111159654198905,
+      "learning_rate": 3.427604557541242e-07,
+      "loss": 0.3118,
+      "step": 3961
+    },
+    {
+      "epoch": 2.6806495263870094,
+      "grad_norm": 1.130495969079568,
+      "learning_rate": 3.4132938565344054e-07,
+      "loss": 0.3022,
+      "step": 3962
+    },
+    {
+      "epoch": 2.6813261163734774,
+      "grad_norm": 1.1229295474030616,
+      "learning_rate": 3.3990120367743074e-07,
+      "loss": 0.2997,
+      "step": 3963
+    },
+    {
+      "epoch": 2.682002706359946,
+      "grad_norm": 1.1437934061736013,
+      "learning_rate": 3.38475910711491e-07,
+      "loss": 0.3105,
+      "step": 3964
+    },
+    {
+      "epoch": 2.6826792963464143,
+      "grad_norm": 1.1404437199848547,
+      "learning_rate": 3.370535076392256e-07,
+      "loss": 0.3138,
+      "step": 3965
+    },
+    {
+      "epoch": 2.6833558863328824,
+      "grad_norm": 1.184558328404599,
+      "learning_rate": 3.356339953424481e-07,
+      "loss": 0.3141,
+      "step": 3966
+    },
+    {
+      "epoch": 2.6840324763193504,
+      "grad_norm": 1.1706284914906093,
+      "learning_rate": 3.342173747011801e-07,
+      "loss": 0.3273,
+      "step": 3967
+    },
+    {
+      "epoch": 2.6847090663058184,
+      "grad_norm": 1.134593138015031,
+      "learning_rate": 3.3280364659364903e-07,
+      "loss": 0.3031,
+      "step": 3968
+    },
+    {
+      "epoch": 2.685385656292287,
+      "grad_norm": 1.1892073800544944,
+      "learning_rate": 3.313928118962906e-07,
+      "loss": 0.3234,
+      "step": 3969
+    },
+    {
+      "epoch": 2.6860622462787553,
+      "grad_norm": 1.1470210857615581,
+      "learning_rate": 3.299848714837473e-07,
+      "loss": 0.3112,
+      "step": 3970
+    },
+    {
+      "epoch": 2.6867388362652234,
+      "grad_norm": 1.1137745498498943,
+      "learning_rate": 3.285798262288653e-07,
+      "loss": 0.3043,
+      "step": 3971
+    },
+    {
+      "epoch": 2.6874154262516914,
+      "grad_norm": 1.1740980544937267,
+      "learning_rate": 3.271776770026963e-07,
+      "loss": 0.3137,
+      "step": 3972
+    },
+    {
+      "epoch": 2.6880920162381594,
+      "grad_norm": 1.1455006056831107,
+      "learning_rate": 3.2577842467449773e-07,
+      "loss": 0.3092,
+      "step": 3973
+    },
+    {
+      "epoch": 2.688768606224628,
+      "grad_norm": 1.149254969088589,
+      "learning_rate": 3.243820701117306e-07,
+      "loss": 0.3188,
+      "step": 3974
+    },
+    {
+      "epoch": 2.6894451962110963,
+      "grad_norm": 1.1212664459809358,
+      "learning_rate": 3.229886141800609e-07,
+      "loss": 0.3115,
+      "step": 3975
+    },
+    {
+      "epoch": 2.6901217861975644,
+      "grad_norm": 1.0991914124526325,
+      "learning_rate": 3.2159805774335364e-07,
+      "loss": 0.3069,
+      "step": 3976
+    },
+    {
+      "epoch": 2.6907983761840324,
+      "grad_norm": 1.1249068903027997,
+      "learning_rate": 3.2021040166368145e-07,
+      "loss": 0.3051,
+      "step": 3977
+    },
+    {
+      "epoch": 2.6914749661705004,
+      "grad_norm": 1.1283914367979384,
+      "learning_rate": 3.18825646801314e-07,
+      "loss": 0.3121,
+      "step": 3978
+    },
+    {
+      "epoch": 2.692151556156969,
+      "grad_norm": 1.118970879435845,
+      "learning_rate": 3.174437940147268e-07,
+      "loss": 0.3076,
+      "step": 3979
+    },
+    {
+      "epoch": 2.6928281461434374,
+      "grad_norm": 1.1018902269904114,
+      "learning_rate": 3.160648441605918e-07,
+      "loss": 0.299,
+      "step": 3980
+    },
+    {
+      "epoch": 2.6935047361299054,
+      "grad_norm": 1.1649370015384704,
+      "learning_rate": 3.146887980937852e-07,
+      "loss": 0.3148,
+      "step": 3981
+    },
+    {
+      "epoch": 2.6941813261163734,
+      "grad_norm": 1.1177723940804327,
+      "learning_rate": 3.133156566673806e-07,
+      "loss": 0.3054,
+      "step": 3982
+    },
+    {
+      "epoch": 2.6948579161028414,
+      "grad_norm": 1.1573227601445006,
+      "learning_rate": 3.119454207326533e-07,
+      "loss": 0.316,
+      "step": 3983
+    },
+    {
+      "epoch": 2.69553450608931,
+      "grad_norm": 1.1521274288041907,
+      "learning_rate": 3.105780911390738e-07,
+      "loss": 0.314,
+      "step": 3984
+    },
+    {
+      "epoch": 2.696211096075778,
+      "grad_norm": 1.1504808865411964,
+      "learning_rate": 3.0921366873431337e-07,
+      "loss": 0.3058,
+      "step": 3985
+    },
+    {
+      "epoch": 2.6968876860622464,
+      "grad_norm": 1.142037633162387,
+      "learning_rate": 3.0785215436423986e-07,
+      "loss": 0.3138,
+      "step": 3986
+    },
+    {
+      "epoch": 2.6975642760487144,
+      "grad_norm": 1.1975045807010414,
+      "learning_rate": 3.0649354887291927e-07,
+      "loss": 0.3181,
+      "step": 3987
+    },
+    {
+      "epoch": 2.6982408660351824,
+      "grad_norm": 1.1534965573568399,
+      "learning_rate": 3.05137853102615e-07,
+      "loss": 0.3243,
+      "step": 3988
+    },
+    {
+      "epoch": 2.698917456021651,
+      "grad_norm": 1.1211567734205146,
+      "learning_rate": 3.037850678937831e-07,
+      "loss": 0.3135,
+      "step": 3989
+    },
+    {
+      "epoch": 2.699594046008119,
+      "grad_norm": 1.1120909076464152,
+      "learning_rate": 3.0243519408507894e-07,
+      "loss": 0.2967,
+      "step": 3990
+    },
+    {
+      "epoch": 2.7002706359945874,
+      "grad_norm": 1.1371430477916977,
+      "learning_rate": 3.0108823251335183e-07,
+      "loss": 0.3094,
+      "step": 3991
+    },
+    {
+      "epoch": 2.7009472259810554,
+      "grad_norm": 1.1738354999672715,
+      "learning_rate": 2.997441840136445e-07,
+      "loss": 0.3186,
+      "step": 3992
+    },
+    {
+      "epoch": 2.7016238159675234,
+      "grad_norm": 1.0896630238956622,
+      "learning_rate": 2.984030494191942e-07,
+      "loss": 0.2936,
+      "step": 3993
+    },
+    {
+      "epoch": 2.702300405953992,
+      "grad_norm": 1.1838252999459855,
+      "learning_rate": 2.97064829561432e-07,
+      "loss": 0.3123,
+      "step": 3994
+    },
+    {
+      "epoch": 2.70297699594046,
+      "grad_norm": 1.1357917165707252,
+      "learning_rate": 2.957295252699832e-07,
+      "loss": 0.31,
+      "step": 3995
+    },
+    {
+      "epoch": 2.7036535859269284,
+      "grad_norm": 1.1507316397604836,
+      "learning_rate": 2.9439713737266504e-07,
+      "loss": 0.3156,
+      "step": 3996
+    },
+    {
+      "epoch": 2.7043301759133964,
+      "grad_norm": 1.1306261457857945,
+      "learning_rate": 2.930676666954846e-07,
+      "loss": 0.2991,
+      "step": 3997
+    },
+    {
+      "epoch": 2.7050067658998644,
+      "grad_norm": 1.1379277221353257,
+      "learning_rate": 2.917411140626425e-07,
+      "loss": 0.314,
+      "step": 3998
+    },
+    {
+      "epoch": 2.705683355886333,
+      "grad_norm": 1.116247752238892,
+      "learning_rate": 2.904174802965293e-07,
+      "loss": 0.2937,
+      "step": 3999
+    },
+    {
+      "epoch": 2.706359945872801,
+      "grad_norm": 1.1468361273310435,
+      "learning_rate": 2.8909676621772853e-07,
+      "loss": 0.3152,
+      "step": 4000
+    },
+    {
+      "epoch": 2.7070365358592694,
+      "grad_norm": 1.139665575219445,
+      "learning_rate": 2.877789726450092e-07,
+      "loss": 0.3176,
+      "step": 4001
+    },
+    {
+      "epoch": 2.7077131258457374,
+      "grad_norm": 1.166998089614663,
+      "learning_rate": 2.864641003953339e-07,
+      "loss": 0.3132,
+      "step": 4002
+    },
+    {
+      "epoch": 2.7083897158322054,
+      "grad_norm": 1.1376079208582652,
+      "learning_rate": 2.8515215028385223e-07,
+      "loss": 0.3099,
+      "step": 4003
+    },
+    {
+      "epoch": 2.709066305818674,
+      "grad_norm": 1.1625053920853172,
+      "learning_rate": 2.8384312312390306e-07,
+      "loss": 0.3088,
+      "step": 4004
+    },
+    {
+      "epoch": 2.709742895805142,
+      "grad_norm": 1.1248582279451893,
+      "learning_rate": 2.8253701972701275e-07,
+      "loss": 0.3066,
+      "step": 4005
+    },
+    {
+      "epoch": 2.7104194857916104,
+      "grad_norm": 1.1216692215532875,
+      "learning_rate": 2.8123384090289307e-07,
+      "loss": 0.2862,
+      "step": 4006
+    },
+    {
+      "epoch": 2.7110960757780784,
+      "grad_norm": 1.1556087009946527,
+      "learning_rate": 2.799335874594461e-07,
+      "loss": 0.3045,
+      "step": 4007
+    },
+    {
+      "epoch": 2.7117726657645465,
+      "grad_norm": 1.1589825412941182,
+      "learning_rate": 2.7863626020275867e-07,
+      "loss": 0.3166,
+      "step": 4008
+    },
+    {
+      "epoch": 2.712449255751015,
+      "grad_norm": 1.0787311167365927,
+      "learning_rate": 2.773418599371047e-07,
+      "loss": 0.2921,
+      "step": 4009
+    },
+    {
+      "epoch": 2.713125845737483,
+      "grad_norm": 1.1914457524471502,
+      "learning_rate": 2.7605038746494063e-07,
+      "loss": 0.3152,
+      "step": 4010
+    },
+    {
+      "epoch": 2.7138024357239514,
+      "grad_norm": 1.1592472178730937,
+      "learning_rate": 2.7476184358691206e-07,
+      "loss": 0.3131,
+      "step": 4011
+    },
+    {
+      "epoch": 2.7144790257104194,
+      "grad_norm": 1.166114882399594,
+      "learning_rate": 2.7347622910184445e-07,
+      "loss": 0.3199,
+      "step": 4012
+    },
+    {
+      "epoch": 2.7151556156968875,
+      "grad_norm": 1.1919275580767696,
+      "learning_rate": 2.7219354480675144e-07,
+      "loss": 0.3391,
+      "step": 4013
+    },
+    {
+      "epoch": 2.715832205683356,
+      "grad_norm": 1.1382464826637717,
+      "learning_rate": 2.7091379149682683e-07,
+      "loss": 0.2977,
+      "step": 4014
+    },
+    {
+      "epoch": 2.716508795669824,
+      "grad_norm": 1.1860334518820388,
+      "learning_rate": 2.696369699654489e-07,
+      "loss": 0.3163,
+      "step": 4015
+    },
+    {
+      "epoch": 2.7171853856562924,
+      "grad_norm": 1.1305080812300905,
+      "learning_rate": 2.6836308100417874e-07,
+      "loss": 0.3179,
+      "step": 4016
+    },
+    {
+      "epoch": 2.7178619756427604,
+      "grad_norm": 1.1478005912251539,
+      "learning_rate": 2.670921254027592e-07,
+      "loss": 0.3179,
+      "step": 4017
+    },
+    {
+      "epoch": 2.7185385656292285,
+      "grad_norm": 1.249165660172556,
+      "learning_rate": 2.6582410394911327e-07,
+      "loss": 0.3037,
+      "step": 4018
+    },
+    {
+      "epoch": 2.719215155615697,
+      "grad_norm": 1.1812469984608873,
+      "learning_rate": 2.6455901742934556e-07,
+      "loss": 0.3291,
+      "step": 4019
+    },
+    {
+      "epoch": 2.719891745602165,
+      "grad_norm": 1.1394091565896398,
+      "learning_rate": 2.6329686662774247e-07,
+      "loss": 0.3102,
+      "step": 4020
