diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,43783 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5,
+  "eval_steps": 500,
+  "global_step": 6250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 8e-05,
+      "grad_norm": 8.342381477355957,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.704,
+      "step": 1
+    },
+    {
+      "epoch": 0.00016,
+      "grad_norm": 13.919276237487793,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.5506,
+      "step": 2
+    },
+    {
+      "epoch": 0.00024,
+      "grad_norm": 9.644050598144531,
+      "learning_rate": 3e-06,
+      "loss": 0.5138,
+      "step": 3
+    },
+    {
+      "epoch": 0.00032,
+      "grad_norm": 9.632057189941406,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.7317,
+      "step": 4
+    },
+    {
+      "epoch": 0.0004,
+      "grad_norm": 8.935699462890625,
+      "learning_rate": 5e-06,
+      "loss": 0.7372,
+      "step": 5
+    },
+    {
+      "epoch": 0.00048,
+      "grad_norm": 9.070698738098145,
+      "learning_rate": 6e-06,
+      "loss": 0.7517,
+      "step": 6
+    },
+    {
+      "epoch": 0.00056,
+      "grad_norm": 9.226506233215332,
+      "learning_rate": 7e-06,
+      "loss": 0.6523,
+      "step": 7
+    },
+    {
+      "epoch": 0.00064,
+      "grad_norm": 11.089458465576172,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.6743,
+      "step": 8
+    },
+    {
+      "epoch": 0.00072,
+      "grad_norm": 7.038223743438721,
+      "learning_rate": 9e-06,
+      "loss": 0.6536,
+      "step": 9
+    },
+    {
+      "epoch": 0.0008,
+      "grad_norm": 12.974882125854492,
+      "learning_rate": 1e-05,
+      "loss": 0.6122,
+      "step": 10
+    },
+    {
+      "epoch": 0.00088,
+      "grad_norm": 8.054683685302734,
+      "learning_rate": 9.999999841833366e-06,
+      "loss": 0.5658,
+      "step": 11
+    },
+    {
+      "epoch": 0.00096,
+      "grad_norm": 15.936073303222656,
+      "learning_rate": 9.99999936733347e-06,
+      "loss": 0.8491,
+      "step": 12
+    },
+    {
+      "epoch": 0.00104,
+      "grad_norm": 4.686435699462891,
+      "learning_rate": 9.999998576500346e-06,
+      "loss": 0.4673,
+      "step": 13
+    },
+    {
+      "epoch": 0.00112,
+      "grad_norm": 9.04387092590332,
+      "learning_rate": 9.999997469334042e-06,
+      "loss": 0.6238,
+      "step": 14
+    },
+    {
+      "epoch": 0.0012,
+      "grad_norm": 10.498541831970215,
+      "learning_rate": 9.999996045834626e-06,
+      "loss": 0.6662,
+      "step": 15
+    },
+    {
+      "epoch": 0.00128,
+      "grad_norm": 5.166777610778809,
+      "learning_rate": 9.999994306002192e-06,
+      "loss": 0.5575,
+      "step": 16
+    },
+    {
+      "epoch": 0.00136,
+      "grad_norm": 7.151340484619141,
+      "learning_rate": 9.99999224983685e-06,
+      "loss": 0.5892,
+      "step": 17
+    },
+    {
+      "epoch": 0.00144,
+      "grad_norm": 5.04518985748291,
+      "learning_rate": 9.999989877338725e-06,
+      "loss": 0.6417,
+      "step": 18
+    },
+    {
+      "epoch": 0.00152,
+      "grad_norm": 5.931295394897461,
+      "learning_rate": 9.999987188507972e-06,
+      "loss": 0.4939,
+      "step": 19
+    },
+    {
+      "epoch": 0.0016,
+      "grad_norm": 5.467217445373535,
+      "learning_rate": 9.99998418334476e-06,
+      "loss": 0.6131,
+      "step": 20
+    },
+    {
+      "epoch": 0.00168,
+      "grad_norm": 5.837316989898682,
+      "learning_rate": 9.999980861849277e-06,
+      "loss": 0.5316,
+      "step": 21
+    },
+    {
+      "epoch": 0.00176,
+      "grad_norm": 4.237757205963135,
+      "learning_rate": 9.999977224021735e-06,
+      "loss": 0.5499,
+      "step": 22
+    },
+    {
+      "epoch": 0.00184,
+      "grad_norm": 6.357263565063477,
+      "learning_rate": 9.999973269862366e-06,
+      "loss": 0.5405,
+      "step": 23
+    },
+    {
+      "epoch": 0.00192,
+      "grad_norm": 6.0218939781188965,
+      "learning_rate": 9.999968999371416e-06,
+      "loss": 0.543,
+      "step": 24
+    },
+    {
+      "epoch": 0.002,
+      "grad_norm": 5.242684841156006,
+      "learning_rate": 9.99996441254916e-06,
+      "loss": 0.7568,
+      "step": 25
+    },
+    {
+      "epoch": 0.00208,
+      "grad_norm": 5.868856906890869,
+      "learning_rate": 9.999959509395884e-06,
+      "loss": 0.5211,
+      "step": 26
+    },
+    {
+      "epoch": 0.00216,
+      "grad_norm": 5.401463985443115,
+      "learning_rate": 9.9999542899119e-06,
+      "loss": 0.4879,
+      "step": 27
+    },
+    {
+      "epoch": 0.00224,
+      "grad_norm": 4.193491458892822,
+      "learning_rate": 9.999948754097538e-06,
+      "loss": 0.409,
+      "step": 28
+    },
+    {
+      "epoch": 0.00232,
+      "grad_norm": 8.131255149841309,
+      "learning_rate": 9.999942901953148e-06,
+      "loss": 0.5406,
+      "step": 29
+    },
+    {
+      "epoch": 0.0024,
+      "grad_norm": 8.522489547729492,
+      "learning_rate": 9.9999367334791e-06,
+      "loss": 0.4171,
+      "step": 30
+    },
+    {
+      "epoch": 0.00248,
+      "grad_norm": 5.474879264831543,
+      "learning_rate": 9.999930248675784e-06,
+      "loss": 0.3518,
+      "step": 31
+    },
+    {
+      "epoch": 0.00256,
+      "grad_norm": 3.1692543029785156,
+      "learning_rate": 9.999923447543614e-06,
+      "loss": 0.3889,
+      "step": 32
+    },
+    {
+      "epoch": 0.00264,
+      "grad_norm": 5.542096138000488,
+      "learning_rate": 9.999916330083015e-06,
+      "loss": 0.4787,
+      "step": 33
+    },
+    {
+      "epoch": 0.00272,
+      "grad_norm": 4.017666339874268,
+      "learning_rate": 9.99990889629444e-06,
+      "loss": 0.4756,
+      "step": 34
+    },
+    {
+      "epoch": 0.0028,
+      "grad_norm": 4.048304080963135,
+      "learning_rate": 9.99990114617836e-06,
+      "loss": 0.4496,
+      "step": 35
+    },
+    {
+      "epoch": 0.00288,
+      "grad_norm": 9.737591743469238,
+      "learning_rate": 9.999893079735262e-06,
+      "loss": 0.5132,
+      "step": 36
+    },
+    {
+      "epoch": 0.00296,
+      "grad_norm": 5.1411051750183105,
+      "learning_rate": 9.99988469696566e-06,
+      "loss": 0.4311,
+      "step": 37
+    },
+    {
+      "epoch": 0.00304,
+      "grad_norm": 7.761783599853516,
+      "learning_rate": 9.999875997870081e-06,
+      "loss": 0.6614,
+      "step": 38
+    },
+    {
+      "epoch": 0.00312,
+      "grad_norm": 2.5379042625427246,
+      "learning_rate": 9.99986698244908e-06,
+      "loss": 0.4899,
+      "step": 39
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 5.286753177642822,
+      "learning_rate": 9.999857650703224e-06,
+      "loss": 0.4297,
+      "step": 40
+    },
+    {
+      "epoch": 0.00328,
+      "grad_norm": 4.113813400268555,
+      "learning_rate": 9.999848002633102e-06,
+      "loss": 0.3484,
+      "step": 41
+    },
+    {
+      "epoch": 0.00336,
+      "grad_norm": 4.206110954284668,
+      "learning_rate": 9.999838038239327e-06,
+      "loss": 0.5235,
+      "step": 42
+    },
+    {
+      "epoch": 0.00344,
+      "grad_norm": 8.220080375671387,
+      "learning_rate": 9.999827757522531e-06,
+      "loss": 0.5051,
+      "step": 43
+    },
+    {
+      "epoch": 0.00352,
+      "grad_norm": 3.16642427444458,
+      "learning_rate": 9.99981716048336e-06,
+      "loss": 0.3753,
+      "step": 44
+    },
+    {
+      "epoch": 0.0036,
+      "grad_norm": 2.503070592880249,
+      "learning_rate": 9.999806247122488e-06,
+      "loss": 0.453,
+      "step": 45
+    },
+    {
+      "epoch": 0.00368,
+      "grad_norm": 4.834280967712402,
+      "learning_rate": 9.999795017440603e-06,
+      "loss": 0.3499,
+      "step": 46
+    },
+    {
+      "epoch": 0.00376,
+      "grad_norm": 2.557096004486084,
+      "learning_rate": 9.999783471438419e-06,
+      "loss": 0.3253,
+      "step": 47
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 3.6029672622680664,
+      "learning_rate": 9.999771609116662e-06,
+      "loss": 0.4115,
+      "step": 48
+    },
+    {
+      "epoch": 0.00392,
+      "grad_norm": 3.1142241954803467,
+      "learning_rate": 9.999759430476084e-06,
+      "loss": 0.5014,
+      "step": 49
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 3.7333784103393555,
+      "learning_rate": 9.999746935517457e-06,
+      "loss": 0.3711,
+      "step": 50
+    },
+    {
+      "epoch": 0.00408,
+      "grad_norm": 3.2080132961273193,
+      "learning_rate": 9.999734124241571e-06,
+      "loss": 0.3347,
+      "step": 51
+    },
+    {
+      "epoch": 0.00416,
+      "grad_norm": 3.805007219314575,
+      "learning_rate": 9.999720996649235e-06,
+      "loss": 0.7125,
+      "step": 52
+    },
+    {
+      "epoch": 0.00424,
+      "grad_norm": 3.14184308052063,
+      "learning_rate": 9.999707552741283e-06,
+      "loss": 0.5121,
+      "step": 53
+    },
+    {
+      "epoch": 0.00432,
+      "grad_norm": 3.540736436843872,
+      "learning_rate": 9.999693792518562e-06,
+      "loss": 0.4294,
+      "step": 54
+    },
+    {
+      "epoch": 0.0044,
+      "grad_norm": 3.0372300148010254,
+      "learning_rate": 9.999679715981942e-06,
+      "loss": 0.4628,
+      "step": 55
+    },
+    {
+      "epoch": 0.00448,
+      "grad_norm": 2.61629056930542,
+      "learning_rate": 9.999665323132317e-06,
+      "loss": 0.4936,
+      "step": 56
+    },
+    {
+      "epoch": 0.00456,
+      "grad_norm": 3.4021859169006348,
+      "learning_rate": 9.999650613970597e-06,
+      "loss": 0.539,
+      "step": 57
+    },
+    {
+      "epoch": 0.00464,
+      "grad_norm": 1.9579095840454102,
+      "learning_rate": 9.99963558849771e-06,
+      "loss": 0.3475,
+      "step": 58
+    },
+    {
+      "epoch": 0.00472,
+      "grad_norm": 3.136955976486206,
+      "learning_rate": 9.999620246714607e-06,
+      "loss": 0.389,
+      "step": 59
+    },
+    {
+      "epoch": 0.0048,
+      "grad_norm": 3.5431883335113525,
+      "learning_rate": 9.999604588622263e-06,
+      "loss": 0.4231,
+      "step": 60
+    },
+    {
+      "epoch": 0.00488,
+      "grad_norm": 4.076716899871826,
+      "learning_rate": 9.999588614221663e-06,
+      "loss": 0.4758,
+      "step": 61
+    },
+    {
+      "epoch": 0.00496,
+      "grad_norm": 2.444188117980957,
+      "learning_rate": 9.99957232351382e-06,
+      "loss": 0.4112,
+      "step": 62
+    },
+    {
+      "epoch": 0.00504,
+      "grad_norm": 8.321596145629883,
+      "learning_rate": 9.999555716499766e-06,
+      "loss": 0.4371,
+      "step": 63
+    },
+    {
+      "epoch": 0.00512,
+      "grad_norm": 2.7783851623535156,
+      "learning_rate": 9.99953879318055e-06,
+      "loss": 0.4137,
+      "step": 64
+    },
+    {
+      "epoch": 0.0052,
+      "grad_norm": 3.005558729171753,
+      "learning_rate": 9.999521553557243e-06,
+      "loss": 0.3781,
+      "step": 65
+    },
+    {
+      "epoch": 0.00528,
+      "grad_norm": 2.853501558303833,
+      "learning_rate": 9.999503997630934e-06,
+      "loss": 0.4312,
+      "step": 66
+    },
+    {
+      "epoch": 0.00536,
+      "grad_norm": 2.694535732269287,
+      "learning_rate": 9.999486125402738e-06,
+      "loss": 0.3815,
+      "step": 67
+    },
+    {
+      "epoch": 0.00544,
+      "grad_norm": 4.3195013999938965,
+      "learning_rate": 9.999467936873783e-06,
+      "loss": 0.4915,
+      "step": 68
+    },
+    {
+      "epoch": 0.00552,
+      "grad_norm": 2.009920358657837,
+      "learning_rate": 9.999449432045218e-06,
+      "loss": 0.4457,
+      "step": 69
+    },
+    {
+      "epoch": 0.0056,
+      "grad_norm": 2.0476291179656982,
+      "learning_rate": 9.999430610918217e-06,
+      "loss": 0.4017,
+      "step": 70
+    },
+    {
+      "epoch": 0.00568,
+      "grad_norm": 3.428873300552368,
+      "learning_rate": 9.99941147349397e-06,
+      "loss": 0.4943,
+      "step": 71
+    },
+    {
+      "epoch": 0.00576,
+      "grad_norm": 4.138948917388916,
+      "learning_rate": 9.999392019773685e-06,
+      "loss": 0.4353,
+      "step": 72
+    },
+    {
+      "epoch": 0.00584,
+      "grad_norm": 3.7678720951080322,
+      "learning_rate": 9.999372249758596e-06,
+      "loss": 0.3594,
+      "step": 73
+    },
+    {
+      "epoch": 0.00592,
+      "grad_norm": 2.1992523670196533,
+      "learning_rate": 9.999352163449954e-06,
+      "loss": 0.4057,
+      "step": 74
+    },
+    {
+      "epoch": 0.006,
+      "grad_norm": 3.9209282398223877,
+      "learning_rate": 9.999331760849028e-06,
+      "loss": 0.3694,
+      "step": 75
+    },
+    {
+      "epoch": 0.00608,
+      "grad_norm": 2.271817684173584,
+      "learning_rate": 9.999311041957109e-06,
+      "loss": 0.3515,
+      "step": 76
+    },
+    {
+      "epoch": 0.00616,
+      "grad_norm": 3.229722023010254,
+      "learning_rate": 9.999290006775507e-06,
+      "loss": 0.3017,
+      "step": 77
+    },
+    {
+      "epoch": 0.00624,
+      "grad_norm": 2.318446397781372,
+      "learning_rate": 9.999268655305556e-06,
+      "loss": 0.3885,
+      "step": 78
+    },
+    {
+      "epoch": 0.00632,
+      "grad_norm": 2.2932703495025635,
+      "learning_rate": 9.999246987548603e-06,
+      "loss": 0.3392,
+      "step": 79
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 2.111154794692993,
+      "learning_rate": 9.999225003506021e-06,
+      "loss": 0.3128,
+      "step": 80
+    },
+    {
+      "epoch": 0.00648,
+      "grad_norm": 4.121140003204346,
+      "learning_rate": 9.9992027031792e-06,
+      "loss": 0.4418,
+      "step": 81
+    },
+    {
+      "epoch": 0.00656,
+      "grad_norm": 1.6190462112426758,
+      "learning_rate": 9.999180086569553e-06,
+      "loss": 0.2905,
+      "step": 82
+    },
+    {
+      "epoch": 0.00664,
+      "grad_norm": 3.0115609169006348,
+      "learning_rate": 9.999157153678509e-06,
+      "loss": 0.3379,
+      "step": 83
+    },
+    {
+      "epoch": 0.00672,
+      "grad_norm": 3.357482671737671,
+      "learning_rate": 9.999133904507518e-06,
+      "loss": 0.4827,
+      "step": 84
+    },
+    {
+      "epoch": 0.0068,
+      "grad_norm": 2.7103617191314697,
+      "learning_rate": 9.99911033905805e-06,
+      "loss": 0.3324,
+      "step": 85
+    },
+    {
+      "epoch": 0.00688,
+      "grad_norm": 2.003753662109375,
+      "learning_rate": 9.999086457331603e-06,
+      "loss": 0.4019,
+      "step": 86
+    },
+    {
+      "epoch": 0.00696,
+      "grad_norm": 2.259552478790283,
+      "learning_rate": 9.999062259329679e-06,
+      "loss": 0.4511,
+      "step": 87
+    },
+    {
+      "epoch": 0.00704,
+      "grad_norm": 2.696129322052002,
+      "learning_rate": 9.999037745053814e-06,
+      "loss": 0.3631,
+      "step": 88
+    },
+    {
+      "epoch": 0.00712,
+      "grad_norm": 2.0996644496917725,
+      "learning_rate": 9.999012914505559e-06,
+      "loss": 0.4356,
+      "step": 89
+    },
+    {
+      "epoch": 0.0072,
+      "grad_norm": 1.694945216178894,
+      "learning_rate": 9.998987767686482e-06,
+      "loss": 0.3558,
+      "step": 90
+    },
+    {
+      "epoch": 0.00728,
+      "grad_norm": 4.748021602630615,
+      "learning_rate": 9.998962304598175e-06,
+      "loss": 0.403,
+      "step": 91
+    },
+    {
+      "epoch": 0.00736,
+      "grad_norm": 2.026329278945923,
+      "learning_rate": 9.998936525242251e-06,
+      "loss": 0.3913,
+      "step": 92
+    },
+    {
+      "epoch": 0.00744,
+      "grad_norm": 1.9248204231262207,
+      "learning_rate": 9.99891042962034e-06,
+      "loss": 0.2967,
+      "step": 93
+    },
+    {
+      "epoch": 0.00752,
+      "grad_norm": 2.36529278755188,
+      "learning_rate": 9.998884017734091e-06,
+      "loss": 0.4356,
+      "step": 94
+    },
+    {
+      "epoch": 0.0076,
+      "grad_norm": 2.3117401599884033,
+      "learning_rate": 9.998857289585177e-06,
+      "loss": 0.362,
+      "step": 95
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 2.4276554584503174,
+      "learning_rate": 9.998830245175288e-06,
+      "loss": 0.5628,
+      "step": 96
+    },
+    {
+      "epoch": 0.00776,
+      "grad_norm": 3.106940984725952,
+      "learning_rate": 9.998802884506136e-06,
+      "loss": 0.3412,
+      "step": 97
+    },
+    {
+      "epoch": 0.00784,
+      "grad_norm": 1.9471877813339233,
+      "learning_rate": 9.998775207579452e-06,
+      "loss": 0.506,
+      "step": 98
+    },
+    {
+      "epoch": 0.00792,
+      "grad_norm": 2.589988946914673,
+      "learning_rate": 9.998747214396987e-06,
+      "loss": 0.5048,
+      "step": 99
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 2.1943140029907227,
+      "learning_rate": 9.998718904960511e-06,
+      "loss": 0.3406,
+      "step": 100
+    },
+    {
+      "epoch": 0.00808,
+      "grad_norm": 2.6359894275665283,
+      "learning_rate": 9.998690279271815e-06,
+      "loss": 0.4702,
+      "step": 101
+    },
+    {
+      "epoch": 0.00816,
+      "grad_norm": 2.8097708225250244,
+      "learning_rate": 9.99866133733271e-06,
+      "loss": 0.2897,
+      "step": 102
+    },
+    {
+      "epoch": 0.00824,
+      "grad_norm": 1.7844816446304321,
+      "learning_rate": 9.99863207914503e-06,
+      "loss": 0.3935,
+      "step": 103
+    },
+    {
+      "epoch": 0.00832,
+      "grad_norm": 2.048438549041748,
+      "learning_rate": 9.998602504710623e-06,
+      "loss": 0.4605,
+      "step": 104
+    },
+    {
+      "epoch": 0.0084,
+      "grad_norm": 2.508671522140503,
+      "learning_rate": 9.99857261403136e-06,
+      "loss": 0.3009,
+      "step": 105
+    },
+    {
+      "epoch": 0.00848,
+      "grad_norm": 1.8178949356079102,
+      "learning_rate": 9.998542407109135e-06,
+      "loss": 0.4019,
+      "step": 106
+    },
+    {
+      "epoch": 0.00856,
+      "grad_norm": 4.293577671051025,
+      "learning_rate": 9.998511883945855e-06,
+      "loss": 0.3646,
+      "step": 107
+    },
+    {
+      "epoch": 0.00864,
+      "grad_norm": 2.485862970352173,
+      "learning_rate": 9.998481044543452e-06,
+      "loss": 0.4891,
+      "step": 108
+    },
+    {
+      "epoch": 0.00872,
+      "grad_norm": 2.0601084232330322,
+      "learning_rate": 9.998449888903881e-06,
+      "loss": 0.3481,
+      "step": 109
+    },
+    {
+      "epoch": 0.0088,
+      "grad_norm": 2.0768215656280518,
+      "learning_rate": 9.99841841702911e-06,
+      "loss": 0.3108,
+      "step": 110
+    },
+    {
+      "epoch": 0.00888,
+      "grad_norm": 1.9751044511795044,
+      "learning_rate": 9.99838662892113e-06,
+      "loss": 0.3349,
+      "step": 111
+    },
+    {
+      "epoch": 0.00896,
+      "grad_norm": 2.3447659015655518,
+      "learning_rate": 9.998354524581953e-06,
+      "loss": 0.4507,
+      "step": 112
+    },
+    {
+      "epoch": 0.00904,
+      "grad_norm": 2.0622479915618896,
+      "learning_rate": 9.998322104013609e-06,
+      "loss": 0.3494,
+      "step": 113
+    },
+    {
+      "epoch": 0.00912,
+      "grad_norm": 2.2754018306732178,
+      "learning_rate": 9.998289367218151e-06,
+      "loss": 0.3852,
+      "step": 114
+    },
+    {
+      "epoch": 0.0092,
+      "grad_norm": 2.0593910217285156,
+      "learning_rate": 9.998256314197648e-06,
+      "loss": 0.3999,
+      "step": 115
+    },
+    {
+      "epoch": 0.00928,
+      "grad_norm": 2.1019093990325928,
+      "learning_rate": 9.998222944954193e-06,
+      "loss": 0.4743,
+      "step": 116
+    },
+    {
+      "epoch": 0.00936,
+      "grad_norm": 1.9430328607559204,
+      "learning_rate": 9.998189259489897e-06,
+      "loss": 0.3751,
+      "step": 117
+    },
+    {
+      "epoch": 0.00944,
+      "grad_norm": 1.9289934635162354,
+      "learning_rate": 9.99815525780689e-06,
+      "loss": 0.3586,
+      "step": 118
+    },
+    {
+      "epoch": 0.00952,
+      "grad_norm": 2.437016248703003,
+      "learning_rate": 9.998120939907323e-06,
+      "loss": 0.4204,
+      "step": 119
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 2.8940718173980713,
+      "learning_rate": 9.998086305793368e-06,
+      "loss": 0.379,
+      "step": 120
+    },
+    {
+      "epoch": 0.00968,
+      "grad_norm": 2.258979558944702,
+      "learning_rate": 9.998051355467215e-06,
+      "loss": 0.4011,
+      "step": 121
+    },
+    {
+      "epoch": 0.00976,
+      "grad_norm": 1.7911226749420166,
+      "learning_rate": 9.99801608893108e-06,
+      "loss": 0.3489,
+      "step": 122
+    },
+    {
+      "epoch": 0.00984,
+      "grad_norm": 3.431490659713745,
+      "learning_rate": 9.997980506187188e-06,
+      "loss": 0.4482,
+      "step": 123
+    },
+    {
+      "epoch": 0.00992,
+      "grad_norm": 2.332395315170288,
+      "learning_rate": 9.997944607237791e-06,
+      "loss": 0.3858,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.693103790283203,
+      "learning_rate": 9.997908392085164e-06,
+      "loss": 0.4227,
+      "step": 125
+    },
+    {
+      "epoch": 0.01008,
+      "grad_norm": 2.643721580505371,
+      "learning_rate": 9.997871860731596e-06,
+      "loss": 0.3627,
+      "step": 126
+    },
+    {
+      "epoch": 0.01016,
+      "grad_norm": 2.5970005989074707,
+      "learning_rate": 9.997835013179397e-06,
+      "loss": 0.5068,
+      "step": 127
+    },
+    {
+      "epoch": 0.01024,
+      "grad_norm": 1.9385316371917725,
+      "learning_rate": 9.997797849430902e-06,
+      "loss": 0.2864,
+      "step": 128
+    },
+    {
+      "epoch": 0.01032,
+      "grad_norm": 1.7153620719909668,
+      "learning_rate": 9.997760369488458e-06,
+      "loss": 0.3723,
+      "step": 129
+    },
+    {
+      "epoch": 0.0104,
+      "grad_norm": 1.6988774538040161,
+      "learning_rate": 9.997722573354438e-06,
+      "loss": 0.3445,
+      "step": 130
+    },
+    {
+      "epoch": 0.01048,
+      "grad_norm": 2.6378743648529053,
+      "learning_rate": 9.997684461031235e-06,
+      "loss": 0.5199,
+      "step": 131
+    },
+    {
+      "epoch": 0.01056,
+      "grad_norm": 2.355621099472046,
+      "learning_rate": 9.997646032521256e-06,
+      "loss": 0.4259,
+      "step": 132
+    },
+    {
+      "epoch": 0.01064,
+      "grad_norm": 2.3169424533843994,
+      "learning_rate": 9.997607287826937e-06,
+      "loss": 0.5117,
+      "step": 133
+    },
+    {
+      "epoch": 0.01072,
+      "grad_norm": 2.7013490200042725,
+      "learning_rate": 9.997568226950723e-06,
+      "loss": 0.6017,
+      "step": 134
+    },
+    {
+      "epoch": 0.0108,
+      "grad_norm": 1.5648651123046875,
+      "learning_rate": 9.997528849895092e-06,
+      "loss": 0.401,
+      "step": 135
+    },
+    {
+      "epoch": 0.01088,
+      "grad_norm": 2.346057415008545,
+      "learning_rate": 9.997489156662532e-06,
+      "loss": 0.6035,
+      "step": 136
+    },
+    {
+      "epoch": 0.01096,
+      "grad_norm": 3.0977964401245117,
+      "learning_rate": 9.997449147255556e-06,
+      "loss": 0.5332,
+      "step": 137
+    },
+    {
+      "epoch": 0.01104,
+      "grad_norm": 2.7511343955993652,
+      "learning_rate": 9.997408821676693e-06,
+      "loss": 0.3371,
+      "step": 138
+    },
+    {
+      "epoch": 0.01112,
+      "grad_norm": 3.267756938934326,
+      "learning_rate": 9.997368179928495e-06,
+      "loss": 0.349,
+      "step": 139
+    },
+    {
+      "epoch": 0.0112,
+      "grad_norm": 2.05903959274292,
+      "learning_rate": 9.997327222013533e-06,
+      "loss": 0.3592,
+      "step": 140
+    },
+    {
+      "epoch": 0.01128,
+      "grad_norm": 1.7902599573135376,
+      "learning_rate": 9.9972859479344e-06,
+      "loss": 0.3498,
+      "step": 141
+    },
+    {
+      "epoch": 0.01136,
+      "grad_norm": 1.8080183267593384,
+      "learning_rate": 9.997244357693704e-06,
+      "loss": 0.3041,
+      "step": 142
+    },
+    {
+      "epoch": 0.01144,
+      "grad_norm": 2.1454615592956543,
+      "learning_rate": 9.99720245129408e-06,
+      "loss": 0.2934,
+      "step": 143
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 2.059626340866089,
+      "learning_rate": 9.997160228738178e-06,
+      "loss": 0.3284,
+      "step": 144
+    },
+    {
+      "epoch": 0.0116,
+      "grad_norm": 2.013929605484009,
+      "learning_rate": 9.997117690028668e-06,
+      "loss": 0.3648,
+      "step": 145
+    },
+    {
+      "epoch": 0.01168,
+      "grad_norm": 2.726135015487671,
+      "learning_rate": 9.997074835168243e-06,
+      "loss": 0.5155,
+      "step": 146
+    },
+    {
+      "epoch": 0.01176,
+      "grad_norm": 1.494078516960144,
+      "learning_rate": 9.997031664159614e-06,
+      "loss": 0.342,
+      "step": 147
+    },
+    {
+      "epoch": 0.01184,
+      "grad_norm": 3.9255640506744385,
+      "learning_rate": 9.996988177005508e-06,
+      "loss": 0.3398,
+      "step": 148
+    },
+    {
+      "epoch": 0.01192,
+      "grad_norm": 1.7292317152023315,
+      "learning_rate": 9.996944373708683e-06,
+      "loss": 0.3688,
+      "step": 149
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 1.9214504957199097,
+      "learning_rate": 9.996900254271909e-06,
+      "loss": 0.3196,
+      "step": 150
+    },
+    {
+      "epoch": 0.01208,
+      "grad_norm": 2.378833293914795,
+      "learning_rate": 9.996855818697973e-06,
+      "loss": 0.4263,
+      "step": 151
+    },
+    {
+      "epoch": 0.01216,
+      "grad_norm": 2.5661027431488037,
+      "learning_rate": 9.996811066989688e-06,
+      "loss": 0.4251,
+      "step": 152
+    },
+    {
+      "epoch": 0.01224,
+      "grad_norm": 2.3186323642730713,
+      "learning_rate": 9.996765999149888e-06,
+      "loss": 0.5136,
+      "step": 153
+    },
+    {
+      "epoch": 0.01232,
+      "grad_norm": 4.518758296966553,
+      "learning_rate": 9.996720615181422e-06,
+      "loss": 0.4732,
+      "step": 154
+    },
+    {
+      "epoch": 0.0124,
+      "grad_norm": 1.8078404664993286,
+      "learning_rate": 9.996674915087161e-06,
+      "loss": 0.342,
+      "step": 155
+    },
+    {
+      "epoch": 0.01248,
+      "grad_norm": 2.0094892978668213,
+      "learning_rate": 9.996628898869998e-06,
+      "loss": 0.3407,
+      "step": 156
+    },
+    {
+      "epoch": 0.01256,
+      "grad_norm": 2.363971471786499,
+      "learning_rate": 9.996582566532844e-06,
+      "loss": 0.5517,
+      "step": 157
+    },
+    {
+      "epoch": 0.01264,
+      "grad_norm": 1.7961527109146118,
+      "learning_rate": 9.99653591807863e-06,
+      "loss": 0.3353,
+      "step": 158
+    },
+    {
+      "epoch": 0.01272,
+      "grad_norm": 2.390897750854492,
+      "learning_rate": 9.996488953510303e-06,
+      "loss": 0.3781,
+      "step": 159
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 2.0351173877716064,
+      "learning_rate": 9.996441672830842e-06,
+      "loss": 0.4534,
+      "step": 160
+    },
+    {
+      "epoch": 0.01288,
+      "grad_norm": 1.8322193622589111,
+      "learning_rate": 9.996394076043235e-06,
+      "loss": 0.3663,
+      "step": 161
+    },
+    {
+      "epoch": 0.01296,
+      "grad_norm": 1.927498459815979,
+      "learning_rate": 9.996346163150489e-06,
+      "loss": 0.4644,
+      "step": 162
+    },
+    {
+      "epoch": 0.01304,
+      "grad_norm": 1.8679307699203491,
+      "learning_rate": 9.996297934155642e-06,
+      "loss": 0.3235,
+      "step": 163
+    },
+    {
+      "epoch": 0.01312,
+      "grad_norm": 2.3288774490356445,
+      "learning_rate": 9.996249389061742e-06,
+      "loss": 0.3555,
+      "step": 164
+    },
+    {
+      "epoch": 0.0132,
+      "grad_norm": 1.7463264465332031,
+      "learning_rate": 9.99620052787186e-06,
+      "loss": 0.3165,
+      "step": 165
+    },
+    {
+      "epoch": 0.01328,
+      "grad_norm": 1.7550216913223267,
+      "learning_rate": 9.996151350589089e-06,
+      "loss": 0.3353,
+      "step": 166
+    },
+    {
+      "epoch": 0.01336,
+      "grad_norm": 1.5060083866119385,
+      "learning_rate": 9.996101857216538e-06,
+      "loss": 0.3295,
+      "step": 167
+    },
+    {
+      "epoch": 0.01344,
+      "grad_norm": 1.9037864208221436,
+      "learning_rate": 9.996052047757342e-06,
+      "loss": 0.4499,
+      "step": 168
+    },
+    {
+      "epoch": 0.01352,
+      "grad_norm": 2.601609468460083,
+      "learning_rate": 9.996001922214646e-06,
+      "loss": 0.4094,
+      "step": 169
+    },
+    {
+      "epoch": 0.0136,
+      "grad_norm": 1.6243177652359009,
+      "learning_rate": 9.995951480591627e-06,
+      "loss": 0.281,
+      "step": 170
+    },
+    {
+      "epoch": 0.01368,
+      "grad_norm": 1.9210420846939087,
+      "learning_rate": 9.995900722891474e-06,
+      "loss": 0.3456,
+      "step": 171
+    },
+    {
+      "epoch": 0.01376,
+      "grad_norm": 2.4466614723205566,
+      "learning_rate": 9.995849649117398e-06,
+      "loss": 0.3608,
+      "step": 172
+    },
+    {
+      "epoch": 0.01384,
+      "grad_norm": 1.706709384918213,
+      "learning_rate": 9.995798259272633e-06,
+      "loss": 0.368,
+      "step": 173
+    },
+    {
+      "epoch": 0.01392,
+      "grad_norm": 2.536444664001465,
+      "learning_rate": 9.995746553360427e-06,
+      "loss": 0.3482,
+      "step": 174
+    },
+    {
+      "epoch": 0.014,
+      "grad_norm": 1.7476345300674438,
+      "learning_rate": 9.995694531384051e-06,
+      "loss": 0.3771,
+      "step": 175
+    },
+    {
+      "epoch": 0.01408,
+      "grad_norm": 2.676931142807007,
+      "learning_rate": 9.9956421933468e-06,
+      "loss": 0.4398,
+      "step": 176
+    },
+    {
+      "epoch": 0.01416,
+      "grad_norm": 2.442253828048706,
+      "learning_rate": 9.99558953925198e-06,
+      "loss": 0.3727,
+      "step": 177
+    },
+    {
+      "epoch": 0.01424,
+      "grad_norm": 1.871006965637207,
+      "learning_rate": 9.995536569102927e-06,
+      "loss": 0.4419,
+      "step": 178
+    },
+    {
+      "epoch": 0.01432,
+      "grad_norm": 1.8206201791763306,
+      "learning_rate": 9.995483282902992e-06,
+      "loss": 0.3648,
+      "step": 179
+    },
+    {
+      "epoch": 0.0144,
+      "grad_norm": 2.077590227127075,
+      "learning_rate": 9.995429680655541e-06,
+      "loss": 0.3741,
+      "step": 180
+    },
+    {
+      "epoch": 0.01448,
+      "grad_norm": 2.4338691234588623,
+      "learning_rate": 9.995375762363972e-06,
+      "loss": 0.3764,
+      "step": 181
+    },
+    {
+      "epoch": 0.01456,
+      "grad_norm": 1.9777390956878662,
+      "learning_rate": 9.995321528031693e-06,
+      "loss": 0.3789,
+      "step": 182
+    },
+    {
+      "epoch": 0.01464,
+      "grad_norm": 1.9252766370773315,
+      "learning_rate": 9.995266977662132e-06,
+      "loss": 0.3909,
+      "step": 183
+    },
+    {
+      "epoch": 0.01472,
+      "grad_norm": 1.6491369009017944,
+      "learning_rate": 9.995212111258745e-06,
+      "loss": 0.4554,
+      "step": 184
+    },
+    {
+      "epoch": 0.0148,
+      "grad_norm": 1.691046953201294,
+      "learning_rate": 9.995156928825003e-06,
+      "loss": 0.4205,
+      "step": 185
+    },
+    {
+      "epoch": 0.01488,
+      "grad_norm": 1.9484405517578125,
+      "learning_rate": 9.995101430364396e-06,
+      "loss": 0.3699,
+      "step": 186
+    },
+    {
+      "epoch": 0.01496,
+      "grad_norm": 2.238854169845581,
+      "learning_rate": 9.995045615880434e-06,
+      "loss": 0.374,
+      "step": 187
+    },
+    {
+      "epoch": 0.01504,
+      "grad_norm": 1.2045531272888184,
+      "learning_rate": 9.99498948537665e-06,
+      "loss": 0.2368,
+      "step": 188
+    },
+    {
+      "epoch": 0.01512,
+      "grad_norm": 1.7461583614349365,
+      "learning_rate": 9.994933038856595e-06,
+      "loss": 0.3133,
+      "step": 189
+    },
+    {
+      "epoch": 0.0152,
+      "grad_norm": 1.7523412704467773,
+      "learning_rate": 9.994876276323839e-06,
+      "loss": 0.3396,
+      "step": 190
+    },
+    {
+      "epoch": 0.01528,
+      "grad_norm": 3.5537517070770264,
+      "learning_rate": 9.994819197781973e-06,
+      "loss": 0.3683,
+      "step": 191
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 1.3427032232284546,
+      "learning_rate": 9.994761803234611e-06,
+      "loss": 0.291,
+      "step": 192
+    },
+    {
+      "epoch": 0.01544,
+      "grad_norm": 1.7434067726135254,
+      "learning_rate": 9.994704092685381e-06,
+      "loss": 0.3493,
+      "step": 193
+    },
+    {
+      "epoch": 0.01552,
+      "grad_norm": 1.7947523593902588,
+      "learning_rate": 9.994646066137937e-06,
+      "loss": 0.4236,
+      "step": 194
+    },
+    {
+      "epoch": 0.0156,
+      "grad_norm": 2.011258125305176,
+      "learning_rate": 9.994587723595946e-06,
+      "loss": 0.3265,
+      "step": 195
+    },
+    {
+      "epoch": 0.01568,
+      "grad_norm": 1.2146482467651367,
+      "learning_rate": 9.994529065063103e-06,
+      "loss": 0.2635,
+      "step": 196
+    },
+    {
+      "epoch": 0.01576,
+      "grad_norm": 2.2841827869415283,
+      "learning_rate": 9.994470090543118e-06,
+      "loss": 0.4389,
+      "step": 197
+    },
+    {
+      "epoch": 0.01584,
+      "grad_norm": 1.4877972602844238,
+      "learning_rate": 9.994410800039721e-06,
+      "loss": 0.4311,
+      "step": 198
+    },
+    {
+      "epoch": 0.01592,
+      "grad_norm": 2.1514832973480225,
+      "learning_rate": 9.994351193556666e-06,
+      "loss": 0.3551,
+      "step": 199
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 1.679862380027771,
+      "learning_rate": 9.99429127109772e-06,
+      "loss": 0.3547,
+      "step": 200
+    },
+    {
+      "epoch": 0.01608,
+      "grad_norm": 2.345980167388916,
+      "learning_rate": 9.994231032666677e-06,
+      "loss": 0.3212,
+      "step": 201
+    },
+    {
+      "epoch": 0.01616,
+      "grad_norm": 1.6985834836959839,
+      "learning_rate": 9.994170478267348e-06,
+      "loss": 0.4209,
+      "step": 202
+    },
+    {
+      "epoch": 0.01624,
+      "grad_norm": 2.21384334564209,
+      "learning_rate": 9.994109607903563e-06,
+      "loss": 0.3629,
+      "step": 203
+    },
+    {
+      "epoch": 0.01632,
+      "grad_norm": 2.7619104385375977,
+      "learning_rate": 9.994048421579173e-06,
+      "loss": 0.5192,
+      "step": 204
+    },
+    {
+      "epoch": 0.0164,
+      "grad_norm": 2.18821382522583,
+      "learning_rate": 9.993986919298049e-06,
+      "loss": 0.388,
+      "step": 205
+    },
+    {
+      "epoch": 0.01648,
+      "grad_norm": 1.5542010068893433,
+      "learning_rate": 9.993925101064084e-06,
+      "loss": 0.3708,
+      "step": 206
+    },
+    {
+      "epoch": 0.01656,
+      "grad_norm": 1.6064860820770264,
+      "learning_rate": 9.993862966881188e-06,
+      "loss": 0.3564,
+      "step": 207
+    },
+    {
+      "epoch": 0.01664,
+      "grad_norm": 1.946254849433899,
+      "learning_rate": 9.993800516753289e-06,
+      "loss": 0.4394,
+      "step": 208
+    },
+    {
+      "epoch": 0.01672,
+      "grad_norm": 1.9601609706878662,
+      "learning_rate": 9.993737750684342e-06,
+      "loss": 0.4078,
+      "step": 209
+    },
+    {
+      "epoch": 0.0168,
+      "grad_norm": 1.7054260969161987,
+      "learning_rate": 9.993674668678316e-06,
+      "loss": 0.3253,
+      "step": 210
+    },
+    {
+      "epoch": 0.01688,
+      "grad_norm": 1.7017536163330078,
+      "learning_rate": 9.993611270739205e-06,
+      "loss": 0.3473,
+      "step": 211
+    },
+    {
+      "epoch": 0.01696,
+      "grad_norm": 2.0936248302459717,
+      "learning_rate": 9.993547556871015e-06,
+      "loss": 0.4329,
+      "step": 212
+    },
+    {
+      "epoch": 0.01704,
+      "grad_norm": 1.91551673412323,
+      "learning_rate": 9.993483527077782e-06,
+      "loss": 0.3773,
+      "step": 213
+    },
+    {
+      "epoch": 0.01712,
+      "grad_norm": 2.9922895431518555,
+      "learning_rate": 9.99341918136355e-06,
+      "loss": 0.3711,
+      "step": 214
+    },
+    {
+      "epoch": 0.0172,
+      "grad_norm": 1.5471163988113403,
+      "learning_rate": 9.993354519732399e-06,
+      "loss": 0.3224,
+      "step": 215
+    },
+    {
+      "epoch": 0.01728,
+      "grad_norm": 1.5325666666030884,
+      "learning_rate": 9.993289542188413e-06,
+      "loss": 0.33,
+      "step": 216
+    },
+    {
+      "epoch": 0.01736,
+      "grad_norm": 1.9776920080184937,
+      "learning_rate": 9.993224248735706e-06,
+      "loss": 0.3761,
+      "step": 217
+    },
+    {
+      "epoch": 0.01744,
+      "grad_norm": 1.6905009746551514,
+      "learning_rate": 9.993158639378408e-06,
+      "loss": 0.3817,
+      "step": 218
+    },
+    {
+      "epoch": 0.01752,
+      "grad_norm": 1.6124826669692993,
+      "learning_rate": 9.993092714120671e-06,
+      "loss": 0.3097,
+      "step": 219
+    },
+    {
+      "epoch": 0.0176,
+      "grad_norm": 1.5717380046844482,
+      "learning_rate": 9.993026472966664e-06,
+      "loss": 0.3696,
+      "step": 220
+    },
+    {
+      "epoch": 0.01768,
+      "grad_norm": 1.8339051008224487,
+      "learning_rate": 9.992959915920579e-06,
+      "loss": 0.3433,
+      "step": 221
+    },
+    {
+      "epoch": 0.01776,
+      "grad_norm": 1.8827868700027466,
+      "learning_rate": 9.992893042986627e-06,
+      "loss": 0.3711,
+      "step": 222
+    },
+    {
+      "epoch": 0.01784,
+      "grad_norm": 1.704789161682129,
+      "learning_rate": 9.992825854169038e-06,
+      "loss": 0.2893,
+      "step": 223
+    },
+    {
+      "epoch": 0.01792,
+      "grad_norm": 1.803351640701294,
+      "learning_rate": 9.992758349472062e-06,
+      "loss": 0.3313,
+      "step": 224
+    },
+    {
+      "epoch": 0.018,
+      "grad_norm": 1.406293511390686,
+      "learning_rate": 9.992690528899972e-06,
+      "loss": 0.3326,
+      "step": 225
+    },
+    {
+      "epoch": 0.01808,
+      "grad_norm": 3.832010269165039,
+      "learning_rate": 9.992622392457058e-06,
+      "loss": 0.4356,
+      "step": 226
+    },
+    {
+      "epoch": 0.01816,
+      "grad_norm": 1.769728660583496,
+      "learning_rate": 9.992553940147631e-06,
+      "loss": 0.3862,
+      "step": 227
+    },
+    {
+      "epoch": 0.01824,
+      "grad_norm": 2.6798720359802246,
+      "learning_rate": 9.99248517197602e-06,
+      "loss": 0.3847,
+      "step": 228
+    },
+    {
+      "epoch": 0.01832,
+      "grad_norm": 1.8670660257339478,
+      "learning_rate": 9.992416087946579e-06,
+      "loss": 0.3875,
+      "step": 229
+    },
+    {
+      "epoch": 0.0184,
+      "grad_norm": 2.3109071254730225,
+      "learning_rate": 9.992346688063676e-06,
+      "loss": 0.4156,
+      "step": 230
+    },
+    {
+      "epoch": 0.01848,
+      "grad_norm": 1.5571988821029663,
+      "learning_rate": 9.992276972331702e-06,
+      "loss": 0.3264,
+      "step": 231
+    },
+    {
+      "epoch": 0.01856,
+      "grad_norm": 2.017947196960449,
+      "learning_rate": 9.992206940755068e-06,
+      "loss": 0.3828,
+      "step": 232
+    },
+    {
+      "epoch": 0.01864,
+      "grad_norm": 1.9046738147735596,
+      "learning_rate": 9.992136593338206e-06,
+      "loss": 0.3952,
+      "step": 233
+    },
+    {
+      "epoch": 0.01872,
+      "grad_norm": 1.9752707481384277,
+      "learning_rate": 9.992065930085564e-06,
+      "loss": 0.3992,
+      "step": 234
+    },
+    {
+      "epoch": 0.0188,
+      "grad_norm": 2.36694598197937,
+      "learning_rate": 9.991994951001616e-06,
+      "loss": 0.4147,
+      "step": 235
+    },
+    {
+      "epoch": 0.01888,
+      "grad_norm": 2.532259464263916,
+      "learning_rate": 9.99192365609085e-06,
+      "loss": 0.4949,
+      "step": 236
+    },
+    {
+      "epoch": 0.01896,
+      "grad_norm": 2.2116692066192627,
+      "learning_rate": 9.991852045357776e-06,
+      "loss": 0.3448,
+      "step": 237
+    },
+    {
+      "epoch": 0.01904,
+      "grad_norm": 1.5159705877304077,
+      "learning_rate": 9.991780118806927e-06,
+      "loss": 0.3287,
+      "step": 238
+    },
+    {
+      "epoch": 0.01912,
+      "grad_norm": 1.7897611856460571,
+      "learning_rate": 9.991707876442851e-06,
+      "loss": 0.331,
+      "step": 239
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 1.7827035188674927,
+      "learning_rate": 9.991635318270123e-06,
+      "loss": 0.3524,
+      "step": 240
+    },
+    {
+      "epoch": 0.01928,
+      "grad_norm": 2.0052402019500732,
+      "learning_rate": 9.991562444293328e-06,
+      "loss": 0.4706,
+      "step": 241
+    },
+    {
+      "epoch": 0.01936,
+      "grad_norm": 3.4093244075775146,
+      "learning_rate": 9.991489254517079e-06,
+      "loss": 0.4795,
+      "step": 242
+    },
+    {
+      "epoch": 0.01944,
+      "grad_norm": 1.9298174381256104,
+      "learning_rate": 9.991415748946007e-06,
+      "loss": 0.3301,
+      "step": 243
+    },
+    {
+      "epoch": 0.01952,
+      "grad_norm": 1.4808518886566162,
+      "learning_rate": 9.991341927584763e-06,
+      "loss": 0.3292,
+      "step": 244
+    },
+    {
+      "epoch": 0.0196,
+      "grad_norm": 1.9954057931900024,
+      "learning_rate": 9.991267790438016e-06,
+      "loss": 0.4463,
+      "step": 245
+    },
+    {
+      "epoch": 0.01968,
+      "grad_norm": 2.5434648990631104,
+      "learning_rate": 9.991193337510455e-06,
+      "loss": 0.4003,
+      "step": 246
+    },
+    {
+      "epoch": 0.01976,
+      "grad_norm": 1.7687448263168335,
+      "learning_rate": 9.991118568806794e-06,
+      "loss": 0.4522,
+      "step": 247
+    },
+    {
+      "epoch": 0.01984,
+      "grad_norm": 1.513317584991455,
+      "learning_rate": 9.99104348433176e-06,
+      "loss": 0.4053,
+      "step": 248
+    },
+    {
+      "epoch": 0.01992,
+      "grad_norm": 1.8949499130249023,
+      "learning_rate": 9.990968084090104e-06,
+      "loss": 0.3524,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4182684421539307,
+      "learning_rate": 9.9908923680866e-06,
+      "loss": 0.3213,
+      "step": 250
+    },
+    {
+      "epoch": 0.02008,
+      "grad_norm": 2.0624146461486816,
+      "learning_rate": 9.990816336326034e-06,
+      "loss": 0.4941,
+      "step": 251
+    },
+    {
+      "epoch": 0.02016,
+      "grad_norm": 1.8073673248291016,
+      "learning_rate": 9.990739988813219e-06,
+      "loss": 0.342,
+      "step": 252
+    },
+    {
+      "epoch": 0.02024,
+      "grad_norm": 1.4276469945907593,
+      "learning_rate": 9.990663325552981e-06,
+      "loss": 0.2914,
+      "step": 253
+    },
+    {
+      "epoch": 0.02032,
+      "grad_norm": 1.8346103429794312,
+      "learning_rate": 9.990586346550175e-06,
+      "loss": 0.3763,
+      "step": 254
+    },
+    {
+      "epoch": 0.0204,
+      "grad_norm": 2.2009072303771973,
+      "learning_rate": 9.99050905180967e-06,
+      "loss": 0.5039,
+      "step": 255
+    },
+    {
+      "epoch": 0.02048,
+      "grad_norm": 1.2374415397644043,
+      "learning_rate": 9.990431441336354e-06,
+      "loss": 0.2549,
+      "step": 256
+    },
+    {
+      "epoch": 0.02056,
+      "grad_norm": 2.125685214996338,
+      "learning_rate": 9.99035351513514e-06,
+      "loss": 0.3265,
+      "step": 257
+    },
+    {
+      "epoch": 0.02064,
+      "grad_norm": 1.3601418733596802,
+      "learning_rate": 9.990275273210958e-06,
+      "loss": 0.3251,
+      "step": 258
+    },
+    {
+      "epoch": 0.02072,
+      "grad_norm": 1.8217127323150635,
+      "learning_rate": 9.990196715568755e-06,
+      "loss": 0.4395,
+      "step": 259
+    },
+    {
+      "epoch": 0.0208,
+      "grad_norm": 2.603867292404175,
+      "learning_rate": 9.990117842213504e-06,
+      "loss": 0.3981,
+      "step": 260
+    },
+    {
+      "epoch": 0.02088,
+      "grad_norm": 1.8914546966552734,
+      "learning_rate": 9.990038653150194e-06,
+      "loss": 0.3568,
+      "step": 261
+    },
+    {
+      "epoch": 0.02096,
+      "grad_norm": 1.4449388980865479,
+      "learning_rate": 9.989959148383834e-06,
+      "loss": 0.4019,
+      "step": 262
+    },
+    {
+      "epoch": 0.02104,
+      "grad_norm": 1.4248216152191162,
+      "learning_rate": 9.989879327919456e-06,
+      "loss": 0.3217,
+      "step": 263
+    },
+    {
+      "epoch": 0.02112,
+      "grad_norm": 2.652259588241577,
+      "learning_rate": 9.98979919176211e-06,
+      "loss": 0.4723,
+      "step": 264
+    },
+    {
+      "epoch": 0.0212,
+      "grad_norm": 1.64741849899292,
+      "learning_rate": 9.989718739916864e-06,
+      "loss": 0.3847,
+      "step": 265
+    },
+    {
+      "epoch": 0.02128,
+      "grad_norm": 1.9805387258529663,
+      "learning_rate": 9.989637972388809e-06,
+      "loss": 0.5401,
+      "step": 266
+    },
+    {
+      "epoch": 0.02136,
+      "grad_norm": 2.0374245643615723,
+      "learning_rate": 9.989556889183055e-06,
+      "loss": 0.4374,
+      "step": 267
+    },
+    {
+      "epoch": 0.02144,
+      "grad_norm": 1.9687358140945435,
+      "learning_rate": 9.989475490304732e-06,
+      "loss": 0.3703,
+      "step": 268
+    },
+    {
+      "epoch": 0.02152,
+      "grad_norm": 1.6070175170898438,
+      "learning_rate": 9.98939377575899e-06,
+      "loss": 0.2785,
+      "step": 269
+    },
+    {
+      "epoch": 0.0216,
+      "grad_norm": 2.3517184257507324,
+      "learning_rate": 9.989311745550997e-06,
+      "loss": 0.3411,
+      "step": 270
+    },
+    {
+      "epoch": 0.02168,
+      "grad_norm": 2.0133581161499023,
+      "learning_rate": 9.989229399685944e-06,
+      "loss": 0.408,
+      "step": 271
+    },
+    {
+      "epoch": 0.02176,
+      "grad_norm": 2.088970422744751,
+      "learning_rate": 9.989146738169042e-06,
+      "loss": 0.361,
+      "step": 272
+    },
+    {
+      "epoch": 0.02184,
+      "grad_norm": 2.6646909713745117,
+      "learning_rate": 9.98906376100552e-06,
+      "loss": 0.4573,
+      "step": 273
+    },
+    {
+      "epoch": 0.02192,
+      "grad_norm": 2.1084513664245605,
+      "learning_rate": 9.988980468200627e-06,
+      "loss": 0.4713,
+      "step": 274
+    },
+    {
+      "epoch": 0.022,
+      "grad_norm": 1.7543795108795166,
+      "learning_rate": 9.988896859759632e-06,
+      "loss": 0.3186,
+      "step": 275
+    },
+    {
+      "epoch": 0.02208,
+      "grad_norm": 1.4450194835662842,
+      "learning_rate": 9.988812935687826e-06,
+      "loss": 0.3079,
+      "step": 276
+    },
+    {
+      "epoch": 0.02216,
+      "grad_norm": 1.684277892112732,
+      "learning_rate": 9.988728695990518e-06,
+      "loss": 0.3892,
+      "step": 277
+    },
+    {
+      "epoch": 0.02224,
+      "grad_norm": 1.5748530626296997,
+      "learning_rate": 9.988644140673038e-06,
+      "loss": 0.3742,
+      "step": 278
+    },
+    {
+      "epoch": 0.02232,
+      "grad_norm": 1.5772830247879028,
+      "learning_rate": 9.988559269740736e-06,
+      "loss": 0.3195,
+      "step": 279
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 1.4628132581710815,
+      "learning_rate": 9.98847408319898e-06,
+      "loss": 0.3627,
+      "step": 280
+    },
+    {
+      "epoch": 0.02248,
+      "grad_norm": 2.2950448989868164,
+      "learning_rate": 9.98838858105316e-06,
+      "loss": 0.4156,
+      "step": 281
+    },
+    {
+      "epoch": 0.02256,
+      "grad_norm": 1.51676344871521,
+      "learning_rate": 9.988302763308686e-06,
+      "loss": 0.2407,
+      "step": 282
+    },
+    {
+      "epoch": 0.02264,
+      "grad_norm": 1.6948258876800537,
+      "learning_rate": 9.988216629970987e-06,
+      "loss": 0.3616,
+      "step": 283
+    },
+    {
+      "epoch": 0.02272,
+      "grad_norm": 1.9347060918807983,
+      "learning_rate": 9.988130181045512e-06,
+      "loss": 0.3056,
+      "step": 284
+    },
+    {
+      "epoch": 0.0228,
+      "grad_norm": 2.000277280807495,
+      "learning_rate": 9.988043416537731e-06,
+      "loss": 0.402,
+      "step": 285
+    },
+    {
+      "epoch": 0.02288,
+      "grad_norm": 1.702262043952942,
+      "learning_rate": 9.987956336453135e-06,
+      "loss": 0.358,
+      "step": 286
+    },
+    {
+      "epoch": 0.02296,
+      "grad_norm": 1.9157183170318604,
+      "learning_rate": 9.98786894079723e-06,
+      "loss": 0.414,
+      "step": 287
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 2.2817223072052,
+      "learning_rate": 9.987781229575547e-06,
+      "loss": 0.4701,
+      "step": 288
+    },
+    {
+      "epoch": 0.02312,
+      "grad_norm": 1.9925824403762817,
+      "learning_rate": 9.987693202793633e-06,
+      "loss": 0.3452,
+      "step": 289
+    },
+    {
+      "epoch": 0.0232,
+      "grad_norm": 1.446832299232483,
+      "learning_rate": 9.98760486045706e-06,
+      "loss": 0.3077,
+      "step": 290
+    },
+    {
+      "epoch": 0.02328,
+      "grad_norm": 1.2968581914901733,
+      "learning_rate": 9.987516202571417e-06,
+      "loss": 0.2845,
+      "step": 291
+    },
+    {
+      "epoch": 0.02336,
+      "grad_norm": 1.7567917108535767,
+      "learning_rate": 9.98742722914231e-06,
+      "loss": 0.4025,
+      "step": 292
+    },
+    {
+      "epoch": 0.02344,
+      "grad_norm": 1.824837327003479,
+      "learning_rate": 9.987337940175371e-06,
+      "loss": 0.3339,
+      "step": 293
+    },
+    {
+      "epoch": 0.02352,
+      "grad_norm": 1.6700366735458374,
+      "learning_rate": 9.98724833567625e-06,
+      "loss": 0.3338,
+      "step": 294
+    },
+    {
+      "epoch": 0.0236,
+      "grad_norm": 1.7511111497879028,
+      "learning_rate": 9.987158415650612e-06,
+      "loss": 0.4037,
+      "step": 295
+    },
+    {
+      "epoch": 0.02368,
+      "grad_norm": 1.5636446475982666,
+      "learning_rate": 9.987068180104148e-06,
+      "loss": 0.3424,
+      "step": 296
+    },
+    {
+      "epoch": 0.02376,
+      "grad_norm": 1.8241537809371948,
+      "learning_rate": 9.986977629042569e-06,
+      "loss": 0.5058,
+      "step": 297
+    },
+    {
+      "epoch": 0.02384,
+      "grad_norm": 2.1625335216522217,
+      "learning_rate": 9.986886762471601e-06,
+      "loss": 0.4052,
+      "step": 298
+    },
+    {
+      "epoch": 0.02392,
+      "grad_norm": 1.7909748554229736,
+      "learning_rate": 9.986795580396994e-06,
+      "loss": 0.3581,
+      "step": 299
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 1.2463935613632202,
+      "learning_rate": 9.986704082824516e-06,
+      "loss": 0.2531,
+      "step": 300
+    },
+    {
+      "epoch": 0.02408,
+      "grad_norm": 1.689233422279358,
+      "learning_rate": 9.986612269759956e-06,
+      "loss": 0.4526,
+      "step": 301
+    },
+    {
+      "epoch": 0.02416,
+      "grad_norm": 2.223651170730591,
+      "learning_rate": 9.986520141209123e-06,
+      "loss": 0.5223,
+      "step": 302
+    },
+    {
+      "epoch": 0.02424,
+      "grad_norm": 2.412198066711426,
+      "learning_rate": 9.986427697177847e-06,
+      "loss": 0.4936,
+      "step": 303
+    },
+    {
+      "epoch": 0.02432,
+      "grad_norm": 1.578535795211792,
+      "learning_rate": 9.986334937671974e-06,
+      "loss": 0.3544,
+      "step": 304
+    },
+    {
+      "epoch": 0.0244,
+      "grad_norm": 1.6475830078125,
+      "learning_rate": 9.986241862697375e-06,
+      "loss": 0.3465,
+      "step": 305
+    },
+    {
+      "epoch": 0.02448,
+      "grad_norm": 1.4819910526275635,
+      "learning_rate": 9.986148472259935e-06,
+      "loss": 0.2884,
+      "step": 306
+    },
+    {
+      "epoch": 0.02456,
+      "grad_norm": 1.7936198711395264,
+      "learning_rate": 9.986054766365566e-06,
+      "loss": 0.5354,
+      "step": 307
+    },
+    {
+      "epoch": 0.02464,
+      "grad_norm": 1.8793686628341675,
+      "learning_rate": 9.985960745020195e-06,
+      "loss": 0.3592,
+      "step": 308
+    },
+    {
+      "epoch": 0.02472,
+      "grad_norm": 1.6589360237121582,
+      "learning_rate": 9.985866408229773e-06,
+      "loss": 0.3107,
+      "step": 309
+    },
+    {
+      "epoch": 0.0248,
+      "grad_norm": 1.6323789358139038,
+      "learning_rate": 9.985771756000264e-06,
+      "loss": 0.3276,
+      "step": 310
+    },
+    {
+      "epoch": 0.02488,
+      "grad_norm": 1.6884311437606812,
+      "learning_rate": 9.98567678833766e-06,
+      "loss": 0.3694,
+      "step": 311
+    },
+    {
+      "epoch": 0.02496,
+      "grad_norm": 1.9537749290466309,
+      "learning_rate": 9.985581505247966e-06,
+      "loss": 0.4009,
+      "step": 312
+    },
+    {
+      "epoch": 0.02504,
+      "grad_norm": 1.6910805702209473,
+      "learning_rate": 9.985485906737212e-06,
+      "loss": 0.3655,
+      "step": 313
+    },
+    {
+      "epoch": 0.02512,
+      "grad_norm": 1.6935604810714722,
+      "learning_rate": 9.985389992811447e-06,
+      "loss": 0.3912,
+      "step": 314
+    },
+    {
+      "epoch": 0.0252,
+      "grad_norm": 1.7497141361236572,
+      "learning_rate": 9.985293763476738e-06,
+      "loss": 0.3744,
+      "step": 315
+    },
+    {
+      "epoch": 0.02528,
+      "grad_norm": 1.6273237466812134,
+      "learning_rate": 9.985197218739173e-06,
+      "loss": 0.316,
+      "step": 316
+    },
+    {
+      "epoch": 0.02536,
+      "grad_norm": 1.482218623161316,
+      "learning_rate": 9.985100358604861e-06,
+      "loss": 0.3293,
+      "step": 317
+    },
+    {
+      "epoch": 0.02544,
+      "grad_norm": 1.6040552854537964,
+      "learning_rate": 9.985003183079929e-06,
+      "loss": 0.3666,
+      "step": 318
+    },
+    {
+      "epoch": 0.02552,
+      "grad_norm": 1.7107411623001099,
+      "learning_rate": 9.984905692170525e-06,
+      "loss": 0.3839,
+      "step": 319
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 1.7257567644119263,
+      "learning_rate": 9.984807885882819e-06,
+      "loss": 0.3729,
+      "step": 320
+    },
+    {
+      "epoch": 0.02568,
+      "grad_norm": 1.6172734498977661,
+      "learning_rate": 9.984709764222997e-06,
+      "loss": 0.3743,
+      "step": 321
+    },
+    {
+      "epoch": 0.02576,
+      "grad_norm": 1.929922342300415,
+      "learning_rate": 9.984611327197267e-06,
+      "loss": 0.3688,
+      "step": 322
+    },
+    {
+      "epoch": 0.02584,
+      "grad_norm": 1.99301278591156,
+      "learning_rate": 9.984512574811857e-06,
+      "loss": 0.3554,
+      "step": 323
+    },
+    {
+      "epoch": 0.02592,
+      "grad_norm": 1.3338180780410767,
+      "learning_rate": 9.984413507073014e-06,
+      "loss": 0.2592,
+      "step": 324
+    },
+    {
+      "epoch": 0.026,
+      "grad_norm": 1.341111183166504,
+      "learning_rate": 9.984314123987006e-06,
+      "loss": 0.2646,
+      "step": 325
+    },
+    {
+      "epoch": 0.02608,
+      "grad_norm": 1.6769309043884277,
+      "learning_rate": 9.984214425560122e-06,
+      "loss": 0.357,
+      "step": 326
+    },
+    {
+      "epoch": 0.02616,
+      "grad_norm": 2.0242223739624023,
+      "learning_rate": 9.984114411798667e-06,
+      "loss": 0.5107,
+      "step": 327
+    },
+    {
+      "epoch": 0.02624,
+      "grad_norm": 1.5169055461883545,
+      "learning_rate": 9.984014082708972e-06,
+      "loss": 0.3416,
+      "step": 328
+    },
+    {
+      "epoch": 0.02632,
+      "grad_norm": 1.7003381252288818,
+      "learning_rate": 9.983913438297381e-06,
+      "loss": 0.3307,
+      "step": 329
+    },
+    {
+      "epoch": 0.0264,
+      "grad_norm": 1.7793561220169067,
+      "learning_rate": 9.983812478570265e-06,
+      "loss": 0.3307,
+      "step": 330
+    },
+    {
+      "epoch": 0.02648,
+      "grad_norm": 1.37666916847229,
+      "learning_rate": 9.983711203534008e-06,
+      "loss": 0.2952,
+      "step": 331
+    },
+    {
+      "epoch": 0.02656,
+      "grad_norm": 1.3145544528961182,
+      "learning_rate": 9.983609613195018e-06,
+      "loss": 0.2823,
+      "step": 332
+    },
+    {
+      "epoch": 0.02664,
+      "grad_norm": 1.7312018871307373,
+      "learning_rate": 9.983507707559724e-06,
+      "loss": 0.3524,
+      "step": 333
+    },
+    {
+      "epoch": 0.02672,
+      "grad_norm": 1.6275979280471802,
+      "learning_rate": 9.983405486634572e-06,
+      "loss": 0.3636,
+      "step": 334
+    },
+    {
+      "epoch": 0.0268,
+      "grad_norm": 1.8315293788909912,
+      "learning_rate": 9.983302950426028e-06,
+      "loss": 0.3742,
+      "step": 335
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 2.2285656929016113,
+      "learning_rate": 9.983200098940582e-06,
+      "loss": 0.5228,
+      "step": 336
+    },
+    {
+      "epoch": 0.02696,
+      "grad_norm": 2.5498311519622803,
+      "learning_rate": 9.98309693218474e-06,
+      "loss": 0.3208,
+      "step": 337
+    },
+    {
+      "epoch": 0.02704,
+      "grad_norm": 1.8318697214126587,
+      "learning_rate": 9.982993450165028e-06,
+      "loss": 0.3807,
+      "step": 338
+    },
+    {
+      "epoch": 0.02712,
+      "grad_norm": 1.5511677265167236,
+      "learning_rate": 9.982889652887992e-06,
+      "loss": 0.3051,
+      "step": 339
+    },
+    {
+      "epoch": 0.0272,
+      "grad_norm": 1.9955275058746338,
+      "learning_rate": 9.982785540360202e-06,
+      "loss": 0.4089,
+      "step": 340
+    },
+    {
+      "epoch": 0.02728,
+      "grad_norm": 1.8035287857055664,
+      "learning_rate": 9.982681112588244e-06,
+      "loss": 0.4043,
+      "step": 341
+    },
+    {
+      "epoch": 0.02736,
+      "grad_norm": 1.9366258382797241,
+      "learning_rate": 9.982576369578724e-06,
+      "loss": 0.3673,
+      "step": 342
+    },
+    {
+      "epoch": 0.02744,
+      "grad_norm": 1.7964532375335693,
+      "learning_rate": 9.982471311338268e-06,
+      "loss": 0.3345,
+      "step": 343
+    },
+    {
+      "epoch": 0.02752,
+      "grad_norm": 1.6594544649124146,
+      "learning_rate": 9.982365937873523e-06,
+      "loss": 0.4043,
+      "step": 344
+    },
+    {
+      "epoch": 0.0276,
+      "grad_norm": 2.0171964168548584,
+      "learning_rate": 9.982260249191159e-06,
+      "loss": 0.4117,
+      "step": 345
+    },
+    {
+      "epoch": 0.02768,
+      "grad_norm": 1.921452283859253,
+      "learning_rate": 9.982154245297856e-06,
+      "loss": 0.4141,
+      "step": 346
+    },
+    {
+      "epoch": 0.02776,
+      "grad_norm": 1.8650532960891724,
+      "learning_rate": 9.982047926200327e-06,
+      "loss": 0.3625,
+      "step": 347
+    },
+    {
+      "epoch": 0.02784,
+      "grad_norm": 1.7884360551834106,
+      "learning_rate": 9.981941291905294e-06,
+      "loss": 0.3626,
+      "step": 348
+    },
+    {
+      "epoch": 0.02792,
+      "grad_norm": 1.35147225856781,
+      "learning_rate": 9.981834342419506e-06,
+      "loss": 0.2929,
+      "step": 349
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 1.7383079528808594,
+      "learning_rate": 9.981727077749727e-06,
+      "loss": 0.378,
+      "step": 350
+    },
+    {
+      "epoch": 0.02808,
+      "grad_norm": 1.9508424997329712,
+      "learning_rate": 9.981619497902746e-06,
+      "loss": 0.3706,
+      "step": 351
+    },
+    {
+      "epoch": 0.02816,
+      "grad_norm": 2.229902744293213,
+      "learning_rate": 9.981511602885368e-06,
+      "loss": 0.4509,
+      "step": 352
+    },
+    {
+      "epoch": 0.02824,
+      "grad_norm": 1.4837419986724854,
+      "learning_rate": 9.981403392704419e-06,
+      "loss": 0.3169,
+      "step": 353
+    },
+    {
+      "epoch": 0.02832,
+      "grad_norm": 1.9721978902816772,
+      "learning_rate": 9.981294867366745e-06,
+      "loss": 0.4006,
+      "step": 354
+    },
+    {
+      "epoch": 0.0284,
+      "grad_norm": 1.4611529111862183,
+      "learning_rate": 9.981186026879212e-06,
+      "loss": 0.2671,
+      "step": 355
+    },
+    {
+      "epoch": 0.02848,
+      "grad_norm": 1.6954723596572876,
+      "learning_rate": 9.981076871248705e-06,
+      "loss": 0.3658,
+      "step": 356
+    },
+    {
+      "epoch": 0.02856,
+      "grad_norm": 1.6720842123031616,
+      "learning_rate": 9.980967400482134e-06,
+      "loss": 0.3438,
+      "step": 357
+    },
+    {
+      "epoch": 0.02864,
+      "grad_norm": 1.6820980310440063,
+      "learning_rate": 9.98085761458642e-06,
+      "loss": 0.3547,
+      "step": 358
+    },
+    {
+      "epoch": 0.02872,
+      "grad_norm": 1.71517813205719,
+      "learning_rate": 9.980747513568511e-06,
+      "loss": 0.3521,
+      "step": 359
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 1.6508595943450928,
+      "learning_rate": 9.980637097435372e-06,
+      "loss": 0.369,
+      "step": 360
+    },
+    {
+      "epoch": 0.02888,
+      "grad_norm": 1.8108190298080444,
+      "learning_rate": 9.98052636619399e-06,
+      "loss": 0.3725,
+      "step": 361
+    },
+    {
+      "epoch": 0.02896,
+      "grad_norm": 1.6566613912582397,
+      "learning_rate": 9.98041531985137e-06,
+      "loss": 0.4799,
+      "step": 362
+    },
+    {
+      "epoch": 0.02904,
+      "grad_norm": 1.784568428993225,
+      "learning_rate": 9.980303958414537e-06,
+      "loss": 0.284,
+      "step": 363
+    },
+    {
+      "epoch": 0.02912,
+      "grad_norm": 1.7334696054458618,
+      "learning_rate": 9.980192281890535e-06,
+      "loss": 0.3531,
+      "step": 364
+    },
+    {
+      "epoch": 0.0292,
+      "grad_norm": 1.55849289894104,
+      "learning_rate": 9.980080290286434e-06,
+      "loss": 0.3523,
+      "step": 365
+    },
+    {
+      "epoch": 0.02928,
+      "grad_norm": 1.8714348077774048,
+      "learning_rate": 9.979967983609313e-06,
+      "loss": 0.315,
+      "step": 366
+    },
+    {
+      "epoch": 0.02936,
+      "grad_norm": 1.5885354280471802,
+      "learning_rate": 9.979855361866283e-06,
+      "loss": 0.3596,
+      "step": 367
+    },
+    {
+      "epoch": 0.02944,
+      "grad_norm": 1.5087720155715942,
+      "learning_rate": 9.979742425064467e-06,
+      "loss": 0.4048,
+      "step": 368
+    },
+    {
+      "epoch": 0.02952,
+      "grad_norm": 1.7668178081512451,
+      "learning_rate": 9.97962917321101e-06,
+      "loss": 0.3881,
+      "step": 369
+    },
+    {
+      "epoch": 0.0296,
+      "grad_norm": 1.5784231424331665,
+      "learning_rate": 9.979515606313074e-06,
+      "loss": 0.3335,
+      "step": 370
+    },
+    {
+      "epoch": 0.02968,
+      "grad_norm": 2.16629695892334,
+      "learning_rate": 9.97940172437785e-06,
+      "loss": 0.5286,
+      "step": 371
+    },
+    {
+      "epoch": 0.02976,
+      "grad_norm": 2.021318197250366,
+      "learning_rate": 9.979287527412541e-06,
+      "loss": 0.3768,
+      "step": 372
+    },
+    {
+      "epoch": 0.02984,
+      "grad_norm": 2.232651710510254,
+      "learning_rate": 9.979173015424369e-06,
+      "loss": 0.517,
+      "step": 373
+    },
+    {
+      "epoch": 0.02992,
+      "grad_norm": 1.7764800786972046,
+      "learning_rate": 9.979058188420581e-06,
+      "loss": 0.3594,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5262402296066284,
+      "learning_rate": 9.978943046408442e-06,
+      "loss": 0.3362,
+      "step": 375
+    },
+    {
+      "epoch": 0.03008,
+      "grad_norm": 1.5017170906066895,
+      "learning_rate": 9.978827589395238e-06,
+      "loss": 0.3163,
+      "step": 376
+    },
+    {
+      "epoch": 0.03016,
+      "grad_norm": 1.7568035125732422,
+      "learning_rate": 9.978711817388266e-06,
+      "loss": 0.3264,
+      "step": 377
+    },
+    {
+      "epoch": 0.03024,
+      "grad_norm": 1.738759160041809,
+      "learning_rate": 9.978595730394861e-06,
+      "loss": 0.4094,
+      "step": 378
+    },
+    {
+      "epoch": 0.03032,
+      "grad_norm": 1.2938789129257202,
+      "learning_rate": 9.978479328422362e-06,
+      "loss": 0.2659,
+      "step": 379
+    },
+    {
+      "epoch": 0.0304,
+      "grad_norm": 1.6452093124389648,
+      "learning_rate": 9.978362611478132e-06,
+      "loss": 0.3264,
+      "step": 380
+    },
+    {
+      "epoch": 0.03048,
+      "grad_norm": 1.610556721687317,
+      "learning_rate": 9.978245579569558e-06,
+      "loss": 0.3351,
+      "step": 381
+    },
+    {
+      "epoch": 0.03056,
+      "grad_norm": 1.4417246580123901,
+      "learning_rate": 9.978128232704044e-06,
+      "loss": 0.3223,
+      "step": 382
+    },
+    {
+      "epoch": 0.03064,
+      "grad_norm": 1.9182772636413574,
+      "learning_rate": 9.978010570889013e-06,
+      "loss": 0.3542,
+      "step": 383
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 2.3410074710845947,
+      "learning_rate": 9.97789259413191e-06,
+      "loss": 0.3877,
+      "step": 384
+    },
+    {
+      "epoch": 0.0308,
+      "grad_norm": 1.9133309125900269,
+      "learning_rate": 9.977774302440199e-06,
+      "loss": 0.43,
+      "step": 385
+    },
+    {
+      "epoch": 0.03088,
+      "grad_norm": 1.8875223398208618,
+      "learning_rate": 9.977655695821362e-06,
+      "loss": 0.4885,
+      "step": 386
+    },
+    {
+      "epoch": 0.03096,
+      "grad_norm": 1.7378370761871338,
+      "learning_rate": 9.977536774282906e-06,
+      "loss": 0.4246,
+      "step": 387
+    },
+    {
+      "epoch": 0.03104,
+      "grad_norm": 1.8936777114868164,
+      "learning_rate": 9.977417537832352e-06,
+      "loss": 0.3571,
+      "step": 388
+    },
+    {
+      "epoch": 0.03112,
+      "grad_norm": 1.7290297746658325,
+      "learning_rate": 9.977297986477246e-06,
+      "loss": 0.3553,
+      "step": 389
+    },
+    {
+      "epoch": 0.0312,
+      "grad_norm": 1.4241451025009155,
+      "learning_rate": 9.977178120225151e-06,
+      "loss": 0.2923,
+      "step": 390
+    },
+    {
+      "epoch": 0.03128,
+      "grad_norm": 1.917623519897461,
+      "learning_rate": 9.977057939083648e-06,
+      "loss": 0.3546,
+      "step": 391
+    },
+    {
+      "epoch": 0.03136,
+      "grad_norm": 1.762268304824829,
+      "learning_rate": 9.976937443060343e-06,
+      "loss": 0.386,
+      "step": 392
+    },
+    {
+      "epoch": 0.03144,
+      "grad_norm": 1.6032752990722656,
+      "learning_rate": 9.97681663216286e-06,
+      "loss": 0.3461,
+      "step": 393
+    },
+    {
+      "epoch": 0.03152,
+      "grad_norm": 1.9459599256515503,
+      "learning_rate": 9.97669550639884e-06,
+      "loss": 0.4753,
+      "step": 394
+    },
+    {
+      "epoch": 0.0316,
+      "grad_norm": 2.0758650302886963,
+      "learning_rate": 9.97657406577595e-06,
+      "loss": 0.4122,
+      "step": 395
+    },
+    {
+      "epoch": 0.03168,
+      "grad_norm": 1.4958903789520264,