+    },
+    {
+      "epoch": 2.7205683355886334,
+      "grad_norm": 1.172131940690211,
+      "learning_rate": 2.620376523267698e-07,
+      "loss": 0.3186,
+      "step": 4021
+    },
+    {
+      "epoch": 2.7212449255751014,
+      "grad_norm": 1.131089064937029,
+      "learning_rate": 2.6078137530707146e-07,
+      "loss": 0.3118,
+      "step": 4022
+    },
+    {
+      "epoch": 2.7219215155615695,
+      "grad_norm": 1.080660036279916,
+      "learning_rate": 2.595280363474717e-07,
+      "loss": 0.2894,
+      "step": 4023
+    },
+    {
+      "epoch": 2.722598105548038,
+      "grad_norm": 1.1520481513191763,
+      "learning_rate": 2.582776362249739e-07,
+      "loss": 0.3103,
+      "step": 4024
+    },
+    {
+      "epoch": 2.723274695534506,
+      "grad_norm": 1.1409844303404006,
+      "learning_rate": 2.5703017571475755e-07,
+      "loss": 0.3115,
+      "step": 4025
+    },
+    {
+      "epoch": 2.7239512855209744,
+      "grad_norm": 1.1487512510352567,
+      "learning_rate": 2.5578565559018276e-07,
+      "loss": 0.3052,
+      "step": 4026
+    },
+    {
+      "epoch": 2.7246278755074425,
+      "grad_norm": 1.1594837220607868,
+      "learning_rate": 2.545440766227825e-07,
+      "loss": 0.3128,
+      "step": 4027
+    },
+    {
+      "epoch": 2.7253044654939105,
+      "grad_norm": 1.1184293351288757,
+      "learning_rate": 2.5330543958227036e-07,
+      "loss": 0.3132,
+      "step": 4028
+    },
+    {
+      "epoch": 2.725981055480379,
+      "grad_norm": 1.1164129239627643,
+      "learning_rate": 2.520697452365345e-07,
+      "loss": 0.2977,
+      "step": 4029
+    },
+    {
+      "epoch": 2.726657645466847,
+      "grad_norm": 1.182607668698762,
+      "learning_rate": 2.508369943516387e-07,
+      "loss": 0.3246,
+      "step": 4030
+    },
+    {
+      "epoch": 2.7273342354533154,
+      "grad_norm": 1.1640272681730328,
+      "learning_rate": 2.4960718769182214e-07,
+      "loss": 0.3267,
+      "step": 4031
+    },
+    {
+      "epoch": 2.7280108254397835,
+      "grad_norm": 1.1600039951419086,
+      "learning_rate": 2.483803260194978e-07,
+      "loss": 0.3112,
+      "step": 4032
+    },
+    {
+      "epoch": 2.7286874154262515,
+      "grad_norm": 1.172072214532289,
+      "learning_rate": 2.4715641009525446e-07,
+      "loss": 0.3123,
+      "step": 4033
+    },
+    {
+      "epoch": 2.72936400541272,
+      "grad_norm": 1.1617898967194897,
+      "learning_rate": 2.459354406778547e-07,
+      "loss": 0.3123,
+      "step": 4034
+    },
+    {
+      "epoch": 2.730040595399188,
+      "grad_norm": 1.1424919602451964,
+      "learning_rate": 2.447174185242324e-07,
+      "loss": 0.3065,
+      "step": 4035
+    },
+    {
+      "epoch": 2.7307171853856564,
+      "grad_norm": 1.1260938582436268,
+      "learning_rate": 2.4350234438949625e-07,
+      "loss": 0.2956,
+      "step": 4036
+    },
+    {
+      "epoch": 2.7313937753721245,
+      "grad_norm": 1.1743016504532113,
+      "learning_rate": 2.4229021902692663e-07,
+      "loss": 0.3195,
+      "step": 4037
+    },
+    {
+      "epoch": 2.7320703653585925,
+      "grad_norm": 1.101385386007145,
+      "learning_rate": 2.4108104318797674e-07,
+      "loss": 0.2995,
+      "step": 4038
+    },
+    {
+      "epoch": 2.732746955345061,
+      "grad_norm": 1.1269255833849812,
+      "learning_rate": 2.3987481762226984e-07,
+      "loss": 0.3074,
+      "step": 4039
+    },
+    {
+      "epoch": 2.733423545331529,
+      "grad_norm": 1.125855092520463,
+      "learning_rate": 2.3867154307759986e-07,
+      "loss": 0.3037,
+      "step": 4040
+    },
+    {
+      "epoch": 2.7341001353179974,
+      "grad_norm": 1.100393441841296,
+      "learning_rate": 2.3747122029993296e-07,
+      "loss": 0.3047,
+      "step": 4041
+    },
+    {
+      "epoch": 2.7347767253044655,
+      "grad_norm": 1.160766957345758,
+      "learning_rate": 2.3627385003340552e-07,
+      "loss": 0.3086,
+      "step": 4042
+    },
+    {
+      "epoch": 2.7354533152909335,
+      "grad_norm": 1.1635387600353782,
+      "learning_rate": 2.3507943302032045e-07,
+      "loss": 0.3233,
+      "step": 4043
+    },
+    {
+      "epoch": 2.736129905277402,
+      "grad_norm": 1.1728497369695257,
+      "learning_rate": 2.3388797000115427e-07,
+      "loss": 0.3234,
+      "step": 4044
+    },
+    {
+      "epoch": 2.73680649526387,
+      "grad_norm": 1.1728148975131911,
+      "learning_rate": 2.3269946171454727e-07,
+      "loss": 0.3157,
+      "step": 4045
+    },
+    {
+      "epoch": 2.7374830852503385,
+      "grad_norm": 1.1095476346839723,
+      "learning_rate": 2.3151390889731285e-07,
+      "loss": 0.3072,
+      "step": 4046
+    },
+    {
+      "epoch": 2.7381596752368065,
+      "grad_norm": 1.0998786334559265,
+      "learning_rate": 2.3033131228442863e-07,
+      "loss": 0.2985,
+      "step": 4047
+    },
+    {
+      "epoch": 2.7388362652232745,
+      "grad_norm": 1.1056490613710948,
+      "learning_rate": 2.2915167260904092e-07,
+      "loss": 0.302,
+      "step": 4048
+    },
+    {
+      "epoch": 2.739512855209743,
+      "grad_norm": 1.1118695733654114,
+      "learning_rate": 2.2797499060246253e-07,
+      "loss": 0.3082,
+      "step": 4049
+    },
+    {
+      "epoch": 2.740189445196211,
+      "grad_norm": 1.1508815373411923,
+      "learning_rate": 2.2680126699417383e-07,
+      "loss": 0.3112,
+      "step": 4050
+    },
+    {
+      "epoch": 2.7408660351826795,
+      "grad_norm": 1.1552472996586618,
+      "learning_rate": 2.256305025118194e-07,
+      "loss": 0.3103,
+      "step": 4051
+    },
+    {
+      "epoch": 2.7415426251691475,
+      "grad_norm": 1.1217024972889118,
+      "learning_rate": 2.244626978812109e-07,
+      "loss": 0.3135,
+      "step": 4052
+    },
+    {
+      "epoch": 2.7422192151556155,
+      "grad_norm": 1.1745396521279396,
+      "learning_rate": 2.2329785382632253e-07,
+      "loss": 0.3235,
+      "step": 4053
+    },
+    {
+      "epoch": 2.742895805142084,
+      "grad_norm": 1.1040598537793167,
+      "learning_rate": 2.2213597106929608e-07,
+      "loss": 0.3004,
+      "step": 4054
+    },
+    {
+      "epoch": 2.743572395128552,
+      "grad_norm": 1.185141353204137,
+      "learning_rate": 2.2097705033043703e-07,
+      "loss": 0.3203,
+      "step": 4055
+    },
+    {
+      "epoch": 2.7442489851150205,
+      "grad_norm": 1.1892370342215453,
+      "learning_rate": 2.198210923282118e-07,
+      "loss": 0.3119,
+      "step": 4056
+    },
+    {
+      "epoch": 2.7449255751014885,
+      "grad_norm": 1.1374531970626292,
+      "learning_rate": 2.1866809777925323e-07,
+      "loss": 0.3144,
+      "step": 4057
+    },
+    {
+      "epoch": 2.7456021650879565,
+      "grad_norm": 1.1202972350254548,
+      "learning_rate": 2.1751806739835624e-07,
+      "loss": 0.2989,
+      "step": 4058
+    },
+    {
+      "epoch": 2.746278755074425,
+      "grad_norm": 1.1168695709129022,
+      "learning_rate": 2.163710018984766e-07,
+      "loss": 0.3048,
+      "step": 4059
+    },
+    {
+      "epoch": 2.746955345060893,
+      "grad_norm": 1.2040695708185165,
+      "learning_rate": 2.1522690199073382e-07,
+      "loss": 0.3256,
+      "step": 4060
+    },
+    {
+      "epoch": 2.7476319350473615,
+      "grad_norm": 1.1507883712785076,
+      "learning_rate": 2.140857683844072e-07,
+      "loss": 0.3148,
+      "step": 4061
+    },
+    {
+      "epoch": 2.7483085250338295,
+      "grad_norm": 1.0805625359510496,
+      "learning_rate": 2.1294760178693918e-07,
+      "loss": 0.2851,
+      "step": 4062
+    },
+    {
+      "epoch": 2.7489851150202975,
+      "grad_norm": 1.0821339895994082,
+      "learning_rate": 2.118124029039309e-07,
+      "loss": 0.2893,
+      "step": 4063
+    },
+    {
+      "epoch": 2.749661705006766,
+      "grad_norm": 1.1465099231279592,
+      "learning_rate": 2.1068017243914663e-07,
+      "loss": 0.3247,
+      "step": 4064
+    },
+    {
+      "epoch": 2.750338294993234,
+      "grad_norm": 1.1445379537695655,
+      "learning_rate": 2.0955091109450488e-07,
+      "loss": 0.3118,
+      "step": 4065
+    },
+    {
+      "epoch": 2.7510148849797025,
+      "grad_norm": 1.1380250051116825,
+      "learning_rate": 2.0842461957008841e-07,
+      "loss": 0.2958,
+      "step": 4066
+    },
+    {
+      "epoch": 2.7516914749661705,
+      "grad_norm": 1.136764005997454,
+      "learning_rate": 2.0730129856413705e-07,
+      "loss": 0.3002,
+      "step": 4067
+    },
+    {
+      "epoch": 2.7523680649526385,
+      "grad_norm": 1.1440071360904445,
+      "learning_rate": 2.061809487730504e-07,
+      "loss": 0.3051,
+      "step": 4068
+    },
+    {
+      "epoch": 2.753044654939107,
+      "grad_norm": 1.1675675818474753,
+      "learning_rate": 2.050635708913834e-07,
+      "loss": 0.3153,
+      "step": 4069
+    },
+    {
+      "epoch": 2.753721244925575,
+      "grad_norm": 1.131607828217165,
+      "learning_rate": 2.0394916561185085e-07,
+      "loss": 0.3136,
+      "step": 4070
+    },
+    {
+      "epoch": 2.7543978349120435,
+      "grad_norm": 1.1720738901947765,
+      "learning_rate": 2.0283773362532455e-07,
+      "loss": 0.3218,
+      "step": 4071
+    },
+    {
+      "epoch": 2.7550744248985115,
+      "grad_norm": 1.166079943261266,
+      "learning_rate": 2.0172927562083056e-07,
+      "loss": 0.3156,
+      "step": 4072
+    },
+    {
+      "epoch": 2.7557510148849795,
+      "grad_norm": 1.0881044885961269,
+      "learning_rate": 2.006237922855553e-07,
+      "loss": 0.2975,
+      "step": 4073
+    },
+    {
+      "epoch": 2.756427604871448,
+      "grad_norm": 1.132434513163992,
+      "learning_rate": 1.9952128430483718e-07,
+      "loss": 0.3097,
+      "step": 4074
+    },
+    {
+      "epoch": 2.757104194857916,
+      "grad_norm": 1.1962221017978052,
+      "learning_rate": 1.9842175236217176e-07,
+      "loss": 0.3232,
+      "step": 4075
+    },
+    {
+      "epoch": 2.7577807848443845,
+      "grad_norm": 1.1314845406806597,
+      "learning_rate": 1.973251971392115e-07,
+      "loss": 0.3042,
+      "step": 4076
+    },
+    {
+      "epoch": 2.7584573748308525,
+      "grad_norm": 1.153865224777945,
+      "learning_rate": 1.962316193157593e-07,
+      "loss": 0.3152,
+      "step": 4077
+    },
+    {
+      "epoch": 2.7591339648173205,
+      "grad_norm": 1.1281332265264288,
+      "learning_rate": 1.9514101956977617e-07,
+      "loss": 0.3109,
+      "step": 4078
+    },
+    {
+      "epoch": 2.759810554803789,
+      "grad_norm": 1.1427027397151777,
+      "learning_rate": 1.9405339857737348e-07,
+      "loss": 0.2993,
+      "step": 4079
+    },
+    {
+      "epoch": 2.760487144790257,
+      "grad_norm": 1.1206618432879296,
+      "learning_rate": 1.9296875701281858e-07,
+      "loss": 0.3061,
+      "step": 4080
+    },
+    {
+      "epoch": 2.7611637347767255,
+      "grad_norm": 1.1269631300621363,
+      "learning_rate": 1.9188709554853137e-07,
+      "loss": 0.2995,
+      "step": 4081
+    },
+    {
+      "epoch": 2.7618403247631935,
+      "grad_norm": 1.1635517077821298,
+      "learning_rate": 1.9080841485508205e-07,
+      "loss": 0.3119,
+      "step": 4082
+    },
+    {
+      "epoch": 2.7625169147496615,
+      "grad_norm": 1.1708843569601985,
+      "learning_rate": 1.8973271560119576e-07,
+      "loss": 0.3186,
+      "step": 4083
+    },
+    {
+      "epoch": 2.76319350473613,
+      "grad_norm": 1.1710324106069068,
+      "learning_rate": 1.8865999845374794e-07,
+      "loss": 0.3067,
+      "step": 4084
+    },
+    {
+      "epoch": 2.763870094722598,
+      "grad_norm": 1.1225365775772305,
+      "learning_rate": 1.8759026407776605e-07,
+      "loss": 0.3107,
+      "step": 4085
+    },
+    {
+      "epoch": 2.7645466847090665,
+      "grad_norm": 1.0651871022090058,
+      "learning_rate": 1.8652351313642568e-07,
+      "loss": 0.2868,
+      "step": 4086
+    },
+    {
+      "epoch": 2.7652232746955345,
+      "grad_norm": 1.116776176522001,
+      "learning_rate": 1.8545974629105624e-07,
+      "loss": 0.3085,
+      "step": 4087
+    },
+    {
+      "epoch": 2.7658998646820026,
+      "grad_norm": 1.1364118546211295,
+      "learning_rate": 1.8439896420113569e-07,
+      "loss": 0.3015,
+      "step": 4088
+    },
+    {
+      "epoch": 2.766576454668471,
+      "grad_norm": 1.227473079179004,
+      "learning_rate": 1.8334116752429243e-07,
+      "loss": 0.3084,
+      "step": 4089
+    },
+    {
+      "epoch": 2.767253044654939,
+      "grad_norm": 1.1578007786831845,
+      "learning_rate": 1.8228635691630191e-07,
+      "loss": 0.312,
+      "step": 4090
+    },
+    {
+      "epoch": 2.7679296346414075,
+      "grad_norm": 1.1333879687887103,
+      "learning_rate": 1.812345330310916e-07,
+      "loss": 0.3163,
+      "step": 4091
+    },
+    {
+      "epoch": 2.7686062246278755,
+      "grad_norm": 1.1771525934157423,
+      "learning_rate": 1.801856965207338e-07,
+      "loss": 0.318,
+      "step": 4092
+    },
+    {
+      "epoch": 2.7692828146143436,
+      "grad_norm": 1.1258856212722625,
+      "learning_rate": 1.791398480354517e-07,
+      "loss": 0.3053,
+      "step": 4093
+    },
+    {
+      "epoch": 2.769959404600812,
+      "grad_norm": 1.1185884402081516,
+      "learning_rate": 1.78096988223615e-07,
+      "loss": 0.2996,
+      "step": 4094
+    },
+    {
+      "epoch": 2.77063599458728,
+      "grad_norm": 1.1276149308890613,
+      "learning_rate": 1.770571177317404e-07,
+      "loss": 0.3137,
+      "step": 4095
+    },
+    {
+      "epoch": 2.7713125845737485,
+      "grad_norm": 1.1124886369718174,
+      "learning_rate": 1.7602023720449114e-07,
+      "loss": 0.2968,
+      "step": 4096
+    },
+    {
+      "epoch": 2.7719891745602165,
+      "grad_norm": 1.1753353901708072,
+      "learning_rate": 1.74986347284678e-07,
+      "loss": 0.3122,
+      "step": 4097
+    },
+    {
+      "epoch": 2.7726657645466846,
+      "grad_norm": 1.1020443639067645,
+      "learning_rate": 1.7395544861325718e-07,
+      "loss": 0.2969,
+      "step": 4098
+    },
+    {
+      "epoch": 2.773342354533153,
+      "grad_norm": 1.180845965506115,
+      "learning_rate": 1.7292754182932914e-07,
+      "loss": 0.3105,
+      "step": 4099
+    },
+    {
+      "epoch": 2.774018944519621,
+      "grad_norm": 1.1508204490656577,
+      "learning_rate": 1.7190262757014076e-07,
+      "loss": 0.3149,
+      "step": 4100
+    },
+    {
+      "epoch": 2.7746955345060895,
+      "grad_norm": 1.1444752421633846,
+      "learning_rate": 1.7088070647108433e-07,
+      "loss": 0.3135,
+      "step": 4101
+    },
+    {
+      "epoch": 2.7753721244925575,
+      "grad_norm": 1.103582666802079,
+      "learning_rate": 1.6986177916569646e-07,
+      "loss": 0.3024,
+      "step": 4102
+    },
+    {
+      "epoch": 2.7760487144790256,
+      "grad_norm": 1.1541460657030975,
+      "learning_rate": 1.688458462856557e-07,
+      "loss": 0.3144,
+      "step": 4103
+    },
+    {
+      "epoch": 2.776725304465494,
+      "grad_norm": 1.1641081686268657,
+      "learning_rate": 1.6783290846078714e-07,
+      "loss": 0.3279,
+      "step": 4104
+    },
+    {
+      "epoch": 2.777401894451962,
+      "grad_norm": 1.0654124923045716,
+      "learning_rate": 1.6682296631905626e-07,
+      "loss": 0.2975,
+      "step": 4105
+    },
+    {
+      "epoch": 2.7780784844384305,
+      "grad_norm": 1.1435681623690126,
+      "learning_rate": 1.6581602048657387e-07,
+      "loss": 0.3078,
+      "step": 4106
+    },
+    {
+      "epoch": 2.7787550744248986,
+      "grad_norm": 1.157534135780399,
+      "learning_rate": 1.648120715875906e-07,
+      "loss": 0.3024,
+      "step": 4107
+    },
+    {
+      "epoch": 2.7794316644113666,
+      "grad_norm": 1.1346443620652469,
+      "learning_rate": 1.6381112024450196e-07,
+      "loss": 0.3096,
+      "step": 4108
+    },
+    {
+      "epoch": 2.780108254397835,
+      "grad_norm": 1.1859527271884611,
+      "learning_rate": 1.6281316707784377e-07,
+      "loss": 0.3219,
+      "step": 4109
+    },
+    {
+      "epoch": 2.780784844384303,
+      "grad_norm": 1.1083880407225266,
+      "learning_rate": 1.618182127062934e-07,
+      "loss": 0.3003,
+      "step": 4110
+    },
+    {
+      "epoch": 2.7814614343707715,
+      "grad_norm": 1.1909965957014754,
+      "learning_rate": 1.6082625774666793e-07,
+      "loss": 0.3149,
+      "step": 4111
+    },
+    {
+      "epoch": 2.7821380243572396,
+      "grad_norm": 1.1109292979066097,
+      "learning_rate": 1.5983730281392663e-07,
+      "loss": 0.3072,
+      "step": 4112
+    },
+    {
+      "epoch": 2.7828146143437076,
+      "grad_norm": 1.1317810099916719,
+      "learning_rate": 1.588513485211679e-07,
+      "loss": 0.3235,
+      "step": 4113
+    },
+    {
+      "epoch": 2.783491204330176,
+      "grad_norm": 1.1060728758762992,
+      "learning_rate": 1.5786839547963008e-07,
+      "loss": 0.3014,
+      "step": 4114
+    },
+    {
+      "epoch": 2.784167794316644,
+      "grad_norm": 1.151548816364964,
+      "learning_rate": 1.5688844429869232e-07,
+      "loss": 0.3159,
+      "step": 4115
+    },
+    {
+      "epoch": 2.7848443843031125,
+      "grad_norm": 1.1795173923464533,
+      "learning_rate": 1.5591149558587037e-07,
+      "loss": 0.3236,
+      "step": 4116
+    },
+    {
+      "epoch": 2.7855209742895806,
+      "grad_norm": 1.1230554910539787,
+      "learning_rate": 1.5493754994681977e-07,
+      "loss": 0.3006,
+      "step": 4117
+    },
+    {
+      "epoch": 2.7861975642760486,
+      "grad_norm": 1.1274673045913184,
+      "learning_rate": 1.539666079853358e-07,
+      "loss": 0.3051,
+      "step": 4118
+    },
+    {
+      "epoch": 2.786874154262517,
+      "grad_norm": 1.093550164196601,
+      "learning_rate": 1.5299867030334815e-07,
+      "loss": 0.3016,
+      "step": 4119
+    },
+    {
+      "epoch": 2.787550744248985,
+      "grad_norm": 1.1354627153320578,
+      "learning_rate": 1.5203373750092676e-07,
+      "loss": 0.2982,
+      "step": 4120
+    },
+    {
+      "epoch": 2.7882273342354535,
+      "grad_norm": 1.1272910263619973,
+      "learning_rate": 1.5107181017627813e-07,
+      "loss": 0.2953,
+      "step": 4121
+    },
+    {
+      "epoch": 2.7889039242219216,
+      "grad_norm": 1.1757951011000707,
+      "learning_rate": 1.5011288892574526e-07,
+      "loss": 0.3242,
+      "step": 4122
+    },
+    {
+      "epoch": 2.7895805142083896,
+      "grad_norm": 1.1516008767284027,
+      "learning_rate": 1.4915697434380816e-07,
+      "loss": 0.3157,
+      "step": 4123
+    },
+    {
+      "epoch": 2.790257104194858,
+      "grad_norm": 1.1053151641072043,
+      "learning_rate": 1.4820406702308165e-07,
+      "loss": 0.3041,
+      "step": 4124
+    },
+    {
+      "epoch": 2.790933694181326,
+      "grad_norm": 1.1537120198382154,
+      "learning_rate": 1.4725416755431655e-07,
+      "loss": 0.3159,
+      "step": 4125
+    },
+    {
+      "epoch": 2.7916102841677946,
+      "grad_norm": 1.1182513124253504,
+      "learning_rate": 1.463072765264001e-07,
+      "loss": 0.3161,
+      "step": 4126
+    },
+    {
+      "epoch": 2.7922868741542626,
+      "grad_norm": 1.1111189281912415,
+      "learning_rate": 1.4536339452635385e-07,
+      "loss": 0.3024,
+      "step": 4127
+    },
+    {
+      "epoch": 2.7929634641407306,
+      "grad_norm": 1.0917773822743122,
+      "learning_rate": 1.444225221393325e-07,
+      "loss": 0.2948,
+      "step": 4128
+    },
+    {
+      "epoch": 2.793640054127199,
+      "grad_norm": 1.1243183403401535,
+      "learning_rate": 1.4348465994862782e-07,
+      "loss": 0.3141,
+      "step": 4129
+    },
+    {
+      "epoch": 2.794316644113667,
+      "grad_norm": 1.1228307517491027,
+      "learning_rate": 1.4254980853566248e-07,
+      "loss": 0.3115,
+      "step": 4130
+    },
+    {
+      "epoch": 2.7949932341001356,
+      "grad_norm": 1.1239836066876274,
+      "learning_rate": 1.4161796847999566e-07,
+      "loss": 0.2951,
+      "step": 4131
+    },
+    {
+      "epoch": 2.7956698240866036,
+      "grad_norm": 1.2078843049461978,
+      "learning_rate": 1.4068914035931635e-07,
+      "loss": 0.3128,
+      "step": 4132
+    },
+    {
+      "epoch": 2.7963464140730716,
+      "grad_norm": 1.0981294995686268,
+      "learning_rate": 1.3976332474944842e-07,
+      "loss": 0.2987,
+      "step": 4133
+    },
+    {
+      "epoch": 2.79702300405954,
+      "grad_norm": 1.1587213188731176,
+      "learning_rate": 1.388405222243472e-07,
+      "loss": 0.3156,
+      "step": 4134
+    },
+    {
+      "epoch": 2.797699594046008,
+      "grad_norm": 1.1846858692026618,
+      "learning_rate": 1.3792073335610111e-07,
+      "loss": 0.3175,
+      "step": 4135
+    },
+    {
+      "epoch": 2.7983761840324766,
+      "grad_norm": 1.1258815737012846,
+      "learning_rate": 1.3700395871493023e-07,
+      "loss": 0.3107,
+      "step": 4136
+    },
+    {
+      "epoch": 2.7990527740189446,
+      "grad_norm": 1.11739619207689,
+      "learning_rate": 1.360901988691843e-07,
+      "loss": 0.3015,
+      "step": 4137
+    },
+    {
+      "epoch": 2.7997293640054126,
+      "grad_norm": 1.1153167164708462,
+      "learning_rate": 1.3517945438534629e-07,
+      "loss": 0.3036,
+      "step": 4138
+    },
+    {
+      "epoch": 2.800405953991881,
+      "grad_norm": 1.148846921294776,
+      "learning_rate": 1.342717258280274e-07,
+      "loss": 0.3078,
+      "step": 4139