+      "learning_rate": 9.976452310301867e-06,
+      "loss": 0.3139,
+      "step": 396
+    },
+    {
+      "epoch": 0.03176,
+      "grad_norm": 1.7065391540527344,
+      "learning_rate": 9.9763302399843e-06,
+      "loss": 0.3258,
+      "step": 397
+    },
+    {
+      "epoch": 0.03184,
+      "grad_norm": 1.7780778408050537,
+      "learning_rate": 9.976207854830968e-06,
+      "loss": 0.3458,
+      "step": 398
+    },
+    {
+      "epoch": 0.03192,
+      "grad_norm": 1.318264365196228,
+      "learning_rate": 9.976085154849617e-06,
+      "loss": 0.3148,
+      "step": 399
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 2.011800527572632,
+      "learning_rate": 9.975962140048007e-06,
+      "loss": 0.3416,
+      "step": 400
+    },
+    {
+      "epoch": 0.03208,
+      "grad_norm": 1.5037428140640259,
+      "learning_rate": 9.975838810433922e-06,
+      "loss": 0.3242,
+      "step": 401
+    },
+    {
+      "epoch": 0.03216,
+      "grad_norm": 1.8600457906723022,
+      "learning_rate": 9.975715166015165e-06,
+      "loss": 0.3643,
+      "step": 402
+    },
+    {
+      "epoch": 0.03224,
+      "grad_norm": 1.61443293094635,
+      "learning_rate": 9.975591206799559e-06,
+      "loss": 0.357,
+      "step": 403
+    },
+    {
+      "epoch": 0.03232,
+      "grad_norm": 1.9129692316055298,
+      "learning_rate": 9.975466932794943e-06,
+      "loss": 0.4214,
+      "step": 404
+    },
+    {
+      "epoch": 0.0324,
+      "grad_norm": 1.9851549863815308,
+      "learning_rate": 9.975342344009186e-06,
+      "loss": 0.4749,
+      "step": 405
+    },
+    {
+      "epoch": 0.03248,
+      "grad_norm": 1.711588978767395,
+      "learning_rate": 9.975217440450164e-06,
+      "loss": 0.32,
+      "step": 406
+    },
+    {
+      "epoch": 0.03256,
+      "grad_norm": 1.4885379076004028,
+      "learning_rate": 9.975092222125783e-06,
+      "loss": 0.3246,
+      "step": 407
+    },
+    {
+      "epoch": 0.03264,
+      "grad_norm": 1.6126595735549927,
+      "learning_rate": 9.974966689043963e-06,
+      "loss": 0.3292,
+      "step": 408
+    },
+    {
+      "epoch": 0.03272,
+      "grad_norm": 1.5870492458343506,
+      "learning_rate": 9.974840841212648e-06,
+      "loss": 0.3787,
+      "step": 409
+    },
+    {
+      "epoch": 0.0328,
+      "grad_norm": 1.4002914428710938,
+      "learning_rate": 9.974714678639797e-06,
+      "loss": 0.2722,
+      "step": 410
+    },
+    {
+      "epoch": 0.03288,
+      "grad_norm": 2.1911635398864746,
+      "learning_rate": 9.974588201333394e-06,
+      "loss": 0.4562,
+      "step": 411
+    },
+    {
+      "epoch": 0.03296,
+      "grad_norm": 1.665370225906372,
+      "learning_rate": 9.974461409301442e-06,
+      "loss": 0.4463,
+      "step": 412
+    },
+    {
+      "epoch": 0.03304,
+      "grad_norm": 4.2547736167907715,
+      "learning_rate": 9.97433430255196e-06,
+      "loss": 0.4829,
+      "step": 413
+    },
+    {
+      "epoch": 0.03312,
+      "grad_norm": 2.003357410430908,
+      "learning_rate": 9.97420688109299e-06,
+      "loss": 0.2167,
+      "step": 414
+    },
+    {
+      "epoch": 0.0332,
+      "grad_norm": 5.256060600280762,
+      "learning_rate": 9.974079144932596e-06,
+      "loss": 0.3008,
+      "step": 415
+    },
+    {
+      "epoch": 0.03328,
+      "grad_norm": 2.828409433364868,
+      "learning_rate": 9.973951094078857e-06,
+      "loss": 0.3806,
+      "step": 416
+    },
+    {
+      "epoch": 0.03336,
+      "grad_norm": 3.722662925720215,
+      "learning_rate": 9.973822728539876e-06,
+      "loss": 0.4364,
+      "step": 417
+    },
+    {
+      "epoch": 0.03344,
+      "grad_norm": 2.073556900024414,
+      "learning_rate": 9.973694048323773e-06,
+      "loss": 0.3991,
+      "step": 418
+    },
+    {
+      "epoch": 0.03352,
+      "grad_norm": 1.1159542798995972,
+      "learning_rate": 9.97356505343869e-06,
+      "loss": 0.2556,
+      "step": 419
+    },
+    {
+      "epoch": 0.0336,
+      "grad_norm": 1.9902396202087402,
+      "learning_rate": 9.973435743892787e-06,
+      "loss": 0.3929,
+      "step": 420
+    },
+    {
+      "epoch": 0.03368,
+      "grad_norm": 1.4526017904281616,
+      "learning_rate": 9.973306119694246e-06,
+      "loss": 0.3209,
+      "step": 421
+    },
+    {
+      "epoch": 0.03376,
+      "grad_norm": 1.8279871940612793,
+      "learning_rate": 9.973176180851267e-06,
+      "loss": 0.4486,
+      "step": 422
+    },
+    {
+      "epoch": 0.03384,
+      "grad_norm": 1.1028552055358887,
+      "learning_rate": 9.973045927372071e-06,
+      "loss": 0.2656,
+      "step": 423
+    },
+    {
+      "epoch": 0.03392,
+      "grad_norm": 1.7597603797912598,
+      "learning_rate": 9.972915359264901e-06,
+      "loss": 0.3717,
+      "step": 424
+    },
+    {
+      "epoch": 0.034,
+      "grad_norm": 1.9462575912475586,
+      "learning_rate": 9.972784476538014e-06,
+      "loss": 0.3828,
+      "step": 425
+    },
+    {
+      "epoch": 0.03408,
+      "grad_norm": 2.132319927215576,
+      "learning_rate": 9.972653279199693e-06,
+      "loss": 0.4109,
+      "step": 426
+    },
+    {
+      "epoch": 0.03416,
+      "grad_norm": 1.7720433473587036,
+      "learning_rate": 9.972521767258236e-06,
+      "loss": 0.3781,
+      "step": 427
+    },
+    {
+      "epoch": 0.03424,
+      "grad_norm": 1.556762933731079,
+      "learning_rate": 9.972389940721967e-06,
+      "loss": 0.3173,
+      "step": 428
+    },
+    {
+      "epoch": 0.03432,
+      "grad_norm": 1.8000061511993408,
+      "learning_rate": 9.972257799599223e-06,
+      "loss": 0.4679,
+      "step": 429
+    },
+    {
+      "epoch": 0.0344,
+      "grad_norm": 1.4809861183166504,
+      "learning_rate": 9.972125343898366e-06,
+      "loss": 0.3235,
+      "step": 430
+    },
+    {
+      "epoch": 0.03448,
+      "grad_norm": 1.4493865966796875,
+      "learning_rate": 9.971992573627775e-06,
+      "loss": 0.2701,
+      "step": 431
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 1.4800419807434082,
+      "learning_rate": 9.971859488795849e-06,
+      "loss": 0.2702,
+      "step": 432
+    },
+    {
+      "epoch": 0.03464,
+      "grad_norm": 1.9442113637924194,
+      "learning_rate": 9.97172608941101e-06,
+      "loss": 0.372,
+      "step": 433
+    },
+    {
+      "epoch": 0.03472,
+      "grad_norm": 2.0416910648345947,
+      "learning_rate": 9.971592375481697e-06,
+      "loss": 0.4425,
+      "step": 434
+    },
+    {
+      "epoch": 0.0348,
+      "grad_norm": 1.7712527513504028,
+      "learning_rate": 9.971458347016369e-06,
+      "loss": 0.4006,
+      "step": 435
+    },
+    {
+      "epoch": 0.03488,
+      "grad_norm": 1.808829426765442,
+      "learning_rate": 9.971324004023506e-06,
+      "loss": 0.3774,
+      "step": 436
+    },
+    {
+      "epoch": 0.03496,
+      "grad_norm": 1.4706016778945923,
+      "learning_rate": 9.971189346511608e-06,
+      "loss": 0.3434,
+      "step": 437
+    },
+    {
+      "epoch": 0.03504,
+      "grad_norm": 1.7320775985717773,
+      "learning_rate": 9.971054374489193e-06,
+      "loss": 0.3452,
+      "step": 438
+    },
+    {
+      "epoch": 0.03512,
+      "grad_norm": 1.6511974334716797,
+      "learning_rate": 9.9709190879648e-06,
+      "loss": 0.3614,
+      "step": 439
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 1.4348496198654175,
+      "learning_rate": 9.970783486946991e-06,
+      "loss": 0.3445,
+      "step": 440
+    },
+    {
+      "epoch": 0.03528,
+      "grad_norm": 1.5991486310958862,
+      "learning_rate": 9.970647571444341e-06,
+      "loss": 0.4998,
+      "step": 441
+    },
+    {
+      "epoch": 0.03536,
+      "grad_norm": 1.3900196552276611,
+      "learning_rate": 9.970511341465453e-06,
+      "loss": 0.3307,
+      "step": 442
+    },
+    {
+      "epoch": 0.03544,
+      "grad_norm": 1.249131202697754,
+      "learning_rate": 9.970374797018942e-06,
+      "loss": 0.2656,
+      "step": 443
+    },
+    {
+      "epoch": 0.03552,
+      "grad_norm": 1.7488571405410767,
+      "learning_rate": 9.97023793811345e-06,
+      "loss": 0.3783,
+      "step": 444
+    },
+    {
+      "epoch": 0.0356,
+      "grad_norm": 1.5800598859786987,
+      "learning_rate": 9.970100764757635e-06,
+      "loss": 0.2901,
+      "step": 445
+    },
+    {
+      "epoch": 0.03568,
+      "grad_norm": 1.4809678792953491,
+      "learning_rate": 9.969963276960173e-06,
+      "loss": 0.3579,
+      "step": 446
+    },
+    {
+      "epoch": 0.03576,
+      "grad_norm": 1.9391019344329834,
+      "learning_rate": 9.969825474729763e-06,
+      "loss": 0.3893,
+      "step": 447
+    },
+    {
+      "epoch": 0.03584,
+      "grad_norm": 1.4065037965774536,
+      "learning_rate": 9.969687358075126e-06,
+      "loss": 0.3532,
+      "step": 448
+    },
+    {
+      "epoch": 0.03592,
+      "grad_norm": 1.735000729560852,
+      "learning_rate": 9.969548927004998e-06,
+      "loss": 0.3295,
+      "step": 449
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 1.9082682132720947,
+      "learning_rate": 9.969410181528138e-06,
+      "loss": 0.3841,
+      "step": 450
+    },
+    {
+      "epoch": 0.03608,
+      "grad_norm": 1.740655779838562,
+      "learning_rate": 9.969271121653323e-06,
+      "loss": 0.3545,
+      "step": 451
+    },
+    {
+      "epoch": 0.03616,
+      "grad_norm": 2.0228629112243652,
+      "learning_rate": 9.96913174738935e-06,
+      "loss": 0.4113,
+      "step": 452
+    },
+    {
+      "epoch": 0.03624,
+      "grad_norm": 1.7781933546066284,
+      "learning_rate": 9.96899205874504e-06,
+      "loss": 0.4382,
+      "step": 453
+    },
+    {
+      "epoch": 0.03632,
+      "grad_norm": 2.111398696899414,
+      "learning_rate": 9.968852055729229e-06,
+      "loss": 0.416,
+      "step": 454
+    },
+    {
+      "epoch": 0.0364,
+      "grad_norm": 1.8158979415893555,
+      "learning_rate": 9.968711738350773e-06,
+      "loss": 0.486,
+      "step": 455
+    },
+    {
+      "epoch": 0.03648,
+      "grad_norm": 1.4151747226715088,
+      "learning_rate": 9.968571106618551e-06,
+      "loss": 0.3059,
+      "step": 456
+    },
+    {
+      "epoch": 0.03656,
+      "grad_norm": 1.4982035160064697,
+      "learning_rate": 9.968430160541461e-06,
+      "loss": 0.282,
+      "step": 457
+    },
+    {
+      "epoch": 0.03664,
+      "grad_norm": 1.3661582469940186,
+      "learning_rate": 9.96828890012842e-06,
+      "loss": 0.3013,
+      "step": 458
+    },
+    {
+      "epoch": 0.03672,
+      "grad_norm": 1.747228980064392,
+      "learning_rate": 9.968147325388363e-06,
+      "loss": 0.3509,
+      "step": 459
+    },
+    {
+      "epoch": 0.0368,
+      "grad_norm": 1.706491231918335,
+      "learning_rate": 9.968005436330246e-06,
+      "loss": 0.4495,
+      "step": 460
+    },
+    {
+      "epoch": 0.03688,
+      "grad_norm": 1.7066677808761597,
+      "learning_rate": 9.96786323296305e-06,
+      "loss": 0.3797,
+      "step": 461
+    },
+    {
+      "epoch": 0.03696,
+      "grad_norm": 2.126333475112915,
+      "learning_rate": 9.96772071529577e-06,
+      "loss": 0.4699,
+      "step": 462
+    },
+    {
+      "epoch": 0.03704,
+      "grad_norm": 2.0231945514678955,
+      "learning_rate": 9.967577883337421e-06,
+      "loss": 0.4655,
+      "step": 463
+    },
+    {
+      "epoch": 0.03712,
+      "grad_norm": 1.7993510961532593,
+      "learning_rate": 9.967434737097043e-06,
+      "loss": 0.3497,
+      "step": 464
+    },
+    {
+      "epoch": 0.0372,
+      "grad_norm": 2.108503580093384,
+      "learning_rate": 9.967291276583688e-06,
+      "loss": 0.3369,
+      "step": 465
+    },
+    {
+      "epoch": 0.03728,
+      "grad_norm": 1.6675549745559692,
+      "learning_rate": 9.967147501806436e-06,
+      "loss": 0.4097,
+      "step": 466
+    },
+    {
+      "epoch": 0.03736,
+      "grad_norm": 1.325800895690918,
+      "learning_rate": 9.967003412774381e-06,
+      "loss": 0.2744,
+      "step": 467
+    },
+    {
+      "epoch": 0.03744,
+      "grad_norm": 1.9661016464233398,
+      "learning_rate": 9.966859009496641e-06,
+      "loss": 0.3642,
+      "step": 468
+    },
+    {
+      "epoch": 0.03752,
+      "grad_norm": 1.5406103134155273,
+      "learning_rate": 9.966714291982349e-06,
+      "loss": 0.2844,
+      "step": 469
+    },
+    {
+      "epoch": 0.0376,
+      "grad_norm": 1.752842903137207,
+      "learning_rate": 9.966569260240664e-06,
+      "loss": 0.5024,
+      "step": 470
+    },
+    {
+      "epoch": 0.03768,
+      "grad_norm": 1.2587329149246216,
+      "learning_rate": 9.966423914280758e-06,
+      "loss": 0.2523,
+      "step": 471
+    },
+    {
+      "epoch": 0.03776,
+      "grad_norm": 1.629905104637146,
+      "learning_rate": 9.96627825411183e-06,
+      "loss": 0.3229,
+      "step": 472
+    },
+    {
+      "epoch": 0.03784,
+      "grad_norm": 1.8891979455947876,
+      "learning_rate": 9.966132279743095e-06,
+      "loss": 0.3933,
+      "step": 473
+    },
+    {
+      "epoch": 0.03792,
+      "grad_norm": 1.3782880306243896,
+      "learning_rate": 9.965985991183787e-06,
+      "loss": 0.3091,
+      "step": 474
+    },
+    {
+      "epoch": 0.038,
+      "grad_norm": 1.4807788133621216,
+      "learning_rate": 9.96583938844316e-06,
+      "loss": 0.2939,
+      "step": 475
+    },
+    {
+      "epoch": 0.03808,
+      "grad_norm": 2.111734390258789,
+      "learning_rate": 9.965692471530492e-06,
+      "loss": 0.3984,
+      "step": 476
+    },
+    {
+      "epoch": 0.03816,
+      "grad_norm": 1.6450163125991821,
+      "learning_rate": 9.965545240455077e-06,
+      "loss": 0.3763,
+      "step": 477
+    },
+    {
+      "epoch": 0.03824,
+      "grad_norm": 1.8153711557388306,
+      "learning_rate": 9.965397695226228e-06,
+      "loss": 0.5331,
+      "step": 478
+    },
+    {
+      "epoch": 0.03832,
+      "grad_norm": 1.3268543481826782,
+      "learning_rate": 9.965249835853281e-06,
+      "loss": 0.3017,
+      "step": 479
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 1.573397159576416,
+      "learning_rate": 9.96510166234559e-06,
+      "loss": 0.3174,
+      "step": 480
+    },
+    {
+      "epoch": 0.03848,
+      "grad_norm": 1.8132137060165405,
+      "learning_rate": 9.964953174712533e-06,
+      "loss": 0.3517,
+      "step": 481
+    },
+    {
+      "epoch": 0.03856,
+      "grad_norm": 1.5276272296905518,
+      "learning_rate": 9.9648043729635e-06,
+      "loss": 0.3427,
+      "step": 482
+    },
+    {
+      "epoch": 0.03864,
+      "grad_norm": 1.4888263940811157,
+      "learning_rate": 9.964655257107906e-06,
+      "loss": 0.3782,
+      "step": 483
+    },
+    {
+      "epoch": 0.03872,
+      "grad_norm": 1.5089031457901,
+      "learning_rate": 9.964505827155186e-06,
+      "loss": 0.3162,
+      "step": 484
+    },
+    {
+      "epoch": 0.0388,
+      "grad_norm": 1.7417253255844116,
+      "learning_rate": 9.964356083114795e-06,
+      "loss": 0.4875,
+      "step": 485
+    },
+    {
+      "epoch": 0.03888,
+      "grad_norm": 1.554205298423767,
+      "learning_rate": 9.964206024996203e-06,
+      "loss": 0.3795,
+      "step": 486
+    },
+    {
+      "epoch": 0.03896,
+      "grad_norm": 2.758025884628296,
+      "learning_rate": 9.964055652808908e-06,
+      "loss": 0.5799,
+      "step": 487
+    },
+    {
+      "epoch": 0.03904,
+      "grad_norm": 1.5787456035614014,
+      "learning_rate": 9.96390496656242e-06,
+      "loss": 0.3235,
+      "step": 488
+    },
+    {
+      "epoch": 0.03912,
+      "grad_norm": 1.6550320386886597,
+      "learning_rate": 9.963753966266276e-06,
+      "loss": 0.3013,
+      "step": 489
+    },
+    {
+      "epoch": 0.0392,
+      "grad_norm": 1.0665961503982544,
+      "learning_rate": 9.963602651930027e-06,
+      "loss": 0.2318,
+      "step": 490
+    },
+    {
+      "epoch": 0.03928,
+      "grad_norm": 1.4649837017059326,
+      "learning_rate": 9.963451023563245e-06,
+      "loss": 0.3583,
+      "step": 491
+    },
+    {
+      "epoch": 0.03936,
+      "grad_norm": 1.462937831878662,
+      "learning_rate": 9.963299081175525e-06,
+      "loss": 0.308,
+      "step": 492
+    },
+    {
+      "epoch": 0.03944,
+      "grad_norm": 1.6257414817810059,
+      "learning_rate": 9.96314682477648e-06,
+      "loss": 0.3596,
+      "step": 493
+    },
+    {
+      "epoch": 0.03952,
+      "grad_norm": 2.3487870693206787,
+      "learning_rate": 9.962994254375742e-06,
+      "loss": 0.4331,
+      "step": 494
+    },
+    {
+      "epoch": 0.0396,
+      "grad_norm": 2.4731879234313965,
+      "learning_rate": 9.962841369982962e-06,
+      "loss": 0.4484,
+      "step": 495
+    },
+    {
+      "epoch": 0.03968,
+      "grad_norm": 1.8673752546310425,
+      "learning_rate": 9.962688171607817e-06,
+      "loss": 0.3284,
+      "step": 496
+    },
+    {
+      "epoch": 0.03976,
+      "grad_norm": 1.7251307964324951,
+      "learning_rate": 9.962534659259995e-06,
+      "loss": 0.2836,
+      "step": 497
+    },
+    {
+      "epoch": 0.03984,
+      "grad_norm": 2.203010320663452,
+      "learning_rate": 9.96238083294921e-06,
+      "loss": 0.3871,
+      "step": 498
+    },
+    {
+      "epoch": 0.03992,
+      "grad_norm": 1.6219152212142944,
+      "learning_rate": 9.962226692685195e-06,
+      "loss": 0.333,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6756365299224854,
+      "learning_rate": 9.962072238477699e-06,
+      "loss": 0.2864,
+      "step": 500
+    },
+    {
+      "epoch": 0.04008,
+      "grad_norm": 1.6424753665924072,
+      "learning_rate": 9.961917470336496e-06,
+      "loss": 0.3253,
+      "step": 501
+    },
+    {
+      "epoch": 0.04016,
+      "grad_norm": 1.5344758033752441,
+      "learning_rate": 9.961762388271378e-06,
+      "loss": 0.3293,
+      "step": 502
+    },
+    {
+      "epoch": 0.04024,
+      "grad_norm": 1.1364933252334595,
+      "learning_rate": 9.961606992292155e-06,
+      "loss": 0.2658,
+      "step": 503
+    },
+    {
+      "epoch": 0.04032,
+      "grad_norm": 1.397683024406433,
+      "learning_rate": 9.96145128240866e-06,
+      "loss": 0.2682,
+      "step": 504
+    },
+    {
+      "epoch": 0.0404,
+      "grad_norm": 1.7218232154846191,
+      "learning_rate": 9.961295258630743e-06,
+      "loss": 0.3327,
+      "step": 505
+    },
+    {
+      "epoch": 0.04048,
+      "grad_norm": 1.204432487487793,
+      "learning_rate": 9.961138920968276e-06,
+      "loss": 0.2728,
+      "step": 506
+    },
+    {
+      "epoch": 0.04056,
+      "grad_norm": 1.273925542831421,
+      "learning_rate": 9.960982269431149e-06,
+      "loss": 0.2972,
+      "step": 507
+    },
+    {
+      "epoch": 0.04064,
+      "grad_norm": 1.682013750076294,
+      "learning_rate": 9.960825304029274e-06,
+      "loss": 0.3276,
+      "step": 508
+    },
+    {
+      "epoch": 0.04072,
+      "grad_norm": 1.4792194366455078,
+      "learning_rate": 9.96066802477258e-06,
+      "loss": 0.4057,
+      "step": 509
+    },
+    {
+      "epoch": 0.0408,
+      "grad_norm": 2.17596173286438,
+      "learning_rate": 9.96051043167102e-06,
+      "loss": 0.5181,
+      "step": 510
+    },
+    {
+      "epoch": 0.04088,
+      "grad_norm": 1.6542383432388306,
+      "learning_rate": 9.960352524734562e-06,
+      "loss": 0.3988,
+      "step": 511
+    },
+    {
+      "epoch": 0.04096,
+      "grad_norm": 1.9632936716079712,
+      "learning_rate": 9.960194303973196e-06,
+      "loss": 0.3267,
+      "step": 512
+    },
+    {
+      "epoch": 0.04104,
+      "grad_norm": 1.4410243034362793,
+      "learning_rate": 9.960035769396935e-06,
+      "loss": 0.2683,
+      "step": 513
+    },
+    {
+      "epoch": 0.04112,
+      "grad_norm": 1.5401276350021362,
+      "learning_rate": 9.959876921015805e-06,
+      "loss": 0.2925,
+      "step": 514
+    },
+    {
+      "epoch": 0.0412,
+      "grad_norm": 1.6564226150512695,
+      "learning_rate": 9.95971775883986e-06,
+      "loss": 0.33,
+      "step": 515
+    },
+    {
+      "epoch": 0.04128,
+      "grad_norm": 1.4229389429092407,
+      "learning_rate": 9.959558282879167e-06,
+      "loss": 0.2565,
+      "step": 516
+    },
+    {
+      "epoch": 0.04136,
+      "grad_norm": 1.524813175201416,
+      "learning_rate": 9.959398493143816e-06,
+      "loss": 0.3145,
+      "step": 517
+    },
+    {
+      "epoch": 0.04144,
+      "grad_norm": 1.274613380432129,
+      "learning_rate": 9.959238389643918e-06,
+      "loss": 0.2743,
+      "step": 518
+    },
+    {
+      "epoch": 0.04152,
+      "grad_norm": 1.9126328229904175,
+      "learning_rate": 9.9590779723896e-06,
+      "loss": 0.3763,
+      "step": 519
+    },
+    {
+      "epoch": 0.0416,
+      "grad_norm": 1.5458030700683594,
+      "learning_rate": 9.95891724139101e-06,
+      "loss": 0.313,
+      "step": 520
+    },
+    {
+      "epoch": 0.04168,
+      "grad_norm": 1.80939781665802,
+      "learning_rate": 9.958756196658321e-06,
+      "loss": 0.3907,
+      "step": 521
+    },
+    {
+      "epoch": 0.04176,
+      "grad_norm": 1.551069974899292,
+      "learning_rate": 9.958594838201719e-06,
+      "loss": 0.3031,
+      "step": 522
+    },
+    {
+      "epoch": 0.04184,
+      "grad_norm": 1.9089337587356567,
+      "learning_rate": 9.958433166031412e-06,
+      "loss": 0.3787,
+      "step": 523
+    },
+    {
+      "epoch": 0.04192,
+      "grad_norm": 1.5611118078231812,
+      "learning_rate": 9.958271180157631e-06,
+      "loss": 0.3589,
+      "step": 524
+    },
+    {
+      "epoch": 0.042,
+      "grad_norm": 1.8825455904006958,
+      "learning_rate": 9.958108880590623e-06,
+      "loss": 0.4088,
+      "step": 525
+    },
+    {
+      "epoch": 0.04208,
+      "grad_norm": 1.9425339698791504,
+      "learning_rate": 9.957946267340655e-06,
+      "loss": 0.2773,
+      "step": 526
+    },
+    {
+      "epoch": 0.04216,
+      "grad_norm": 1.547389030456543,
+      "learning_rate": 9.957783340418015e-06,
+      "loss": 0.2913,
+      "step": 527
+    },
+    {
+      "epoch": 0.04224,
+      "grad_norm": 1.622795820236206,
+      "learning_rate": 9.957620099833013e-06,
+      "loss": 0.3456,
+      "step": 528
+    },
+    {
+      "epoch": 0.04232,
+      "grad_norm": 1.9623744487762451,
+      "learning_rate": 9.957456545595977e-06,
+      "loss": 0.3924,
+      "step": 529
+    },
+    {
+      "epoch": 0.0424,
+      "grad_norm": 1.489332675933838,
+      "learning_rate": 9.95729267771725e-06,
+      "loss": 0.3155,
+      "step": 530
+    },
+    {
+      "epoch": 0.04248,
+      "grad_norm": 1.6495463848114014,
+      "learning_rate": 9.957128496207205e-06,
+      "loss": 0.3544,
+      "step": 531
+    },
+    {
+      "epoch": 0.04256,
+      "grad_norm": 1.5633338689804077,
+      "learning_rate": 9.956964001076224e-06,
+      "loss": 0.3408,
+      "step": 532
+    },
+    {
+      "epoch": 0.04264,
+      "grad_norm": 2.2354142665863037,
+      "learning_rate": 9.95679919233472e-06,
+      "loss": 0.4222,
+      "step": 533
+    },
+    {
+      "epoch": 0.04272,
+      "grad_norm": 1.6407157182693481,
+      "learning_rate": 9.956634069993114e-06,
+      "loss": 0.4222,
+      "step": 534
+    },
+    {
+      "epoch": 0.0428,
+      "grad_norm": 1.805367112159729,
+      "learning_rate": 9.956468634061857e-06,
+      "loss": 0.3485,
+      "step": 535
+    },
+    {
+      "epoch": 0.04288,
+      "grad_norm": 1.9591755867004395,
+      "learning_rate": 9.956302884551413e-06,
+      "loss": 0.397,
+      "step": 536
+    },
+    {
+      "epoch": 0.04296,
+      "grad_norm": 2.0101656913757324,
+      "learning_rate": 9.956136821472269e-06,
+      "loss": 0.5719,
+      "step": 537
+    },
+    {
+      "epoch": 0.04304,
+      "grad_norm": 2.0787875652313232,
+      "learning_rate": 9.955970444834933e-06,
+      "loss": 0.4123,
+      "step": 538
+    },
+    {
+      "epoch": 0.04312,
+      "grad_norm": 1.7366102933883667,
+      "learning_rate": 9.955803754649929e-06,
+      "loss": 0.3547,
+      "step": 539
+    },
+    {
+      "epoch": 0.0432,
+      "grad_norm": 1.8972814083099365,
+      "learning_rate": 9.955636750927803e-06,
+      "loss": 0.3922,
+      "step": 540
+    },
+    {
+      "epoch": 0.04328,
+      "grad_norm": 1.79804265499115,
+      "learning_rate": 9.95546943367912e-06,
+      "loss": 0.4213,
+      "step": 541
+    },
+    {
+      "epoch": 0.04336,
+      "grad_norm": 1.6599233150482178,
+      "learning_rate": 9.955301802914471e-06,
+      "loss": 0.3559,
+      "step": 542
+    },
+    {
+      "epoch": 0.04344,
+      "grad_norm": 1.4300049543380737,
+      "learning_rate": 9.955133858644455e-06,
+      "loss": 0.3159,
+      "step": 543
+    },
+    {
+      "epoch": 0.04352,
+      "grad_norm": 1.2100526094436646,
+      "learning_rate": 9.9549656008797e-06,
+      "loss": 0.2633,
+      "step": 544
+    },
+    {
+      "epoch": 0.0436,
+      "grad_norm": 1.8038049936294556,
+      "learning_rate": 9.95479702963085e-06,
+      "loss": 0.4075,
+      "step": 545
+    },
+    {
+      "epoch": 0.04368,
+      "grad_norm": 2.2004456520080566,
+      "learning_rate": 9.954628144908573e-06,
+      "loss": 0.516,
+      "step": 546
+    },
+    {
+      "epoch": 0.04376,
+      "grad_norm": 1.4208917617797852,
+      "learning_rate": 9.95445894672355e-06,
+      "loss": 0.3567,
+      "step": 547
+    },
+    {
+      "epoch": 0.04384,
+      "grad_norm": 1.8718032836914062,
+      "learning_rate": 9.954289435086487e-06,
+      "loss": 0.4532,
+      "step": 548
+    },
+    {
+      "epoch": 0.04392,
+      "grad_norm": 1.7566274404525757,
+      "learning_rate": 9.954119610008108e-06,
+      "loss": 0.3248,
+      "step": 549
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 1.7727574110031128,
+      "learning_rate": 9.953949471499157e-06,
+      "loss": 0.3495,
+      "step": 550
+    },
+    {
+      "epoch": 0.04408,
+      "grad_norm": 1.5641764402389526,
+      "learning_rate": 9.953779019570402e-06,
+      "loss": 0.3757,
+      "step": 551
+    },
+    {
+      "epoch": 0.04416,
+      "grad_norm": 1.4524203538894653,
+      "learning_rate": 9.953608254232622e-06,
+      "loss": 0.3353,
+      "step": 552
+    },
+    {
+      "epoch": 0.04424,
+      "grad_norm": 1.6970480680465698,
+      "learning_rate": 9.953437175496622e-06,
+      "loss": 0.3662,
+      "step": 553
+    },
+    {
+      "epoch": 0.04432,
+      "grad_norm": 2.238062858581543,
+      "learning_rate": 9.953265783373227e-06,
+      "loss": 0.4972,
+      "step": 554
+    },
+    {
+      "epoch": 0.0444,
+      "grad_norm": 1.676147699356079,
+      "learning_rate": 9.95309407787328e-06,
+      "loss": 0.3475,
+      "step": 555
+    },
+    {
+      "epoch": 0.04448,
+      "grad_norm": 1.4633108377456665,
+      "learning_rate": 9.952922059007643e-06,
+      "loss": 0.3406,
+      "step": 556
+    },
+    {
+      "epoch": 0.04456,
+      "grad_norm": 1.5241601467132568,
+      "learning_rate": 9.952749726787201e-06,
+      "loss": 0.2829,
+      "step": 557
+    },
+    {
+      "epoch": 0.04464,
+      "grad_norm": 2.0277199745178223,
+      "learning_rate": 9.952577081222854e-06,
+      "loss": 0.3451,
+      "step": 558
+    },
+    {
+      "epoch": 0.04472,
+      "grad_norm": 1.724655032157898,
+      "learning_rate": 9.952404122325528e-06,
+      "loss": 0.4065,
+      "step": 559
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 1.4190020561218262,
+      "learning_rate": 9.952230850106164e-06,
+      "loss": 0.3292,
+      "step": 560
+    },
+    {
+      "epoch": 0.04488,
+      "grad_norm": 1.4401750564575195,
+      "learning_rate": 9.952057264575723e-06,
+      "loss": 0.3097,
+      "step": 561
+    },
+    {
+      "epoch": 0.04496,
+      "grad_norm": 2.137566328048706,
+      "learning_rate": 9.95188336574519e-06,
+      "loss": 0.4065,
+      "step": 562
+    },
+    {
+      "epoch": 0.04504,
+      "grad_norm": 1.5885168313980103,
+      "learning_rate": 9.951709153625564e-06,
+      "loss": 0.2592,
+      "step": 563
+    },
+    {
+      "epoch": 0.04512,
+      "grad_norm": 1.431435227394104,
+      "learning_rate": 9.951534628227868e-06,
+      "loss": 0.4273,
+      "step": 564
+    },
+    {
+      "epoch": 0.0452,
+      "grad_norm": 1.6501264572143555,
+      "learning_rate": 9.951359789563145e-06,
+      "loss": 0.3777,
+      "step": 565
+    },
+    {
+      "epoch": 0.04528,
+      "grad_norm": 2.1749095916748047,
+      "learning_rate": 9.951184637642456e-06,
+      "loss": 0.461,
+      "step": 566
+    },
+    {
+      "epoch": 0.04536,
+      "grad_norm": 1.7694364786148071,
+      "learning_rate": 9.95100917247688e-06,
+      "loss": 0.3626,
+      "step": 567
+    },
+    {
+      "epoch": 0.04544,
+      "grad_norm": 1.836769938468933,
+      "learning_rate": 9.950833394077522e-06,
+      "loss": 0.3411,
+      "step": 568
+    },
+    {
+      "epoch": 0.04552,
+      "grad_norm": 1.6859595775604248,
+      "learning_rate": 9.9506573024555e-06,
+      "loss": 0.3591,
+      "step": 569
+    },
+    {
+      "epoch": 0.0456,
+      "grad_norm": 2.330047130584717,
+      "learning_rate": 9.950480897621955e-06,
+      "loss": 0.6001,
+      "step": 570
+    },
+    {
+      "epoch": 0.04568,
+      "grad_norm": 1.5222753286361694,
+      "learning_rate": 9.950304179588047e-06,
+      "loss": 0.3455,
+      "step": 571
+    },
+    {
+      "epoch": 0.04576,
+      "grad_norm": 1.7251989841461182,
+      "learning_rate": 9.950127148364958e-06,
+      "loss": 0.365,
+      "step": 572
+    },
+    {
+      "epoch": 0.04584,
+      "grad_norm": 1.6580973863601685,
+      "learning_rate": 9.949949803963887e-06,
+      "loss": 0.3802,
+      "step": 573
+    },
+    {
+      "epoch": 0.04592,
+      "grad_norm": 1.8021025657653809,
+      "learning_rate": 9.949772146396056e-06,
+      "loss": 0.4321,
+      "step": 574
+    },
+    {
+      "epoch": 0.046,
+      "grad_norm": 1.5739644765853882,
+      "learning_rate": 9.949594175672703e-06,
+      "loss": 0.3246,
+      "step": 575
+    },
+    {
+      "epoch": 0.04608,
+      "grad_norm": 2.065218925476074,
+      "learning_rate": 9.949415891805087e-06,
+      "loss": 0.4126,
+      "step": 576
+    },
+    {
+      "epoch": 0.04616,
+      "grad_norm": 1.7216342687606812,
+      "learning_rate": 9.949237294804489e-06,
+      "loss": 0.4876,
+      "step": 577
+    },
+    {
+      "epoch": 0.04624,
+      "grad_norm": 1.3691469430923462,
+      "learning_rate": 9.949058384682206e-06,
+      "loss": 0.2763,
+      "step": 578
+    },
+    {
+      "epoch": 0.04632,
+      "grad_norm": 1.4548450708389282,
+      "learning_rate": 9.94887916144956e-06,
+      "loss": 0.3798,
+      "step": 579
+    },
+    {
+      "epoch": 0.0464,
+      "grad_norm": 1.7984998226165771,
+      "learning_rate": 9.948699625117888e-06,
+      "loss": 0.4941,
+      "step": 580
+    },
+    {
+      "epoch": 0.04648,
+      "grad_norm": 1.7592239379882812,
+      "learning_rate": 9.948519775698551e-06,
+      "loss": 0.3491,
+      "step": 581
+    },
+    {
+      "epoch": 0.04656,
+      "grad_norm": 1.7368760108947754,
+      "learning_rate": 9.948339613202923e-06,
+      "loss": 0.401,
+      "step": 582
+    },
+    {
+      "epoch": 0.04664,
+      "grad_norm": 1.7027363777160645,
+      "learning_rate": 9.948159137642407e-06,
+      "loss": 0.38,
+      "step": 583
+    },
+    {
+      "epoch": 0.04672,
+      "grad_norm": 2.045701503753662,
+      "learning_rate": 9.947978349028418e-06,
+      "loss": 0.4475,
+      "step": 584
+    },
+    {
+      "epoch": 0.0468,
+      "grad_norm": 1.5346065759658813,
+      "learning_rate": 9.947797247372394e-06,
+      "loss": 0.3696,
+      "step": 585
+    },
+    {
+      "epoch": 0.04688,
+      "grad_norm": 1.5219029188156128,
+      "learning_rate": 9.947615832685795e-06,
+      "loss": 0.2784,
+      "step": 586
+    },
+    {
+      "epoch": 0.04696,
+      "grad_norm": 1.4893606901168823,
+      "learning_rate": 9.947434104980097e-06,
+      "loss": 0.33,
+      "step": 587
+    },
+    {
+      "epoch": 0.04704,
+      "grad_norm": 1.7347660064697266,
+      "learning_rate": 9.947252064266796e-06,
+      "loss": 0.3985,
+      "step": 588
+    },
+    {
+      "epoch": 0.04712,
+      "grad_norm": 1.6877951622009277,
+      "learning_rate": 9.947069710557412e-06,
+      "loss": 0.3609,
+      "step": 589
+    },
+    {
+      "epoch": 0.0472,
+      "grad_norm": 1.5663824081420898,
+      "learning_rate": 9.946887043863478e-06,
+      "loss": 0.3961,
+      "step": 590
+    },
+    {
+      "epoch": 0.04728,
+      "grad_norm": 1.6489958763122559,
+      "learning_rate": 9.946704064196554e-06,
+      "loss": 0.4324,
+      "step": 591
+    },
+    {
+      "epoch": 0.04736,
+      "grad_norm": 1.8338820934295654,
+      "learning_rate": 9.946520771568217e-06,
+      "loss": 0.3133,
+      "step": 592
+    },
+    {
+      "epoch": 0.04744,
+      "grad_norm": 1.592786192893982,
+      "learning_rate": 9.946337165990061e-06,
+      "loss": 0.3564,
+      "step": 593
+    },
+    {
+      "epoch": 0.04752,
+      "grad_norm": 1.8404992818832397,
+      "learning_rate": 9.946153247473702e-06,
+      "loss": 0.3641,
+      "step": 594
+    },
+    {
+      "epoch": 0.0476,
+      "grad_norm": 1.990356206893921,
+      "learning_rate": 9.945969016030779e-06,
+      "loss": 0.4124,
+      "step": 595
+    },
+    {
+      "epoch": 0.04768,
+      "grad_norm": 1.7861665487289429,
+      "learning_rate": 9.945784471672943e-06,
+      "loss": 0.3725,
+      "step": 596
+    },
+    {
+      "epoch": 0.04776,
+      "grad_norm": 2.0100088119506836,
+      "learning_rate": 9.945599614411875e-06,
+      "loss": 0.4765,
+      "step": 597
+    },
+    {
+      "epoch": 0.04784,
+      "grad_norm": 1.882002830505371,
+      "learning_rate": 9.945414444259267e-06,
+      "loss": 0.3974,
+      "step": 598
+    },
+    {
+      "epoch": 0.04792,
+      "grad_norm": 1.8743993043899536,
+      "learning_rate": 9.945228961226832e-06,
+      "loss": 0.3773,
+      "step": 599
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 1.5349911451339722,
+      "learning_rate": 9.94504316532631e-06,
+      "loss": 0.2909,
+      "step": 600
+    },
+    {
+      "epoch": 0.04808,
+      "grad_norm": 1.8333827257156372,
+      "learning_rate": 9.944857056569452e-06,
+      "loss": 0.3577,
+      "step": 601
+    },
+    {
+      "epoch": 0.04816,
+      "grad_norm": 1.9618743658065796,
+      "learning_rate": 9.944670634968033e-06,
+      "loss": 0.4357,
+      "step": 602
+    },
+    {
+      "epoch": 0.04824,
+      "grad_norm": 1.9149047136306763,
+      "learning_rate": 9.944483900533848e-06,
+      "loss": 0.4081,
+      "step": 603
+    },
+    {
+      "epoch": 0.04832,
+      "grad_norm": 2.0370891094207764,
+      "learning_rate": 9.944296853278712e-06,
+      "loss": 0.3549,
+      "step": 604
+    },
+    {
+      "epoch": 0.0484,
+      "grad_norm": 2.002598524093628,
+      "learning_rate": 9.944109493214458e-06,
+      "loss": 0.3807,
+      "step": 605
+    },
+    {
+      "epoch": 0.04848,
+      "grad_norm": 1.196305274963379,
+      "learning_rate": 9.943921820352938e-06,
+      "loss": 0.2734,
+      "step": 606
+    },
+    {
+      "epoch": 0.04856,
+      "grad_norm": 1.8980247974395752,
+      "learning_rate": 9.943733834706027e-06,
+      "loss": 0.45,
+      "step": 607
+    },
+    {
+      "epoch": 0.04864,
+      "grad_norm": 1.8210073709487915,
+      "learning_rate": 9.94354553628562e-06,
+      "loss": 0.4916,
+      "step": 608
+    },
+    {
+      "epoch": 0.04872,
+      "grad_norm": 1.4517135620117188,
+      "learning_rate": 9.943356925103625e-06,
+      "loss": 0.3107,
+      "step": 609
+    },
+    {
+      "epoch": 0.0488,
+      "grad_norm": 1.8347890377044678,
+      "learning_rate": 9.94316800117198e-06,
+      "loss": 0.4305,
+      "step": 610
+    },
+    {
+      "epoch": 0.04888,
+      "grad_norm": 1.664490818977356,
+      "learning_rate": 9.942978764502635e-06,
+      "loss": 0.3694,
+      "step": 611
+    },
+    {
+      "epoch": 0.04896,
+      "grad_norm": 1.5147761106491089,
+      "learning_rate": 9.942789215107562e-06,
+      "loss": 0.3826,
+      "step": 612
+    },
+    {
+      "epoch": 0.04904,
+      "grad_norm": 1.5997848510742188,
+      "learning_rate": 9.942599352998756e-06,
+      "loss": 0.3296,
+      "step": 613
+    },
+    {
+      "epoch": 0.04912,
+      "grad_norm": 1.7910867929458618,
+      "learning_rate": 9.942409178188225e-06,
+      "loss": 0.354,
+      "step": 614
+    },
+    {
+      "epoch": 0.0492,
+      "grad_norm": 1.7603079080581665,
+      "learning_rate": 9.942218690688004e-06,
+      "loss": 0.486,
+      "step": 615
+    },
+    {
+      "epoch": 0.04928,
+      "grad_norm": 1.4481785297393799,
+      "learning_rate": 9.942027890510142e-06,
+      "loss": 0.4026,
+      "step": 616
+    },
+    {
+      "epoch": 0.04936,
+      "grad_norm": 1.7464784383773804,
+      "learning_rate": 9.94183677766671e-06,
+      "loss": 0.3778,
+      "step": 617
+    },
+    {
+      "epoch": 0.04944,
+      "grad_norm": 1.5305639505386353,
+      "learning_rate": 9.941645352169804e-06,
+      "loss": 0.3281,
+      "step": 618
+    },
+    {
+      "epoch": 0.04952,
+      "grad_norm": 1.8555560111999512,
+      "learning_rate": 9.941453614031528e-06,
+      "loss": 0.3575,
+      "step": 619
+    },
+    {
+      "epoch": 0.0496,
+      "grad_norm": 1.9013909101486206,
+      "learning_rate": 9.941261563264019e-06,
+      "loss": 0.3648,
+      "step": 620
+    },
+    {
+      "epoch": 0.04968,
+      "grad_norm": 1.7493908405303955,
+      "learning_rate": 9.941069199879424e-06,
+      "loss": 0.3552,
+      "step": 621
+    },
+    {
+      "epoch": 0.04976,
+      "grad_norm": 1.6803377866744995,
+      "learning_rate": 9.940876523889911e-06,
+      "loss": 0.4728,
+      "step": 622
+    },
+    {
+      "epoch": 0.04984,
+      "grad_norm": 1.9179245233535767,
+      "learning_rate": 9.940683535307675e-06,
+      "loss": 0.357,
+      "step": 623
+    },
+    {
+      "epoch": 0.04992,
+      "grad_norm": 2.1097536087036133,
+      "learning_rate": 9.940490234144923e-06,
+      "loss": 0.4825,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8075186014175415,
+      "learning_rate": 9.940296620413884e-06,
+      "loss": 0.4851,
+      "step": 625
+    },
+    {
+      "epoch": 0.05008,
+      "grad_norm": 1.5507980585098267,
+      "learning_rate": 9.940102694126809e-06,
+      "loss": 0.3254,
+      "step": 626
+    },
+    {
+      "epoch": 0.05016,
+      "grad_norm": 2.0220351219177246,
+      "learning_rate": 9.939908455295965e-06,
+      "loss": 0.405,
+      "step": 627
+    },
+    {
+      "epoch": 0.05024,
+      "grad_norm": 1.5802489519119263,
+      "learning_rate": 9.939713903933644e-06,
+      "loss": 0.3536,
+      "step": 628
+    },
+    {
+      "epoch": 0.05032,
+      "grad_norm": 1.7642475366592407,
+      "learning_rate": 9.93951904005215e-06,
+      "loss": 0.3986,
+      "step": 629
+    },
+    {
+      "epoch": 0.0504,
+      "grad_norm": 1.5196278095245361,
+      "learning_rate": 9.939323863663817e-06,
+      "loss": 0.3779,
+      "step": 630
+    },
+    {
+      "epoch": 0.05048,
+      "grad_norm": 1.995604395866394,
+      "learning_rate": 9.939128374780987e-06,
+      "loss": 0.4138,
+      "step": 631
+    },
+    {
+      "epoch": 0.05056,
+      "grad_norm": 1.300185203552246,
+      "learning_rate": 9.938932573416033e-06,
+      "loss": 0.3161,
+      "step": 632
+    },
+    {
+      "epoch": 0.05064,
+      "grad_norm": 1.9059289693832397,
+      "learning_rate": 9.938736459581341e-06,
+      "loss": 0.3545,
+      "step": 633
+    },
+    {
+      "epoch": 0.05072,
+      "grad_norm": 2.062704563140869,
+      "learning_rate": 9.938540033289317e-06,
+      "loss": 0.3853,
+      "step": 634
+    },
+    {
+      "epoch": 0.0508,
+      "grad_norm": 1.7212554216384888,
+      "learning_rate": 9.93834329455239e-06,
+      "loss": 0.3571,
+      "step": 635
+    },
+    {
+      "epoch": 0.05088,
+      "grad_norm": 1.2724170684814453,
+      "learning_rate": 9.938146243383006e-06,
+      "loss": 0.3076,
+      "step": 636
+    },
+    {
+      "epoch": 0.05096,
+      "grad_norm": 1.4473823308944702,
+      "learning_rate": 9.937948879793634e-06,
+      "loss": 0.328,
+      "step": 637
+    },
+    {
+      "epoch": 0.05104,
+      "grad_norm": 1.304731011390686,
+      "learning_rate": 9.937751203796758e-06,
+      "loss": 0.3388,
+      "step": 638
+    },
+    {
+      "epoch": 0.05112,
+      "grad_norm": 1.8237764835357666,
+      "learning_rate": 9.937553215404884e-06,
+      "loss": 0.3978,
+      "step": 639
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 1.4053940773010254,
+      "learning_rate": 9.937354914630542e-06,
+      "loss": 0.2694,
+      "step": 640
+    },
+    {
+      "epoch": 0.05128,
+      "grad_norm": 1.79597806930542,
+      "learning_rate": 9.937156301486273e-06,
+      "loss": 0.4567,
+      "step": 641
+    },
+    {
+      "epoch": 0.05136,
+      "grad_norm": 1.2079006433486938,
+      "learning_rate": 9.936957375984644e-06,
+      "loss": 0.2878,
+      "step": 642
+    },
+    {
+      "epoch": 0.05144,
+      "grad_norm": 1.547292947769165,
+      "learning_rate": 9.93675813813824e-06,
+      "loss": 0.3161,
+      "step": 643
+    },
+    {
+      "epoch": 0.05152,
+      "grad_norm": 1.8033218383789062,
+      "learning_rate": 9.936558587959671e-06,
+      "loss": 0.3995,
+      "step": 644
+    },
+    {
+      "epoch": 0.0516,
+      "grad_norm": 1.5613384246826172,
+      "learning_rate": 9.936358725461555e-06,
+      "loss": 0.2709,
+      "step": 645
+    },
+    {
+      "epoch": 0.05168,
+      "grad_norm": 1.7098636627197266,
+      "learning_rate": 9.936158550656539e-06,
+      "loss": 0.3536,
+      "step": 646
+    },
+    {
+      "epoch": 0.05176,
+      "grad_norm": 1.7437068223953247,
+      "learning_rate": 9.93595806355729e-06,
+      "loss": 0.3261,
+      "step": 647
+    },
+    {
+      "epoch": 0.05184,
+      "grad_norm": 1.75038480758667,
+      "learning_rate": 9.93575726417649e-06,
+      "loss": 0.3398,
+      "step": 648
+    },
+    {
+      "epoch": 0.05192,
+      "grad_norm": 1.4043598175048828,
+      "learning_rate": 9.93555615252684e-06,
+      "loss": 0.3377,
+      "step": 649
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 1.877252221107483,
+      "learning_rate": 9.935354728621069e-06,
+      "loss": 0.3839,
+      "step": 650
+    },
+    {
+      "epoch": 0.05208,
+      "grad_norm": 1.5518877506256104,
+      "learning_rate": 9.935152992471918e-06,
+      "loss": 0.4235,
+      "step": 651
+    },
+    {
+      "epoch": 0.05216,
+      "grad_norm": 1.384873390197754,
+      "learning_rate": 9.93495094409215e-06,
+      "loss": 0.3017,
+      "step": 652
+    },
+    {
+      "epoch": 0.05224,
+      "grad_norm": 2.235605001449585,
+      "learning_rate": 9.93474858349455e-06,
+      "loss": 0.3975,
+      "step": 653
+    },
+    {
+      "epoch": 0.05232,
+      "grad_norm": 2.175873279571533,
+      "learning_rate": 9.934545910691914e-06,
+      "loss": 0.4333,
+      "step": 654
+    },
+    {
+      "epoch": 0.0524,
+      "grad_norm": 2.141580581665039,
+      "learning_rate": 9.934342925697074e-06,
+      "loss": 0.4373,
+      "step": 655
+    },
+    {
+      "epoch": 0.05248,
+      "grad_norm": 1.8927725553512573,
+      "learning_rate": 9.934139628522865e-06,
+      "loss": 0.3006,
+      "step": 656
+    },
+    {
+      "epoch": 0.05256,
+      "grad_norm": 1.7776702642440796,
+      "learning_rate": 9.933936019182154e-06,
+      "loss": 0.4212,
+      "step": 657
+    },
+    {
+      "epoch": 0.05264,
+      "grad_norm": 1.5397083759307861,
+      "learning_rate": 9.933732097687817e-06,
+      "loss": 0.3998,
+      "step": 658
+    },
+    {
+      "epoch": 0.05272,
+      "grad_norm": 1.476570963859558,
+      "learning_rate": 9.93352786405276e-06,
+      "loss": 0.2903,
+      "step": 659
+    },
+    {
+      "epoch": 0.0528,
+      "grad_norm": 2.0092062950134277,
+      "learning_rate": 9.933323318289902e-06,
+      "loss": 0.3786,
+      "step": 660
+    },
+    {
+      "epoch": 0.05288,
+      "grad_norm": 1.8279545307159424,
+      "learning_rate": 9.933118460412186e-06,
+      "loss": 0.3953,
+      "step": 661
+    },
+    {
+      "epoch": 0.05296,
+      "grad_norm": 1.9125584363937378,
+      "learning_rate": 9.93291329043257e-06,
+      "loss": 0.3686,
+      "step": 662
+    },
+    {
+      "epoch": 0.05304,
+      "grad_norm": 1.951283574104309,
+      "learning_rate": 9.932707808364035e-06,
+      "loss": 0.4019,
+      "step": 663
+    },
+    {
+      "epoch": 0.05312,
+      "grad_norm": 1.3012398481369019,
+      "learning_rate": 9.932502014219583e-06,
+      "loss": 0.2866,
+      "step": 664
+    },
+    {
+      "epoch": 0.0532,
+      "grad_norm": 1.913560390472412,
+      "learning_rate": 9.932295908012233e-06,
+      "loss": 0.3671,
+      "step": 665
+    },
+    {
+      "epoch": 0.05328,
+      "grad_norm": 1.9930202960968018,
+      "learning_rate": 9.932089489755024e-06,
+      "loss": 0.4171,
+      "step": 666
+    },
+    {
+      "epoch": 0.05336,
+      "grad_norm": 1.7586143016815186,
+      "learning_rate": 9.931882759461016e-06,
+      "loss": 0.4016,
+      "step": 667
+    },
+    {
+      "epoch": 0.05344,
+      "grad_norm": 1.564494013786316,
+      "learning_rate": 9.931675717143288e-06,
+      "loss": 0.3041,
+      "step": 668
+    },
+    {
+      "epoch": 0.05352,
+      "grad_norm": 1.6268788576126099,
+      "learning_rate": 9.931468362814937e-06,
+      "loss": 0.3104,
+      "step": 669
+    },
+    {
+      "epoch": 0.0536,
+      "grad_norm": 1.5686761140823364,
+      "learning_rate": 9.931260696489085e-06,
+      "loss": 0.2989,
+      "step": 670
+    },
+    {
+      "epoch": 0.05368,
+      "grad_norm": 1.2307571172714233,
+      "learning_rate": 9.931052718178869e-06,
+      "loss": 0.2887,
+      "step": 671
+    },
+    {
+      "epoch": 0.05376,
+      "grad_norm": 1.586047887802124,
+      "learning_rate": 9.930844427897447e-06,
+      "loss": 0.3362,
+      "step": 672
+    },
+    {
+      "epoch": 0.05384,
+      "grad_norm": 1.9724785089492798,
+      "learning_rate": 9.930635825657996e-06,
+      "loss": 0.3575,
+      "step": 673
+    },
+    {
+      "epoch": 0.05392,
+      "grad_norm": 1.7712501287460327,
+      "learning_rate": 9.930426911473715e-06,
+      "loss": 0.3853,
+      "step": 674
+    },
+    {
+      "epoch": 0.054,
+      "grad_norm": 1.4114841222763062,
+      "learning_rate": 9.93021768535782e-06,
+      "loss": 0.3587,
+      "step": 675
+    },
+    {
+      "epoch": 0.05408,
+      "grad_norm": 2.0432698726654053,
+      "learning_rate": 9.93000814732355e-06,
+      "loss": 0.4619,
+      "step": 676
+    },
+    {
+      "epoch": 0.05416,
+      "grad_norm": 1.5945602655410767,
+      "learning_rate": 9.92979829738416e-06,
+      "loss": 0.2833,
+      "step": 677
+    },
+    {
+      "epoch": 0.05424,
+      "grad_norm": 2.2913811206817627,
+      "learning_rate": 9.929588135552925e-06,
+      "loss": 0.4524,
+      "step": 678
+    },
+    {
+      "epoch": 0.05432,
+      "grad_norm": 1.2987146377563477,
+      "learning_rate": 9.929377661843143e-06,
+      "loss": 0.2484,
+      "step": 679
+    },
+    {
+      "epoch": 0.0544,
+      "grad_norm": 1.3887112140655518,
+      "learning_rate": 9.929166876268132e-06,
+      "loss": 0.3099,
+      "step": 680
+    },
+    {
+      "epoch": 0.05448,
+      "grad_norm": 1.4816182851791382,
+      "learning_rate": 9.928955778841224e-06,
+      "loss": 0.3065,
+      "step": 681
+    },
+    {
+      "epoch": 0.05456,
+      "grad_norm": 1.4863247871398926,
+      "learning_rate": 9.928744369575778e-06,
+      "loss": 0.3439,
+      "step": 682
+    },
+    {
+      "epoch": 0.05464,
+      "grad_norm": 2.332185745239258,
+      "learning_rate": 9.928532648485168e-06,
+      "loss": 0.5691,
+      "step": 683
+    },
+    {
+      "epoch": 0.05472,
+      "grad_norm": 1.8246127367019653,
+      "learning_rate": 9.928320615582784e-06,
+      "loss": 0.3214,
+      "step": 684
+    },
+    {
+      "epoch": 0.0548,
+      "grad_norm": 1.643385648727417,
+      "learning_rate": 9.928108270882049e-06,
+      "loss": 0.2813,
+      "step": 685
+    },
+    {
+      "epoch": 0.05488,
+      "grad_norm": 1.6711945533752441,
+      "learning_rate": 9.927895614396392e-06,
+      "loss": 0.3232,
+      "step": 686
+    },
+    {
+      "epoch": 0.05496,
+      "grad_norm": 1.6612136363983154,
+      "learning_rate": 9.927682646139269e-06,
+      "loss": 0.3504,
+      "step": 687
+    },
+    {
+      "epoch": 0.05504,
+      "grad_norm": 1.4387818574905396,
+      "learning_rate": 9.927469366124152e-06,
+      "loss": 0.3185,
+      "step": 688
+    },
+    {
+      "epoch": 0.05512,
+      "grad_norm": 1.5262702703475952,
+      "learning_rate": 9.927255774364535e-06,
+      "loss": 0.3075,
+      "step": 689
+    },
+    {
+      "epoch": 0.0552,
+      "grad_norm": 1.6111522912979126,
+      "learning_rate": 9.927041870873932e-06,
+      "loss": 0.3691,
+      "step": 690
+    },
+    {
+      "epoch": 0.05528,
+      "grad_norm": 1.7556387186050415,
+      "learning_rate": 9.926827655665878e-06,
+      "loss": 0.4307,
+      "step": 691
+    },
+    {
+      "epoch": 0.05536,
+      "grad_norm": 1.449609637260437,
+      "learning_rate": 9.926613128753922e-06,
+      "loss": 0.3468,
+      "step": 692
+    },
+    {
+      "epoch": 0.05544,
+      "grad_norm": 1.6672128438949585,
+      "learning_rate": 9.926398290151637e-06,
+      "loss": 0.3852,
+      "step": 693
+    },
+    {
+      "epoch": 0.05552,
+      "grad_norm": 1.932892084121704,
+      "learning_rate": 9.926183139872616e-06,
+      "loss": 0.5628,
+      "step": 694
+    },
+    {
+      "epoch": 0.0556,
+      "grad_norm": 1.3282470703125,
+      "learning_rate": 9.925967677930472e-06,
+      "loss": 0.3317,
+      "step": 695
+    },
+    {
+      "epoch": 0.05568,
+      "grad_norm": 1.645140528678894,
+      "learning_rate": 9.925751904338834e-06,
+      "loss": 0.3647,
+      "step": 696
+    },
+    {
+      "epoch": 0.05576,
+      "grad_norm": 1.6291844844818115,
+      "learning_rate": 9.925535819111356e-06,
+      "loss": 0.2835,
+      "step": 697
+    },
+    {
+      "epoch": 0.05584,
+      "grad_norm": 1.8840759992599487,
+      "learning_rate": 9.925319422261708e-06,
+      "loss": 0.3661,
+      "step": 698
+    },
+    {
+      "epoch": 0.05592,
+      "grad_norm": 1.7590316534042358,
+      "learning_rate": 9.925102713803579e-06,
+      "loss": 0.3905,
+      "step": 699
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 1.4774218797683716,
+      "learning_rate": 9.924885693750681e-06,
+      "loss": 0.491,
+      "step": 700
+    },
+    {
+      "epoch": 0.05608,
+      "grad_norm": 1.8233764171600342,
+      "learning_rate": 9.924668362116743e-06,
+      "loss": 0.4349,
+      "step": 701
+    },
+    {
+      "epoch": 0.05616,
+      "grad_norm": 2.0210931301116943,
+      "learning_rate": 9.924450718915517e-06,
+      "loss": 0.3602,
+      "step": 702
+    },
+    {
+      "epoch": 0.05624,
+      "grad_norm": 1.9228521585464478,
+      "learning_rate": 9.92423276416077e-06,
+      "loss": 0.3999,
+      "step": 703
+    },
+    {
+      "epoch": 0.05632,
+      "grad_norm": 1.6303220987319946,
+      "learning_rate": 9.924014497866295e-06,
+      "loss": 0.3443,
+      "step": 704
+    },
+    {
+      "epoch": 0.0564,
+      "grad_norm": 1.7699851989746094,
+      "learning_rate": 9.923795920045896e-06,
+      "loss": 0.3392,
+      "step": 705
+    },
+    {
+      "epoch": 0.05648,
+      "grad_norm": 1.4135288000106812,
+      "learning_rate": 9.923577030713406e-06,
+      "loss": 0.2955,
+      "step": 706
+    },
+    {
+      "epoch": 0.05656,
+      "grad_norm": 1.8755179643630981,
+      "learning_rate": 9.923357829882671e-06,
+      "loss": 0.4674,
+      "step": 707
+    },
+    {
+      "epoch": 0.05664,
+      "grad_norm": 1.4635642766952515,
+      "learning_rate": 9.92313831756756e-06,
+      "loss": 0.2897,
+      "step": 708
+    },
+    {
+      "epoch": 0.05672,
+      "grad_norm": 1.7681899070739746,
+      "learning_rate": 9.922918493781958e-06,
+      "loss": 0.3768,
+      "step": 709
+    },
+    {
+      "epoch": 0.0568,
+      "grad_norm": 1.4016932249069214,
+      "learning_rate": 9.92269835853978e-06,
+      "loss": 0.2945,
+      "step": 710
+    },
+    {
+      "epoch": 0.05688,
+      "grad_norm": 2.1971428394317627,
+      "learning_rate": 9.922477911854945e-06,
+      "loss": 0.4537,
+      "step": 711
+    },
+    {
+      "epoch": 0.05696,
+      "grad_norm": 1.4524810314178467,
+      "learning_rate": 9.922257153741402e-06,
+      "loss": 0.3269,
+      "step": 712
+    },
+    {
+      "epoch": 0.05704,
+      "grad_norm": 1.7311736345291138,
+      "learning_rate": 9.92203608421312e-06,
+      "loss": 0.415,
+      "step": 713
+    },
+    {
+      "epoch": 0.05712,
+      "grad_norm": 2.3954567909240723,
+      "learning_rate": 9.921814703284086e-06,
+      "loss": 0.4937,
+      "step": 714
+    },
+    {
+      "epoch": 0.0572,
+      "grad_norm": 1.8570671081542969,
+      "learning_rate": 9.921593010968302e-06,
+      "loss": 0.3431,
+      "step": 715
+    },
+    {
+      "epoch": 0.05728,
+      "grad_norm": 1.7509444952011108,
+      "learning_rate": 9.921371007279796e-06,
+      "loss": 0.4205,
+      "step": 716
+    },
+    {
+      "epoch": 0.05736,
+      "grad_norm": 1.6266485452651978,
+      "learning_rate": 9.921148692232615e-06,
+      "loss": 0.3455,
+      "step": 717
+    },
+    {
+      "epoch": 0.05744,
+      "grad_norm": 1.3018132448196411,
+      "learning_rate": 9.920926065840821e-06,
+      "loss": 0.2517,
+      "step": 718
+    },
+    {
+      "epoch": 0.05752,
+      "grad_norm": 1.8658397197723389,
+      "learning_rate": 9.9207031281185e-06,
+      "loss": 0.4211,
+      "step": 719
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 1.8189882040023804,
+      "learning_rate": 9.920479879079758e-06,
+      "loss": 0.3201,
+      "step": 720
+    },
+    {
+      "epoch": 0.05768,
+      "grad_norm": 1.655189871788025,
+      "learning_rate": 9.920256318738717e-06,
+      "loss": 0.3993,
+      "step": 721
+    },
+    {
+      "epoch": 0.05776,
+      "grad_norm": 1.7855379581451416,
+      "learning_rate": 9.920032447109522e-06,
+      "loss": 0.3763,
+      "step": 722
+    },
+    {
+      "epoch": 0.05784,
+      "grad_norm": 1.5921399593353271,
+      "learning_rate": 9.919808264206339e-06,
+      "loss": 0.3259,
+      "step": 723
+    },
+    {
+      "epoch": 0.05792,
+      "grad_norm": 1.2852582931518555,
+      "learning_rate": 9.919583770043347e-06,
+      "loss": 0.2554,
+      "step": 724
+    },
+    {
+      "epoch": 0.058,
+      "grad_norm": 1.9973994493484497,
+      "learning_rate": 9.91935896463475e-06,
+      "loss": 0.4433,
+      "step": 725
+    },
+    {
+      "epoch": 0.05808,
+      "grad_norm": 1.5259404182434082,
+      "learning_rate": 9.91913384799477e-06,
+      "loss": 0.391,
+      "step": 726
+    },
+    {
+      "epoch": 0.05816,
+      "grad_norm": 1.7888422012329102,
+      "learning_rate": 9.918908420137654e-06,
+      "loss": 0.3253,
+      "step": 727
+    },
+    {
+      "epoch": 0.05824,
+      "grad_norm": 1.7182378768920898,
+      "learning_rate": 9.91868268107766e-06,
+      "loss": 0.3694,
+      "step": 728
+    },
+    {
+      "epoch": 0.05832,
+      "grad_norm": 1.8759191036224365,
+      "learning_rate": 9.918456630829071e-06,
+      "loss": 0.3199,
+      "step": 729
+    },
+    {
+      "epoch": 0.0584,
+      "grad_norm": 1.4170511960983276,
+      "learning_rate": 9.918230269406188e-06,
+      "loss": 0.3506,
+      "step": 730
+    },
+    {
+      "epoch": 0.05848,
+      "grad_norm": 1.6913896799087524,
+      "learning_rate": 9.918003596823333e-06,
+      "loss": 0.4956,
+      "step": 731
+    },
+    {
+      "epoch": 0.05856,
+      "grad_norm": 1.5514379739761353,
+      "learning_rate": 9.917776613094846e-06,
+      "loss": 0.3401,
+      "step": 732
+    },
+    {
+      "epoch": 0.05864,
+      "grad_norm": 1.7475535869598389,
+      "learning_rate": 9.917549318235086e-06,
+      "loss": 0.3692,
+      "step": 733
+    },
+    {
+      "epoch": 0.05872,
+      "grad_norm": 1.3284653425216675,
+      "learning_rate": 9.917321712258436e-06,
+      "loss": 0.29,
+      "step": 734
+    },
+    {
+      "epoch": 0.0588,
+      "grad_norm": 1.3704982995986938,
+      "learning_rate": 9.917093795179294e-06,
+      "loss": 0.288,
+      "step": 735
+    },
+    {
+      "epoch": 0.05888,
+      "grad_norm": 1.5144543647766113,
+      "learning_rate": 9.916865567012082e-06,
+      "loss": 0.3907,
+      "step": 736
+    },
+    {
+      "epoch": 0.05896,
+      "grad_norm": 1.4970391988754272,
+      "learning_rate": 9.916637027771236e-06,
+      "loss": 0.3982,
+      "step": 737
+    },
+    {
+      "epoch": 0.05904,
+      "grad_norm": 1.6594504117965698,
+      "learning_rate": 9.916408177471216e-06,
+      "loss": 0.3547,
+      "step": 738
+    },
+    {
+      "epoch": 0.05912,
+      "grad_norm": 1.6561534404754639,
+      "learning_rate": 9.916179016126502e-06,
+      "loss": 0.4931,
+      "step": 739
+    },
+    {
+      "epoch": 0.0592,
+      "grad_norm": 1.0301389694213867,
+      "learning_rate": 9.915949543751591e-06,
+      "loss": 0.1981,
+      "step": 740
+    },
+    {
+      "epoch": 0.05928,
+      "grad_norm": 1.5365773439407349,
+      "learning_rate": 9.915719760361e-06,
+      "loss": 0.3218,
+      "step": 741
+    },
+    {
+      "epoch": 0.05936,
+      "grad_norm": 1.725466251373291,
+      "learning_rate": 9.915489665969269e-06,
+      "loss": 0.3497,
+      "step": 742
+    },
+    {
+      "epoch": 0.05944,
+      "grad_norm": 1.3347848653793335,
+      "learning_rate": 9.915259260590954e-06,
+      "loss": 0.29,
+      "step": 743
+    },
+    {
+      "epoch": 0.05952,
+      "grad_norm": 2.047433853149414,
+      "learning_rate": 9.915028544240633e-06,
+      "loss": 0.3546,
+      "step": 744
+    },
+    {
+      "epoch": 0.0596,
+      "grad_norm": 1.5531580448150635,
+      "learning_rate": 9.914797516932899e-06,
+      "loss": 0.2984,
+      "step": 745
+    },
+    {
+      "epoch": 0.05968,
+      "grad_norm": 1.964288353919983,
+      "learning_rate": 9.914566178682373e-06,
+      "loss": 0.3893,
+      "step": 746
+    },
+    {
+      "epoch": 0.05976,
+      "grad_norm": 1.6539465188980103,
+      "learning_rate": 9.91433452950369e-06,
+      "loss": 0.3372,
+      "step": 747
+    },
+    {
+      "epoch": 0.05984,
+      "grad_norm": 1.79845130443573,
+      "learning_rate": 9.914102569411503e-06,
+      "loss": 0.379,
+      "step": 748
+    },
+    {
+      "epoch": 0.05992,
+      "grad_norm": 1.5639078617095947,
+      "learning_rate": 9.91387029842049e-06,
+      "loss": 0.3265,
+      "step": 749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.675551414489746,
+      "learning_rate": 9.913637716545344e-06,
+      "loss": 0.3333,
+      "step": 750
+    },
+    {
+      "epoch": 0.06008,
+      "grad_norm": 1.6994894742965698,
+      "learning_rate": 9.913404823800782e-06,
+      "loss": 0.3565,
+      "step": 751
+    },
+    {
+      "epoch": 0.06016,
+      "grad_norm": 1.9031753540039062,
+      "learning_rate": 9.913171620201536e-06,
+      "loss": 0.4082,
+      "step": 752
+    },
+    {
+      "epoch": 0.06024,
+      "grad_norm": 1.789406657218933,
+      "learning_rate": 9.912938105762362e-06,
+      "loss": 0.3834,
+      "step": 753
+    },
+    {
+      "epoch": 0.06032,
+      "grad_norm": 1.7641888856887817,
+      "learning_rate": 9.912704280498032e-06,
+      "loss": 0.3464,
+      "step": 754
+    },
+    {
+      "epoch": 0.0604,
+      "grad_norm": 1.8987987041473389,
+      "learning_rate": 9.91247014442334e-06,
+      "loss": 0.4163,
+      "step": 755
+    },
+    {
+      "epoch": 0.06048,
+      "grad_norm": 1.8092420101165771,
+      "learning_rate": 9.912235697553101e-06,
+      "loss": 0.433,
+      "step": 756
+    },
+    {
+      "epoch": 0.06056,
+      "grad_norm": 1.20881187915802,
+      "learning_rate": 9.912000939902144e-06,
+      "loss": 0.251,
+      "step": 757
+    },
+    {
+      "epoch": 0.06064,
+      "grad_norm": 1.4415931701660156,
+      "learning_rate": 9.911765871485325e-06,
+      "loss": 0.3483,
+      "step": 758
+    },
+    {
+      "epoch": 0.06072,
+      "grad_norm": 2.056685209274292,
+      "learning_rate": 9.911530492317511e-06,
+      "loss": 0.4327,
+      "step": 759
+    },
+    {
+      "epoch": 0.0608,
+      "grad_norm": 2.1116669178009033,
+      "learning_rate": 9.9112948024136e-06,
+      "loss": 0.549,
+      "step": 760
+    },
+    {
+      "epoch": 0.06088,
+      "grad_norm": 1.765018343925476,
+      "learning_rate": 9.911058801788499e-06,
+      "loss": 0.3743,
+      "step": 761
+    },
+    {
+      "epoch": 0.06096,
+      "grad_norm": 1.5852882862091064,
+      "learning_rate": 9.910822490457139e-06,
+      "loss": 0.356,
+      "step": 762
+    },
+    {
+      "epoch": 0.06104,
+      "grad_norm": 1.6404635906219482,
+      "learning_rate": 9.910585868434473e-06,
+      "loss": 0.3543,
+      "step": 763
+    },
+    {
+      "epoch": 0.06112,
+      "grad_norm": 1.5035934448242188,
+      "learning_rate": 9.91034893573547e-06,
+      "loss": 0.3572,
+      "step": 764
+    },
+    {
+      "epoch": 0.0612,
+      "grad_norm": 1.9684243202209473,
+      "learning_rate": 9.910111692375118e-06,
+      "loss": 0.4283,
+      "step": 765
+    },
+    {
+      "epoch": 0.06128,
+      "grad_norm": 1.9752951860427856,
+      "learning_rate": 9.90987413836843e-06,
+      "loss": 0.4133,
+      "step": 766
+    },
+    {
+      "epoch": 0.06136,
+      "grad_norm": 1.6825329065322876,
+      "learning_rate": 9.909636273730434e-06,
+      "loss": 0.4326,
+      "step": 767
+    },
+    {
+      "epoch": 0.06144,
+      "grad_norm": 1.7126015424728394,
+      "learning_rate": 9.909398098476177e-06,
+      "loss": 0.3714,
+      "step": 768
+    },
+    {
+      "epoch": 0.06152,
+      "grad_norm": 1.39950692653656,
+      "learning_rate": 9.90915961262073e-06,
+      "loss": 0.3208,
+      "step": 769
+    },
+    {
+      "epoch": 0.0616,
+      "grad_norm": 1.8275245428085327,
+      "learning_rate": 9.908920816179182e-06,
+      "loss": 0.3871,
+      "step": 770
+    },
+    {
+      "epoch": 0.06168,
+      "grad_norm": 1.6578397750854492,
+      "learning_rate": 9.90868170916664e-06,
+      "loss": 0.335,
+      "step": 771
+    },
+    {
+      "epoch": 0.06176,
+      "grad_norm": 1.8626042604446411,
+      "learning_rate": 9.908442291598227e-06,
+      "loss": 0.3591,
+      "step": 772
+    },
+    {
+      "epoch": 0.06184,
+      "grad_norm": 1.6682721376419067,
+      "learning_rate": 9.908202563489095e-06,
+      "loss": 0.3483,
+      "step": 773
+    },
+    {
+      "epoch": 0.06192,
+      "grad_norm": 1.5909526348114014,
+      "learning_rate": 9.907962524854411e-06,
+      "loss": 0.4166,
+      "step": 774
+    },
+    {
+      "epoch": 0.062,
+      "grad_norm": 1.7059555053710938,
+      "learning_rate": 9.90772217570936e-06,
+      "loss": 0.3201,
+      "step": 775
+    },
+    {
+      "epoch": 0.06208,
+      "grad_norm": 1.5793256759643555,
+      "learning_rate": 9.907481516069149e-06,
+      "loss": 0.3644,
+      "step": 776
+    },
+    {
+      "epoch": 0.06216,
+      "grad_norm": 1.5775402784347534,
+      "learning_rate": 9.907240545949001e-06,
+      "loss": 0.2876,
+      "step": 777
+    },
+    {
+      "epoch": 0.06224,
+      "grad_norm": 1.9473414421081543,
+      "learning_rate": 9.906999265364163e-06,
+      "loss": 0.367,
+      "step": 778
+    },
+    {
+      "epoch": 0.06232,
+      "grad_norm": 1.9545674324035645,
+      "learning_rate": 9.906757674329903e-06,
+      "loss": 0.4359,
+      "step": 779
+    },
+    {
+      "epoch": 0.0624,
+      "grad_norm": 1.605450987815857,
+      "learning_rate": 9.906515772861501e-06,
+      "loss": 0.3206,
+      "step": 780
+    },
+    {
+      "epoch": 0.06248,
+      "grad_norm": 1.5518044233322144,
+      "learning_rate": 9.906273560974264e-06,
+      "loss": 0.3723,
+      "step": 781
+    },
+    {
+      "epoch": 0.06256,
+      "grad_norm": 1.6524393558502197,
+      "learning_rate": 9.906031038683515e-06,
+      "loss": 0.3181,
+      "step": 782
+    },
+    {
+      "epoch": 0.06264,
+      "grad_norm": 1.3769612312316895,
+      "learning_rate": 9.9057882060046e-06,
+      "loss": 0.3077,
+      "step": 783
+    },
+    {
+      "epoch": 0.06272,
+      "grad_norm": 1.5857384204864502,
+      "learning_rate": 9.905545062952876e-06,
+      "loss": 0.3109,
+      "step": 784
+    },
+    {
+      "epoch": 0.0628,
+      "grad_norm": 1.549974799156189,
+      "learning_rate": 9.90530160954373e-06,
+      "loss": 0.3216,
+      "step": 785
+    },
+    {
+      "epoch": 0.06288,
+      "grad_norm": 1.836879849433899,
+      "learning_rate": 9.905057845792568e-06,
+      "loss": 0.3937,
+      "step": 786
+    },
+    {
+      "epoch": 0.06296,
+      "grad_norm": 1.6906849145889282,
+      "learning_rate": 9.904813771714806e-06,
+      "loss": 0.3045,
+      "step": 787
+    },
+    {
+      "epoch": 0.06304,
+      "grad_norm": 1.4917078018188477,
+      "learning_rate": 9.904569387325888e-06,
+      "loss": 0.2654,
+      "step": 788
+    },
+    {
+      "epoch": 0.06312,
+      "grad_norm": 1.9071767330169678,
+      "learning_rate": 9.904324692641279e-06,
+      "loss": 0.4201,
+      "step": 789
+    },
+    {
+      "epoch": 0.0632,
+      "grad_norm": 1.6818795204162598,
+      "learning_rate": 9.904079687676453e-06,
+      "loss": 0.3577,
+      "step": 790
+    },
+    {
+      "epoch": 0.06328,
+      "grad_norm": 1.5190314054489136,
+      "learning_rate": 9.903834372446914e-06,
+      "loss": 0.3156,
+      "step": 791
+    },
+    {
+      "epoch": 0.06336,
+      "grad_norm": 1.2045843601226807,
+      "learning_rate": 9.903588746968185e-06,
+      "loss": 0.2759,
+      "step": 792
+    },
+    {
+      "epoch": 0.06344,
+      "grad_norm": 1.6211133003234863,
+      "learning_rate": 9.903342811255802e-06,
+      "loss": 0.3412,
+      "step": 793
+    },
+    {
+      "epoch": 0.06352,
+      "grad_norm": 1.6240700483322144,
+      "learning_rate": 9.903096565325326e-06,
+      "loss": 0.2906,
+      "step": 794
+    },
+    {
+      "epoch": 0.0636,
+      "grad_norm": 1.5824121236801147,
+      "learning_rate": 9.902850009192338e-06,
+      "loss": 0.3956,
+      "step": 795
+    },
+    {
+      "epoch": 0.06368,
+      "grad_norm": 1.8506085872650146,
+      "learning_rate": 9.902603142872433e-06,
+      "loss": 0.3616,
+      "step": 796
+    },
+    {
+      "epoch": 0.06376,
+      "grad_norm": 1.4106500148773193,
+      "learning_rate": 9.90235596638123e-06,
+      "loss": 0.3068,
+      "step": 797
+    },
+    {
+      "epoch": 0.06384,
+      "grad_norm": 1.5966297388076782,
+      "learning_rate": 9.902108479734372e-06,
+      "loss": 0.4436,
+      "step": 798
+    },
+    {
+      "epoch": 0.06392,
+      "grad_norm": 1.908096194267273,
+      "learning_rate": 9.90186068294751e-06,
+      "loss": 0.3364,
+      "step": 799
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 1.6867671012878418,
+      "learning_rate": 9.901612576036326e-06,
+      "loss": 0.377,
+      "step": 800
+    },
+    {
+      "epoch": 0.06408,
+      "grad_norm": 1.7174458503723145,
+      "learning_rate": 9.901364159016514e-06,
+      "loss": 0.4152,
+      "step": 801
+    },
+    {
+      "epoch": 0.06416,
+      "grad_norm": 1.4476027488708496,
+      "learning_rate": 9.901115431903792e-06,
+      "loss": 0.3267,
+      "step": 802
+    },
+    {
+      "epoch": 0.06424,
+      "grad_norm": 1.7989741563796997,
+      "learning_rate": 9.900866394713895e-06,
+      "loss": 0.3113,
+      "step": 803
+    },
+    {
+      "epoch": 0.06432,
+      "grad_norm": 2.0357306003570557,
+      "learning_rate": 9.900617047462581e-06,
+      "loss": 0.3554,
+      "step": 804
+    },
+    {
+      "epoch": 0.0644,
+      "grad_norm": 1.6617554426193237,
+      "learning_rate": 9.900367390165623e-06,
+      "loss": 0.3521,
+      "step": 805
+    },
+    {
+      "epoch": 0.06448,
+      "grad_norm": 1.6434162855148315,
+      "learning_rate": 9.900117422838817e-06,
+      "loss": 0.3676,
+      "step": 806
+    },
+    {
+      "epoch": 0.06456,
+      "grad_norm": 1.570143699645996,
+      "learning_rate": 9.899867145497978e-06,
+      "loss": 0.3857,
+      "step": 807
+    },
+    {
+      "epoch": 0.06464,
+      "grad_norm": 1.8630397319793701,
+      "learning_rate": 9.89961655815894e-06,
+      "loss": 0.4321,
+      "step": 808
+    },
+    {
+      "epoch": 0.06472,
+      "grad_norm": 1.3922876119613647,
+      "learning_rate": 9.899365660837555e-06,
+      "loss": 0.3751,
+      "step": 809
+    },
+    {
+      "epoch": 0.0648,
+      "grad_norm": 1.575070858001709,
+      "learning_rate": 9.899114453549699e-06,
+      "loss": 0.3384,
+      "step": 810
+    },
+    {
+      "epoch": 0.06488,
+      "grad_norm": 1.494339108467102,
+      "learning_rate": 9.898862936311264e-06,
+      "loss": 0.3632,
+      "step": 811
+    },
+    {
+      "epoch": 0.06496,
+      "grad_norm": 1.5550836324691772,
+      "learning_rate": 9.898611109138163e-06,
+      "loss": 0.3129,
+      "step": 812
+    },
+    {
+      "epoch": 0.06504,
+      "grad_norm": 1.3468754291534424,
+      "learning_rate": 9.898358972046327e-06,
+      "loss": 0.3435,
+      "step": 813
+    },
+    {
+      "epoch": 0.06512,
+      "grad_norm": 1.2275058031082153,
+      "learning_rate": 9.89810652505171e-06,
+      "loss": 0.2591,
+      "step": 814
+    },
+    {
+      "epoch": 0.0652,
+      "grad_norm": 1.689713954925537,
+      "learning_rate": 9.897853768170282e-06,
+      "loss": 0.4462,
+      "step": 815
+    },
+    {
+      "epoch": 0.06528,
+      "grad_norm": 1.4738130569458008,
+      "learning_rate": 9.897600701418033e-06,
+      "loss": 0.2962,
+      "step": 816
+    },
+    {
+      "epoch": 0.06536,
+      "grad_norm": 1.8294364213943481,
+      "learning_rate": 9.897347324810977e-06,
+      "loss": 0.3394,
+      "step": 817
+    },
+    {
+      "epoch": 0.06544,
+      "grad_norm": 2.3864986896514893,
+      "learning_rate": 9.89709363836514e-06,
+      "loss": 0.5214,
+      "step": 818
+    },
+    {
+      "epoch": 0.06552,
+      "grad_norm": 1.4256991147994995,
+      "learning_rate": 9.896839642096576e-06,
+      "loss": 0.3178,
+      "step": 819
+    },
+    {
+      "epoch": 0.0656,
+      "grad_norm": 1.8996535539627075,
+      "learning_rate": 9.896585336021353e-06,
+      "loss": 0.53,
+      "step": 820
+    },
+    {
+      "epoch": 0.06568,
+      "grad_norm": 1.7665208578109741,
+      "learning_rate": 9.896330720155558e-06,
+      "loss": 0.4027,
+      "step": 821
+    },
+    {
+      "epoch": 0.06576,
+      "grad_norm": 1.2968679666519165,
+      "learning_rate": 9.896075794515304e-06,
+      "loss": 0.2658,
+      "step": 822
+    },
+    {
+      "epoch": 0.06584,
+      "grad_norm": 2.179189682006836,
+      "learning_rate": 9.895820559116715e-06,
+      "loss": 0.3901,
+      "step": 823
+    },
+    {
+      "epoch": 0.06592,
+      "grad_norm": 1.4469057321548462,
+      "learning_rate": 9.89556501397594e-06,
+      "loss": 0.3504,
+      "step": 824
+    },
+    {
+      "epoch": 0.066,
+      "grad_norm": 1.4145433902740479,
+      "learning_rate": 9.89530915910915e-06,
+      "loss": 0.3422,
+      "step": 825
+    },
+    {
+      "epoch": 0.06608,
+      "grad_norm": 1.6971393823623657,
+      "learning_rate": 9.895052994532527e-06,
+      "loss": 0.3849,
+      "step": 826
+    },
+    {
+      "epoch": 0.06616,
+      "grad_norm": 1.416899561882019,
+      "learning_rate": 9.89479652026228e-06,
+      "loss": 0.2804,
+      "step": 827
+    },
+    {
+      "epoch": 0.06624,
+      "grad_norm": 1.758364200592041,
+      "learning_rate": 9.894539736314636e-06,
+      "loss": 0.3886,
+      "step": 828
+    },
+    {
+      "epoch": 0.06632,
+      "grad_norm": 1.8403048515319824,
+      "learning_rate": 9.894282642705839e-06,
+      "loss": 0.3514,
+      "step": 829
+    },
+    {
+      "epoch": 0.0664,
+      "grad_norm": 2.0701754093170166,
+      "learning_rate": 9.894025239452156e-06,
+      "loss": 0.3829,
+      "step": 830
+    },
+    {
+      "epoch": 0.06648,
+      "grad_norm": 1.6465054750442505,
+      "learning_rate": 9.893767526569873e-06,
+      "loss": 0.4353,
+      "step": 831
+    },
+    {
+      "epoch": 0.06656,
+      "grad_norm": 1.3121788501739502,
+      "learning_rate": 9.893509504075291e-06,
+      "loss": 0.2722,
+      "step": 832
+    },
+    {
+      "epoch": 0.06664,
+      "grad_norm": 1.453292727470398,
+      "learning_rate": 9.893251171984737e-06,
+      "loss": 0.3116,
+      "step": 833
+    },
+    {
+      "epoch": 0.06672,
+      "grad_norm": 1.511839747428894,
+      "learning_rate": 9.892992530314556e-06,
+      "loss": 0.3135,
+      "step": 834
+    },
+    {
+      "epoch": 0.0668,
+      "grad_norm": 1.5262436866760254,
+      "learning_rate": 9.892733579081108e-06,
+      "loss": 0.321,
+      "step": 835
+    },
+    {
+      "epoch": 0.06688,
+      "grad_norm": 2.0322422981262207,
+      "learning_rate": 9.892474318300778e-06,
+      "loss": 0.5337,
+      "step": 836
+    },
+    {
+      "epoch": 0.06696,
+      "grad_norm": 1.5687463283538818,
+      "learning_rate": 9.892214747989967e-06,
+      "loss": 0.2962,
+      "step": 837
+    },
+    {
+      "epoch": 0.06704,
+      "grad_norm": 2.1740803718566895,
+      "learning_rate": 9.8919548681651e-06,
+      "loss": 0.4359,
+      "step": 838
+    },
+    {
+      "epoch": 0.06712,
+      "grad_norm": 1.6020910739898682,
+      "learning_rate": 9.891694678842617e-06,
+      "loss": 0.3122,
+      "step": 839
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 1.6646548509597778,
+      "learning_rate": 9.891434180038979e-06,
+      "loss": 0.4588,
+      "step": 840
+    },
+    {
+      "epoch": 0.06728,
+      "grad_norm": 1.4084759950637817,
+      "learning_rate": 9.891173371770666e-06,
+      "loss": 0.2731,
+      "step": 841
+    },
+    {
+      "epoch": 0.06736,
+      "grad_norm": 1.433922290802002,
+      "learning_rate": 9.890912254054182e-06,
+      "loss": 0.3111,
+      "step": 842
+    },
+    {
+      "epoch": 0.06744,
+      "grad_norm": 1.3869508504867554,
+      "learning_rate": 9.890650826906042e-06,
+      "loss": 0.3775,
+      "step": 843
+    },
+    {
+      "epoch": 0.06752,
+      "grad_norm": 1.6502546072006226,
+      "learning_rate": 9.890389090342789e-06,
+      "loss": 0.3441,
+      "step": 844
+    },
+    {
+      "epoch": 0.0676,
+      "grad_norm": 1.827817678451538,
+      "learning_rate": 9.890127044380983e-06,
+      "loss": 0.4,
+      "step": 845
+    },
+    {
+      "epoch": 0.06768,
+      "grad_norm": 1.306138038635254,
+      "learning_rate": 9.889864689037201e-06,
+      "loss": 0.3183,
+      "step": 846
+    },
+    {
+      "epoch": 0.06776,
+      "grad_norm": 2.2315001487731934,
+      "learning_rate": 9.88960202432804e-06,
+      "loss": 0.4797,
+      "step": 847
+    },
+    {
+      "epoch": 0.06784,
+      "grad_norm": 1.777730941772461,
+      "learning_rate": 9.889339050270122e-06,
+      "loss": 0.4378,
+      "step": 848
+    },
+    {
+      "epoch": 0.06792,
+      "grad_norm": 1.909468173980713,
+      "learning_rate": 9.88907576688008e-06,
+      "loss": 0.4438,
+      "step": 849
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 1.5327777862548828,
+      "learning_rate": 9.888812174174574e-06,
+      "loss": 0.3694,
+      "step": 850
+    },
+    {
+      "epoch": 0.06808,
+      "grad_norm": 2.512308120727539,
+      "learning_rate": 9.88854827217028e-06,
+      "loss": 0.5076,
+      "step": 851
+    },
+    {
+      "epoch": 0.06816,
+      "grad_norm": 1.4551916122436523,
+      "learning_rate": 9.888284060883892e-06,
+      "loss": 0.3042,
+      "step": 852
+    },
+    {
+      "epoch": 0.06824,
+      "grad_norm": 1.8381390571594238,
+      "learning_rate": 9.88801954033213e-06,
+      "loss": 0.4247,
+      "step": 853
+    },
+    {
+      "epoch": 0.06832,
+      "grad_norm": 1.5407111644744873,
+      "learning_rate": 9.887754710531727e-06,
+      "loss": 0.3213,
+      "step": 854
+    },
+    {
+      "epoch": 0.0684,
+      "grad_norm": 1.478391408920288,
+      "learning_rate": 9.887489571499438e-06,
+      "loss": 0.3527,
+      "step": 855
+    },
+    {
+      "epoch": 0.06848,
+      "grad_norm": 1.9624853134155273,
+      "learning_rate": 9.887224123252037e-06,
+      "loss": 0.4039,
+      "step": 856
+    },
+    {
+      "epoch": 0.06856,
+      "grad_norm": 1.6501410007476807,
+      "learning_rate": 9.886958365806317e-06,
+      "loss": 0.4169,
+      "step": 857
+    },
+    {
+      "epoch": 0.06864,
+      "grad_norm": 1.4560359716415405,
+      "learning_rate": 9.886692299179094e-06,
+      "loss": 0.2993,
+      "step": 858
+    },
+    {
+      "epoch": 0.06872,
+      "grad_norm": 1.681654930114746,
+      "learning_rate": 9.8864259233872e-06,
+      "loss": 0.405,
+      "step": 859
+    },
+    {
+      "epoch": 0.0688,
+      "grad_norm": 1.7231281995773315,
+      "learning_rate": 9.886159238447488e-06,
+      "loss": 0.3455,
+      "step": 860
+    },
+    {
+      "epoch": 0.06888,
+      "grad_norm": 1.5479826927185059,
+      "learning_rate": 9.885892244376831e-06,
+      "loss": 0.3774,
+      "step": 861
+    },
+    {
+      "epoch": 0.06896,
+      "grad_norm": 1.5777459144592285,
+      "learning_rate": 9.88562494119212e-06,
+      "loss": 0.3301,
+      "step": 862
+    },
+    {
+      "epoch": 0.06904,
+      "grad_norm": 2.2783584594726562,
+      "learning_rate": 9.885357328910265e-06,
+      "loss": 0.4368,
+      "step": 863
+    },
+    {
+      "epoch": 0.06912,
+      "grad_norm": 1.8769716024398804,
+      "learning_rate": 9.885089407548198e-06,
+      "loss": 0.3725,
+      "step": 864
+    },
+    {
+      "epoch": 0.0692,
+      "grad_norm": 1.3858102560043335,
+      "learning_rate": 9.884821177122871e-06,
+      "loss": 0.2691,
+      "step": 865
+    },
+    {
+      "epoch": 0.06928,
+      "grad_norm": 1.477070689201355,
+      "learning_rate": 9.884552637651252e-06,
+      "loss": 0.2952,
+      "step": 866
+    },
+    {
+      "epoch": 0.06936,
+      "grad_norm": 1.43630051612854,
+      "learning_rate": 9.884283789150332e-06,
+      "loss": 0.3235,
+      "step": 867
+    },
+    {
+      "epoch": 0.06944,
+      "grad_norm": 1.65440833568573,
+      "learning_rate": 9.88401463163712e-06,
+      "loss": 0.3179,
+      "step": 868
+    },
+    {
+      "epoch": 0.06952,
+      "grad_norm": 1.6520179510116577,
+      "learning_rate": 9.883745165128642e-06,
+      "loss": 0.3051,
+      "step": 869
+    },
+    {
+      "epoch": 0.0696,
+      "grad_norm": 1.9777926206588745,
+      "learning_rate": 9.883475389641952e-06,
+      "loss": 0.4753,
+      "step": 870
+    },
+    {
+      "epoch": 0.06968,
+      "grad_norm": 1.5092403888702393,
+      "learning_rate": 9.883205305194112e-06,
+      "loss": 0.3216,
+      "step": 871
+    },
+    {
+      "epoch": 0.06976,
+      "grad_norm": 1.4967854022979736,
+      "learning_rate": 9.88293491180221e-06,
+      "loss": 0.3259,
+      "step": 872
+    },
+    {
+      "epoch": 0.06984,
+      "grad_norm": 1.7150827646255493,
+      "learning_rate": 9.882664209483356e-06,
+      "loss": 0.4019,
+      "step": 873
+    },
+    {
+      "epoch": 0.06992,
+      "grad_norm": 1.9008959531784058,
+      "learning_rate": 9.882393198254676e-06,
+      "loss": 0.5442,
+      "step": 874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.919569492340088,
+      "learning_rate": 9.882121878133314e-06,
+      "loss": 0.501,
+      "step": 875
+    },
+    {
+      "epoch": 0.07008,
+      "grad_norm": 1.6029773950576782,
+      "learning_rate": 9.881850249136438e-06,
+      "loss": 0.3842,
+      "step": 876
+    },
+    {
+      "epoch": 0.07016,
+      "grad_norm": 1.4710146188735962,
+      "learning_rate": 9.881578311281229e-06,
+      "loss": 0.3785,
+      "step": 877
+    },
+    {
+      "epoch": 0.07024,
+      "grad_norm": 1.135948896408081,
+      "learning_rate": 9.881306064584895e-06,
+      "loss": 0.201,
+      "step": 878
+    },
+    {
+      "epoch": 0.07032,
+      "grad_norm": 1.733964443206787,
+      "learning_rate": 9.88103350906466e-06,
+      "loss": 0.327,
+      "step": 879
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 1.3882209062576294,
+      "learning_rate": 9.880760644737765e-06,
+      "loss": 0.3169,
+      "step": 880
+    },
+    {
+      "epoch": 0.07048,
+      "grad_norm": 1.7301795482635498,
+      "learning_rate": 9.880487471621476e-06,
+      "loss": 0.3786,
+      "step": 881
+    },
+    {
+      "epoch": 0.07056,
+      "grad_norm": 1.1793718338012695,
+      "learning_rate": 9.880213989733077e-06,
+      "loss": 0.2577,
+      "step": 882
+    },
+    {
+      "epoch": 0.07064,
+      "grad_norm": 1.4494835138320923,
+      "learning_rate": 9.879940199089864e-06,
+      "loss": 0.2908,
+      "step": 883
+    },
+    {
+      "epoch": 0.07072,
+      "grad_norm": 1.7559984922409058,
+      "learning_rate": 9.879666099709166e-06,
+      "loss": 0.3739,
+      "step": 884
+    },
+    {
+      "epoch": 0.0708,
+      "grad_norm": 1.729697823524475,
+      "learning_rate": 9.87939169160832e-06,
+      "loss": 0.4134,
+      "step": 885
+    },
+    {
+      "epoch": 0.07088,
+      "grad_norm": 1.867188572883606,
+      "learning_rate": 9.879116974804688e-06,
+      "loss": 0.4629,
+      "step": 886
+    },
+    {
+      "epoch": 0.07096,
+      "grad_norm": 1.5233434438705444,
+      "learning_rate": 9.878841949315652e-06,
+      "loss": 0.3819,
+      "step": 887
+    },
+    {
+      "epoch": 0.07104,
+      "grad_norm": 1.6903587579727173,
+      "learning_rate": 9.878566615158609e-06,
+      "loss": 0.3121,
+      "step": 888
+    },
+    {
+      "epoch": 0.07112,
+      "grad_norm": 1.3567419052124023,
+      "learning_rate": 9.878290972350981e-06,
+      "loss": 0.2761,
+      "step": 889
+    },
+    {
+      "epoch": 0.0712,
+      "grad_norm": 1.6625133752822876,
+      "learning_rate": 9.878015020910205e-06,
+      "loss": 0.3199,
+      "step": 890
+    },
+    {
+      "epoch": 0.07128,
+      "grad_norm": 1.4875622987747192,
+      "learning_rate": 9.877738760853741e-06,
+      "loss": 0.3499,
+      "step": 891
+    },
+    {
+      "epoch": 0.07136,
+      "grad_norm": 1.3312549591064453,
+      "learning_rate": 9.877462192199068e-06,
+      "loss": 0.2261,
+      "step": 892
+    },
+    {
+      "epoch": 0.07144,
+      "grad_norm": 1.9580225944519043,
+      "learning_rate": 9.87718531496368e-06,
+      "loss": 0.3892,
+      "step": 893
+    },
+    {
+      "epoch": 0.07152,
+      "grad_norm": 1.421797275543213,
+      "learning_rate": 9.876908129165096e-06,
+      "loss": 0.3268,
+      "step": 894
+    },
+    {
+      "epoch": 0.0716,
+      "grad_norm": 2.2238597869873047,
+      "learning_rate": 9.876630634820853e-06,
+      "loss": 0.5997,
+      "step": 895
+    },
+    {
+      "epoch": 0.07168,
+      "grad_norm": 1.9498764276504517,
+      "learning_rate": 9.87635283194851e-06,
+      "loss": 0.3721,
+      "step": 896
+    },
+    {
+      "epoch": 0.07176,
+      "grad_norm": 1.8909319639205933,
+      "learning_rate": 9.876074720565637e-06,
+      "loss": 0.4237,
+      "step": 897
+    },
+    {
+      "epoch": 0.07184,
+      "grad_norm": 1.432112455368042,
+      "learning_rate": 9.875796300689832e-06,
+      "loss": 0.2579,
+      "step": 898
+    },
+    {
+      "epoch": 0.07192,
+      "grad_norm": 1.9798771142959595,
+      "learning_rate": 9.875517572338711e-06,
+      "loss": 0.3328,
+      "step": 899
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 1.692348837852478,
+      "learning_rate": 9.875238535529905e-06,
+      "loss": 0.3862,
+      "step": 900
+    },
+    {
+      "epoch": 0.07208,
+      "grad_norm": 1.7320940494537354,
+      "learning_rate": 9.87495919028107e-06,
+      "loss": 0.4359,
+      "step": 901
+    },
+    {
+      "epoch": 0.07216,
+      "grad_norm": 1.4372979402542114,
+      "learning_rate": 9.87467953660988e-06,
+      "loss": 0.3045,
+      "step": 902
+    },
+    {
+      "epoch": 0.07224,
+      "grad_norm": 1.5734885931015015,
+      "learning_rate": 9.874399574534024e-06,
+      "loss": 0.3442,
+      "step": 903
+    },
+    {
+      "epoch": 0.07232,
+      "grad_norm": 1.7863317728042603,
+      "learning_rate": 9.874119304071217e-06,
+      "loss": 0.3815,
+      "step": 904
+    },
+    {
+      "epoch": 0.0724,
+      "grad_norm": 1.580787181854248,
+      "learning_rate": 9.873838725239192e-06,
+      "loss": 0.3385,
+      "step": 905
+    },
+    {
+      "epoch": 0.07248,
+      "grad_norm": 1.7870442867279053,
+      "learning_rate": 9.873557838055698e-06,
+      "loss": 0.3777,
+      "step": 906
+    },
+    {
+      "epoch": 0.07256,
+      "grad_norm": 1.8872705698013306,
+      "learning_rate": 9.873276642538508e-06,
+      "loss": 0.3434,
+      "step": 907
+    },
+    {
+      "epoch": 0.07264,
+      "grad_norm": 1.5185067653656006,
+      "learning_rate": 9.872995138705407e-06,
+      "loss": 0.3182,
+      "step": 908
+    },
+    {
+      "epoch": 0.07272,
+      "grad_norm": 1.175097942352295,
+      "learning_rate": 9.872713326574212e-06,
+      "loss": 0.2708,
+      "step": 909
+    },
+    {
+      "epoch": 0.0728,
+      "grad_norm": 1.4773153066635132,
+      "learning_rate": 9.87243120616275e-06,
+      "loss": 0.3242,
+      "step": 910
+    },
+    {
+      "epoch": 0.07288,
+      "grad_norm": 1.9224504232406616,
+      "learning_rate": 9.872148777488865e-06,
+      "loss": 0.5271,
+      "step": 911
+    },
+    {
+      "epoch": 0.07296,
+      "grad_norm": 1.5207834243774414,
+      "learning_rate": 9.871866040570432e-06,
+      "loss": 0.3758,
+      "step": 912
+    },
+    {
+      "epoch": 0.07304,
+      "grad_norm": 1.730446696281433,
+      "learning_rate": 9.871582995425335e-06,
+      "loss": 0.4164,
+      "step": 913
+    },
+    {
+      "epoch": 0.07312,
+      "grad_norm": 1.8168655633926392,
+      "learning_rate": 9.871299642071483e-06,
+      "loss": 0.3849,
+      "step": 914
+    },
+    {
+      "epoch": 0.0732,
+      "grad_norm": 2.1133110523223877,
+      "learning_rate": 9.871015980526802e-06,
+      "loss": 0.4274,
+      "step": 915
+    },
+    {
+      "epoch": 0.07328,
+      "grad_norm": 1.3025873899459839,
+      "learning_rate": 9.870732010809236e-06,
+      "loss": 0.2901,
+      "step": 916
+    },
+    {
+      "epoch": 0.07336,
+      "grad_norm": 1.259355068206787,
+      "learning_rate": 9.870447732936755e-06,
+      "loss": 0.3114,
+      "step": 917
+    },
+    {
+      "epoch": 0.07344,
+      "grad_norm": 1.4465082883834839,
+      "learning_rate": 9.870163146927343e-06,
+      "loss": 0.3957,
+      "step": 918
+    },
+    {
+      "epoch": 0.07352,
+      "grad_norm": 1.7125508785247803,
+      "learning_rate": 9.869878252799004e-06,
+      "loss": 0.3449,
+      "step": 919
+    },
+    {
+      "epoch": 0.0736,
+      "grad_norm": 2.206003189086914,
+      "learning_rate": 9.869593050569761e-06,
+      "loss": 0.4511,
+      "step": 920
+    },
+    {
+      "epoch": 0.07368,
+      "grad_norm": 1.9376089572906494,
+      "learning_rate": 9.869307540257663e-06,
+      "loss": 0.3655,
+      "step": 921
+    },
+    {
+      "epoch": 0.07376,
+      "grad_norm": 1.8399741649627686,
+      "learning_rate": 9.869021721880765e-06,
+      "loss": 0.4562,
+      "step": 922
+    },
+    {
+      "epoch": 0.07384,
+      "grad_norm": 1.5403954982757568,
+      "learning_rate": 9.868735595457157e-06,
+      "loss": 0.4594,
+      "step": 923
+    },
+    {
+      "epoch": 0.07392,
+      "grad_norm": 1.8176610469818115,
+      "learning_rate": 9.86844916100494e-06,
+      "loss": 0.3866,
+      "step": 924
+    },
+    {
+      "epoch": 0.074,
+      "grad_norm": 1.6104621887207031,
+      "learning_rate": 9.868162418542233e-06,
+      "loss": 0.3379,
+      "step": 925
+    },
+    {
+      "epoch": 0.07408,
+      "grad_norm": 1.7758665084838867,
+      "learning_rate": 9.867875368087179e-06,
+      "loss": 0.3828,
+      "step": 926
+    },
+    {
+      "epoch": 0.07416,
+      "grad_norm": 1.398337960243225,
+      "learning_rate": 9.867588009657938e-06,
+      "loss": 0.3024,
+      "step": 927
+    },
+    {
+      "epoch": 0.07424,
+      "grad_norm": 1.812811017036438,
+      "learning_rate": 9.86730034327269e-06,
+      "loss": 0.5139,
+      "step": 928
+    },
+    {
+      "epoch": 0.07432,
+      "grad_norm": 1.4070488214492798,
+      "learning_rate": 9.867012368949637e-06,
+      "loss": 0.325,
+      "step": 929
+    },
+    {
+      "epoch": 0.0744,
+      "grad_norm": 1.56523597240448,
+      "learning_rate": 9.866724086706996e-06,
+      "loss": 0.4111,
+      "step": 930
+    },
+    {
+      "epoch": 0.07448,
+      "grad_norm": 2.1064064502716064,
+      "learning_rate": 9.866435496563004e-06,
+      "loss": 0.4322,
+      "step": 931
+    },
+    {
+      "epoch": 0.07456,
+      "grad_norm": 1.599244236946106,
+      "learning_rate": 9.866146598535925e-06,
+      "loss": 0.3167,
+      "step": 932
+    },
+    {
+      "epoch": 0.07464,
+      "grad_norm": 1.8283414840698242,
+      "learning_rate": 9.865857392644029e-06,
+      "loss": 0.4891,
+      "step": 933
+    },
+    {
+      "epoch": 0.07472,
+      "grad_norm": 1.8395085334777832,
+      "learning_rate": 9.86556787890562e-06,
+      "loss": 0.4302,
+      "step": 934
+    },
+    {
+      "epoch": 0.0748,
+      "grad_norm": 1.6834166049957275,
+      "learning_rate": 9.865278057339011e-06,
+      "loss": 0.3058,
+      "step": 935
+    },
+    {
+      "epoch": 0.07488,
+      "grad_norm": 1.350382924079895,
+      "learning_rate": 9.864987927962536e-06,
+      "loss": 0.3224,
+      "step": 936
+    },
+    {
+      "epoch": 0.07496,
+      "grad_norm": 2.0382933616638184,
+      "learning_rate": 9.864697490794556e-06,
+      "loss": 0.3888,
+      "step": 937
+    },
+    {
+      "epoch": 0.07504,
+      "grad_norm": 1.7430832386016846,
+      "learning_rate": 9.864406745853443e-06,
+      "loss": 0.3893,
+      "step": 938
+    },
+    {
+      "epoch": 0.07512,
+      "grad_norm": 2.0541799068450928,
+      "learning_rate": 9.86411569315759e-06,
+      "loss": 0.4582,
+      "step": 939
+    },
+    {
+      "epoch": 0.0752,
+      "grad_norm": 1.439474105834961,
+      "learning_rate": 9.863824332725413e-06,
+      "loss": 0.3332,
+      "step": 940
+    },
+    {
+      "epoch": 0.07528,
+      "grad_norm": 1.6011489629745483,
+      "learning_rate": 9.863532664575346e-06,
+      "loss": 0.3182,
+      "step": 941
+    },
+    {
+      "epoch": 0.07536,
+      "grad_norm": 1.7721103429794312,
+      "learning_rate": 9.863240688725839e-06,
+      "loss": 0.2911,
+      "step": 942
+    },
+    {
+      "epoch": 0.07544,
+      "grad_norm": 1.915481448173523,
+      "learning_rate": 9.862948405195367e-06,
+      "loss": 0.4128,
+      "step": 943
+    },
+    {
+      "epoch": 0.07552,
+      "grad_norm": 1.525109887123108,
+      "learning_rate": 9.862655814002421e-06,
+      "loss": 0.2976,
+      "step": 944
+    },
+    {
+      "epoch": 0.0756,
+      "grad_norm": 1.5049928426742554,
+      "learning_rate": 9.862362915165513e-06,
+      "loss": 0.3732,
+      "step": 945
+    },
+    {
+      "epoch": 0.07568,
+      "grad_norm": 1.1610263586044312,
+      "learning_rate": 9.862069708703172e-06,
+      "loss": 0.2572,
+      "step": 946
+    },
+    {
+      "epoch": 0.07576,
+      "grad_norm": 1.5599390268325806,
+      "learning_rate": 9.861776194633948e-06,
+      "loss": 0.3857,
+      "step": 947
+    },
+    {
+      "epoch": 0.07584,
+      "grad_norm": 1.5780138969421387,
+      "learning_rate": 9.861482372976413e-06,
+      "loss": 0.3915,
+      "step": 948
+    },
+    {
+      "epoch": 0.07592,
+      "grad_norm": 1.3370410203933716,
+      "learning_rate": 9.861188243749154e-06,
+      "loss": 0.3541,
+      "step": 949
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 1.5612949132919312,
+      "learning_rate": 9.86089380697078e-06,
+      "loss": 0.3142,
+      "step": 950
+    },
+    {
+      "epoch": 0.07608,
+      "grad_norm": 2.2083637714385986,
+      "learning_rate": 9.860599062659922e-06,
+      "loss": 0.4862,
+      "step": 951
+    },
+    {
+      "epoch": 0.07616,
+      "grad_norm": 1.7525696754455566,
+      "learning_rate": 9.860304010835222e-06,
+      "loss": 0.3501,
+      "step": 952
+    },
+    {
+      "epoch": 0.07624,
+      "grad_norm": 1.5072181224822998,
+      "learning_rate": 9.860008651515352e-06,
+      "loss": 0.3569,
+      "step": 953
+    },
+    {
+      "epoch": 0.07632,
+      "grad_norm": 2.169581890106201,
+      "learning_rate": 9.859712984718994e-06,
+      "loss": 0.4334,
+      "step": 954
+    },
+    {
+      "epoch": 0.0764,
+      "grad_norm": 1.740846037864685,
+      "learning_rate": 9.859417010464857e-06,
+      "loss": 0.3597,
+      "step": 955
+    },
+    {
+      "epoch": 0.07648,
+      "grad_norm": 1.516270637512207,
+      "learning_rate": 9.859120728771667e-06,
+      "loss": 0.3523,
+      "step": 956
+    },
+    {
+      "epoch": 0.07656,
+      "grad_norm": 2.101482629776001,
+      "learning_rate": 9.858824139658166e-06,
+      "loss": 0.4112,
+      "step": 957
+    },
+    {
+      "epoch": 0.07664,
+      "grad_norm": 1.6379002332687378,
+      "learning_rate": 9.85852724314312e-06,
+      "loss": 0.3163,
+      "step": 958
+    },
+    {
+      "epoch": 0.07672,
+      "grad_norm": 1.6774054765701294,
+      "learning_rate": 9.858230039245312e-06,
+      "loss": 0.2885,
+      "step": 959
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 1.9318816661834717,
+      "learning_rate": 9.857932527983544e-06,
+      "loss": 0.4156,
+      "step": 960
+    },
+    {
+      "epoch": 0.07688,
+      "grad_norm": 1.6440109014511108,
+      "learning_rate": 9.85763470937664e-06,
+      "loss": 0.397,
+      "step": 961
+    },
+    {
+      "epoch": 0.07696,
+      "grad_norm": 1.682336449623108,
+      "learning_rate": 9.857336583443441e-06,
+      "loss": 0.3695,
+      "step": 962
+    },
+    {
+      "epoch": 0.07704,
+      "grad_norm": 1.555406093597412,
+      "learning_rate": 9.85703815020281e-06,
+      "loss": 0.304,
+      "step": 963
+    },
+    {
+      "epoch": 0.07712,
+      "grad_norm": 1.527746319770813,
+      "learning_rate": 9.856739409673628e-06,
+      "loss": 0.3563,
+      "step": 964
+    },
+    {
+      "epoch": 0.0772,
+      "grad_norm": 2.2929794788360596,
+      "learning_rate": 9.856440361874791e-06,
+      "loss": 0.449,
+      "step": 965
+    },
+    {
+      "epoch": 0.07728,
+      "grad_norm": 1.6270774602890015,
+      "learning_rate": 9.856141006825225e-06,
+      "loss": 0.3649,
+      "step": 966
+    },
+    {
+      "epoch": 0.07736,
+      "grad_norm": 1.5416566133499146,
+      "learning_rate": 9.855841344543865e-06,
+      "loss": 0.3216,
+      "step": 967
+    },
+    {
+      "epoch": 0.07744,
+      "grad_norm": 1.646708607673645,
+      "learning_rate": 9.855541375049671e-06,
+      "loss": 0.4936,
+      "step": 968
+    },
+    {
+      "epoch": 0.07752,
+      "grad_norm": 1.7581204175949097,
+      "learning_rate": 9.85524109836162e-06,
+      "loss": 0.3516,
+      "step": 969
+    },
+    {
+      "epoch": 0.0776,
+      "grad_norm": 1.6109693050384521,
+      "learning_rate": 9.854940514498712e-06,
+      "loss": 0.3182,
+      "step": 970
+    },
+    {
+      "epoch": 0.07768,
+      "grad_norm": 1.38651442527771,
+      "learning_rate": 9.854639623479962e-06,
+      "loss": 0.2693,
+      "step": 971
+    },
+    {
+      "epoch": 0.07776,
+      "grad_norm": 1.5283217430114746,
+      "learning_rate": 9.854338425324405e-06,
+      "loss": 0.3448,
+      "step": 972
+    },
+    {
+      "epoch": 0.07784,
+      "grad_norm": 1.3900861740112305,
+      "learning_rate": 9.854036920051102e-06,
+      "loss": 0.2597,
+      "step": 973
+    },
+    {
+      "epoch": 0.07792,
+      "grad_norm": 1.3767924308776855,
+      "learning_rate": 9.85373510767912e-06,
+      "loss": 0.3011,
+      "step": 974
+    },
+    {
+      "epoch": 0.078,
+      "grad_norm": 1.4808169603347778,
+      "learning_rate": 9.853432988227563e-06,
+      "loss": 0.3063,
+      "step": 975
+    },
+    {
+      "epoch": 0.07808,
+      "grad_norm": 1.500436782836914,
+      "learning_rate": 9.853130561715538e-06,
+      "loss": 0.3745,
+      "step": 976
+    },
+    {
+      "epoch": 0.07816,
+      "grad_norm": 1.8564033508300781,
+      "learning_rate": 9.852827828162182e-06,
+      "loss": 0.4051,
+      "step": 977
+    },
+    {
+      "epoch": 0.07824,
+      "grad_norm": 2.207338809967041,
+      "learning_rate": 9.852524787586645e-06,
+      "loss": 0.4495,
+      "step": 978
+    },
+    {
+      "epoch": 0.07832,
+      "grad_norm": 1.2589378356933594,
+      "learning_rate": 9.852221440008103e-06,
+      "loss": 0.3034,
+      "step": 979
+    },
+    {
+      "epoch": 0.0784,
+      "grad_norm": 1.5934374332427979,
+      "learning_rate": 9.851917785445745e-06,
+      "loss": 0.3168,
+      "step": 980
+    },
+    {
+      "epoch": 0.07848,
+      "grad_norm": 1.6377482414245605,
+      "learning_rate": 9.851613823918785e-06,
+      "loss": 0.3054,
+      "step": 981
+    },
+    {
+      "epoch": 0.07856,
+      "grad_norm": 1.398160696029663,
+      "learning_rate": 9.85130955544645e-06,
+      "loss": 0.3111,
+      "step": 982
+    },
+    {
+      "epoch": 0.07864,
+      "grad_norm": 1.5453648567199707,
+      "learning_rate": 9.851004980047993e-06,
+      "loss": 0.3174,
+      "step": 983
+    },
+    {
+      "epoch": 0.07872,
+      "grad_norm": 1.4585448503494263,
+      "learning_rate": 9.850700097742683e-06,
+      "loss": 0.3859,
+      "step": 984
+    },
+    {
+      "epoch": 0.0788,
+      "grad_norm": 1.838578224182129,
+      "learning_rate": 9.850394908549808e-06,
+      "loss": 0.431,
+      "step": 985
+    },
+    {
+      "epoch": 0.07888,
+      "grad_norm": 1.7750165462493896,
+      "learning_rate": 9.850089412488676e-06,
+      "loss": 0.3653,
+      "step": 986
+    },
+    {
+      "epoch": 0.07896,
+      "grad_norm": 1.7502179145812988,
+      "learning_rate": 9.849783609578616e-06,
+      "loss": 0.4266,
+      "step": 987
+    },
+    {
+      "epoch": 0.07904,
+      "grad_norm": 1.66646409034729,
+      "learning_rate": 9.849477499838974e-06,
+      "loss": 0.3501,
+      "step": 988
+    },
+    {
+      "epoch": 0.07912,
+      "grad_norm": 2.072134256362915,
+      "learning_rate": 9.849171083289117e-06,
+      "loss": 0.4357,
+      "step": 989
+    },
+    {
+      "epoch": 0.0792,
+      "grad_norm": 1.6113394498825073,
+      "learning_rate": 9.84886435994843e-06,
+      "loss": 0.3318,
+      "step": 990
+    },
+    {
+      "epoch": 0.07928,
+      "grad_norm": 1.862763524055481,
+      "learning_rate": 9.84855732983632e-06,
+      "loss": 0.34,
+      "step": 991
+    },
+    {
+      "epoch": 0.07936,
+      "grad_norm": 1.8371633291244507,
+      "learning_rate": 9.848249992972212e-06,
+      "loss": 0.3153,
+      "step": 992
+    },
+    {
+      "epoch": 0.07944,
+      "grad_norm": 1.5195534229278564,
+      "learning_rate": 9.847942349375549e-06,
+      "loss": 0.3188,
+      "step": 993
+    },
+    {
+      "epoch": 0.07952,
+      "grad_norm": 1.642816185951233,
+      "learning_rate": 9.847634399065794e-06,
+      "loss": 0.3539,
+      "step": 994
+    },
+    {
+      "epoch": 0.0796,
+      "grad_norm": 1.4042744636535645,
+      "learning_rate": 9.84732614206243e-06,
+      "loss": 0.3128,
+      "step": 995
+    },
+    {
+      "epoch": 0.07968,
+      "grad_norm": 1.9033617973327637,
+      "learning_rate": 9.847017578384961e-06,
+      "loss": 0.395,
+      "step": 996
+    },
+    {
+      "epoch": 0.07976,
+      "grad_norm": 2.087599039077759,
+      "learning_rate": 9.846708708052908e-06,
+      "loss": 0.3815,
+      "step": 997
+    },
+    {
+      "epoch": 0.07984,
+      "grad_norm": 1.8923081159591675,
+      "learning_rate": 9.846399531085812e-06,
+      "loss": 0.3115,
+      "step": 998
+    },
+    {
+      "epoch": 0.07992,
+      "grad_norm": 1.654727816581726,
+      "learning_rate": 9.846090047503235e-06,
+      "loss": 0.2935,
+      "step": 999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.230875015258789,
+      "learning_rate": 9.845780257324755e-06,
+      "loss": 0.3925,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08008,
+      "grad_norm": 1.9971858263015747,
+      "learning_rate": 9.845470160569973e-06,
+      "loss": 0.3956,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08016,
+      "grad_norm": 1.6523634195327759,
+      "learning_rate": 9.845159757258505e-06,
+      "loss": 0.443,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08024,
+      "grad_norm": 2.058262586593628,
+      "learning_rate": 9.844849047409993e-06,
+      "loss": 0.3834,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08032,
+      "grad_norm": 1.6032230854034424,
+      "learning_rate": 9.844538031044092e-06,
+      "loss": 0.3474,
+      "step": 1004
+    },
+    {
+      "epoch": 0.0804,
+      "grad_norm": 1.7125388383865356,
+      "learning_rate": 9.84422670818048e-06,
+      "loss": 0.4696,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08048,
+      "grad_norm": 1.2934495210647583,
+      "learning_rate": 9.843915078838852e-06,
+      "loss": 0.3361,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08056,
+      "grad_norm": 1.6751182079315186,
+      "learning_rate": 9.843603143038925e-06,
+      "loss": 0.3279,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08064,
+      "grad_norm": 1.739035964012146,
+      "learning_rate": 9.843290900800436e-06,
+      "loss": 0.3179,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08072,
+      "grad_norm": 1.8006422519683838,
+      "learning_rate": 9.842978352143133e-06,
+      "loss": 0.3985,
+      "step": 1009
+    },
+    {
+      "epoch": 0.0808,
+      "grad_norm": 1.3257513046264648,
+      "learning_rate": 9.842665497086798e-06,
+      "loss": 0.3005,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08088,
+      "grad_norm": 1.31377112865448,
+      "learning_rate": 9.84235233565122e-06,
+      "loss": 0.2922,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08096,
+      "grad_norm": 1.3837974071502686,
+      "learning_rate": 9.842038867856211e-06,
+      "loss": 0.2946,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08104,
+      "grad_norm": 1.7588199377059937,
+      "learning_rate": 9.841725093721606e-06,
+      "loss": 0.4315,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08112,
+      "grad_norm": 1.4498939514160156,
+      "learning_rate": 9.841411013267252e-06,
+      "loss": 0.3537,
+      "step": 1014
+    },
+    {
+      "epoch": 0.0812,
+      "grad_norm": 1.9055722951889038,
+      "learning_rate": 9.841096626513024e-06,
+      "loss": 0.4242,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08128,
+      "grad_norm": 2.0020415782928467,
+      "learning_rate": 9.840781933478813e-06,
+      "loss": 0.4011,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08136,
+      "grad_norm": 1.7482478618621826,
+      "learning_rate": 9.840466934184525e-06,
+      "loss": 0.3524,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08144,
+      "grad_norm": 1.8049051761627197,
+      "learning_rate": 9.84015162865009e-06,
+      "loss": 0.3151,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08152,
+      "grad_norm": 1.8534055948257446,
+      "learning_rate": 9.839836016895457e-06,
+      "loss": 0.3722,
+      "step": 1019
+    },
+    {
+      "epoch": 0.0816,
+      "grad_norm": 1.7496310472488403,
+      "learning_rate": 9.839520098940593e-06,
+      "loss": 0.4035,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08168,
+      "grad_norm": 2.2628374099731445,
+      "learning_rate": 9.839203874805486e-06,
+      "loss": 0.4903,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08176,
+      "grad_norm": 1.4022674560546875,
+      "learning_rate": 9.838887344510139e-06,
+      "loss": 0.314,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08184,
+      "grad_norm": 1.7159093618392944,
+      "learning_rate": 9.838570508074584e-06,
+      "loss": 0.4232,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08192,
+      "grad_norm": 1.4128509759902954,
+      "learning_rate": 9.838253365518862e-06,
+      "loss": 0.3225,
+      "step": 1024
+    },
+    {
+      "epoch": 0.082,
+      "grad_norm": 1.8229515552520752,
+      "learning_rate": 9.837935916863038e-06,
+      "loss": 0.3784,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08208,
+      "grad_norm": 1.3460335731506348,
+      "learning_rate": 9.837618162127196e-06,
+      "loss": 0.2683,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08216,
+      "grad_norm": 1.5191764831542969,
+      "learning_rate": 9.83730010133144e-06,
+      "loss": 0.3418,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08224,
+      "grad_norm": 1.3885241746902466,
+      "learning_rate": 9.836981734495895e-06,
+      "loss": 0.2815,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08232,
+      "grad_norm": 2.1471521854400635,
+      "learning_rate": 9.836663061640697e-06,
+      "loss": 0.3671,
+      "step": 1029
+    },
+    {
+      "epoch": 0.0824,
+      "grad_norm": 1.2107166051864624,
+      "learning_rate": 9.83634408278601e-06,
+      "loss": 0.2604,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08248,
+      "grad_norm": 1.9330739974975586,
+      "learning_rate": 9.836024797952017e-06,
+      "loss": 0.5037,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08256,
+      "grad_norm": 1.541945219039917,
+      "learning_rate": 9.835705207158916e-06,
+      "loss": 0.3424,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08264,
+      "grad_norm": 2.0090341567993164,
+      "learning_rate": 9.835385310426928e-06,
+      "loss": 0.4266,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08272,
+      "grad_norm": 1.8531626462936401,
+      "learning_rate": 9.835065107776289e-06,
+      "loss": 0.4106,
+      "step": 1034
+    },
+    {
+      "epoch": 0.0828,
+      "grad_norm": 1.9602084159851074,
+      "learning_rate": 9.83474459922726e-06,
+      "loss": 0.3546,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08288,
+      "grad_norm": 1.5967477560043335,
+      "learning_rate": 9.834423784800115e-06,
+      "loss": 0.4163,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08296,
+      "grad_norm": 1.9444454908370972,
+      "learning_rate": 9.834102664515155e-06,
+      "loss": 0.396,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08304,
+      "grad_norm": 1.4539134502410889,
+      "learning_rate": 9.833781238392695e-06,
+      "loss": 0.3087,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08312,
+      "grad_norm": 1.5706698894500732,
+      "learning_rate": 9.833459506453069e-06,
+      "loss": 0.3336,
+      "step": 1039
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 1.93034827709198,
+      "learning_rate": 9.833137468716634e-06,
+      "loss": 0.402,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08328,
+      "grad_norm": 1.7693796157836914,
+      "learning_rate": 9.832815125203761e-06,
+      "loss": 0.4466,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08336,
+      "grad_norm": 1.506583571434021,
+      "learning_rate": 9.832492475934848e-06,
+      "loss": 0.3304,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08344,
+      "grad_norm": 1.3750256299972534,
+      "learning_rate": 9.832169520930303e-06,
+      "loss": 0.2565,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08352,
+      "grad_norm": 1.999240517616272,
+      "learning_rate": 9.831846260210563e-06,
+      "loss": 0.486,
+      "step": 1044
+    },
+    {
+      "epoch": 0.0836,
+      "grad_norm": 1.6762970685958862,
+      "learning_rate": 9.831522693796077e-06,
+      "loss": 0.3922,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08368,
+      "grad_norm": 1.6641404628753662,
+      "learning_rate": 9.831198821707316e-06,
+      "loss": 0.3165,
+      "step": 1046
+    },
+    {
+      "epoch": 0.08376,
+      "grad_norm": 1.7150557041168213,
+      "learning_rate": 9.83087464396477e-06,
+      "loss": 0.3909,
+      "step": 1047
+    },
+    {
+      "epoch": 0.08384,
+      "grad_norm": 1.5680177211761475,
+      "learning_rate": 9.830550160588951e-06,
+      "loss": 0.4403,
+      "step": 1048
+    },
+    {
+      "epoch": 0.08392,
+      "grad_norm": 1.7904253005981445,
+      "learning_rate": 9.830225371600386e-06,
+      "loss": 0.3263,
+      "step": 1049
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 1.3248636722564697,
+      "learning_rate": 9.829900277019624e-06,
+      "loss": 0.2865,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08408,
+      "grad_norm": 1.825286865234375,
+      "learning_rate": 9.829574876867232e-06,
+      "loss": 0.4188,
+      "step": 1051
+    },
+    {
+      "epoch": 0.08416,
+      "grad_norm": 1.7181260585784912,
+      "learning_rate": 9.829249171163798e-06,
+      "loss": 0.3737,
+      "step": 1052
+    },
+    {
+      "epoch": 0.08424,
+      "grad_norm": 2.284151077270508,
+      "learning_rate": 9.828923159929927e-06,
+      "loss": 0.4879,
+      "step": 1053
+    },
+    {
+      "epoch": 0.08432,
+      "grad_norm": 1.6576015949249268,
+      "learning_rate": 9.828596843186244e-06,
+      "loss": 0.3157,
+      "step": 1054
+    },
+    {
+      "epoch": 0.0844,
+      "grad_norm": 1.5160009860992432,
+      "learning_rate": 9.828270220953398e-06,
+      "loss": 0.2853,
+      "step": 1055
+    },
+    {
+      "epoch": 0.08448,
+      "grad_norm": 1.5151618719100952,
+      "learning_rate": 9.827943293252048e-06,
+      "loss": 0.4335,
+      "step": 1056
+    },
+    {
+      "epoch": 0.08456,
+      "grad_norm": 1.3789680004119873,
+      "learning_rate": 9.82761606010288e-06,
+      "loss": 0.3956,
+      "step": 1057
+    },
+    {
+      "epoch": 0.08464,
+      "grad_norm": 2.0407497882843018,
+      "learning_rate": 9.8272885215266e-06,
+      "loss": 0.4707,
+      "step": 1058
+    },
+    {
+      "epoch": 0.08472,
+      "grad_norm": 1.4480390548706055,
+      "learning_rate": 9.826960677543926e-06,
+      "loss": 0.3389,
+      "step": 1059
+    },
+    {
+      "epoch": 0.0848,
+      "grad_norm": 1.438524603843689,
+      "learning_rate": 9.8266325281756e-06,
+      "loss": 0.2841,
+      "step": 1060
+    },
+    {
+      "epoch": 0.08488,
+      "grad_norm": 1.9115839004516602,
+      "learning_rate": 9.826304073442385e-06,
+      "loss": 0.3926,
+      "step": 1061
+    },
+    {
+      "epoch": 0.08496,
+      "grad_norm": 2.102163314819336,
+      "learning_rate": 9.82597531336506e-06,
+      "loss": 0.4392,
+      "step": 1062
+    },
+    {
+      "epoch": 0.08504,
+      "grad_norm": 1.6580944061279297,
+      "learning_rate": 9.825646247964425e-06,
+      "loss": 0.3485,
+      "step": 1063
+    },
+    {
+      "epoch": 0.08512,
+      "grad_norm": 1.469714879989624,
+      "learning_rate": 9.825316877261298e-06,
+      "loss": 0.3193,
+      "step": 1064
+    },
+    {
+      "epoch": 0.0852,
+      "grad_norm": 1.4311915636062622,
+      "learning_rate": 9.824987201276519e-06,
+      "loss": 0.2758,
+      "step": 1065
+    },
+    {
+      "epoch": 0.08528,
+      "grad_norm": 1.8289291858673096,
+      "learning_rate": 9.824657220030942e-06,
+      "loss": 0.4056,
+      "step": 1066
+    },
+    {
+      "epoch": 0.08536,
+      "grad_norm": 1.6348884105682373,
+      "learning_rate": 9.824326933545448e-06,
+      "loss": 0.2958,
+      "step": 1067
+    },
+    {
+      "epoch": 0.08544,
+      "grad_norm": 1.3245985507965088,
+      "learning_rate": 9.823996341840929e-06,
+      "loss": 0.2905,
+      "step": 1068
+    },
+    {
+      "epoch": 0.08552,
+      "grad_norm": 1.329889178276062,
+      "learning_rate": 9.823665444938304e-06,
+      "loss": 0.3558,
+      "step": 1069
+    },
+    {
+      "epoch": 0.0856,
+      "grad_norm": 1.8176319599151611,
+      "learning_rate": 9.823334242858506e-06,
+      "loss": 0.3617,
+      "step": 1070
+    },
+    {
+      "epoch": 0.08568,
+      "grad_norm": 2.28464674949646,
+      "learning_rate": 9.82300273562249e-06,
+      "loss": 0.4581,
+      "step": 1071
+    },
+    {
+      "epoch": 0.08576,
+      "grad_norm": 1.3536391258239746,
+      "learning_rate": 9.822670923251228e-06,
+      "loss": 0.3201,
+      "step": 1072
+    },
+    {
+      "epoch": 0.08584,
+      "grad_norm": 1.6658986806869507,
+      "learning_rate": 9.822338805765714e-06,
+      "loss": 0.4,
+      "step": 1073
+    },
+    {
+      "epoch": 0.08592,
+      "grad_norm": 1.4204604625701904,
+      "learning_rate": 9.82200638318696e-06,
+      "loss": 0.3206,
+      "step": 1074
+    },
+    {
+      "epoch": 0.086,
+      "grad_norm": 1.6750638484954834,
+      "learning_rate": 9.821673655535995e-06,
+      "loss": 0.3092,
+      "step": 1075
+    },
+    {
+      "epoch": 0.08608,
+      "grad_norm": 1.8198531866073608,
+      "learning_rate": 9.821340622833873e-06,
+      "loss": 0.3966,
+      "step": 1076
+    },
+    {
+      "epoch": 0.08616,
+      "grad_norm": 1.710455298423767,
+      "learning_rate": 9.82100728510166e-06,
+      "loss": 0.2862,
+      "step": 1077
+    },
+    {
+      "epoch": 0.08624,
+      "grad_norm": 1.3838942050933838,
+      "learning_rate": 9.820673642360448e-06,
+      "loss": 0.3097,
+      "step": 1078
+    },
+    {
+      "epoch": 0.08632,
+      "grad_norm": 1.588562250137329,
+      "learning_rate": 9.820339694631345e-06,
+      "loss": 0.4134,
+      "step": 1079
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 1.470871925354004,
+      "learning_rate": 9.820005441935479e-06,
+      "loss": 0.2925,
+      "step": 1080
+    },
+    {
+      "epoch": 0.08648,
+      "grad_norm": 1.691702127456665,
+      "learning_rate": 9.819670884293994e-06,
+      "loss": 0.3222,
+      "step": 1081
+    },
+    {
+      "epoch": 0.08656,
+      "grad_norm": 1.7869762182235718,
+      "learning_rate": 9.819336021728062e-06,
+      "loss": 0.3606,
+      "step": 1082
+    },
+    {
+      "epoch": 0.08664,
+      "grad_norm": 1.437152624130249,
+      "learning_rate": 9.819000854258864e-06,
+      "loss": 0.2756,
+      "step": 1083
+    },
+    {
+      "epoch": 0.08672,
+      "grad_norm": 1.7799644470214844,
+      "learning_rate": 9.818665381907605e-06,
+      "loss": 0.3988,
+      "step": 1084
+    },
+    {
+      "epoch": 0.0868,
+      "grad_norm": 1.8911408185958862,
+      "learning_rate": 9.818329604695513e-06,
+      "loss": 0.4247,
+      "step": 1085
+    },
+    {
+      "epoch": 0.08688,
+      "grad_norm": 2.2056448459625244,
+      "learning_rate": 9.817993522643827e-06,
+      "loss": 0.3675,
+      "step": 1086
+    },
+    {
+      "epoch": 0.08696,
+      "grad_norm": 1.294008731842041,
+      "learning_rate": 9.817657135773813e-06,
+      "loss": 0.3124,
+      "step": 1087
+    },
+    {
+      "epoch": 0.08704,
+      "grad_norm": 1.4826951026916504,
+      "learning_rate": 9.817320444106753e-06,
+      "loss": 0.3155,
+      "step": 1088
+    },
+    {
+      "epoch": 0.08712,
+      "grad_norm": 1.5381442308425903,
+      "learning_rate": 9.816983447663946e-06,
+      "loss": 0.4008,
+      "step": 1089
+    },
+    {
+      "epoch": 0.0872,
+      "grad_norm": 1.5871357917785645,
+      "learning_rate": 9.816646146466714e-06,
+      "loss": 0.333,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08728,
+      "grad_norm": 1.603145718574524,
+      "learning_rate": 9.816308540536396e-06,
+      "loss": 0.3314,
+      "step": 1091
+    },
+    {
+      "epoch": 0.08736,
+      "grad_norm": 1.7113319635391235,
+      "learning_rate": 9.815970629894354e-06,
+      "loss": 0.408,
+      "step": 1092
+    },
+    {
+      "epoch": 0.08744,
+      "grad_norm": 1.4441734552383423,
+      "learning_rate": 9.815632414561964e-06,
+      "loss": 0.2921,
+      "step": 1093
+    },
+    {
+      "epoch": 0.08752,
+      "grad_norm": 1.7782763242721558,
+      "learning_rate": 9.815293894560623e-06,
+      "loss": 0.3528,
+      "step": 1094
+    },
+    {
+      "epoch": 0.0876,
+      "grad_norm": 1.4475657939910889,
+      "learning_rate": 9.814955069911752e-06,
+      "loss": 0.3025,
+      "step": 1095
+    },
+    {
+      "epoch": 0.08768,
+      "grad_norm": 1.3749080896377563,
+      "learning_rate": 9.814615940636781e-06,
+      "loss": 0.277,
+      "step": 1096
+    },
+    {
+      "epoch": 0.08776,
+      "grad_norm": 1.6869237422943115,
+      "learning_rate": 9.814276506757172e-06,
+      "loss": 0.4035,
+      "step": 1097
+    },
+    {
+      "epoch": 0.08784,
+      "grad_norm": 1.6214927434921265,
+      "learning_rate": 9.813936768294397e-06,
+      "loss": 0.3524,
+      "step": 1098
+    },
+    {
+      "epoch": 0.08792,
+      "grad_norm": 2.009049654006958,
+      "learning_rate": 9.813596725269948e-06,
+      "loss": 0.3494,
+      "step": 1099
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 1.901945948600769,
+      "learning_rate": 9.813256377705341e-06,
+      "loss": 0.4302,
+      "step": 1100
+    },
+    {
+      "epoch": 0.08808,
+      "grad_norm": 2.3681435585021973,
+      "learning_rate": 9.812915725622109e-06,
+      "loss": 0.5687,
+      "step": 1101
+    },
+    {
+      "epoch": 0.08816,
+      "grad_norm": 2.2910284996032715,
+      "learning_rate": 9.812574769041805e-06,
+      "loss": 0.4986,
+      "step": 1102
+    },
+    {
+      "epoch": 0.08824,
+      "grad_norm": 1.6809210777282715,
+      "learning_rate": 9.812233507985995e-06,
+      "loss": 0.3687,
+      "step": 1103
+    },
+    {
+      "epoch": 0.08832,
+      "grad_norm": 1.779154658317566,
+      "learning_rate": 9.811891942476275e-06,
+      "loss": 0.4001,
+      "step": 1104
+    },
+    {
+      "epoch": 0.0884,
+      "grad_norm": 1.6831563711166382,
+      "learning_rate": 9.811550072534251e-06,
+      "loss": 0.3435,
+      "step": 1105
+    },
+    {
+      "epoch": 0.08848,
+      "grad_norm": 1.9863594770431519,
+      "learning_rate": 9.811207898181555e-06,
+      "loss": 0.3946,
+      "step": 1106
+    },
+    {
+      "epoch": 0.08856,
+      "grad_norm": 1.761527419090271,
+      "learning_rate": 9.81086541943983e-06,
+      "loss": 0.316,
+      "step": 1107
+    },
+    {
+      "epoch": 0.08864,
+      "grad_norm": 1.4106063842773438,
+      "learning_rate": 9.810522636330751e-06,
+      "loss": 0.2762,
+      "step": 1108
+    },
+    {
+      "epoch": 0.08872,
+      "grad_norm": 1.7829062938690186,
+      "learning_rate": 9.810179548875999e-06,
+      "loss": 0.3734,
+      "step": 1109
+    },
+    {
+      "epoch": 0.0888,
+      "grad_norm": 1.5982229709625244,
+      "learning_rate": 9.809836157097282e-06,
+      "loss": 0.3701,
+      "step": 1110
+    },
+    {
+      "epoch": 0.08888,
+      "grad_norm": 1.4014500379562378,
+      "learning_rate": 9.809492461016326e-06,
+      "loss": 0.3202,
+      "step": 1111
+    },
+    {
+      "epoch": 0.08896,
+      "grad_norm": 1.6743066310882568,
+      "learning_rate": 9.809148460654874e-06,
+      "loss": 0.2995,
+      "step": 1112
+    },
+    {
+      "epoch": 0.08904,
+      "grad_norm": 1.6122390031814575,
+      "learning_rate": 9.80880415603469e-06,
+      "loss": 0.3507,
+      "step": 1113
+    },
+    {
+      "epoch": 0.08912,
+      "grad_norm": 1.8418755531311035,
+      "learning_rate": 9.808459547177559e-06,
+      "loss": 0.4162,
+      "step": 1114
+    },
+    {
+      "epoch": 0.0892,
+      "grad_norm": 1.4559040069580078,
+      "learning_rate": 9.808114634105278e-06,
+      "loss": 0.3797,
+      "step": 1115
+    },
+    {
+      "epoch": 0.08928,
+      "grad_norm": 1.9068504571914673,
+      "learning_rate": 9.807769416839677e-06,
+      "loss": 0.4154,
+      "step": 1116
+    },
+    {
+      "epoch": 0.08936,
+      "grad_norm": 1.5469180345535278,
+      "learning_rate": 9.807423895402587e-06,
+      "loss": 0.4295,
+      "step": 1117
+    },
+    {
+      "epoch": 0.08944,
+      "grad_norm": 1.7214394807815552,
+      "learning_rate": 9.807078069815877e-06,
+      "loss": 0.3601,
+      "step": 1118
+    },
+    {
+      "epoch": 0.08952,
+      "grad_norm": 1.7878152132034302,
+      "learning_rate": 9.80673194010142e-06,
+      "loss": 0.3575,
+      "step": 1119
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 1.857308030128479,
+      "learning_rate": 9.806385506281117e-06,
+      "loss": 0.4363,
+      "step": 1120
+    },
+    {
+      "epoch": 0.08968,
+      "grad_norm": 1.6544005870819092,
+      "learning_rate": 9.806038768376885e-06,
+      "loss": 0.3477,
+      "step": 1121
+    },
+    {
+      "epoch": 0.08976,
+      "grad_norm": 1.5055336952209473,
+      "learning_rate": 9.80569172641066e-06,
+      "loss": 0.2681,
+      "step": 1122
+    },
+    {
+      "epoch": 0.08984,
+      "grad_norm": 1.9073718786239624,
+      "learning_rate": 9.8053443804044e-06,
+      "loss": 0.4285,
+      "step": 1123
+    },
+    {
+      "epoch": 0.08992,
+      "grad_norm": 1.2498080730438232,
+      "learning_rate": 9.80499673038008e-06,
+      "loss": 0.2639,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5033406019210815,
+      "learning_rate": 9.804648776359695e-06,
+      "loss": 0.3395,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09008,
+      "grad_norm": 1.9122231006622314,
+      "learning_rate": 9.80430051836526e-06,
+      "loss": 0.3571,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09016,
+      "grad_norm": 2.6971442699432373,
+      "learning_rate": 9.803951956418803e-06,
+      "loss": 0.4251,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09024,
+      "grad_norm": 1.4795504808425903,
+      "learning_rate": 9.803603090542381e-06,
+      "loss": 0.3596,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09032,
+      "grad_norm": 1.6504251956939697,
+      "learning_rate": 9.803253920758064e-06,
+      "loss": 0.3429,
+      "step": 1129
+    },
+    {
+      "epoch": 0.0904,
+      "grad_norm": 2.284956693649292,
+      "learning_rate": 9.802904447087945e-06,
+      "loss": 0.6959,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09048,
+      "grad_norm": 1.777529001235962,
+      "learning_rate": 9.802554669554131e-06,
+      "loss": 0.3375,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09056,
+      "grad_norm": 1.831652045249939,
+      "learning_rate": 9.802204588178752e-06,
+      "loss": 0.4083,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09064,
+      "grad_norm": 2.0457098484039307,
+      "learning_rate": 9.801854202983957e-06,
+      "loss": 0.3995,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09072,
+      "grad_norm": 1.3770190477371216,
+      "learning_rate": 9.801503513991914e-06,
+      "loss": 0.2523,
+      "step": 1134
+    },
+    {
+      "epoch": 0.0908,
+      "grad_norm": 1.600197434425354,
+      "learning_rate": 9.80115252122481e-06,
+      "loss": 0.3447,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09088,
+      "grad_norm": 1.7964802980422974,
+      "learning_rate": 9.800801224704851e-06,
+      "loss": 0.4085,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09096,
+      "grad_norm": 1.558449149131775,
+      "learning_rate": 9.800449624454262e-06,
+      "loss": 0.3532,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09104,
+      "grad_norm": 1.844244360923767,
+      "learning_rate": 9.800097720495286e-06,
+      "loss": 0.3174,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09112,
+      "grad_norm": 1.8442306518554688,
+      "learning_rate": 9.79974551285019e-06,
+      "loss": 0.3615,
+      "step": 1139
+    },
+    {
+      "epoch": 0.0912,
+      "grad_norm": 1.5331751108169556,
+      "learning_rate": 9.799393001541255e-06,
+      "loss": 0.3207,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09128,
+      "grad_norm": 1.578279972076416,
+      "learning_rate": 9.799040186590782e-06,
+      "loss": 0.4399,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09136,
+      "grad_norm": 1.4967174530029297,
+      "learning_rate": 9.798687068021095e-06,
+      "loss": 0.3297,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09144,
+      "grad_norm": 2.0132367610931396,
+      "learning_rate": 9.798333645854536e-06,
+      "loss": 0.3207,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09152,
+      "grad_norm": 1.6801670789718628,
+      "learning_rate": 9.79797992011346e-06,
+      "loss": 0.3193,
+      "step": 1144
+    },
+    {
+      "epoch": 0.0916,
+      "grad_norm": 2.1940245628356934,
+      "learning_rate": 9.797625890820249e-06,
+      "loss": 0.4476,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09168,
+      "grad_norm": 1.625232458114624,
+      "learning_rate": 9.7972715579973e-06,
+      "loss": 0.2663,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09176,
+      "grad_norm": 2.0687153339385986,
+      "learning_rate": 9.796916921667033e-06,
+      "loss": 0.3539,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09184,
+      "grad_norm": 1.9847198724746704,
+      "learning_rate": 9.796561981851882e-06,
+      "loss": 0.4134,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09192,
+      "grad_norm": 1.644768238067627,
+      "learning_rate": 9.796206738574303e-06,
+      "loss": 0.2589,
+      "step": 1149
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 1.431854486465454,
+      "learning_rate": 9.795851191856774e-06,
+      "loss": 0.3136,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09208,
+      "grad_norm": 1.3832859992980957,
+      "learning_rate": 9.795495341721784e-06,
+      "loss": 0.3832,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09216,
+      "grad_norm": 1.497043251991272,
+      "learning_rate": 9.795139188191851e-06,
+      "loss": 0.367,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09224,
+      "grad_norm": 1.3032704591751099,
+      "learning_rate": 9.794782731289507e-06,
+      "loss": 0.2557,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09232,
+      "grad_norm": 1.7582857608795166,
+      "learning_rate": 9.794425971037303e-06,
+      "loss": 0.3647,
+      "step": 1154
+    },
+    {
+      "epoch": 0.0924,
+      "grad_norm": 1.4522085189819336,
+      "learning_rate": 9.794068907457809e-06,
+      "loss": 0.2847,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09248,
+      "grad_norm": 1.3870216608047485,
+      "learning_rate": 9.793711540573616e-06,
+      "loss": 0.2989,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09256,
+      "grad_norm": 1.3415980339050293,
+      "learning_rate": 9.793353870407335e-06,
+      "loss": 0.2967,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09264,
+      "grad_norm": 1.8516277074813843,
+      "learning_rate": 9.792995896981591e-06,
+      "loss": 0.3464,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09272,
+      "grad_norm": 1.5445516109466553,
+      "learning_rate": 9.792637620319037e-06,
+      "loss": 0.3213,
+      "step": 1159
+    },
+    {
+      "epoch": 0.0928,
+      "grad_norm": 1.8748204708099365,
+      "learning_rate": 9.792279040442334e-06,
+      "loss": 0.4657,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09288,
+      "grad_norm": 1.5904078483581543,
+      "learning_rate": 9.791920157374173e-06,
+      "loss": 0.3865,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09296,
+      "grad_norm": 1.2793481349945068,
+      "learning_rate": 9.791560971137257e-06,
+      "loss": 0.3927,
+      "step": 1162
+    },
+    {
+      "epoch": 0.09304,
+      "grad_norm": 1.3844211101531982,
+      "learning_rate": 9.791201481754312e-06,
+      "loss": 0.2778,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09312,
+      "grad_norm": 1.733962059020996,
+      "learning_rate": 9.790841689248078e-06,
+      "loss": 0.3492,
+      "step": 1164
+    },
+    {
+      "epoch": 0.0932,
+      "grad_norm": 1.4930529594421387,
+      "learning_rate": 9.790481593641324e-06,
+      "loss": 0.2962,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09328,
+      "grad_norm": 1.9883126020431519,
+      "learning_rate": 9.790121194956825e-06,
+      "loss": 0.3775,
+      "step": 1166
+    },
+    {
+      "epoch": 0.09336,
+      "grad_norm": 1.5635162591934204,
+      "learning_rate": 9.789760493217388e-06,
+      "loss": 0.327,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09344,
+      "grad_norm": 1.5043566226959229,
+      "learning_rate": 9.78939948844583e-06,
+      "loss": 0.2745,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09352,
+      "grad_norm": 1.3067989349365234,
+      "learning_rate": 9.789038180664994e-06,
+      "loss": 0.2482,
+      "step": 1169
+    },
+    {
+      "epoch": 0.0936,
+      "grad_norm": 1.678215503692627,
+      "learning_rate": 9.788676569897734e-06,
+      "loss": 0.316,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09368,
+      "grad_norm": 1.8970040082931519,
+      "learning_rate": 9.788314656166931e-06,
+      "loss": 0.4448,
+      "step": 1171
+    },
+    {
+      "epoch": 0.09376,
+      "grad_norm": 1.6337584257125854,
+      "learning_rate": 9.787952439495481e-06,
+      "loss": 0.402,
+      "step": 1172
+    },
+    {
+      "epoch": 0.09384,
+      "grad_norm": 2.054110050201416,
+      "learning_rate": 9.787589919906301e-06,
+      "loss": 0.4844,
+      "step": 1173
+    },
+    {
+      "epoch": 0.09392,
+      "grad_norm": 1.683643102645874,
+      "learning_rate": 9.787227097422327e-06,
+      "loss": 0.4237,
+      "step": 1174
+    },
+    {
+      "epoch": 0.094,
+      "grad_norm": 1.800292730331421,
+      "learning_rate": 9.786863972066515e-06,
+      "loss": 0.4866,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09408,
+      "grad_norm": 1.3618955612182617,
+      "learning_rate": 9.786500543861833e-06,
+      "loss": 0.3326,
+      "step": 1176
+    },
+    {
+      "epoch": 0.09416,
+      "grad_norm": 1.6615434885025024,
+      "learning_rate": 9.786136812831276e-06,
+      "loss": 0.3495,
+      "step": 1177
+    },
+    {
+      "epoch": 0.09424,
+      "grad_norm": 1.6406981945037842,
+      "learning_rate": 9.78577277899786e-06,
+      "loss": 0.3495,
+      "step": 1178
+    },
+    {
+      "epoch": 0.09432,
+      "grad_norm": 1.4927983283996582,
+      "learning_rate": 9.785408442384612e-06,
+      "loss": 0.3832,
+      "step": 1179
+    },
+    {
+      "epoch": 0.0944,
+      "grad_norm": 1.31818425655365,
+      "learning_rate": 9.785043803014584e-06,
+      "loss": 0.3047,
+      "step": 1180
+    },
+    {
+      "epoch": 0.09448,
+      "grad_norm": 1.7068977355957031,
+      "learning_rate": 9.784678860910846e-06,
+      "loss": 0.3715,
+      "step": 1181
+    },
+    {
+      "epoch": 0.09456,
+      "grad_norm": 2.2386581897735596,
+      "learning_rate": 9.784313616096486e-06,
+      "loss": 0.3983,
+      "step": 1182
+    },
+    {
+      "epoch": 0.09464,
+      "grad_norm": 1.369645595550537,
+      "learning_rate": 9.783948068594613e-06,
+      "loss": 0.3607,
+      "step": 1183
+    },
+    {
+      "epoch": 0.09472,
+      "grad_norm": 1.6752623319625854,