+    },
+    {
+      "epoch": 2.801082543978349,
+      "grad_norm": 1.1362173734992982,
+      "learning_rate": 1.333670137599713e-07,
+      "loss": 0.3078,
+      "step": 4140
+    },
+    {
+      "epoch": 2.8017591339648176,
+      "grad_norm": 1.145705430890723,
+      "learning_rate": 1.3246531874204994e-07,
+      "loss": 0.3106,
+      "step": 4141
+    },
+    {
+      "epoch": 2.8024357239512856,
+      "grad_norm": 1.1355565467776323,
+      "learning_rate": 1.3156664133326614e-07,
+      "loss": 0.3106,
+      "step": 4142
+    },
+    {
+      "epoch": 2.8031123139377536,
+      "grad_norm": 1.0968762107052787,
+      "learning_rate": 1.3067098209075202e-07,
+      "loss": 0.2931,
+      "step": 4143
+    },
+    {
+      "epoch": 2.803788903924222,
+      "grad_norm": 1.181016506230049,
+      "learning_rate": 1.2977834156976733e-07,
+      "loss": 0.3183,
+      "step": 4144
+    },
+    {
+      "epoch": 2.80446549391069,
+      "grad_norm": 1.168940594706016,
+      "learning_rate": 1.2888872032370103e-07,
+      "loss": 0.3142,
+      "step": 4145
+    },
+    {
+      "epoch": 2.8051420838971586,
+      "grad_norm": 1.1187113158210438,
+      "learning_rate": 1.280021189040709e-07,
+      "loss": 0.3065,
+      "step": 4146
+    },
+    {
+      "epoch": 2.8058186738836266,
+      "grad_norm": 1.1074702361941224,
+      "learning_rate": 1.2711853786052108e-07,
+      "loss": 0.2937,
+      "step": 4147
+    },
+    {
+      "epoch": 2.8064952638700946,
+      "grad_norm": 1.1815011731635956,
+      "learning_rate": 1.2623797774082514e-07,
+      "loss": 0.3145,
+      "step": 4148
+    },
+    {
+      "epoch": 2.8071718538565626,
+      "grad_norm": 1.1651985673228868,
+      "learning_rate": 1.253604390908819e-07,
+      "loss": 0.322,
+      "step": 4149
+    },
+    {
+      "epoch": 2.807848443843031,
+      "grad_norm": 1.1611545199311877,
+      "learning_rate": 1.2448592245471903e-07,
+      "loss": 0.3175,
+      "step": 4150
+    },
+    {
+      "epoch": 2.8085250338294996,
+      "grad_norm": 1.1423329228733659,
+      "learning_rate": 1.2361442837449e-07,
+      "loss": 0.3126,
+      "step": 4151
+    },
+    {
+      "epoch": 2.8092016238159676,
+      "grad_norm": 1.094420894467688,
+      "learning_rate": 1.2274595739047267e-07,
+      "loss": 0.2927,
+      "step": 4152
+    },
+    {
+      "epoch": 2.8098782138024356,
+      "grad_norm": 1.1482770593230631,
+      "learning_rate": 1.2188051004107305e-07,
+      "loss": 0.311,
+      "step": 4153
+    },
+    {
+      "epoch": 2.8105548037889037,
+      "grad_norm": 1.1870469052793697,
+      "learning_rate": 1.210180868628219e-07,
+      "loss": 0.3095,
+      "step": 4154
+    },
+    {
+      "epoch": 2.811231393775372,
+      "grad_norm": 1.1079696291691443,
+      "learning_rate": 1.2015868839037492e-07,
+      "loss": 0.3073,
+      "step": 4155
+    },
+    {
+      "epoch": 2.8119079837618406,
+      "grad_norm": 1.151093810299083,
+      "learning_rate": 1.1930231515651313e-07,
+      "loss": 0.3084,
+      "step": 4156
+    },
+    {
+      "epoch": 2.8125845737483086,
+      "grad_norm": 1.1479766364856205,
+      "learning_rate": 1.1844896769214187e-07,
+      "loss": 0.306,
+      "step": 4157
+    },
+    {
+      "epoch": 2.8132611637347766,
+      "grad_norm": 1.194479911283576,
+      "learning_rate": 1.1759864652629072e-07,
+      "loss": 0.3264,
+      "step": 4158
+    },
+    {
+      "epoch": 2.8139377537212447,
+      "grad_norm": 1.142204995609788,
+      "learning_rate": 1.1675135218611188e-07,
+      "loss": 0.3127,
+      "step": 4159
+    },
+    {
+      "epoch": 2.814614343707713,
+      "grad_norm": 1.136898076727429,
+      "learning_rate": 1.1590708519688243e-07,
+      "loss": 0.3113,
+      "step": 4160
+    },
+    {
+      "epoch": 2.8152909336941816,
+      "grad_norm": 1.1563326571565744,
+      "learning_rate": 1.1506584608200366e-07,
+      "loss": 0.325,
+      "step": 4161
+    },
+    {
+      "epoch": 2.8159675236806496,
+      "grad_norm": 1.1685954336342999,
+      "learning_rate": 1.142276353629973e-07,
+      "loss": 0.3256,
+      "step": 4162
+    },
+    {
+      "epoch": 2.8166441136671176,
+      "grad_norm": 1.1592850359604618,
+      "learning_rate": 1.1339245355950934e-07,
+      "loss": 0.3117,
+      "step": 4163
+    },
+    {
+      "epoch": 2.8173207036535857,
+      "grad_norm": 1.1657630261390621,
+      "learning_rate": 1.1256030118930727e-07,
+      "loss": 0.3068,
+      "step": 4164
+    },
+    {
+      "epoch": 2.817997293640054,
+      "grad_norm": 1.1398606878363755,
+      "learning_rate": 1.1173117876828066e-07,
+      "loss": 0.3019,
+      "step": 4165
+    },
+    {
+      "epoch": 2.8186738836265226,
+      "grad_norm": 1.1695017410378856,
+      "learning_rate": 1.1090508681044055e-07,
+      "loss": 0.3173,
+      "step": 4166
+    },
+    {
+      "epoch": 2.8193504736129906,
+      "grad_norm": 1.1425891144598073,
+      "learning_rate": 1.1008202582792005e-07,
+      "loss": 0.296,
+      "step": 4167
+    },
+    {
+      "epoch": 2.8200270635994586,
+      "grad_norm": 1.1568237647767319,
+      "learning_rate": 1.0926199633097156e-07,
+      "loss": 0.3043,
+      "step": 4168
+    },
+    {
+      "epoch": 2.8207036535859267,
+      "grad_norm": 1.1625379556250555,
+      "learning_rate": 1.0844499882797011e-07,
+      "loss": 0.3052,
+      "step": 4169
+    },
+    {
+      "epoch": 2.821380243572395,
+      "grad_norm": 1.1568963498525002,
+      "learning_rate": 1.0763103382541052e-07,
+      "loss": 0.312,
+      "step": 4170
+    },
+    {
+      "epoch": 2.822056833558863,
+      "grad_norm": 1.146058362233935,
+      "learning_rate": 1.0682010182790637e-07,
+      "loss": 0.3071,
+      "step": 4171
+    },
+    {
+      "epoch": 2.8227334235453316,
+      "grad_norm": 1.1541650012035944,
+      "learning_rate": 1.0601220333819162e-07,
+      "loss": 0.307,
+      "step": 4172
+    },
+    {
+      "epoch": 2.8234100135317997,
+      "grad_norm": 1.1650657993777809,
+      "learning_rate": 1.0520733885712008e-07,
+      "loss": 0.3219,
+      "step": 4173
+    },
+    {
+      "epoch": 2.8240866035182677,
+      "grad_norm": 1.189695651515764,
+      "learning_rate": 1.0440550888366485e-07,
+      "loss": 0.3261,
+      "step": 4174
+    },
+    {
+      "epoch": 2.824763193504736,
+      "grad_norm": 1.1410751873083764,
+      "learning_rate": 1.0360671391491606e-07,
+      "loss": 0.3048,
+      "step": 4175
+    },
+    {
+      "epoch": 2.825439783491204,
+      "grad_norm": 1.181271198859864,
+      "learning_rate": 1.0281095444608425e-07,
+      "loss": 0.3228,
+      "step": 4176
+    },
+    {
+      "epoch": 2.8261163734776726,
+      "grad_norm": 1.0996530026726692,
+      "learning_rate": 1.0201823097049812e-07,
+      "loss": 0.3094,
+      "step": 4177
+    },
+    {
+      "epoch": 2.8267929634641407,
+      "grad_norm": 1.1427832540400786,
+      "learning_rate": 1.0122854397960292e-07,
+      "loss": 0.306,
+      "step": 4178
+    },
+    {
+      "epoch": 2.8274695534506087,
+      "grad_norm": 1.1551453432551007,
+      "learning_rate": 1.0044189396296144e-07,
+      "loss": 0.3134,
+      "step": 4179
+    },
+    {
+      "epoch": 2.828146143437077,
+      "grad_norm": 1.142606923697455,
+      "learning_rate": 9.965828140825529e-08,
+      "loss": 0.3079,
+      "step": 4180
+    },
+    {
+      "epoch": 2.828822733423545,
+      "grad_norm": 1.095129495826332,
+      "learning_rate": 9.887770680128083e-08,
+      "loss": 0.2974,
+      "step": 4181
+    },
+    {
+      "epoch": 2.8294993234100136,
+      "grad_norm": 1.15517421321211,
+      "learning_rate": 9.810017062595322e-08,
+      "loss": 0.3167,
+      "step": 4182
+    },
+    {
+      "epoch": 2.8301759133964817,
+      "grad_norm": 1.1648548269503585,
+      "learning_rate": 9.732567336430298e-08,
+      "loss": 0.3122,
+      "step": 4183
+    },
+    {
+      "epoch": 2.8308525033829497,
+      "grad_norm": 1.1182338854160827,
+      "learning_rate": 9.655421549647603e-08,
+      "loss": 0.3177,
+      "step": 4184
+    },
+    {
+      "epoch": 2.831529093369418,
+      "grad_norm": 1.1405031371526768,
+      "learning_rate": 9.57857975007348e-08,
+      "loss": 0.3172,
+      "step": 4185
+    },
+    {
+      "epoch": 2.832205683355886,
+      "grad_norm": 1.1440733212189316,
+      "learning_rate": 9.502041985345766e-08,
+      "loss": 0.3128,
+      "step": 4186
+    },
+    {
+      "epoch": 2.8328822733423547,
+      "grad_norm": 1.1361539508525604,
+      "learning_rate": 9.42580830291373e-08,
+      "loss": 0.3085,
+      "step": 4187
+    },
+    {
+      "epoch": 2.8335588633288227,
+      "grad_norm": 1.1509590013050146,
+      "learning_rate": 9.349878750038067e-08,
+      "loss": 0.2973,
+      "step": 4188
+    },
+    {
+      "epoch": 2.8342354533152907,
+      "grad_norm": 1.1329417952007672,
+      "learning_rate": 9.274253373791064e-08,
+      "loss": 0.3182,
+      "step": 4189
+    },
+    {
+      "epoch": 2.834912043301759,
+      "grad_norm": 1.1376827475730935,
+      "learning_rate": 9.198932221056333e-08,
+      "loss": 0.3124,
+      "step": 4190
+    },
+    {
+      "epoch": 2.835588633288227,
+      "grad_norm": 1.1393405169013877,
+      "learning_rate": 9.123915338529132e-08,
+      "loss": 0.3027,
+      "step": 4191
+    },
+    {
+      "epoch": 2.8362652232746957,
+      "grad_norm": 1.098878898100899,
+      "learning_rate": 9.049202772715593e-08,
+      "loss": 0.2994,
+      "step": 4192
+    },
+    {
+      "epoch": 2.8369418132611637,
+      "grad_norm": 1.158227816546315,
+      "learning_rate": 8.974794569933609e-08,
+      "loss": 0.3149,
+      "step": 4193
+    },
+    {
+      "epoch": 2.8376184032476317,
+      "grad_norm": 1.1326083176134083,
+      "learning_rate": 8.900690776312282e-08,
+      "loss": 0.3083,
+      "step": 4194
+    },
+    {
+      "epoch": 2.8382949932341,
+      "grad_norm": 1.1070998118615791,
+      "learning_rate": 8.826891437791974e-08,
+      "loss": 0.3114,
+      "step": 4195
+    },
+    {
+      "epoch": 2.838971583220568,
+      "grad_norm": 1.1564043460596638,
+      "learning_rate": 8.753396600124254e-08,
+      "loss": 0.3241,
+      "step": 4196
+    },
+    {
+      "epoch": 2.8396481732070367,
+      "grad_norm": 1.1348179095928945,
+      "learning_rate": 8.680206308871953e-08,
+      "loss": 0.3148,
+      "step": 4197
+    },
+    {
+      "epoch": 2.8403247631935047,
+      "grad_norm": 1.2031507738983054,
+      "learning_rate": 8.607320609409165e-08,
+      "loss": 0.3157,
+      "step": 4198
+    },
+    {
+      "epoch": 2.8410013531799727,
+      "grad_norm": 1.1558767804874057,