+      "learning_rate": 9.783582218428352e-06,
+      "loss": 0.4052,
+      "step": 1184
+    },
+    {
+      "epoch": 0.0948,
+      "grad_norm": 1.740079641342163,
+      "learning_rate": 9.783216065620849e-06,
+      "loss": 0.3649,
+      "step": 1185
+    },
+    {
+      "epoch": 0.09488,
+      "grad_norm": 1.514141321182251,
+      "learning_rate": 9.78284961019527e-06,
+      "loss": 0.3816,
+      "step": 1186
+    },
+    {
+      "epoch": 0.09496,
+      "grad_norm": 1.8777161836624146,
+      "learning_rate": 9.782482852174802e-06,
+      "loss": 0.3526,
+      "step": 1187
+    },
+    {
+      "epoch": 0.09504,
+      "grad_norm": 2.190656900405884,
+      "learning_rate": 9.782115791582644e-06,
+      "loss": 0.4483,
+      "step": 1188
+    },
+    {
+      "epoch": 0.09512,
+      "grad_norm": 1.8506275415420532,
+      "learning_rate": 9.781748428442022e-06,
+      "loss": 0.3753,
+      "step": 1189
+    },
+    {
+      "epoch": 0.0952,
+      "grad_norm": 1.515964150428772,
+      "learning_rate": 9.781380762776176e-06,
+      "loss": 0.4088,
+      "step": 1190
+    },
+    {
+      "epoch": 0.09528,
+      "grad_norm": 1.1206365823745728,
+      "learning_rate": 9.781012794608368e-06,
+      "loss": 0.2067,
+      "step": 1191
+    },
+    {
+      "epoch": 0.09536,
+      "grad_norm": 1.286902666091919,
+      "learning_rate": 9.780644523961877e-06,
+      "loss": 0.2787,
+      "step": 1192
+    },
+    {
+      "epoch": 0.09544,
+      "grad_norm": 1.765989065170288,
+      "learning_rate": 9.780275950860005e-06,
+      "loss": 0.3091,
+      "step": 1193
+    },
+    {
+      "epoch": 0.09552,
+      "grad_norm": 2.0813465118408203,
+      "learning_rate": 9.779907075326066e-06,
+      "loss": 0.3626,
+      "step": 1194
+    },
+    {
+      "epoch": 0.0956,
+      "grad_norm": 1.8357547521591187,
+      "learning_rate": 9.779537897383403e-06,
+      "loss": 0.3385,
+      "step": 1195
+    },
+    {
+      "epoch": 0.09568,
+      "grad_norm": 2.120861530303955,
+      "learning_rate": 9.779168417055368e-06,
+      "loss": 0.462,
+      "step": 1196
+    },
+    {
+      "epoch": 0.09576,
+      "grad_norm": 1.4538586139678955,
+      "learning_rate": 9.778798634365336e-06,
+      "loss": 0.3281,
+      "step": 1197
+    },
+    {
+      "epoch": 0.09584,
+      "grad_norm": 2.0800697803497314,
+      "learning_rate": 9.778428549336707e-06,
+      "loss": 0.4369,
+      "step": 1198
+    },
+    {
+      "epoch": 0.09592,
+      "grad_norm": 1.2680668830871582,
+      "learning_rate": 9.778058161992892e-06,
+      "loss": 0.2548,
+      "step": 1199
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 1.8447930812835693,
+      "learning_rate": 9.777687472357324e-06,
+      "loss": 0.5501,
+      "step": 1200
+    },
+    {
+      "epoch": 0.09608,
+      "grad_norm": 2.0600554943084717,
+      "learning_rate": 9.777316480453457e-06,
+      "loss": 0.604,
+      "step": 1201
+    },
+    {
+      "epoch": 0.09616,
+      "grad_norm": 1.6585683822631836,
+      "learning_rate": 9.77694518630476e-06,
+      "loss": 0.3472,
+      "step": 1202
+    },
+    {
+      "epoch": 0.09624,
+      "grad_norm": 1.3858237266540527,
+      "learning_rate": 9.776573589934726e-06,
+      "loss": 0.3476,
+      "step": 1203
+    },
+    {
+      "epoch": 0.09632,
+      "grad_norm": 1.6310840845108032,
+      "learning_rate": 9.776201691366863e-06,
+      "loss": 0.3252,
+      "step": 1204
+    },
+    {
+      "epoch": 0.0964,
+      "grad_norm": 1.4285833835601807,
+      "learning_rate": 9.775829490624698e-06,
+      "loss": 0.2611,
+      "step": 1205
+    },
+    {
+      "epoch": 0.09648,
+      "grad_norm": 1.5389498472213745,
+      "learning_rate": 9.775456987731784e-06,
+      "loss": 0.3895,
+      "step": 1206
+    },
+    {
+      "epoch": 0.09656,
+      "grad_norm": 1.5493495464324951,
+      "learning_rate": 9.775084182711683e-06,
+      "loss": 0.3191,
+      "step": 1207
+    },
+    {
+      "epoch": 0.09664,
+      "grad_norm": 1.5432543754577637,
+      "learning_rate": 9.774711075587985e-06,
+      "loss": 0.4227,
+      "step": 1208
+    },
+    {
+      "epoch": 0.09672,
+      "grad_norm": 1.4613178968429565,
+      "learning_rate": 9.774337666384293e-06,
+      "loss": 0.3943,
+      "step": 1209
+    },
+    {
+      "epoch": 0.0968,
+      "grad_norm": 2.0416183471679688,
+      "learning_rate": 9.773963955124232e-06,
+      "loss": 0.5029,
+      "step": 1210
+    },
+    {
+      "epoch": 0.09688,
+      "grad_norm": 1.3715733289718628,
+      "learning_rate": 9.773589941831446e-06,
+      "loss": 0.3508,
+      "step": 1211
+    },
+    {
+      "epoch": 0.09696,
+      "grad_norm": 1.3506369590759277,
+      "learning_rate": 9.773215626529596e-06,
+      "loss": 0.3365,
+      "step": 1212
+    },
+    {
+      "epoch": 0.09704,
+      "grad_norm": 1.867976427078247,
+      "learning_rate": 9.772841009242362e-06,
+      "loss": 0.3752,
+      "step": 1213
+    },
+    {
+      "epoch": 0.09712,
+      "grad_norm": 1.6031577587127686,
+      "learning_rate": 9.772466089993451e-06,
+      "loss": 0.407,
+      "step": 1214
+    },
+    {
+      "epoch": 0.0972,
+      "grad_norm": 1.4785380363464355,
+      "learning_rate": 9.772090868806578e-06,
+      "loss": 0.2738,
+      "step": 1215
+    },
+    {
+      "epoch": 0.09728,
+      "grad_norm": 1.7080274820327759,
+      "learning_rate": 9.771715345705482e-06,
+      "loss": 0.3905,
+      "step": 1216
+    },
+    {
+      "epoch": 0.09736,
+      "grad_norm": 1.4409105777740479,
+      "learning_rate": 9.771339520713924e-06,
+      "loss": 0.2903,
+      "step": 1217
+    },
+    {
+      "epoch": 0.09744,
+      "grad_norm": 2.0454001426696777,
+      "learning_rate": 9.77096339385568e-06,
+      "loss": 0.4558,
+      "step": 1218
+    },
+    {
+      "epoch": 0.09752,
+      "grad_norm": 1.4832075834274292,
+      "learning_rate": 9.770586965154542e-06,
+      "loss": 0.3588,
+      "step": 1219
+    },
+    {
+      "epoch": 0.0976,
+      "grad_norm": 1.7369167804718018,
+      "learning_rate": 9.770210234634333e-06,
+      "loss": 0.3038,
+      "step": 1220
+    },
+    {
+      "epoch": 0.09768,
+      "grad_norm": 1.412338137626648,
+      "learning_rate": 9.769833202318882e-06,
+      "loss": 0.27,
+      "step": 1221
+    },
+    {
+      "epoch": 0.09776,
+      "grad_norm": 2.267824649810791,
+      "learning_rate": 9.769455868232044e-06,
+      "loss": 0.6121,
+      "step": 1222
+    },
+    {
+      "epoch": 0.09784,
+      "grad_norm": 1.5539761781692505,
+      "learning_rate": 9.769078232397693e-06,
+      "loss": 0.3358,
+      "step": 1223
+    },
+    {
+      "epoch": 0.09792,
+      "grad_norm": 1.4364866018295288,
+      "learning_rate": 9.76870029483972e-06,
+      "loss": 0.3932,
+      "step": 1224
+    },
+    {
+      "epoch": 0.098,
+      "grad_norm": 1.2992466688156128,
+      "learning_rate": 9.768322055582034e-06,
+      "loss": 0.2755,
+      "step": 1225
+    },
+    {
+      "epoch": 0.09808,
+      "grad_norm": 1.7094968557357788,
+      "learning_rate": 9.767943514648567e-06,
+      "loss": 0.3883,
+      "step": 1226
+    },
+    {
+      "epoch": 0.09816,
+      "grad_norm": 1.9164812564849854,
+      "learning_rate": 9.767564672063268e-06,
+      "loss": 0.4088,
+      "step": 1227
+    },
+    {
+      "epoch": 0.09824,
+      "grad_norm": 2.1447086334228516,
+      "learning_rate": 9.767185527850103e-06,
+      "loss": 0.4737,
+      "step": 1228
+    },
+    {
+      "epoch": 0.09832,
+      "grad_norm": 1.7146050930023193,
+      "learning_rate": 9.766806082033061e-06,
+      "loss": 0.3838,
+      "step": 1229
+    },
+    {
+      "epoch": 0.0984,
+      "grad_norm": 1.4320570230484009,
+      "learning_rate": 9.766426334636149e-06,
+      "loss": 0.3106,
+      "step": 1230
+    },
+    {
+      "epoch": 0.09848,
+      "grad_norm": 1.5674022436141968,
+      "learning_rate": 9.76604628568339e-06,
+      "loss": 0.3787,
+      "step": 1231
+    },
+    {
+      "epoch": 0.09856,
+      "grad_norm": 1.2451162338256836,
+      "learning_rate": 9.765665935198831e-06,
+      "loss": 0.3102,
+      "step": 1232
+    },
+    {
+      "epoch": 0.09864,
+      "grad_norm": 1.3635990619659424,
+      "learning_rate": 9.765285283206533e-06,
+      "loss": 0.2954,
+      "step": 1233
+    },
+    {
+      "epoch": 0.09872,
+      "grad_norm": 2.4409143924713135,
+      "learning_rate": 9.764904329730583e-06,
+      "loss": 0.3997,
+      "step": 1234
+    },
+    {
+      "epoch": 0.0988,
+      "grad_norm": 1.705710530281067,
+      "learning_rate": 9.764523074795077e-06,
+      "loss": 0.4332,
+      "step": 1235
+    },
+    {
+      "epoch": 0.09888,
+      "grad_norm": 1.3760895729064941,
+      "learning_rate": 9.764141518424138e-06,
+      "loss": 0.3509,
+      "step": 1236
+    },
+    {
+      "epoch": 0.09896,
+      "grad_norm": 1.7472193241119385,
+      "learning_rate": 9.763759660641905e-06,
+      "loss": 0.3773,
+      "step": 1237
+    },
+    {
+      "epoch": 0.09904,
+      "grad_norm": 1.6968291997909546,
+      "learning_rate": 9.76337750147254e-06,
+      "loss": 0.3579,
+      "step": 1238
+    },
+    {
+      "epoch": 0.09912,
+      "grad_norm": 1.7409708499908447,
+      "learning_rate": 9.762995040940217e-06,
+      "loss": 0.2909,
+      "step": 1239
+    },
+    {
+      "epoch": 0.0992,
+      "grad_norm": 1.7769598960876465,
+      "learning_rate": 9.762612279069136e-06,
+      "loss": 0.3435,
+      "step": 1240
+    },
+    {
+      "epoch": 0.09928,
+      "grad_norm": 1.4426788091659546,
+      "learning_rate": 9.762229215883511e-06,
+      "loss": 0.3389,
+      "step": 1241
+    },
+    {
+      "epoch": 0.09936,
+      "grad_norm": 1.7437098026275635,
+      "learning_rate": 9.76184585140758e-06,
+      "loss": 0.3582,
+      "step": 1242
+    },
+    {
+      "epoch": 0.09944,
+      "grad_norm": 1.8013086318969727,
+      "learning_rate": 9.761462185665593e-06,
+      "loss": 0.3392,
+      "step": 1243
+    },
+    {
+      "epoch": 0.09952,
+      "grad_norm": 1.4805794954299927,
+      "learning_rate": 9.761078218681827e-06,
+      "loss": 0.2881,
+      "step": 1244
+    },
+    {
+      "epoch": 0.0996,
+      "grad_norm": 1.754407525062561,
+      "learning_rate": 9.760693950480572e-06,
+      "loss": 0.3884,
+      "step": 1245
+    },
+    {
+      "epoch": 0.09968,
+      "grad_norm": 1.5800998210906982,
+      "learning_rate": 9.760309381086139e-06,
+      "loss": 0.3402,
+      "step": 1246
+    },
+    {
+      "epoch": 0.09976,
+      "grad_norm": 1.5772699117660522,
+      "learning_rate": 9.759924510522861e-06,
+      "loss": 0.3164,
+      "step": 1247
+    },
+    {
+      "epoch": 0.09984,
+      "grad_norm": 1.4671974182128906,
+      "learning_rate": 9.759539338815085e-06,
+      "loss": 0.3409,
+      "step": 1248
+    },
+    {
+      "epoch": 0.09992,
+      "grad_norm": 1.8091286420822144,
+      "learning_rate": 9.75915386598718e-06,
+      "loss": 0.3817,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8199232816696167,
+      "learning_rate": 9.758768092063536e-06,
+      "loss": 0.3736,
+      "step": 1250
+    },
+    {
+      "epoch": 0.10008,
+      "grad_norm": 1.4588650465011597,
+      "learning_rate": 9.758382017068558e-06,
+      "loss": 0.3103,
+      "step": 1251
+    },
+    {
+      "epoch": 0.10016,
+      "grad_norm": 2.0028064250946045,
+      "learning_rate": 9.757995641026669e-06,
+      "loss": 0.4277,
+      "step": 1252
+    },
+    {
+      "epoch": 0.10024,
+      "grad_norm": 1.8617825508117676,
+      "learning_rate": 9.757608963962317e-06,
+      "loss": 0.3659,
+      "step": 1253
+    },
+    {
+      "epoch": 0.10032,
+      "grad_norm": 1.7988715171813965,
+      "learning_rate": 9.757221985899965e-06,
+      "loss": 0.3815,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1004,
+      "grad_norm": 1.3747351169586182,
+      "learning_rate": 9.756834706864096e-06,
+      "loss": 0.3078,
+      "step": 1255
+    },
+    {
+      "epoch": 0.10048,
+      "grad_norm": 1.8129006624221802,
+      "learning_rate": 9.756447126879212e-06,
+      "loss": 0.3561,
+      "step": 1256
+    },
+    {
+      "epoch": 0.10056,
+      "grad_norm": 1.850455641746521,
+      "learning_rate": 9.756059245969832e-06,
+      "loss": 0.4652,
+      "step": 1257
+    },
+    {
+      "epoch": 0.10064,
+      "grad_norm": 1.7385430335998535,
+      "learning_rate": 9.755671064160499e-06,
+      "loss": 0.3239,
+      "step": 1258
+    },
+    {
+      "epoch": 0.10072,
+      "grad_norm": 1.8935037851333618,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.4487,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1008,
+      "grad_norm": 1.537541389465332,
+      "learning_rate": 9.754893797940222e-06,
+      "loss": 0.2818,
+      "step": 1260
+    },
+    {
+      "epoch": 0.10088,
+      "grad_norm": 1.770788550376892,
+      "learning_rate": 9.754504713578453e-06,
+      "loss": 0.3322,
+      "step": 1261
+    },
+    {
+      "epoch": 0.10096,
+      "grad_norm": 1.3864850997924805,
+      "learning_rate": 9.75411532841508e-06,
+      "loss": 0.2596,
+      "step": 1262
+    },
+    {
+      "epoch": 0.10104,
+      "grad_norm": 1.4967447519302368,
+      "learning_rate": 9.753725642474739e-06,
+      "loss": 0.3948,
+      "step": 1263
+    },
+    {
+      "epoch": 0.10112,
+      "grad_norm": 1.707395315170288,
+      "learning_rate": 9.75333565578208e-06,
+      "loss": 0.4174,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1012,
+      "grad_norm": 1.876741886138916,
+      "learning_rate": 9.752945368361782e-06,
+      "loss": 0.4069,
+      "step": 1265
+    },
+    {
+      "epoch": 0.10128,
+      "grad_norm": 1.4767177104949951,
+      "learning_rate": 9.75255478023853e-06,
+      "loss": 0.3826,
+      "step": 1266
+    },
+    {
+      "epoch": 0.10136,
+      "grad_norm": 2.0608136653900146,
+      "learning_rate": 9.752163891437042e-06,
+      "loss": 0.4269,
+      "step": 1267
+    },
+    {
+      "epoch": 0.10144,
+      "grad_norm": 1.9878028631210327,
+      "learning_rate": 9.751772701982045e-06,
+      "loss": 0.4665,
+      "step": 1268
+    },
+    {
+      "epoch": 0.10152,
+      "grad_norm": 1.6732946634292603,
+      "learning_rate": 9.751381211898288e-06,
+      "loss": 0.3332,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1016,
+      "grad_norm": 1.626977562904358,
+      "learning_rate": 9.75098942121054e-06,
+      "loss": 0.4315,
+      "step": 1270
+    },
+    {
+      "epoch": 0.10168,
+      "grad_norm": 1.6960049867630005,
+      "learning_rate": 9.750597329943588e-06,
+      "loss": 0.328,
+      "step": 1271
+    },
+    {
+      "epoch": 0.10176,
+      "grad_norm": 1.4753273725509644,
+      "learning_rate": 9.75020493812224e-06,
+      "loss": 0.3749,
+      "step": 1272
+    },
+    {
+      "epoch": 0.10184,
+      "grad_norm": 1.60029935836792,
+      "learning_rate": 9.749812245771318e-06,
+      "loss": 0.37,
+      "step": 1273
+    },
+    {
+      "epoch": 0.10192,
+      "grad_norm": 1.379223346710205,
+      "learning_rate": 9.749419252915668e-06,
+      "loss": 0.2877,
+      "step": 1274
+    },
+    {
+      "epoch": 0.102,
+      "grad_norm": 1.588037371635437,
+      "learning_rate": 9.749025959580156e-06,
+      "loss": 0.3533,
+      "step": 1275
+    },
+    {
+      "epoch": 0.10208,
+      "grad_norm": 1.8960047960281372,
+      "learning_rate": 9.748632365789658e-06,
+      "loss": 0.5091,
+      "step": 1276
+    },
+    {
+      "epoch": 0.10216,
+      "grad_norm": 1.6755867004394531,
+      "learning_rate": 9.748238471569083e-06,
+      "loss": 0.3203,
+      "step": 1277
+    },
+    {
+      "epoch": 0.10224,
+      "grad_norm": 1.5891623497009277,
+      "learning_rate": 9.747844276943345e-06,
+      "loss": 0.4389,
+      "step": 1278
+    },
+    {
+      "epoch": 0.10232,
+      "grad_norm": 1.7560958862304688,
+      "learning_rate": 9.747449781937388e-06,
+      "loss": 0.281,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 1.9748413562774658,
+      "learning_rate": 9.747054986576165e-06,
+      "loss": 0.5097,
+      "step": 1280
+    },
+    {
+      "epoch": 0.10248,
+      "grad_norm": 1.5240323543548584,
+      "learning_rate": 9.74665989088466e-06,
+      "loss": 0.3438,
+      "step": 1281
+    },
+    {
+      "epoch": 0.10256,
+      "grad_norm": 1.2884881496429443,
+      "learning_rate": 9.746264494887865e-06,
+      "loss": 0.3211,
+      "step": 1282
+    },
+    {
+      "epoch": 0.10264,
+      "grad_norm": 2.2344231605529785,
+      "learning_rate": 9.745868798610796e-06,
+      "loss": 0.3818,
+      "step": 1283
+    },
+    {
+      "epoch": 0.10272,
+      "grad_norm": 1.9735519886016846,
+      "learning_rate": 9.745472802078488e-06,
+      "loss": 0.35,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1028,
+      "grad_norm": 1.7784732580184937,
+      "learning_rate": 9.745076505315994e-06,
+      "loss": 0.4224,
+      "step": 1285
+    },
+    {
+      "epoch": 0.10288,
+      "grad_norm": 1.9323874711990356,
+      "learning_rate": 9.744679908348386e-06,
+      "loss": 0.4125,
+      "step": 1286
+    },
+    {
+      "epoch": 0.10296,
+      "grad_norm": 2.1199615001678467,
+      "learning_rate": 9.74428301120076e-06,
+      "loss": 0.4652,
+      "step": 1287
+    },
+    {
+      "epoch": 0.10304,
+      "grad_norm": 1.5500082969665527,
+      "learning_rate": 9.743885813898217e-06,
+      "loss": 0.3081,
+      "step": 1288
+    },
+    {
+      "epoch": 0.10312,
+      "grad_norm": 1.9076387882232666,
+      "learning_rate": 9.743488316465895e-06,
+      "loss": 0.4051,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1032,
+      "grad_norm": 1.48446524143219,
+      "learning_rate": 9.743090518928937e-06,
+      "loss": 0.3234,
+      "step": 1290
+    },
+    {
+      "epoch": 0.10328,
+      "grad_norm": 2.2340919971466064,
+      "learning_rate": 9.742692421312515e-06,
+      "loss": 0.4365,
+      "step": 1291
+    },
+    {
+      "epoch": 0.10336,
+      "grad_norm": 1.6090425252914429,
+      "learning_rate": 9.74229402364181e-06,
+      "loss": 0.3622,
+      "step": 1292
+    },
+    {
+      "epoch": 0.10344,
+      "grad_norm": 1.7896775007247925,
+      "learning_rate": 9.74189532594203e-06,
+      "loss": 0.388,
+      "step": 1293
+    },
+    {
+      "epoch": 0.10352,
+      "grad_norm": 1.6529676914215088,
+      "learning_rate": 9.7414963282384e-06,
+      "loss": 0.2979,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1036,
+      "grad_norm": 1.5958153009414673,
+      "learning_rate": 9.741097030556162e-06,
+      "loss": 0.3039,
+      "step": 1295
+    },
+    {
+      "epoch": 0.10368,
+      "grad_norm": 2.0807502269744873,
+      "learning_rate": 9.740697432920579e-06,
+      "loss": 0.4092,
+      "step": 1296
+    },
+    {
+      "epoch": 0.10376,
+      "grad_norm": 1.3676483631134033,
+      "learning_rate": 9.740297535356931e-06,
+      "loss": 0.2693,
+      "step": 1297
+    },
+    {
+      "epoch": 0.10384,
+      "grad_norm": 2.016907215118408,
+      "learning_rate": 9.739897337890521e-06,
+      "loss": 0.3837,
+      "step": 1298
+    },
+    {
+      "epoch": 0.10392,
+      "grad_norm": 1.2759778499603271,
+      "learning_rate": 9.739496840546663e-06,
+      "loss": 0.2593,
+      "step": 1299
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 1.7022594213485718,
+      "learning_rate": 9.7390960433507e-06,
+      "loss": 0.3582,
+      "step": 1300
+    },
+    {
+      "epoch": 0.10408,
+      "grad_norm": 1.783856987953186,
+      "learning_rate": 9.738694946327988e-06,
+      "loss": 0.4139,
+      "step": 1301
+    },
+    {
+      "epoch": 0.10416,
+      "grad_norm": 1.6456480026245117,
+      "learning_rate": 9.738293549503902e-06,
+      "loss": 0.3018,
+      "step": 1302
+    },
+    {
+      "epoch": 0.10424,
+      "grad_norm": 1.666909098625183,
+      "learning_rate": 9.737891852903838e-06,
+      "loss": 0.3686,
+      "step": 1303
+    },
+    {
+      "epoch": 0.10432,
+      "grad_norm": 1.3941386938095093,
+      "learning_rate": 9.737489856553209e-06,
+      "loss": 0.4,
+      "step": 1304
+    },
+    {
+      "epoch": 0.1044,
+      "grad_norm": 1.745068907737732,
+      "learning_rate": 9.737087560477449e-06,
+      "loss": 0.3254,
+      "step": 1305
+    },
+    {
+      "epoch": 0.10448,
+      "grad_norm": 1.4899107217788696,
+      "learning_rate": 9.736684964702008e-06,
+      "loss": 0.3617,
+      "step": 1306
+    },
+    {
+      "epoch": 0.10456,
+      "grad_norm": 1.8580790758132935,
+      "learning_rate": 9.736282069252358e-06,
+      "loss": 0.368,
+      "step": 1307
+    },
+    {
+      "epoch": 0.10464,
+      "grad_norm": 1.2298766374588013,
+      "learning_rate": 9.735878874153993e-06,
+      "loss": 0.293,
+      "step": 1308
+    },
+    {
+      "epoch": 0.10472,
+      "grad_norm": 1.4866394996643066,
+      "learning_rate": 9.735475379432414e-06,
+      "loss": 0.3504,
+      "step": 1309
+    },
+    {
+      "epoch": 0.1048,
+      "grad_norm": 1.508781909942627,
+      "learning_rate": 9.735071585113153e-06,
+      "loss": 0.3257,
+      "step": 1310
+    },
+    {
+      "epoch": 0.10488,
+      "grad_norm": 1.6359803676605225,
+      "learning_rate": 9.734667491221758e-06,
+      "loss": 0.3885,
+      "step": 1311
+    },
+    {
+      "epoch": 0.10496,
+      "grad_norm": 1.50149405002594,
+      "learning_rate": 9.734263097783792e-06,
+      "loss": 0.3049,
+      "step": 1312
+    },
+    {
+      "epoch": 0.10504,
+      "grad_norm": 1.7158201932907104,
+      "learning_rate": 9.73385840482484e-06,
+      "loss": 0.4445,
+      "step": 1313
+    },
+    {
+      "epoch": 0.10512,
+      "grad_norm": 1.7673109769821167,
+      "learning_rate": 9.733453412370508e-06,
+      "loss": 0.3097,
+      "step": 1314
+    },
+    {
+      "epoch": 0.1052,
+      "grad_norm": 1.5726984739303589,
+      "learning_rate": 9.733048120446416e-06,
+      "loss": 0.2946,
+      "step": 1315
+    },
+    {
+      "epoch": 0.10528,
+      "grad_norm": 1.9193859100341797,
+      "learning_rate": 9.732642529078206e-06,
+      "loss": 0.4489,
+      "step": 1316
+    },
+    {
+      "epoch": 0.10536,
+      "grad_norm": 1.6108776330947876,
+      "learning_rate": 9.73223663829154e-06,
+      "loss": 0.3922,
+      "step": 1317
+    },
+    {
+      "epoch": 0.10544,
+      "grad_norm": 1.4639272689819336,
+      "learning_rate": 9.731830448112096e-06,
+      "loss": 0.3506,
+      "step": 1318
+    },
+    {
+      "epoch": 0.10552,
+      "grad_norm": 1.342962622642517,
+      "learning_rate": 9.731423958565571e-06,
+      "loss": 0.3522,
+      "step": 1319
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 1.7002488374710083,
+      "learning_rate": 9.731017169677683e-06,
+      "loss": 0.4434,
+      "step": 1320
+    },
+    {
+      "epoch": 0.10568,
+      "grad_norm": 1.669387698173523,
+      "learning_rate": 9.73061008147417e-06,
+      "loss": 0.3224,
+      "step": 1321
+    },
+    {
+      "epoch": 0.10576,
+      "grad_norm": 1.4158101081848145,
+      "learning_rate": 9.730202693980786e-06,
+      "loss": 0.3246,
+      "step": 1322
+    },
+    {
+      "epoch": 0.10584,
+      "grad_norm": 1.3361846208572388,
+      "learning_rate": 9.729795007223303e-06,
+      "loss": 0.269,
+      "step": 1323
+    },
+    {
+      "epoch": 0.10592,
+      "grad_norm": 1.3655201196670532,
+      "learning_rate": 9.729387021227518e-06,
+      "loss": 0.3433,
+      "step": 1324
+    },
+    {
+      "epoch": 0.106,
+      "grad_norm": 1.260786771774292,
+      "learning_rate": 9.728978736019238e-06,
+      "loss": 0.2933,
+      "step": 1325
+    },
+    {
+      "epoch": 0.10608,
+      "grad_norm": 1.3825486898422241,
+      "learning_rate": 9.7285701516243e-06,
+      "loss": 0.3286,
+      "step": 1326
+    },
+    {
+      "epoch": 0.10616,
+      "grad_norm": 1.707014799118042,
+      "learning_rate": 9.72816126806855e-06,
+      "loss": 0.447,
+      "step": 1327
+    },
+    {
+      "epoch": 0.10624,
+      "grad_norm": 1.8020687103271484,
+      "learning_rate": 9.727752085377855e-06,
+      "loss": 0.4374,
+      "step": 1328
+    },
+    {
+      "epoch": 0.10632,
+      "grad_norm": 1.697378158569336,
+      "learning_rate": 9.727342603578105e-06,
+      "loss": 0.3393,
+      "step": 1329
+    },
+    {
+      "epoch": 0.1064,
+      "grad_norm": 1.4379756450653076,
+      "learning_rate": 9.726932822695208e-06,
+      "loss": 0.2879,
+      "step": 1330
+    },
+    {
+      "epoch": 0.10648,
+      "grad_norm": 2.371371269226074,
+      "learning_rate": 9.726522742755085e-06,
+      "loss": 0.4398,
+      "step": 1331
+    },
+    {
+      "epoch": 0.10656,
+      "grad_norm": 1.6934208869934082,
+      "learning_rate": 9.726112363783684e-06,
+      "loss": 0.3344,
+      "step": 1332
+    },
+    {
+      "epoch": 0.10664,
+      "grad_norm": 1.745850682258606,
+      "learning_rate": 9.725701685806968e-06,
+      "loss": 0.4684,
+      "step": 1333
+    },
+    {
+      "epoch": 0.10672,
+      "grad_norm": 1.6199334859848022,
+      "learning_rate": 9.725290708850919e-06,
+      "loss": 0.3236,
+      "step": 1334
+    },
+    {
+      "epoch": 0.1068,
+      "grad_norm": 1.4844225645065308,
+      "learning_rate": 9.724879432941536e-06,
+      "loss": 0.3124,
+      "step": 1335
+    },
+    {
+      "epoch": 0.10688,
+      "grad_norm": 1.5500150918960571,
+      "learning_rate": 9.724467858104843e-06,
+      "loss": 0.3436,
+      "step": 1336
+    },
+    {
+      "epoch": 0.10696,
+      "grad_norm": 1.6310945749282837,
+      "learning_rate": 9.724055984366876e-06,
+      "loss": 0.3663,
+      "step": 1337
+    },
+    {
+      "epoch": 0.10704,
+      "grad_norm": 1.5722607374191284,
+      "learning_rate": 9.723643811753693e-06,
+      "loss": 0.3498,
+      "step": 1338
+    },
+    {
+      "epoch": 0.10712,
+      "grad_norm": 1.2501575946807861,
+      "learning_rate": 9.723231340291372e-06,
+      "loss": 0.2408,
+      "step": 1339
+    },
+    {
+      "epoch": 0.1072,
+      "grad_norm": 1.5553503036499023,
+      "learning_rate": 9.722818570006008e-06,
+      "loss": 0.4481,
+      "step": 1340
+    },
+    {
+      "epoch": 0.10728,
+      "grad_norm": 2.1203765869140625,
+      "learning_rate": 9.722405500923715e-06,
+      "loss": 0.4438,
+      "step": 1341
+    },
+    {
+      "epoch": 0.10736,
+      "grad_norm": 1.671330213546753,
+      "learning_rate": 9.721992133070627e-06,
+      "loss": 0.3637,
+      "step": 1342
+    },
+    {
+      "epoch": 0.10744,
+      "grad_norm": 1.697939157485962,
+      "learning_rate": 9.721578466472896e-06,
+      "loss": 0.3393,
+      "step": 1343
+    },
+    {
+      "epoch": 0.10752,
+      "grad_norm": 1.2914625406265259,
+      "learning_rate": 9.721164501156697e-06,
+      "loss": 0.2479,
+      "step": 1344
+    },
+    {
+      "epoch": 0.1076,
+      "grad_norm": 1.7447216510772705,
+      "learning_rate": 9.720750237148214e-06,
+      "loss": 0.5241,
+      "step": 1345
+    },
+    {
+      "epoch": 0.10768,
+      "grad_norm": 1.8008346557617188,
+      "learning_rate": 9.72033567447366e-06,
+      "loss": 0.4449,
+      "step": 1346
+    },
+    {
+      "epoch": 0.10776,
+      "grad_norm": 1.2198455333709717,
+      "learning_rate": 9.719920813159262e-06,
+      "loss": 0.2726,
+      "step": 1347
+    },
+    {
+      "epoch": 0.10784,
+      "grad_norm": 1.7773081064224243,
+      "learning_rate": 9.719505653231268e-06,
+      "loss": 0.3292,
+      "step": 1348
+    },
+    {
+      "epoch": 0.10792,
+      "grad_norm": 1.6153675317764282,
+      "learning_rate": 9.719090194715943e-06,
+      "loss": 0.404,
+      "step": 1349
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 1.9209851026535034,
+      "learning_rate": 9.71867443763957e-06,
+      "loss": 0.4205,
+      "step": 1350
+    },
+    {
+      "epoch": 0.10808,
+      "grad_norm": 1.8615111112594604,
+      "learning_rate": 9.718258382028456e-06,
+      "loss": 0.4368,
+      "step": 1351
+    },
+    {
+      "epoch": 0.10816,
+      "grad_norm": 1.647263526916504,
+      "learning_rate": 9.71784202790892e-06,
+      "loss": 0.5561,
+      "step": 1352
+    },
+    {
+      "epoch": 0.10824,
+      "grad_norm": 1.804128646850586,
+      "learning_rate": 9.717425375307305e-06,
+      "loss": 0.3763,
+      "step": 1353
+    },
+    {
+      "epoch": 0.10832,
+      "grad_norm": 1.5256234407424927,
+      "learning_rate": 9.717008424249973e-06,
+      "loss": 0.3653,
+      "step": 1354
+    },
+    {
+      "epoch": 0.1084,
+      "grad_norm": 1.6643016338348389,
+      "learning_rate": 9.716591174763297e-06,
+      "loss": 0.3654,
+      "step": 1355
+    },
+    {
+      "epoch": 0.10848,
+      "grad_norm": 1.531507968902588,
+      "learning_rate": 9.716173626873682e-06,
+      "loss": 0.2859,
+      "step": 1356
+    },
+    {
+      "epoch": 0.10856,
+      "grad_norm": 1.4687572717666626,
+      "learning_rate": 9.71575578060754e-06,
+      "loss": 0.3194,
+      "step": 1357
+    },
+    {
+      "epoch": 0.10864,
+      "grad_norm": 1.7269558906555176,
+      "learning_rate": 9.715337635991312e-06,
+      "loss": 0.4283,
+      "step": 1358
+    },
+    {
+      "epoch": 0.10872,
+      "grad_norm": 1.9255378246307373,
+      "learning_rate": 9.714919193051448e-06,
+      "loss": 0.3628,
+      "step": 1359
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 1.5584735870361328,
+      "learning_rate": 9.714500451814421e-06,
+      "loss": 0.2875,
+      "step": 1360
+    },
+    {
+      "epoch": 0.10888,
+      "grad_norm": 1.3883323669433594,
+      "learning_rate": 9.714081412306728e-06,
+      "loss": 0.2614,
+      "step": 1361
+    },
+    {
+      "epoch": 0.10896,
+      "grad_norm": 1.938321590423584,
+      "learning_rate": 9.713662074554875e-06,
+      "loss": 0.3478,
+      "step": 1362
+    },
+    {
+      "epoch": 0.10904,
+      "grad_norm": 1.7196362018585205,
+      "learning_rate": 9.713242438585397e-06,
+      "loss": 0.3891,
+      "step": 1363
+    },
+    {
+      "epoch": 0.10912,
+      "grad_norm": 1.8739652633666992,
+      "learning_rate": 9.712822504424839e-06,
+      "loss": 0.4158,
+      "step": 1364
+    },
+    {
+      "epoch": 0.1092,
+      "grad_norm": 1.669812560081482,
+      "learning_rate": 9.71240227209977e-06,
+      "loss": 0.3483,
+      "step": 1365
+    },
+    {
+      "epoch": 0.10928,
+      "grad_norm": 1.673043131828308,
+      "learning_rate": 9.711981741636777e-06,
+      "loss": 0.3858,
+      "step": 1366
+    },
+    {
+      "epoch": 0.10936,
+      "grad_norm": 1.8101592063903809,
+      "learning_rate": 9.711560913062465e-06,
+      "loss": 0.3462,
+      "step": 1367
+    },
+    {
+      "epoch": 0.10944,
+      "grad_norm": 1.4530932903289795,
+      "learning_rate": 9.711139786403461e-06,
+      "loss": 0.2484,
+      "step": 1368
+    },
+    {
+      "epoch": 0.10952,
+      "grad_norm": 1.6506141424179077,
+      "learning_rate": 9.710718361686405e-06,
+      "loss": 0.3754,
+      "step": 1369
+    },
+    {
+      "epoch": 0.1096,
+      "grad_norm": 1.7748082876205444,
+      "learning_rate": 9.71029663893796e-06,
+      "loss": 0.3619,
+      "step": 1370
+    },
+    {
+      "epoch": 0.10968,
+      "grad_norm": 1.5540876388549805,
+      "learning_rate": 9.709874618184808e-06,
+      "loss": 0.3668,
+      "step": 1371
+    },
+    {
+      "epoch": 0.10976,
+      "grad_norm": 1.5610040426254272,
+      "learning_rate": 9.709452299453648e-06,
+      "loss": 0.3208,
+      "step": 1372
+    },
+    {
+      "epoch": 0.10984,
+      "grad_norm": 1.334820032119751,
+      "learning_rate": 9.709029682771198e-06,
+      "loss": 0.2853,
+      "step": 1373
+    },
+    {
+      "epoch": 0.10992,
+      "grad_norm": 1.774716854095459,
+      "learning_rate": 9.708606768164199e-06,
+      "loss": 0.406,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4674115180969238,
+      "learning_rate": 9.708183555659404e-06,
+      "loss": 0.3442,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11008,
+      "grad_norm": 1.6923162937164307,
+      "learning_rate": 9.707760045283587e-06,
+      "loss": 0.4355,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11016,
+      "grad_norm": 1.6160029172897339,
+      "learning_rate": 9.707336237063546e-06,
+      "loss": 0.3239,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11024,
+      "grad_norm": 1.5265469551086426,
+      "learning_rate": 9.70691213102609e-06,
+      "loss": 0.3529,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11032,
+      "grad_norm": 1.5923312902450562,
+      "learning_rate": 9.706487727198055e-06,
+      "loss": 0.3371,
+      "step": 1379
+    },
+    {
+      "epoch": 0.1104,
+      "grad_norm": 1.9004793167114258,
+      "learning_rate": 9.706063025606288e-06,
+      "loss": 0.3516,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11048,
+      "grad_norm": 1.440619945526123,
+      "learning_rate": 9.70563802627766e-06,
+      "loss": 0.2806,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11056,
+      "grad_norm": 1.896283507347107,
+      "learning_rate": 9.705212729239061e-06,
+      "loss": 0.3541,
+      "step": 1382
+    },
+    {
+      "epoch": 0.11064,
+      "grad_norm": 2.1123244762420654,
+      "learning_rate": 9.704787134517396e-06,
+      "loss": 0.3817,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11072,
+      "grad_norm": 2.1019601821899414,
+      "learning_rate": 9.704361242139589e-06,
+      "loss": 0.3864,
+      "step": 1384
+    },
+    {
+      "epoch": 0.1108,
+      "grad_norm": 1.6103894710540771,
+      "learning_rate": 9.703935052132589e-06,
+      "loss": 0.3911,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11088,
+      "grad_norm": 1.6342370510101318,
+      "learning_rate": 9.703508564523356e-06,
+      "loss": 0.3514,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11096,
+      "grad_norm": 1.655190348625183,
+      "learning_rate": 9.703081779338877e-06,
+      "loss": 0.3249,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11104,
+      "grad_norm": 1.3274317979812622,
+      "learning_rate": 9.702654696606147e-06,
+      "loss": 0.3422,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11112,
+      "grad_norm": 1.903498888015747,
+      "learning_rate": 9.702227316352192e-06,
+      "loss": 0.4509,
+      "step": 1389
+    },
+    {
+      "epoch": 0.1112,
+      "grad_norm": 1.5915570259094238,
+      "learning_rate": 9.701799638604048e-06,
+      "loss": 0.3718,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11128,
+      "grad_norm": 1.9381896257400513,
+      "learning_rate": 9.701371663388771e-06,
+      "loss": 0.3438,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11136,
+      "grad_norm": 1.6756565570831299,
+      "learning_rate": 9.700943390733442e-06,
+      "loss": 0.3565,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11144,
+      "grad_norm": 1.537050485610962,
+      "learning_rate": 9.700514820665153e-06,
+      "loss": 0.2721,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11152,
+      "grad_norm": 1.410021424293518,
+      "learning_rate": 9.70008595321102e-06,
+      "loss": 0.2962,
+      "step": 1394
+    },
+    {
+      "epoch": 0.1116,
+      "grad_norm": 1.8756940364837646,
+      "learning_rate": 9.699656788398175e-06,
+      "loss": 0.4885,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11168,
+      "grad_norm": 1.6304752826690674,
+      "learning_rate": 9.69922732625377e-06,
+      "loss": 0.382,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11176,
+      "grad_norm": 1.274293303489685,
+      "learning_rate": 9.698797566804976e-06,
+      "loss": 0.2796,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11184,
+      "grad_norm": 1.7127177715301514,
+      "learning_rate": 9.69836751007898e-06,
+      "loss": 0.3734,
+      "step": 1398
+    },
+    {
+      "epoch": 0.11192,
+      "grad_norm": 1.886913776397705,
+      "learning_rate": 9.697937156102997e-06,
+      "loss": 0.3908,
+      "step": 1399
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 1.5846264362335205,
+      "learning_rate": 9.697506504904246e-06,
+      "loss": 0.3985,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11208,
+      "grad_norm": 1.3665478229522705,
+      "learning_rate": 9.697075556509978e-06,
+      "loss": 0.2967,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11216,
+      "grad_norm": 1.4394631385803223,
+      "learning_rate": 9.696644310947453e-06,
+      "loss": 0.3561,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11224,
+      "grad_norm": 1.8952381610870361,
+      "learning_rate": 9.69621276824396e-06,
+      "loss": 0.5919,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11232,
+      "grad_norm": 1.3404344320297241,
+      "learning_rate": 9.6957809284268e-06,
+      "loss": 0.3089,
+      "step": 1404
+    },
+    {
+      "epoch": 0.1124,
+      "grad_norm": 1.6374577283859253,
+      "learning_rate": 9.69534879152329e-06,
+      "loss": 0.3373,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11248,
+      "grad_norm": 1.6670361757278442,
+      "learning_rate": 9.694916357560774e-06,
+      "loss": 0.4697,
+      "step": 1406
+    },
+    {
+      "epoch": 0.11256,
+      "grad_norm": 2.1049678325653076,
+      "learning_rate": 9.694483626566609e-06,
+      "loss": 0.3929,
+      "step": 1407
+    },
+    {
+      "epoch": 0.11264,
+      "grad_norm": 2.02200984954834,
+      "learning_rate": 9.694050598568173e-06,
+      "loss": 0.503,
+      "step": 1408
+    },
+    {
+      "epoch": 0.11272,
+      "grad_norm": 1.2237039804458618,
+      "learning_rate": 9.693617273592861e-06,
+      "loss": 0.2909,
+      "step": 1409
+    },
+    {
+      "epoch": 0.1128,
+      "grad_norm": 1.4623360633850098,
+      "learning_rate": 9.69318365166809e-06,
+      "loss": 0.2837,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11288,
+      "grad_norm": 1.7014158964157104,
+      "learning_rate": 9.692749732821293e-06,
+      "loss": 0.2977,
+      "step": 1411
+    },
+    {
+      "epoch": 0.11296,
+      "grad_norm": 1.7711082696914673,
+      "learning_rate": 9.692315517079922e-06,
+      "loss": 0.3118,
+      "step": 1412
+    },
+    {
+      "epoch": 0.11304,
+      "grad_norm": 1.5991653203964233,
+      "learning_rate": 9.691881004471449e-06,
+      "loss": 0.3468,
+      "step": 1413
+    },
+    {
+      "epoch": 0.11312,
+      "grad_norm": 1.7637615203857422,
+      "learning_rate": 9.691446195023364e-06,
+      "loss": 0.3192,
+      "step": 1414
+    },
+    {
+      "epoch": 0.1132,
+      "grad_norm": 1.9800078868865967,
+      "learning_rate": 9.691011088763175e-06,
+      "loss": 0.3971,
+      "step": 1415
+    },
+    {
+      "epoch": 0.11328,
+      "grad_norm": 1.2780556678771973,
+      "learning_rate": 9.69057568571841e-06,
+      "loss": 0.2921,
+      "step": 1416
+    },
+    {
+      "epoch": 0.11336,
+      "grad_norm": 1.2494962215423584,
+      "learning_rate": 9.690139985916619e-06,
+      "loss": 0.283,
+      "step": 1417
+    },
+    {
+      "epoch": 0.11344,
+      "grad_norm": 1.7476698160171509,
+      "learning_rate": 9.689703989385362e-06,
+      "loss": 0.3595,
+      "step": 1418
+    },
+    {
+      "epoch": 0.11352,
+      "grad_norm": 1.494022011756897,
+      "learning_rate": 9.689267696152226e-06,
+      "loss": 0.2728,
+      "step": 1419
+    },
+    {
+      "epoch": 0.1136,
+      "grad_norm": 1.702022671699524,
+      "learning_rate": 9.688831106244814e-06,
+      "loss": 0.3349,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11368,
+      "grad_norm": 1.6980208158493042,
+      "learning_rate": 9.688394219690745e-06,
+      "loss": 0.3074,
+      "step": 1421
+    },
+    {
+      "epoch": 0.11376,
+      "grad_norm": 1.3088799715042114,
+      "learning_rate": 9.687957036517662e-06,
+      "loss": 0.2807,
+      "step": 1422
+    },
+    {
+      "epoch": 0.11384,
+      "grad_norm": 1.4283772706985474,
+      "learning_rate": 9.687519556753225e-06,
+      "loss": 0.3256,
+      "step": 1423
+    },
+    {
+      "epoch": 0.11392,
+      "grad_norm": 1.6759427785873413,
+      "learning_rate": 9.687081780425108e-06,
+      "loss": 0.3734,
+      "step": 1424
+    },
+    {
+      "epoch": 0.114,
+      "grad_norm": 1.7384825944900513,
+      "learning_rate": 9.68664370756101e-06,
+      "loss": 0.396,
+      "step": 1425
+    },
+    {
+      "epoch": 0.11408,
+      "grad_norm": 1.8031843900680542,
+      "learning_rate": 9.686205338188645e-06,
+      "loss": 0.4733,
+      "step": 1426
+    },
+    {
+      "epoch": 0.11416,
+      "grad_norm": 1.8591822385787964,
+      "learning_rate": 9.68576667233575e-06,
+      "loss": 0.4139,
+      "step": 1427
+    },
+    {
+      "epoch": 0.11424,
+      "grad_norm": 1.8857415914535522,
+      "learning_rate": 9.685327710030077e-06,
+      "loss": 0.3974,
+      "step": 1428
+    },
+    {
+      "epoch": 0.11432,
+      "grad_norm": 1.1718003749847412,
+      "learning_rate": 9.684888451299396e-06,
+      "loss": 0.2668,
+      "step": 1429
+    },
+    {
+      "epoch": 0.1144,
+      "grad_norm": 1.7708724737167358,
+      "learning_rate": 9.684448896171498e-06,
+      "loss": 0.3957,
+      "step": 1430
+    },
+    {
+      "epoch": 0.11448,
+      "grad_norm": 1.4460155963897705,
+      "learning_rate": 9.684009044674193e-06,
+      "loss": 0.3137,
+      "step": 1431
+    },
+    {
+      "epoch": 0.11456,
+      "grad_norm": 1.4484257698059082,
+      "learning_rate": 9.683568896835309e-06,
+      "loss": 0.33,
+      "step": 1432
+    },
+    {
+      "epoch": 0.11464,
+      "grad_norm": 2.073709487915039,
+      "learning_rate": 9.683128452682692e-06,
+      "loss": 0.4239,
+      "step": 1433
+    },
+    {
+      "epoch": 0.11472,
+      "grad_norm": 1.4383269548416138,
+      "learning_rate": 9.682687712244205e-06,
+      "loss": 0.269,
+      "step": 1434
+    },
+    {
+      "epoch": 0.1148,
+      "grad_norm": 1.5897904634475708,
+      "learning_rate": 9.682246675547737e-06,
+      "loss": 0.3171,
+      "step": 1435
+    },
+    {
+      "epoch": 0.11488,
+      "grad_norm": 1.4807348251342773,
+      "learning_rate": 9.68180534262119e-06,
+      "loss": 0.2884,
+      "step": 1436
+    },
+    {
+      "epoch": 0.11496,
+      "grad_norm": 1.8860735893249512,
+      "learning_rate": 9.681363713492483e-06,
+      "loss": 0.449,
+      "step": 1437
+    },
+    {
+      "epoch": 0.11504,
+      "grad_norm": 1.6362406015396118,
+      "learning_rate": 9.680921788189556e-06,
+      "loss": 0.3038,
+      "step": 1438
+    },
+    {
+      "epoch": 0.11512,
+      "grad_norm": 1.7273122072219849,
+      "learning_rate": 9.680479566740373e-06,
+      "loss": 0.3497,
+      "step": 1439
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 1.6315315961837769,
+      "learning_rate": 9.680037049172907e-06,
+      "loss": 0.3365,
+      "step": 1440
+    },
+    {
+      "epoch": 0.11528,
+      "grad_norm": 1.381567358970642,
+      "learning_rate": 9.679594235515158e-06,
+      "loss": 0.2703,
+      "step": 1441
+    },
+    {
+      "epoch": 0.11536,
+      "grad_norm": 2.0206964015960693,
+      "learning_rate": 9.679151125795136e-06,
+      "loss": 0.4859,
+      "step": 1442
+    },
+    {
+      "epoch": 0.11544,
+      "grad_norm": 1.7309236526489258,
+      "learning_rate": 9.678707720040882e-06,
+      "loss": 0.4465,
+      "step": 1443
+    },
+    {
+      "epoch": 0.11552,
+      "grad_norm": 1.8960282802581787,
+      "learning_rate": 9.678264018280445e-06,
+      "loss": 0.4757,
+      "step": 1444
+    },
+    {
+      "epoch": 0.1156,
+      "grad_norm": 1.276365041732788,
+      "learning_rate": 9.677820020541898e-06,
+      "loss": 0.2635,
+      "step": 1445
+    },
+    {
+      "epoch": 0.11568,
+      "grad_norm": 1.4210036993026733,
+      "learning_rate": 9.677375726853327e-06,
+      "loss": 0.3031,
+      "step": 1446
+    },
+    {
+      "epoch": 0.11576,
+      "grad_norm": 1.6796081066131592,
+      "learning_rate": 9.676931137242846e-06,
+      "loss": 0.4549,
+      "step": 1447
+    },
+    {
+      "epoch": 0.11584,
+      "grad_norm": 1.4270731210708618,
+      "learning_rate": 9.676486251738581e-06,
+      "loss": 0.3278,
+      "step": 1448
+    },
+    {
+      "epoch": 0.11592,
+      "grad_norm": 1.5556656122207642,
+      "learning_rate": 9.67604107036868e-06,
+      "loss": 0.2951,
+      "step": 1449
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 1.3497978448867798,
+      "learning_rate": 9.675595593161305e-06,
+      "loss": 0.357,
+      "step": 1450
+    },
+    {
+      "epoch": 0.11608,
+      "grad_norm": 1.3413703441619873,
+      "learning_rate": 9.675149820144643e-06,
+      "loss": 0.3007,
+      "step": 1451
+    },
+    {
+      "epoch": 0.11616,
+      "grad_norm": 1.4101148843765259,
+      "learning_rate": 9.674703751346893e-06,
+      "loss": 0.3115,
+      "step": 1452
+    },
+    {
+      "epoch": 0.11624,
+      "grad_norm": 1.480543851852417,
+      "learning_rate": 9.67425738679628e-06,
+      "loss": 0.3169,
+      "step": 1453
+    },
+    {
+      "epoch": 0.11632,
+      "grad_norm": 2.082014799118042,
+      "learning_rate": 9.67381072652104e-06,
+      "loss": 0.4067,
+      "step": 1454
+    },
+    {
+      "epoch": 0.1164,
+      "grad_norm": 1.5257582664489746,
+      "learning_rate": 9.673363770549435e-06,
+      "loss": 0.3861,
+      "step": 1455
+    },
+    {
+      "epoch": 0.11648,
+      "grad_norm": 1.7837907075881958,
+      "learning_rate": 9.672916518909743e-06,
+      "loss": 0.4652,
+      "step": 1456
+    },
+    {
+      "epoch": 0.11656,
+      "grad_norm": 1.7206987142562866,
+      "learning_rate": 9.672468971630256e-06,
+      "loss": 0.4114,
+      "step": 1457
+    },
+    {
+      "epoch": 0.11664,
+      "grad_norm": 1.51250159740448,
+      "learning_rate": 9.672021128739293e-06,
+      "loss": 0.4297,
+      "step": 1458
+    },
+    {
+      "epoch": 0.11672,
+      "grad_norm": 1.856751799583435,
+      "learning_rate": 9.671572990265186e-06,
+      "loss": 0.4375,
+      "step": 1459
+    },
+    {
+      "epoch": 0.1168,
+      "grad_norm": 2.219356060028076,
+      "learning_rate": 9.671124556236284e-06,
+      "loss": 0.5495,
+      "step": 1460
+    },
+    {
+      "epoch": 0.11688,
+      "grad_norm": 1.8340235948562622,
+      "learning_rate": 9.670675826680963e-06,
+      "loss": 0.361,
+      "step": 1461
+    },
+    {
+      "epoch": 0.11696,
+      "grad_norm": 1.3916457891464233,
+      "learning_rate": 9.670226801627611e-06,
+      "loss": 0.3085,
+      "step": 1462
+    },
+    {
+      "epoch": 0.11704,
+      "grad_norm": 1.3472083806991577,
+      "learning_rate": 9.669777481104637e-06,
+      "loss": 0.3098,
+      "step": 1463
+    },
+    {
+      "epoch": 0.11712,
+      "grad_norm": 1.739112138748169,
+      "learning_rate": 9.669327865140465e-06,
+      "loss": 0.3132,
+      "step": 1464
+    },
+    {
+      "epoch": 0.1172,
+      "grad_norm": 1.6176772117614746,
+      "learning_rate": 9.668877953763544e-06,
+      "loss": 0.4131,
+      "step": 1465
+    },
+    {
+      "epoch": 0.11728,
+      "grad_norm": 1.2607818841934204,
+      "learning_rate": 9.668427747002337e-06,
+      "loss": 0.3161,
+      "step": 1466
+    },
+    {
+      "epoch": 0.11736,
+      "grad_norm": 1.6247892379760742,
+      "learning_rate": 9.667977244885324e-06,
+      "loss": 0.3631,
+      "step": 1467
+    },
+    {
+      "epoch": 0.11744,
+      "grad_norm": 1.8417178392410278,
+      "learning_rate": 9.667526447441012e-06,
+      "loss": 0.3967,
+      "step": 1468
+    },
+    {
+      "epoch": 0.11752,
+      "grad_norm": 1.448082447052002,
+      "learning_rate": 9.667075354697919e-06,
+      "loss": 0.2738,
+      "step": 1469
+    },
+    {
+      "epoch": 0.1176,
+      "grad_norm": 1.4578620195388794,
+      "learning_rate": 9.666623966684585e-06,
+      "loss": 0.437,
+      "step": 1470
+    },
+    {
+      "epoch": 0.11768,
+      "grad_norm": 1.3342931270599365,
+      "learning_rate": 9.666172283429566e-06,
+      "loss": 0.3525,
+      "step": 1471
+    },
+    {
+      "epoch": 0.11776,
+      "grad_norm": 1.410962700843811,
+      "learning_rate": 9.66572030496144e-06,
+      "loss": 0.3757,
+      "step": 1472
+    },
+    {
+      "epoch": 0.11784,
+      "grad_norm": 1.3203762769699097,
+      "learning_rate": 9.665268031308804e-06,
+      "loss": 0.2784,
+      "step": 1473
+    },
+    {
+      "epoch": 0.11792,
+      "grad_norm": 1.5164921283721924,
+      "learning_rate": 9.664815462500268e-06,
+      "loss": 0.3143,
+      "step": 1474
+    },
+    {
+      "epoch": 0.118,
+      "grad_norm": 1.555022120475769,
+      "learning_rate": 9.664362598564466e-06,
+      "loss": 0.3576,
+      "step": 1475
+    },
+    {
+      "epoch": 0.11808,
+      "grad_norm": 1.5586203336715698,
+      "learning_rate": 9.66390943953005e-06,
+      "loss": 0.2553,
+      "step": 1476
+    },
+    {
+      "epoch": 0.11816,
+      "grad_norm": 1.8795398473739624,
+      "learning_rate": 9.66345598542569e-06,
+      "loss": 0.3442,
+      "step": 1477
+    },
+    {
+      "epoch": 0.11824,
+      "grad_norm": 1.5754092931747437,
+      "learning_rate": 9.663002236280072e-06,
+      "loss": 0.2897,
+      "step": 1478
+    },
+    {
+      "epoch": 0.11832,
+      "grad_norm": 1.3736915588378906,
+      "learning_rate": 9.662548192121905e-06,
+      "loss": 0.3255,
+      "step": 1479
+    },
+    {
+      "epoch": 0.1184,
+      "grad_norm": 1.4949525594711304,
+      "learning_rate": 9.662093852979916e-06,
+      "loss": 0.2846,
+      "step": 1480
+    },
+    {
+      "epoch": 0.11848,
+      "grad_norm": 1.5100445747375488,
+      "learning_rate": 9.661639218882849e-06,
+      "loss": 0.3407,
+      "step": 1481
+    },
+    {
+      "epoch": 0.11856,
+      "grad_norm": 1.6961963176727295,
+      "learning_rate": 9.661184289859465e-06,
+      "loss": 0.3679,
+      "step": 1482
+    },
+    {
+      "epoch": 0.11864,
+      "grad_norm": 1.5878902673721313,
+      "learning_rate": 9.660729065938547e-06,
+      "loss": 0.2542,
+      "step": 1483
+    },
+    {
+      "epoch": 0.11872,
+      "grad_norm": 1.5185855627059937,
+      "learning_rate": 9.660273547148897e-06,
+      "loss": 0.3301,
+      "step": 1484
+    },
+    {
+      "epoch": 0.1188,
+      "grad_norm": 1.7931679487228394,
+      "learning_rate": 9.659817733519333e-06,
+      "loss": 0.3853,
+      "step": 1485
+    },
+    {
+      "epoch": 0.11888,
+      "grad_norm": 1.5024420022964478,
+      "learning_rate": 9.659361625078691e-06,
+      "loss": 0.2664,
+      "step": 1486
+    },
+    {
+      "epoch": 0.11896,
+      "grad_norm": 1.6533669233322144,
+      "learning_rate": 9.65890522185583e-06,
+      "loss": 0.3301,
+      "step": 1487
+    },
+    {
+      "epoch": 0.11904,
+      "grad_norm": 2.171912908554077,
+      "learning_rate": 9.658448523879626e-06,
+      "loss": 0.396,
+      "step": 1488
+    },
+    {
+      "epoch": 0.11912,
+      "grad_norm": 1.7563323974609375,
+      "learning_rate": 9.65799153117897e-06,
+      "loss": 0.4444,
+      "step": 1489
+    },
+    {
+      "epoch": 0.1192,
+      "grad_norm": 1.1641733646392822,
+      "learning_rate": 9.657534243782775e-06,
+      "loss": 0.2883,
+      "step": 1490
+    },
+    {
+      "epoch": 0.11928,
+      "grad_norm": 1.5933700799942017,
+      "learning_rate": 9.657076661719972e-06,
+      "loss": 0.336,
+      "step": 1491
+    },
+    {
+      "epoch": 0.11936,
+      "grad_norm": 1.411896824836731,
+      "learning_rate": 9.656618785019513e-06,
+      "loss": 0.2531,
+      "step": 1492
+    },
+    {
+      "epoch": 0.11944,
+      "grad_norm": 1.3056985139846802,
+      "learning_rate": 9.656160613710364e-06,
+      "loss": 0.2732,
+      "step": 1493
+    },
+    {
+      "epoch": 0.11952,
+      "grad_norm": 1.6541974544525146,
+      "learning_rate": 9.655702147821514e-06,
+      "loss": 0.3592,
+      "step": 1494
+    },
+    {
+      "epoch": 0.1196,
+      "grad_norm": 1.6240873336791992,
+      "learning_rate": 9.655243387381965e-06,
+      "loss": 0.3601,
+      "step": 1495
+    },
+    {
+      "epoch": 0.11968,
+      "grad_norm": 1.4387151002883911,
+      "learning_rate": 9.654784332420744e-06,
+      "loss": 0.274,
+      "step": 1496
+    },
+    {
+      "epoch": 0.11976,
+      "grad_norm": 1.491258144378662,
+      "learning_rate": 9.654324982966891e-06,
+      "loss": 0.3344,
+      "step": 1497
+    },
+    {
+      "epoch": 0.11984,
+      "grad_norm": 1.6981704235076904,
+      "learning_rate": 9.653865339049472e-06,
+      "loss": 0.5099,
+      "step": 1498
+    },
+    {
+      "epoch": 0.11992,
+      "grad_norm": 1.6840261220932007,
+      "learning_rate": 9.653405400697567e-06,
+      "loss": 0.3375,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4480576515197754,
+      "learning_rate": 9.65294516794027e-06,
+      "loss": 0.3806,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12008,
+      "grad_norm": 1.518832802772522,
+      "learning_rate": 9.6524846408067e-06,
+      "loss": 0.3553,
+      "step": 1501
+    },
+    {
+      "epoch": 0.12016,
+      "grad_norm": 1.694017767906189,
+      "learning_rate": 9.652023819325998e-06,
+      "loss": 0.3129,
+      "step": 1502
+    },
+    {
+      "epoch": 0.12024,
+      "grad_norm": 1.3871276378631592,
+      "learning_rate": 9.651562703527311e-06,
+      "loss": 0.3113,
+      "step": 1503
+    },
+    {
+      "epoch": 0.12032,
+      "grad_norm": 1.6968255043029785,
+      "learning_rate": 9.651101293439817e-06,
+      "loss": 0.3511,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1204,
+      "grad_norm": 1.6635756492614746,
+      "learning_rate": 9.650639589092708e-06,
+      "loss": 0.354,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12048,
+      "grad_norm": 1.5392272472381592,
+      "learning_rate": 9.65017759051519e-06,
+      "loss": 0.3186,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12056,
+      "grad_norm": 1.2892400026321411,
+      "learning_rate": 9.649715297736499e-06,
+      "loss": 0.2914,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12064,
+      "grad_norm": 1.6253796815872192,
+      "learning_rate": 9.649252710785876e-06,
+      "loss": 0.3478,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12072,
+      "grad_norm": 1.8582258224487305,
+      "learning_rate": 9.648789829692594e-06,
+      "loss": 0.4373,
+      "step": 1509
+    },
+    {
+      "epoch": 0.1208,
+      "grad_norm": 1.683634877204895,
+      "learning_rate": 9.648326654485931e-06,
+      "loss": 0.3919,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12088,
+      "grad_norm": 2.1171162128448486,
+      "learning_rate": 9.647863185195197e-06,
+      "loss": 0.4234,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12096,
+      "grad_norm": 1.5578632354736328,
+      "learning_rate": 9.647399421849708e-06,
+      "loss": 0.3344,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12104,
+      "grad_norm": 1.7130169868469238,
+      "learning_rate": 9.64693536447881e-06,
+      "loss": 0.4795,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12112,
+      "grad_norm": 1.5698732137680054,
+      "learning_rate": 9.64647101311186e-06,
+      "loss": 0.3961,
+      "step": 1514
+    },
+    {
+      "epoch": 0.1212,
+      "grad_norm": 1.7023333311080933,
+      "learning_rate": 9.646006367778235e-06,
+      "loss": 0.3476,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12128,
+      "grad_norm": 1.8466098308563232,
+      "learning_rate": 9.645541428507334e-06,
+      "loss": 0.4728,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12136,
+      "grad_norm": 1.5508915185928345,
+      "learning_rate": 9.64507619532857e-06,
+      "loss": 0.4352,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12144,
+      "grad_norm": 1.3814231157302856,
+      "learning_rate": 9.644610668271377e-06,
+      "loss": 0.3032,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12152,
+      "grad_norm": 1.73944890499115,
+      "learning_rate": 9.64414484736521e-06,
+      "loss": 0.3277,
+      "step": 1519
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 1.2703596353530884,
+      "learning_rate": 9.643678732639537e-06,
+      "loss": 0.2786,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12168,
+      "grad_norm": 1.4023946523666382,
+      "learning_rate": 9.643212324123848e-06,
+      "loss": 0.2626,
+      "step": 1521
+    },
+    {
+      "epoch": 0.12176,
+      "grad_norm": 1.5729879140853882,
+      "learning_rate": 9.64274562184765e-06,
+      "loss": 0.3552,
+      "step": 1522
+    },
+    {
+      "epoch": 0.12184,
+      "grad_norm": 1.6716023683547974,
+      "learning_rate": 9.642278625840473e-06,
+      "loss": 0.3472,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12192,
+      "grad_norm": 1.9659554958343506,
+      "learning_rate": 9.64181133613186e-06,
+      "loss": 0.3494,
+      "step": 1524
+    },
+    {
+      "epoch": 0.122,
+      "grad_norm": 1.7329281568527222,
+      "learning_rate": 9.641343752751375e-06,
+      "loss": 0.3407,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12208,
+      "grad_norm": 1.9919676780700684,
+      "learning_rate": 9.640875875728602e-06,
+      "loss": 0.461,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12216,
+      "grad_norm": 1.6600706577301025,
+      "learning_rate": 9.64040770509314e-06,
+      "loss": 0.3863,
+      "step": 1527
+    },
+    {
+      "epoch": 0.12224,
+      "grad_norm": 1.8697322607040405,
+      "learning_rate": 9.639939240874609e-06,
+      "loss": 0.4213,
+      "step": 1528
+    },
+    {
+      "epoch": 0.12232,
+      "grad_norm": 1.6128782033920288,
+      "learning_rate": 9.639470483102647e-06,
+      "loss": 0.2803,
+      "step": 1529
+    },
+    {
+      "epoch": 0.1224,
+      "grad_norm": 1.698266625404358,
+      "learning_rate": 9.639001431806912e-06,
+      "loss": 0.3881,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12248,
+      "grad_norm": 1.2810356616973877,
+      "learning_rate": 9.638532087017079e-06,
+      "loss": 0.2839,
+      "step": 1531
+    },
+    {
+      "epoch": 0.12256,
+      "grad_norm": 1.382637619972229,
+      "learning_rate": 9.638062448762842e-06,
+      "loss": 0.2794,
+      "step": 1532
+    },
+    {
+      "epoch": 0.12264,
+      "grad_norm": 1.7451645135879517,
+      "learning_rate": 9.637592517073911e-06,
+      "loss": 0.3944,
+      "step": 1533
+    },
+    {
+      "epoch": 0.12272,
+      "grad_norm": 1.6562174558639526,
+      "learning_rate": 9.63712229198002e-06,
+      "loss": 0.3808,
+      "step": 1534
+    },
+    {
+      "epoch": 0.1228,
+      "grad_norm": 1.47275972366333,
+      "learning_rate": 9.636651773510917e-06,
+      "loss": 0.3047,
+      "step": 1535
+    },
+    {
+      "epoch": 0.12288,
+      "grad_norm": 1.6594504117965698,
+      "learning_rate": 9.636180961696371e-06,
+      "loss": 0.3467,
+      "step": 1536
+    },
+    {
+      "epoch": 0.12296,
+      "grad_norm": 1.4340310096740723,
+      "learning_rate": 9.635709856566167e-06,
+      "loss": 0.2981,
+      "step": 1537
+    },
+    {
+      "epoch": 0.12304,
+      "grad_norm": 1.524653673171997,
+      "learning_rate": 9.635238458150114e-06,
+      "loss": 0.2904,
+      "step": 1538
+    },
+    {
+      "epoch": 0.12312,
+      "grad_norm": 1.5759027004241943,
+      "learning_rate": 9.634766766478032e-06,
+      "loss": 0.3737,
+      "step": 1539
+    },
+    {
+      "epoch": 0.1232,
+      "grad_norm": 1.283862590789795,
+      "learning_rate": 9.634294781579764e-06,
+      "loss": 0.2563,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12328,
+      "grad_norm": 1.5947818756103516,
+      "learning_rate": 9.633822503485172e-06,
+      "loss": 0.3229,
+      "step": 1541
+    },
+    {
+      "epoch": 0.12336,
+      "grad_norm": 1.2302602529525757,
+      "learning_rate": 9.633349932224135e-06,
+      "loss": 0.2566,
+      "step": 1542
+    },
+    {
+      "epoch": 0.12344,
+      "grad_norm": 1.265610694885254,
+      "learning_rate": 9.632877067826552e-06,
+      "loss": 0.25,
+      "step": 1543
+    },
+    {
+      "epoch": 0.12352,
+      "grad_norm": 1.6030651330947876,
+      "learning_rate": 9.632403910322337e-06,
+      "loss": 0.3573,
+      "step": 1544
+    },
+    {
+      "epoch": 0.1236,
+      "grad_norm": 1.4760074615478516,
+      "learning_rate": 9.631930459741427e-06,
+      "loss": 0.3391,
+      "step": 1545
+    },
+    {
+      "epoch": 0.12368,
+      "grad_norm": 1.7136142253875732,
+      "learning_rate": 9.631456716113777e-06,
+      "loss": 0.3732,
+      "step": 1546
+    },
+    {
+      "epoch": 0.12376,
+      "grad_norm": 1.8622156381607056,
+      "learning_rate": 9.630982679469355e-06,
+      "loss": 0.3882,
+      "step": 1547
+    },
+    {
+      "epoch": 0.12384,
+      "grad_norm": 1.7958600521087646,
+      "learning_rate": 9.630508349838155e-06,
+      "loss": 0.3191,
+      "step": 1548
+    },
+    {
+      "epoch": 0.12392,
+      "grad_norm": 1.9841034412384033,
+      "learning_rate": 9.630033727250186e-06,
+      "loss": 0.4149,
+      "step": 1549
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 1.6318235397338867,
+      "learning_rate": 9.629558811735475e-06,
+      "loss": 0.349,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12408,
+      "grad_norm": 1.3617398738861084,
+      "learning_rate": 9.62908360332407e-06,
+      "loss": 0.3016,
+      "step": 1551
+    },
+    {
+      "epoch": 0.12416,
+      "grad_norm": 1.6608842611312866,
+      "learning_rate": 9.628608102046032e-06,
+      "loss": 0.3923,
+      "step": 1552
+    },
+    {
+      "epoch": 0.12424,
+      "grad_norm": 1.6326426267623901,
+      "learning_rate": 9.628132307931446e-06,
+      "loss": 0.3796,
+      "step": 1553
+    },
+    {
+      "epoch": 0.12432,
+      "grad_norm": 1.473404049873352,
+      "learning_rate": 9.627656221010417e-06,
+      "loss": 0.3165,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1244,
+      "grad_norm": 1.6658129692077637,
+      "learning_rate": 9.627179841313063e-06,
+      "loss": 0.3675,
+      "step": 1555
+    },
+    {
+      "epoch": 0.12448,
+      "grad_norm": 1.455764889717102,
+      "learning_rate": 9.626703168869522e-06,
+      "loss": 0.3633,
+      "step": 1556
+    },
+    {
+      "epoch": 0.12456,
+      "grad_norm": 1.4597536325454712,
+      "learning_rate": 9.626226203709954e-06,
+      "loss": 0.3412,
+      "step": 1557
+    },
+    {
+      "epoch": 0.12464,
+      "grad_norm": 1.6526191234588623,
+      "learning_rate": 9.625748945864531e-06,
+      "loss": 0.4253,
+      "step": 1558
+    },
+    {
+      "epoch": 0.12472,
+      "grad_norm": 1.3261500597000122,
+      "learning_rate": 9.625271395363453e-06,
+      "loss": 0.2575,
+      "step": 1559
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 1.5173916816711426,
+      "learning_rate": 9.624793552236927e-06,
+      "loss": 0.3687,
+      "step": 1560
+    },
+    {
+      "epoch": 0.12488,
+      "grad_norm": 1.2248857021331787,
+      "learning_rate": 9.62431541651519e-06,
+      "loss": 0.2559,
+      "step": 1561
+    },
+    {
+      "epoch": 0.12496,
+      "grad_norm": 1.925079584121704,
+      "learning_rate": 9.623836988228487e-06,
+      "loss": 0.365,
+      "step": 1562
+    },
+    {
+      "epoch": 0.12504,
+      "grad_norm": 1.8328197002410889,
+      "learning_rate": 9.623358267407092e-06,
+      "loss": 0.3888,
+      "step": 1563
+    },
+    {
+      "epoch": 0.12512,
+      "grad_norm": 1.723613977432251,
+      "learning_rate": 9.622879254081288e-06,
+      "loss": 0.4945,
+      "step": 1564
+    },
+    {
+      "epoch": 0.1252,
+      "grad_norm": 1.787847638130188,
+      "learning_rate": 9.622399948281382e-06,
+      "loss": 0.424,
+      "step": 1565
+    },
+    {
+      "epoch": 0.12528,
+      "grad_norm": 1.1664930582046509,
+      "learning_rate": 9.621920350037697e-06,
+      "loss": 0.2628,
+      "step": 1566
+    },
+    {
+      "epoch": 0.12536,
+      "grad_norm": 1.3959739208221436,
+      "learning_rate": 9.621440459380577e-06,
+      "loss": 0.2896,
+      "step": 1567
+    },
+    {
+      "epoch": 0.12544,
+      "grad_norm": 1.3110196590423584,
+      "learning_rate": 9.620960276340383e-06,
+      "loss": 0.2658,
+      "step": 1568
+    },
+    {
+      "epoch": 0.12552,
+      "grad_norm": 1.2237164974212646,
+      "learning_rate": 9.620479800947494e-06,
+      "loss": 0.2624,
+      "step": 1569
+    },
+    {
+      "epoch": 0.1256,
+      "grad_norm": 1.7535439729690552,
+      "learning_rate": 9.619999033232308e-06,
+      "loss": 0.3613,
+      "step": 1570
+    },
+    {
+      "epoch": 0.12568,
+      "grad_norm": 1.5596789121627808,
+      "learning_rate": 9.61951797322524e-06,
+      "loss": 0.3552,
+      "step": 1571
+    },
+    {
+      "epoch": 0.12576,
+      "grad_norm": 2.2569661140441895,
+      "learning_rate": 9.61903662095673e-06,
+      "loss": 0.4619,
+      "step": 1572
+    },
+    {
+      "epoch": 0.12584,
+      "grad_norm": 1.540496826171875,
+      "learning_rate": 9.618554976457226e-06,
+      "loss": 0.2943,
+      "step": 1573
+    },
+    {
+      "epoch": 0.12592,
+      "grad_norm": 1.417845606803894,
+      "learning_rate": 9.618073039757204e-06,
+      "loss": 0.2505,
+      "step": 1574
+    },
+    {
+      "epoch": 0.126,
+      "grad_norm": 1.3284822702407837,
+      "learning_rate": 9.617590810887151e-06,
+      "loss": 0.2476,
+      "step": 1575
+    },
+    {
+      "epoch": 0.12608,
+      "grad_norm": 1.5151894092559814,
+      "learning_rate": 9.617108289877578e-06,
+      "loss": 0.3971,
+      "step": 1576
+    },
+    {
+      "epoch": 0.12616,
+      "grad_norm": 1.6880301237106323,
+      "learning_rate": 9.616625476759014e-06,
+      "loss": 0.3476,
+      "step": 1577
+    },
+    {
+      "epoch": 0.12624,
+      "grad_norm": 1.7845243215560913,
+      "learning_rate": 9.616142371562003e-06,
+      "loss": 0.4271,
+      "step": 1578
+    },
+    {
+      "epoch": 0.12632,
+      "grad_norm": 1.6483310461044312,
+      "learning_rate": 9.61565897431711e-06,
+      "loss": 0.3685,
+      "step": 1579
+    },
+    {
+      "epoch": 0.1264,
+      "grad_norm": 1.4874074459075928,
+      "learning_rate": 9.615175285054916e-06,
+      "loss": 0.3274,
+      "step": 1580
+    },
+    {
+      "epoch": 0.12648,
+      "grad_norm": 1.4330966472625732,
+      "learning_rate": 9.614691303806027e-06,
+      "loss": 0.3611,
+      "step": 1581
+    },
+    {
+      "epoch": 0.12656,
+      "grad_norm": 1.424462914466858,
+      "learning_rate": 9.614207030601057e-06,
+      "loss": 0.2903,
+      "step": 1582
+    },
+    {
+      "epoch": 0.12664,
+      "grad_norm": 1.2430349588394165,
+      "learning_rate": 9.61372246547065e-06,
+      "loss": 0.2505,
+      "step": 1583
+    },
+    {
+      "epoch": 0.12672,
+      "grad_norm": 1.5988134145736694,
+      "learning_rate": 9.613237608445458e-06,
+      "loss": 0.3123,
+      "step": 1584
+    },
+    {
+      "epoch": 0.1268,
+      "grad_norm": 1.6877517700195312,
+      "learning_rate": 9.612752459556161e-06,
+      "loss": 0.4546,
+      "step": 1585
+    },
+    {
+      "epoch": 0.12688,
+      "grad_norm": 1.350681185722351,
+      "learning_rate": 9.612267018833448e-06,
+      "loss": 0.3087,
+      "step": 1586
+    },
+    {
+      "epoch": 0.12696,
+      "grad_norm": 1.260632038116455,
+      "learning_rate": 9.611781286308032e-06,
+      "loss": 0.2645,
+      "step": 1587
+    },
+    {
+      "epoch": 0.12704,
+      "grad_norm": 1.2421963214874268,
+      "learning_rate": 9.611295262010649e-06,
+      "loss": 0.296,
+      "step": 1588
+    },
+    {
+      "epoch": 0.12712,
+      "grad_norm": 1.5727829933166504,
+      "learning_rate": 9.610808945972042e-06,
+      "loss": 0.3084,
+      "step": 1589
+    },
+    {
+      "epoch": 0.1272,
+      "grad_norm": 1.3248558044433594,
+      "learning_rate": 9.610322338222982e-06,
+      "loss": 0.3195,
+      "step": 1590
+    },
+    {
+      "epoch": 0.12728,
+      "grad_norm": 1.5574604272842407,
+      "learning_rate": 9.60983543879425e-06,
+      "loss": 0.3364,
+      "step": 1591
+    },
+    {
+      "epoch": 0.12736,
+      "grad_norm": 1.849393367767334,
+      "learning_rate": 9.609348247716658e-06,
+      "loss": 0.3784,
+      "step": 1592
+    },
+    {
+      "epoch": 0.12744,
+      "grad_norm": 1.5725646018981934,
+      "learning_rate": 9.608860765021025e-06,
+      "loss": 0.3448,
+      "step": 1593
+    },
+    {
+      "epoch": 0.12752,
+      "grad_norm": 1.3313325643539429,
+      "learning_rate": 9.608372990738193e-06,
+      "loss": 0.2394,
+      "step": 1594
+    },
+    {
+      "epoch": 0.1276,
+      "grad_norm": 1.4957760572433472,
+      "learning_rate": 9.60788492489902e-06,
+      "loss": 0.3165,
+      "step": 1595
+    },
+    {
+      "epoch": 0.12768,
+      "grad_norm": 1.8721736669540405,
+      "learning_rate": 9.607396567534387e-06,
+      "loss": 0.3877,
+      "step": 1596
+    },
+    {
+      "epoch": 0.12776,
+      "grad_norm": 1.5303232669830322,
+      "learning_rate": 9.606907918675189e-06,
+      "loss": 0.3717,
+      "step": 1597
+    },
+    {
+      "epoch": 0.12784,
+      "grad_norm": 1.6356817483901978,
+      "learning_rate": 9.606418978352342e-06,
+      "loss": 0.4317,
+      "step": 1598
+    },
+    {
+      "epoch": 0.12792,
+      "grad_norm": 1.7951912879943848,
+      "learning_rate": 9.60592974659678e-06,
+      "loss": 0.3572,
+      "step": 1599
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 1.7281938791275024,
+      "learning_rate": 9.605440223439452e-06,
+      "loss": 0.3721,
+      "step": 1600
+    },
+    {
+      "epoch": 0.12808,
+      "grad_norm": 1.4484968185424805,
+      "learning_rate": 9.604950408911334e-06,
+      "loss": 0.2891,
+      "step": 1601
+    },
+    {
+      "epoch": 0.12816,
+      "grad_norm": 1.6097825765609741,
+      "learning_rate": 9.604460303043411e-06,
+      "loss": 0.3225,
+      "step": 1602
+    },
+    {
+      "epoch": 0.12824,
+      "grad_norm": 1.80912184715271,
+      "learning_rate": 9.60396990586669e-06,
+      "loss": 0.4023,
+      "step": 1603
+    },
+    {
+      "epoch": 0.12832,
+      "grad_norm": 1.5157806873321533,
+      "learning_rate": 9.6034792174122e-06,
+      "loss": 0.3546,
+      "step": 1604
+    },
+    {
+      "epoch": 0.1284,
+      "grad_norm": 1.6665889024734497,
+      "learning_rate": 9.60298823771098e-06,
+      "loss": 0.3968,
+      "step": 1605
+    },
+    {
+      "epoch": 0.12848,
+      "grad_norm": 1.4822949171066284,
+      "learning_rate": 9.602496966794098e-06,
+      "loss": 0.3122,
+      "step": 1606
+    },
+    {
+      "epoch": 0.12856,
+      "grad_norm": 1.6601344347000122,
+      "learning_rate": 9.602005404692633e-06,
+      "loss": 0.3554,
+      "step": 1607
+    },
+    {
+      "epoch": 0.12864,
+      "grad_norm": 1.8020589351654053,
+      "learning_rate": 9.601513551437685e-06,
+      "loss": 0.4218,
+      "step": 1608
+    },
+    {
+      "epoch": 0.12872,
+      "grad_norm": 1.8085490465164185,
+      "learning_rate": 9.60102140706037e-06,
+      "loss": 0.5042,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1288,
+      "grad_norm": 1.8596782684326172,
+      "learning_rate": 9.600528971591824e-06,
+      "loss": 0.4824,
+      "step": 1610
+    },
+    {
+      "epoch": 0.12888,
+      "grad_norm": 1.5251644849777222,
+      "learning_rate": 9.600036245063206e-06,
+      "loss": 0.298,
+      "step": 1611
+    },
+    {
+      "epoch": 0.12896,
+      "grad_norm": 1.4406249523162842,
+      "learning_rate": 9.599543227505685e-06,
+      "loss": 0.286,
+      "step": 1612
+    },
+    {
+      "epoch": 0.12904,
+      "grad_norm": 2.1787021160125732,
+      "learning_rate": 9.599049918950456e-06,
+      "loss": 0.4677,
+      "step": 1613
+    },
+    {
+      "epoch": 0.12912,
+      "grad_norm": 1.8443048000335693,
+      "learning_rate": 9.598556319428726e-06,
+      "loss": 0.3842,
+      "step": 1614
+    },
+    {
+      "epoch": 0.1292,
+      "grad_norm": 2.4744701385498047,
+      "learning_rate": 9.598062428971725e-06,
+      "loss": 0.5183,
+      "step": 1615
+    },
+    {
+      "epoch": 0.12928,
+      "grad_norm": 1.6642011404037476,
+      "learning_rate": 9.597568247610699e-06,
+      "loss": 0.3329,
+      "step": 1616
+    },
+    {
+      "epoch": 0.12936,
+      "grad_norm": 1.8446506261825562,
+      "learning_rate": 9.597073775376912e-06,
+      "loss": 0.4378,
+      "step": 1617
+    },
+    {
+      "epoch": 0.12944,
+      "grad_norm": 2.154025077819824,
+      "learning_rate": 9.596579012301652e-06,
+      "loss": 0.4079,
+      "step": 1618
+    },
+    {
+      "epoch": 0.12952,
+      "grad_norm": 1.8335460424423218,
+      "learning_rate": 9.596083958416216e-06,
+      "loss": 0.3403,
+      "step": 1619
+    },
+    {
+      "epoch": 0.1296,
+      "grad_norm": 1.4928038120269775,
+      "learning_rate": 9.595588613751927e-06,
+      "loss": 0.3134,
+      "step": 1620
+    },
+    {
+      "epoch": 0.12968,
+      "grad_norm": 1.566196322441101,
+      "learning_rate": 9.595092978340124e-06,
+      "loss": 0.3698,
+      "step": 1621
+    },
+    {
+      "epoch": 0.12976,
+      "grad_norm": 1.6201937198638916,
+      "learning_rate": 9.594597052212163e-06,
+      "loss": 0.3139,
+      "step": 1622
+    },
+    {
+      "epoch": 0.12984,
+      "grad_norm": 1.5580354928970337,
+      "learning_rate": 9.59410083539942e-06,
+      "loss": 0.3374,
+      "step": 1623
+    },
+    {
+      "epoch": 0.12992,
+      "grad_norm": 1.472684621810913,
+      "learning_rate": 9.593604327933288e-06,
+      "loss": 0.3058,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9004932641983032,
+      "learning_rate": 9.59310752984518e-06,
+      "loss": 0.3661,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13008,
+      "grad_norm": 1.7129523754119873,
+      "learning_rate": 9.59261044116653e-06,
+      "loss": 0.4236,
+      "step": 1626
+    },
+    {
+      "epoch": 0.13016,
+      "grad_norm": 1.9962881803512573,
+      "learning_rate": 9.592113061928783e-06,
+      "loss": 0.4657,
+      "step": 1627
+    },
+    {
+      "epoch": 0.13024,
+      "grad_norm": 1.0669331550598145,
+      "learning_rate": 9.591615392163408e-06,
+      "loss": 0.2477,
+      "step": 1628
+    },
+    {
+      "epoch": 0.13032,
+      "grad_norm": 1.2336879968643188,
+      "learning_rate": 9.59111743190189e-06,
+      "loss": 0.2468,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1304,
+      "grad_norm": 2.120313882827759,
+      "learning_rate": 9.590619181175736e-06,
+      "loss": 0.423,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13048,
+      "grad_norm": 2.0639867782592773,
+      "learning_rate": 9.590120640016463e-06,
+      "loss": 0.3888,
+      "step": 1631
+    },
+    {
+      "epoch": 0.13056,
+      "grad_norm": 1.6942673921585083,
+      "learning_rate": 9.589621808455617e-06,
+      "loss": 0.377,
+      "step": 1632
+    },
+    {
+      "epoch": 0.13064,
+      "grad_norm": 1.8577876091003418,
+      "learning_rate": 9.589122686524759e-06,
+      "loss": 0.3929,
+      "step": 1633
+    },
+    {
+      "epoch": 0.13072,
+      "grad_norm": 1.837852954864502,
+      "learning_rate": 9.588623274255461e-06,
+      "loss": 0.3284,
+      "step": 1634
+    },
+    {
+      "epoch": 0.1308,
+      "grad_norm": 1.546615481376648,
+      "learning_rate": 9.588123571679323e-06,
+      "loss": 0.3227,
+      "step": 1635
+    },
+    {
+      "epoch": 0.13088,
+      "grad_norm": 1.5524723529815674,
+      "learning_rate": 9.587623578827958e-06,
+      "loss": 0.3375,
+      "step": 1636
+    },
+    {
+      "epoch": 0.13096,
+      "grad_norm": 1.36067533493042,
+      "learning_rate": 9.587123295733e-06,
+      "loss": 0.394,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13104,
+      "grad_norm": 1.860670804977417,
+      "learning_rate": 9.5866227224261e-06,
+      "loss": 0.563,
+      "step": 1638
+    },
+    {
+      "epoch": 0.13112,
+      "grad_norm": 1.4262064695358276,
+      "learning_rate": 9.586121858938926e-06,
+      "loss": 0.3446,
+      "step": 1639
+    },
+    {
+      "epoch": 0.1312,
+      "grad_norm": 1.6558672189712524,
+      "learning_rate": 9.585620705303168e-06,
+      "loss": 0.3452,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13128,
+      "grad_norm": 1.3890858888626099,
+      "learning_rate": 9.585119261550531e-06,
+      "loss": 0.3047,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13136,
+      "grad_norm": 1.5068423748016357,
+      "learning_rate": 9.58461752771274e-06,
+      "loss": 0.3401,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13144,
+      "grad_norm": 1.6645478010177612,
+      "learning_rate": 9.584115503821538e-06,
+      "loss": 0.2867,
+      "step": 1643
+    },
+    {
+      "epoch": 0.13152,
+      "grad_norm": 1.6663644313812256,
+      "learning_rate": 9.583613189908688e-06,
+      "loss": 0.3418,
+      "step": 1644
+    },
+    {
+      "epoch": 0.1316,
+      "grad_norm": 1.356171727180481,
+      "learning_rate": 9.583110586005969e-06,
+      "loss": 0.2561,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13168,
+      "grad_norm": 2.0098698139190674,
+      "learning_rate": 9.582607692145176e-06,
+      "loss": 0.3869,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13176,
+      "grad_norm": 1.6945312023162842,
+      "learning_rate": 9.582104508358128e-06,
+      "loss": 0.3589,
+      "step": 1647
+    },
+    {
+      "epoch": 0.13184,
+      "grad_norm": 1.593902826309204,
+      "learning_rate": 9.58160103467666e-06,
+      "loss": 0.367,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13192,
+      "grad_norm": 1.8232020139694214,
+      "learning_rate": 9.581097271132626e-06,
+      "loss": 0.4105,
+      "step": 1649
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 1.5520724058151245,
+      "learning_rate": 9.580593217757893e-06,
+      "loss": 0.3554,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13208,
+      "grad_norm": 1.5328576564788818,
+      "learning_rate": 9.580088874584356e-06,
+      "loss": 0.3405,
+      "step": 1651
+    },
+    {
+      "epoch": 0.13216,
+      "grad_norm": 1.6017850637435913,
+      "learning_rate": 9.579584241643923e-06,
+      "loss": 0.3577,
+      "step": 1652
+    },
+    {
+      "epoch": 0.13224,
+      "grad_norm": 1.6488410234451294,
+      "learning_rate": 9.579079318968514e-06,
+      "loss": 0.3528,
+      "step": 1653
+    },
+    {
+      "epoch": 0.13232,
+      "grad_norm": 2.3161511421203613,
+      "learning_rate": 9.578574106590081e-06,
+      "loss": 0.5027,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1324,
+      "grad_norm": 1.536486029624939,
+      "learning_rate": 9.578068604540582e-06,
+      "loss": 0.3476,
+      "step": 1655
+    },
+    {
+      "epoch": 0.13248,
+      "grad_norm": 1.761263132095337,
+      "learning_rate": 9.577562812852004e-06,
+      "loss": 0.36,
+      "step": 1656
+    },
+    {
+      "epoch": 0.13256,
+      "grad_norm": 1.8455753326416016,
+      "learning_rate": 9.57705673155634e-06,
+      "loss": 0.4576,
+      "step": 1657
+    },
+    {
+      "epoch": 0.13264,
+      "grad_norm": 1.7750699520111084,
+      "learning_rate": 9.576550360685613e-06,
+      "loss": 0.3432,
+      "step": 1658
+    },
+    {
+      "epoch": 0.13272,
+      "grad_norm": 1.4518996477127075,
+      "learning_rate": 9.576043700271857e-06,
+      "loss": 0.3352,
+      "step": 1659
+    },
+    {
+      "epoch": 0.1328,
+      "grad_norm": 1.6674631834030151,
+      "learning_rate": 9.57553675034713e-06,
+      "loss": 0.3836,
+      "step": 1660
+    },
+    {
+      "epoch": 0.13288,
+      "grad_norm": 1.7204508781433105,
+      "learning_rate": 9.575029510943501e-06,
+      "loss": 0.3045,
+      "step": 1661
+    },
+    {
+      "epoch": 0.13296,
+      "grad_norm": 1.37985098361969,
+      "learning_rate": 9.574521982093063e-06,
+      "loss": 0.2761,
+      "step": 1662
+    },
+    {
+      "epoch": 0.13304,
+      "grad_norm": 1.4991388320922852,
+      "learning_rate": 9.574014163827926e-06,
+      "loss": 0.3735,
+      "step": 1663
+    },
+    {
+      "epoch": 0.13312,
+      "grad_norm": 1.4861139059066772,
+      "learning_rate": 9.573506056180215e-06,
+      "loss": 0.3132,
+      "step": 1664
+    },
+    {
+      "epoch": 0.1332,
+      "grad_norm": 1.6248459815979004,
+      "learning_rate": 9.572997659182081e-06,
+      "loss": 0.3233,
+      "step": 1665
+    },
+    {
+      "epoch": 0.13328,
+      "grad_norm": 1.3340996503829956,
+      "learning_rate": 9.572488972865686e-06,
+      "loss": 0.3419,
+      "step": 1666
+    },
+    {
+      "epoch": 0.13336,
+      "grad_norm": 1.760284423828125,
+      "learning_rate": 9.571979997263214e-06,
+      "loss": 0.3719,
+      "step": 1667
+    },
+    {
+      "epoch": 0.13344,
+      "grad_norm": 1.4957592487335205,
+      "learning_rate": 9.571470732406865e-06,
+      "loss": 0.3708,
+      "step": 1668
+    },
+    {
+      "epoch": 0.13352,
+      "grad_norm": 1.7483258247375488,
+      "learning_rate": 9.570961178328859e-06,
+      "loss": 0.4577,
+      "step": 1669
+    },
+    {
+      "epoch": 0.1336,
+      "grad_norm": 1.3776280879974365,
+      "learning_rate": 9.570451335061433e-06,
+      "loss": 0.288,
+      "step": 1670
+    },
+    {
+      "epoch": 0.13368,
+      "grad_norm": 1.427151083946228,
+      "learning_rate": 9.569941202636846e-06,
+      "loss": 0.3302,
+      "step": 1671
+    },
+    {
+      "epoch": 0.13376,
+      "grad_norm": 1.7294217348098755,
+      "learning_rate": 9.569430781087367e-06,
+      "loss": 0.3842,
+      "step": 1672
+    },
+    {
+      "epoch": 0.13384,
+      "grad_norm": 1.1407139301300049,
+      "learning_rate": 9.568920070445295e-06,
+      "loss": 0.2215,
+      "step": 1673
+    },
+    {
+      "epoch": 0.13392,
+      "grad_norm": 1.2754679918289185,
+      "learning_rate": 9.568409070742936e-06,
+      "loss": 0.2493,
+      "step": 1674
+    },
+    {
+      "epoch": 0.134,
+      "grad_norm": 1.5319108963012695,
+      "learning_rate": 9.56789778201262e-06,
+      "loss": 0.3972,
+      "step": 1675
+    },
+    {
+      "epoch": 0.13408,
+      "grad_norm": 1.3693424463272095,
+      "learning_rate": 9.567386204286697e-06,
+      "loss": 0.261,
+      "step": 1676
+    },
+    {
+      "epoch": 0.13416,
+      "grad_norm": 1.649755597114563,
+      "learning_rate": 9.566874337597533e-06,
+      "loss": 0.3777,
+      "step": 1677
+    },
+    {
+      "epoch": 0.13424,
+      "grad_norm": 1.7307039499282837,
+      "learning_rate": 9.566362181977509e-06,
+      "loss": 0.4339,
+      "step": 1678
+    },
+    {
+      "epoch": 0.13432,
+      "grad_norm": 1.3721426725387573,
+      "learning_rate": 9.565849737459027e-06,
+      "loss": 0.2845,
+      "step": 1679
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 2.235292434692383,
+      "learning_rate": 9.565337004074512e-06,
+      "loss": 0.4784,
+      "step": 1680
+    },
+    {
+      "epoch": 0.13448,
+      "grad_norm": 1.6313210725784302,
+      "learning_rate": 9.5648239818564e-06,
+      "loss": 0.3986,
+      "step": 1681
+    },
+    {
+      "epoch": 0.13456,
+      "grad_norm": 1.9515478610992432,
+      "learning_rate": 9.564310670837146e-06,
+      "loss": 0.4105,
+      "step": 1682
+    },
+    {
+      "epoch": 0.13464,
+      "grad_norm": 1.635183572769165,
+      "learning_rate": 9.563797071049232e-06,
+      "loss": 0.3187,
+      "step": 1683
+    },
+    {
+      "epoch": 0.13472,
+      "grad_norm": 1.5895339250564575,
+      "learning_rate": 9.563283182525145e-06,
+      "loss": 0.3844,
+      "step": 1684
+    },
+    {
+      "epoch": 0.1348,
+      "grad_norm": 1.718421459197998,
+      "learning_rate": 9.562769005297401e-06,
+      "loss": 0.3469,
+      "step": 1685
+    },
+    {
+      "epoch": 0.13488,
+      "grad_norm": 1.5541642904281616,
+      "learning_rate": 9.56225453939853e-06,
+      "loss": 0.3621,
+      "step": 1686
+    },
+    {
+      "epoch": 0.13496,
+      "grad_norm": 1.4598628282546997,
+      "learning_rate": 9.561739784861077e-06,
+      "loss": 0.3712,
+      "step": 1687
+    },
+    {
+      "epoch": 0.13504,
+      "grad_norm": 1.4597030878067017,
+      "learning_rate": 9.561224741717614e-06,
+      "loss": 0.343,
+      "step": 1688
+    },
+    {
+      "epoch": 0.13512,
+      "grad_norm": 1.3350636959075928,
+      "learning_rate": 9.560709410000722e-06,
+      "loss": 0.2609,
+      "step": 1689
+    },
+    {
+      "epoch": 0.1352,
+      "grad_norm": 1.1671631336212158,
+      "learning_rate": 9.560193789743006e-06,
+      "loss": 0.2378,
+      "step": 1690
+    },
+    {
+      "epoch": 0.13528,
+      "grad_norm": 1.94452965259552,
+      "learning_rate": 9.55967788097709e-06,
+      "loss": 0.3756,
+      "step": 1691
+    },
+    {
+      "epoch": 0.13536,
+      "grad_norm": 2.2720422744750977,
+      "learning_rate": 9.559161683735607e-06,
+      "loss": 0.4896,
+      "step": 1692
+    },
+    {
+      "epoch": 0.13544,
+      "grad_norm": 1.2150167226791382,
+      "learning_rate": 9.558645198051221e-06,
+      "loss": 0.2944,
+      "step": 1693
+    },
+    {
+      "epoch": 0.13552,
+      "grad_norm": 1.321184754371643,
+      "learning_rate": 9.558128423956608e-06,
+      "loss": 0.3323,
+      "step": 1694
+    },
+    {
+      "epoch": 0.1356,
+      "grad_norm": 1.6017346382141113,
+      "learning_rate": 9.55761136148446e-06,
+      "loss": 0.3454,
+      "step": 1695
+    },
+    {
+      "epoch": 0.13568,
+      "grad_norm": 1.4574471712112427,
+      "learning_rate": 9.55709401066749e-06,
+      "loss": 0.3351,
+      "step": 1696
+    },
+    {
+      "epoch": 0.13576,
+      "grad_norm": 1.7760628461837769,
+      "learning_rate": 9.556576371538431e-06,
+      "loss": 0.3757,
+      "step": 1697
+    },
+    {
+      "epoch": 0.13584,
+      "grad_norm": 1.3665014505386353,
+      "learning_rate": 9.556058444130032e-06,
+      "loss": 0.2966,
+      "step": 1698
+    },
+    {
+      "epoch": 0.13592,
+      "grad_norm": 1.6557552814483643,
+      "learning_rate": 9.555540228475058e-06,
+      "loss": 0.3721,
+      "step": 1699
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 1.6838151216506958,
+      "learning_rate": 9.555021724606298e-06,
+      "loss": 0.4326,
+      "step": 1700
+    },
+    {
+      "epoch": 0.13608,
+      "grad_norm": 1.3219002485275269,
+      "learning_rate": 9.554502932556555e-06,
+      "loss": 0.3573,
+      "step": 1701
+    },
+    {
+      "epoch": 0.13616,
+      "grad_norm": 1.4627076387405396,
+      "learning_rate": 9.55398385235865e-06,
+      "loss": 0.3406,
+      "step": 1702
+    },
+    {
+      "epoch": 0.13624,
+      "grad_norm": 1.6964737176895142,
+      "learning_rate": 9.553464484045425e-06,
+      "loss": 0.3365,
+      "step": 1703
+    },
+    {
+      "epoch": 0.13632,
+      "grad_norm": 1.4238336086273193,
+      "learning_rate": 9.552944827649737e-06,
+      "loss": 0.3474,
+      "step": 1704
+    },
+    {
+      "epoch": 0.1364,
+      "grad_norm": 1.3716814517974854,
+      "learning_rate": 9.552424883204465e-06,
+      "loss": 0.3811,
+      "step": 1705
+    },
+    {
+      "epoch": 0.13648,
+      "grad_norm": 1.5849741697311401,
+      "learning_rate": 9.551904650742503e-06,
+      "loss": 0.3416,
+      "step": 1706
+    },
+    {
+      "epoch": 0.13656,
+      "grad_norm": 1.7011810541152954,
+      "learning_rate": 9.551384130296763e-06,
+      "loss": 0.3839,
+      "step": 1707
+    },
+    {
+      "epoch": 0.13664,
+      "grad_norm": 1.6651289463043213,
+      "learning_rate": 9.55086332190018e-06,
+      "loss": 0.3876,
+      "step": 1708
+    },
+    {
+      "epoch": 0.13672,
+      "grad_norm": 1.391045331954956,
+      "learning_rate": 9.5503422255857e-06,
+      "loss": 0.2749,
+      "step": 1709
+    },
+    {
+      "epoch": 0.1368,
+      "grad_norm": 1.50571870803833,
+      "learning_rate": 9.549820841386295e-06,
+      "loss": 0.3082,
+      "step": 1710
+    },
+    {
+      "epoch": 0.13688,
+      "grad_norm": 1.1734155416488647,
+      "learning_rate": 9.549299169334948e-06,
+      "loss": 0.3251,
+      "step": 1711
+    },
+    {
+      "epoch": 0.13696,
+      "grad_norm": 1.4512368440628052,
+      "learning_rate": 9.548777209464664e-06,
+      "loss": 0.3287,
+      "step": 1712
+    },
+    {
+      "epoch": 0.13704,
+      "grad_norm": 1.5819345712661743,
+      "learning_rate": 9.548254961808467e-06,
+      "loss": 0.3006,
+      "step": 1713
+    },
+    {
+      "epoch": 0.13712,
+      "grad_norm": 1.6825000047683716,
+      "learning_rate": 9.547732426399397e-06,
+      "loss": 0.3654,
+      "step": 1714
+    },
+    {
+      "epoch": 0.1372,
+      "grad_norm": 2.002753496170044,
+      "learning_rate": 9.547209603270513e-06,
+      "loss": 0.4603,
+      "step": 1715
+    },
+    {
+      "epoch": 0.13728,
+      "grad_norm": 1.3878475427627563,
+      "learning_rate": 9.546686492454892e-06,
+      "loss": 0.3205,
+      "step": 1716
+    },
+    {
+      "epoch": 0.13736,
+      "grad_norm": 1.9172121286392212,
+      "learning_rate": 9.546163093985631e-06,
+      "loss": 0.4037,
+      "step": 1717
+    },
+    {
+      "epoch": 0.13744,
+      "grad_norm": 1.4683332443237305,
+      "learning_rate": 9.545639407895842e-06,
+      "loss": 0.3113,
+      "step": 1718
+    },
+    {
+      "epoch": 0.13752,
+      "grad_norm": 1.4814456701278687,
+      "learning_rate": 9.545115434218658e-06,
+      "loss": 0.3195,
+      "step": 1719
+    },
+    {
+      "epoch": 0.1376,
+      "grad_norm": 1.6814812421798706,
+      "learning_rate": 9.544591172987227e-06,
+      "loss": 0.3734,
+      "step": 1720
+    },
+    {
+      "epoch": 0.13768,
+      "grad_norm": 1.4139302968978882,
+      "learning_rate": 9.54406662423472e-06,
+      "loss": 0.3155,
+      "step": 1721
+    },
+    {
+      "epoch": 0.13776,
+      "grad_norm": 1.6127204895019531,
+      "learning_rate": 9.543541787994322e-06,
+      "loss": 0.3476,
+      "step": 1722
+    },
+    {
+      "epoch": 0.13784,
+      "grad_norm": 1.9391632080078125,
+      "learning_rate": 9.543016664299237e-06,
+      "loss": 0.4613,
+      "step": 1723
+    },
+    {
+      "epoch": 0.13792,
+      "grad_norm": 1.070672869682312,
+      "learning_rate": 9.542491253182689e-06,
+      "loss": 0.2503,
+      "step": 1724
+    },
+    {
+      "epoch": 0.138,
+      "grad_norm": 1.7871367931365967,
+      "learning_rate": 9.541965554677918e-06,
+      "loss": 0.4031,
+      "step": 1725
+    },
+    {
+      "epoch": 0.13808,
+      "grad_norm": 1.4849225282669067,
+      "learning_rate": 9.541439568818186e-06,
+      "loss": 0.3372,
+      "step": 1726
+    },
+    {
+      "epoch": 0.13816,
+      "grad_norm": 1.5674102306365967,
+      "learning_rate": 9.540913295636766e-06,
+      "loss": 0.4731,
+      "step": 1727
+    },
+    {
+      "epoch": 0.13824,
+      "grad_norm": 1.6845797300338745,
+      "learning_rate": 9.540386735166957e-06,
+      "loss": 0.4774,
+      "step": 1728
+    },
+    {
+      "epoch": 0.13832,
+      "grad_norm": 1.2045738697052002,
+      "learning_rate": 9.539859887442071e-06,
+      "loss": 0.3388,
+      "step": 1729
+    },
+    {
+      "epoch": 0.1384,
+      "grad_norm": 1.78014075756073,
+      "learning_rate": 9.53933275249544e-06,
+      "loss": 0.3886,
+      "step": 1730
+    },
+    {
+      "epoch": 0.13848,
+      "grad_norm": 1.5145901441574097,
+      "learning_rate": 9.538805330360415e-06,
+      "loss": 0.3352,
+      "step": 1731
+    },
+    {
+      "epoch": 0.13856,
+      "grad_norm": 1.681176781654358,
+      "learning_rate": 9.538277621070363e-06,
+      "loss": 0.3815,
+      "step": 1732
+    },
+    {
+      "epoch": 0.13864,
+      "grad_norm": 1.7225366830825806,
+      "learning_rate": 9.537749624658671e-06,
+      "loss": 0.3022,
+      "step": 1733
+    },
+    {
+      "epoch": 0.13872,
+      "grad_norm": 1.4362283945083618,
+      "learning_rate": 9.537221341158745e-06,
+      "loss": 0.3086,
+      "step": 1734
+    },
+    {
+      "epoch": 0.1388,
+      "grad_norm": 1.7611305713653564,
+      "learning_rate": 9.536692770604005e-06,
+      "loss": 0.3078,
+      "step": 1735
+    },
+    {
+      "epoch": 0.13888,
+      "grad_norm": 2.2746381759643555,
+      "learning_rate": 9.536163913027894e-06,
+      "loss": 0.4661,
+      "step": 1736
+    },
+    {
+      "epoch": 0.13896,
+      "grad_norm": 1.6022064685821533,
+      "learning_rate": 9.535634768463869e-06,
+      "loss": 0.3514,
+      "step": 1737
+    },
+    {
+      "epoch": 0.13904,
+      "grad_norm": 2.1435439586639404,
+      "learning_rate": 9.53510533694541e-06,
+      "loss": 0.4979,
+      "step": 1738
+    },
+    {
+      "epoch": 0.13912,
+      "grad_norm": 1.5919454097747803,
+      "learning_rate": 9.53457561850601e-06,
+      "loss": 0.387,
+      "step": 1739
+    },
+    {
+      "epoch": 0.1392,
+      "grad_norm": 1.475101351737976,
+      "learning_rate": 9.534045613179184e-06,
+      "loss": 0.3359,
+      "step": 1740
+    },
+    {
+      "epoch": 0.13928,
+      "grad_norm": 1.5171840190887451,
+      "learning_rate": 9.533515320998462e-06,
+      "loss": 0.3717,
+      "step": 1741
+    },
+    {
+      "epoch": 0.13936,
+      "grad_norm": 1.7646937370300293,
+      "learning_rate": 9.532984741997395e-06,
+      "loss": 0.3868,
+      "step": 1742
+    },
+    {
+      "epoch": 0.13944,
+      "grad_norm": 2.229182720184326,
+      "learning_rate": 9.532453876209551e-06,
+      "loss": 0.4725,
+      "step": 1743
+    },
+    {
+      "epoch": 0.13952,
+      "grad_norm": 1.8388185501098633,
+      "learning_rate": 9.531922723668517e-06,
+      "loss": 0.3709,
+      "step": 1744
+    },
+    {
+      "epoch": 0.1396,
+      "grad_norm": 1.9653486013412476,
+      "learning_rate": 9.531391284407896e-06,
+      "loss": 0.4021,
+      "step": 1745
+    },
+    {
+      "epoch": 0.13968,
+      "grad_norm": 1.6100776195526123,
+      "learning_rate": 9.530859558461309e-06,
+      "loss": 0.3279,
+      "step": 1746
+    },
+    {
+      "epoch": 0.13976,
+      "grad_norm": 1.587827205657959,
+      "learning_rate": 9.530327545862398e-06,
+      "loss": 0.3501,
+      "step": 1747
+    },
+    {
+      "epoch": 0.13984,
+      "grad_norm": 1.6959279775619507,
+      "learning_rate": 9.529795246644821e-06,
+      "loss": 0.3558,
+      "step": 1748
+    },
+    {
+      "epoch": 0.13992,
+      "grad_norm": 1.7590053081512451,
+      "learning_rate": 9.529262660842257e-06,
+      "loss": 0.3838,
+      "step": 1749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2163288593292236,
+      "learning_rate": 9.5287297884884e-06,
+      "loss": 0.4245,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14008,
+      "grad_norm": 1.510589599609375,
+      "learning_rate": 9.528196629616963e-06,
+      "loss": 0.3794,
+      "step": 1751
+    },
+    {
+      "epoch": 0.14016,
+      "grad_norm": 1.942276954650879,
+      "learning_rate": 9.527663184261674e-06,
+      "loss": 0.4612,
+      "step": 1752
+    },
+    {
+      "epoch": 0.14024,
+      "grad_norm": 1.3821239471435547,
+      "learning_rate": 9.527129452456288e-06,
+      "loss": 0.3189,
+      "step": 1753
+    },
+    {
+      "epoch": 0.14032,
+      "grad_norm": 2.456770181655884,
+      "learning_rate": 9.526595434234567e-06,
+      "loss": 0.5582,
+      "step": 1754
+    },
+    {
+      "epoch": 0.1404,
+      "grad_norm": 1.4780219793319702,
+      "learning_rate": 9.5260611296303e-06,
+      "loss": 0.373,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14048,
+      "grad_norm": 1.8656375408172607,
+      "learning_rate": 9.52552653867729e-06,
+      "loss": 0.3918,
+      "step": 1756
+    },
+    {
+      "epoch": 0.14056,
+      "grad_norm": 1.6014630794525146,
+      "learning_rate": 9.524991661409356e-06,
+      "loss": 0.3833,
+      "step": 1757
+    },
+    {
+      "epoch": 0.14064,
+      "grad_norm": 1.5639004707336426,
+      "learning_rate": 9.524456497860342e-06,
+      "loss": 0.4446,
+      "step": 1758
+    },
+    {
+      "epoch": 0.14072,
+      "grad_norm": 1.6596323251724243,
+      "learning_rate": 9.523921048064105e-06,
+      "loss": 0.3756,
+      "step": 1759
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 1.7615050077438354,
+      "learning_rate": 9.523385312054519e-06,
+      "loss": 0.355,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14088,
+      "grad_norm": 1.968165636062622,
+      "learning_rate": 9.52284928986548e-06,
+      "loss": 0.4518,
+      "step": 1761
+    },
+    {
+      "epoch": 0.14096,
+      "grad_norm": 1.9906011819839478,
+      "learning_rate": 9.5223129815309e-06,
+      "loss": 0.4838,
+      "step": 1762
+    },
+    {
+      "epoch": 0.14104,
+      "grad_norm": 1.559403419494629,
+      "learning_rate": 9.52177638708471e-06,
+      "loss": 0.3133,
+      "step": 1763
+    },
+    {
+      "epoch": 0.14112,
+      "grad_norm": 1.639631748199463,
+      "learning_rate": 9.521239506560856e-06,
+      "loss": 0.3229,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1412,
+      "grad_norm": 1.5165354013442993,
+      "learning_rate": 9.520702339993308e-06,
+      "loss": 0.2711,
+      "step": 1765
+    },
+    {
+      "epoch": 0.14128,
+      "grad_norm": 1.5883980989456177,
+      "learning_rate": 9.520164887416048e-06,
+      "loss": 0.3004,
+      "step": 1766
+    },
+    {
+      "epoch": 0.14136,
+      "grad_norm": 2.106351852416992,
+      "learning_rate": 9.519627148863083e-06,
+      "loss": 0.3873,
+      "step": 1767
+    },
+    {
+      "epoch": 0.14144,
+      "grad_norm": 1.6328215599060059,
+      "learning_rate": 9.519089124368428e-06,
+      "loss": 0.3724,
+      "step": 1768
+    },
+    {
+      "epoch": 0.14152,
+      "grad_norm": 1.9253731966018677,
+      "learning_rate": 9.518550813966127e-06,
+      "loss": 0.4179,
+      "step": 1769
+    },
+    {
+      "epoch": 0.1416,
+      "grad_norm": 1.5602757930755615,
+      "learning_rate": 9.518012217690233e-06,
+      "loss": 0.3443,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14168,
+      "grad_norm": 0.9548719525337219,
+      "learning_rate": 9.517473335574826e-06,
+      "loss": 0.2493,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14176,
+      "grad_norm": 1.837083339691162,
+      "learning_rate": 9.516934167653995e-06,
+      "loss": 0.4071,
+      "step": 1772
+    },
+    {
+      "epoch": 0.14184,
+      "grad_norm": 1.7493815422058105,
+      "learning_rate": 9.516394713961851e-06,
+      "loss": 0.3572,
+      "step": 1773
+    },
+    {
+      "epoch": 0.14192,
+      "grad_norm": 1.7509686946868896,
+      "learning_rate": 9.51585497453253e-06,
+      "loss": 0.38,
+      "step": 1774
+    },
+    {
+      "epoch": 0.142,
+      "grad_norm": 1.8923393487930298,
+      "learning_rate": 9.515314949400172e-06,
+      "loss": 0.3667,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14208,
+      "grad_norm": 1.5569912195205688,
+      "learning_rate": 9.514774638598945e-06,
+      "loss": 0.3337,
+      "step": 1776
+    },
+    {
+      "epoch": 0.14216,
+      "grad_norm": 1.260204792022705,
+      "learning_rate": 9.514234042163033e-06,
+      "loss": 0.2805,
+      "step": 1777
+    },
+    {
+      "epoch": 0.14224,
+      "grad_norm": 1.447977066040039,
+      "learning_rate": 9.51369316012664e-06,
+      "loss": 0.3253,
+      "step": 1778
+    },
+    {
+      "epoch": 0.14232,
+      "grad_norm": 1.598922610282898,
+      "learning_rate": 9.513151992523982e-06,
+      "loss": 0.3362,
+      "step": 1779
+    },
+    {
+      "epoch": 0.1424,
+      "grad_norm": 1.7492644786834717,
+      "learning_rate": 9.512610539389297e-06,
+      "loss": 0.3523,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14248,
+      "grad_norm": 1.6594542264938354,
+      "learning_rate": 9.512068800756845e-06,
+      "loss": 0.3077,
+      "step": 1781
+    },
+    {
+      "epoch": 0.14256,
+      "grad_norm": 1.445451259613037,
+      "learning_rate": 9.511526776660898e-06,
+      "loss": 0.2962,
+      "step": 1782
+    },
+    {
+      "epoch": 0.14264,
+      "grad_norm": 1.554978609085083,
+      "learning_rate": 9.510984467135744e-06,
+      "loss": 0.4125,
+      "step": 1783
+    },
+    {
+      "epoch": 0.14272,
+      "grad_norm": 1.4923337697982788,
+      "learning_rate": 9.5104418722157e-06,
+      "loss": 0.2499,
+      "step": 1784
+    },
+    {
+      "epoch": 0.1428,
+      "grad_norm": 1.4580963850021362,
+      "learning_rate": 9.509898991935088e-06,
+      "loss": 0.3109,
+      "step": 1785
+    },
+    {
+      "epoch": 0.14288,
+      "grad_norm": 1.3246641159057617,
+      "learning_rate": 9.50935582632826e-06,
+      "loss": 0.2577,
+      "step": 1786
+    },
+    {
+      "epoch": 0.14296,
+      "grad_norm": 1.401862382888794,
+      "learning_rate": 9.508812375429575e-06,
+      "loss": 0.34,
+      "step": 1787
+    },
+    {
+      "epoch": 0.14304,
+      "grad_norm": 1.5076367855072021,
+      "learning_rate": 9.508268639273417e-06,
+      "loss": 0.4019,
+      "step": 1788
+    },
+    {
+      "epoch": 0.14312,
+      "grad_norm": 1.6799163818359375,
+      "learning_rate": 9.507724617894188e-06,
+      "loss": 0.4145,
+      "step": 1789
+    },
+    {
+      "epoch": 0.1432,
+      "grad_norm": 1.2659521102905273,
+      "learning_rate": 9.507180311326306e-06,
+      "loss": 0.2803,
+      "step": 1790
+    },
+    {
+      "epoch": 0.14328,
+      "grad_norm": 2.103095054626465,
+      "learning_rate": 9.506635719604207e-06,
+      "loss": 0.4016,
+      "step": 1791
+    },
+    {
+      "epoch": 0.14336,
+      "grad_norm": 1.4720572233200073,
+      "learning_rate": 9.506090842762344e-06,
+      "loss": 0.3029,
+      "step": 1792
+    },
+    {
+      "epoch": 0.14344,
+      "grad_norm": 1.6554877758026123,
+      "learning_rate": 9.50554568083519e-06,
+      "loss": 0.3414,
+      "step": 1793
+    },
+    {
+      "epoch": 0.14352,
+      "grad_norm": 1.4517250061035156,
+      "learning_rate": 9.505000233857238e-06,
+      "loss": 0.423,
+      "step": 1794
+    },
+    {
+      "epoch": 0.1436,
+      "grad_norm": 1.9899773597717285,
+      "learning_rate": 9.504454501862994e-06,
+      "loss": 0.3435,
+      "step": 1795
+    },
+    {
+      "epoch": 0.14368,
+      "grad_norm": 1.4385102987289429,
+      "learning_rate": 9.503908484886986e-06,
+      "loss": 0.3223,
+      "step": 1796
+    },
+    {
+      "epoch": 0.14376,
+      "grad_norm": 1.4701497554779053,
+      "learning_rate": 9.503362182963757e-06,
+      "loss": 0.3363,
+      "step": 1797
+    },
+    {
+      "epoch": 0.14384,
+      "grad_norm": 2.0168707370758057,
+      "learning_rate": 9.502815596127874e-06,
+      "loss": 0.5457,
+      "step": 1798
+    },
+    {
+      "epoch": 0.14392,
+      "grad_norm": 1.6727696657180786,
+      "learning_rate": 9.502268724413913e-06,
+      "loss": 0.3096,
+      "step": 1799
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 1.4367340803146362,
+      "learning_rate": 9.501721567856475e-06,
+      "loss": 0.2831,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14408,
+      "grad_norm": 1.5396634340286255,
+      "learning_rate": 9.501174126490176e-06,
+      "loss": 0.3791,
+      "step": 1801
+    },
+    {
+      "epoch": 0.14416,
+      "grad_norm": 1.8134430646896362,
+      "learning_rate": 9.500626400349651e-06,
+      "loss": 0.5339,
+      "step": 1802
+    },
+    {
+      "epoch": 0.14424,
+      "grad_norm": 1.4883620738983154,
+      "learning_rate": 9.500078389469551e-06,
+      "loss": 0.3966,
+      "step": 1803
+    },
+    {
+      "epoch": 0.14432,
+      "grad_norm": 1.666069746017456,
+      "learning_rate": 9.49953009388455e-06,
+      "loss": 0.3484,
+      "step": 1804
+    },
+    {
+      "epoch": 0.1444,
+      "grad_norm": 1.8850048780441284,
+      "learning_rate": 9.498981513629336e-06,
+      "loss": 0.3768,
+      "step": 1805
+    },
+    {
+      "epoch": 0.14448,
+      "grad_norm": 1.885728359222412,
+      "learning_rate": 9.498432648738616e-06,
+      "loss": 0.413,
+      "step": 1806
+    },
+    {
+      "epoch": 0.14456,
+      "grad_norm": 1.6938875913619995,
+      "learning_rate": 9.497883499247112e-06,
+      "loss": 0.3417,
+      "step": 1807
+    },
+    {
+      "epoch": 0.14464,
+      "grad_norm": 1.6095532178878784,
+      "learning_rate": 9.49733406518957e-06,
+      "loss": 0.3333,
+      "step": 1808
+    },
+    {
+      "epoch": 0.14472,
+      "grad_norm": 2.2393977642059326,
+      "learning_rate": 9.496784346600749e-06,
+      "loss": 0.4968,
+      "step": 1809
+    },
+    {
+      "epoch": 0.1448,
+      "grad_norm": 1.5077526569366455,
+      "learning_rate": 9.496234343515428e-06,
+      "loss": 0.3411,
+      "step": 1810
+    },
+    {
+      "epoch": 0.14488,
+      "grad_norm": 1.496593713760376,
+      "learning_rate": 9.495684055968408e-06,
+      "loss": 0.468,
+      "step": 1811
+    },
+    {
+      "epoch": 0.14496,
+      "grad_norm": 1.466931700706482,
+      "learning_rate": 9.495133483994498e-06,
+      "loss": 0.3441,
+      "step": 1812
+    },
+    {
+      "epoch": 0.14504,
+      "grad_norm": 1.3516722917556763,
+      "learning_rate": 9.494582627628533e-06,
+      "loss": 0.2619,
+      "step": 1813
+    },
+    {
+      "epoch": 0.14512,
+      "grad_norm": 2.093080520629883,
+      "learning_rate": 9.494031486905366e-06,
+      "loss": 0.4629,
+      "step": 1814
+    },
+    {
+      "epoch": 0.1452,
+      "grad_norm": 1.777840256690979,
+      "learning_rate": 9.493480061859861e-06,
+      "loss": 0.3284,
+      "step": 1815
+    },
+    {
+      "epoch": 0.14528,
+      "grad_norm": 1.5537855625152588,
+      "learning_rate": 9.492928352526908e-06,
+      "loss": 0.3827,
+      "step": 1816
+    },
+    {
+      "epoch": 0.14536,
+      "grad_norm": 1.2391389608383179,
+      "learning_rate": 9.492376358941414e-06,
+      "loss": 0.3373,
+      "step": 1817
+    },
+    {
+      "epoch": 0.14544,
+      "grad_norm": 1.3341397047042847,
+      "learning_rate": 9.4918240811383e-06,
+      "loss": 0.2814,
+      "step": 1818
+    },
+    {
+      "epoch": 0.14552,
+      "grad_norm": 1.5178688764572144,
+      "learning_rate": 9.491271519152503e-06,
+      "loss": 0.3376,
+      "step": 1819
+    },
+    {
+      "epoch": 0.1456,
+      "grad_norm": 1.4326900243759155,
+      "learning_rate": 9.490718673018986e-06,
+      "loss": 0.3139,
+      "step": 1820
+    },
+    {
+      "epoch": 0.14568,
+      "grad_norm": 1.5329378843307495,
+      "learning_rate": 9.490165542772724e-06,
+      "loss": 0.3327,
+      "step": 1821
+    },
+    {
+      "epoch": 0.14576,
+      "grad_norm": 1.8199474811553955,
+      "learning_rate": 9.489612128448714e-06,
+      "loss": 0.4225,
+      "step": 1822
+    },
+    {
+      "epoch": 0.14584,
+      "grad_norm": 1.8470107316970825,
+      "learning_rate": 9.489058430081964e-06,
+      "loss": 0.3648,
+      "step": 1823
+    },
+    {
+      "epoch": 0.14592,
+      "grad_norm": 1.2022221088409424,
+      "learning_rate": 9.48850444770751e-06,
+      "loss": 0.304,
+      "step": 1824
+    },
+    {
+      "epoch": 0.146,
+      "grad_norm": 1.4264217615127563,
+      "learning_rate": 9.487950181360397e-06,
+      "loss": 0.3224,
+      "step": 1825
+    },
+    {
+      "epoch": 0.14608,
+      "grad_norm": 2.1026673316955566,
+      "learning_rate": 9.487395631075693e-06,
+      "loss": 0.4528,
+      "step": 1826
+    },
+    {
+      "epoch": 0.14616,
+      "grad_norm": 1.5366865396499634,
+      "learning_rate": 9.486840796888483e-06,
+      "loss": 0.2714,
+      "step": 1827
+    },
+    {
+      "epoch": 0.14624,
+      "grad_norm": 1.836358666419983,
+      "learning_rate": 9.48628567883387e-06,
+      "loss": 0.3978,
+      "step": 1828
+    },
+    {
+      "epoch": 0.14632,
+      "grad_norm": 1.6860939264297485,
+      "learning_rate": 9.48573027694697e-06,
+      "loss": 0.3714,
+      "step": 1829
+    },
+    {
+      "epoch": 0.1464,
+      "grad_norm": 1.3524024486541748,
+      "learning_rate": 9.485174591262925e-06,
+      "loss": 0.2514,
+      "step": 1830
+    },
+    {
+      "epoch": 0.14648,
+      "grad_norm": 1.6706621646881104,
+      "learning_rate": 9.484618621816892e-06,
+      "loss": 0.3619,
+      "step": 1831
+    },
+    {
+      "epoch": 0.14656,
+      "grad_norm": 1.2413554191589355,
+      "learning_rate": 9.484062368644045e-06,
+      "loss": 0.3066,
+      "step": 1832
+    },
+    {
+      "epoch": 0.14664,
+      "grad_norm": 1.7254809141159058,
+      "learning_rate": 9.483505831779577e-06,
+      "loss": 0.3962,
+      "step": 1833
+    },
+    {
+      "epoch": 0.14672,
+      "grad_norm": 1.8376824855804443,
+      "learning_rate": 9.482949011258693e-06,
+      "loss": 0.5107,
+      "step": 1834
+    },
+    {
+      "epoch": 0.1468,
+      "grad_norm": 1.2581253051757812,
+      "learning_rate": 9.482391907116628e-06,
+      "loss": 0.2733,
+      "step": 1835
+    },
+    {
+      "epoch": 0.14688,
+      "grad_norm": 1.1894922256469727,
+      "learning_rate": 9.481834519388624e-06,
+      "loss": 0.2422,
+      "step": 1836
+    },
+    {
+      "epoch": 0.14696,
+      "grad_norm": 1.6918854713439941,
+      "learning_rate": 9.481276848109947e-06,
+      "loss": 0.3263,
+      "step": 1837
+    },
+    {
+      "epoch": 0.14704,
+      "grad_norm": 1.6978294849395752,
+      "learning_rate": 9.480718893315876e-06,
+      "loss": 0.3933,
+      "step": 1838
+    },
+    {
+      "epoch": 0.14712,
+      "grad_norm": 1.701881766319275,
+      "learning_rate": 9.480160655041717e-06,
+      "loss": 0.3639,
+      "step": 1839
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 1.7398412227630615,
+      "learning_rate": 9.479602133322781e-06,
+      "loss": 0.3151,
+      "step": 1840
+    },
+    {
+      "epoch": 0.14728,
+      "grad_norm": 1.8988409042358398,
+      "learning_rate": 9.479043328194409e-06,
+      "loss": 0.4829,
+      "step": 1841
+    },
+    {
+      "epoch": 0.14736,
+      "grad_norm": 1.7794743776321411,
+      "learning_rate": 9.47848423969195e-06,
+      "loss": 0.3512,
+      "step": 1842
+    },
+    {
+      "epoch": 0.14744,
+      "grad_norm": 1.6518741846084595,
+      "learning_rate": 9.477924867850781e-06,
+      "loss": 0.35,
+      "step": 1843
+    },
+    {
+      "epoch": 0.14752,
+      "grad_norm": 2.0679967403411865,
+      "learning_rate": 9.477365212706286e-06,
+      "loss": 0.3591,
+      "step": 1844
+    },
+    {
+      "epoch": 0.1476,
+      "grad_norm": 1.650707721710205,
+      "learning_rate": 9.476805274293877e-06,
+      "loss": 0.2558,
+      "step": 1845
+    },
+    {
+      "epoch": 0.14768,
+      "grad_norm": 1.2669817209243774,
+      "learning_rate": 9.476245052648978e-06,
+      "loss": 0.2441,
+      "step": 1846
+    },
+    {
+      "epoch": 0.14776,
+      "grad_norm": 1.504056692123413,
+      "learning_rate": 9.475684547807032e-06,
+      "loss": 0.2845,
+      "step": 1847
+    },
+    {
+      "epoch": 0.14784,
+      "grad_norm": 0.9670877456665039,
+      "learning_rate": 9.4751237598035e-06,
+      "loss": 0.1881,
+      "step": 1848
+    },
+    {
+      "epoch": 0.14792,
+      "grad_norm": 1.458354115486145,
+      "learning_rate": 9.474562688673861e-06,
+      "loss": 0.3841,
+      "step": 1849
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 1.9649925231933594,
+      "learning_rate": 9.474001334453613e-06,
+      "loss": 0.3908,
+      "step": 1850
+    },
+    {
+      "epoch": 0.14808,
+      "grad_norm": 1.5312343835830688,
+      "learning_rate": 9.47343969717827e-06,
+      "loss": 0.3213,
+      "step": 1851
+    },
+    {
+      "epoch": 0.14816,
+      "grad_norm": 1.8272303342819214,
+      "learning_rate": 9.472877776883365e-06,
+      "loss": 0.3786,
+      "step": 1852
+    },
+    {
+      "epoch": 0.14824,
+      "grad_norm": 1.590427041053772,
+      "learning_rate": 9.47231557360445e-06,
+      "loss": 0.3035,
+      "step": 1853
+    },
+    {
+      "epoch": 0.14832,
+      "grad_norm": 1.4117902517318726,
+      "learning_rate": 9.471753087377094e-06,
+      "loss": 0.4148,
+      "step": 1854
+    },
+    {
+      "epoch": 0.1484,
+      "grad_norm": 2.0782604217529297,
+      "learning_rate": 9.471190318236883e-06,
+      "loss": 0.7431,
+      "step": 1855
+    },
+    {
+      "epoch": 0.14848,
+      "grad_norm": 1.819188117980957,
+      "learning_rate": 9.47062726621942e-06,
+      "loss": 0.3289,
+      "step": 1856
+    },
+    {
+      "epoch": 0.14856,
+      "grad_norm": 1.4712084531784058,
+      "learning_rate": 9.470063931360329e-06,
+      "loss": 0.3727,
+      "step": 1857
+    },
+    {
+      "epoch": 0.14864,
+      "grad_norm": 1.7734615802764893,
+      "learning_rate": 9.46950031369525e-06,
+      "loss": 0.3527,
+      "step": 1858
+    },
+    {
+      "epoch": 0.14872,
+      "grad_norm": 1.6210514307022095,
+      "learning_rate": 9.468936413259842e-06,
+      "loss": 0.3714,
+      "step": 1859
+    },
+    {
+      "epoch": 0.1488,
+      "grad_norm": 1.4350985288619995,
+      "learning_rate": 9.468372230089779e-06,
+      "loss": 0.3185,
+      "step": 1860
+    },
+    {
+      "epoch": 0.14888,
+      "grad_norm": 1.2719461917877197,
+      "learning_rate": 9.467807764220757e-06,
+      "loss": 0.2664,
+      "step": 1861
+    },
+    {
+      "epoch": 0.14896,
+      "grad_norm": 1.8364416360855103,
+      "learning_rate": 9.467243015688486e-06,
+      "loss": 0.5466,
+      "step": 1862
+    },
+    {
+      "epoch": 0.14904,
+      "grad_norm": 1.5372499227523804,
+      "learning_rate": 9.466677984528698e-06,
+      "loss": 0.3587,
+      "step": 1863
+    },
+    {
+      "epoch": 0.14912,
+      "grad_norm": 1.268711805343628,
+      "learning_rate": 9.46611267077714e-06,
+      "loss": 0.2545,
+      "step": 1864
+    },
+    {
+      "epoch": 0.1492,
+      "grad_norm": 1.8258861303329468,
+      "learning_rate": 9.465547074469576e-06,
+      "loss": 0.4055,
+      "step": 1865
+    },
+    {
+      "epoch": 0.14928,
+      "grad_norm": 1.428105354309082,
+      "learning_rate": 9.46498119564179e-06,
+      "loss": 0.3134,
+      "step": 1866
+    },
+    {
+      "epoch": 0.14936,
+      "grad_norm": 1.4180129766464233,
+      "learning_rate": 9.464415034329584e-06,
+      "loss": 0.3236,
+      "step": 1867
+    },
+    {
+      "epoch": 0.14944,
+      "grad_norm": 1.5813634395599365,
+      "learning_rate": 9.463848590568776e-06,
+      "loss": 0.3679,
+      "step": 1868
+    },
+    {
+      "epoch": 0.14952,
+      "grad_norm": 1.8351691961288452,
+      "learning_rate": 9.463281864395204e-06,
+      "loss": 0.3565,
+      "step": 1869
+    },
+    {
+      "epoch": 0.1496,
+      "grad_norm": 1.5723137855529785,
+      "learning_rate": 9.462714855844724e-06,
+      "loss": 0.3112,
+      "step": 1870
+    },
+    {
+      "epoch": 0.14968,
+      "grad_norm": 1.319880723953247,
+      "learning_rate": 9.462147564953206e-06,
+      "loss": 0.2729,
+      "step": 1871
+    },
+    {
+      "epoch": 0.14976,
+      "grad_norm": 1.9398796558380127,
+      "learning_rate": 9.461579991756543e-06,
+      "loss": 0.4355,
+      "step": 1872
+    },
+    {
+      "epoch": 0.14984,
+      "grad_norm": 1.74547278881073,
+      "learning_rate": 9.461012136290641e-06,
+      "loss": 0.4414,
+      "step": 1873
+    },
+    {
+      "epoch": 0.14992,
+      "grad_norm": 1.329624891281128,
+      "learning_rate": 9.460443998591429e-06,
+      "loss": 0.3745,
+      "step": 1874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7463513612747192,
+      "learning_rate": 9.45987557869485e-06,
+      "loss": 0.5341,
+      "step": 1875
+    },
+    {
+      "epoch": 0.15008,
+      "grad_norm": 2.0810306072235107,
+      "learning_rate": 9.459306876636865e-06,
+      "loss": 0.5546,
+      "step": 1876
+    },
+    {
+      "epoch": 0.15016,
+      "grad_norm": 1.8790583610534668,
+      "learning_rate": 9.458737892453455e-06,
+      "loss": 0.5832,
+      "step": 1877
+    },
+    {
+      "epoch": 0.15024,
+      "grad_norm": 1.392000675201416,
+      "learning_rate": 9.458168626180619e-06,
+      "loss": 0.2942,
+      "step": 1878
+    },
+    {
+      "epoch": 0.15032,
+      "grad_norm": 1.9226977825164795,
+      "learning_rate": 9.457599077854369e-06,
+      "loss": 0.3703,
+      "step": 1879
+    },
+    {
+      "epoch": 0.1504,
+      "grad_norm": 1.9512839317321777,
+      "learning_rate": 9.457029247510742e-06,
+      "loss": 0.3894,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15048,
+      "grad_norm": 1.148292899131775,
+      "learning_rate": 9.456459135185787e-06,
+      "loss": 0.2978,
+      "step": 1881
+    },
+    {
+      "epoch": 0.15056,
+      "grad_norm": 1.5737361907958984,
+      "learning_rate": 9.455888740915573e-06,
+      "loss": 0.3149,
+      "step": 1882
+    },
+    {
+      "epoch": 0.15064,
+      "grad_norm": 1.5501254796981812,
+      "learning_rate": 9.45531806473619e-06,
+      "loss": 0.3572,
+      "step": 1883
+    },
+    {
+      "epoch": 0.15072,
+      "grad_norm": 1.52022385597229,
+      "learning_rate": 9.45474710668374e-06,
+      "loss": 0.2995,
+      "step": 1884
+    },
+    {
+      "epoch": 0.1508,
+      "grad_norm": 2.051806926727295,
+      "learning_rate": 9.454175866794344e-06,
+      "loss": 0.4449,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15088,
+      "grad_norm": 1.4593029022216797,
+      "learning_rate": 9.453604345104146e-06,
+      "loss": 0.3873,
+      "step": 1886
+    },
+    {
+      "epoch": 0.15096,
+      "grad_norm": 1.7019308805465698,
+      "learning_rate": 9.453032541649304e-06,
+      "loss": 0.3382,
+      "step": 1887
+    },
+    {
+      "epoch": 0.15104,
+      "grad_norm": 1.7120949029922485,
+      "learning_rate": 9.452460456465991e-06,
+      "loss": 0.3859,
+      "step": 1888
+    },
+    {
+      "epoch": 0.15112,
+      "grad_norm": 1.4484901428222656,
+      "learning_rate": 9.451888089590404e-06,
+      "loss": 0.3352,
+      "step": 1889
+    },
+    {
+      "epoch": 0.1512,
+      "grad_norm": 1.4985060691833496,
+      "learning_rate": 9.451315441058754e-06,
+      "loss": 0.3172,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15128,
+      "grad_norm": 1.5723661184310913,
+      "learning_rate": 9.45074251090727e-06,
+      "loss": 0.394,
+      "step": 1891
+    },
+    {
+      "epoch": 0.15136,
+      "grad_norm": 1.131264567375183,
+      "learning_rate": 9.450169299172201e-06,
+      "loss": 0.2274,
+      "step": 1892
+    },
+    {
+      "epoch": 0.15144,
+      "grad_norm": 1.2636674642562866,
+      "learning_rate": 9.449595805889809e-06,
+      "loss": 0.3056,
+      "step": 1893
+    },
+    {
+      "epoch": 0.15152,
+      "grad_norm": 1.8149745464324951,
+      "learning_rate": 9.449022031096378e-06,
+      "loss": 0.3782,
+      "step": 1894
+    },
+    {
+      "epoch": 0.1516,
+      "grad_norm": 2.029224157333374,
+      "learning_rate": 9.448447974828209e-06,
+      "loss": 0.3975,
+      "step": 1895
+    },
+    {
+      "epoch": 0.15168,
+      "grad_norm": 1.4619930982589722,
+      "learning_rate": 9.447873637121624e-06,
+      "loss": 0.3157,
+      "step": 1896
+    },
+    {
+      "epoch": 0.15176,
+      "grad_norm": 1.570886254310608,
+      "learning_rate": 9.447299018012954e-06,
+      "loss": 0.372,
+      "step": 1897
+    },
+    {
+      "epoch": 0.15184,
+      "grad_norm": 1.4011813402175903,
+      "learning_rate": 9.446724117538559e-06,
+      "loss": 0.3174,
+      "step": 1898
+    },
+    {
+      "epoch": 0.15192,
+      "grad_norm": 1.6975699663162231,
+      "learning_rate": 9.446148935734804e-06,
+      "loss": 0.3275,
+      "step": 1899
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 3.675276041030884,
+      "learning_rate": 9.445573472638085e-06,
+      "loss": 0.4014,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15208,
+      "grad_norm": 1.3942408561706543,
+      "learning_rate": 9.444997728284808e-06,
+      "loss": 0.3226,
+      "step": 1901
+    },
+    {
+      "epoch": 0.15216,
+      "grad_norm": 1.4078660011291504,
+      "learning_rate": 9.444421702711397e-06,
+      "loss": 0.2997,
+      "step": 1902
+    },
+    {
+      "epoch": 0.15224,
+      "grad_norm": 1.1963045597076416,
+      "learning_rate": 9.443845395954295e-06,
+      "loss": 0.2498,
+      "step": 1903
+    },
+    {
+      "epoch": 0.15232,
+      "grad_norm": 1.53505277633667,
+      "learning_rate": 9.443268808049966e-06,
+      "loss": 0.3383,
+      "step": 1904
+    },
+    {
+      "epoch": 0.1524,
+      "grad_norm": 1.4499201774597168,
+      "learning_rate": 9.442691939034885e-06,
+      "loss": 0.2753,
+      "step": 1905
+    },
+    {
+      "epoch": 0.15248,
+      "grad_norm": 1.238735556602478,
+      "learning_rate": 9.44211478894555e-06,
+      "loss": 0.2528,
+      "step": 1906
+    },
+    {
+      "epoch": 0.15256,
+      "grad_norm": 1.7521767616271973,
+      "learning_rate": 9.441537357818476e-06,
+      "loss": 0.4218,
+      "step": 1907
+    },
+    {
+      "epoch": 0.15264,
+      "grad_norm": 1.695391297340393,
+      "learning_rate": 9.440959645690195e-06,
+      "loss": 0.3643,
+      "step": 1908
+    },
+    {
+      "epoch": 0.15272,
+      "grad_norm": 1.8289469480514526,
+      "learning_rate": 9.440381652597258e-06,
+      "loss": 0.3741,
+      "step": 1909
+    },
+    {
+      "epoch": 0.1528,
+      "grad_norm": 1.3207136392593384,
+      "learning_rate": 9.43980337857623e-06,
+      "loss": 0.3747,
+      "step": 1910
+    },
+    {
+      "epoch": 0.15288,
+      "grad_norm": 1.8912609815597534,
+      "learning_rate": 9.439224823663698e-06,
+      "loss": 0.4154,
+      "step": 1911
+    },
+    {
+      "epoch": 0.15296,
+      "grad_norm": 1.8537139892578125,
+      "learning_rate": 9.438645987896264e-06,
+      "loss": 0.4575,
+      "step": 1912
+    },
+    {
+      "epoch": 0.15304,
+      "grad_norm": 1.393486499786377,
+      "learning_rate": 9.438066871310552e-06,
+      "loss": 0.2722,
+      "step": 1913
+    },
+    {
+      "epoch": 0.15312,
+      "grad_norm": 1.4133944511413574,
+      "learning_rate": 9.437487473943198e-06,
+      "loss": 0.335,
+      "step": 1914
+    },
+    {
+      "epoch": 0.1532,
+      "grad_norm": 1.4364306926727295,
+      "learning_rate": 9.436907795830861e-06,
+      "loss": 0.302,
+      "step": 1915
+    },
+    {
+      "epoch": 0.15328,
+      "grad_norm": 1.730425477027893,
+      "learning_rate": 9.43632783701021e-06,
+      "loss": 0.3615,
+      "step": 1916
+    },
+    {
+      "epoch": 0.15336,
+      "grad_norm": 1.5723671913146973,
+      "learning_rate": 9.435747597517943e-06,
+      "loss": 0.2994,
+      "step": 1917
+    },
+    {
+      "epoch": 0.15344,
+      "grad_norm": 1.4602887630462646,
+      "learning_rate": 9.435167077390768e-06,
+      "loss": 0.3276,
+      "step": 1918
+    },
+    {
+      "epoch": 0.15352,
+      "grad_norm": 1.5619248151779175,
+      "learning_rate": 9.434586276665412e-06,
+      "loss": 0.4369,
+      "step": 1919
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 1.581910252571106,
+      "learning_rate": 9.434005195378622e-06,
+      "loss": 0.3914,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15368,
+      "grad_norm": 1.3012194633483887,
+      "learning_rate": 9.433423833567156e-06,
+      "loss": 0.2824,
+      "step": 1921
+    },
+    {
+      "epoch": 0.15376,
+      "grad_norm": 1.5758806467056274,
+      "learning_rate": 9.432842191267802e-06,
+      "loss": 0.2919,
+      "step": 1922
+    },
+    {
+      "epoch": 0.15384,
+      "grad_norm": 2.18058180809021,
+      "learning_rate": 9.432260268517352e-06,
+      "loss": 0.3899,
+      "step": 1923
+    },
+    {
+      "epoch": 0.15392,
+      "grad_norm": 1.6730529069900513,
+      "learning_rate": 9.431678065352625e-06,
+      "loss": 0.5427,
+      "step": 1924
+    },
+    {
+      "epoch": 0.154,
+      "grad_norm": 1.690724492073059,
+      "learning_rate": 9.431095581810457e-06,
+      "loss": 0.3699,
+      "step": 1925
+    },
+    {
+      "epoch": 0.15408,
+      "grad_norm": 1.7529829740524292,
+      "learning_rate": 9.430512817927698e-06,
+      "loss": 0.3121,
+      "step": 1926
+    },
+    {
+      "epoch": 0.15416,
+      "grad_norm": 1.5383788347244263,
+      "learning_rate": 9.429929773741216e-06,
+      "loss": 0.389,
+      "step": 1927
+    },
+    {
+      "epoch": 0.15424,
+      "grad_norm": 2.017871141433716,
+      "learning_rate": 9.429346449287902e-06,
+      "loss": 0.531,
+      "step": 1928
+    },
+    {
+      "epoch": 0.15432,
+      "grad_norm": 1.6736904382705688,
+      "learning_rate": 9.428762844604658e-06,
+      "loss": 0.3876,
+      "step": 1929
+    },
+    {
+      "epoch": 0.1544,
+      "grad_norm": 1.5975745916366577,
+      "learning_rate": 9.428178959728406e-06,
+      "loss": 0.408,
+      "step": 1930
+    },
+    {
+      "epoch": 0.15448,
+      "grad_norm": 1.3620975017547607,
+      "learning_rate": 9.427594794696089e-06,
+      "loss": 0.2987,
+      "step": 1931
+    },
+    {
+      "epoch": 0.15456,
+      "grad_norm": 1.9585412740707397,
+      "learning_rate": 9.427010349544665e-06,
+      "loss": 0.4924,
+      "step": 1932
+    },
+    {
+      "epoch": 0.15464,
+      "grad_norm": 1.616124153137207,
+      "learning_rate": 9.426425624311107e-06,
+      "loss": 0.3609,
+      "step": 1933
+    },
+    {
+      "epoch": 0.15472,
+      "grad_norm": 1.6203333139419556,
+      "learning_rate": 9.425840619032411e-06,
+      "loss": 0.343,
+      "step": 1934
+    },
+    {
+      "epoch": 0.1548,
+      "grad_norm": 1.8410626649856567,
+      "learning_rate": 9.42525533374559e-06,
+      "loss": 0.3722,
+      "step": 1935
+    },
+    {
+      "epoch": 0.15488,
+      "grad_norm": 1.4306490421295166,
+      "learning_rate": 9.424669768487668e-06,
+      "loss": 0.3013,
+      "step": 1936
+    },
+    {
+      "epoch": 0.15496,
+      "grad_norm": 1.5979857444763184,
+      "learning_rate": 9.424083923295698e-06,
+      "loss": 0.4501,
+      "step": 1937
+    },
+    {
+      "epoch": 0.15504,
+      "grad_norm": 1.3215088844299316,
+      "learning_rate": 9.42349779820674e-06,
+      "loss": 0.3814,
+      "step": 1938
+    },
+    {
+      "epoch": 0.15512,
+      "grad_norm": 1.462342619895935,
+      "learning_rate": 9.422911393257876e-06,
+      "loss": 0.3256,
+      "step": 1939
+    },
+    {
+      "epoch": 0.1552,
+      "grad_norm": 1.2837486267089844,
+      "learning_rate": 9.422324708486208e-06,
+      "loss": 0.3029,
+      "step": 1940
+    },
+    {
+      "epoch": 0.15528,
+      "grad_norm": 1.586371660232544,
+      "learning_rate": 9.421737743928854e-06,
+      "loss": 0.4278,
+      "step": 1941
+    },
+    {
+      "epoch": 0.15536,
+      "grad_norm": 1.5170694589614868,
+      "learning_rate": 9.421150499622947e-06,
+      "loss": 0.3776,
+      "step": 1942
+    },
+    {
+      "epoch": 0.15544,
+      "grad_norm": 1.640691876411438,
+      "learning_rate": 9.42056297560564e-06,
+      "loss": 0.2805,
+      "step": 1943
+    },
+    {
+      "epoch": 0.15552,
+      "grad_norm": 1.7329473495483398,
+      "learning_rate": 9.419975171914108e-06,
+      "loss": 0.4477,
+      "step": 1944
+    },
+    {
+      "epoch": 0.1556,
+      "grad_norm": 1.7486817836761475,
+      "learning_rate": 9.419387088585534e-06,
+      "loss": 0.3196,
+      "step": 1945
+    },
+    {
+      "epoch": 0.15568,
+      "grad_norm": 1.9604579210281372,
+      "learning_rate": 9.418798725657125e-06,
+      "loss": 0.373,
+      "step": 1946
+    },
+    {
+      "epoch": 0.15576,
+      "grad_norm": 1.8238966464996338,
+      "learning_rate": 9.418210083166108e-06,
+      "loss": 0.4417,
+      "step": 1947
+    },
+    {
+      "epoch": 0.15584,
+      "grad_norm": 1.3136991262435913,
+      "learning_rate": 9.417621161149723e-06,
+      "loss": 0.2475,
+      "step": 1948
+    },
+    {
+      "epoch": 0.15592,
+      "grad_norm": 2.024016857147217,
+      "learning_rate": 9.417031959645227e-06,
+      "loss": 0.3924,
+      "step": 1949
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 1.4279539585113525,
+      "learning_rate": 9.416442478689898e-06,
+      "loss": 0.3131,
+      "step": 1950
+    },
+    {
+      "epoch": 0.15608,
+      "grad_norm": 1.5639196634292603,
+      "learning_rate": 9.415852718321032e-06,
+      "loss": 0.3048,
+      "step": 1951
+    },
+    {
+      "epoch": 0.15616,
+      "grad_norm": 1.6544080972671509,
+      "learning_rate": 9.41526267857594e-06,
+      "loss": 0.3371,
+      "step": 1952
+    },
+    {
+      "epoch": 0.15624,
+      "grad_norm": 1.7706680297851562,
+      "learning_rate": 9.414672359491952e-06,
+      "loss": 0.3639,
+      "step": 1953
+    },
+    {
+      "epoch": 0.15632,
+      "grad_norm": 1.6847538948059082,
+      "learning_rate": 9.414081761106413e-06,
+      "loss": 0.5029,
+      "step": 1954
+    },
+    {
+      "epoch": 0.1564,
+      "grad_norm": 1.432637095451355,
+      "learning_rate": 9.413490883456694e-06,
+      "loss": 0.2809,
+      "step": 1955
+    },
+    {
+      "epoch": 0.15648,
+      "grad_norm": 1.5604770183563232,
+      "learning_rate": 9.412899726580171e-06,
+      "loss": 0.2855,
+      "step": 1956
+    },
+    {
+      "epoch": 0.15656,
+      "grad_norm": 1.5877360105514526,
+      "learning_rate": 9.41230829051425e-06,
+      "loss": 0.3401,
+      "step": 1957
+    },
+    {
+      "epoch": 0.15664,
+      "grad_norm": 2.027043581008911,
+      "learning_rate": 9.411716575296349e-06,
+      "loss": 0.4226,
+      "step": 1958
+    },
+    {
+      "epoch": 0.15672,
+      "grad_norm": 1.7963011264801025,
+      "learning_rate": 9.411124580963897e-06,
+      "loss": 0.3433,
+      "step": 1959
+    },
+    {
+      "epoch": 0.1568,
+      "grad_norm": 1.6322389841079712,
+      "learning_rate": 9.410532307554356e-06,
+      "loss": 0.3208,
+      "step": 1960
+    },
+    {
+      "epoch": 0.15688,
+      "grad_norm": 1.4425139427185059,
+      "learning_rate": 9.409939755105193e-06,
+      "loss": 0.3305,
+      "step": 1961
+    },
+    {
+      "epoch": 0.15696,
+      "grad_norm": 2.181800365447998,
+      "learning_rate": 9.409346923653897e-06,
+      "loss": 0.428,
+      "step": 1962
+    },
+    {
+      "epoch": 0.15704,
+      "grad_norm": 1.3570939302444458,
+      "learning_rate": 9.408753813237974e-06,
+      "loss": 0.3384,
+      "step": 1963
+    },
+    {
+      "epoch": 0.15712,
+      "grad_norm": 1.302597999572754,
+      "learning_rate": 9.40816042389495e-06,
+      "loss": 0.2947,
+      "step": 1964
+    },
+    {
+      "epoch": 0.1572,
+      "grad_norm": 1.9911285638809204,
+      "learning_rate": 9.407566755662366e-06,
+      "loss": 0.4879,
+      "step": 1965
+    },
+    {
+      "epoch": 0.15728,
+      "grad_norm": 1.518795132637024,
+      "learning_rate": 9.406972808577782e-06,
+      "loss": 0.4356,
+      "step": 1966
+    },
+    {
+      "epoch": 0.15736,
+      "grad_norm": 1.8553831577301025,
+      "learning_rate": 9.406378582678772e-06,
+      "loss": 0.407,
+      "step": 1967
+    },
+    {
+      "epoch": 0.15744,
+      "grad_norm": 1.409943699836731,
+      "learning_rate": 9.405784078002932e-06,
+      "loss": 0.3465,
+      "step": 1968
+    },
+    {