+      "learning_rate": 8.534739546921023e-08,
+      "loss": 0.3131,
+      "step": 4199
+    },
+    {
+      "epoch": 2.841677943166441,
+      "grad_norm": 1.1134903575631645,
+      "learning_rate": 8.462463166403978e-08,
+      "loss": 0.303,
+      "step": 4200
+    },
+    {
+      "epoch": 2.842354533152909,
+      "grad_norm": 1.14888777254408,
+      "learning_rate": 8.390491512665355e-08,
+      "loss": 0.3129,
+      "step": 4201
+    },
+    {
+      "epoch": 2.8430311231393777,
+      "grad_norm": 1.1558301195406901,
+      "learning_rate": 8.318824630323741e-08,
+      "loss": 0.3096,
+      "step": 4202
+    },
+    {
+      "epoch": 2.8437077131258457,
+      "grad_norm": 1.151349654889666,
+      "learning_rate": 8.247462563808816e-08,
+      "loss": 0.321,
+      "step": 4203
+    },
+    {
+      "epoch": 2.8443843031123137,
+      "grad_norm": 1.1615776626594305,
+      "learning_rate": 8.176405357361194e-08,
+      "loss": 0.3128,
+      "step": 4204
+    },
+    {
+      "epoch": 2.845060893098782,
+      "grad_norm": 1.1550974909648855,
+      "learning_rate": 8.105653055032415e-08,
+      "loss": 0.3149,
+      "step": 4205
+    },
+    {
+      "epoch": 2.84573748308525,
+      "grad_norm": 1.1331678946526835,
+      "learning_rate": 8.035205700685167e-08,
+      "loss": 0.3181,
+      "step": 4206
+    },
+    {
+      "epoch": 2.8464140730717187,
+      "grad_norm": 1.1529725830179045,
+      "learning_rate": 7.965063337993018e-08,
+      "loss": 0.3099,
+      "step": 4207
+    },
+    {
+      "epoch": 2.8470906630581867,
+      "grad_norm": 1.167392401492405,
+      "learning_rate": 7.89522601044046e-08,
+      "loss": 0.3123,
+      "step": 4208
+    },
+    {
+      "epoch": 2.8477672530446547,
+      "grad_norm": 1.117294557253064,
+      "learning_rate": 7.825693761322861e-08,
+      "loss": 0.3133,
+      "step": 4209
+    },
+    {
+      "epoch": 2.848443843031123,
+      "grad_norm": 1.206116459835629,
+      "learning_rate": 7.756466633746407e-08,
+      "loss": 0.3229,
+      "step": 4210
+    },
+    {
+      "epoch": 2.849120433017591,
+      "grad_norm": 1.1468698271960456,
+      "learning_rate": 7.687544670628267e-08,
+      "loss": 0.3129,
+      "step": 4211
+    },
+    {
+      "epoch": 2.8497970230040597,
+      "grad_norm": 1.1055694300739007,
+      "learning_rate": 7.618927914696372e-08,
+      "loss": 0.3044,
+      "step": 4212
+    },
+    {
+      "epoch": 2.8504736129905277,
+      "grad_norm": 1.1838200930266098,
+      "learning_rate": 7.550616408489253e-08,
+      "loss": 0.3287,
+      "step": 4213
+    },
+    {
+      "epoch": 2.8511502029769957,
+      "grad_norm": 1.1044320622923602,
+      "learning_rate": 7.482610194356477e-08,
+      "loss": 0.3069,
+      "step": 4214
+    },
+    {
+      "epoch": 2.851826792963464,
+      "grad_norm": 1.1455420346697114,
+      "learning_rate": 7.414909314458263e-08,
+      "loss": 0.3084,
+      "step": 4215
+    },
+    {
+      "epoch": 2.852503382949932,
+      "grad_norm": 1.1684253547397263,
+      "learning_rate": 7.347513810765427e-08,
+      "loss": 0.3204,
+      "step": 4216
+    },
+    {
+      "epoch": 2.8531799729364007,
+      "grad_norm": 1.1195624570775045,
+      "learning_rate": 7.280423725059604e-08,
+      "loss": 0.3079,
+      "step": 4217
+    },
+    {
+      "epoch": 2.8538565629228687,
+      "grad_norm": 1.1385479529534084,
+      "learning_rate": 7.213639098933022e-08,
+      "loss": 0.3084,
+      "step": 4218
+    },
+    {
+      "epoch": 2.8545331529093367,
+      "grad_norm": 1.1331069585572051,
+      "learning_rate": 7.147159973788508e-08,
+      "loss": 0.3079,
+      "step": 4219
+    },
+    {
+      "epoch": 2.855209742895805,
+      "grad_norm": 1.1465147205919115,
+      "learning_rate": 7.080986390839539e-08,
+      "loss": 0.3184,
+      "step": 4220
+    },
+    {
+      "epoch": 2.8558863328822732,
+      "grad_norm": 1.1791051770711483,
+      "learning_rate": 7.015118391110299e-08,
+      "loss": 0.3359,
+      "step": 4221
+    },
+    {
+      "epoch": 2.8565629228687417,
+      "grad_norm": 1.1773610438147977,
+      "learning_rate": 6.949556015435178e-08,
+      "loss": 0.324,
+      "step": 4222
+    },
+    {
+      "epoch": 2.8572395128552097,
+      "grad_norm": 1.1946754374606703,
+      "learning_rate": 6.884299304459497e-08,
+      "loss": 0.3171,
+      "step": 4223
+    },
+    {
+      "epoch": 2.8579161028416777,
+      "grad_norm": 1.211898989104555,
+      "learning_rate": 6.819348298638839e-08,
+      "loss": 0.3238,
+      "step": 4224
+    },
+    {
+      "epoch": 2.858592692828146,
+      "grad_norm": 1.1204644983851304,
+      "learning_rate": 6.75470303823933e-08,
+      "loss": 0.306,
+      "step": 4225
+    },
+    {
+      "epoch": 2.8592692828146142,
+      "grad_norm": 1.1183598532251589,
+      "learning_rate": 6.690363563337466e-08,
+      "loss": 0.3043,
+      "step": 4226
+    },
+    {
+      "epoch": 2.8599458728010827,
+      "grad_norm": 1.1104493994712024,
+      "learning_rate": 6.626329913820339e-08,
+      "loss": 0.3217,
+      "step": 4227
+    },
+    {
+      "epoch": 2.8606224627875507,
+      "grad_norm": 1.1383910628787934,
+      "learning_rate": 6.562602129385365e-08,
+      "loss": 0.3103,
+      "step": 4228
+    },
+    {
+      "epoch": 2.8612990527740187,
+      "grad_norm": 1.1456320872057297,
+      "learning_rate": 6.499180249540382e-08,
+      "loss": 0.3264,
+      "step": 4229
+    },
+    {
+      "epoch": 2.861975642760487,
+      "grad_norm": 1.0972800260668512,
+      "learning_rate": 6.436064313603385e-08,
+      "loss": 0.2902,
+      "step": 4230
+    },
+    {
+      "epoch": 2.8626522327469552,
+      "grad_norm": 1.1457624918306197,
+      "learning_rate": 6.373254360703019e-08,
+      "loss": 0.3108,
+      "step": 4231
+    },
+    {
+      "epoch": 2.8633288227334237,
+      "grad_norm": 1.1740680025880672,
+      "learning_rate": 6.310750429777912e-08,
+      "loss": 0.3209,
+      "step": 4232
+    },
+    {
+      "epoch": 2.8640054127198917,
+      "grad_norm": 1.156124442542606,
+      "learning_rate": 6.248552559577292e-08,
+      "loss": 0.3127,
+      "step": 4233
+    },
+    {
+      "epoch": 2.8646820027063598,
+      "grad_norm": 1.111208197867936,
+      "learning_rate": 6.186660788660315e-08,
+      "loss": 0.3053,
+      "step": 4234
+    },
+    {
+      "epoch": 2.865358592692828,
+      "grad_norm": 1.1541823175218264,
+      "learning_rate": 6.125075155396675e-08,
+      "loss": 0.3137,
+      "step": 4235
+    },
+    {
+      "epoch": 2.8660351826792962,
+      "grad_norm": 1.1556398548370799,
+      "learning_rate": 6.063795697966057e-08,
+      "loss": 0.3099,
+      "step": 4236
+    },
+    {
+      "epoch": 2.8667117726657647,
+      "grad_norm": 1.1474723616040594,
+      "learning_rate": 6.00282245435857e-08,
+      "loss": 0.3032,
+      "step": 4237
+    },
+    {
+      "epoch": 2.8673883626522327,
+      "grad_norm": 1.1636513063535923,
+      "learning_rate": 5.9421554623742e-08,
+      "loss": 0.3204,
+      "step": 4238
+    },
+    {
+      "epoch": 2.8680649526387008,
+      "grad_norm": 1.151528450500134,
+      "learning_rate": 5.881794759623194e-08,
+      "loss": 0.3102,
+      "step": 4239
+    },
+    {
+      "epoch": 2.8687415426251692,
+      "grad_norm": 1.1227225002615908,
+      "learning_rate": 5.8217403835260086e-08,
+      "loss": 0.3077,
+      "step": 4240
+    },
+    {
+      "epoch": 2.8694181326116373,
+      "grad_norm": 1.1314870892298636,
+      "learning_rate": 5.7619923713130857e-08,
+      "loss": 0.314,
+      "step": 4241
+    },
+    {
+      "epoch": 2.8700947225981057,
+      "grad_norm": 1.1341287369437187,
+      "learning_rate": 5.7025507600250165e-08,
+      "loss": 0.3124,
+      "step": 4242
+    },
+    {
+      "epoch": 2.8707713125845737,
+      "grad_norm": 1.1221479027133288,
+      "learning_rate": 5.643415586512324e-08,
+      "loss": 0.3027,
+      "step": 4243
+    },
+    {
+      "epoch": 2.8714479025710418,
+      "grad_norm": 1.1445270725231738,
+      "learning_rate": 5.584586887435739e-08,
+      "loss": 0.3194,
+      "step": 4244
+    },
+    {
+      "epoch": 2.8721244925575102,
+      "grad_norm": 1.106224431618639,
+      "learning_rate": 5.526064699265754e-08,
+      "loss": 0.3021,
+      "step": 4245
+    },
+    {
+      "epoch": 2.8728010825439783,
+      "grad_norm": 1.144408374954783,
+      "learning_rate": 5.4678490582830704e-08,
+      "loss": 0.308,
+      "step": 4246
+    },
+    {
+      "epoch": 2.8734776725304467,
+      "grad_norm": 1.1519235955199802,
+      "learning_rate": 5.409940000578207e-08,
+      "loss": 0.3254,
+      "step": 4247
+    },
+    {
+      "epoch": 2.8741542625169147,
+      "grad_norm": 1.154899227868188,
+      "learning_rate": 5.352337562051613e-08,
+      "loss": 0.3223,
+      "step": 4248
+    },
+    {
+      "epoch": 2.8748308525033828,
+      "grad_norm": 1.162412920809482,
+      "learning_rate": 5.2950417784137785e-08,
+      "loss": 0.3165,
+      "step": 4249
+    },
+    {
+      "epoch": 2.8755074424898512,
+      "grad_norm": 1.1324873167755383,
+      "learning_rate": 5.2380526851850135e-08,
+      "loss": 0.3004,
+      "step": 4250
+    },
+    {
+      "epoch": 2.8761840324763193,
+      "grad_norm": 1.1099522242217954,
+      "learning_rate": 5.181370317695389e-08,
+      "loss": 0.3053,
+      "step": 4251
+    },
+    {
+      "epoch": 2.8768606224627877,
+      "grad_norm": 1.1489724733135365,
+      "learning_rate": 5.124994711084963e-08,
+      "loss": 0.3036,
+      "step": 4252
+    },
+    {
+      "epoch": 2.8775372124492558,
+      "grad_norm": 1.115590005852624,
+      "learning_rate": 5.0689259003035566e-08,
+      "loss": 0.3109,
+      "step": 4253
+    },
+    {
+      "epoch": 2.878213802435724,
+      "grad_norm": 1.1517167262264398,
+      "learning_rate": 5.013163920110864e-08,
+      "loss": 0.3099,
+      "step": 4254
+    },
+    {
+      "epoch": 2.8788903924221922,
+      "grad_norm": 1.1240083042703695,
+      "learning_rate": 4.9577088050762337e-08,
+      "loss": 0.3031,
+      "step": 4255
+    },
+    {
+      "epoch": 2.8795669824086603,
+      "grad_norm": 1.1318330550373505,
+      "learning_rate": 4.9025605895788867e-08,
+      "loss": 0.3164,
+      "step": 4256
+    },
+    {
+      "epoch": 2.8802435723951287,
+      "grad_norm": 1.239141698502644,
+      "learning_rate": 4.847719307807752e-08,
+      "loss": 0.3306,
+      "step": 4257
+    },
+    {
+      "epoch": 2.8809201623815968,
+      "grad_norm": 1.134785816446629,
+      "learning_rate": 4.793184993761468e-08,
+      "loss": 0.2994,
+      "step": 4258
+    },
+    {
+      "epoch": 2.881596752368065,