+      "epoch": 0.15752,
+      "grad_norm": 1.3879483938217163,
+      "learning_rate": 9.405189294587879e-06,
+      "loss": 0.3013,
+      "step": 1969
+    },
+    {
+      "epoch": 0.1576,
+      "grad_norm": 1.4256548881530762,
+      "learning_rate": 9.404594232471237e-06,
+      "loss": 0.2841,
+      "step": 1970
+    },
+    {
+      "epoch": 0.15768,
+      "grad_norm": 1.50364351272583,
+      "learning_rate": 9.403998891690655e-06,
+      "loss": 0.3085,
+      "step": 1971
+    },
+    {
+      "epoch": 0.15776,
+      "grad_norm": 1.7137510776519775,
+      "learning_rate": 9.4034032722838e-06,
+      "loss": 0.3598,
+      "step": 1972
+    },
+    {
+      "epoch": 0.15784,
+      "grad_norm": 1.9388797283172607,
+      "learning_rate": 9.402807374288354e-06,
+      "loss": 0.505,
+      "step": 1973
+    },
+    {
+      "epoch": 0.15792,
+      "grad_norm": 1.7556933164596558,
+      "learning_rate": 9.402211197742016e-06,
+      "loss": 0.3934,
+      "step": 1974
+    },
+    {
+      "epoch": 0.158,
+      "grad_norm": 1.7362114191055298,
+      "learning_rate": 9.401614742682508e-06,
+      "loss": 0.3839,
+      "step": 1975
+    },
+    {
+      "epoch": 0.15808,
+      "grad_norm": 1.413800597190857,
+      "learning_rate": 9.40101800914756e-06,
+      "loss": 0.3378,
+      "step": 1976
+    },
+    {
+      "epoch": 0.15816,
+      "grad_norm": 1.6335464715957642,
+      "learning_rate": 9.40042099717493e-06,
+      "loss": 0.4107,
+      "step": 1977
+    },
+    {
+      "epoch": 0.15824,
+      "grad_norm": 1.9229556322097778,
+      "learning_rate": 9.399823706802386e-06,
+      "loss": 0.5094,
+      "step": 1978
+    },
+    {
+      "epoch": 0.15832,
+      "grad_norm": 1.2669042348861694,
+      "learning_rate": 9.399226138067721e-06,
+      "loss": 0.304,
+      "step": 1979
+    },
+    {
+      "epoch": 0.1584,
+      "grad_norm": 1.4909025430679321,
+      "learning_rate": 9.398628291008735e-06,
+      "loss": 0.3361,
+      "step": 1980
+    },
+    {
+      "epoch": 0.15848,
+      "grad_norm": 1.4694348573684692,
+      "learning_rate": 9.398030165663257e-06,
+      "loss": 0.3228,
+      "step": 1981
+    },
+    {
+      "epoch": 0.15856,
+      "grad_norm": 1.7894591093063354,
+      "learning_rate": 9.397431762069124e-06,
+      "loss": 0.4964,
+      "step": 1982
+    },
+    {
+      "epoch": 0.15864,
+      "grad_norm": 1.3913213014602661,
+      "learning_rate": 9.396833080264198e-06,
+      "loss": 0.3159,
+      "step": 1983
+    },
+    {
+      "epoch": 0.15872,
+      "grad_norm": 1.4307069778442383,
+      "learning_rate": 9.396234120286356e-06,
+      "loss": 0.4041,
+      "step": 1984
+    },
+    {
+      "epoch": 0.1588,
+      "grad_norm": 1.663399577140808,
+      "learning_rate": 9.39563488217349e-06,
+      "loss": 0.3734,
+      "step": 1985
+    },
+    {
+      "epoch": 0.15888,
+      "grad_norm": 2.0649588108062744,
+      "learning_rate": 9.395035365963514e-06,
+      "loss": 0.4112,
+      "step": 1986
+    },
+    {
+      "epoch": 0.15896,
+      "grad_norm": 1.6770638227462769,
+      "learning_rate": 9.394435571694356e-06,
+      "loss": 0.3357,
+      "step": 1987
+    },
+    {
+      "epoch": 0.15904,
+      "grad_norm": 2.003188371658325,
+      "learning_rate": 9.393835499403963e-06,
+      "loss": 0.4054,
+      "step": 1988
+    },
+    {
+      "epoch": 0.15912,
+      "grad_norm": 1.657409906387329,
+      "learning_rate": 9.393235149130299e-06,
+      "loss": 0.3624,
+      "step": 1989
+    },
+    {
+      "epoch": 0.1592,
+      "grad_norm": 1.6382983922958374,
+      "learning_rate": 9.392634520911348e-06,
+      "loss": 0.4006,
+      "step": 1990
+    },
+    {
+      "epoch": 0.15928,
+      "grad_norm": 2.041346549987793,
+      "learning_rate": 9.392033614785108e-06,
+      "loss": 0.3929,
+      "step": 1991
+    },
+    {
+      "epoch": 0.15936,
+      "grad_norm": 1.2810925245285034,
+      "learning_rate": 9.391432430789597e-06,
+      "loss": 0.2469,
+      "step": 1992
+    },
+    {
+      "epoch": 0.15944,
+      "grad_norm": 1.6841486692428589,
+      "learning_rate": 9.390830968962849e-06,
+      "loss": 0.3423,
+      "step": 1993
+    },
+    {
+      "epoch": 0.15952,
+      "grad_norm": 1.7549775838851929,
+      "learning_rate": 9.390229229342918e-06,
+      "loss": 0.3976,
+      "step": 1994
+    },
+    {
+      "epoch": 0.1596,
+      "grad_norm": 1.5098248720169067,
+      "learning_rate": 9.389627211967874e-06,
+      "loss": 0.3533,
+      "step": 1995
+    },
+    {
+      "epoch": 0.15968,
+      "grad_norm": 1.392735242843628,
+      "learning_rate": 9.389024916875805e-06,
+      "loss": 0.4128,
+      "step": 1996
+    },
+    {
+      "epoch": 0.15976,
+      "grad_norm": 1.8983796834945679,
+      "learning_rate": 9.388422344104812e-06,
+      "loss": 0.4626,
+      "step": 1997
+    },
+    {
+      "epoch": 0.15984,
+      "grad_norm": 1.4891161918640137,
+      "learning_rate": 9.387819493693025e-06,
+      "loss": 0.3736,
+      "step": 1998
+    },
+    {
+      "epoch": 0.15992,
+      "grad_norm": 1.8602221012115479,
+      "learning_rate": 9.387216365678578e-06,
+      "loss": 0.4016,
+      "step": 1999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5026849508285522,
+      "learning_rate": 9.38661296009963e-06,
+      "loss": 0.3062,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16008,
+      "grad_norm": 1.7527680397033691,
+      "learning_rate": 9.38600927699436e-06,
+      "loss": 0.3501,
+      "step": 2001
+    },
+    {
+      "epoch": 0.16016,
+      "grad_norm": 1.6725231409072876,
+      "learning_rate": 9.385405316400957e-06,
+      "loss": 0.4831,
+      "step": 2002
+    },
+    {
+      "epoch": 0.16024,
+      "grad_norm": 1.6204967498779297,
+      "learning_rate": 9.384801078357635e-06,
+      "loss": 0.3016,
+      "step": 2003
+    },
+    {
+      "epoch": 0.16032,
+      "grad_norm": 1.6982675790786743,
+      "learning_rate": 9.38419656290262e-06,
+      "loss": 0.5008,
+      "step": 2004
+    },
+    {
+      "epoch": 0.1604,
+      "grad_norm": 1.5643174648284912,
+      "learning_rate": 9.383591770074156e-06,
+      "loss": 0.302,
+      "step": 2005
+    },
+    {
+      "epoch": 0.16048,
+      "grad_norm": 1.7776808738708496,
+      "learning_rate": 9.38298669991051e-06,
+      "loss": 0.3796,
+      "step": 2006
+    },
+    {
+      "epoch": 0.16056,
+      "grad_norm": 1.6092027425765991,
+      "learning_rate": 9.38238135244996e-06,
+      "loss": 0.3388,
+      "step": 2007
+    },
+    {
+      "epoch": 0.16064,
+      "grad_norm": 1.8656328916549683,
+      "learning_rate": 9.381775727730807e-06,
+      "loss": 0.3574,
+      "step": 2008
+    },
+    {
+      "epoch": 0.16072,
+      "grad_norm": 1.6534117460250854,
+      "learning_rate": 9.381169825791364e-06,
+      "loss": 0.4165,
+      "step": 2009
+    },
+    {
+      "epoch": 0.1608,
+      "grad_norm": 1.7371861934661865,
+      "learning_rate": 9.380563646669967e-06,
+      "loss": 0.41,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16088,
+      "grad_norm": 1.6235700845718384,
+      "learning_rate": 9.379957190404966e-06,
+      "loss": 0.3675,
+      "step": 2011
+    },
+    {
+      "epoch": 0.16096,
+      "grad_norm": 1.9247312545776367,
+      "learning_rate": 9.379350457034726e-06,
+      "loss": 0.4458,
+      "step": 2012
+    },
+    {
+      "epoch": 0.16104,
+      "grad_norm": 1.2080788612365723,
+      "learning_rate": 9.378743446597635e-06,
+      "loss": 0.2824,
+      "step": 2013
+    },
+    {
+      "epoch": 0.16112,
+      "grad_norm": 1.8663572072982788,
+      "learning_rate": 9.378136159132101e-06,
+      "loss": 0.3204,
+      "step": 2014
+    },
+    {
+      "epoch": 0.1612,
+      "grad_norm": 1.4532349109649658,
+      "learning_rate": 9.37752859467654e-06,
+      "loss": 0.3409,
+      "step": 2015
+    },
+    {
+      "epoch": 0.16128,
+      "grad_norm": 1.467343807220459,
+      "learning_rate": 9.376920753269391e-06,
+      "loss": 0.3027,
+      "step": 2016
+    },
+    {
+      "epoch": 0.16136,
+      "grad_norm": 1.7154619693756104,
+      "learning_rate": 9.376312634949114e-06,
+      "loss": 0.3916,
+      "step": 2017
+    },
+    {
+      "epoch": 0.16144,
+      "grad_norm": 1.4569969177246094,
+      "learning_rate": 9.375704239754178e-06,
+      "loss": 0.316,
+      "step": 2018
+    },
+    {
+      "epoch": 0.16152,
+      "grad_norm": 1.6265618801116943,
+      "learning_rate": 9.375095567723076e-06,
+      "loss": 0.29,
+      "step": 2019
+    },
+    {
+      "epoch": 0.1616,
+      "grad_norm": 1.7646688222885132,
+      "learning_rate": 9.374486618894316e-06,
+      "loss": 0.4374,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16168,
+      "grad_norm": 1.7426738739013672,
+      "learning_rate": 9.373877393306426e-06,
+      "loss": 0.3223,
+      "step": 2021
+    },
+    {
+      "epoch": 0.16176,
+      "grad_norm": 1.8202601671218872,
+      "learning_rate": 9.373267890997949e-06,
+      "loss": 0.3962,
+      "step": 2022
+    },
+    {
+      "epoch": 0.16184,
+      "grad_norm": 1.5177546739578247,
+      "learning_rate": 9.372658112007442e-06,
+      "loss": 0.2866,
+      "step": 2023
+    },
+    {
+      "epoch": 0.16192,
+      "grad_norm": 1.7819550037384033,
+      "learning_rate": 9.37204805637349e-06,
+      "loss": 0.3551,
+      "step": 2024
+    },
+    {
+      "epoch": 0.162,
+      "grad_norm": 1.8583236932754517,
+      "learning_rate": 9.371437724134687e-06,
+      "loss": 0.3201,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16208,
+      "grad_norm": 1.4678624868392944,
+      "learning_rate": 9.370827115329644e-06,
+      "loss": 0.3144,
+      "step": 2026
+    },
+    {
+      "epoch": 0.16216,
+      "grad_norm": 1.5797638893127441,
+      "learning_rate": 9.370216229996995e-06,
+      "loss": 0.2976,
+      "step": 2027
+    },
+    {
+      "epoch": 0.16224,
+      "grad_norm": 1.471345067024231,
+      "learning_rate": 9.369605068175388e-06,
+      "loss": 0.2832,
+      "step": 2028
+    },
+    {
+      "epoch": 0.16232,
+      "grad_norm": 2.2264575958251953,
+      "learning_rate": 9.368993629903489e-06,
+      "loss": 0.4478,
+      "step": 2029
+    },
+    {
+      "epoch": 0.1624,
+      "grad_norm": 2.1455178260803223,
+      "learning_rate": 9.368381915219982e-06,
+      "loss": 0.455,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16248,
+      "grad_norm": 1.47684907913208,
+      "learning_rate": 9.367769924163568e-06,
+      "loss": 0.2801,
+      "step": 2031
+    },
+    {
+      "epoch": 0.16256,
+      "grad_norm": 1.5183435678482056,
+      "learning_rate": 9.367157656772965e-06,
+      "loss": 0.367,
+      "step": 2032
+    },
+    {
+      "epoch": 0.16264,
+      "grad_norm": 1.7493641376495361,
+      "learning_rate": 9.366545113086909e-06,
+      "loss": 0.3876,
+      "step": 2033
+    },
+    {
+      "epoch": 0.16272,
+      "grad_norm": 1.9206058979034424,
+      "learning_rate": 9.365932293144156e-06,
+      "loss": 0.3943,
+      "step": 2034
+    },
+    {
+      "epoch": 0.1628,
+      "grad_norm": 1.4696857929229736,
+      "learning_rate": 9.365319196983474e-06,
+      "loss": 0.316,
+      "step": 2035
+    },
+    {
+      "epoch": 0.16288,
+      "grad_norm": 1.558495044708252,
+      "learning_rate": 9.364705824643653e-06,
+      "loss": 0.3151,
+      "step": 2036
+    },
+    {
+      "epoch": 0.16296,
+      "grad_norm": 1.9205808639526367,
+      "learning_rate": 9.364092176163499e-06,
+      "loss": 0.3735,
+      "step": 2037
+    },
+    {
+      "epoch": 0.16304,
+      "grad_norm": 2.3546855449676514,
+      "learning_rate": 9.363478251581834e-06,
+      "loss": 0.4612,
+      "step": 2038
+    },
+    {
+      "epoch": 0.16312,
+      "grad_norm": 1.5820739269256592,
+      "learning_rate": 9.362864050937503e-06,
+      "loss": 0.3454,
+      "step": 2039
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 1.4024479389190674,
+      "learning_rate": 9.36224957426936e-06,
+      "loss": 0.284,
+      "step": 2040
+    },
+    {
+      "epoch": 0.16328,
+      "grad_norm": 1.346354603767395,
+      "learning_rate": 9.361634821616284e-06,
+      "loss": 0.3249,
+      "step": 2041
+    },
+    {
+      "epoch": 0.16336,
+      "grad_norm": 1.496208667755127,
+      "learning_rate": 9.361019793017164e-06,
+      "loss": 0.4657,
+      "step": 2042
+    },
+    {
+      "epoch": 0.16344,
+      "grad_norm": 1.449812650680542,
+      "learning_rate": 9.360404488510916e-06,
+      "loss": 0.332,
+      "step": 2043
+    },
+    {
+      "epoch": 0.16352,
+      "grad_norm": 1.5065429210662842,
+      "learning_rate": 9.359788908136464e-06,
+      "loss": 0.3212,
+      "step": 2044
+    },
+    {
+      "epoch": 0.1636,
+      "grad_norm": 1.5385485887527466,
+      "learning_rate": 9.359173051932758e-06,
+      "loss": 0.35,
+      "step": 2045
+    },
+    {
+      "epoch": 0.16368,
+      "grad_norm": 1.4125738143920898,
+      "learning_rate": 9.358556919938759e-06,
+      "loss": 0.3793,
+      "step": 2046
+    },
+    {
+      "epoch": 0.16376,
+      "grad_norm": 1.4307483434677124,
+      "learning_rate": 9.357940512193446e-06,
+      "loss": 0.3367,
+      "step": 2047
+    },
+    {
+      "epoch": 0.16384,
+      "grad_norm": 1.469079613685608,
+      "learning_rate": 9.357323828735818e-06,
+      "loss": 0.3201,
+      "step": 2048
+    },
+    {
+      "epoch": 0.16392,
+      "grad_norm": 1.462432861328125,
+      "learning_rate": 9.356706869604892e-06,
+      "loss": 0.2956,
+      "step": 2049
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 1.8476369380950928,
+      "learning_rate": 9.3560896348397e-06,
+      "loss": 0.4874,
+      "step": 2050
+    },
+    {
+      "epoch": 0.16408,
+      "grad_norm": 1.6221674680709839,
+      "learning_rate": 9.355472124479292e-06,
+      "loss": 0.2875,
+      "step": 2051
+    },
+    {
+      "epoch": 0.16416,
+      "grad_norm": 1.772452473640442,
+      "learning_rate": 9.354854338562735e-06,
+      "loss": 0.3395,
+      "step": 2052
+    },
+    {
+      "epoch": 0.16424,
+      "grad_norm": 1.8067185878753662,
+      "learning_rate": 9.354236277129119e-06,
+      "loss": 0.4001,
+      "step": 2053
+    },
+    {
+      "epoch": 0.16432,
+      "grad_norm": 1.7058254480361938,
+      "learning_rate": 9.35361794021754e-06,
+      "loss": 0.48,
+      "step": 2054
+    },
+    {
+      "epoch": 0.1644,
+      "grad_norm": 1.4677443504333496,
+      "learning_rate": 9.352999327867122e-06,
+      "loss": 0.268,
+      "step": 2055
+    },
+    {
+      "epoch": 0.16448,
+      "grad_norm": 1.252163052558899,
+      "learning_rate": 9.352380440117002e-06,
+      "loss": 0.2939,
+      "step": 2056
+    },
+    {
+      "epoch": 0.16456,
+      "grad_norm": 1.5425444841384888,
+      "learning_rate": 9.351761277006332e-06,
+      "loss": 0.3803,
+      "step": 2057
+    },
+    {
+      "epoch": 0.16464,
+      "grad_norm": 1.7046698331832886,
+      "learning_rate": 9.351141838574291e-06,
+      "loss": 0.3808,
+      "step": 2058
+    },
+    {
+      "epoch": 0.16472,
+      "grad_norm": 1.4355201721191406,
+      "learning_rate": 9.350522124860063e-06,
+      "loss": 0.2472,
+      "step": 2059
+    },
+    {
+      "epoch": 0.1648,
+      "grad_norm": 1.3044764995574951,
+      "learning_rate": 9.349902135902857e-06,
+      "loss": 0.27,
+      "step": 2060
+    },
+    {
+      "epoch": 0.16488,
+      "grad_norm": 1.7062314748764038,
+      "learning_rate": 9.349281871741898e-06,
+      "loss": 0.3372,
+      "step": 2061
+    },
+    {
+      "epoch": 0.16496,
+      "grad_norm": 1.5979080200195312,
+      "learning_rate": 9.348661332416428e-06,
+      "loss": 0.3478,
+      "step": 2062
+    },
+    {
+      "epoch": 0.16504,
+      "grad_norm": 1.3202630281448364,
+      "learning_rate": 9.348040517965704e-06,
+      "loss": 0.2525,
+      "step": 2063
+    },
+    {
+      "epoch": 0.16512,
+      "grad_norm": 1.7079370021820068,
+      "learning_rate": 9.347419428429007e-06,
+      "loss": 0.4255,
+      "step": 2064
+    },
+    {
+      "epoch": 0.1652,
+      "grad_norm": 2.1258163452148438,
+      "learning_rate": 9.34679806384563e-06,
+      "loss": 0.4046,
+      "step": 2065
+    },
+    {
+      "epoch": 0.16528,
+      "grad_norm": 1.4599392414093018,
+      "learning_rate": 9.346176424254883e-06,
+      "loss": 0.3442,
+      "step": 2066
+    },
+    {
+      "epoch": 0.16536,
+      "grad_norm": 1.5100538730621338,
+      "learning_rate": 9.345554509696096e-06,
+      "loss": 0.314,
+      "step": 2067
+    },
+    {
+      "epoch": 0.16544,
+      "grad_norm": 1.472424030303955,
+      "learning_rate": 9.344932320208615e-06,
+      "loss": 0.3324,
+      "step": 2068
+    },
+    {
+      "epoch": 0.16552,
+      "grad_norm": 1.7998536825180054,
+      "learning_rate": 9.344309855831806e-06,
+      "loss": 0.4959,
+      "step": 2069
+    },
+    {
+      "epoch": 0.1656,
+      "grad_norm": 1.8450273275375366,
+      "learning_rate": 9.343687116605045e-06,
+      "loss": 0.3331,
+      "step": 2070
+    },
+    {
+      "epoch": 0.16568,
+      "grad_norm": 1.7891483306884766,
+      "learning_rate": 9.343064102567738e-06,
+      "loss": 0.3411,
+      "step": 2071
+    },
+    {
+      "epoch": 0.16576,
+      "grad_norm": 1.251073956489563,
+      "learning_rate": 9.342440813759294e-06,
+      "loss": 0.2926,
+      "step": 2072
+    },
+    {
+      "epoch": 0.16584,
+      "grad_norm": 1.9193722009658813,
+      "learning_rate": 9.341817250219153e-06,
+      "loss": 0.4983,
+      "step": 2073
+    },
+    {
+      "epoch": 0.16592,
+      "grad_norm": 1.5404998064041138,
+      "learning_rate": 9.34119341198676e-06,
+      "loss": 0.4395,
+      "step": 2074
+    },
+    {
+      "epoch": 0.166,
+      "grad_norm": 1.6261802911758423,
+      "learning_rate": 9.340569299101584e-06,
+      "loss": 0.382,
+      "step": 2075
+    },
+    {
+      "epoch": 0.16608,
+      "grad_norm": 1.4044370651245117,
+      "learning_rate": 9.339944911603116e-06,
+      "loss": 0.3091,
+      "step": 2076
+    },
+    {
+      "epoch": 0.16616,
+      "grad_norm": 1.707542061805725,
+      "learning_rate": 9.339320249530851e-06,
+      "loss": 0.3379,
+      "step": 2077
+    },
+    {
+      "epoch": 0.16624,
+      "grad_norm": 2.2420706748962402,
+      "learning_rate": 9.338695312924317e-06,
+      "loss": 0.4235,
+      "step": 2078
+    },
+    {
+      "epoch": 0.16632,
+      "grad_norm": 1.7224174737930298,
+      "learning_rate": 9.338070101823046e-06,
+      "loss": 0.3715,
+      "step": 2079
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 1.4423843622207642,
+      "learning_rate": 9.337444616266595e-06,
+      "loss": 0.3355,
+      "step": 2080
+    },
+    {
+      "epoch": 0.16648,
+      "grad_norm": 1.2990514039993286,
+      "learning_rate": 9.336818856294535e-06,
+      "loss": 0.2665,
+      "step": 2081
+    },
+    {
+      "epoch": 0.16656,
+      "grad_norm": 1.7929611206054688,
+      "learning_rate": 9.336192821946459e-06,
+      "loss": 0.34,
+      "step": 2082
+    },
+    {
+      "epoch": 0.16664,
+      "grad_norm": 1.742226243019104,
+      "learning_rate": 9.33556651326197e-06,
+      "loss": 0.3516,
+      "step": 2083
+    },
+    {
+      "epoch": 0.16672,
+      "grad_norm": 1.5436040163040161,
+      "learning_rate": 9.334939930280698e-06,
+      "loss": 0.2721,
+      "step": 2084
+    },
+    {
+      "epoch": 0.1668,
+      "grad_norm": 2.2682416439056396,
+      "learning_rate": 9.334313073042279e-06,
+      "loss": 0.4935,
+      "step": 2085
+    },
+    {
+      "epoch": 0.16688,
+      "grad_norm": 1.4342129230499268,
+      "learning_rate": 9.333685941586375e-06,
+      "loss": 0.382,
+      "step": 2086
+    },
+    {
+      "epoch": 0.16696,
+      "grad_norm": 1.3763701915740967,
+      "learning_rate": 9.333058535952661e-06,
+      "loss": 0.2874,
+      "step": 2087
+    },
+    {
+      "epoch": 0.16704,
+      "grad_norm": 1.4731640815734863,
+      "learning_rate": 9.332430856180831e-06,
+      "loss": 0.3329,
+      "step": 2088
+    },
+    {
+      "epoch": 0.16712,
+      "grad_norm": 1.6573126316070557,
+      "learning_rate": 9.3318029023106e-06,
+      "loss": 0.4924,
+      "step": 2089
+    },
+    {
+      "epoch": 0.1672,
+      "grad_norm": 1.5075602531433105,
+      "learning_rate": 9.331174674381692e-06,
+      "loss": 0.3175,
+      "step": 2090
+    },
+    {
+      "epoch": 0.16728,
+      "grad_norm": 1.938504934310913,
+      "learning_rate": 9.330546172433855e-06,
+      "loss": 0.4211,
+      "step": 2091
+    },
+    {
+      "epoch": 0.16736,
+      "grad_norm": 1.2835865020751953,
+      "learning_rate": 9.329917396506851e-06,
+      "loss": 0.2607,
+      "step": 2092
+    },
+    {
+      "epoch": 0.16744,
+      "grad_norm": 1.6493488550186157,
+      "learning_rate": 9.329288346640462e-06,
+      "loss": 0.3207,
+      "step": 2093
+    },
+    {
+      "epoch": 0.16752,
+      "grad_norm": 1.4919483661651611,
+      "learning_rate": 9.328659022874486e-06,
+      "loss": 0.3066,
+      "step": 2094
+    },
+    {
+      "epoch": 0.1676,
+      "grad_norm": 1.7278863191604614,
+      "learning_rate": 9.328029425248736e-06,
+      "loss": 0.4222,
+      "step": 2095
+    },
+    {
+      "epoch": 0.16768,
+      "grad_norm": 1.574093222618103,
+      "learning_rate": 9.327399553803047e-06,
+      "loss": 0.3571,
+      "step": 2096
+    },
+    {
+      "epoch": 0.16776,
+      "grad_norm": 1.94347083568573,
+      "learning_rate": 9.326769408577266e-06,
+      "loss": 0.4714,
+      "step": 2097
+    },
+    {
+      "epoch": 0.16784,
+      "grad_norm": 1.8207807540893555,
+      "learning_rate": 9.326138989611265e-06,
+      "loss": 0.3273,
+      "step": 2098
+    },
+    {
+      "epoch": 0.16792,
+      "grad_norm": 1.5437902212142944,
+      "learning_rate": 9.325508296944922e-06,
+      "loss": 0.3561,
+      "step": 2099
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 1.638817310333252,
+      "learning_rate": 9.324877330618143e-06,
+      "loss": 0.3146,
+      "step": 2100
+    },
+    {
+      "epoch": 0.16808,
+      "grad_norm": 1.3537306785583496,
+      "learning_rate": 9.324246090670847e-06,
+      "loss": 0.263,
+      "step": 2101
+    },
+    {
+      "epoch": 0.16816,
+      "grad_norm": 1.4852831363677979,
+      "learning_rate": 9.32361457714297e-06,
+      "loss": 0.2881,
+      "step": 2102
+    },
+    {
+      "epoch": 0.16824,
+      "grad_norm": 1.9877256155014038,
+      "learning_rate": 9.322982790074467e-06,
+      "loss": 0.425,
+      "step": 2103
+    },
+    {
+      "epoch": 0.16832,
+      "grad_norm": 1.4754945039749146,
+      "learning_rate": 9.322350729505305e-06,
+      "loss": 0.2753,
+      "step": 2104
+    },
+    {
+      "epoch": 0.1684,
+      "grad_norm": 1.820770502090454,
+      "learning_rate": 9.321718395475475e-06,
+      "loss": 0.3354,
+      "step": 2105
+    },
+    {
+      "epoch": 0.16848,
+      "grad_norm": 1.3193517923355103,
+      "learning_rate": 9.321085788024983e-06,
+      "loss": 0.2598,
+      "step": 2106
+    },
+    {
+      "epoch": 0.16856,
+      "grad_norm": 1.3211908340454102,
+      "learning_rate": 9.320452907193854e-06,
+      "loss": 0.2607,
+      "step": 2107
+    },
+    {
+      "epoch": 0.16864,
+      "grad_norm": 2.045539140701294,
+      "learning_rate": 9.319819753022123e-06,
+      "loss": 0.3862,
+      "step": 2108
+    },
+    {
+      "epoch": 0.16872,
+      "grad_norm": 1.5828810930252075,
+      "learning_rate": 9.319186325549851e-06,
+      "loss": 0.478,
+      "step": 2109
+    },
+    {
+      "epoch": 0.1688,
+      "grad_norm": 1.4170411825180054,
+      "learning_rate": 9.318552624817114e-06,
+      "loss": 0.323,
+      "step": 2110
+    },
+    {
+      "epoch": 0.16888,
+      "grad_norm": 1.40409255027771,
+      "learning_rate": 9.317918650864e-06,
+      "loss": 0.2817,
+      "step": 2111
+    },
+    {
+      "epoch": 0.16896,
+      "grad_norm": 1.6362119913101196,
+      "learning_rate": 9.317284403730622e-06,
+      "loss": 0.3455,
+      "step": 2112
+    },
+    {
+      "epoch": 0.16904,
+      "grad_norm": 1.4493852853775024,
+      "learning_rate": 9.316649883457104e-06,
+      "loss": 0.3114,
+      "step": 2113
+    },
+    {
+      "epoch": 0.16912,
+      "grad_norm": 1.9049961566925049,
+      "learning_rate": 9.316015090083595e-06,
+      "loss": 0.3715,
+      "step": 2114
+    },
+    {
+      "epoch": 0.1692,
+      "grad_norm": 1.3683356046676636,
+      "learning_rate": 9.315380023650248e-06,
+      "loss": 0.2945,
+      "step": 2115
+    },
+    {
+      "epoch": 0.16928,
+      "grad_norm": 2.296933889389038,
+      "learning_rate": 9.31474468419725e-06,
+      "loss": 0.445,
+      "step": 2116
+    },
+    {
+      "epoch": 0.16936,
+      "grad_norm": 1.487442970275879,
+      "learning_rate": 9.314109071764793e-06,
+      "loss": 0.2976,
+      "step": 2117
+    },
+    {
+      "epoch": 0.16944,
+      "grad_norm": 1.581235647201538,
+      "learning_rate": 9.313473186393087e-06,
+      "loss": 0.3736,
+      "step": 2118
+    },
+    {
+      "epoch": 0.16952,
+      "grad_norm": 1.685365080833435,
+      "learning_rate": 9.312837028122368e-06,
+      "loss": 0.3519,
+      "step": 2119
+    },
+    {
+      "epoch": 0.1696,
+      "grad_norm": 1.5775948762893677,
+      "learning_rate": 9.312200596992879e-06,
+      "loss": 0.3245,
+      "step": 2120
+    },
+    {
+      "epoch": 0.16968,
+      "grad_norm": 1.5085936784744263,
+      "learning_rate": 9.311563893044888e-06,
+      "loss": 0.2838,
+      "step": 2121
+    },
+    {
+      "epoch": 0.16976,
+      "grad_norm": 1.522886037826538,
+      "learning_rate": 9.310926916318677e-06,
+      "loss": 0.3361,
+      "step": 2122
+    },
+    {
+      "epoch": 0.16984,
+      "grad_norm": 1.5970796346664429,
+      "learning_rate": 9.310289666854543e-06,
+      "loss": 0.4171,
+      "step": 2123
+    },
+    {
+      "epoch": 0.16992,
+      "grad_norm": 1.9791122674942017,
+      "learning_rate": 9.309652144692804e-06,
+      "loss": 0.4402,
+      "step": 2124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1077051162719727,
+      "learning_rate": 9.309014349873795e-06,
+      "loss": 0.5385,
+      "step": 2125
+    },
+    {
+      "epoch": 0.17008,
+      "grad_norm": 1.4200516939163208,
+      "learning_rate": 9.308376282437866e-06,
+      "loss": 0.306,
+      "step": 2126
+    },
+    {
+      "epoch": 0.17016,
+      "grad_norm": 1.521388053894043,
+      "learning_rate": 9.307737942425385e-06,
+      "loss": 0.3083,
+      "step": 2127
+    },
+    {
+      "epoch": 0.17024,
+      "grad_norm": 2.1909584999084473,
+      "learning_rate": 9.307099329876736e-06,
+      "loss": 0.3912,
+      "step": 2128
+    },
+    {
+      "epoch": 0.17032,
+      "grad_norm": 1.6918203830718994,
+      "learning_rate": 9.306460444832327e-06,
+      "loss": 0.3992,
+      "step": 2129
+    },
+    {
+      "epoch": 0.1704,
+      "grad_norm": 1.7204002141952515,
+      "learning_rate": 9.305821287332575e-06,
+      "loss": 0.3076,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17048,
+      "grad_norm": 1.634256362915039,
+      "learning_rate": 9.305181857417917e-06,
+      "loss": 0.3281,
+      "step": 2131
+    },
+    {
+      "epoch": 0.17056,
+      "grad_norm": 1.616892695426941,
+      "learning_rate": 9.304542155128806e-06,
+      "loss": 0.3026,
+      "step": 2132
+    },
+    {
+      "epoch": 0.17064,
+      "grad_norm": 1.3723448514938354,
+      "learning_rate": 9.30390218050572e-06,
+      "loss": 0.3969,
+      "step": 2133
+    },
+    {
+      "epoch": 0.17072,
+      "grad_norm": 1.7421725988388062,
+      "learning_rate": 9.303261933589141e-06,
+      "loss": 0.3514,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1708,
+      "grad_norm": 1.6025031805038452,
+      "learning_rate": 9.302621414419577e-06,
+      "loss": 0.3093,
+      "step": 2135
+    },
+    {
+      "epoch": 0.17088,
+      "grad_norm": 1.4779335260391235,
+      "learning_rate": 9.301980623037556e-06,
+      "loss": 0.2883,
+      "step": 2136
+    },
+    {
+      "epoch": 0.17096,
+      "grad_norm": 1.6824702024459839,
+      "learning_rate": 9.301339559483614e-06,
+      "loss": 0.3437,
+      "step": 2137
+    },
+    {
+      "epoch": 0.17104,
+      "grad_norm": 1.7231637239456177,
+      "learning_rate": 9.30069822379831e-06,
+      "loss": 0.5239,
+      "step": 2138
+    },
+    {
+      "epoch": 0.17112,
+      "grad_norm": 1.4851043224334717,
+      "learning_rate": 9.30005661602222e-06,
+      "loss": 0.3925,
+      "step": 2139
+    },
+    {
+      "epoch": 0.1712,
+      "grad_norm": 1.611282229423523,
+      "learning_rate": 9.299414736195936e-06,
+      "loss": 0.316,
+      "step": 2140
+    },
+    {
+      "epoch": 0.17128,
+      "grad_norm": 1.8075153827667236,
+      "learning_rate": 9.298772584360068e-06,
+      "loss": 0.4018,
+      "step": 2141
+    },
+    {
+      "epoch": 0.17136,
+      "grad_norm": 1.6185144186019897,
+      "learning_rate": 9.298130160555241e-06,
+      "loss": 0.3886,
+      "step": 2142
+    },
+    {
+      "epoch": 0.17144,
+      "grad_norm": 1.4245575666427612,
+      "learning_rate": 9.297487464822101e-06,
+      "loss": 0.2893,
+      "step": 2143
+    },
+    {
+      "epoch": 0.17152,
+      "grad_norm": 1.7352980375289917,
+      "learning_rate": 9.296844497201309e-06,
+      "loss": 0.3129,
+      "step": 2144
+    },
+    {
+      "epoch": 0.1716,
+      "grad_norm": 1.7901087999343872,
+      "learning_rate": 9.296201257733542e-06,
+      "loss": 0.4773,
+      "step": 2145
+    },
+    {
+      "epoch": 0.17168,
+      "grad_norm": 1.4795407056808472,
+      "learning_rate": 9.295557746459498e-06,
+      "loss": 0.2931,
+      "step": 2146
+    },
+    {
+      "epoch": 0.17176,
+      "grad_norm": 1.349307894706726,
+      "learning_rate": 9.294913963419887e-06,
+      "loss": 0.2764,
+      "step": 2147
+    },
+    {
+      "epoch": 0.17184,
+      "grad_norm": 1.5399342775344849,
+      "learning_rate": 9.29426990865544e-06,
+      "loss": 0.3619,
+      "step": 2148
+    },
+    {
+      "epoch": 0.17192,
+      "grad_norm": 1.2979024648666382,
+      "learning_rate": 9.293625582206907e-06,
+      "loss": 0.2828,
+      "step": 2149
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 1.999293565750122,
+      "learning_rate": 9.292980984115048e-06,
+      "loss": 0.4908,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17208,
+      "grad_norm": 1.910706877708435,
+      "learning_rate": 9.292336114420645e-06,
+      "loss": 0.3672,
+      "step": 2151
+    },
+    {
+      "epoch": 0.17216,
+      "grad_norm": 1.7465935945510864,
+      "learning_rate": 9.2916909731645e-06,
+      "loss": 0.447,
+      "step": 2152
+    },
+    {
+      "epoch": 0.17224,
+      "grad_norm": 1.3406716585159302,
+      "learning_rate": 9.291045560387428e-06,
+      "loss": 0.3121,
+      "step": 2153
+    },
+    {
+      "epoch": 0.17232,
+      "grad_norm": 1.4294350147247314,
+      "learning_rate": 9.290399876130261e-06,
+      "loss": 0.3287,
+      "step": 2154
+    },
+    {
+      "epoch": 0.1724,
+      "grad_norm": 1.1393554210662842,
+      "learning_rate": 9.289753920433848e-06,
+      "loss": 0.2707,
+      "step": 2155
+    },
+    {
+      "epoch": 0.17248,
+      "grad_norm": 1.8590903282165527,
+      "learning_rate": 9.28910769333906e-06,
+      "loss": 0.4045,
+      "step": 2156
+    },
+    {
+      "epoch": 0.17256,
+      "grad_norm": 1.8060451745986938,
+      "learning_rate": 9.288461194886778e-06,
+      "loss": 0.4324,
+      "step": 2157
+    },
+    {
+      "epoch": 0.17264,
+      "grad_norm": 1.4243879318237305,
+      "learning_rate": 9.287814425117907e-06,
+      "loss": 0.2985,
+      "step": 2158
+    },
+    {
+      "epoch": 0.17272,
+      "grad_norm": 1.8122689723968506,
+      "learning_rate": 9.287167384073364e-06,
+      "loss": 0.4555,
+      "step": 2159
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 1.4533016681671143,
+      "learning_rate": 9.286520071794085e-06,
+      "loss": 0.2583,
+      "step": 2160
+    },
+    {
+      "epoch": 0.17288,
+      "grad_norm": 1.7077553272247314,
+      "learning_rate": 9.285872488321023e-06,
+      "loss": 0.475,
+      "step": 2161
+    },
+    {
+      "epoch": 0.17296,
+      "grad_norm": 1.8486095666885376,
+      "learning_rate": 9.285224633695151e-06,
+      "loss": 0.4598,
+      "step": 2162
+    },
+    {
+      "epoch": 0.17304,
+      "grad_norm": 1.693713903427124,
+      "learning_rate": 9.284576507957454e-06,
+      "loss": 0.3301,
+      "step": 2163
+    },
+    {
+      "epoch": 0.17312,
+      "grad_norm": 1.5107775926589966,
+      "learning_rate": 9.283928111148937e-06,
+      "loss": 0.3226,
+      "step": 2164
+    },
+    {
+      "epoch": 0.1732,
+      "grad_norm": 1.395003318786621,
+      "learning_rate": 9.283279443310623e-06,
+      "loss": 0.3354,
+      "step": 2165
+    },
+    {
+      "epoch": 0.17328,
+      "grad_norm": 1.4116548299789429,
+      "learning_rate": 9.28263050448355e-06,
+      "loss": 0.2977,
+      "step": 2166
+    },
+    {
+      "epoch": 0.17336,
+      "grad_norm": 1.5333601236343384,
+      "learning_rate": 9.281981294708775e-06,
+      "loss": 0.4171,
+      "step": 2167
+    },
+    {
+      "epoch": 0.17344,
+      "grad_norm": 1.510820746421814,
+      "learning_rate": 9.281331814027372e-06,
+      "loss": 0.3506,
+      "step": 2168
+    },
+    {
+      "epoch": 0.17352,
+      "grad_norm": 1.3123133182525635,
+      "learning_rate": 9.28068206248043e-06,
+      "loss": 0.2977,
+      "step": 2169
+    },
+    {
+      "epoch": 0.1736,
+      "grad_norm": 1.606671929359436,
+      "learning_rate": 9.280032040109057e-06,
+      "loss": 0.3311,
+      "step": 2170
+    },
+    {
+      "epoch": 0.17368,
+      "grad_norm": 1.8891907930374146,
+      "learning_rate": 9.279381746954378e-06,
+      "loss": 0.5067,
+      "step": 2171
+    },
+    {
+      "epoch": 0.17376,
+      "grad_norm": 1.3156824111938477,
+      "learning_rate": 9.278731183057533e-06,
+      "loss": 0.3077,
+      "step": 2172
+    },
+    {
+      "epoch": 0.17384,
+      "grad_norm": 1.8756389617919922,
+      "learning_rate": 9.278080348459684e-06,
+      "loss": 0.3977,
+      "step": 2173
+    },
+    {
+      "epoch": 0.17392,
+      "grad_norm": 1.0538018941879272,
+      "learning_rate": 9.277429243202007e-06,
+      "loss": 0.244,
+      "step": 2174
+    },
+    {
+      "epoch": 0.174,
+      "grad_norm": 1.7108577489852905,
+      "learning_rate": 9.276777867325693e-06,
+      "loss": 0.4015,
+      "step": 2175
+    },
+    {
+      "epoch": 0.17408,
+      "grad_norm": 1.6555908918380737,
+      "learning_rate": 9.276126220871952e-06,
+      "loss": 0.4131,
+      "step": 2176
+    },
+    {
+      "epoch": 0.17416,
+      "grad_norm": 1.5826455354690552,
+      "learning_rate": 9.275474303882016e-06,
+      "loss": 0.3234,
+      "step": 2177
+    },
+    {
+      "epoch": 0.17424,
+      "grad_norm": 1.338919758796692,
+      "learning_rate": 9.274822116397124e-06,
+      "loss": 0.3157,
+      "step": 2178
+    },
+    {
+      "epoch": 0.17432,
+      "grad_norm": 1.8238811492919922,
+      "learning_rate": 9.274169658458543e-06,
+      "loss": 0.3202,
+      "step": 2179
+    },
+    {
+      "epoch": 0.1744,
+      "grad_norm": 1.4607664346694946,
+      "learning_rate": 9.273516930107547e-06,
+      "loss": 0.3517,
+      "step": 2180
+    },
+    {
+      "epoch": 0.17448,
+      "grad_norm": 1.8293615579605103,
+      "learning_rate": 9.272863931385434e-06,
+      "loss": 0.3754,
+      "step": 2181
+    },
+    {
+      "epoch": 0.17456,
+      "grad_norm": 2.1700384616851807,
+      "learning_rate": 9.272210662333518e-06,
+      "loss": 0.5072,
+      "step": 2182
+    },
+    {
+      "epoch": 0.17464,
+      "grad_norm": 1.8814268112182617,
+      "learning_rate": 9.27155712299313e-06,
+      "loss": 0.366,
+      "step": 2183
+    },
+    {
+      "epoch": 0.17472,
+      "grad_norm": 1.32944655418396,
+      "learning_rate": 9.270903313405612e-06,
+      "loss": 0.287,
+      "step": 2184
+    },
+    {
+      "epoch": 0.1748,
+      "grad_norm": 1.4097766876220703,
+      "learning_rate": 9.270249233612334e-06,
+      "loss": 0.3646,
+      "step": 2185
+    },
+    {
+      "epoch": 0.17488,
+      "grad_norm": 2.0853865146636963,
+      "learning_rate": 9.269594883654673e-06,
+      "loss": 0.383,
+      "step": 2186
+    },
+    {
+      "epoch": 0.17496,
+      "grad_norm": 2.0160350799560547,
+      "learning_rate": 9.268940263574032e-06,
+      "loss": 0.5086,
+      "step": 2187
+    },
+    {
+      "epoch": 0.17504,
+      "grad_norm": 1.8854643106460571,
+      "learning_rate": 9.268285373411823e-06,
+      "loss": 0.4704,
+      "step": 2188
+    },
+    {
+      "epoch": 0.17512,
+      "grad_norm": 1.7649178504943848,
+      "learning_rate": 9.267630213209482e-06,
+      "loss": 0.3193,
+      "step": 2189
+    },
+    {
+      "epoch": 0.1752,
+      "grad_norm": 1.8833215236663818,
+      "learning_rate": 9.266974783008456e-06,
+      "loss": 0.4477,
+      "step": 2190
+    },
+    {
+      "epoch": 0.17528,
+      "grad_norm": 1.488447666168213,
+      "learning_rate": 9.266319082850212e-06,
+      "loss": 0.3464,
+      "step": 2191
+    },
+    {
+      "epoch": 0.17536,
+      "grad_norm": 1.2886351346969604,
+      "learning_rate": 9.265663112776237e-06,
+      "loss": 0.3003,
+      "step": 2192
+    },
+    {
+      "epoch": 0.17544,
+      "grad_norm": 2.210524320602417,
+      "learning_rate": 9.265006872828028e-06,
+      "loss": 0.4336,
+      "step": 2193
+    },
+    {
+      "epoch": 0.17552,
+      "grad_norm": 1.7764190435409546,
+      "learning_rate": 9.264350363047105e-06,
+      "loss": 0.3219,
+      "step": 2194
+    },
+    {
+      "epoch": 0.1756,
+      "grad_norm": 1.7255445718765259,
+      "learning_rate": 9.263693583475003e-06,
+      "loss": 0.4066,
+      "step": 2195
+    },
+    {
+      "epoch": 0.17568,
+      "grad_norm": 1.7819242477416992,
+      "learning_rate": 9.263036534153276e-06,
+      "loss": 0.3992,
+      "step": 2196
+    },
+    {
+      "epoch": 0.17576,
+      "grad_norm": 1.5859037637710571,
+      "learning_rate": 9.262379215123489e-06,
+      "loss": 0.4164,
+      "step": 2197
+    },
+    {
+      "epoch": 0.17584,
+      "grad_norm": 1.5518395900726318,
+      "learning_rate": 9.261721626427233e-06,
+      "loss": 0.2977,
+      "step": 2198
+    },
+    {
+      "epoch": 0.17592,
+      "grad_norm": 1.689637541770935,
+      "learning_rate": 9.26106376810611e-06,
+      "loss": 0.4156,
+      "step": 2199
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 1.8921533823013306,
+      "learning_rate": 9.260405640201737e-06,
+      "loss": 0.4744,
+      "step": 2200
+    },
+    {
+      "epoch": 0.17608,
+      "grad_norm": 1.4645209312438965,
+      "learning_rate": 9.259747242755757e-06,
+      "loss": 0.3006,
+      "step": 2201
+    },
+    {
+      "epoch": 0.17616,
+      "grad_norm": 1.5621235370635986,
+      "learning_rate": 9.25908857580982e-06,
+      "loss": 0.3133,
+      "step": 2202
+    },
+    {
+      "epoch": 0.17624,
+      "grad_norm": 1.5191184282302856,
+      "learning_rate": 9.258429639405602e-06,
+      "loss": 0.3818,
+      "step": 2203
+    },
+    {
+      "epoch": 0.17632,
+      "grad_norm": 1.3126143217086792,
+      "learning_rate": 9.25777043358479e-06,
+      "loss": 0.2928,
+      "step": 2204
+    },
+    {
+      "epoch": 0.1764,
+      "grad_norm": 1.6664056777954102,
+      "learning_rate": 9.257110958389088e-06,
+      "loss": 0.3457,
+      "step": 2205
+    },
+    {
+      "epoch": 0.17648,
+      "grad_norm": 1.7404066324234009,
+      "learning_rate": 9.25645121386022e-06,
+      "loss": 0.5175,
+      "step": 2206
+    },
+    {
+      "epoch": 0.17656,
+      "grad_norm": 1.5736061334609985,
+      "learning_rate": 9.255791200039925e-06,
+      "loss": 0.3344,
+      "step": 2207
+    },
+    {
+      "epoch": 0.17664,
+      "grad_norm": 1.7708841562271118,
+      "learning_rate": 9.255130916969962e-06,
+      "loss": 0.3816,
+      "step": 2208
+    },
+    {
+      "epoch": 0.17672,
+      "grad_norm": 1.5940806865692139,
+      "learning_rate": 9.254470364692103e-06,
+      "loss": 0.2876,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1768,
+      "grad_norm": 1.2727887630462646,
+      "learning_rate": 9.253809543248139e-06,
+      "loss": 0.3499,
+      "step": 2210
+    },
+    {
+      "epoch": 0.17688,
+      "grad_norm": 1.7963303327560425,
+      "learning_rate": 9.253148452679878e-06,
+      "loss": 0.4021,
+      "step": 2211
+    },
+    {
+      "epoch": 0.17696,
+      "grad_norm": 1.2776786088943481,
+      "learning_rate": 9.252487093029149e-06,
+      "loss": 0.2827,
+      "step": 2212
+    },
+    {
+      "epoch": 0.17704,
+      "grad_norm": 1.625144600868225,
+      "learning_rate": 9.251825464337785e-06,
+      "loss": 0.3478,
+      "step": 2213
+    },
+    {
+      "epoch": 0.17712,
+      "grad_norm": 1.4690876007080078,
+      "learning_rate": 9.251163566647655e-06,
+      "loss": 0.3588,
+      "step": 2214
+    },
+    {
+      "epoch": 0.1772,
+      "grad_norm": 1.4391189813613892,
+      "learning_rate": 9.250501400000628e-06,
+      "loss": 0.3249,
+      "step": 2215
+    },
+    {
+      "epoch": 0.17728,
+      "grad_norm": 1.5366137027740479,
+      "learning_rate": 9.249838964438602e-06,
+      "loss": 0.35,
+      "step": 2216
+    },
+    {
+      "epoch": 0.17736,
+      "grad_norm": 1.5445330142974854,
+      "learning_rate": 9.249176260003482e-06,
+      "loss": 0.3531,
+      "step": 2217
+    },
+    {
+      "epoch": 0.17744,
+      "grad_norm": 1.41277277469635,
+      "learning_rate": 9.248513286737199e-06,
+      "loss": 0.3252,
+      "step": 2218
+    },
+    {
+      "epoch": 0.17752,
+      "grad_norm": 1.4927526712417603,
+      "learning_rate": 9.247850044681698e-06,
+      "loss": 0.3272,
+      "step": 2219
+    },
+    {
+      "epoch": 0.1776,
+      "grad_norm": 1.9397213459014893,
+      "learning_rate": 9.247186533878936e-06,
+      "loss": 0.4147,
+      "step": 2220
+    },
+    {
+      "epoch": 0.17768,
+      "grad_norm": 1.734050989151001,
+      "learning_rate": 9.246522754370893e-06,
+      "loss": 0.3629,
+      "step": 2221
+    },
+    {
+      "epoch": 0.17776,
+      "grad_norm": 1.57426118850708,
+      "learning_rate": 9.245858706199565e-06,
+      "loss": 0.3241,
+      "step": 2222
+    },
+    {
+      "epoch": 0.17784,
+      "grad_norm": 1.3740696907043457,
+      "learning_rate": 9.245194389406961e-06,
+      "loss": 0.2866,
+      "step": 2223
+    },
+    {
+      "epoch": 0.17792,
+      "grad_norm": 1.3346598148345947,
+      "learning_rate": 9.244529804035116e-06,
+      "loss": 0.2516,
+      "step": 2224
+    },
+    {
+      "epoch": 0.178,
+      "grad_norm": 1.5931025743484497,
+      "learning_rate": 9.24386495012607e-06,
+      "loss": 0.4599,
+      "step": 2225
+    },
+    {
+      "epoch": 0.17808,
+      "grad_norm": 1.608982801437378,
+      "learning_rate": 9.24319982772189e-06,
+      "loss": 0.3688,
+      "step": 2226
+    },
+    {
+      "epoch": 0.17816,
+      "grad_norm": 1.6699520349502563,
+      "learning_rate": 9.242534436864654e-06,
+      "loss": 0.3289,
+      "step": 2227
+    },
+    {
+      "epoch": 0.17824,
+      "grad_norm": 1.524670958518982,
+      "learning_rate": 9.24186877759646e-06,
+      "loss": 0.3446,
+      "step": 2228
+    },
+    {
+      "epoch": 0.17832,
+      "grad_norm": 2.181579113006592,
+      "learning_rate": 9.241202849959422e-06,
+      "loss": 0.4242,
+      "step": 2229
+    },
+    {
+      "epoch": 0.1784,
+      "grad_norm": 1.4129284620285034,
+      "learning_rate": 9.240536653995671e-06,
+      "loss": 0.3339,
+      "step": 2230
+    },
+    {
+      "epoch": 0.17848,
+      "grad_norm": 1.5664548873901367,
+      "learning_rate": 9.239870189747355e-06,
+      "loss": 0.3472,
+      "step": 2231
+    },
+    {
+      "epoch": 0.17856,
+      "grad_norm": 2.093280076980591,
+      "learning_rate": 9.239203457256636e-06,
+      "loss": 0.4444,
+      "step": 2232
+    },
+    {
+      "epoch": 0.17864,
+      "grad_norm": 1.6080608367919922,
+      "learning_rate": 9.238536456565702e-06,
+      "loss": 0.4602,
+      "step": 2233
+    },
+    {
+      "epoch": 0.17872,
+      "grad_norm": 1.671555519104004,
+      "learning_rate": 9.237869187716747e-06,
+      "loss": 0.4864,
+      "step": 2234
+    },
+    {
+      "epoch": 0.1788,
+      "grad_norm": 1.5570496320724487,
+      "learning_rate": 9.237201650751987e-06,
+      "loss": 0.4289,
+      "step": 2235
+    },
+    {
+      "epoch": 0.17888,
+      "grad_norm": 1.9579259157180786,
+      "learning_rate": 9.23653384571366e-06,
+      "loss": 0.373,
+      "step": 2236
+    },
+    {
+      "epoch": 0.17896,
+      "grad_norm": 1.2836438417434692,
+      "learning_rate": 9.23586577264401e-06,
+      "loss": 0.2913,
+      "step": 2237
+    },
+    {
+      "epoch": 0.17904,
+      "grad_norm": 1.651989221572876,
+      "learning_rate": 9.235197431585305e-06,
+      "loss": 0.3415,
+      "step": 2238
+    },
+    {
+      "epoch": 0.17912,
+      "grad_norm": 1.5639806985855103,
+      "learning_rate": 9.23452882257983e-06,
+      "loss": 0.4252,
+      "step": 2239
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 1.5605380535125732,
+      "learning_rate": 9.233859945669888e-06,
+      "loss": 0.303,
+      "step": 2240
+    },
+    {
+      "epoch": 0.17928,
+      "grad_norm": 1.6003406047821045,
+      "learning_rate": 9.23319080089779e-06,
+      "loss": 0.3035,
+      "step": 2241
+    },
+    {
+      "epoch": 0.17936,
+      "grad_norm": 1.3260300159454346,
+      "learning_rate": 9.232521388305876e-06,
+      "loss": 0.2764,
+      "step": 2242
+    },
+    {
+      "epoch": 0.17944,
+      "grad_norm": 1.5889723300933838,
+      "learning_rate": 9.231851707936495e-06,
+      "loss": 0.3757,
+      "step": 2243
+    },
+    {
+      "epoch": 0.17952,
+      "grad_norm": 1.5519212484359741,
+      "learning_rate": 9.231181759832017e-06,
+      "loss": 0.3367,
+      "step": 2244
+    },
+    {
+      "epoch": 0.1796,
+      "grad_norm": 1.8289942741394043,
+      "learning_rate": 9.230511544034826e-06,
+      "loss": 0.3846,
+      "step": 2245
+    },
+    {
+      "epoch": 0.17968,
+      "grad_norm": 1.6226454973220825,
+      "learning_rate": 9.229841060587326e-06,
+      "loss": 0.3169,
+      "step": 2246
+    },
+    {
+      "epoch": 0.17976,
+      "grad_norm": 1.7149393558502197,
+      "learning_rate": 9.229170309531934e-06,
+      "loss": 0.3524,
+      "step": 2247
+    },
+    {
+      "epoch": 0.17984,
+      "grad_norm": 1.6193126440048218,
+      "learning_rate": 9.228499290911088e-06,
+      "loss": 0.3369,
+      "step": 2248
+    },
+    {
+      "epoch": 0.17992,
+      "grad_norm": 2.0542750358581543,
+      "learning_rate": 9.22782800476724e-06,
+      "loss": 0.3477,
+      "step": 2249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3201754093170166,
+      "learning_rate": 9.227156451142863e-06,
+      "loss": 0.3066,
+      "step": 2250
+    },
+    {
+      "epoch": 0.18008,
+      "grad_norm": 1.7556428909301758,
+      "learning_rate": 9.226484630080439e-06,
+      "loss": 0.3791,
+      "step": 2251
+    },
+    {
+      "epoch": 0.18016,
+      "grad_norm": 1.8350175619125366,
+      "learning_rate": 9.225812541622474e-06,
+      "loss": 0.3668,
+      "step": 2252
+    },
+    {
+      "epoch": 0.18024,
+      "grad_norm": 1.2991888523101807,
+      "learning_rate": 9.22514018581149e-06,
+      "loss": 0.2949,
+      "step": 2253
+    },
+    {
+      "epoch": 0.18032,
+      "grad_norm": 1.7130646705627441,
+      "learning_rate": 9.224467562690022e-06,
+      "loss": 0.381,
+      "step": 2254
+    },
+    {
+      "epoch": 0.1804,
+      "grad_norm": 1.361568570137024,
+      "learning_rate": 9.22379467230063e-06,
+      "loss": 0.2958,
+      "step": 2255
+    },
+    {
+      "epoch": 0.18048,
+      "grad_norm": 1.8371490240097046,
+      "learning_rate": 9.22312151468588e-06,
+      "loss": 0.3656,
+      "step": 2256
+    },
+    {
+      "epoch": 0.18056,
+      "grad_norm": 1.5410816669464111,
+      "learning_rate": 9.22244808988836e-06,
+      "loss": 0.3638,
+      "step": 2257
+    },
+    {
+      "epoch": 0.18064,
+      "grad_norm": 1.7891284227371216,
+      "learning_rate": 9.22177439795068e-06,
+      "loss": 0.4046,
+      "step": 2258
+    },
+    {
+      "epoch": 0.18072,
+      "grad_norm": 1.0117267370224,
+      "learning_rate": 9.221100438915462e-06,
+      "loss": 0.2083,
+      "step": 2259
+    },
+    {
+      "epoch": 0.1808,
+      "grad_norm": 1.6379694938659668,
+      "learning_rate": 9.22042621282534e-06,
+      "loss": 0.3407,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18088,
+      "grad_norm": 1.8293073177337646,
+      "learning_rate": 9.219751719722974e-06,
+      "loss": 0.3578,
+      "step": 2261
+    },
+    {
+      "epoch": 0.18096,
+      "grad_norm": 1.6874363422393799,
+      "learning_rate": 9.219076959651037e-06,
+      "loss": 0.438,
+      "step": 2262
+    },
+    {
+      "epoch": 0.18104,
+      "grad_norm": 1.3425878286361694,
+      "learning_rate": 9.218401932652217e-06,
+      "loss": 0.2945,
+      "step": 2263
+    },
+    {
+      "epoch": 0.18112,
+      "grad_norm": 1.3295488357543945,
+      "learning_rate": 9.21772663876922e-06,
+      "loss": 0.3008,
+      "step": 2264
+    },
+    {
+      "epoch": 0.1812,
+      "grad_norm": 1.479942798614502,
+      "learning_rate": 9.217051078044773e-06,
+      "loss": 0.3512,
+      "step": 2265
+    },
+    {
+      "epoch": 0.18128,
+      "grad_norm": 1.6754070520401,
+      "learning_rate": 9.216375250521614e-06,
+      "loss": 0.3106,
+      "step": 2266
+    },
+    {
+      "epoch": 0.18136,
+      "grad_norm": 1.2535821199417114,
+      "learning_rate": 9.215699156242501e-06,
+      "loss": 0.298,
+      "step": 2267
+    },
+    {
+      "epoch": 0.18144,
+      "grad_norm": 1.4116896390914917,
+      "learning_rate": 9.215022795250209e-06,
+      "loss": 0.3316,
+      "step": 2268
+    },
+    {
+      "epoch": 0.18152,
+      "grad_norm": 1.704809546470642,
+      "learning_rate": 9.214346167587529e-06,
+      "loss": 0.4056,
+      "step": 2269
+    },
+    {
+      "epoch": 0.1816,
+      "grad_norm": 1.3820918798446655,
+      "learning_rate": 9.213669273297266e-06,
+      "loss": 0.3467,
+      "step": 2270
+    },
+    {
+      "epoch": 0.18168,
+      "grad_norm": 1.1921701431274414,
+      "learning_rate": 9.212992112422248e-06,
+      "loss": 0.277,
+      "step": 2271
+    },
+    {
+      "epoch": 0.18176,
+      "grad_norm": 1.5377929210662842,
+      "learning_rate": 9.212314685005314e-06,
+      "loss": 0.3292,
+      "step": 2272
+    },
+    {
+      "epoch": 0.18184,
+      "grad_norm": 1.6128257513046265,
+      "learning_rate": 9.211636991089328e-06,
+      "loss": 0.3684,
+      "step": 2273
+    },
+    {
+      "epoch": 0.18192,
+      "grad_norm": 2.1811368465423584,
+      "learning_rate": 9.210959030717158e-06,
+      "loss": 0.4364,
+      "step": 2274
+    },
+    {
+      "epoch": 0.182,
+      "grad_norm": 1.9046738147735596,
+      "learning_rate": 9.210280803931702e-06,
+      "loss": 0.3347,
+      "step": 2275
+    },
+    {
+      "epoch": 0.18208,
+      "grad_norm": 1.7492754459381104,
+      "learning_rate": 9.209602310775868e-06,
+      "loss": 0.3851,
+      "step": 2276
+    },
+    {
+      "epoch": 0.18216,
+      "grad_norm": 1.6498044729232788,
+      "learning_rate": 9.208923551292578e-06,
+      "loss": 0.3078,
+      "step": 2277
+    },
+    {
+      "epoch": 0.18224,
+      "grad_norm": 1.6144391298294067,
+      "learning_rate": 9.208244525524782e-06,
+      "loss": 0.4138,
+      "step": 2278
+    },
+    {
+      "epoch": 0.18232,
+      "grad_norm": 1.871065616607666,
+      "learning_rate": 9.207565233515434e-06,
+      "loss": 0.4951,
+      "step": 2279
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 1.9937913417816162,
+      "learning_rate": 9.20688567530751e-06,
+      "loss": 0.4494,
+      "step": 2280
+    },
+    {
+      "epoch": 0.18248,
+      "grad_norm": 1.3968439102172852,
+      "learning_rate": 9.206205850944009e-06,
+      "loss": 0.2798,
+      "step": 2281
+    },
+    {
+      "epoch": 0.18256,
+      "grad_norm": 1.554082989692688,
+      "learning_rate": 9.205525760467937e-06,
+      "loss": 0.3656,
+      "step": 2282
+    },
+    {
+      "epoch": 0.18264,
+      "grad_norm": 2.032961130142212,
+      "learning_rate": 9.204845403922321e-06,
+      "loss": 0.4478,
+      "step": 2283
+    },
+    {
+      "epoch": 0.18272,
+      "grad_norm": 1.621554970741272,
+      "learning_rate": 9.204164781350207e-06,
+      "loss": 0.3502,
+      "step": 2284
+    },
+    {
+      "epoch": 0.1828,
+      "grad_norm": 1.4644954204559326,
+      "learning_rate": 9.203483892794652e-06,
+      "loss": 0.3379,
+      "step": 2285
+    },
+    {
+      "epoch": 0.18288,
+      "grad_norm": 1.294154167175293,
+      "learning_rate": 9.202802738298738e-06,
+      "loss": 0.2404,
+      "step": 2286
+    },
+    {
+      "epoch": 0.18296,
+      "grad_norm": 2.0851991176605225,
+      "learning_rate": 9.202121317905557e-06,
+      "loss": 0.3982,
+      "step": 2287
+    },
+    {
+      "epoch": 0.18304,
+      "grad_norm": 1.1855297088623047,
+      "learning_rate": 9.20143963165822e-06,
+      "loss": 0.2492,
+      "step": 2288
+    },
+    {
+      "epoch": 0.18312,
+      "grad_norm": 1.5839331150054932,
+      "learning_rate": 9.200757679599857e-06,
+      "loss": 0.2824,
+      "step": 2289
+    },
+    {
+      "epoch": 0.1832,
+      "grad_norm": 1.5208096504211426,
+      "learning_rate": 9.20007546177361e-06,
+      "loss": 0.3692,
+      "step": 2290
+    },
+    {
+      "epoch": 0.18328,
+      "grad_norm": 2.2178423404693604,
+      "learning_rate": 9.199392978222644e-06,
+      "loss": 0.436,
+      "step": 2291
+    },
+    {
+      "epoch": 0.18336,
+      "grad_norm": 1.7429447174072266,
+      "learning_rate": 9.198710228990132e-06,
+      "loss": 0.4171,
+      "step": 2292
+    },
+    {
+      "epoch": 0.18344,
+      "grad_norm": 1.6876320838928223,
+      "learning_rate": 9.198027214119275e-06,
+      "loss": 0.4079,
+      "step": 2293
+    },
+    {
+      "epoch": 0.18352,
+      "grad_norm": 2.200977087020874,
+      "learning_rate": 9.197343933653283e-06,
+      "loss": 0.4152,
+      "step": 2294
+    },
+    {
+      "epoch": 0.1836,
+      "grad_norm": 1.5130945444107056,
+      "learning_rate": 9.196660387635384e-06,
+      "loss": 0.2694,
+      "step": 2295
+    },
+    {
+      "epoch": 0.18368,
+      "grad_norm": 1.4931762218475342,
+      "learning_rate": 9.195976576108825e-06,
+      "loss": 0.4342,
+      "step": 2296
+    },
+    {
+      "epoch": 0.18376,
+      "grad_norm": 1.4749504327774048,
+      "learning_rate": 9.195292499116868e-06,
+      "loss": 0.3459,
+      "step": 2297
+    },
+    {
+      "epoch": 0.18384,
+      "grad_norm": 1.4674580097198486,
+      "learning_rate": 9.19460815670279e-06,
+      "loss": 0.3511,
+      "step": 2298
+    },
+    {
+      "epoch": 0.18392,
+      "grad_norm": 1.4509178400039673,
+      "learning_rate": 9.193923548909891e-06,
+      "loss": 0.279,
+      "step": 2299
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 1.370303988456726,
+      "learning_rate": 9.193238675781482e-06,
+      "loss": 0.2597,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18408,
+      "grad_norm": 1.7368767261505127,
+      "learning_rate": 9.19255353736089e-06,
+      "loss": 0.387,
+      "step": 2301
+    },
+    {
+      "epoch": 0.18416,
+      "grad_norm": 1.8166148662567139,
+      "learning_rate": 9.191868133691467e-06,
+      "loss": 0.3739,
+      "step": 2302
+    },
+    {
+      "epoch": 0.18424,
+      "grad_norm": 1.7067928314208984,
+      "learning_rate": 9.191182464816572e-06,
+      "loss": 0.3617,
+      "step": 2303
+    },
+    {
+      "epoch": 0.18432,
+      "grad_norm": 1.0539000034332275,
+      "learning_rate": 9.190496530779587e-06,
+      "loss": 0.2726,
+      "step": 2304
+    },
+    {
+      "epoch": 0.1844,
+      "grad_norm": 1.555287480354309,
+      "learning_rate": 9.189810331623908e-06,
+      "loss": 0.332,
+      "step": 2305
+    },
+    {
+      "epoch": 0.18448,
+      "grad_norm": 1.7620913982391357,
+      "learning_rate": 9.189123867392947e-06,
+      "loss": 0.4206,
+      "step": 2306
+    },
+    {
+      "epoch": 0.18456,
+      "grad_norm": 1.1607882976531982,
+      "learning_rate": 9.188437138130138e-06,
+      "loss": 0.242,
+      "step": 2307
+    },
+    {
+      "epoch": 0.18464,
+      "grad_norm": 1.4528453350067139,
+      "learning_rate": 9.187750143878924e-06,
+      "loss": 0.2846,
+      "step": 2308
+    },
+    {
+      "epoch": 0.18472,
+      "grad_norm": 1.9288702011108398,
+      "learning_rate": 9.187062884682772e-06,
+      "loss": 0.4231,
+      "step": 2309
+    },
+    {
+      "epoch": 0.1848,
+      "grad_norm": 1.4445855617523193,
+      "learning_rate": 9.186375360585159e-06,
+      "loss": 0.3683,
+      "step": 2310
+    },
+    {
+      "epoch": 0.18488,
+      "grad_norm": 1.6469016075134277,
+      "learning_rate": 9.185687571629587e-06,
+      "loss": 0.4188,
+      "step": 2311
+    },
+    {
+      "epoch": 0.18496,
+      "grad_norm": 1.7436647415161133,
+      "learning_rate": 9.184999517859566e-06,
+      "loss": 0.3646,
+      "step": 2312
+    },
+    {
+      "epoch": 0.18504,
+      "grad_norm": 2.044245719909668,
+      "learning_rate": 9.18431119931863e-06,
+      "loss": 0.4894,
+      "step": 2313
+    },
+    {
+      "epoch": 0.18512,
+      "grad_norm": 1.396704077720642,
+      "learning_rate": 9.183622616050323e-06,
+      "loss": 0.3058,
+      "step": 2314
+    },
+    {
+      "epoch": 0.1852,
+      "grad_norm": 1.5335050821304321,
+      "learning_rate": 9.182933768098213e-06,
+      "loss": 0.3617,
+      "step": 2315
+    },
+    {
+      "epoch": 0.18528,
+      "grad_norm": 1.3849726915359497,
+      "learning_rate": 9.18224465550588e-06,
+      "loss": 0.2886,
+      "step": 2316
+    },
+    {
+      "epoch": 0.18536,
+      "grad_norm": 1.3755298852920532,
+      "learning_rate": 9.181555278316921e-06,
+      "loss": 0.2749,
+      "step": 2317
+    },
+    {
+      "epoch": 0.18544,
+      "grad_norm": 1.4264538288116455,
+      "learning_rate": 9.180865636574951e-06,
+      "loss": 0.369,
+      "step": 2318
+    },
+    {
+      "epoch": 0.18552,
+      "grad_norm": 1.5905702114105225,
+      "learning_rate": 9.180175730323602e-06,
+      "loss": 0.3422,
+      "step": 2319
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 2.0537757873535156,
+      "learning_rate": 9.17948555960652e-06,
+      "loss": 0.3696,
+      "step": 2320
+    },
+    {
+      "epoch": 0.18568,
+      "grad_norm": 1.9900577068328857,
+      "learning_rate": 9.178795124467372e-06,
+      "loss": 0.5154,
+      "step": 2321
+    },
+    {
+      "epoch": 0.18576,
+      "grad_norm": 1.8288344144821167,
+      "learning_rate": 9.17810442494984e-06,
+      "loss": 0.3817,
+      "step": 2322
+    },
+    {
+      "epoch": 0.18584,
+      "grad_norm": 2.499269723892212,
+      "learning_rate": 9.17741346109762e-06,
+      "loss": 0.4914,
+      "step": 2323
+    },
+    {
+      "epoch": 0.18592,
+      "grad_norm": 1.8195888996124268,
+      "learning_rate": 9.176722232954426e-06,
+      "loss": 0.3946,
+      "step": 2324
+    },
+    {
+      "epoch": 0.186,
+      "grad_norm": 2.1742544174194336,
+      "learning_rate": 9.176030740563994e-06,
+      "loss": 0.465,
+      "step": 2325
+    },
+    {
+      "epoch": 0.18608,
+      "grad_norm": 2.0490312576293945,
+      "learning_rate": 9.175338983970071e-06,
+      "loss": 0.3448,
+      "step": 2326
+    },
+    {
+      "epoch": 0.18616,
+      "grad_norm": 1.3807387351989746,
+      "learning_rate": 9.17464696321642e-06,
+      "loss": 0.3351,
+      "step": 2327
+    },
+    {
+      "epoch": 0.18624,
+      "grad_norm": 1.89095938205719,
+      "learning_rate": 9.173954678346823e-06,
+      "loss": 0.336,
+      "step": 2328
+    },
+    {
+      "epoch": 0.18632,
+      "grad_norm": 1.5240287780761719,
+      "learning_rate": 9.173262129405081e-06,
+      "loss": 0.3218,
+      "step": 2329
+    },
+    {
+      "epoch": 0.1864,
+      "grad_norm": 1.6663240194320679,
+      "learning_rate": 9.172569316435008e-06,
+      "loss": 0.3771,
+      "step": 2330
+    },
+    {
+      "epoch": 0.18648,
+      "grad_norm": 1.4801119565963745,
+      "learning_rate": 9.171876239480435e-06,
+      "loss": 0.3371,
+      "step": 2331
+    },
+    {
+      "epoch": 0.18656,
+      "grad_norm": 1.4088448286056519,
+      "learning_rate": 9.17118289858521e-06,
+      "loss": 0.351,
+      "step": 2332
+    },
+    {
+      "epoch": 0.18664,
+      "grad_norm": 1.6157206296920776,
+      "learning_rate": 9.170489293793203e-06,
+      "loss": 0.3401,
+      "step": 2333
+    },
+    {
+      "epoch": 0.18672,
+      "grad_norm": 1.7214024066925049,
+      "learning_rate": 9.16979542514829e-06,
+      "loss": 0.3182,
+      "step": 2334
+    },
+    {
+      "epoch": 0.1868,
+      "grad_norm": 1.4012519121170044,
+      "learning_rate": 9.169101292694376e-06,
+      "loss": 0.3879,
+      "step": 2335
+    },
+    {
+      "epoch": 0.18688,
+      "grad_norm": 1.5938009023666382,
+      "learning_rate": 9.168406896475372e-06,
+      "loss": 0.3759,
+      "step": 2336
+    },
+    {
+      "epoch": 0.18696,
+      "grad_norm": 1.4147008657455444,
+      "learning_rate": 9.167712236535209e-06,
+      "loss": 0.2956,
+      "step": 2337
+    },
+    {
+      "epoch": 0.18704,
+      "grad_norm": 1.8677122592926025,
+      "learning_rate": 9.16701731291784e-06,
+      "loss": 0.4634,
+      "step": 2338
+    },
+    {
+      "epoch": 0.18712,
+      "grad_norm": 1.4628612995147705,
+      "learning_rate": 9.166322125667229e-06,
+      "loss": 0.2652,
+      "step": 2339
+    },
+    {
+      "epoch": 0.1872,
+      "grad_norm": 1.6055920124053955,
+      "learning_rate": 9.165626674827355e-06,
+      "loss": 0.4288,
+      "step": 2340
+    },
+    {
+      "epoch": 0.18728,
+      "grad_norm": 1.2751909494400024,
+      "learning_rate": 9.164930960442222e-06,
+      "loss": 0.249,
+      "step": 2341
+    },
+    {
+      "epoch": 0.18736,
+      "grad_norm": 1.565497875213623,
+      "learning_rate": 9.164234982555841e-06,
+      "loss": 0.3797,
+      "step": 2342
+    },
+    {
+      "epoch": 0.18744,
+      "grad_norm": 1.4904969930648804,
+      "learning_rate": 9.163538741212247e-06,
+      "loss": 0.3217,
+      "step": 2343
+    },
+    {
+      "epoch": 0.18752,
+      "grad_norm": 1.5334428548812866,
+      "learning_rate": 9.16284223645549e-06,
+      "loss": 0.3089,
+      "step": 2344
+    },
+    {
+      "epoch": 0.1876,
+      "grad_norm": 1.3491618633270264,
+      "learning_rate": 9.16214546832963e-06,
+      "loss": 0.3263,
+      "step": 2345
+    },
+    {
+      "epoch": 0.18768,
+      "grad_norm": 1.757620930671692,
+      "learning_rate": 9.161448436878755e-06,
+      "loss": 0.3392,
+      "step": 2346
+    },
+    {
+      "epoch": 0.18776,
+      "grad_norm": 1.591558814048767,
+      "learning_rate": 9.160751142146962e-06,
+      "loss": 0.3605,
+      "step": 2347
+    },
+    {
+      "epoch": 0.18784,
+      "grad_norm": 1.4992103576660156,
+      "learning_rate": 9.160053584178365e-06,
+      "loss": 0.3162,
+      "step": 2348
+    },
+    {
+      "epoch": 0.18792,
+      "grad_norm": 1.847265601158142,
+      "learning_rate": 9.159355763017099e-06,
+      "loss": 0.4105,
+      "step": 2349
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 1.6127371788024902,
+      "learning_rate": 9.15865767870731e-06,
+      "loss": 0.345,
+      "step": 2350
+    },
+    {
+      "epoch": 0.18808,
+      "grad_norm": 1.6684445142745972,
+      "learning_rate": 9.157959331293165e-06,
+      "loss": 0.3925,
+      "step": 2351
+    },
+    {
+      "epoch": 0.18816,
+      "grad_norm": 1.3808155059814453,
+      "learning_rate": 9.157260720818843e-06,
+      "loss": 0.3334,
+      "step": 2352
+    },
+    {
+      "epoch": 0.18824,
+      "grad_norm": 1.176571249961853,
+      "learning_rate": 9.15656184732855e-06,
+      "loss": 0.2693,
+      "step": 2353
+    },
+    {
+      "epoch": 0.18832,
+      "grad_norm": 1.5287673473358154,
+      "learning_rate": 9.155862710866493e-06,
+      "loss": 0.3012,
+      "step": 2354
+    },
+    {
+      "epoch": 0.1884,
+      "grad_norm": 2.0192978382110596,
+      "learning_rate": 9.15516331147691e-06,
+      "loss": 0.5501,
+      "step": 2355
+    },
+    {
+      "epoch": 0.18848,
+      "grad_norm": 1.7251996994018555,
+      "learning_rate": 9.154463649204046e-06,
+      "loss": 0.4322,
+      "step": 2356
+    },
+    {
+      "epoch": 0.18856,
+      "grad_norm": 1.4777201414108276,
+      "learning_rate": 9.153763724092169e-06,
+      "loss": 0.3243,
+      "step": 2357
+    },
+    {
+      "epoch": 0.18864,
+      "grad_norm": 1.573727011680603,
+      "learning_rate": 9.15306353618556e-06,
+      "loss": 0.3615,
+      "step": 2358
+    },
+    {
+      "epoch": 0.18872,
+      "grad_norm": 1.976193904876709,
+      "learning_rate": 9.152363085528516e-06,
+      "loss": 0.3525,
+      "step": 2359
+    },
+    {
+      "epoch": 0.1888,
+      "grad_norm": 1.72635817527771,
+      "learning_rate": 9.151662372165354e-06,
+      "loss": 0.357,
+      "step": 2360
+    },
+    {