+      "grad_norm": 1.1317951031414402,
+      "learning_rate": 4.73895768124838e-08,
+      "loss": 0.3067,
+      "step": 4259
+    },
+    {
+      "epoch": 2.8822733423545333,
+      "grad_norm": 1.1435087101163084,
+      "learning_rate": 4.685037403886483e-08,
+      "loss": 0.3139,
+      "step": 4260
+    },
+    {
+      "epoch": 2.8829499323410013,
+      "grad_norm": 1.1418542042945439,
+      "learning_rate": 4.631424195103373e-08,
+      "loss": 0.3044,
+      "step": 4261
+    },
+    {
+      "epoch": 2.8836265223274697,
+      "grad_norm": 1.195661538243185,
+      "learning_rate": 4.578118088136463e-08,
+      "loss": 0.3119,
+      "step": 4262
+    },
+    {
+      "epoch": 2.8843031123139378,
+      "grad_norm": 1.099512951186891,
+      "learning_rate": 4.52511911603265e-08,
+      "loss": 0.2981,
+      "step": 4263
+    },
+    {
+      "epoch": 2.884979702300406,
+      "grad_norm": 1.1341522644029463,
+      "learning_rate": 4.4724273116483754e-08,
+      "loss": 0.3063,
+      "step": 4264
+    },
+    {
+      "epoch": 2.8856562922868743,
+      "grad_norm": 1.1028914095176316,
+      "learning_rate": 4.42004270764973e-08,
+      "loss": 0.3045,
+      "step": 4265
+    },
+    {
+      "epoch": 2.8863328822733423,
+      "grad_norm": 1.1768256494467735,
+      "learning_rate": 4.367965336512403e-08,
+      "loss": 0.3246,
+      "step": 4266
+    },
+    {
+      "epoch": 2.8870094722598107,
+      "grad_norm": 1.1216950161209396,
+      "learning_rate": 4.316195230521514e-08,
+      "loss": 0.3068,
+      "step": 4267
+    },
+    {
+      "epoch": 2.8876860622462788,
+      "grad_norm": 1.1218076507246368,
+      "learning_rate": 4.264732421771722e-08,
+      "loss": 0.3044,
+      "step": 4268
+    },
+    {
+      "epoch": 2.888362652232747,
+      "grad_norm": 1.159660336650667,
+      "learning_rate": 4.21357694216723e-08,
+      "loss": 0.3052,
+      "step": 4269
+    },
+    {
+      "epoch": 2.8890392422192153,
+      "grad_norm": 1.1531078466110107,
+      "learning_rate": 4.162728823421669e-08,
+      "loss": 0.3077,
+      "step": 4270
+    },
+    {
+      "epoch": 2.8897158322056833,
+      "grad_norm": 1.140202671945342,
+      "learning_rate": 4.112188097058156e-08,
+      "loss": 0.3162,
+      "step": 4271
+    },
+    {
+      "epoch": 2.8903924221921518,
+      "grad_norm": 1.1292119945054573,
+      "learning_rate": 4.061954794409184e-08,
+      "loss": 0.3163,
+      "step": 4272
+    },
+    {
+      "epoch": 2.89106901217862,
+      "grad_norm": 1.1191609652179115,
+      "learning_rate": 4.0120289466166754e-08,
+      "loss": 0.3037,
+      "step": 4273
+    },
+    {
+      "epoch": 2.891745602165088,
+      "grad_norm": 1.2125190697405077,
+      "learning_rate": 3.9624105846319813e-08,
+      "loss": 0.3298,
+      "step": 4274
+    },
+    {
+      "epoch": 2.8924221921515563,
+      "grad_norm": 1.1198285600941194,
+      "learning_rate": 3.9130997392157756e-08,
+      "loss": 0.3079,
+      "step": 4275
+    },
+    {
+      "epoch": 2.8930987821380243,
+      "grad_norm": 1.1031892741599705,
+      "learning_rate": 3.86409644093827e-08,
+      "loss": 0.2921,
+      "step": 4276
+    },
+    {
+      "epoch": 2.8937753721244928,
+      "grad_norm": 1.1909455252020806,
+      "learning_rate": 3.8154007201787194e-08,
+      "loss": 0.3146,
+      "step": 4277
+    },
+    {
+      "epoch": 2.894451962110961,
+      "grad_norm": 1.1639128358691913,
+      "learning_rate": 3.7670126071259194e-08,
+      "loss": 0.3294,
+      "step": 4278
+    },
+    {
+      "epoch": 2.895128552097429,
+      "grad_norm": 1.1423374219976659,
+      "learning_rate": 3.718932131777819e-08,
+      "loss": 0.3059,
+      "step": 4279
+    },
+    {
+      "epoch": 2.8958051420838973,
+      "grad_norm": 1.1339918716128774,
+      "learning_rate": 3.6711593239417976e-08,
+      "loss": 0.3018,
+      "step": 4280
+    },
+    {
+      "epoch": 2.8964817320703653,
+      "grad_norm": 1.144619379367415,
+      "learning_rate": 3.62369421323433e-08,
+      "loss": 0.3066,
+      "step": 4281
+    },
+    {
+      "epoch": 2.8971583220568338,
+      "grad_norm": 1.157453603164695,
+      "learning_rate": 3.576536829081323e-08,
+      "loss": 0.3232,
+      "step": 4282
+    },
+    {
+      "epoch": 2.897834912043302,
+      "grad_norm": 1.1663898548339142,
+      "learning_rate": 3.52968720071778e-08,
+      "loss": 0.3121,
+      "step": 4283
+    },
+    {
+      "epoch": 2.89851150202977,
+      "grad_norm": 1.151140435547747,
+      "learning_rate": 3.483145357187967e-08,
+      "loss": 0.3109,
+      "step": 4284
+    },
+    {
+      "epoch": 2.8991880920162383,
+      "grad_norm": 1.1582013388566426,
+      "learning_rate": 3.436911327345305e-08,
+      "loss": 0.2984,
+      "step": 4285
+    },
+    {
+      "epoch": 2.8998646820027063,
+      "grad_norm": 1.1412644655425839,
+      "learning_rate": 3.3909851398523654e-08,
+      "loss": 0.308,
+      "step": 4286
+    },
+    {
+      "epoch": 2.9005412719891748,
+      "grad_norm": 1.1645084434101423,
+      "learning_rate": 3.345366823180929e-08,
+      "loss": 0.3206,
+      "step": 4287
+    },
+    {
+      "epoch": 2.901217861975643,
+      "grad_norm": 1.149063910594397,
+      "learning_rate": 3.300056405611873e-08,
+      "loss": 0.321,
+      "step": 4288
+    },
+    {
+      "epoch": 2.901894451962111,
+      "grad_norm": 1.1572085115744013,
+      "learning_rate": 3.2550539152352845e-08,
+      "loss": 0.3093,
+      "step": 4289
+    },
+    {
+      "epoch": 2.9025710419485793,
+      "grad_norm": 1.1516695356643691,
+      "learning_rate": 3.2103593799501786e-08,
+      "loss": 0.3087,
+      "step": 4290
+    },
+    {
+      "epoch": 2.9032476319350473,
+      "grad_norm": 1.1729018449936988,
+      "learning_rate": 3.165972827464892e-08,
+      "loss": 0.309,
+      "step": 4291
+    },
+    {
+      "epoch": 2.903924221921516,
+      "grad_norm": 1.123975545818059,
+      "learning_rate": 3.1218942852965226e-08,
+      "loss": 0.3037,
+      "step": 4292
+    },
+    {
+      "epoch": 2.904600811907984,
+      "grad_norm": 1.1420041050641399,
+      "learning_rate": 3.078123780771602e-08,
+      "loss": 0.3076,
+      "step": 4293
+    },
+    {
+      "epoch": 2.905277401894452,
+      "grad_norm": 1.186079651681517,
+      "learning_rate": 3.034661341025258e-08,
+      "loss": 0.321,
+      "step": 4294
+    },
+    {
+      "epoch": 2.9059539918809203,
+      "grad_norm": 1.1526191444306315,
+      "learning_rate": 2.9915069930019914e-08,
+      "loss": 0.3091,
+      "step": 4295
+    },
+    {
+      "epoch": 2.9066305818673883,
+      "grad_norm": 1.1565920873433775,
+      "learning_rate": 2.94866076345518e-08,
+      "loss": 0.3118,
+      "step": 4296
+    },
+    {
+      "epoch": 2.907307171853857,
+      "grad_norm": 1.1453368313546646,
+      "learning_rate": 2.9061226789471874e-08,
+      "loss": 0.314,
+      "step": 4297
+    },
+    {
+      "epoch": 2.907983761840325,
+      "grad_norm": 1.1340941402221867,
+      "learning_rate": 2.863892765849252e-08,
+      "loss": 0.3064,
+      "step": 4298
+    },
+    {
+      "epoch": 2.908660351826793,
+      "grad_norm": 1.171677798973917,
+      "learning_rate": 2.8219710503416543e-08,
+      "loss": 0.3268,
+      "step": 4299
+    },
+    {
+      "epoch": 2.9093369418132613,
+      "grad_norm": 1.1474207981438231,
+      "learning_rate": 2.78035755841366e-08,
+      "loss": 0.2955,
+      "step": 4300
+    },
+    {
+      "epoch": 2.9100135317997293,
+      "grad_norm": 1.1276142697099485,
+      "learning_rate": 2.7390523158633552e-08,
+      "loss": 0.3141,
+      "step": 4301
+    },
+    {
+      "epoch": 2.910690121786198,
+      "grad_norm": 1.134055949016168,
+      "learning_rate": 2.6980553482977566e-08,
+      "loss": 0.3136,
+      "step": 4302
+    },
+    {
+      "epoch": 2.911366711772666,
+      "grad_norm": 1.1643805303472392,
+      "learning_rate": 2.657366681132756e-08,
+      "loss": 0.319,
+      "step": 4303
+    },
+    {
+      "epoch": 2.912043301759134,
+      "grad_norm": 1.160709431692687,
+      "learning_rate": 2.6169863395932304e-08,
+      "loss": 0.3146,
+      "step": 4304
+    },
+    {
+      "epoch": 2.9127198917456023,
+      "grad_norm": 1.1510206817058268,
+      "learning_rate": 2.5769143487127113e-08,
+      "loss": 0.3083,
+      "step": 4305
+    },
+    {
+      "epoch": 2.9133964817320703,
+      "grad_norm": 1.1919954516471527,
+      "learning_rate": 2.5371507333337153e-08,
+      "loss": 0.3149,
+      "step": 4306
+    },
+    {
+      "epoch": 2.914073071718539,
+      "grad_norm": 1.166119466811266,
+      "learning_rate": 2.497695518107579e-08,
+      "loss": 0.3185,
+      "step": 4307
+    },
+    {
+      "epoch": 2.914749661705007,
+      "grad_norm": 1.1456687779806747,
+      "learning_rate": 2.4585487274942922e-08,
+      "loss": 0.3206,
+      "step": 4308
+    },
+    {
+      "epoch": 2.915426251691475,
+      "grad_norm": 1.1323663586397024,
+      "learning_rate": 2.4197103857628858e-08,
+      "loss": 0.3102,
+      "step": 4309
+    },
+    {
+      "epoch": 2.9161028416779433,
+      "grad_norm": 1.1033632191595115,
+      "learning_rate": 2.381180516990933e-08,
+      "loss": 0.3029,
+      "step": 4310
+    },
+    {
+      "epoch": 2.9167794316644113,
+      "grad_norm": 1.1557013695686562,
+      "learning_rate": 2.3429591450649934e-08,
+      "loss": 0.3131,
+      "step": 4311
+    },
+    {
+      "epoch": 2.91745602165088,
+      "grad_norm": 1.131307734017743,
+      "learning_rate": 2.305046293680113e-08,
+      "loss": 0.3084,
+      "step": 4312
+    },
+    {
+      "epoch": 2.918132611637348,
+      "grad_norm": 1.1290122102509739,
+      "learning_rate": 2.267441986340324e-08,
+      "loss": 0.3136,
+      "step": 4313
+    },
+    {
+      "epoch": 2.918809201623816,
+      "grad_norm": 1.1217149604078114,
+      "learning_rate": 2.230146246358256e-08,
+      "loss": 0.2991,
+      "step": 4314
+    },
+    {
+      "epoch": 2.9194857916102843,
+      "grad_norm": 1.15371748816963,
+      "learning_rate": 2.193159096855191e-08,
+      "loss": 0.3202,
+      "step": 4315
+    },
+    {
+      "epoch": 2.9201623815967523,
+      "grad_norm": 1.1426381301277648,
+      "learning_rate": 2.1564805607612317e-08,
+      "loss": 0.3074,
+      "step": 4316
+    },
+    {
+      "epoch": 2.920838971583221,
+      "grad_norm": 1.1490085281905362,
+      "learning_rate": 2.120110660815078e-08,
+      "loss": 0.3042,
+      "step": 4317
+    },
+    {
+      "epoch": 2.921515561569689,
+      "grad_norm": 1.166632241591695,
+      "learning_rate": 2.0840494195641382e-08,
+      "loss": 0.3216,
+      "step": 4318