+      "epoch": 0.18888,
+      "grad_norm": 1.5015325546264648,
+      "learning_rate": 9.150961396140405e-06,
+      "loss": 0.3098,
+      "step": 2361
+    },
+    {
+      "epoch": 0.18896,
+      "grad_norm": 1.459358811378479,
+      "learning_rate": 9.15026015749802e-06,
+      "loss": 0.3554,
+      "step": 2362
+    },
+    {
+      "epoch": 0.18904,
+      "grad_norm": 1.7636947631835938,
+      "learning_rate": 9.149558656282557e-06,
+      "loss": 0.4678,
+      "step": 2363
+    },
+    {
+      "epoch": 0.18912,
+      "grad_norm": 1.615004539489746,
+      "learning_rate": 9.148856892538406e-06,
+      "loss": 0.3788,
+      "step": 2364
+    },
+    {
+      "epoch": 0.1892,
+      "grad_norm": 1.5762476921081543,
+      "learning_rate": 9.14815486630996e-06,
+      "loss": 0.3102,
+      "step": 2365
+    },
+    {
+      "epoch": 0.18928,
+      "grad_norm": 1.6073107719421387,
+      "learning_rate": 9.147452577641635e-06,
+      "loss": 0.3938,
+      "step": 2366
+    },
+    {
+      "epoch": 0.18936,
+      "grad_norm": 1.3423787355422974,
+      "learning_rate": 9.146750026577865e-06,
+      "loss": 0.2977,
+      "step": 2367
+    },
+    {
+      "epoch": 0.18944,
+      "grad_norm": 1.3909164667129517,
+      "learning_rate": 9.146047213163094e-06,
+      "loss": 0.256,
+      "step": 2368
+    },
+    {
+      "epoch": 0.18952,
+      "grad_norm": 1.6559034585952759,
+      "learning_rate": 9.14534413744179e-06,
+      "loss": 0.3731,
+      "step": 2369
+    },
+    {
+      "epoch": 0.1896,
+      "grad_norm": 1.2788715362548828,
+      "learning_rate": 9.14464079945843e-06,
+      "loss": 0.2787,
+      "step": 2370
+    },
+    {
+      "epoch": 0.18968,
+      "grad_norm": 1.9036865234375,
+      "learning_rate": 9.143937199257518e-06,
+      "loss": 0.3465,
+      "step": 2371
+    },
+    {
+      "epoch": 0.18976,
+      "grad_norm": 1.7238149642944336,
+      "learning_rate": 9.143233336883563e-06,
+      "loss": 0.3865,
+      "step": 2372
+    },
+    {
+      "epoch": 0.18984,
+      "grad_norm": 1.2368091344833374,
+      "learning_rate": 9.142529212381098e-06,
+      "loss": 0.2267,
+      "step": 2373
+    },
+    {
+      "epoch": 0.18992,
+      "grad_norm": 1.7011198997497559,
+      "learning_rate": 9.141824825794672e-06,
+      "loss": 0.3792,
+      "step": 2374
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9078706502914429,
+      "learning_rate": 9.141120177168846e-06,
+      "loss": 0.3637,
+      "step": 2375
+    },
+    {
+      "epoch": 0.19008,
+      "grad_norm": 2.0864484310150146,
+      "learning_rate": 9.140415266548203e-06,
+      "loss": 0.4524,
+      "step": 2376
+    },
+    {
+      "epoch": 0.19016,
+      "grad_norm": 1.4254752397537231,
+      "learning_rate": 9.139710093977342e-06,
+      "loss": 0.2662,
+      "step": 2377
+    },
+    {
+      "epoch": 0.19024,
+      "grad_norm": 1.6436759233474731,
+      "learning_rate": 9.139004659500874e-06,
+      "loss": 0.3361,
+      "step": 2378
+    },
+    {
+      "epoch": 0.19032,
+      "grad_norm": 2.105926752090454,
+      "learning_rate": 9.138298963163429e-06,
+      "loss": 0.3339,
+      "step": 2379
+    },
+    {
+      "epoch": 0.1904,
+      "grad_norm": 1.5537519454956055,
+      "learning_rate": 9.137593005009657e-06,
+      "loss": 0.3319,
+      "step": 2380
+    },
+    {
+      "epoch": 0.19048,
+      "grad_norm": 2.272524833679199,
+      "learning_rate": 9.136886785084217e-06,
+      "loss": 0.4195,
+      "step": 2381
+    },
+    {
+      "epoch": 0.19056,
+      "grad_norm": 1.2556356191635132,
+      "learning_rate": 9.136180303431797e-06,
+      "loss": 0.3392,
+      "step": 2382
+    },
+    {
+      "epoch": 0.19064,
+      "grad_norm": 1.5762921571731567,
+      "learning_rate": 9.135473560097086e-06,
+      "loss": 0.3597,
+      "step": 2383
+    },
+    {
+      "epoch": 0.19072,
+      "grad_norm": 1.3007557392120361,
+      "learning_rate": 9.1347665551248e-06,
+      "loss": 0.2813,
+      "step": 2384
+    },
+    {
+      "epoch": 0.1908,
+      "grad_norm": 1.295035719871521,
+      "learning_rate": 9.134059288559669e-06,
+      "loss": 0.3211,
+      "step": 2385
+    },
+    {
+      "epoch": 0.19088,
+      "grad_norm": 1.5391051769256592,
+      "learning_rate": 9.13335176044644e-06,
+      "loss": 0.3058,
+      "step": 2386
+    },
+    {
+      "epoch": 0.19096,
+      "grad_norm": 1.6031337976455688,
+      "learning_rate": 9.132643970829876e-06,
+      "loss": 0.3478,
+      "step": 2387
+    },
+    {
+      "epoch": 0.19104,
+      "grad_norm": 1.572896122932434,
+      "learning_rate": 9.131935919754755e-06,
+      "loss": 0.4024,
+      "step": 2388
+    },
+    {
+      "epoch": 0.19112,
+      "grad_norm": 1.4550867080688477,
+      "learning_rate": 9.131227607265874e-06,
+      "loss": 0.2725,
+      "step": 2389
+    },
+    {
+      "epoch": 0.1912,
+      "grad_norm": 2.0746469497680664,
+      "learning_rate": 9.130519033408045e-06,
+      "loss": 0.4477,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19128,
+      "grad_norm": 1.5469022989273071,
+      "learning_rate": 9.129810198226099e-06,
+      "loss": 0.3219,
+      "step": 2391
+    },
+    {
+      "epoch": 0.19136,
+      "grad_norm": 1.4501341581344604,
+      "learning_rate": 9.12910110176488e-06,
+      "loss": 0.4573,
+      "step": 2392
+    },
+    {
+      "epoch": 0.19144,
+      "grad_norm": 1.1760106086730957,
+      "learning_rate": 9.128391744069248e-06,
+      "loss": 0.2149,
+      "step": 2393
+    },
+    {
+      "epoch": 0.19152,
+      "grad_norm": 1.8944004774093628,
+      "learning_rate": 9.127682125184085e-06,
+      "loss": 0.4243,
+      "step": 2394
+    },
+    {
+      "epoch": 0.1916,
+      "grad_norm": 1.382332682609558,
+      "learning_rate": 9.126972245154287e-06,
+      "loss": 0.2722,
+      "step": 2395
+    },
+    {
+      "epoch": 0.19168,
+      "grad_norm": 1.5882385969161987,
+      "learning_rate": 9.12626210402476e-06,
+      "loss": 0.4348,
+      "step": 2396
+    },
+    {
+      "epoch": 0.19176,
+      "grad_norm": 1.7748380899429321,
+      "learning_rate": 9.12555170184044e-06,
+      "loss": 0.3083,
+      "step": 2397
+    },
+    {
+      "epoch": 0.19184,
+      "grad_norm": 1.7226957082748413,
+      "learning_rate": 9.124841038646268e-06,
+      "loss": 0.3241,
+      "step": 2398
+    },
+    {
+      "epoch": 0.19192,
+      "grad_norm": 1.642120599746704,
+      "learning_rate": 9.124130114487203e-06,
+      "loss": 0.3724,
+      "step": 2399
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 1.916398048400879,
+      "learning_rate": 9.123418929408225e-06,
+      "loss": 0.4925,
+      "step": 2400
+    },
+    {
+      "epoch": 0.19208,
+      "grad_norm": 1.2868858575820923,
+      "learning_rate": 9.12270748345433e-06,
+      "loss": 0.3048,
+      "step": 2401
+    },
+    {
+      "epoch": 0.19216,
+      "grad_norm": 1.6964247226715088,
+      "learning_rate": 9.121995776670527e-06,
+      "loss": 0.3746,
+      "step": 2402
+    },
+    {
+      "epoch": 0.19224,
+      "grad_norm": 1.710965871810913,
+      "learning_rate": 9.121283809101843e-06,
+      "loss": 0.4557,
+      "step": 2403
+    },
+    {
+      "epoch": 0.19232,
+      "grad_norm": 1.4680863618850708,
+      "learning_rate": 9.120571580793322e-06,
+      "loss": 0.303,
+      "step": 2404
+    },
+    {
+      "epoch": 0.1924,
+      "grad_norm": 1.802053451538086,
+      "learning_rate": 9.119859091790025e-06,
+      "loss": 0.3215,
+      "step": 2405
+    },
+    {
+      "epoch": 0.19248,
+      "grad_norm": 1.676236867904663,
+      "learning_rate": 9.119146342137029e-06,
+      "loss": 0.4425,
+      "step": 2406
+    },
+    {
+      "epoch": 0.19256,
+      "grad_norm": 1.6125231981277466,
+      "learning_rate": 9.118433331879424e-06,
+      "loss": 0.3079,
+      "step": 2407
+    },
+    {
+      "epoch": 0.19264,
+      "grad_norm": 1.313044548034668,
+      "learning_rate": 9.117720061062324e-06,
+      "loss": 0.3317,
+      "step": 2408
+    },
+    {
+      "epoch": 0.19272,
+      "grad_norm": 1.5678797960281372,
+      "learning_rate": 9.117006529730853e-06,
+      "loss": 0.311,
+      "step": 2409
+    },
+    {
+      "epoch": 0.1928,
+      "grad_norm": 1.5004569292068481,
+      "learning_rate": 9.116292737930156e-06,
+      "loss": 0.3152,
+      "step": 2410
+    },
+    {
+      "epoch": 0.19288,
+      "grad_norm": 1.86920964717865,
+      "learning_rate": 9.115578685705391e-06,
+      "loss": 0.3537,
+      "step": 2411
+    },
+    {
+      "epoch": 0.19296,
+      "grad_norm": 1.8753119707107544,
+      "learning_rate": 9.114864373101733e-06,
+      "loss": 0.4428,
+      "step": 2412
+    },
+    {
+      "epoch": 0.19304,
+      "grad_norm": 1.3418453931808472,
+      "learning_rate": 9.114149800164372e-06,
+      "loss": 0.2649,
+      "step": 2413
+    },
+    {
+      "epoch": 0.19312,
+      "grad_norm": 1.6189359426498413,
+      "learning_rate": 9.11343496693852e-06,
+      "loss": 0.3543,
+      "step": 2414
+    },
+    {
+      "epoch": 0.1932,
+      "grad_norm": 1.6323224306106567,
+      "learning_rate": 9.112719873469403e-06,
+      "loss": 0.3309,
+      "step": 2415
+    },
+    {
+      "epoch": 0.19328,
+      "grad_norm": 1.9282023906707764,
+      "learning_rate": 9.11200451980226e-06,
+      "loss": 0.3991,
+      "step": 2416
+    },
+    {
+      "epoch": 0.19336,
+      "grad_norm": 1.5333116054534912,
+      "learning_rate": 9.11128890598235e-06,
+      "loss": 0.3514,
+      "step": 2417
+    },
+    {
+      "epoch": 0.19344,
+      "grad_norm": 1.4084820747375488,
+      "learning_rate": 9.110573032054947e-06,
+      "loss": 0.3691,
+      "step": 2418
+    },
+    {
+      "epoch": 0.19352,
+      "grad_norm": 1.6476877927780151,
+      "learning_rate": 9.109856898065343e-06,
+      "loss": 0.3209,
+      "step": 2419
+    },
+    {
+      "epoch": 0.1936,
+      "grad_norm": 1.7577415704727173,
+      "learning_rate": 9.109140504058843e-06,
+      "loss": 0.3974,
+      "step": 2420
+    },
+    {
+      "epoch": 0.19368,
+      "grad_norm": 1.454849123954773,
+      "learning_rate": 9.108423850080774e-06,
+      "loss": 0.2629,
+      "step": 2421
+    },
+    {
+      "epoch": 0.19376,
+      "grad_norm": 1.6696566343307495,
+      "learning_rate": 9.107706936176474e-06,
+      "loss": 0.3151,
+      "step": 2422
+    },
+    {
+      "epoch": 0.19384,
+      "grad_norm": 2.0093162059783936,
+      "learning_rate": 9.1069897623913e-06,
+      "loss": 0.3786,
+      "step": 2423
+    },
+    {
+      "epoch": 0.19392,
+      "grad_norm": 1.7907012701034546,
+      "learning_rate": 9.106272328770627e-06,
+      "loss": 0.3346,
+      "step": 2424
+    },
+    {
+      "epoch": 0.194,
+      "grad_norm": 1.5353763103485107,
+      "learning_rate": 9.105554635359843e-06,
+      "loss": 0.3275,
+      "step": 2425
+    },
+    {
+      "epoch": 0.19408,
+      "grad_norm": 1.6136126518249512,
+      "learning_rate": 9.104836682204354e-06,
+      "loss": 0.3316,
+      "step": 2426
+    },
+    {
+      "epoch": 0.19416,
+      "grad_norm": 1.877288579940796,
+      "learning_rate": 9.104118469349585e-06,
+      "loss": 0.3735,
+      "step": 2427
+    },
+    {
+      "epoch": 0.19424,
+      "grad_norm": 1.869982361793518,
+      "learning_rate": 9.103399996840972e-06,
+      "loss": 0.4767,
+      "step": 2428
+    },
+    {
+      "epoch": 0.19432,
+      "grad_norm": 1.6566518545150757,
+      "learning_rate": 9.102681264723969e-06,
+      "loss": 0.3472,
+      "step": 2429
+    },
+    {
+      "epoch": 0.1944,
+      "grad_norm": 1.631504774093628,
+      "learning_rate": 9.101962273044053e-06,
+      "loss": 0.3334,
+      "step": 2430
+    },
+    {
+      "epoch": 0.19448,
+      "grad_norm": 1.7339836359024048,
+      "learning_rate": 9.101243021846705e-06,
+      "loss": 0.4071,
+      "step": 2431
+    },
+    {
+      "epoch": 0.19456,
+      "grad_norm": 1.6289665699005127,
+      "learning_rate": 9.10052351117744e-06,
+      "loss": 0.4196,
+      "step": 2432
+    },
+    {
+      "epoch": 0.19464,
+      "grad_norm": 1.76044762134552,
+      "learning_rate": 9.099803741081767e-06,
+      "loss": 0.4187,
+      "step": 2433
+    },
+    {
+      "epoch": 0.19472,
+      "grad_norm": 1.4212559461593628,
+      "learning_rate": 9.099083711605233e-06,
+      "loss": 0.3407,
+      "step": 2434
+    },
+    {
+      "epoch": 0.1948,
+      "grad_norm": 1.439171314239502,
+      "learning_rate": 9.098363422793387e-06,
+      "loss": 0.2902,
+      "step": 2435
+    },
+    {
+      "epoch": 0.19488,
+      "grad_norm": 1.539538025856018,
+      "learning_rate": 9.0976428746918e-06,
+      "loss": 0.3686,
+      "step": 2436
+    },
+    {
+      "epoch": 0.19496,
+      "grad_norm": 1.2634087800979614,
+      "learning_rate": 9.09692206734606e-06,
+      "loss": 0.2648,
+      "step": 2437
+    },
+    {
+      "epoch": 0.19504,
+      "grad_norm": 1.2132208347320557,
+      "learning_rate": 9.096201000801768e-06,
+      "loss": 0.2667,
+      "step": 2438
+    },
+    {
+      "epoch": 0.19512,
+      "grad_norm": 1.5921686887741089,
+      "learning_rate": 9.095479675104543e-06,
+      "loss": 0.3618,
+      "step": 2439
+    },
+    {
+      "epoch": 0.1952,
+      "grad_norm": 1.5952152013778687,
+      "learning_rate": 9.094758090300026e-06,
+      "loss": 0.3662,
+      "step": 2440
+    },
+    {
+      "epoch": 0.19528,
+      "grad_norm": 1.3034418821334839,
+      "learning_rate": 9.094036246433863e-06,
+      "loss": 0.2761,
+      "step": 2441
+    },
+    {
+      "epoch": 0.19536,
+      "grad_norm": 1.6235853433609009,
+      "learning_rate": 9.093314143551728e-06,
+      "loss": 0.4082,
+      "step": 2442
+    },
+    {
+      "epoch": 0.19544,
+      "grad_norm": 1.4469619989395142,
+      "learning_rate": 9.092591781699302e-06,
+      "loss": 0.3711,
+      "step": 2443
+    },
+    {
+      "epoch": 0.19552,
+      "grad_norm": 1.574036717414856,
+      "learning_rate": 9.09186916092229e-06,
+      "loss": 0.399,
+      "step": 2444
+    },
+    {
+      "epoch": 0.1956,
+      "grad_norm": 1.363416314125061,
+      "learning_rate": 9.091146281266403e-06,
+      "loss": 0.3059,
+      "step": 2445
+    },
+    {
+      "epoch": 0.19568,
+      "grad_norm": 1.4807307720184326,
+      "learning_rate": 9.090423142777383e-06,
+      "loss": 0.3728,
+      "step": 2446
+    },
+    {
+      "epoch": 0.19576,
+      "grad_norm": 1.6184738874435425,
+      "learning_rate": 9.089699745500977e-06,
+      "loss": 0.3375,
+      "step": 2447
+    },
+    {
+      "epoch": 0.19584,
+      "grad_norm": 1.5262713432312012,
+      "learning_rate": 9.08897608948295e-06,
+      "loss": 0.3389,
+      "step": 2448
+    },
+    {
+      "epoch": 0.19592,
+      "grad_norm": 1.2660731077194214,
+      "learning_rate": 9.088252174769092e-06,
+      "loss": 0.2845,
+      "step": 2449
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 2.0074665546417236,
+      "learning_rate": 9.087528001405194e-06,
+      "loss": 0.3709,
+      "step": 2450
+    },
+    {
+      "epoch": 0.19608,
+      "grad_norm": 1.678610920906067,
+      "learning_rate": 9.08680356943708e-06,
+      "loss": 0.3842,
+      "step": 2451
+    },
+    {
+      "epoch": 0.19616,
+      "grad_norm": 1.604011058807373,
+      "learning_rate": 9.086078878910576e-06,
+      "loss": 0.305,
+      "step": 2452
+    },
+    {
+      "epoch": 0.19624,
+      "grad_norm": 1.7445040941238403,
+      "learning_rate": 9.085353929871534e-06,
+      "loss": 0.3833,
+      "step": 2453
+    },
+    {
+      "epoch": 0.19632,
+      "grad_norm": 1.6903281211853027,
+      "learning_rate": 9.084628722365817e-06,
+      "loss": 0.3579,
+      "step": 2454
+    },
+    {
+      "epoch": 0.1964,
+      "grad_norm": 1.587699294090271,
+      "learning_rate": 9.08390325643931e-06,
+      "loss": 0.324,
+      "step": 2455
+    },
+    {
+      "epoch": 0.19648,
+      "grad_norm": 1.681836724281311,
+      "learning_rate": 9.083177532137909e-06,
+      "loss": 0.3006,
+      "step": 2456
+    },
+    {
+      "epoch": 0.19656,
+      "grad_norm": 1.259889006614685,
+      "learning_rate": 9.082451549507528e-06,
+      "loss": 0.3757,
+      "step": 2457
+    },
+    {
+      "epoch": 0.19664,
+      "grad_norm": 1.4303218126296997,
+      "learning_rate": 9.081725308594096e-06,
+      "loss": 0.3122,
+      "step": 2458
+    },
+    {
+      "epoch": 0.19672,
+      "grad_norm": 1.956416130065918,
+      "learning_rate": 9.080998809443563e-06,
+      "loss": 0.5083,
+      "step": 2459
+    },
+    {
+      "epoch": 0.1968,
+      "grad_norm": 1.0173094272613525,
+      "learning_rate": 9.080272052101888e-06,
+      "loss": 0.204,
+      "step": 2460
+    },
+    {
+      "epoch": 0.19688,
+      "grad_norm": 1.6032788753509521,
+      "learning_rate": 9.079545036615054e-06,
+      "loss": 0.4081,
+      "step": 2461
+    },
+    {
+      "epoch": 0.19696,
+      "grad_norm": 1.881056547164917,
+      "learning_rate": 9.078817763029054e-06,
+      "loss": 0.4511,
+      "step": 2462
+    },
+    {
+      "epoch": 0.19704,
+      "grad_norm": 1.7417298555374146,
+      "learning_rate": 9.078090231389904e-06,
+      "loss": 0.307,
+      "step": 2463
+    },
+    {
+      "epoch": 0.19712,
+      "grad_norm": 1.7411683797836304,
+      "learning_rate": 9.077362441743632e-06,
+      "loss": 0.3968,
+      "step": 2464
+    },
+    {
+      "epoch": 0.1972,
+      "grad_norm": 1.9373418092727661,
+      "learning_rate": 9.076634394136279e-06,
+      "loss": 0.5019,
+      "step": 2465
+    },
+    {
+      "epoch": 0.19728,
+      "grad_norm": 1.9146742820739746,
+      "learning_rate": 9.075906088613909e-06,
+      "loss": 0.421,
+      "step": 2466
+    },
+    {
+      "epoch": 0.19736,
+      "grad_norm": 1.7911179065704346,
+      "learning_rate": 9.075177525222597e-06,
+      "loss": 0.3551,
+      "step": 2467
+    },
+    {
+      "epoch": 0.19744,
+      "grad_norm": 2.1474406719207764,
+      "learning_rate": 9.074448704008441e-06,
+      "loss": 0.4482,
+      "step": 2468
+    },
+    {
+      "epoch": 0.19752,
+      "grad_norm": 1.610866665840149,
+      "learning_rate": 9.073719625017548e-06,
+      "loss": 0.4025,
+      "step": 2469
+    },
+    {
+      "epoch": 0.1976,
+      "grad_norm": 1.5338889360427856,
+      "learning_rate": 9.072990288296044e-06,
+      "loss": 0.3913,
+      "step": 2470
+    },
+    {
+      "epoch": 0.19768,
+      "grad_norm": 1.8097115755081177,
+      "learning_rate": 9.072260693890073e-06,
+      "loss": 0.4809,
+      "step": 2471
+    },
+    {
+      "epoch": 0.19776,
+      "grad_norm": 1.3825757503509521,
+      "learning_rate": 9.071530841845794e-06,
+      "loss": 0.3793,
+      "step": 2472
+    },
+    {
+      "epoch": 0.19784,
+      "grad_norm": 1.5398048162460327,
+      "learning_rate": 9.070800732209382e-06,
+      "loss": 0.3685,
+      "step": 2473
+    },
+    {
+      "epoch": 0.19792,
+      "grad_norm": 1.705269694328308,
+      "learning_rate": 9.070070365027029e-06,
+      "loss": 0.5052,
+      "step": 2474
+    },
+    {
+      "epoch": 0.198,
+      "grad_norm": 1.5767617225646973,
+      "learning_rate": 9.069339740344943e-06,
+      "loss": 0.3937,
+      "step": 2475
+    },
+    {
+      "epoch": 0.19808,
+      "grad_norm": 1.8536933660507202,
+      "learning_rate": 9.068608858209347e-06,
+      "loss": 0.5109,
+      "step": 2476
+    },
+    {
+      "epoch": 0.19816,
+      "grad_norm": 1.6715956926345825,
+      "learning_rate": 9.067877718666482e-06,
+      "loss": 0.3558,
+      "step": 2477
+    },
+    {
+      "epoch": 0.19824,
+      "grad_norm": 1.6765884160995483,
+      "learning_rate": 9.067146321762603e-06,
+      "loss": 0.3681,
+      "step": 2478
+    },
+    {
+      "epoch": 0.19832,
+      "grad_norm": 1.7231676578521729,
+      "learning_rate": 9.066414667543988e-06,
+      "loss": 0.3274,
+      "step": 2479
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 1.4541349411010742,
+      "learning_rate": 9.065682756056922e-06,
+      "loss": 0.3247,
+      "step": 2480
+    },
+    {
+      "epoch": 0.19848,
+      "grad_norm": 1.4403581619262695,
+      "learning_rate": 9.064950587347711e-06,
+      "loss": 0.4417,
+      "step": 2481
+    },
+    {
+      "epoch": 0.19856,
+      "grad_norm": 1.7160850763320923,
+      "learning_rate": 9.06421816146268e-06,
+      "loss": 0.3751,
+      "step": 2482
+    },
+    {
+      "epoch": 0.19864,
+      "grad_norm": 1.7260886430740356,
+      "learning_rate": 9.063485478448164e-06,
+      "loss": 0.3181,
+      "step": 2483
+    },
+    {
+      "epoch": 0.19872,
+      "grad_norm": 2.1682159900665283,
+      "learning_rate": 9.062752538350517e-06,
+      "loss": 0.4809,
+      "step": 2484
+    },
+    {
+      "epoch": 0.1988,
+      "grad_norm": 1.7218812704086304,
+      "learning_rate": 9.062019341216112e-06,
+      "loss": 0.3777,
+      "step": 2485
+    },
+    {
+      "epoch": 0.19888,
+      "grad_norm": 1.914481520652771,
+      "learning_rate": 9.061285887091334e-06,
+      "loss": 0.4097,
+      "step": 2486
+    },
+    {
+      "epoch": 0.19896,
+      "grad_norm": 1.4902898073196411,
+      "learning_rate": 9.060552176022587e-06,
+      "loss": 0.3261,
+      "step": 2487
+    },
+    {
+      "epoch": 0.19904,
+      "grad_norm": 1.306275725364685,
+      "learning_rate": 9.059818208056293e-06,
+      "loss": 0.2747,
+      "step": 2488
+    },
+    {
+      "epoch": 0.19912,
+      "grad_norm": 1.6835832595825195,
+      "learning_rate": 9.059083983238882e-06,
+      "loss": 0.3671,
+      "step": 2489
+    },
+    {
+      "epoch": 0.1992,
+      "grad_norm": 1.5236090421676636,
+      "learning_rate": 9.05834950161681e-06,
+      "loss": 0.359,
+      "step": 2490
+    },
+    {
+      "epoch": 0.19928,
+      "grad_norm": 1.5683189630508423,
+      "learning_rate": 9.057614763236545e-06,
+      "loss": 0.3818,
+      "step": 2491
+    },
+    {
+      "epoch": 0.19936,
+      "grad_norm": 1.9629745483398438,
+      "learning_rate": 9.056879768144572e-06,
+      "loss": 0.4597,
+      "step": 2492
+    },
+    {
+      "epoch": 0.19944,
+      "grad_norm": 1.8202420473098755,
+      "learning_rate": 9.056144516387387e-06,
+      "loss": 0.3821,
+      "step": 2493
+    },
+    {
+      "epoch": 0.19952,
+      "grad_norm": 2.0477683544158936,
+      "learning_rate": 9.055409008011513e-06,
+      "loss": 0.5704,
+      "step": 2494
+    },
+    {
+      "epoch": 0.1996,
+      "grad_norm": 1.5534504652023315,
+      "learning_rate": 9.05467324306348e-06,
+      "loss": 0.2988,
+      "step": 2495
+    },
+    {
+      "epoch": 0.19968,
+      "grad_norm": 1.8985226154327393,
+      "learning_rate": 9.053937221589837e-06,
+      "loss": 0.4126,
+      "step": 2496
+    },
+    {
+      "epoch": 0.19976,
+      "grad_norm": 1.7892488241195679,
+      "learning_rate": 9.05320094363715e-06,
+      "loss": 0.4152,
+      "step": 2497
+    },
+    {
+      "epoch": 0.19984,
+      "grad_norm": 1.563428282737732,
+      "learning_rate": 9.052464409252003e-06,
+      "loss": 0.3499,
+      "step": 2498
+    },
+    {
+      "epoch": 0.19992,
+      "grad_norm": 1.7737313508987427,
+      "learning_rate": 9.051727618480992e-06,
+      "loss": 0.4627,
+      "step": 2499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.461615800857544,
+      "learning_rate": 9.050990571370731e-06,
+      "loss": 0.3934,
+      "step": 2500
+    },
+    {
+      "epoch": 0.20008,
+      "grad_norm": 1.1548340320587158,
+      "learning_rate": 9.050253267967852e-06,
+      "loss": 0.2507,
+      "step": 2501
+    },
+    {
+      "epoch": 0.20016,
+      "grad_norm": 1.709346055984497,
+      "learning_rate": 9.049515708319001e-06,
+      "loss": 0.3237,
+      "step": 2502
+    },
+    {
+      "epoch": 0.20024,
+      "grad_norm": 1.4969236850738525,
+      "learning_rate": 9.048777892470841e-06,
+      "loss": 0.2986,
+      "step": 2503
+    },
+    {
+      "epoch": 0.20032,
+      "grad_norm": 1.687313199043274,
+      "learning_rate": 9.048039820470049e-06,
+      "loss": 0.3521,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2004,
+      "grad_norm": 1.6837893724441528,
+      "learning_rate": 9.047301492363325e-06,
+      "loss": 0.376,
+      "step": 2505
+    },
+    {
+      "epoch": 0.20048,
+      "grad_norm": 1.6461684703826904,
+      "learning_rate": 9.046562908197376e-06,
+      "loss": 0.3472,
+      "step": 2506
+    },
+    {
+      "epoch": 0.20056,
+      "grad_norm": 1.2622382640838623,
+      "learning_rate": 9.045824068018934e-06,
+      "loss": 0.259,
+      "step": 2507
+    },
+    {
+      "epoch": 0.20064,
+      "grad_norm": 1.7232669591903687,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.4805,
+      "step": 2508
+    },
+    {
+      "epoch": 0.20072,
+      "grad_norm": 1.287067174911499,
+      "learning_rate": 9.044345619811552e-06,
+      "loss": 0.2888,
+      "step": 2509
+    },
+    {
+      "epoch": 0.2008,
+      "grad_norm": 1.4580618143081665,
+      "learning_rate": 9.04360601187615e-06,
+      "loss": 0.3821,
+      "step": 2510
+    },
+    {
+      "epoch": 0.20088,
+      "grad_norm": 1.244520902633667,
+      "learning_rate": 9.042866148115325e-06,
+      "loss": 0.276,
+      "step": 2511
+    },
+    {
+      "epoch": 0.20096,
+      "grad_norm": 1.3449219465255737,
+      "learning_rate": 9.042126028575889e-06,
+      "loss": 0.2915,
+      "step": 2512
+    },
+    {
+      "epoch": 0.20104,
+      "grad_norm": 1.5387424230575562,
+      "learning_rate": 9.041385653304664e-06,
+      "loss": 0.3284,
+      "step": 2513
+    },
+    {
+      "epoch": 0.20112,
+      "grad_norm": 2.1138205528259277,
+      "learning_rate": 9.04064502234849e-06,
+      "loss": 0.4097,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2012,
+      "grad_norm": 1.5106236934661865,
+      "learning_rate": 9.039904135754225e-06,
+      "loss": 0.3061,
+      "step": 2515
+    },
+    {
+      "epoch": 0.20128,
+      "grad_norm": 1.434637427330017,
+      "learning_rate": 9.039162993568743e-06,
+      "loss": 0.3543,
+      "step": 2516
+    },
+    {
+      "epoch": 0.20136,
+      "grad_norm": 1.5437572002410889,
+      "learning_rate": 9.038421595838934e-06,
+      "loss": 0.3407,
+      "step": 2517
+    },
+    {
+      "epoch": 0.20144,
+      "grad_norm": 1.3405263423919678,
+      "learning_rate": 9.037679942611704e-06,
+      "loss": 0.2682,
+      "step": 2518
+    },
+    {
+      "epoch": 0.20152,
+      "grad_norm": 1.4590741395950317,
+      "learning_rate": 9.036938033933973e-06,
+      "loss": 0.3135,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 1.43552565574646,
+      "learning_rate": 9.03619586985268e-06,
+      "loss": 0.2935,
+      "step": 2520
+    },
+    {
+      "epoch": 0.20168,
+      "grad_norm": 1.7763866186141968,
+      "learning_rate": 9.035453450414779e-06,
+      "loss": 0.3757,
+      "step": 2521
+    },
+    {
+      "epoch": 0.20176,
+      "grad_norm": 1.8434540033340454,
+      "learning_rate": 9.034710775667242e-06,
+      "loss": 0.4078,
+      "step": 2522
+    },
+    {
+      "epoch": 0.20184,
+      "grad_norm": 1.806249737739563,
+      "learning_rate": 9.033967845657054e-06,
+      "loss": 0.4245,
+      "step": 2523
+    },
+    {
+      "epoch": 0.20192,
+      "grad_norm": 1.451335072517395,
+      "learning_rate": 9.033224660431219e-06,
+      "loss": 0.3327,
+      "step": 2524
+    },
+    {
+      "epoch": 0.202,
+      "grad_norm": 1.4611728191375732,
+      "learning_rate": 9.032481220036754e-06,
+      "loss": 0.3051,
+      "step": 2525
+    },
+    {
+      "epoch": 0.20208,
+      "grad_norm": 1.524274230003357,
+      "learning_rate": 9.031737524520697e-06,
+      "loss": 0.3235,
+      "step": 2526
+    },
+    {
+      "epoch": 0.20216,
+      "grad_norm": 1.7024296522140503,
+      "learning_rate": 9.030993573930094e-06,
+      "loss": 0.3785,
+      "step": 2527
+    },
+    {
+      "epoch": 0.20224,
+      "grad_norm": 0.9294398427009583,
+      "learning_rate": 9.030249368312015e-06,
+      "loss": 0.1819,
+      "step": 2528
+    },
+    {
+      "epoch": 0.20232,
+      "grad_norm": 1.809221625328064,
+      "learning_rate": 9.029504907713547e-06,
+      "loss": 0.3686,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2024,
+      "grad_norm": 1.8182200193405151,
+      "learning_rate": 9.028760192181785e-06,
+      "loss": 0.3476,
+      "step": 2530
+    },
+    {
+      "epoch": 0.20248,
+      "grad_norm": 1.5775073766708374,
+      "learning_rate": 9.028015221763844e-06,
+      "loss": 0.3117,
+      "step": 2531
+    },
+    {
+      "epoch": 0.20256,
+      "grad_norm": 1.9079387187957764,
+      "learning_rate": 9.02726999650686e-06,
+      "loss": 0.4509,
+      "step": 2532
+    },
+    {
+      "epoch": 0.20264,
+      "grad_norm": 1.451041579246521,
+      "learning_rate": 9.026524516457977e-06,
+      "loss": 0.3606,
+      "step": 2533
+    },
+    {
+      "epoch": 0.20272,
+      "grad_norm": 1.1723933219909668,
+      "learning_rate": 9.025778781664361e-06,
+      "loss": 0.2418,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2028,
+      "grad_norm": 1.693428635597229,
+      "learning_rate": 9.025032792173193e-06,
+      "loss": 0.3352,
+      "step": 2535
+    },
+    {
+      "epoch": 0.20288,
+      "grad_norm": 1.216301441192627,
+      "learning_rate": 9.024286548031666e-06,
+      "loss": 0.334,
+      "step": 2536
+    },
+    {
+      "epoch": 0.20296,
+      "grad_norm": 1.6559041738510132,
+      "learning_rate": 9.023540049286996e-06,
+      "loss": 0.3532,
+      "step": 2537
+    },
+    {
+      "epoch": 0.20304,
+      "grad_norm": 1.55605149269104,
+      "learning_rate": 9.02279329598641e-06,
+      "loss": 0.3522,
+      "step": 2538
+    },
+    {
+      "epoch": 0.20312,
+      "grad_norm": 1.8087921142578125,
+      "learning_rate": 9.022046288177153e-06,
+      "loss": 0.3933,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2032,
+      "grad_norm": 1.6193270683288574,
+      "learning_rate": 9.021299025906482e-06,
+      "loss": 0.3471,
+      "step": 2540
+    },
+    {
+      "epoch": 0.20328,
+      "grad_norm": 1.2611744403839111,
+      "learning_rate": 9.02055150922168e-06,
+      "loss": 0.2635,
+      "step": 2541
+    },
+    {
+      "epoch": 0.20336,
+      "grad_norm": 1.7183457612991333,
+      "learning_rate": 9.019803738170036e-06,
+      "loss": 0.3218,
+      "step": 2542
+    },
+    {
+      "epoch": 0.20344,
+      "grad_norm": 1.3315821886062622,
+      "learning_rate": 9.01905571279886e-06,
+      "loss": 0.2472,
+      "step": 2543
+    },
+    {
+      "epoch": 0.20352,
+      "grad_norm": 1.7462687492370605,
+      "learning_rate": 9.018307433155477e-06,
+      "loss": 0.3837,
+      "step": 2544
+    },
+    {
+      "epoch": 0.2036,
+      "grad_norm": 2.0826213359832764,
+      "learning_rate": 9.017558899287226e-06,
+      "loss": 0.4974,
+      "step": 2545
+    },
+    {
+      "epoch": 0.20368,
+      "grad_norm": 1.5876009464263916,
+      "learning_rate": 9.01681011124147e-06,
+      "loss": 0.3448,
+      "step": 2546
+    },
+    {
+      "epoch": 0.20376,
+      "grad_norm": 1.6092311143875122,
+      "learning_rate": 9.016061069065576e-06,
+      "loss": 0.2966,
+      "step": 2547
+    },
+    {
+      "epoch": 0.20384,
+      "grad_norm": 1.5083656311035156,
+      "learning_rate": 9.015311772806937e-06,
+      "loss": 0.3039,
+      "step": 2548
+    },
+    {
+      "epoch": 0.20392,
+      "grad_norm": 1.6449142694473267,
+      "learning_rate": 9.014562222512954e-06,
+      "loss": 0.3372,
+      "step": 2549
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 1.5135823488235474,
+      "learning_rate": 9.013812418231055e-06,
+      "loss": 0.2915,
+      "step": 2550
+    },
+    {
+      "epoch": 0.20408,
+      "grad_norm": 2.0886871814727783,
+      "learning_rate": 9.013062360008675e-06,
+      "loss": 0.3916,
+      "step": 2551
+    },
+    {
+      "epoch": 0.20416,
+      "grad_norm": 1.2576608657836914,
+      "learning_rate": 9.012312047893265e-06,
+      "loss": 0.2639,
+      "step": 2552
+    },
+    {
+      "epoch": 0.20424,
+      "grad_norm": 1.7450748682022095,
+      "learning_rate": 9.011561481932301e-06,
+      "loss": 0.3559,
+      "step": 2553
+    },
+    {
+      "epoch": 0.20432,
+      "grad_norm": 1.6265859603881836,
+      "learning_rate": 9.010810662173262e-06,
+      "loss": 0.3201,
+      "step": 2554
+    },
+    {
+      "epoch": 0.2044,
+      "grad_norm": 1.57313072681427,
+      "learning_rate": 9.010059588663651e-06,
+      "loss": 0.4484,
+      "step": 2555
+    },
+    {
+      "epoch": 0.20448,
+      "grad_norm": 1.5142831802368164,
+      "learning_rate": 9.00930826145099e-06,
+      "loss": 0.3172,
+      "step": 2556
+    },
+    {
+      "epoch": 0.20456,
+      "grad_norm": 1.7187414169311523,
+      "learning_rate": 9.00855668058281e-06,
+      "loss": 0.3237,
+      "step": 2557
+    },
+    {
+      "epoch": 0.20464,
+      "grad_norm": 1.5038609504699707,
+      "learning_rate": 9.007804846106662e-06,
+      "loss": 0.3068,
+      "step": 2558
+    },
+    {
+      "epoch": 0.20472,
+      "grad_norm": 1.3688982725143433,
+      "learning_rate": 9.007052758070111e-06,
+      "loss": 0.2904,
+      "step": 2559
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 1.5706552267074585,
+      "learning_rate": 9.00630041652074e-06,
+      "loss": 0.3624,
+      "step": 2560
+    },
+    {
+      "epoch": 0.20488,
+      "grad_norm": 1.5629266500473022,
+      "learning_rate": 9.005547821506145e-06,
+      "loss": 0.3069,
+      "step": 2561
+    },
+    {
+      "epoch": 0.20496,
+      "grad_norm": 1.3994823694229126,
+      "learning_rate": 9.004794973073943e-06,
+      "loss": 0.2774,
+      "step": 2562
+    },
+    {
+      "epoch": 0.20504,
+      "grad_norm": 1.459088921546936,
+      "learning_rate": 9.004041871271763e-06,
+      "loss": 0.3553,
+      "step": 2563
+    },
+    {
+      "epoch": 0.20512,
+      "grad_norm": 1.7317893505096436,
+      "learning_rate": 9.003288516147253e-06,
+      "loss": 0.3853,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2052,
+      "grad_norm": 2.0651421546936035,
+      "learning_rate": 9.002534907748071e-06,
+      "loss": 0.6532,
+      "step": 2565
+    },
+    {
+      "epoch": 0.20528,
+      "grad_norm": 1.663630485534668,
+      "learning_rate": 9.0017810461219e-06,
+      "loss": 0.3526,
+      "step": 2566
+    },
+    {
+      "epoch": 0.20536,
+      "grad_norm": 1.511698842048645,
+      "learning_rate": 9.00102693131643e-06,
+      "loss": 0.2987,
+      "step": 2567
+    },
+    {
+      "epoch": 0.20544,
+      "grad_norm": 1.929552674293518,
+      "learning_rate": 9.000272563379375e-06,
+      "loss": 0.4855,
+      "step": 2568
+    },
+    {
+      "epoch": 0.20552,
+      "grad_norm": 1.6176848411560059,
+      "learning_rate": 8.99951794235846e-06,
+      "loss": 0.3513,
+      "step": 2569
+    },
+    {
+      "epoch": 0.2056,
+      "grad_norm": 1.647544264793396,
+      "learning_rate": 8.998763068301428e-06,
+      "loss": 0.4105,
+      "step": 2570
+    },
+    {
+      "epoch": 0.20568,
+      "grad_norm": 1.4232338666915894,
+      "learning_rate": 8.998007941256035e-06,
+      "loss": 0.3615,
+      "step": 2571
+    },
+    {
+      "epoch": 0.20576,
+      "grad_norm": 1.769791841506958,
+      "learning_rate": 8.997252561270058e-06,
+      "loss": 0.3636,
+      "step": 2572
+    },
+    {
+      "epoch": 0.20584,
+      "grad_norm": 1.6207672357559204,
+      "learning_rate": 8.996496928391285e-06,
+      "loss": 0.3445,
+      "step": 2573
+    },
+    {
+      "epoch": 0.20592,
+      "grad_norm": 1.8386582136154175,
+      "learning_rate": 8.995741042667524e-06,
+      "loss": 0.3672,
+      "step": 2574
+    },
+    {
+      "epoch": 0.206,
+      "grad_norm": 1.6057279109954834,
+      "learning_rate": 8.994984904146599e-06,
+      "loss": 0.344,
+      "step": 2575
+    },
+    {
+      "epoch": 0.20608,
+      "grad_norm": 1.5600706338882446,
+      "learning_rate": 8.994228512876345e-06,
+      "loss": 0.346,
+      "step": 2576
+    },
+    {
+      "epoch": 0.20616,
+      "grad_norm": 1.798862338066101,
+      "learning_rate": 8.993471868904617e-06,
+      "loss": 0.3838,
+      "step": 2577
+    },
+    {
+      "epoch": 0.20624,
+      "grad_norm": 1.6833524703979492,
+      "learning_rate": 8.992714972279285e-06,
+      "loss": 0.4041,
+      "step": 2578
+    },
+    {
+      "epoch": 0.20632,
+      "grad_norm": 1.3995176553726196,
+      "learning_rate": 8.99195782304824e-06,
+      "loss": 0.3753,
+      "step": 2579
+    },
+    {
+      "epoch": 0.2064,
+      "grad_norm": 1.242598295211792,
+      "learning_rate": 8.991200421259378e-06,
+      "loss": 0.2893,
+      "step": 2580
+    },
+    {
+      "epoch": 0.20648,
+      "grad_norm": 1.356940746307373,
+      "learning_rate": 8.990442766960622e-06,
+      "loss": 0.2868,
+      "step": 2581
+    },
+    {
+      "epoch": 0.20656,
+      "grad_norm": 1.8150062561035156,
+      "learning_rate": 8.989684860199903e-06,
+      "loss": 0.4231,
+      "step": 2582
+    },
+    {
+      "epoch": 0.20664,
+      "grad_norm": 1.3813666105270386,
+      "learning_rate": 8.988926701025171e-06,
+      "loss": 0.33,
+      "step": 2583
+    },
+    {
+      "epoch": 0.20672,
+      "grad_norm": 1.7560433149337769,
+      "learning_rate": 8.988168289484396e-06,
+      "loss": 0.34,
+      "step": 2584
+    },
+    {
+      "epoch": 0.2068,
+      "grad_norm": 1.6420555114746094,
+      "learning_rate": 8.987409625625556e-06,
+      "loss": 0.3671,
+      "step": 2585
+    },
+    {
+      "epoch": 0.20688,
+      "grad_norm": 1.5787367820739746,
+      "learning_rate": 8.986650709496652e-06,
+      "loss": 0.3619,
+      "step": 2586
+    },
+    {
+      "epoch": 0.20696,
+      "grad_norm": 1.6930251121520996,
+      "learning_rate": 8.985891541145696e-06,
+      "loss": 0.3442,
+      "step": 2587
+    },
+    {
+      "epoch": 0.20704,
+      "grad_norm": 1.488715410232544,
+      "learning_rate": 8.98513212062072e-06,
+      "loss": 0.4164,
+      "step": 2588
+    },
+    {
+      "epoch": 0.20712,
+      "grad_norm": 1.5280767679214478,
+      "learning_rate": 8.98437244796977e-06,
+      "loss": 0.2627,
+      "step": 2589
+    },
+    {
+      "epoch": 0.2072,
+      "grad_norm": 1.939414381980896,
+      "learning_rate": 8.983612523240903e-06,
+      "loss": 0.3893,
+      "step": 2590
+    },
+    {
+      "epoch": 0.20728,
+      "grad_norm": 1.3220232725143433,
+      "learning_rate": 8.982852346482205e-06,
+      "loss": 0.3041,
+      "step": 2591
+    },
+    {
+      "epoch": 0.20736,
+      "grad_norm": 1.585889458656311,
+      "learning_rate": 8.982091917741764e-06,
+      "loss": 0.3324,
+      "step": 2592
+    },
+    {
+      "epoch": 0.20744,
+      "grad_norm": 1.9471094608306885,
+      "learning_rate": 8.981331237067691e-06,
+      "loss": 0.4718,
+      "step": 2593
+    },
+    {
+      "epoch": 0.20752,
+      "grad_norm": 1.783488392829895,
+      "learning_rate": 8.980570304508114e-06,
+      "loss": 0.4395,
+      "step": 2594
+    },
+    {
+      "epoch": 0.2076,
+      "grad_norm": 1.2568445205688477,
+      "learning_rate": 8.97980912011117e-06,
+      "loss": 0.2889,
+      "step": 2595
+    },
+    {
+      "epoch": 0.20768,
+      "grad_norm": 1.5438101291656494,
+      "learning_rate": 8.979047683925022e-06,
+      "loss": 0.4005,
+      "step": 2596
+    },
+    {
+      "epoch": 0.20776,
+      "grad_norm": 1.58918297290802,
+      "learning_rate": 8.978285995997839e-06,
+      "loss": 0.2668,
+      "step": 2597
+    },
+    {
+      "epoch": 0.20784,
+      "grad_norm": 1.4388453960418701,
+      "learning_rate": 8.977524056377814e-06,
+      "loss": 0.2873,
+      "step": 2598
+    },
+    {
+      "epoch": 0.20792,
+      "grad_norm": 2.0037522315979004,
+      "learning_rate": 8.97676186511315e-06,
+      "loss": 0.3798,
+      "step": 2599
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 1.6372380256652832,
+      "learning_rate": 8.975999422252071e-06,
+      "loss": 0.3634,
+      "step": 2600
+    },
+    {
+      "epoch": 0.20808,
+      "grad_norm": 1.2802543640136719,
+      "learning_rate": 8.97523672784281e-06,
+      "loss": 0.3416,
+      "step": 2601
+    },
+    {
+      "epoch": 0.20816,
+      "grad_norm": 1.6268035173416138,
+      "learning_rate": 8.974473781933623e-06,
+      "loss": 0.3532,
+      "step": 2602
+    },
+    {
+      "epoch": 0.20824,
+      "grad_norm": 1.604090690612793,
+      "learning_rate": 8.97371058457278e-06,
+      "loss": 0.3794,
+      "step": 2603
+    },
+    {
+      "epoch": 0.20832,
+      "grad_norm": 1.903481125831604,
+      "learning_rate": 8.97294713580856e-06,
+      "loss": 0.4268,
+      "step": 2604
+    },
+    {
+      "epoch": 0.2084,
+      "grad_norm": 1.351362705230713,
+      "learning_rate": 8.972183435689273e-06,
+      "loss": 0.2761,
+      "step": 2605
+    },
+    {
+      "epoch": 0.20848,
+      "grad_norm": 1.6017656326293945,
+      "learning_rate": 8.97141948426323e-06,
+      "loss": 0.3673,
+      "step": 2606
+    },
+    {
+      "epoch": 0.20856,
+      "grad_norm": 1.882969856262207,
+      "learning_rate": 8.970655281578762e-06,
+      "loss": 0.3843,
+      "step": 2607
+    },
+    {
+      "epoch": 0.20864,
+      "grad_norm": 2.003662347793579,
+      "learning_rate": 8.969890827684222e-06,
+      "loss": 0.4264,
+      "step": 2608
+    },
+    {
+      "epoch": 0.20872,
+      "grad_norm": 1.6131091117858887,
+      "learning_rate": 8.969126122627973e-06,
+      "loss": 0.3647,
+      "step": 2609
+    },
+    {
+      "epoch": 0.2088,
+      "grad_norm": 1.922102451324463,
+      "learning_rate": 8.968361166458395e-06,
+      "loss": 0.3897,
+      "step": 2610
+    },
+    {
+      "epoch": 0.20888,
+      "grad_norm": 1.5826460123062134,
+      "learning_rate": 8.967595959223882e-06,
+      "loss": 0.3342,
+      "step": 2611
+    },
+    {
+      "epoch": 0.20896,
+      "grad_norm": 1.6929792165756226,
+      "learning_rate": 8.966830500972852e-06,
+      "loss": 0.4384,
+      "step": 2612
+    },
+    {
+      "epoch": 0.20904,
+      "grad_norm": 1.5595133304595947,
+      "learning_rate": 8.966064791753727e-06,
+      "loss": 0.3149,
+      "step": 2613
+    },
+    {
+      "epoch": 0.20912,
+      "grad_norm": 1.4859492778778076,
+      "learning_rate": 8.965298831614952e-06,
+      "loss": 0.2991,
+      "step": 2614
+    },
+    {
+      "epoch": 0.2092,
+      "grad_norm": 2.3546998500823975,
+      "learning_rate": 8.96453262060499e-06,
+      "loss": 0.6323,
+      "step": 2615
+    },
+    {
+      "epoch": 0.20928,
+      "grad_norm": 1.4650485515594482,
+      "learning_rate": 8.963766158772314e-06,
+      "loss": 0.2931,
+      "step": 2616
+    },
+    {
+      "epoch": 0.20936,
+      "grad_norm": 1.8523716926574707,
+      "learning_rate": 8.962999446165417e-06,
+      "loss": 0.3536,
+      "step": 2617
+    },
+    {
+      "epoch": 0.20944,
+      "grad_norm": 1.4977569580078125,
+      "learning_rate": 8.962232482832803e-06,
+      "loss": 0.356,
+      "step": 2618
+    },
+    {
+      "epoch": 0.20952,
+      "grad_norm": 1.681130051612854,
+      "learning_rate": 8.961465268822997e-06,
+      "loss": 0.2998,
+      "step": 2619
+    },
+    {
+      "epoch": 0.2096,
+      "grad_norm": 1.994041919708252,
+      "learning_rate": 8.960697804184541e-06,
+      "loss": 0.4286,
+      "step": 2620
+    },
+    {
+      "epoch": 0.20968,
+      "grad_norm": 1.5844112634658813,
+      "learning_rate": 8.959930088965987e-06,
+      "loss": 0.4088,
+      "step": 2621
+    },
+    {
+      "epoch": 0.20976,
+      "grad_norm": 1.7651921510696411,
+      "learning_rate": 8.959162123215906e-06,
+      "loss": 0.4029,
+      "step": 2622
+    },
+    {
+      "epoch": 0.20984,
+      "grad_norm": 1.6215308904647827,
+      "learning_rate": 8.958393906982885e-06,
+      "loss": 0.4116,
+      "step": 2623
+    },
+    {
+      "epoch": 0.20992,
+      "grad_norm": 1.4633090496063232,
+      "learning_rate": 8.957625440315524e-06,
+      "loss": 0.4603,
+      "step": 2624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4941697120666504,
+      "learning_rate": 8.956856723262445e-06,
+      "loss": 0.3767,
+      "step": 2625
+    },
+    {
+      "epoch": 0.21008,
+      "grad_norm": 2.0093963146209717,
+      "learning_rate": 8.956087755872283e-06,
+      "loss": 0.3451,
+      "step": 2626
+    },
+    {
+      "epoch": 0.21016,
+      "grad_norm": 1.8792312145233154,
+      "learning_rate": 8.955318538193684e-06,
+      "loss": 0.3411,
+      "step": 2627
+    },
+    {
+      "epoch": 0.21024,
+      "grad_norm": 1.4969063997268677,
+      "learning_rate": 8.954549070275316e-06,
+      "loss": 0.331,
+      "step": 2628
+    },
+    {
+      "epoch": 0.21032,
+      "grad_norm": 1.5657908916473389,
+      "learning_rate": 8.953779352165859e-06,
+      "loss": 0.4029,
+      "step": 2629
+    },
+    {
+      "epoch": 0.2104,
+      "grad_norm": 1.5333744287490845,
+      "learning_rate": 8.953009383914012e-06,
+      "loss": 0.2624,
+      "step": 2630
+    },
+    {
+      "epoch": 0.21048,
+      "grad_norm": 1.338257908821106,
+      "learning_rate": 8.95223916556849e-06,
+      "loss": 0.3209,
+      "step": 2631
+    },
+    {
+      "epoch": 0.21056,
+      "grad_norm": 1.6348015069961548,
+      "learning_rate": 8.95146869717802e-06,
+      "loss": 0.3488,
+      "step": 2632
+    },
+    {
+      "epoch": 0.21064,
+      "grad_norm": 1.488411784172058,
+      "learning_rate": 8.950697978791345e-06,
+      "loss": 0.3626,
+      "step": 2633
+    },
+    {
+      "epoch": 0.21072,
+      "grad_norm": 1.8199793100357056,
+      "learning_rate": 8.94992701045723e-06,
+      "loss": 0.3665,
+      "step": 2634
+    },
+    {
+      "epoch": 0.2108,
+      "grad_norm": 1.5185476541519165,
+      "learning_rate": 8.949155792224448e-06,
+      "loss": 0.3566,
+      "step": 2635
+    },
+    {
+      "epoch": 0.21088,
+      "grad_norm": 1.959179401397705,
+      "learning_rate": 8.948384324141794e-06,
+      "loss": 0.3719,
+      "step": 2636
+    },
+    {
+      "epoch": 0.21096,
+      "grad_norm": 1.65220046043396,
+      "learning_rate": 8.947612606258076e-06,
+      "loss": 0.3149,
+      "step": 2637
+    },
+    {
+      "epoch": 0.21104,
+      "grad_norm": 1.392369270324707,
+      "learning_rate": 8.946840638622117e-06,
+      "loss": 0.2671,
+      "step": 2638
+    },
+    {
+      "epoch": 0.21112,
+      "grad_norm": 1.6883777379989624,
+      "learning_rate": 8.946068421282754e-06,
+      "loss": 0.4213,
+      "step": 2639
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 1.4826557636260986,
+      "learning_rate": 8.945295954288848e-06,
+      "loss": 0.312,
+      "step": 2640
+    },
+    {
+      "epoch": 0.21128,
+      "grad_norm": 1.6768203973770142,
+      "learning_rate": 8.944523237689268e-06,
+      "loss": 0.3766,
+      "step": 2641
+    },
+    {
+      "epoch": 0.21136,
+      "grad_norm": 1.9561296701431274,
+      "learning_rate": 8.9437502715329e-06,
+      "loss": 0.3921,
+      "step": 2642
+    },
+    {
+      "epoch": 0.21144,
+      "grad_norm": 1.2287523746490479,
+      "learning_rate": 8.94297705586865e-06,
+      "loss": 0.3746,
+      "step": 2643
+    },
+    {
+      "epoch": 0.21152,
+      "grad_norm": 1.383736252784729,
+      "learning_rate": 8.942203590745433e-06,
+      "loss": 0.3747,
+      "step": 2644
+    },
+    {
+      "epoch": 0.2116,
+      "grad_norm": 2.0053927898406982,
+      "learning_rate": 8.941429876212187e-06,
+      "loss": 0.2964,
+      "step": 2645
+    },
+    {
+      "epoch": 0.21168,
+      "grad_norm": 1.3449758291244507,
+      "learning_rate": 8.94065591231786e-06,
+      "loss": 0.2413,
+      "step": 2646
+    },
+    {
+      "epoch": 0.21176,
+      "grad_norm": 1.6508148908615112,
+      "learning_rate": 8.939881699111418e-06,
+      "loss": 0.3681,
+      "step": 2647
+    },
+    {
+      "epoch": 0.21184,
+      "grad_norm": 1.292067050933838,
+      "learning_rate": 8.939107236641845e-06,
+      "loss": 0.3007,
+      "step": 2648
+    },
+    {
+      "epoch": 0.21192,
+      "grad_norm": 1.5809190273284912,
+      "learning_rate": 8.938332524958137e-06,
+      "loss": 0.3361,
+      "step": 2649
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 1.343364953994751,
+      "learning_rate": 8.937557564109307e-06,
+      "loss": 0.2596,
+      "step": 2650
+    },
+    {
+      "epoch": 0.21208,
+      "grad_norm": 1.7212300300598145,
+      "learning_rate": 8.936782354144387e-06,
+      "loss": 0.3525,
+      "step": 2651
+    },
+    {
+      "epoch": 0.21216,
+      "grad_norm": 2.104867935180664,
+      "learning_rate": 8.93600689511242e-06,
+      "loss": 0.4524,
+      "step": 2652
+    },
+    {
+      "epoch": 0.21224,
+      "grad_norm": 1.942893624305725,
+      "learning_rate": 8.935231187062465e-06,
+      "loss": 0.4223,
+      "step": 2653
+    },
+    {
+      "epoch": 0.21232,
+      "grad_norm": 1.5138806104660034,
+      "learning_rate": 8.9344552300436e-06,
+      "loss": 0.2939,
+      "step": 2654
+    },
+    {
+      "epoch": 0.2124,
+      "grad_norm": 1.6802589893341064,
+      "learning_rate": 8.93367902410492e-06,
+      "loss": 0.3564,
+      "step": 2655
+    },
+    {
+      "epoch": 0.21248,
+      "grad_norm": 1.7636982202529907,
+      "learning_rate": 8.932902569295527e-06,
+      "loss": 0.4802,
+      "step": 2656
+    },
+    {
+      "epoch": 0.21256,
+      "grad_norm": 1.8842812776565552,
+      "learning_rate": 8.932125865664549e-06,
+      "loss": 0.4784,
+      "step": 2657
+    },
+    {
+      "epoch": 0.21264,
+      "grad_norm": 1.495451807975769,
+      "learning_rate": 8.931348913261125e-06,
+      "loss": 0.2886,
+      "step": 2658
+    },
+    {
+      "epoch": 0.21272,
+      "grad_norm": 1.4396910667419434,
+      "learning_rate": 8.93057171213441e-06,
+      "loss": 0.2712,
+      "step": 2659
+    },
+    {
+      "epoch": 0.2128,
+      "grad_norm": 1.7895922660827637,
+      "learning_rate": 8.929794262333574e-06,
+      "loss": 0.3548,
+      "step": 2660
+    },
+    {
+      "epoch": 0.21288,
+      "grad_norm": 2.094775438308716,
+      "learning_rate": 8.929016563907805e-06,
+      "loss": 0.4313,
+      "step": 2661
+    },
+    {
+      "epoch": 0.21296,
+      "grad_norm": 1.7981663942337036,
+      "learning_rate": 8.928238616906302e-06,
+      "loss": 0.3359,
+      "step": 2662
+    },
+    {
+      "epoch": 0.21304,
+      "grad_norm": 1.6360666751861572,
+      "learning_rate": 8.927460421378287e-06,
+      "loss": 0.3905,
+      "step": 2663
+    },
+    {
+      "epoch": 0.21312,
+      "grad_norm": 2.0702054500579834,
+      "learning_rate": 8.926681977372993e-06,
+      "loss": 0.4009,
+      "step": 2664
+    },
+    {
+      "epoch": 0.2132,
+      "grad_norm": 1.4153721332550049,
+      "learning_rate": 8.92590328493967e-06,
+      "loss": 0.3303,
+      "step": 2665
+    },
+    {
+      "epoch": 0.21328,
+      "grad_norm": 1.5334463119506836,
+      "learning_rate": 8.92512434412758e-06,
+      "loss": 0.3581,
+      "step": 2666
+    },
+    {
+      "epoch": 0.21336,
+      "grad_norm": 1.445220708847046,
+      "learning_rate": 8.924345154986008e-06,
+      "loss": 0.2794,
+      "step": 2667
+    },
+    {
+      "epoch": 0.21344,
+      "grad_norm": 1.4006540775299072,
+      "learning_rate": 8.923565717564247e-06,
+      "loss": 0.2452,
+      "step": 2668
+    },
+    {
+      "epoch": 0.21352,
+      "grad_norm": 1.3207964897155762,
+      "learning_rate": 8.922786031911613e-06,
+      "loss": 0.2685,
+      "step": 2669
+    },
+    {
+      "epoch": 0.2136,
+      "grad_norm": 1.587815523147583,
+      "learning_rate": 8.922006098077432e-06,
+      "loss": 0.3331,
+      "step": 2670
+    },
+    {
+      "epoch": 0.21368,
+      "grad_norm": 1.743464469909668,
+      "learning_rate": 8.921225916111048e-06,
+      "loss": 0.3479,
+      "step": 2671
+    },
+    {
+      "epoch": 0.21376,
+      "grad_norm": 1.6636502742767334,
+      "learning_rate": 8.920445486061822e-06,
+      "loss": 0.3878,
+      "step": 2672
+    },
+    {
+      "epoch": 0.21384,
+      "grad_norm": 1.9191046953201294,
+      "learning_rate": 8.919664807979126e-06,
+      "loss": 0.4291,
+      "step": 2673
+    },
+    {
+      "epoch": 0.21392,
+      "grad_norm": 1.7954356670379639,
+      "learning_rate": 8.918883881912353e-06,
+      "loss": 0.3258,
+      "step": 2674
+    },
+    {
+      "epoch": 0.214,
+      "grad_norm": 1.4741268157958984,
+      "learning_rate": 8.91810270791091e-06,
+      "loss": 0.2867,
+      "step": 2675
+    },
+    {
+      "epoch": 0.21408,
+      "grad_norm": 1.7667266130447388,
+      "learning_rate": 8.917321286024218e-06,
+      "loss": 0.4209,
+      "step": 2676
+    },
+    {
+      "epoch": 0.21416,
+      "grad_norm": 1.924230694770813,
+      "learning_rate": 8.916539616301718e-06,
+      "loss": 0.3504,
+      "step": 2677
+    },
+    {
+      "epoch": 0.21424,
+      "grad_norm": 1.4865533113479614,
+      "learning_rate": 8.91575769879286e-06,
+      "loss": 0.3856,
+      "step": 2678
+    },
+    {
+      "epoch": 0.21432,
+      "grad_norm": 1.7114523649215698,
+      "learning_rate": 8.914975533547114e-06,
+      "loss": 0.4292,
+      "step": 2679
+    },
+    {
+      "epoch": 0.2144,
+      "grad_norm": 1.5713534355163574,
+      "learning_rate": 8.914193120613966e-06,
+      "loss": 0.3665,
+      "step": 2680
+    },
+    {
+      "epoch": 0.21448,
+      "grad_norm": 1.5431712865829468,
+      "learning_rate": 8.913410460042915e-06,
+      "loss": 0.3065,
+      "step": 2681
+    },
+    {
+      "epoch": 0.21456,
+      "grad_norm": 1.625646710395813,
+      "learning_rate": 8.91262755188348e-06,
+      "loss": 0.4117,
+      "step": 2682
+    },
+    {
+      "epoch": 0.21464,
+      "grad_norm": 1.5041875839233398,
+      "learning_rate": 8.911844396185192e-06,
+      "loss": 0.3816,
+      "step": 2683
+    },
+    {
+      "epoch": 0.21472,
+      "grad_norm": 1.8179059028625488,
+      "learning_rate": 8.911060992997596e-06,
+      "loss": 0.3031,
+      "step": 2684
+    },
+    {
+      "epoch": 0.2148,
+      "grad_norm": 1.7004882097244263,
+      "learning_rate": 8.910277342370259e-06,
+      "loss": 0.4804,
+      "step": 2685
+    },
+    {
+      "epoch": 0.21488,
+      "grad_norm": 2.1051254272460938,
+      "learning_rate": 8.909493444352757e-06,
+      "loss": 0.4295,
+      "step": 2686
+    },
+    {
+      "epoch": 0.21496,
+      "grad_norm": 1.3784528970718384,
+      "learning_rate": 8.908709298994686e-06,
+      "loss": 0.3112,
+      "step": 2687
+    },
+    {
+      "epoch": 0.21504,
+      "grad_norm": 1.7142976522445679,
+      "learning_rate": 8.907924906345659e-06,
+      "loss": 0.3797,
+      "step": 2688
+    },
+    {
+      "epoch": 0.21512,
+      "grad_norm": 1.7396079301834106,
+      "learning_rate": 8.907140266455297e-06,
+      "loss": 0.4081,
+      "step": 2689
+    },
+    {
+      "epoch": 0.2152,
+      "grad_norm": 0.8471266627311707,
+      "learning_rate": 8.906355379373243e-06,
+      "loss": 0.2378,
+      "step": 2690
+    },
+    {
+      "epoch": 0.21528,
+      "grad_norm": 1.6248892545700073,
+      "learning_rate": 8.905570245149156e-06,
+      "loss": 0.3695,
+      "step": 2691
+    },
+    {
+      "epoch": 0.21536,
+      "grad_norm": 1.4380123615264893,
+      "learning_rate": 8.904784863832708e-06,
+      "loss": 0.3377,
+      "step": 2692
+    },
+    {
+      "epoch": 0.21544,
+      "grad_norm": 1.8028794527053833,
+      "learning_rate": 8.903999235473586e-06,
+      "loss": 0.5013,
+      "step": 2693
+    },
+    {
+      "epoch": 0.21552,
+      "grad_norm": 1.2009968757629395,
+      "learning_rate": 8.903213360121496e-06,
+      "loss": 0.2609,
+      "step": 2694
+    },
+    {
+      "epoch": 0.2156,
+      "grad_norm": 1.5965559482574463,
+      "learning_rate": 8.902427237826157e-06,
+      "loss": 0.3731,
+      "step": 2695
+    },
+    {
+      "epoch": 0.21568,
+      "grad_norm": 1.466230869293213,
+      "learning_rate": 8.901640868637304e-06,
+      "loss": 0.3941,
+      "step": 2696
+    },
+    {
+      "epoch": 0.21576,
+      "grad_norm": 1.538789987564087,
+      "learning_rate": 8.900854252604689e-06,
+      "loss": 0.2974,
+      "step": 2697
+    },
+    {
+      "epoch": 0.21584,
+      "grad_norm": 1.5418365001678467,
+      "learning_rate": 8.900067389778075e-06,
+      "loss": 0.38,
+      "step": 2698
+    },
+    {
+      "epoch": 0.21592,
+      "grad_norm": 1.4428279399871826,
+      "learning_rate": 8.89928028020725e-06,
+      "loss": 0.3617,
+      "step": 2699
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 1.5055745840072632,
+      "learning_rate": 8.898492923942007e-06,
+      "loss": 0.2906,
+      "step": 2700
+    },
+    {
+      "epoch": 0.21608,
+      "grad_norm": 1.5170435905456543,
+      "learning_rate": 8.897705321032162e-06,
+      "loss": 0.3001,
+      "step": 2701
+    },
+    {
+      "epoch": 0.21616,
+      "grad_norm": 1.5084747076034546,
+      "learning_rate": 8.896917471527542e-06,
+      "loss": 0.3537,
+      "step": 2702
+    },
+    {
+      "epoch": 0.21624,
+      "grad_norm": 1.590378761291504,
+      "learning_rate": 8.896129375477993e-06,
+      "loss": 0.2969,
+      "step": 2703
+    },
+    {
+      "epoch": 0.21632,
+      "grad_norm": 1.5651493072509766,
+      "learning_rate": 8.895341032933376e-06,
+      "loss": 0.3104,
+      "step": 2704
+    },
+    {
+      "epoch": 0.2164,
+      "grad_norm": 1.545362949371338,
+      "learning_rate": 8.894552443943564e-06,
+      "loss": 0.3274,
+      "step": 2705
+    },
+    {
+      "epoch": 0.21648,
+      "grad_norm": 1.7417821884155273,
+      "learning_rate": 8.893763608558453e-06,
+      "loss": 0.4584,
+      "step": 2706
+    },
+    {
+      "epoch": 0.21656,
+      "grad_norm": 1.2758499383926392,
+      "learning_rate": 8.892974526827944e-06,
+      "loss": 0.2997,
+      "step": 2707
+    },
+    {
+      "epoch": 0.21664,
+      "grad_norm": 1.4422128200531006,
+      "learning_rate": 8.892185198801963e-06,
+      "loss": 0.3044,
+      "step": 2708
+    },
+    {
+      "epoch": 0.21672,
+      "grad_norm": 1.5123414993286133,
+      "learning_rate": 8.891395624530449e-06,
+      "loss": 0.3265,
+      "step": 2709
+    },
+    {
+      "epoch": 0.2168,
+      "grad_norm": 1.5700615644454956,
+      "learning_rate": 8.890605804063353e-06,
+      "loss": 0.3064,
+      "step": 2710
+    },
+    {
+      "epoch": 0.21688,
+      "grad_norm": 1.971044659614563,
+      "learning_rate": 8.889815737450648e-06,
+      "loss": 0.3514,
+      "step": 2711
+    },
+    {
+      "epoch": 0.21696,
+      "grad_norm": 1.5621892213821411,
+      "learning_rate": 8.889025424742314e-06,
+      "loss": 0.3242,
+      "step": 2712
+    },
+    {
+      "epoch": 0.21704,
+      "grad_norm": 1.4305284023284912,
+      "learning_rate": 8.888234865988356e-06,
+      "loss": 0.3518,
+      "step": 2713
+    },
+    {
+      "epoch": 0.21712,
+      "grad_norm": 1.436153769493103,
+      "learning_rate": 8.887444061238787e-06,
+      "loss": 0.3249,
+      "step": 2714
+    },
+    {
+      "epoch": 0.2172,
+      "grad_norm": 1.4724273681640625,
+      "learning_rate": 8.886653010543641e-06,
+      "loss": 0.303,
+      "step": 2715
+    },
+    {
+      "epoch": 0.21728,
+      "grad_norm": 2.188133716583252,
+      "learning_rate": 8.885861713952964e-06,
+      "loss": 0.4513,
+      "step": 2716
+    },
+    {
+      "epoch": 0.21736,
+      "grad_norm": 1.3821355104446411,
+      "learning_rate": 8.885070171516816e-06,
+      "loss": 0.2599,
+      "step": 2717
+    },
+    {
+      "epoch": 0.21744,
+      "grad_norm": 1.943583369255066,
+      "learning_rate": 8.88427838328528e-06,
+      "loss": 0.5347,
+      "step": 2718
+    },
+    {
+      "epoch": 0.21752,
+      "grad_norm": 1.9702650308609009,
+      "learning_rate": 8.883486349308446e-06,
+      "loss": 0.387,
+      "step": 2719
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 1.7490559816360474,
+      "learning_rate": 8.882694069636426e-06,
+      "loss": 0.3813,
+      "step": 2720
+    },
+    {
+      "epoch": 0.21768,
+      "grad_norm": 1.3750137090682983,
+      "learning_rate": 8.881901544319345e-06,
+      "loss": 0.3268,
+      "step": 2721
+    },
+    {
+      "epoch": 0.21776,
+      "grad_norm": 1.3257077932357788,
+      "learning_rate": 8.881108773407338e-06,
+      "loss": 0.2837,
+      "step": 2722
+    },
+    {
+      "epoch": 0.21784,
+      "grad_norm": 1.7465240955352783,
+      "learning_rate": 8.88031575695057e-06,
+      "loss": 0.3157,
+      "step": 2723
+    },
+    {
+      "epoch": 0.21792,
+      "grad_norm": 1.619113564491272,
+      "learning_rate": 8.879522494999204e-06,
+      "loss": 0.3192,
+      "step": 2724
+    },
+    {
+      "epoch": 0.218,
+      "grad_norm": 1.9502321481704712,
+      "learning_rate": 8.878728987603433e-06,
+      "loss": 0.3442,
+      "step": 2725
+    },
+    {
+      "epoch": 0.21808,
+      "grad_norm": 1.1939575672149658,
+      "learning_rate": 8.877935234813455e-06,
+      "loss": 0.2337,
+      "step": 2726
+    },
+    {
+      "epoch": 0.21816,
+      "grad_norm": 1.9584698677062988,
+      "learning_rate": 8.877141236679492e-06,
+      "loss": 0.3656,
+      "step": 2727
+    },
+    {
+      "epoch": 0.21824,
+      "grad_norm": 1.442341685295105,
+      "learning_rate": 8.876346993251777e-06,
+      "loss": 0.352,
+      "step": 2728
+    },
+    {
+      "epoch": 0.21832,
+      "grad_norm": 1.8673537969589233,
+      "learning_rate": 8.875552504580556e-06,
+      "loss": 0.391,
+      "step": 2729
+    },
+    {
+      "epoch": 0.2184,
+      "grad_norm": 1.3063445091247559,
+      "learning_rate": 8.874757770716096e-06,
+      "loss": 0.3345,
+      "step": 2730
+    },
+    {
+      "epoch": 0.21848,
+      "grad_norm": 1.615033507347107,
+      "learning_rate": 8.873962791708676e-06,
+      "loss": 0.3744,
+      "step": 2731
+    },
+    {
+      "epoch": 0.21856,
+      "grad_norm": 1.8220925331115723,
+      "learning_rate": 8.873167567608594e-06,
+      "loss": 0.3211,
+      "step": 2732
+    },
+    {
+      "epoch": 0.21864,
+      "grad_norm": 1.4469846487045288,
+      "learning_rate": 8.872372098466159e-06,
+      "loss": 0.4197,
+      "step": 2733
+    },
+    {
+      "epoch": 0.21872,
+      "grad_norm": 1.1662476062774658,
+      "learning_rate": 8.871576384331699e-06,
+      "loss": 0.263,
+      "step": 2734
+    },
+    {
+      "epoch": 0.2188,
+      "grad_norm": 1.2721196413040161,
+      "learning_rate": 8.870780425255554e-06,
+      "loss": 0.2601,
+      "step": 2735
+    },
+    {
+      "epoch": 0.21888,
+      "grad_norm": 1.9340859651565552,
+      "learning_rate": 8.869984221288085e-06,
+      "loss": 0.3647,
+      "step": 2736
+    },
+    {
+      "epoch": 0.21896,
+      "grad_norm": 1.5645490884780884,
+      "learning_rate": 8.869187772479661e-06,
+      "loss": 0.3446,
+      "step": 2737
+    },
+    {
+      "epoch": 0.21904,
+      "grad_norm": 1.3666660785675049,
+      "learning_rate": 8.868391078880677e-06,
+      "loss": 0.3243,
+      "step": 2738
+    },
+    {
+      "epoch": 0.21912,
+      "grad_norm": 1.340316653251648,
+      "learning_rate": 8.86759414054153e-06,
+      "loss": 0.2668,
+      "step": 2739
+    },
+    {
+      "epoch": 0.2192,
+      "grad_norm": 1.5360743999481201,
+      "learning_rate": 8.866796957512642e-06,
+      "loss": 0.3216,
+      "step": 2740
+    },
+    {
+      "epoch": 0.21928,
+      "grad_norm": 1.6655339002609253,
+      "learning_rate": 8.865999529844452e-06,
+      "loss": 0.3079,
+      "step": 2741
+    },
+    {
+      "epoch": 0.21936,
+      "grad_norm": 1.4910820722579956,
+      "learning_rate": 8.865201857587405e-06,
+      "loss": 0.2896,
+      "step": 2742
+    },
+    {
+      "epoch": 0.21944,
+      "grad_norm": 1.905479907989502,
+      "learning_rate": 8.864403940791969e-06,
+      "loss": 0.4271,
+      "step": 2743
+    },
+    {
+      "epoch": 0.21952,
+      "grad_norm": 1.5210912227630615,
+      "learning_rate": 8.863605779508627e-06,
+      "loss": 0.319,
+      "step": 2744
+    },
+    {
+      "epoch": 0.2196,
+      "grad_norm": 1.7995033264160156,
+      "learning_rate": 8.862807373787876e-06,
+      "loss": 0.3828,
+      "step": 2745
+    },
+    {
+      "epoch": 0.21968,
+      "grad_norm": 1.853249430656433,
+      "learning_rate": 8.862008723680225e-06,
+      "loss": 0.374,
+      "step": 2746
+    },
+    {
+      "epoch": 0.21976,
+      "grad_norm": 1.831558108329773,
+      "learning_rate": 8.861209829236206e-06,
+      "loss": 0.4955,
+      "step": 2747
+    },
+    {
+      "epoch": 0.21984,
+      "grad_norm": 1.945634365081787,
+      "learning_rate": 8.860410690506361e-06,
+      "loss": 0.3997,
+      "step": 2748
+    },
+    {
+      "epoch": 0.21992,
+      "grad_norm": 1.5275148153305054,
+      "learning_rate": 8.859611307541247e-06,
+      "loss": 0.3866,
+      "step": 2749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7615596055984497,
+      "learning_rate": 8.858811680391442e-06,
+      "loss": 0.3599,
+      "step": 2750
+    },
+    {
+      "epoch": 0.22008,
+      "grad_norm": 1.431030511856079,
+      "learning_rate": 8.858011809107532e-06,
+      "loss": 0.2883,
+      "step": 2751
+    },
+    {
+      "epoch": 0.22016,
+      "grad_norm": 1.2879825830459595,
+      "learning_rate": 8.857211693740125e-06,
+      "loss": 0.317,
+      "step": 2752
+    },
+    {