+    },
+    {
+      "epoch": 2.922192151556157,
+      "grad_norm": 1.1147929426270626,
+      "learning_rate": 2.0482968593643625e-08,
+      "loss": 0.3073,
+      "step": 4319
+    },
+    {
+      "epoch": 2.9228687415426253,
+      "grad_norm": 1.1575877238634333,
+      "learning_rate": 2.012853002380466e-08,
+      "loss": 0.3004,
+      "step": 4320
+    },
+    {
+      "epoch": 2.9235453315290933,
+      "grad_norm": 1.1135864591516014,
+      "learning_rate": 1.97771787058576e-08,
+      "loss": 0.3066,
+      "step": 4321
+    },
+    {
+      "epoch": 2.924221921515562,
+      "grad_norm": 1.1322713277733414,
+      "learning_rate": 1.942891485762044e-08,
+      "loss": 0.3082,
+      "step": 4322
+    },
+    {
+      "epoch": 2.92489851150203,
+      "grad_norm": 1.1568277979539667,
+      "learning_rate": 1.9083738694998798e-08,
+      "loss": 0.3214,
+      "step": 4323
+    },
+    {
+      "epoch": 2.925575101488498,
+      "grad_norm": 1.0867872016253546,
+      "learning_rate": 1.8741650431982615e-08,
+      "loss": 0.2924,
+      "step": 4324
+    },
+    {
+      "epoch": 2.9262516914749663,
+      "grad_norm": 1.1458071757826465,
+      "learning_rate": 1.8402650280648916e-08,
+      "loss": 0.3112,
+      "step": 4325
+    },
+    {
+      "epoch": 2.9269282814614344,
+      "grad_norm": 1.0777452737503344,
+      "learning_rate": 1.8066738451159028e-08,
+      "loss": 0.2902,
+      "step": 4326
+    },
+    {
+      "epoch": 2.927604871447903,
+      "grad_norm": 1.1488526020025729,
+      "learning_rate": 1.773391515176026e-08,
+      "loss": 0.3093,
+      "step": 4327
+    },
+    {
+      "epoch": 2.928281461434371,
+      "grad_norm": 1.1562641729153602,
+      "learning_rate": 1.740418058878479e-08,
+      "loss": 0.3121,
+      "step": 4328
+    },
+    {
+      "epoch": 2.928958051420839,
+      "grad_norm": 1.1746849763786804,
+      "learning_rate": 1.7077534966650767e-08,
+      "loss": 0.3068,
+      "step": 4329
+    },
+    {
+      "epoch": 2.9296346414073073,
+      "grad_norm": 1.1733716856475764,
+      "learning_rate": 1.6753978487860645e-08,
+      "loss": 0.3254,
+      "step": 4330
+    },
+    {
+      "epoch": 2.9303112313937754,
+      "grad_norm": 1.1052256437749923,
+      "learning_rate": 1.6433511353002863e-08,
+      "loss": 0.29,
+      "step": 4331
+    },
+    {
+      "epoch": 2.930987821380244,
+      "grad_norm": 1.0968689832486906,
+      "learning_rate": 1.6116133760747944e-08,
+      "loss": 0.3048,
+      "step": 4332
+    },
+    {
+      "epoch": 2.931664411366712,
+      "grad_norm": 1.1139674432670768,
+      "learning_rate": 1.5801845907854606e-08,
+      "loss": 0.3056,
+      "step": 4333
+    },
+    {
+      "epoch": 2.93234100135318,
+      "grad_norm": 1.0980939854228233,
+      "learning_rate": 1.549064798916311e-08,
+      "loss": 0.3006,
+      "step": 4334
+    },
+    {
+      "epoch": 2.933017591339648,
+      "grad_norm": 1.1462567878563057,
+      "learning_rate": 1.5182540197600237e-08,
+      "loss": 0.3251,
+      "step": 4335
+    },
+    {
+      "epoch": 2.9336941813261164,
+      "grad_norm": 1.1233414413943121,
+      "learning_rate": 1.4877522724175974e-08,
+      "loss": 0.3155,
+      "step": 4336
+    },
+    {
+      "epoch": 2.934370771312585,
+      "grad_norm": 1.1694248394037319,
+      "learning_rate": 1.4575595757985172e-08,
+      "loss": 0.3128,
+      "step": 4337
+    },
+    {
+      "epoch": 2.935047361299053,
+      "grad_norm": 1.1554047473605722,
+      "learning_rate": 1.4276759486205328e-08,
+      "loss": 0.3214,
+      "step": 4338
+    },
+    {
+      "epoch": 2.935723951285521,
+      "grad_norm": 1.1722479445058565,
+      "learning_rate": 1.3981014094099354e-08,
+      "loss": 0.3129,
+      "step": 4339
+    },
+    {
+      "epoch": 2.936400541271989,
+      "grad_norm": 1.1309022463962741,
+      "learning_rate": 1.368835976501337e-08,
+      "loss": 0.308,
+      "step": 4340
+    },
+    {
+      "epoch": 2.9370771312584574,
+      "grad_norm": 1.1506270946384547,
+      "learning_rate": 1.3398796680377245e-08,
+      "loss": 0.3101,
+      "step": 4341
+    },
+    {
+      "epoch": 2.937753721244926,
+      "grad_norm": 1.1441809234601779,
+      "learning_rate": 1.3112325019704608e-08,
+      "loss": 0.3084,
+      "step": 4342
+    },
+    {
+      "epoch": 2.938430311231394,
+      "grad_norm": 1.129286698477764,
+      "learning_rate": 1.2828944960592837e-08,
+      "loss": 0.3036,
+      "step": 4343
+    },
+    {
+      "epoch": 2.939106901217862,
+      "grad_norm": 1.13276012436253,
+      "learning_rate": 1.2548656678721404e-08,
+      "loss": 0.3151,
+      "step": 4344
+    },
+    {
+      "epoch": 2.93978349120433,
+      "grad_norm": 1.1405521841051598,
+      "learning_rate": 1.2271460347854091e-08,
+      "loss": 0.3028,
+      "step": 4345
+    },
+    {
+      "epoch": 2.9404600811907984,
+      "grad_norm": 1.161498100591003,
+      "learning_rate": 1.1997356139838434e-08,
+      "loss": 0.3122,
+      "step": 4346
+    },
+    {
+      "epoch": 2.941136671177267,
+      "grad_norm": 1.1523731008191445,
+      "learning_rate": 1.1726344224603504e-08,
+      "loss": 0.2991,
+      "step": 4347
+    },
+    {
+      "epoch": 2.941813261163735,
+      "grad_norm": 1.1710962219756096,
+      "learning_rate": 1.145842477016268e-08,
+      "loss": 0.3295,
+      "step": 4348
+    },
+    {
+      "epoch": 2.942489851150203,
+      "grad_norm": 1.1348381341768248,
+      "learning_rate": 1.119359794261088e-08,
+      "loss": 0.3043,
+      "step": 4349
+    },
+    {
+      "epoch": 2.943166441136671,
+      "grad_norm": 1.119796205001435,
+      "learning_rate": 1.0931863906127327e-08,
+      "loss": 0.3114,
+      "step": 4350
+    },
+    {
+      "epoch": 2.9438430311231394,
+      "grad_norm": 1.1489488407168515,
+      "learning_rate": 1.0673222822972229e-08,
+      "loss": 0.3168,
+      "step": 4351
+    },
+    {
+      "epoch": 2.944519621109608,
+      "grad_norm": 1.110150794084272,
+      "learning_rate": 1.0417674853489545e-08,
+      "loss": 0.2983,
+      "step": 4352
+    },
+    {
+      "epoch": 2.945196211096076,
+      "grad_norm": 1.1683257612732658,
+      "learning_rate": 1.0165220156105326e-08,
+      "loss": 0.3234,
+      "step": 4353
+    },
+    {
+      "epoch": 2.945872801082544,
+      "grad_norm": 1.109813392704684,
+      "learning_rate": 9.915858887327157e-09,
+      "loss": 0.2923,
+      "step": 4354
+    },
+    {
+      "epoch": 2.946549391069012,
+      "grad_norm": 1.196883300808739,
+      "learning_rate": 9.669591201746375e-09,
+      "loss": 0.3164,
+      "step": 4355
+    },
+    {
+      "epoch": 2.9472259810554804,
+      "grad_norm": 1.1456394101942902,
+      "learning_rate": 9.426417252035858e-09,
+      "loss": 0.3086,
+      "step": 4356
+    },
+    {
+      "epoch": 2.9479025710419484,
+      "grad_norm": 1.1331249915686332,
+      "learning_rate": 9.186337188949456e-09,
+      "loss": 0.307,
+      "step": 4357
+    },
+    {
+      "epoch": 2.948579161028417,
+      "grad_norm": 1.2310369862585375,
+      "learning_rate": 8.949351161324227e-09,
+      "loss": 0.3324,
+      "step": 4358
+    },
+    {
+      "epoch": 2.949255751014885,
+      "grad_norm": 1.1533518432060552,
+      "learning_rate": 8.715459316078756e-09,
+      "loss": 0.3172,
+      "step": 4359
+    },
+    {
+      "epoch": 2.949932341001353,
+      "grad_norm": 1.1373796418007207,
+      "learning_rate": 8.484661798213723e-09,
+      "loss": 0.3084,
+      "step": 4360
+    },
+    {
+      "epoch": 2.9506089309878214,
+      "grad_norm": 1.117765115517499,
+      "learning_rate": 8.256958750810784e-09,
+      "loss": 0.2974,
+      "step": 4361
+    },
+    {
+      "epoch": 2.9512855209742894,
+      "grad_norm": 1.156417502178171,
+      "learning_rate": 8.032350315033688e-09,
+      "loss": 0.3067,
+      "step": 4362
+    },
+    {
+      "epoch": 2.951962110960758,
+      "grad_norm": 1.1304037754137621,
+      "learning_rate": 7.810836630127717e-09,
+      "loss": 0.307,
+      "step": 4363
+    },
+    {
+      "epoch": 2.952638700947226,
+      "grad_norm": 1.138766354028754,
+      "learning_rate": 7.59241783341913e-09,
+      "loss": 0.3139,
+      "step": 4364
+    },
+    {
+      "epoch": 2.953315290933694,
+      "grad_norm": 1.1439120799276494,
+      "learning_rate": 7.377094060315726e-09,
+      "loss": 0.3162,
+      "step": 4365
+    },
+    {
+      "epoch": 2.9539918809201624,
+      "grad_norm": 1.1107857012151061,
+      "learning_rate": 7.164865444306834e-09,
+      "loss": 0.2997,
+      "step": 4366
+    },
+    {
+      "epoch": 2.9546684709066304,
+      "grad_norm": 1.1851367102888233,
+      "learning_rate": 6.9557321169622105e-09,
+      "loss": 0.3235,
+      "step": 4367
+    },
+    {
+      "epoch": 2.955345060893099,
+      "grad_norm": 1.1398386073520308,
+      "learning_rate": 6.7496942079342546e-09,
+      "loss": 0.3102,
+      "step": 4368
+    },
+    {
+      "epoch": 2.956021650879567,
+      "grad_norm": 1.1828735338696095,
+      "learning_rate": 6.546751844955235e-09,
+      "loss": 0.3164,
+      "step": 4369
+    },
+    {
+      "epoch": 2.956698240866035,
+      "grad_norm": 1.1264165638073909,
+      "learning_rate": 6.346905153837291e-09,
+      "loss": 0.3194,
+      "step": 4370
+    },
+    {
+      "epoch": 2.9573748308525034,
+      "grad_norm": 1.116024016414339,
+      "learning_rate": 6.150154258476315e-09,
+      "loss": 0.2981,
+      "step": 4371
+    },
+    {
+      "epoch": 2.9580514208389714,
+      "grad_norm": 1.0984584846507182,
+      "learning_rate": 5.956499280845851e-09,
+      "loss": 0.3011,
+      "step": 4372
+    },
+    {
+      "epoch": 2.95872801082544,
+      "grad_norm": 1.1119161048589243,
+      "learning_rate": 5.765940341002085e-09,
+      "loss": 0.3099,
+      "step": 4373
+    },
+    {
+      "epoch": 2.959404600811908,
+      "grad_norm": 1.1637728366343703,
+      "learning_rate": 5.578477557081074e-09,
+      "loss": 0.3202,
+      "step": 4374
+    },
+    {
+      "epoch": 2.960081190798376,
+      "grad_norm": 1.1234941277175854,
+      "learning_rate": 5.394111045299855e-09,
+      "loss": 0.3018,
+      "step": 4375
+    },
+    {
+      "epoch": 2.9607577807848444,
+      "grad_norm": 1.1741162886088914,
+      "learning_rate": 5.212840919955886e-09,
+      "loss": 0.3299,
+      "step": 4376
+    },
+    {
+      "epoch": 2.9614343707713124,
+      "grad_norm": 1.1440273787574575,
+      "learning_rate": 5.034667293427053e-09,
+      "loss": 0.3133,
+      "step": 4377
+    },
+    {
+      "epoch": 2.962110960757781,
+      "grad_norm": 1.124868236628393,
+      "learning_rate": 4.859590276170556e-09,
+      "loss": 0.3146,
+      "step": 4378
+    },
+    {
+      "epoch": 2.962787550744249,
+      "grad_norm": 1.1031333776064192,
+      "learning_rate": 4.687609976725127e-09,
+      "loss": 0.2942,
+      "step": 4379
+    },
+    {
+      "epoch": 2.963464140730717,
+      "grad_norm": 1.1387413183402901,
+      "learning_rate": 4.51872650170937e-09,
+      "loss": 0.3078,
+      "step": 4380
+    },
+    {
+      "epoch": 2.9641407307171854,
+      "grad_norm": 1.0601478577033863,
+      "learning_rate": 4.352939955822311e-09,
+      "loss": 0.2847,
+      "step": 4381
+    },
+    {
+      "epoch": 2.9648173207036534,
+      "grad_norm": 1.1175635320768091,
+      "learning_rate": 4.190250441841737e-09,
+      "loss": 0.3004,
+      "step": 4382
+    },
+    {
+      "epoch": 2.965493910690122,
+      "grad_norm": 1.15277509463722,
+      "learning_rate": 4.030658060626969e-09,
+      "loss": 0.3106,
+      "step": 4383
+    },
+    {
+      "epoch": 2.96617050067659,
+      "grad_norm": 1.1296122280853835,
+      "learning_rate": 3.874162911117196e-09,
+      "loss": 0.3107,
+      "step": 4384
+    },
+    {
+      "epoch": 2.966847090663058,
+      "grad_norm": 1.1385531535031743,
+      "learning_rate": 3.7207650903298143e-09,
+      "loss": 0.2958,
+      "step": 4385
+    },
+    {
+      "epoch": 2.9675236806495264,
+      "grad_norm": 1.1876739786980457,
+      "learning_rate": 3.570464693364306e-09,
+      "loss": 0.3181,
+      "step": 4386
+    },
+    {
+      "epoch": 2.9682002706359945,
+      "grad_norm": 1.1089642461299507,
+      "learning_rate": 3.4232618133978044e-09,
+      "loss": 0.2913,
+      "step": 4387
+    },
+    {
+      "epoch": 2.968876860622463,
+      "grad_norm": 1.1483076805904084,
+      "learning_rate": 3.279156541688422e-09,
+      "loss": 0.313,
+      "step": 4388
+    },
+    {
+      "epoch": 2.969553450608931,
+      "grad_norm": 1.0951418894610812,
+      "learning_rate": 3.1381489675746946e-09,
+      "loss": 0.3035,
+      "step": 4389
+    },
+    {
+      "epoch": 2.970230040595399,
+      "grad_norm": 1.1389919207006642,
+      "learning_rate": 3.000239178472253e-09,
+      "loss": 0.3259,
+      "step": 4390
+    },
+    {
+      "epoch": 2.9709066305818674,
+      "grad_norm": 1.1079950747464211,
+      "learning_rate": 2.8654272598788167e-09,
+      "loss": 0.3055,
+      "step": 4391
+    },
+    {
+      "epoch": 2.9715832205683355,
+      "grad_norm": 1.1359130773351196,
+      "learning_rate": 2.7337132953697555e-09,
+      "loss": 0.3095,
+      "step": 4392
+    },
+    {
+      "epoch": 2.972259810554804,
+      "grad_norm": 1.1636770999387485,
+      "learning_rate": 2.605097366601417e-09,
+      "loss": 0.3192,
+      "step": 4393
+    },
+    {
+      "epoch": 2.972936400541272,
+      "grad_norm": 1.1317737119272322,
+      "learning_rate": 2.479579553307798e-09,
+      "loss": 0.31,
+      "step": 4394
+    },
+    {
+      "epoch": 2.97361299052774,
+      "grad_norm": 1.1862242556681604,
+      "learning_rate": 2.3571599333038765e-09,
+      "loss": 0.3254,
+      "step": 4395
+    },
+    {
+      "epoch": 2.9742895805142084,
+      "grad_norm": 1.1509284888371165,
+      "learning_rate": 2.237838582483387e-09,
+      "loss": 0.3043,
+      "step": 4396
+    },
+    {
+      "epoch": 2.9749661705006765,
+      "grad_norm": 1.1426201542871595,
+      "learning_rate": 2.12161557481827e-09,
+      "loss": 0.3213,
+      "step": 4397
+    },
+    {
+      "epoch": 2.975642760487145,
+      "grad_norm": 1.1252494999790472,
+      "learning_rate": 2.008490982360889e-09,
+      "loss": 0.2965,
+      "step": 4398
+    },
+    {
+      "epoch": 2.976319350473613,
+      "grad_norm": 1.1204951487459502,
+      "learning_rate": 1.8984648752429222e-09,
+      "loss": 0.2978,
+      "step": 4399
+    },
+    {
+      "epoch": 2.976995940460081,
+      "grad_norm": 1.1695668395379417,
+      "learning_rate": 1.7915373216742527e-09,
+      "loss": 0.3115,
+      "step": 4400
+    },
+    {
+      "epoch": 2.9776725304465494,
+      "grad_norm": 1.119696402677367,
+      "learning_rate": 1.687708387944076e-09,
+      "loss": 0.2968,
+      "step": 4401
+    },
+    {
+      "epoch": 2.9783491204330175,
+      "grad_norm": 1.1047732478283987,
+      "learning_rate": 1.5869781384203475e-09,
+      "loss": 0.3021,
+      "step": 4402
+    },
+    {
+      "epoch": 2.979025710419486,
+      "grad_norm": 1.1364559701770647,
+      "learning_rate": 1.4893466355514474e-09,
+      "loss": 0.3064,
+      "step": 4403
+    },
+    {
+      "epoch": 2.979702300405954,
+      "grad_norm": 1.1148899633374867,
+      "learning_rate": 1.3948139398628492e-09,
+      "loss": 0.3093,
+      "step": 4404
+    },
+    {
+      "epoch": 2.980378890392422,
+      "grad_norm": 1.12969911338335,
+      "learning_rate": 1.3033801099598954e-09,
+      "loss": 0.3005,
+      "step": 4405
+    },
+    {
+      "epoch": 2.9810554803788905,
+      "grad_norm": 1.1212118242990712,
+      "learning_rate": 1.215045202527243e-09,
+      "loss": 0.3044,
+      "step": 4406
+    },
+    {
+      "epoch": 2.9817320703653585,
+      "grad_norm": 1.117365484245686,
+      "learning_rate": 1.1298092723266429e-09,
+      "loss": 0.3001,
+      "step": 4407
+    },
+    {
+      "epoch": 2.982408660351827,
+      "grad_norm": 1.0904502565464471,
+      "learning_rate": 1.0476723722002702e-09,
+      "loss": 0.3028,
+      "step": 4408
+    },
+    {
+      "epoch": 2.983085250338295,
+      "grad_norm": 1.145864536129481,
+      "learning_rate": 9.686345530690589e-10,
+      "loss": 0.3101,
+      "step": 4409
+    },
+    {
+      "epoch": 2.983761840324763,
+      "grad_norm": 1.116492308235731,
+      "learning_rate": 8.926958639315919e-10,
+      "loss": 0.3075,
+      "step": 4410
+    },
+    {
+      "epoch": 2.9844384303112315,
+      "grad_norm": 1.15049666392336,
+      "learning_rate": 8.198563518657665e-10,
+      "loss": 0.3058,
+      "step": 4411
+    },
+    {
+      "epoch": 2.9851150202976995,
+      "grad_norm": 1.0942773671352757,
+      "learning_rate": 7.50116062028794e-10,
+      "loss": 0.2933,
+      "step": 4412
+    },
+    {
+      "epoch": 2.985791610284168,
+      "grad_norm": 1.1922095560073203,
+      "learning_rate": 6.834750376549793e-10,
+      "loss": 0.3336,
+      "step": 4413
+    },
+    {
+      "epoch": 2.986468200270636,
+      "grad_norm": 1.1380529184851402,
+      "learning_rate": 6.199333200590519e-10,
+      "loss": 0.3169,
+      "step": 4414
+    },
+    {
+      "epoch": 2.987144790257104,
+      "grad_norm": 1.1053295877919185,
+      "learning_rate": 5.594909486328348e-10,
+      "loss": 0.2981,
+      "step": 4415
+    },
+    {
+      "epoch": 2.9878213802435725,
+      "grad_norm": 1.1197045497023874,
+      "learning_rate": 5.021479608474655e-10,
+      "loss": 0.3141,
+      "step": 4416
+    },
+    {
+      "epoch": 2.9884979702300405,
+      "grad_norm": 1.1186182970793708,
+      "learning_rate": 4.4790439225284034e-10,
+      "loss": 0.3034,
+      "step": 4417
+    },
+    {
+      "epoch": 2.989174560216509,
+      "grad_norm": 1.1731055612079,
+      "learning_rate": 3.967602764770595e-10,
+      "loss": 0.3054,
+      "step": 4418
+    },
+    {
+      "epoch": 2.989851150202977,
+      "grad_norm": 1.121528681282999,
+      "learning_rate": 3.487156452258722e-10,
+      "loss": 0.2953,
+      "step": 4419
+    },
+    {
+      "epoch": 2.990527740189445,
+      "grad_norm": 1.1943003012372644,
+      "learning_rate": 3.0377052828489684e-10,
+      "loss": 0.3188,
+      "step": 4420
+    },
+    {
+      "epoch": 2.9912043301759135,
+      "grad_norm": 1.1217151073214495,
+      "learning_rate": 2.6192495351795576e-10,
+      "loss": 0.2999,
+      "step": 4421
+    },
+    {
+      "epoch": 2.9918809201623815,
+      "grad_norm": 1.1422225996726174,
+      "learning_rate": 2.231789468670753e-10,
+      "loss": 0.2982,
+      "step": 4422
+    },
+    {
+      "epoch": 2.99255751014885,
+      "grad_norm": 1.1261024883457165,
+      "learning_rate": 1.8753253235248568e-10,
+      "loss": 0.2966,
+      "step": 4423
+    },
+    {
+      "epoch": 2.993234100135318,
+      "grad_norm": 1.175016815498028,
+      "learning_rate": 1.5498573207262112e-10,
+      "loss": 0.3207,
+      "step": 4424
+    },
+    {
+      "epoch": 2.993910690121786,
+      "grad_norm": 1.142588901617758,
+      "learning_rate": 1.2553856620522997e-10,
+      "loss": 0.3187,
+      "step": 4425
+    },
+    {
+      "epoch": 2.9945872801082545,
+      "grad_norm": 1.1462723401963262,
+      "learning_rate": 9.919105300570941e-11,
+      "loss": 0.3166,
+      "step": 4426
+    },
+    {
+      "epoch": 2.9952638700947225,
+      "grad_norm": 1.134317040194697,
+      "learning_rate": 7.59432088082157e-11,
+      "loss": 0.3095,
+      "step": 4427
+    },
+    {
+      "epoch": 2.995940460081191,
+      "grad_norm": 1.1462902638304833,
+      "learning_rate": 5.579504802566416e-11,
+      "loss": 0.3164,
+      "step": 4428
+    },
+    {
+      "epoch": 2.996617050067659,
+      "grad_norm": 1.1741380922743976,
+      "learning_rate": 3.8746583148063786e-11,
+      "loss": 0.3096,
+      "step": 4429
+    },
+    {
+      "epoch": 2.997293640054127,
+      "grad_norm": 1.1566532627269621,
+      "learning_rate": 2.4797824744737797e-11,
+      "loss": 0.3142,
+      "step": 4430
+    },
+    {
+      "epoch": 2.9979702300405955,
+      "grad_norm": 1.185825918050989,
+      "learning_rate": 1.3948781463213324e-11,
+      "loss": 0.3286,
+      "step": 4431
+    },
+    {
+      "epoch": 2.9986468200270635,
+      "grad_norm": 1.1129394991701362,
+      "learning_rate": 6.199460029221449e-12,
+      "loss": 0.2988,
+      "step": 4432
+    },
+    {
+      "epoch": 2.999323410013532,
+      "grad_norm": 1.1564904886316638,
+      "learning_rate": 1.549865247807425e-12,
+      "loss": 0.3123,
+      "step": 4433
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.068601292043698,
+      "learning_rate": 0.0,
+      "loss": 0.2835,
+      "step": 4434
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.39919164776802063,
+      "eval_runtime": 431.5509,
+      "eval_samples_per_second": 23.068,
+      "eval_steps_per_second": 0.723,
+      "step": 4434
+    },
+    {
+      "epoch": 3.0,
+      "step": 4434,
+      "total_flos": 1.6090113089910866e+18,
+      "train_loss": 0.69218219782406,
+      "train_runtime": 92434.0859,
+      "train_samples_per_second": 6.138,
+      "train_steps_per_second": 0.048
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4434,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.6090113089910866e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}