+      "epoch": 0.22024,
+      "grad_norm": 1.4592652320861816,
+      "learning_rate": 8.856411334339841e-06,
+      "loss": 0.2936,
+      "step": 2753
+    },
+    {
+      "epoch": 0.22032,
+      "grad_norm": 1.5514662265777588,
+      "learning_rate": 8.855610730957313e-06,
+      "loss": 0.4303,
+      "step": 2754
+    },
+    {
+      "epoch": 0.2204,
+      "grad_norm": 1.5528576374053955,
+      "learning_rate": 8.854809883643197e-06,
+      "loss": 0.3552,
+      "step": 2755
+    },
+    {
+      "epoch": 0.22048,
+      "grad_norm": 1.8523775339126587,
+      "learning_rate": 8.854008792448156e-06,
+      "loss": 0.4036,
+      "step": 2756
+    },
+    {
+      "epoch": 0.22056,
+      "grad_norm": 1.6846541166305542,
+      "learning_rate": 8.853207457422877e-06,
+      "loss": 0.3821,
+      "step": 2757
+    },
+    {
+      "epoch": 0.22064,
+      "grad_norm": 1.7963849306106567,
+      "learning_rate": 8.852405878618052e-06,
+      "loss": 0.3123,
+      "step": 2758
+    },
+    {
+      "epoch": 0.22072,
+      "grad_norm": 1.451002597808838,
+      "learning_rate": 8.8516040560844e-06,
+      "loss": 0.3076,
+      "step": 2759
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 1.808860421180725,
+      "learning_rate": 8.850801989872644e-06,
+      "loss": 0.3846,
+      "step": 2760
+    },
+    {
+      "epoch": 0.22088,
+      "grad_norm": 1.5843877792358398,
+      "learning_rate": 8.849999680033535e-06,
+      "loss": 0.286,
+      "step": 2761
+    },
+    {
+      "epoch": 0.22096,
+      "grad_norm": 1.2339478731155396,
+      "learning_rate": 8.849197126617824e-06,
+      "loss": 0.3054,
+      "step": 2762
+    },
+    {
+      "epoch": 0.22104,
+      "grad_norm": 1.9607781171798706,
+      "learning_rate": 8.848394329676294e-06,
+      "loss": 0.3739,
+      "step": 2763
+    },
+    {
+      "epoch": 0.22112,
+      "grad_norm": 1.4922035932540894,
+      "learning_rate": 8.847591289259729e-06,
+      "loss": 0.3911,
+      "step": 2764
+    },
+    {
+      "epoch": 0.2212,
+      "grad_norm": 1.7909282445907593,
+      "learning_rate": 8.846788005418938e-06,
+      "loss": 0.401,
+      "step": 2765
+    },
+    {
+      "epoch": 0.22128,
+      "grad_norm": 1.6844979524612427,
+      "learning_rate": 8.845984478204742e-06,
+      "loss": 0.4258,
+      "step": 2766
+    },
+    {
+      "epoch": 0.22136,
+      "grad_norm": 1.6638482809066772,
+      "learning_rate": 8.845180707667975e-06,
+      "loss": 0.3188,
+      "step": 2767
+    },
+    {
+      "epoch": 0.22144,
+      "grad_norm": 1.4212157726287842,
+      "learning_rate": 8.84437669385949e-06,
+      "loss": 0.3071,
+      "step": 2768
+    },
+    {
+      "epoch": 0.22152,
+      "grad_norm": 1.4176979064941406,
+      "learning_rate": 8.843572436830157e-06,
+      "loss": 0.3029,
+      "step": 2769
+    },
+    {
+      "epoch": 0.2216,
+      "grad_norm": 1.8087478876113892,
+      "learning_rate": 8.842767936630857e-06,
+      "loss": 0.4011,
+      "step": 2770
+    },
+    {
+      "epoch": 0.22168,
+      "grad_norm": 1.3983772993087769,
+      "learning_rate": 8.841963193312487e-06,
+      "loss": 0.347,
+      "step": 2771
+    },
+    {
+      "epoch": 0.22176,
+      "grad_norm": 1.5463216304779053,
+      "learning_rate": 8.841158206925959e-06,
+      "loss": 0.3349,
+      "step": 2772
+    },
+    {
+      "epoch": 0.22184,
+      "grad_norm": 1.3587409257888794,
+      "learning_rate": 8.840352977522206e-06,
+      "loss": 0.3083,
+      "step": 2773
+    },
+    {
+      "epoch": 0.22192,
+      "grad_norm": 1.3786388635635376,
+      "learning_rate": 8.83954750515217e-06,
+      "loss": 0.3388,
+      "step": 2774
+    },
+    {
+      "epoch": 0.222,
+      "grad_norm": 1.750858187675476,
+      "learning_rate": 8.83874178986681e-06,
+      "loss": 0.4101,
+      "step": 2775
+    },
+    {
+      "epoch": 0.22208,
+      "grad_norm": 1.5796808004379272,
+      "learning_rate": 8.837935831717102e-06,
+      "loss": 0.3572,
+      "step": 2776
+    },
+    {
+      "epoch": 0.22216,
+      "grad_norm": 1.397741675376892,
+      "learning_rate": 8.837129630754034e-06,
+      "loss": 0.3428,
+      "step": 2777
+    },
+    {
+      "epoch": 0.22224,
+      "grad_norm": 1.4873077869415283,
+      "learning_rate": 8.836323187028615e-06,
+      "loss": 0.3411,
+      "step": 2778
+    },
+    {
+      "epoch": 0.22232,
+      "grad_norm": 1.351516604423523,
+      "learning_rate": 8.835516500591863e-06,
+      "loss": 0.3341,
+      "step": 2779
+    },
+    {
+      "epoch": 0.2224,
+      "grad_norm": 1.303882360458374,
+      "learning_rate": 8.834709571494817e-06,
+      "loss": 0.3025,
+      "step": 2780
+    },
+    {
+      "epoch": 0.22248,
+      "grad_norm": 1.4679828882217407,
+      "learning_rate": 8.833902399788527e-06,
+      "loss": 0.3012,
+      "step": 2781
+    },
+    {
+      "epoch": 0.22256,
+      "grad_norm": 1.4130665063858032,
+      "learning_rate": 8.83309498552406e-06,
+      "loss": 0.2994,
+      "step": 2782
+    },
+    {
+      "epoch": 0.22264,
+      "grad_norm": 1.885979413986206,
+      "learning_rate": 8.832287328752499e-06,
+      "loss": 0.4008,
+      "step": 2783
+    },
+    {
+      "epoch": 0.22272,
+      "grad_norm": 2.2439332008361816,
+      "learning_rate": 8.83147942952494e-06,
+      "loss": 0.5255,
+      "step": 2784
+    },
+    {
+      "epoch": 0.2228,
+      "grad_norm": 1.7513844966888428,
+      "learning_rate": 8.8306712878925e-06,
+      "loss": 0.3906,
+      "step": 2785
+    },
+    {
+      "epoch": 0.22288,
+      "grad_norm": 1.585735559463501,
+      "learning_rate": 8.829862903906306e-06,
+      "loss": 0.3534,
+      "step": 2786
+    },
+    {
+      "epoch": 0.22296,
+      "grad_norm": 1.6071237325668335,
+      "learning_rate": 8.829054277617499e-06,
+      "loss": 0.3327,
+      "step": 2787
+    },
+    {
+      "epoch": 0.22304,
+      "grad_norm": 1.4853163957595825,
+      "learning_rate": 8.828245409077241e-06,
+      "loss": 0.3353,
+      "step": 2788
+    },
+    {
+      "epoch": 0.22312,
+      "grad_norm": 1.8716695308685303,
+      "learning_rate": 8.827436298336703e-06,
+      "loss": 0.4146,
+      "step": 2789
+    },
+    {
+      "epoch": 0.2232,
+      "grad_norm": 1.6231932640075684,
+      "learning_rate": 8.826626945447079e-06,
+      "loss": 0.3451,
+      "step": 2790
+    },
+    {
+      "epoch": 0.22328,
+      "grad_norm": 1.5309462547302246,
+      "learning_rate": 8.825817350459571e-06,
+      "loss": 0.3078,
+      "step": 2791
+    },
+    {
+      "epoch": 0.22336,
+      "grad_norm": 1.340057134628296,
+      "learning_rate": 8.825007513425401e-06,
+      "loss": 0.2812,
+      "step": 2792
+    },
+    {
+      "epoch": 0.22344,
+      "grad_norm": 1.6794257164001465,
+      "learning_rate": 8.824197434395805e-06,
+      "loss": 0.3899,
+      "step": 2793
+    },
+    {
+      "epoch": 0.22352,
+      "grad_norm": 1.566836953163147,
+      "learning_rate": 8.823387113422034e-06,
+      "loss": 0.4213,
+      "step": 2794
+    },
+    {
+      "epoch": 0.2236,
+      "grad_norm": 1.4850705862045288,
+      "learning_rate": 8.82257655055535e-06,
+      "loss": 0.3569,
+      "step": 2795
+    },
+    {
+      "epoch": 0.22368,
+      "grad_norm": 1.5087475776672363,
+      "learning_rate": 8.82176574584704e-06,
+      "loss": 0.3534,
+      "step": 2796
+    },
+    {
+      "epoch": 0.22376,
+      "grad_norm": 1.5350658893585205,
+      "learning_rate": 8.820954699348399e-06,
+      "loss": 0.3551,
+      "step": 2797
+    },
+    {
+      "epoch": 0.22384,
+      "grad_norm": 1.1860790252685547,
+      "learning_rate": 8.820143411110737e-06,
+      "loss": 0.214,
+      "step": 2798
+    },
+    {
+      "epoch": 0.22392,
+      "grad_norm": 2.052689552307129,
+      "learning_rate": 8.819331881185387e-06,
+      "loss": 0.5796,
+      "step": 2799
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 1.7889010906219482,
+      "learning_rate": 8.818520109623687e-06,
+      "loss": 0.441,
+      "step": 2800
+    },
+    {
+      "epoch": 0.22408,
+      "grad_norm": 1.6740270853042603,
+      "learning_rate": 8.817708096476996e-06,
+      "loss": 0.3098,
+      "step": 2801
+    },
+    {
+      "epoch": 0.22416,
+      "grad_norm": 1.4243390560150146,
+      "learning_rate": 8.81689584179669e-06,
+      "loss": 0.2933,
+      "step": 2802
+    },
+    {
+      "epoch": 0.22424,
+      "grad_norm": 1.7533442974090576,
+      "learning_rate": 8.816083345634153e-06,
+      "loss": 0.3185,
+      "step": 2803
+    },
+    {
+      "epoch": 0.22432,
+      "grad_norm": 1.8434312343597412,
+      "learning_rate": 8.815270608040792e-06,
+      "loss": 0.4669,
+      "step": 2804
+    },
+    {
+      "epoch": 0.2244,
+      "grad_norm": 1.609260082244873,
+      "learning_rate": 8.814457629068025e-06,
+      "loss": 0.304,
+      "step": 2805
+    },
+    {
+      "epoch": 0.22448,
+      "grad_norm": 1.258373737335205,
+      "learning_rate": 8.813644408767287e-06,
+      "loss": 0.2606,
+      "step": 2806
+    },
+    {
+      "epoch": 0.22456,
+      "grad_norm": 1.3866522312164307,
+      "learning_rate": 8.812830947190028e-06,
+      "loss": 0.2768,
+      "step": 2807
+    },
+    {
+      "epoch": 0.22464,
+      "grad_norm": 1.794108271598816,
+      "learning_rate": 8.812017244387714e-06,
+      "loss": 0.4018,
+      "step": 2808
+    },
+    {
+      "epoch": 0.22472,
+      "grad_norm": 1.3862345218658447,
+      "learning_rate": 8.811203300411823e-06,
+      "loss": 0.3217,
+      "step": 2809
+    },
+    {
+      "epoch": 0.2248,
+      "grad_norm": 2.0301730632781982,
+      "learning_rate": 8.81038911531385e-06,
+      "loss": 0.3777,
+      "step": 2810
+    },
+    {
+      "epoch": 0.22488,
+      "grad_norm": 1.378750205039978,
+      "learning_rate": 8.80957468914531e-06,
+      "loss": 0.3752,
+      "step": 2811
+    },
+    {
+      "epoch": 0.22496,
+      "grad_norm": 1.7173678874969482,
+      "learning_rate": 8.808760021957725e-06,
+      "loss": 0.3415,
+      "step": 2812
+    },
+    {
+      "epoch": 0.22504,
+      "grad_norm": 1.7305114269256592,
+      "learning_rate": 8.807945113802638e-06,
+      "loss": 0.4127,
+      "step": 2813
+    },
+    {
+      "epoch": 0.22512,
+      "grad_norm": 1.6595813035964966,
+      "learning_rate": 8.807129964731604e-06,
+      "loss": 0.3818,
+      "step": 2814
+    },
+    {
+      "epoch": 0.2252,
+      "grad_norm": 1.9219270944595337,
+      "learning_rate": 8.806314574796198e-06,
+      "loss": 0.4096,
+      "step": 2815
+    },
+    {
+      "epoch": 0.22528,
+      "grad_norm": 1.5241590738296509,
+      "learning_rate": 8.805498944048003e-06,
+      "loss": 0.3071,
+      "step": 2816
+    },
+    {
+      "epoch": 0.22536,
+      "grad_norm": 1.5931185483932495,
+      "learning_rate": 8.804683072538623e-06,
+      "loss": 0.3009,
+      "step": 2817
+    },
+    {
+      "epoch": 0.22544,
+      "grad_norm": 1.7457307577133179,
+      "learning_rate": 8.803866960319676e-06,
+      "loss": 0.2971,
+      "step": 2818
+    },
+    {
+      "epoch": 0.22552,
+      "grad_norm": 1.769567847251892,
+      "learning_rate": 8.803050607442794e-06,
+      "loss": 0.4444,
+      "step": 2819
+    },
+    {
+      "epoch": 0.2256,
+      "grad_norm": 2.1283857822418213,
+      "learning_rate": 8.802234013959626e-06,
+      "loss": 0.4656,
+      "step": 2820
+    },
+    {
+      "epoch": 0.22568,
+      "grad_norm": 1.2949507236480713,
+      "learning_rate": 8.801417179921834e-06,
+      "loss": 0.2726,
+      "step": 2821
+    },
+    {
+      "epoch": 0.22576,
+      "grad_norm": 1.8464694023132324,
+      "learning_rate": 8.800600105381097e-06,
+      "loss": 0.4021,
+      "step": 2822
+    },
+    {
+      "epoch": 0.22584,
+      "grad_norm": 1.674856185913086,
+      "learning_rate": 8.799782790389107e-06,
+      "loss": 0.3484,
+      "step": 2823
+    },
+    {
+      "epoch": 0.22592,
+      "grad_norm": 1.2917641401290894,
+      "learning_rate": 8.798965234997574e-06,
+      "loss": 0.2873,
+      "step": 2824
+    },
+    {
+      "epoch": 0.226,
+      "grad_norm": 1.5360853672027588,
+      "learning_rate": 8.798147439258222e-06,
+      "loss": 0.3182,
+      "step": 2825
+    },
+    {
+      "epoch": 0.22608,
+      "grad_norm": 1.5402098894119263,
+      "learning_rate": 8.797329403222791e-06,
+      "loss": 0.3192,
+      "step": 2826
+    },
+    {
+      "epoch": 0.22616,
+      "grad_norm": 1.3484301567077637,
+      "learning_rate": 8.796511126943032e-06,
+      "loss": 0.2512,
+      "step": 2827
+    },
+    {
+      "epoch": 0.22624,
+      "grad_norm": 1.4177839756011963,
+      "learning_rate": 8.79569261047072e-06,
+      "loss": 0.3263,
+      "step": 2828
+    },
+    {
+      "epoch": 0.22632,
+      "grad_norm": 1.4625004529953003,
+      "learning_rate": 8.794873853857638e-06,
+      "loss": 0.3261,
+      "step": 2829
+    },
+    {
+      "epoch": 0.2264,
+      "grad_norm": 1.3922655582427979,
+      "learning_rate": 8.794054857155582e-06,
+      "loss": 0.3243,
+      "step": 2830
+    },
+    {
+      "epoch": 0.22648,
+      "grad_norm": 1.4068520069122314,
+      "learning_rate": 8.793235620416372e-06,
+      "loss": 0.2632,
+      "step": 2831
+    },
+    {
+      "epoch": 0.22656,
+      "grad_norm": 1.9339982271194458,
+      "learning_rate": 8.792416143691836e-06,
+      "loss": 0.4937,
+      "step": 2832
+    },
+    {
+      "epoch": 0.22664,
+      "grad_norm": 1.5683056116104126,
+      "learning_rate": 8.791596427033818e-06,
+      "loss": 0.3712,
+      "step": 2833
+    },
+    {
+      "epoch": 0.22672,
+      "grad_norm": 1.811155915260315,
+      "learning_rate": 8.790776470494183e-06,
+      "loss": 0.3811,
+      "step": 2834
+    },
+    {
+      "epoch": 0.2268,
+      "grad_norm": 1.855803370475769,
+      "learning_rate": 8.789956274124805e-06,
+      "loss": 0.4298,
+      "step": 2835
+    },
+    {
+      "epoch": 0.22688,
+      "grad_norm": 2.0151405334472656,
+      "learning_rate": 8.789135837977573e-06,
+      "loss": 0.4112,
+      "step": 2836
+    },
+    {
+      "epoch": 0.22696,
+      "grad_norm": 1.684455394744873,
+      "learning_rate": 8.788315162104396e-06,
+      "loss": 0.2819,
+      "step": 2837
+    },
+    {
+      "epoch": 0.22704,
+      "grad_norm": 1.5724467039108276,
+      "learning_rate": 8.787494246557195e-06,
+      "loss": 0.3237,
+      "step": 2838
+    },
+    {
+      "epoch": 0.22712,
+      "grad_norm": 1.4349114894866943,
+      "learning_rate": 8.786673091387906e-06,
+      "loss": 0.3576,
+      "step": 2839
+    },
+    {
+      "epoch": 0.2272,
+      "grad_norm": 1.4084806442260742,
+      "learning_rate": 8.78585169664848e-06,
+      "loss": 0.2697,
+      "step": 2840
+    },
+    {
+      "epoch": 0.22728,
+      "grad_norm": 1.615369200706482,
+      "learning_rate": 8.785030062390885e-06,
+      "loss": 0.3478,
+      "step": 2841
+    },
+    {
+      "epoch": 0.22736,
+      "grad_norm": 1.2227230072021484,
+      "learning_rate": 8.784208188667102e-06,
+      "loss": 0.2652,
+      "step": 2842
+    },
+    {
+      "epoch": 0.22744,
+      "grad_norm": 1.4531816244125366,
+      "learning_rate": 8.78338607552913e-06,
+      "loss": 0.3219,
+      "step": 2843
+    },
+    {
+      "epoch": 0.22752,
+      "grad_norm": 1.391142725944519,
+      "learning_rate": 8.782563723028979e-06,
+      "loss": 0.3095,
+      "step": 2844
+    },
+    {
+      "epoch": 0.2276,
+      "grad_norm": 1.6999260187149048,
+      "learning_rate": 8.781741131218678e-06,
+      "loss": 0.3403,
+      "step": 2845
+    },
+    {
+      "epoch": 0.22768,
+      "grad_norm": 1.1833946704864502,
+      "learning_rate": 8.78091830015027e-06,
+      "loss": 0.2708,
+      "step": 2846
+    },
+    {
+      "epoch": 0.22776,
+      "grad_norm": 1.4600259065628052,
+      "learning_rate": 8.780095229875813e-06,
+      "loss": 0.4186,
+      "step": 2847
+    },
+    {
+      "epoch": 0.22784,
+      "grad_norm": 1.6836893558502197,
+      "learning_rate": 8.779271920447378e-06,
+      "loss": 0.4746,
+      "step": 2848
+    },
+    {
+      "epoch": 0.22792,
+      "grad_norm": 1.6797387599945068,
+      "learning_rate": 8.778448371917055e-06,
+      "loss": 0.3224,
+      "step": 2849
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 1.6203384399414062,
+      "learning_rate": 8.777624584336944e-06,
+      "loss": 0.4012,
+      "step": 2850
+    },
+    {
+      "epoch": 0.22808,
+      "grad_norm": 1.385426640510559,
+      "learning_rate": 8.776800557759167e-06,
+      "loss": 0.3187,
+      "step": 2851
+    },
+    {
+      "epoch": 0.22816,
+      "grad_norm": 1.3712928295135498,
+      "learning_rate": 8.775976292235857e-06,
+      "loss": 0.3654,
+      "step": 2852
+    },
+    {
+      "epoch": 0.22824,
+      "grad_norm": 1.6508076190948486,
+      "learning_rate": 8.775151787819159e-06,
+      "loss": 0.4184,
+      "step": 2853
+    },
+    {
+      "epoch": 0.22832,
+      "grad_norm": 1.793189525604248,
+      "learning_rate": 8.77432704456124e-06,
+      "loss": 0.4786,
+      "step": 2854
+    },
+    {
+      "epoch": 0.2284,
+      "grad_norm": 1.7186195850372314,
+      "learning_rate": 8.77350206251428e-06,
+      "loss": 0.3742,
+      "step": 2855
+    },
+    {
+      "epoch": 0.22848,
+      "grad_norm": 1.5993332862854004,
+      "learning_rate": 8.772676841730468e-06,
+      "loss": 0.397,
+      "step": 2856
+    },
+    {
+      "epoch": 0.22856,
+      "grad_norm": 1.4774996042251587,
+      "learning_rate": 8.771851382262016e-06,
+      "loss": 0.3349,
+      "step": 2857
+    },
+    {
+      "epoch": 0.22864,
+      "grad_norm": 1.5428051948547363,
+      "learning_rate": 8.771025684161147e-06,
+      "loss": 0.3738,
+      "step": 2858
+    },
+    {
+      "epoch": 0.22872,
+      "grad_norm": 1.4334566593170166,
+      "learning_rate": 8.770199747480105e-06,
+      "loss": 0.2782,
+      "step": 2859
+    },
+    {
+      "epoch": 0.2288,
+      "grad_norm": 2.1138930320739746,
+      "learning_rate": 8.769373572271137e-06,
+      "loss": 0.4068,
+      "step": 2860
+    },
+    {
+      "epoch": 0.22888,
+      "grad_norm": 1.580213189125061,
+      "learning_rate": 8.768547158586514e-06,
+      "loss": 0.304,
+      "step": 2861
+    },
+    {
+      "epoch": 0.22896,
+      "grad_norm": 1.601637840270996,
+      "learning_rate": 8.767720506478523e-06,
+      "loss": 0.3489,
+      "step": 2862
+    },
+    {
+      "epoch": 0.22904,
+      "grad_norm": 1.425022840499878,
+      "learning_rate": 8.766893615999463e-06,
+      "loss": 0.2925,
+      "step": 2863
+    },
+    {
+      "epoch": 0.22912,
+      "grad_norm": 1.401037573814392,
+      "learning_rate": 8.766066487201648e-06,
+      "loss": 0.3418,
+      "step": 2864
+    },
+    {
+      "epoch": 0.2292,
+      "grad_norm": 1.5454732179641724,
+      "learning_rate": 8.765239120137407e-06,
+      "loss": 0.3042,
+      "step": 2865
+    },
+    {
+      "epoch": 0.22928,
+      "grad_norm": 1.3919222354888916,
+      "learning_rate": 8.764411514859086e-06,
+      "loss": 0.3628,
+      "step": 2866
+    },
+    {
+      "epoch": 0.22936,
+      "grad_norm": 1.6965492963790894,
+      "learning_rate": 8.763583671419045e-06,
+      "loss": 0.3109,
+      "step": 2867
+    },
+    {
+      "epoch": 0.22944,
+      "grad_norm": 1.665880799293518,
+      "learning_rate": 8.762755589869655e-06,
+      "loss": 0.3084,
+      "step": 2868
+    },
+    {
+      "epoch": 0.22952,
+      "grad_norm": 1.277369737625122,
+      "learning_rate": 8.761927270263313e-06,
+      "loss": 0.2969,
+      "step": 2869
+    },
+    {
+      "epoch": 0.2296,
+      "grad_norm": 1.5863888263702393,
+      "learning_rate": 8.761098712652418e-06,
+      "loss": 0.342,
+      "step": 2870
+    },
+    {
+      "epoch": 0.22968,
+      "grad_norm": 1.9805275201797485,
+      "learning_rate": 8.760269917089392e-06,
+      "loss": 0.4074,
+      "step": 2871
+    },
+    {
+      "epoch": 0.22976,
+      "grad_norm": 1.5686451196670532,
+      "learning_rate": 8.75944088362667e-06,
+      "loss": 0.3272,
+      "step": 2872
+    },
+    {
+      "epoch": 0.22984,
+      "grad_norm": 1.4256458282470703,
+      "learning_rate": 8.758611612316704e-06,
+      "loss": 0.2976,
+      "step": 2873
+    },
+    {
+      "epoch": 0.22992,
+      "grad_norm": 1.0142706632614136,
+      "learning_rate": 8.757782103211958e-06,
+      "loss": 0.2594,
+      "step": 2874
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2783145904541016,
+      "learning_rate": 8.756952356364909e-06,
+      "loss": 0.3521,
+      "step": 2875
+    },
+    {
+      "epoch": 0.23008,
+      "grad_norm": 1.5316811800003052,
+      "learning_rate": 8.756122371828058e-06,
+      "loss": 0.3306,
+      "step": 2876
+    },
+    {
+      "epoch": 0.23016,
+      "grad_norm": 1.5814586877822876,
+      "learning_rate": 8.75529214965391e-06,
+      "loss": 0.2806,
+      "step": 2877
+    },
+    {
+      "epoch": 0.23024,
+      "grad_norm": 1.5847948789596558,
+      "learning_rate": 8.754461689894995e-06,
+      "loss": 0.4181,
+      "step": 2878
+    },
+    {
+      "epoch": 0.23032,
+      "grad_norm": 1.3571077585220337,
+      "learning_rate": 8.75363099260385e-06,
+      "loss": 0.3639,
+      "step": 2879
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 1.354897379875183,
+      "learning_rate": 8.752800057833033e-06,
+      "loss": 0.3231,
+      "step": 2880
+    },
+    {
+      "epoch": 0.23048,
+      "grad_norm": 2.0537118911743164,
+      "learning_rate": 8.751968885635115e-06,
+      "loss": 0.3662,
+      "step": 2881
+    },
+    {
+      "epoch": 0.23056,
+      "grad_norm": 1.585959792137146,
+      "learning_rate": 8.751137476062677e-06,
+      "loss": 0.392,
+      "step": 2882
+    },
+    {
+      "epoch": 0.23064,
+      "grad_norm": 2.2385966777801514,
+      "learning_rate": 8.750305829168321e-06,
+      "loss": 0.4388,
+      "step": 2883
+    },
+    {
+      "epoch": 0.23072,
+      "grad_norm": 1.3772207498550415,
+      "learning_rate": 8.749473945004665e-06,
+      "loss": 0.3189,
+      "step": 2884
+    },
+    {
+      "epoch": 0.2308,
+      "grad_norm": 1.298640251159668,
+      "learning_rate": 8.74864182362434e-06,
+      "loss": 0.2753,
+      "step": 2885
+    },
+    {
+      "epoch": 0.23088,
+      "grad_norm": 1.3288183212280273,
+      "learning_rate": 8.747809465079988e-06,
+      "loss": 0.297,
+      "step": 2886
+    },
+    {
+      "epoch": 0.23096,
+      "grad_norm": 1.2760984897613525,
+      "learning_rate": 8.746976869424272e-06,
+      "loss": 0.2857,
+      "step": 2887
+    },
+    {
+      "epoch": 0.23104,
+      "grad_norm": 1.6551486253738403,
+      "learning_rate": 8.746144036709867e-06,
+      "loss": 0.3029,
+      "step": 2888
+    },
+    {
+      "epoch": 0.23112,
+      "grad_norm": 1.5374372005462646,
+      "learning_rate": 8.745310966989463e-06,
+      "loss": 0.3214,
+      "step": 2889
+    },
+    {
+      "epoch": 0.2312,
+      "grad_norm": 2.0101284980773926,
+      "learning_rate": 8.744477660315767e-06,
+      "loss": 0.3779,
+      "step": 2890
+    },
+    {
+      "epoch": 0.23128,
+      "grad_norm": 1.6186890602111816,
+      "learning_rate": 8.743644116741497e-06,
+      "loss": 0.3475,
+      "step": 2891
+    },
+    {
+      "epoch": 0.23136,
+      "grad_norm": 1.7061872482299805,
+      "learning_rate": 8.742810336319391e-06,
+      "loss": 0.3468,
+      "step": 2892
+    },
+    {
+      "epoch": 0.23144,
+      "grad_norm": 1.7789872884750366,
+      "learning_rate": 8.741976319102198e-06,
+      "loss": 0.4333,
+      "step": 2893
+    },
+    {
+      "epoch": 0.23152,
+      "grad_norm": 1.6100202798843384,
+      "learning_rate": 8.741142065142683e-06,
+      "loss": 0.39,
+      "step": 2894
+    },
+    {
+      "epoch": 0.2316,
+      "grad_norm": 1.5869725942611694,
+      "learning_rate": 8.74030757449363e-06,
+      "loss": 0.3075,
+      "step": 2895
+    },
+    {
+      "epoch": 0.23168,
+      "grad_norm": 1.3948324918746948,
+      "learning_rate": 8.73947284720783e-06,
+      "loss": 0.2776,
+      "step": 2896
+    },
+    {
+      "epoch": 0.23176,
+      "grad_norm": 2.0415258407592773,
+      "learning_rate": 8.738637883338097e-06,
+      "loss": 0.3917,
+      "step": 2897
+    },
+    {
+      "epoch": 0.23184,
+      "grad_norm": 1.994744062423706,
+      "learning_rate": 8.737802682937253e-06,
+      "loss": 0.407,
+      "step": 2898
+    },
+    {
+      "epoch": 0.23192,
+      "grad_norm": 1.5355931520462036,
+      "learning_rate": 8.736967246058139e-06,
+      "loss": 0.3687,
+      "step": 2899
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 1.7060344219207764,
+      "learning_rate": 8.736131572753614e-06,
+      "loss": 0.326,
+      "step": 2900
+    },
+    {
+      "epoch": 0.23208,
+      "grad_norm": 1.27048921585083,
+      "learning_rate": 8.735295663076545e-06,
+      "loss": 0.2716,
+      "step": 2901
+    },
+    {
+      "epoch": 0.23216,
+      "grad_norm": 1.5977073907852173,
+      "learning_rate": 8.734459517079815e-06,
+      "loss": 0.3342,
+      "step": 2902
+    },
+    {
+      "epoch": 0.23224,
+      "grad_norm": 2.4220876693725586,
+      "learning_rate": 8.733623134816329e-06,
+      "loss": 0.4348,
+      "step": 2903
+    },
+    {
+      "epoch": 0.23232,
+      "grad_norm": 1.7467727661132812,
+      "learning_rate": 8.732786516339e-06,
+      "loss": 0.346,
+      "step": 2904
+    },
+    {
+      "epoch": 0.2324,
+      "grad_norm": 1.4733282327651978,
+      "learning_rate": 8.731949661700759e-06,
+      "loss": 0.3384,
+      "step": 2905
+    },
+    {
+      "epoch": 0.23248,
+      "grad_norm": 1.3760995864868164,
+      "learning_rate": 8.731112570954547e-06,
+      "loss": 0.3029,
+      "step": 2906
+    },
+    {
+      "epoch": 0.23256,
+      "grad_norm": 1.9718873500823975,
+      "learning_rate": 8.73027524415333e-06,
+      "loss": 0.395,
+      "step": 2907
+    },
+    {
+      "epoch": 0.23264,
+      "grad_norm": 1.4080829620361328,
+      "learning_rate": 8.729437681350078e-06,
+      "loss": 0.4309,
+      "step": 2908
+    },
+    {
+      "epoch": 0.23272,
+      "grad_norm": 1.7443673610687256,
+      "learning_rate": 8.728599882597784e-06,
+      "loss": 0.373,
+      "step": 2909
+    },
+    {
+      "epoch": 0.2328,
+      "grad_norm": 1.4104093313217163,
+      "learning_rate": 8.72776184794945e-06,
+      "loss": 0.2877,
+      "step": 2910
+    },
+    {
+      "epoch": 0.23288,
+      "grad_norm": 1.8802305459976196,
+      "learning_rate": 8.726923577458097e-06,
+      "loss": 0.2943,
+      "step": 2911
+    },
+    {
+      "epoch": 0.23296,
+      "grad_norm": 1.4886568784713745,
+      "learning_rate": 8.726085071176761e-06,
+      "loss": 0.3797,
+      "step": 2912
+    },
+    {
+      "epoch": 0.23304,
+      "grad_norm": 1.259099006652832,
+      "learning_rate": 8.72524632915849e-06,
+      "loss": 0.3082,
+      "step": 2913
+    },
+    {
+      "epoch": 0.23312,
+      "grad_norm": 1.6729494333267212,
+      "learning_rate": 8.724407351456348e-06,
+      "loss": 0.3901,
+      "step": 2914
+    },
+    {
+      "epoch": 0.2332,
+      "grad_norm": 2.0305817127227783,
+      "learning_rate": 8.723568138123414e-06,
+      "loss": 0.6386,
+      "step": 2915
+    },
+    {
+      "epoch": 0.23328,
+      "grad_norm": 1.870473027229309,
+      "learning_rate": 8.722728689212785e-06,
+      "loss": 0.4489,
+      "step": 2916
+    },
+    {
+      "epoch": 0.23336,
+      "grad_norm": 1.5763081312179565,
+      "learning_rate": 8.721889004777566e-06,
+      "loss": 0.3733,
+      "step": 2917
+    },
+    {
+      "epoch": 0.23344,
+      "grad_norm": 1.4396545886993408,
+      "learning_rate": 8.721049084870883e-06,
+      "loss": 0.3985,
+      "step": 2918
+    },
+    {
+      "epoch": 0.23352,
+      "grad_norm": 1.4830892086029053,
+      "learning_rate": 8.720208929545876e-06,
+      "loss": 0.2998,
+      "step": 2919
+    },
+    {
+      "epoch": 0.2336,
+      "grad_norm": 1.6304142475128174,
+      "learning_rate": 8.719368538855699e-06,
+      "loss": 0.4229,
+      "step": 2920
+    },
+    {
+      "epoch": 0.23368,
+      "grad_norm": 1.3104946613311768,
+      "learning_rate": 8.718527912853518e-06,
+      "loss": 0.2697,
+      "step": 2921
+    },
+    {
+      "epoch": 0.23376,
+      "grad_norm": 1.498120665550232,
+      "learning_rate": 8.717687051592518e-06,
+      "loss": 0.3604,
+      "step": 2922
+    },
+    {
+      "epoch": 0.23384,
+      "grad_norm": 2.0636587142944336,
+      "learning_rate": 8.716845955125899e-06,
+      "loss": 0.3904,
+      "step": 2923
+    },
+    {
+      "epoch": 0.23392,
+      "grad_norm": 1.651100516319275,
+      "learning_rate": 8.716004623506872e-06,
+      "loss": 0.3936,
+      "step": 2924
+    },
+    {
+      "epoch": 0.234,
+      "grad_norm": 1.8705461025238037,
+      "learning_rate": 8.715163056788666e-06,
+      "loss": 0.4764,
+      "step": 2925
+    },
+    {
+      "epoch": 0.23408,
+      "grad_norm": 1.6736024618148804,
+      "learning_rate": 8.714321255024525e-06,
+      "loss": 0.3558,
+      "step": 2926
+    },
+    {
+      "epoch": 0.23416,
+      "grad_norm": 1.4450464248657227,
+      "learning_rate": 8.713479218267707e-06,
+      "loss": 0.3201,
+      "step": 2927
+    },
+    {
+      "epoch": 0.23424,
+      "grad_norm": 1.6315438747406006,
+      "learning_rate": 8.712636946571484e-06,
+      "loss": 0.3774,
+      "step": 2928
+    },
+    {
+      "epoch": 0.23432,
+      "grad_norm": 2.0019845962524414,
+      "learning_rate": 8.711794439989142e-06,
+      "loss": 0.3845,
+      "step": 2929
+    },
+    {
+      "epoch": 0.2344,
+      "grad_norm": 1.9739534854888916,
+      "learning_rate": 8.710951698573987e-06,
+      "loss": 0.3859,
+      "step": 2930
+    },
+    {
+      "epoch": 0.23448,
+      "grad_norm": 1.6644498109817505,
+      "learning_rate": 8.710108722379335e-06,
+      "loss": 0.3413,
+      "step": 2931
+    },
+    {
+      "epoch": 0.23456,
+      "grad_norm": 1.7164329290390015,
+      "learning_rate": 8.709265511458518e-06,
+      "loss": 0.3603,
+      "step": 2932
+    },
+    {
+      "epoch": 0.23464,
+      "grad_norm": 1.5060267448425293,
+      "learning_rate": 8.708422065864884e-06,
+      "loss": 0.3308,
+      "step": 2933
+    },
+    {
+      "epoch": 0.23472,
+      "grad_norm": 1.650381326675415,
+      "learning_rate": 8.707578385651795e-06,
+      "loss": 0.3483,
+      "step": 2934
+    },
+    {
+      "epoch": 0.2348,
+      "grad_norm": 1.5644901990890503,
+      "learning_rate": 8.706734470872624e-06,
+      "loss": 0.2845,
+      "step": 2935
+    },
+    {
+      "epoch": 0.23488,
+      "grad_norm": 1.9320694208145142,
+      "learning_rate": 8.705890321580768e-06,
+      "loss": 0.3227,
+      "step": 2936
+    },
+    {
+      "epoch": 0.23496,
+      "grad_norm": 1.542862892150879,
+      "learning_rate": 8.70504593782963e-06,
+      "loss": 0.2855,
+      "step": 2937
+    },
+    {
+      "epoch": 0.23504,
+      "grad_norm": 1.6313329935073853,
+      "learning_rate": 8.704201319672635e-06,
+      "loss": 0.3229,
+      "step": 2938
+    },
+    {
+      "epoch": 0.23512,
+      "grad_norm": 2.1460344791412354,
+      "learning_rate": 8.703356467163214e-06,
+      "loss": 0.5213,
+      "step": 2939
+    },
+    {
+      "epoch": 0.2352,
+      "grad_norm": 1.938667893409729,
+      "learning_rate": 8.702511380354822e-06,
+      "loss": 0.4258,
+      "step": 2940
+    },
+    {
+      "epoch": 0.23528,
+      "grad_norm": 1.58048415184021,
+      "learning_rate": 8.701666059300924e-06,
+      "loss": 0.3398,
+      "step": 2941
+    },
+    {
+      "epoch": 0.23536,
+      "grad_norm": 1.4565937519073486,
+      "learning_rate": 8.700820504055e-06,
+      "loss": 0.3913,
+      "step": 2942
+    },
+    {
+      "epoch": 0.23544,
+      "grad_norm": 1.457868218421936,
+      "learning_rate": 8.699974714670544e-06,
+      "loss": 0.3067,
+      "step": 2943
+    },
+    {
+      "epoch": 0.23552,
+      "grad_norm": 1.6874083280563354,
+      "learning_rate": 8.699128691201071e-06,
+      "loss": 0.3866,
+      "step": 2944
+    },
+    {
+      "epoch": 0.2356,
+      "grad_norm": 1.3682913780212402,
+      "learning_rate": 8.698282433700102e-06,
+      "loss": 0.2925,
+      "step": 2945
+    },
+    {
+      "epoch": 0.23568,
+      "grad_norm": 1.2929877042770386,
+      "learning_rate": 8.697435942221178e-06,
+      "loss": 0.2813,
+      "step": 2946
+    },
+    {
+      "epoch": 0.23576,
+      "grad_norm": 1.2776330709457397,
+      "learning_rate": 8.696589216817852e-06,
+      "loss": 0.2464,
+      "step": 2947
+    },
+    {
+      "epoch": 0.23584,
+      "grad_norm": 1.6202374696731567,
+      "learning_rate": 8.695742257543697e-06,
+      "loss": 0.3886,
+      "step": 2948
+    },
+    {
+      "epoch": 0.23592,
+      "grad_norm": 1.200810432434082,
+      "learning_rate": 8.694895064452294e-06,
+      "loss": 0.3024,
+      "step": 2949
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 1.6930382251739502,
+      "learning_rate": 8.694047637597245e-06,
+      "loss": 0.3178,
+      "step": 2950
+    },
+    {
+      "epoch": 0.23608,
+      "grad_norm": 1.5066219568252563,
+      "learning_rate": 8.693199977032161e-06,
+      "loss": 0.322,
+      "step": 2951
+    },
+    {
+      "epoch": 0.23616,
+      "grad_norm": 1.304224967956543,
+      "learning_rate": 8.692352082810673e-06,
+      "loss": 0.3451,
+      "step": 2952
+    },
+    {
+      "epoch": 0.23624,
+      "grad_norm": 1.763218641281128,
+      "learning_rate": 8.691503954986422e-06,
+      "loss": 0.4435,
+      "step": 2953
+    },
+    {
+      "epoch": 0.23632,
+      "grad_norm": 1.316672444343567,
+      "learning_rate": 8.690655593613068e-06,
+      "loss": 0.3426,
+      "step": 2954
+    },
+    {
+      "epoch": 0.2364,
+      "grad_norm": 1.6360975503921509,
+      "learning_rate": 8.689806998744284e-06,
+      "loss": 0.3094,
+      "step": 2955
+    },
+    {
+      "epoch": 0.23648,
+      "grad_norm": 1.7397687435150146,
+      "learning_rate": 8.688958170433757e-06,
+      "loss": 0.4201,
+      "step": 2956
+    },
+    {
+      "epoch": 0.23656,
+      "grad_norm": 1.3750033378601074,
+      "learning_rate": 8.68810910873519e-06,
+      "loss": 0.2625,
+      "step": 2957
+    },
+    {
+      "epoch": 0.23664,
+      "grad_norm": 1.3705966472625732,
+      "learning_rate": 8.687259813702301e-06,
+      "loss": 0.3107,
+      "step": 2958
+    },
+    {
+      "epoch": 0.23672,
+      "grad_norm": 1.367472529411316,
+      "learning_rate": 8.686410285388818e-06,
+      "loss": 0.2897,
+      "step": 2959
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 1.3883439302444458,
+      "learning_rate": 8.685560523848494e-06,
+      "loss": 0.3076,
+      "step": 2960
+    },
+    {
+      "epoch": 0.23688,
+      "grad_norm": 1.2411656379699707,
+      "learning_rate": 8.684710529135088e-06,
+      "loss": 0.3676,
+      "step": 2961
+    },
+    {
+      "epoch": 0.23696,
+      "grad_norm": 1.4780466556549072,
+      "learning_rate": 8.683860301302373e-06,
+      "loss": 0.2878,
+      "step": 2962
+    },
+    {
+      "epoch": 0.23704,
+      "grad_norm": 1.2274631261825562,
+      "learning_rate": 8.683009840404145e-06,
+      "loss": 0.3174,
+      "step": 2963
+    },
+    {
+      "epoch": 0.23712,
+      "grad_norm": 1.504817247390747,
+      "learning_rate": 8.682159146494208e-06,
+      "loss": 0.291,
+      "step": 2964
+    },
+    {
+      "epoch": 0.2372,
+      "grad_norm": 1.6749637126922607,
+      "learning_rate": 8.681308219626381e-06,
+      "loss": 0.3828,
+      "step": 2965
+    },
+    {
+      "epoch": 0.23728,
+      "grad_norm": 1.835526943206787,
+      "learning_rate": 8.680457059854502e-06,
+      "loss": 0.3636,
+      "step": 2966
+    },
+    {
+      "epoch": 0.23736,
+      "grad_norm": 1.364960789680481,
+      "learning_rate": 8.679605667232421e-06,
+      "loss": 0.3038,
+      "step": 2967
+    },
+    {
+      "epoch": 0.23744,
+      "grad_norm": 1.8709944486618042,
+      "learning_rate": 8.678754041813996e-06,
+      "loss": 0.4753,
+      "step": 2968
+    },
+    {
+      "epoch": 0.23752,
+      "grad_norm": 1.5966829061508179,
+      "learning_rate": 8.677902183653117e-06,
+      "loss": 0.3359,
+      "step": 2969
+    },
+    {
+      "epoch": 0.2376,
+      "grad_norm": 1.4219958782196045,
+      "learning_rate": 8.677050092803671e-06,
+      "loss": 0.2726,
+      "step": 2970
+    },
+    {
+      "epoch": 0.23768,
+      "grad_norm": 1.6112638711929321,
+      "learning_rate": 8.67619776931957e-06,
+      "loss": 0.4072,
+      "step": 2971
+    },
+    {
+      "epoch": 0.23776,
+      "grad_norm": 1.330270767211914,
+      "learning_rate": 8.675345213254739e-06,
+      "loss": 0.3817,
+      "step": 2972
+    },
+    {
+      "epoch": 0.23784,
+      "grad_norm": 1.7107291221618652,
+      "learning_rate": 8.67449242466311e-06,
+      "loss": 0.3458,
+      "step": 2973
+    },
+    {
+      "epoch": 0.23792,
+      "grad_norm": 1.2672098875045776,
+      "learning_rate": 8.67363940359864e-06,
+      "loss": 0.2455,
+      "step": 2974
+    },
+    {
+      "epoch": 0.238,
+      "grad_norm": 1.589351773262024,
+      "learning_rate": 8.6727861501153e-06,
+      "loss": 0.3935,
+      "step": 2975
+    },
+    {
+      "epoch": 0.23808,
+      "grad_norm": 1.4220237731933594,
+      "learning_rate": 8.67193266426707e-06,
+      "loss": 0.3488,
+      "step": 2976
+    },
+    {
+      "epoch": 0.23816,
+      "grad_norm": 1.346812129020691,
+      "learning_rate": 8.671078946107942e-06,
+      "loss": 0.2996,
+      "step": 2977
+    },
+    {
+      "epoch": 0.23824,
+      "grad_norm": 1.5532695055007935,
+      "learning_rate": 8.670224995691937e-06,
+      "loss": 0.2827,
+      "step": 2978
+    },
+    {
+      "epoch": 0.23832,
+      "grad_norm": 1.6328967809677124,
+      "learning_rate": 8.669370813073076e-06,
+      "loss": 0.3164,
+      "step": 2979
+    },
+    {
+      "epoch": 0.2384,
+      "grad_norm": 1.4983983039855957,
+      "learning_rate": 8.6685163983054e-06,
+      "loss": 0.3014,
+      "step": 2980
+    },
+    {
+      "epoch": 0.23848,
+      "grad_norm": 1.8855558633804321,
+      "learning_rate": 8.667661751442967e-06,
+      "loss": 0.3764,
+      "step": 2981
+    },
+    {
+      "epoch": 0.23856,
+      "grad_norm": 1.2733697891235352,
+      "learning_rate": 8.666806872539848e-06,
+      "loss": 0.2521,
+      "step": 2982
+    },
+    {
+      "epoch": 0.23864,
+      "grad_norm": 1.7222651243209839,
+      "learning_rate": 8.665951761650126e-06,
+      "loss": 0.451,
+      "step": 2983
+    },
+    {
+      "epoch": 0.23872,
+      "grad_norm": 1.8912527561187744,
+      "learning_rate": 8.665096418827902e-06,
+      "loss": 0.4282,
+      "step": 2984
+    },
+    {
+      "epoch": 0.2388,
+      "grad_norm": 1.8017947673797607,
+      "learning_rate": 8.664240844127294e-06,
+      "loss": 0.414,
+      "step": 2985
+    },
+    {
+      "epoch": 0.23888,
+      "grad_norm": 1.7017481327056885,
+      "learning_rate": 8.663385037602425e-06,
+      "loss": 0.3998,
+      "step": 2986
+    },
+    {
+      "epoch": 0.23896,
+      "grad_norm": 1.7164137363433838,
+      "learning_rate": 8.662528999307445e-06,
+      "loss": 0.4701,
+      "step": 2987
+    },
+    {
+      "epoch": 0.23904,
+      "grad_norm": 1.832329273223877,
+      "learning_rate": 8.661672729296508e-06,
+      "loss": 0.368,
+      "step": 2988
+    },
+    {
+      "epoch": 0.23912,
+      "grad_norm": 1.7810895442962646,
+      "learning_rate": 8.660816227623791e-06,
+      "loss": 0.3759,
+      "step": 2989
+    },
+    {
+      "epoch": 0.2392,
+      "grad_norm": 1.597602128982544,
+      "learning_rate": 8.65995949434348e-06,
+      "loss": 0.3215,
+      "step": 2990
+    },
+    {
+      "epoch": 0.23928,
+      "grad_norm": 1.6633431911468506,
+      "learning_rate": 8.659102529509777e-06,
+      "loss": 0.3171,
+      "step": 2991
+    },
+    {
+      "epoch": 0.23936,
+      "grad_norm": 1.16764497756958,
+      "learning_rate": 8.6582453331769e-06,
+      "loss": 0.2655,
+      "step": 2992
+    },
+    {
+      "epoch": 0.23944,
+      "grad_norm": 1.6470662355422974,
+      "learning_rate": 8.657387905399085e-06,
+      "loss": 0.3766,
+      "step": 2993
+    },
+    {
+      "epoch": 0.23952,
+      "grad_norm": 1.707756519317627,
+      "learning_rate": 8.65653024623057e-06,
+      "loss": 0.3319,
+      "step": 2994
+    },
+    {
+      "epoch": 0.2396,
+      "grad_norm": 1.4462788105010986,
+      "learning_rate": 8.655672355725624e-06,
+      "loss": 0.3719,
+      "step": 2995
+    },
+    {
+      "epoch": 0.23968,
+      "grad_norm": 1.4448779821395874,
+      "learning_rate": 8.65481423393852e-06,
+      "loss": 0.3045,
+      "step": 2996
+    },
+    {
+      "epoch": 0.23976,
+      "grad_norm": 1.3186711072921753,
+      "learning_rate": 8.653955880923548e-06,
+      "loss": 0.2988,
+      "step": 2997
+    },
+    {
+      "epoch": 0.23984,
+      "grad_norm": 1.8978686332702637,
+      "learning_rate": 8.653097296735013e-06,
+      "loss": 0.4307,
+      "step": 2998
+    },
+    {
+      "epoch": 0.23992,
+      "grad_norm": 1.8591535091400146,
+      "learning_rate": 8.652238481427236e-06,
+      "loss": 0.4093,
+      "step": 2999
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7549595832824707,
+      "learning_rate": 8.65137943505455e-06,
+      "loss": 0.3208,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24008,
+      "grad_norm": 1.3385145664215088,
+      "learning_rate": 8.650520157671305e-06,
+      "loss": 0.3473,
+      "step": 3001
+    },
+    {
+      "epoch": 0.24016,
+      "grad_norm": 1.6653660535812378,
+      "learning_rate": 8.649660649331866e-06,
+      "loss": 0.4244,
+      "step": 3002
+    },
+    {
+      "epoch": 0.24024,
+      "grad_norm": 1.7460511922836304,
+      "learning_rate": 8.648800910090607e-06,
+      "loss": 0.4825,
+      "step": 3003
+    },
+    {
+      "epoch": 0.24032,
+      "grad_norm": 1.650795817375183,
+      "learning_rate": 8.647940940001925e-06,
+      "loss": 0.3525,
+      "step": 3004
+    },
+    {
+      "epoch": 0.2404,
+      "grad_norm": 1.116976261138916,
+      "learning_rate": 8.647080739120224e-06,
+      "loss": 0.2398,
+      "step": 3005
+    },
+    {
+      "epoch": 0.24048,
+      "grad_norm": 2.5539703369140625,
+      "learning_rate": 8.64622030749993e-06,
+      "loss": 0.5062,
+      "step": 3006
+    },
+    {
+      "epoch": 0.24056,
+      "grad_norm": 1.4344453811645508,
+      "learning_rate": 8.645359645195475e-06,
+      "loss": 0.4,
+      "step": 3007
+    },
+    {
+      "epoch": 0.24064,
+      "grad_norm": 1.8538168668746948,
+      "learning_rate": 8.644498752261314e-06,
+      "loss": 0.3884,
+      "step": 3008
+    },
+    {
+      "epoch": 0.24072,
+      "grad_norm": 1.4812930822372437,
+      "learning_rate": 8.643637628751912e-06,
+      "loss": 0.379,
+      "step": 3009
+    },
+    {
+      "epoch": 0.2408,
+      "grad_norm": 2.2699289321899414,
+      "learning_rate": 8.642776274721747e-06,
+      "loss": 0.4227,
+      "step": 3010
+    },
+    {
+      "epoch": 0.24088,
+      "grad_norm": 1.7083232402801514,
+      "learning_rate": 8.64191469022532e-06,
+      "loss": 0.4064,
+      "step": 3011
+    },
+    {
+      "epoch": 0.24096,
+      "grad_norm": 1.5670645236968994,
+      "learning_rate": 8.641052875317134e-06,
+      "loss": 0.4143,
+      "step": 3012
+    },
+    {
+      "epoch": 0.24104,
+      "grad_norm": 1.5622729063034058,
+      "learning_rate": 8.640190830051714e-06,
+      "loss": 0.3417,
+      "step": 3013
+    },
+    {
+      "epoch": 0.24112,
+      "grad_norm": 1.4791356325149536,
+      "learning_rate": 8.639328554483602e-06,
+      "loss": 0.3159,
+      "step": 3014
+    },
+    {
+      "epoch": 0.2412,
+      "grad_norm": 1.7441151142120361,
+      "learning_rate": 8.63846604866735e-06,
+      "loss": 0.3908,
+      "step": 3015
+    },
+    {
+      "epoch": 0.24128,
+      "grad_norm": 1.3369641304016113,
+      "learning_rate": 8.637603312657523e-06,
+      "loss": 0.3106,
+      "step": 3016
+    },
+    {
+      "epoch": 0.24136,
+      "grad_norm": 1.8151953220367432,
+      "learning_rate": 8.636740346508708e-06,
+      "loss": 0.3418,
+      "step": 3017
+    },
+    {
+      "epoch": 0.24144,
+      "grad_norm": 1.3240052461624146,
+      "learning_rate": 8.635877150275498e-06,
+      "loss": 0.2562,
+      "step": 3018
+    },
+    {
+      "epoch": 0.24152,
+      "grad_norm": 1.6432435512542725,
+      "learning_rate": 8.63501372401251e-06,
+      "loss": 0.4784,
+      "step": 3019
+    },
+    {
+      "epoch": 0.2416,
+      "grad_norm": 1.8531707525253296,
+      "learning_rate": 8.634150067774363e-06,
+      "loss": 0.4161,
+      "step": 3020
+    },
+    {
+      "epoch": 0.24168,
+      "grad_norm": 1.6908509731292725,
+      "learning_rate": 8.633286181615701e-06,
+      "loss": 0.3549,
+      "step": 3021
+    },
+    {
+      "epoch": 0.24176,
+      "grad_norm": 1.5330978631973267,
+      "learning_rate": 8.632422065591181e-06,
+      "loss": 0.3241,
+      "step": 3022
+    },
+    {
+      "epoch": 0.24184,
+      "grad_norm": 1.633192777633667,
+      "learning_rate": 8.63155771975547e-06,
+      "loss": 0.3212,
+      "step": 3023
+    },
+    {
+      "epoch": 0.24192,
+      "grad_norm": 1.6121563911437988,
+      "learning_rate": 8.630693144163255e-06,
+      "loss": 0.3842,
+      "step": 3024
+    },
+    {
+      "epoch": 0.242,
+      "grad_norm": 1.674614667892456,
+      "learning_rate": 8.629828338869232e-06,
+      "loss": 0.3152,
+      "step": 3025
+    },
+    {
+      "epoch": 0.24208,
+      "grad_norm": 1.8287391662597656,
+      "learning_rate": 8.628963303928115e-06,
+      "loss": 0.5112,
+      "step": 3026
+    },
+    {
+      "epoch": 0.24216,
+      "grad_norm": 1.853173017501831,
+      "learning_rate": 8.628098039394632e-06,
+      "loss": 0.3848,
+      "step": 3027
+    },
+    {
+      "epoch": 0.24224,
+      "grad_norm": 1.7024120092391968,
+      "learning_rate": 8.627232545323527e-06,
+      "loss": 0.3737,
+      "step": 3028
+    },
+    {
+      "epoch": 0.24232,
+      "grad_norm": 1.4395607709884644,
+      "learning_rate": 8.626366821769556e-06,
+      "loss": 0.3088,
+      "step": 3029
+    },
+    {
+      "epoch": 0.2424,
+      "grad_norm": 1.6949063539505005,
+      "learning_rate": 8.625500868787488e-06,
+      "loss": 0.4567,
+      "step": 3030
+    },
+    {
+      "epoch": 0.24248,
+      "grad_norm": 2.09765887260437,
+      "learning_rate": 8.624634686432112e-06,
+      "loss": 0.2887,
+      "step": 3031
+    },
+    {
+      "epoch": 0.24256,
+      "grad_norm": 1.5703338384628296,
+      "learning_rate": 8.623768274758228e-06,
+      "loss": 0.3327,
+      "step": 3032
+    },
+    {
+      "epoch": 0.24264,
+      "grad_norm": 1.6307145357131958,
+      "learning_rate": 8.62290163382065e-06,
+      "loss": 0.3305,
+      "step": 3033
+    },
+    {
+      "epoch": 0.24272,
+      "grad_norm": 1.5243542194366455,
+      "learning_rate": 8.622034763674207e-06,
+      "loss": 0.3326,
+      "step": 3034
+    },
+    {
+      "epoch": 0.2428,
+      "grad_norm": 1.9055639505386353,
+      "learning_rate": 8.621167664373746e-06,
+      "loss": 0.4019,
+      "step": 3035
+    },
+    {
+      "epoch": 0.24288,
+      "grad_norm": 1.4484702348709106,
+      "learning_rate": 8.620300335974122e-06,
+      "loss": 0.2865,
+      "step": 3036
+    },
+    {
+      "epoch": 0.24296,
+      "grad_norm": 1.721147894859314,
+      "learning_rate": 8.61943277853021e-06,
+      "loss": 0.4413,
+      "step": 3037
+    },
+    {
+      "epoch": 0.24304,
+      "grad_norm": 1.7040313482284546,
+      "learning_rate": 8.618564992096896e-06,
+      "loss": 0.2903,
+      "step": 3038
+    },
+    {
+      "epoch": 0.24312,
+      "grad_norm": 1.6943026781082153,
+      "learning_rate": 8.617696976729082e-06,
+      "loss": 0.3307,
+      "step": 3039
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 1.560989260673523,
+      "learning_rate": 8.616828732481687e-06,
+      "loss": 0.4324,
+      "step": 3040
+    },
+    {
+      "epoch": 0.24328,
+      "grad_norm": 1.2807221412658691,
+      "learning_rate": 8.61596025940964e-06,
+      "loss": 0.3329,
+      "step": 3041
+    },
+    {
+      "epoch": 0.24336,
+      "grad_norm": 1.504556655883789,
+      "learning_rate": 8.615091557567884e-06,
+      "loss": 0.303,
+      "step": 3042
+    },
+    {
+      "epoch": 0.24344,
+      "grad_norm": 1.388826847076416,
+      "learning_rate": 8.614222627011382e-06,
+      "loss": 0.3595,
+      "step": 3043
+    },
+    {
+      "epoch": 0.24352,
+      "grad_norm": 1.3632359504699707,
+      "learning_rate": 8.613353467795108e-06,
+      "loss": 0.2863,
+      "step": 3044
+    },
+    {
+      "epoch": 0.2436,
+      "grad_norm": 1.6326030492782593,
+      "learning_rate": 8.612484079974053e-06,
+      "loss": 0.3433,
+      "step": 3045
+    },
+    {
+      "epoch": 0.24368,
+      "grad_norm": 1.4086010456085205,
+      "learning_rate": 8.611614463603215e-06,
+      "loss": 0.4176,
+      "step": 3046
+    },
+    {
+      "epoch": 0.24376,
+      "grad_norm": 1.719990611076355,
+      "learning_rate": 8.610744618737614e-06,
+      "loss": 0.4066,
+      "step": 3047
+    },
+    {
+      "epoch": 0.24384,
+      "grad_norm": 1.6791647672653198,
+      "learning_rate": 8.609874545432285e-06,
+      "loss": 0.3398,
+      "step": 3048
+    },
+    {
+      "epoch": 0.24392,
+      "grad_norm": 1.7768301963806152,
+      "learning_rate": 8.60900424374227e-06,
+      "loss": 0.4071,
+      "step": 3049
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 1.4605388641357422,
+      "learning_rate": 8.608133713722631e-06,
+      "loss": 0.3426,
+      "step": 3050
+    },
+    {
+      "epoch": 0.24408,
+      "grad_norm": 1.4044535160064697,
+      "learning_rate": 8.607262955428449e-06,
+      "loss": 0.3431,
+      "step": 3051
+    },
+    {
+      "epoch": 0.24416,
+      "grad_norm": 1.5170437097549438,
+      "learning_rate": 8.606391968914807e-06,
+      "loss": 0.3865,
+      "step": 3052
+    },
+    {
+      "epoch": 0.24424,
+      "grad_norm": 1.5312057733535767,
+      "learning_rate": 8.60552075423681e-06,
+      "loss": 0.391,
+      "step": 3053
+    },
+    {
+      "epoch": 0.24432,
+      "grad_norm": 2.062711715698242,
+      "learning_rate": 8.604649311449583e-06,
+      "loss": 0.5652,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2444,
+      "grad_norm": 1.4739586114883423,
+      "learning_rate": 8.603777640608254e-06,
+      "loss": 0.3014,
+      "step": 3055
+    },
+    {
+      "epoch": 0.24448,
+      "grad_norm": 1.3225152492523193,
+      "learning_rate": 8.60290574176797e-06,
+      "loss": 0.311,
+      "step": 3056
+    },
+    {
+      "epoch": 0.24456,
+      "grad_norm": 1.7784897089004517,
+      "learning_rate": 8.602033614983898e-06,
+      "loss": 0.3726,
+      "step": 3057
+    },
+    {
+      "epoch": 0.24464,
+      "grad_norm": 1.7232739925384521,
+      "learning_rate": 8.60116126031121e-06,
+      "loss": 0.3162,
+      "step": 3058
+    },
+    {
+      "epoch": 0.24472,
+      "grad_norm": 1.1945064067840576,
+      "learning_rate": 8.6002886778051e-06,
+      "loss": 0.2356,
+      "step": 3059
+    },
+    {
+      "epoch": 0.2448,
+      "grad_norm": 1.5598511695861816,
+      "learning_rate": 8.59941586752077e-06,
+      "loss": 0.3743,
+      "step": 3060
+    },
+    {
+      "epoch": 0.24488,
+      "grad_norm": 1.2169926166534424,
+      "learning_rate": 8.598542829513444e-06,
+      "loss": 0.2614,
+      "step": 3061
+    },
+    {
+      "epoch": 0.24496,
+      "grad_norm": 1.6568950414657593,
+      "learning_rate": 8.597669563838353e-06,
+      "loss": 0.3015,
+      "step": 3062
+    },
+    {
+      "epoch": 0.24504,
+      "grad_norm": 1.6429921388626099,
+      "learning_rate": 8.596796070550746e-06,
+      "loss": 0.3532,
+      "step": 3063
+    },
+    {
+      "epoch": 0.24512,
+      "grad_norm": 1.6444114446640015,
+      "learning_rate": 8.595922349705886e-06,
+      "loss": 0.4508,
+      "step": 3064
+    },
+    {
+      "epoch": 0.2452,
+      "grad_norm": 1.5652337074279785,
+      "learning_rate": 8.595048401359052e-06,
+      "loss": 0.2787,
+      "step": 3065
+    },
+    {
+      "epoch": 0.24528,
+      "grad_norm": 2.0153632164001465,
+      "learning_rate": 8.594174225565535e-06,
+      "loss": 0.3774,
+      "step": 3066
+    },
+    {
+      "epoch": 0.24536,
+      "grad_norm": 1.5180140733718872,
+      "learning_rate": 8.59329982238064e-06,
+      "loss": 0.3183,
+      "step": 3067
+    },
+    {
+      "epoch": 0.24544,
+      "grad_norm": 1.8348703384399414,
+      "learning_rate": 8.592425191859687e-06,
+      "loss": 0.4139,
+      "step": 3068
+    },
+    {
+      "epoch": 0.24552,
+      "grad_norm": 1.6924543380737305,
+      "learning_rate": 8.591550334058015e-06,
+      "loss": 0.3641,
+      "step": 3069
+    },
+    {
+      "epoch": 0.2456,
+      "grad_norm": 1.5413908958435059,
+      "learning_rate": 8.59067524903097e-06,
+      "loss": 0.3124,
+      "step": 3070
+    },
+    {
+      "epoch": 0.24568,
+      "grad_norm": 1.7611223459243774,
+      "learning_rate": 8.589799936833916e-06,
+      "loss": 0.397,
+      "step": 3071
+    },
+    {
+      "epoch": 0.24576,
+      "grad_norm": 1.5739551782608032,
+      "learning_rate": 8.588924397522231e-06,
+      "loss": 0.2906,
+      "step": 3072
+    },
+    {
+      "epoch": 0.24584,
+      "grad_norm": 1.6537935733795166,
+      "learning_rate": 8.58804863115131e-06,
+      "loss": 0.3932,
+      "step": 3073
+    },
+    {
+      "epoch": 0.24592,
+      "grad_norm": 1.7853214740753174,
+      "learning_rate": 8.587172637776558e-06,
+      "loss": 0.3289,
+      "step": 3074
+    },
+    {
+      "epoch": 0.246,
+      "grad_norm": 1.2469711303710938,
+      "learning_rate": 8.586296417453393e-06,
+      "loss": 0.3285,
+      "step": 3075
+    },
+    {
+      "epoch": 0.24608,
+      "grad_norm": 1.7327982187271118,
+      "learning_rate": 8.585419970237255e-06,
+      "loss": 0.5739,
+      "step": 3076
+    },
+    {
+      "epoch": 0.24616,
+      "grad_norm": 1.298710823059082,
+      "learning_rate": 8.584543296183591e-06,
+      "loss": 0.2513,
+      "step": 3077
+    },
+    {
+      "epoch": 0.24624,
+      "grad_norm": 1.2274887561798096,
+      "learning_rate": 8.583666395347869e-06,
+      "loss": 0.2949,
+      "step": 3078
+    },
+    {
+      "epoch": 0.24632,
+      "grad_norm": 1.9516730308532715,
+      "learning_rate": 8.582789267785563e-06,
+      "loss": 0.4306,
+      "step": 3079
+    },
+    {
+      "epoch": 0.2464,
+      "grad_norm": 1.424416422843933,
+      "learning_rate": 8.58191191355217e-06,
+      "loss": 0.3194,
+      "step": 3080
+    },
+    {
+      "epoch": 0.24648,
+      "grad_norm": 1.8527755737304688,
+      "learning_rate": 8.581034332703194e-06,
+      "loss": 0.3808,
+      "step": 3081
+    },
+    {
+      "epoch": 0.24656,
+      "grad_norm": 1.4947212934494019,
+      "learning_rate": 8.58015652529416e-06,
+      "loss": 0.2791,
+      "step": 3082
+    },
+    {
+      "epoch": 0.24664,
+      "grad_norm": 1.3371108770370483,
+      "learning_rate": 8.579278491380598e-06,
+      "loss": 0.2267,
+      "step": 3083
+    },
+    {
+      "epoch": 0.24672,
+      "grad_norm": 1.5755013227462769,
+      "learning_rate": 8.578400231018064e-06,
+      "loss": 0.3123,
+      "step": 3084
+    },
+    {
+      "epoch": 0.2468,
+      "grad_norm": 1.653031349182129,
+      "learning_rate": 8.577521744262123e-06,
+      "loss": 0.4084,
+      "step": 3085
+    },
+    {
+      "epoch": 0.24688,
+      "grad_norm": 1.827981948852539,
+      "learning_rate": 8.576643031168349e-06,
+      "loss": 0.3714,
+      "step": 3086
+    },
+    {
+      "epoch": 0.24696,
+      "grad_norm": 1.6476799249649048,
+      "learning_rate": 8.575764091792339e-06,
+      "loss": 0.2948,
+      "step": 3087
+    },
+    {
+      "epoch": 0.24704,
+      "grad_norm": 1.5807143449783325,
+      "learning_rate": 8.574884926189699e-06,
+      "loss": 0.4274,
+      "step": 3088
+    },
+    {
+      "epoch": 0.24712,
+      "grad_norm": 1.801981806755066,
+      "learning_rate": 8.57400553441605e-06,
+      "loss": 0.4747,
+      "step": 3089
+    },
+    {
+      "epoch": 0.2472,
+      "grad_norm": 1.1952424049377441,
+      "learning_rate": 8.573125916527031e-06,
+      "loss": 0.2432,
+      "step": 3090
+    },
+    {
+      "epoch": 0.24728,
+      "grad_norm": 1.6386176347732544,
+      "learning_rate": 8.572246072578292e-06,
+      "loss": 0.3894,
+      "step": 3091
+    },
+    {
+      "epoch": 0.24736,
+      "grad_norm": 1.36851966381073,
+      "learning_rate": 8.571366002625495e-06,
+      "loss": 0.2686,
+      "step": 3092
+    },
+    {
+      "epoch": 0.24744,
+      "grad_norm": 1.3977208137512207,
+      "learning_rate": 8.570485706724322e-06,
+      "loss": 0.3077,
+      "step": 3093
+    },
+    {
+      "epoch": 0.24752,
+      "grad_norm": 1.9915634393692017,
+      "learning_rate": 8.569605184930466e-06,
+      "loss": 0.3775,
+      "step": 3094
+    },
+    {
+      "epoch": 0.2476,
+      "grad_norm": 1.7627999782562256,
+      "learning_rate": 8.568724437299631e-06,
+      "loss": 0.3882,
+      "step": 3095
+    },
+    {
+      "epoch": 0.24768,
+      "grad_norm": 1.5895828008651733,
+      "learning_rate": 8.567843463887543e-06,
+      "loss": 0.2905,
+      "step": 3096
+    },
+    {
+      "epoch": 0.24776,
+      "grad_norm": 1.7350879907608032,
+      "learning_rate": 8.566962264749938e-06,
+      "loss": 0.3652,
+      "step": 3097
+    },
+    {
+      "epoch": 0.24784,
+      "grad_norm": 1.7022144794464111,
+      "learning_rate": 8.566080839942566e-06,
+      "loss": 0.3455,
+      "step": 3098
+    },
+    {
+      "epoch": 0.24792,
+      "grad_norm": 1.8170467615127563,
+      "learning_rate": 8.565199189521189e-06,
+      "loss": 0.3867,
+      "step": 3099
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 1.3393418788909912,
+      "learning_rate": 8.56431731354159e-06,
+      "loss": 0.2957,
+      "step": 3100
+    },
+    {
+      "epoch": 0.24808,
+      "grad_norm": 1.4289052486419678,
+      "learning_rate": 8.563435212059561e-06,
+      "loss": 0.3263,
+      "step": 3101
+    },
+    {
+      "epoch": 0.24816,
+      "grad_norm": 1.3951023817062378,
+      "learning_rate": 8.56255288513091e-06,
+      "loss": 0.2906,
+      "step": 3102
+    },
+    {
+      "epoch": 0.24824,
+      "grad_norm": 1.709173560142517,
+      "learning_rate": 8.561670332811458e-06,
+      "loss": 0.2987,
+      "step": 3103
+    },
+    {
+      "epoch": 0.24832,
+      "grad_norm": 2.1143503189086914,
+      "learning_rate": 8.56078755515704e-06,
+      "loss": 0.4867,
+      "step": 3104
+    },
+    {
+      "epoch": 0.2484,
+      "grad_norm": 1.202492356300354,
+      "learning_rate": 8.55990455222351e-06,
+      "loss": 0.2769,
+      "step": 3105
+    },
+    {
+      "epoch": 0.24848,
+      "grad_norm": 1.666141390800476,
+      "learning_rate": 8.559021324066728e-06,
+      "loss": 0.3279,
+      "step": 3106
+    },
+    {
+      "epoch": 0.24856,
+      "grad_norm": 1.3046966791152954,
+      "learning_rate": 8.558137870742578e-06,
+      "loss": 0.3452,
+      "step": 3107
+    },
+    {
+      "epoch": 0.24864,
+      "grad_norm": 1.476283073425293,
+      "learning_rate": 8.557254192306948e-06,
+      "loss": 0.3225,
+      "step": 3108
+    },
+    {
+      "epoch": 0.24872,
+      "grad_norm": 1.7318915128707886,
+      "learning_rate": 8.55637028881575e-06,
+      "loss": 0.3913,
+      "step": 3109
+    },
+    {
+      "epoch": 0.2488,
+      "grad_norm": 1.5203393697738647,
+      "learning_rate": 8.555486160324902e-06,
+      "loss": 0.3077,
+      "step": 3110
+    },
+    {
+      "epoch": 0.24888,
+      "grad_norm": 1.610252857208252,
+      "learning_rate": 8.554601806890342e-06,
+      "loss": 0.3474,
+      "step": 3111
+    },
+    {
+      "epoch": 0.24896,
+      "grad_norm": 1.8684039115905762,
+      "learning_rate": 8.553717228568018e-06,
+      "loss": 0.361,
+      "step": 3112
+    },
+    {
+      "epoch": 0.24904,
+      "grad_norm": 1.8766143321990967,
+      "learning_rate": 8.552832425413897e-06,
+      "loss": 0.3453,
+      "step": 3113
+    },
+    {
+      "epoch": 0.24912,
+      "grad_norm": 1.7946739196777344,
+      "learning_rate": 8.551947397483957e-06,
+      "loss": 0.4859,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2492,
+      "grad_norm": 1.4358093738555908,
+      "learning_rate": 8.551062144834189e-06,
+      "loss": 0.3495,
+      "step": 3115
+    },
+    {
+      "epoch": 0.24928,
+      "grad_norm": 1.4946582317352295,
+      "learning_rate": 8.5501766675206e-06,
+      "loss": 0.361,
+      "step": 3116
+    },
+    {
+      "epoch": 0.24936,
+      "grad_norm": 1.362602710723877,
+      "learning_rate": 8.549290965599214e-06,
+      "loss": 0.3362,
+      "step": 3117
+    },
+    {
+      "epoch": 0.24944,
+      "grad_norm": 1.9473742246627808,
+      "learning_rate": 8.548405039126064e-06,
+      "loss": 0.4263,
+      "step": 3118
+    },
+    {
+      "epoch": 0.24952,
+      "grad_norm": 1.4761608839035034,
+      "learning_rate": 8.5475188881572e-06,
+      "loss": 0.3398,
+      "step": 3119
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 1.3862004280090332,
+      "learning_rate": 8.546632512748685e-06,
+      "loss": 0.361,
+      "step": 3120
+    },
+    {
+      "epoch": 0.24968,
+      "grad_norm": 1.507232904434204,
+      "learning_rate": 8.5457459129566e-06,
+      "loss": 0.2987,
+      "step": 3121
+    },
+    {
+      "epoch": 0.24976,
+      "grad_norm": 1.8091908693313599,
+      "learning_rate": 8.544859088837034e-06,
+      "loss": 0.3306,
+      "step": 3122
+    },
+    {
+      "epoch": 0.24984,
+      "grad_norm": 1.4990782737731934,
+      "learning_rate": 8.543972040446094e-06,
+      "loss": 0.2953,
+      "step": 3123
+    },
+    {
+      "epoch": 0.24992,
+      "grad_norm": 1.6368714570999146,
+      "learning_rate": 8.543084767839903e-06,
+      "loss": 0.2681,
+      "step": 3124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.792515516281128,
+      "learning_rate": 8.542197271074593e-06,
+      "loss": 0.3714,
+      "step": 3125
+    },
+    {
+      "epoch": 0.25008,
+      "grad_norm": 1.680100679397583,
+      "learning_rate": 8.541309550206314e-06,
+      "loss": 0.4091,
+      "step": 3126
+    },
+    {
+      "epoch": 0.25016,
+      "grad_norm": 1.2095801830291748,
+      "learning_rate": 8.540421605291228e-06,
+      "loss": 0.2428,
+      "step": 3127
+    },
+    {
+      "epoch": 0.25024,
+      "grad_norm": 1.3905729055404663,
+      "learning_rate": 8.539533436385515e-06,
+      "loss": 0.2728,
+      "step": 3128
+    },
+    {
+      "epoch": 0.25032,
+      "grad_norm": 1.7047653198242188,
+      "learning_rate": 8.538645043545364e-06,
+      "loss": 0.3631,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2504,
+      "grad_norm": 1.328326940536499,
+      "learning_rate": 8.537756426826981e-06,
+      "loss": 0.2764,
+      "step": 3130
+    },
+    {
+      "epoch": 0.25048,
+      "grad_norm": 1.6265760660171509,
+      "learning_rate": 8.536867586286586e-06,
+      "loss": 0.3045,
+      "step": 3131
+    },
+    {
+      "epoch": 0.25056,
+      "grad_norm": 1.6370584964752197,
+      "learning_rate": 8.535978521980414e-06,
+      "loss": 0.3782,
+      "step": 3132
+    },
+    {
+      "epoch": 0.25064,
+      "grad_norm": 1.4649031162261963,
+      "learning_rate": 8.535089233964712e-06,
+      "loss": 0.3088,
+      "step": 3133
+    },
+    {
+      "epoch": 0.25072,
+      "grad_norm": 1.2999662160873413,
+      "learning_rate": 8.534199722295744e-06,
+      "loss": 0.2541,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2508,
+      "grad_norm": 1.5978385210037231,
+      "learning_rate": 8.533309987029783e-06,
+      "loss": 0.3905,
+      "step": 3135
+    },
+    {
+      "epoch": 0.25088,
+      "grad_norm": 1.4170856475830078,
+      "learning_rate": 8.532420028223122e-06,
+      "loss": 0.3053,
+      "step": 3136
+    },
+    {
+      "epoch": 0.25096,
+      "grad_norm": 1.6570608615875244,
+      "learning_rate": 8.531529845932066e-06,
+      "loss": 0.3299,
+      "step": 3137
+    },
+    {
+      "epoch": 0.25104,
+      "grad_norm": 1.5469874143600464,
+      "learning_rate": 8.530639440212934e-06,
+      "loss": 0.3261,
+      "step": 3138
+    },
+    {
+      "epoch": 0.25112,
+      "grad_norm": 1.5680407285690308,
+      "learning_rate": 8.529748811122056e-06,
+      "loss": 0.3176,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2512,
+      "grad_norm": 1.4539591073989868,
+      "learning_rate": 8.528857958715783e-06,
+      "loss": 0.3613,
+      "step": 3140
+    },
+    {
+      "epoch": 0.25128,
+      "grad_norm": 1.8934903144836426,
+      "learning_rate": 8.527966883050474e-06,
+      "loss": 0.445,
+      "step": 3141
+    },
+    {
+      "epoch": 0.25136,
+      "grad_norm": 1.4876171350479126,
+      "learning_rate": 8.527075584182505e-06,
+      "loss": 0.2891,
+      "step": 3142
+    },
+    {
+      "epoch": 0.25144,
+      "grad_norm": 1.2558088302612305,
+      "learning_rate": 8.526184062168264e-06,
+      "loss": 0.3297,
+      "step": 3143
+    },
+    {
+      "epoch": 0.25152,
+      "grad_norm": 1.8519783020019531,
+      "learning_rate": 8.525292317064157e-06,
+      "loss": 0.3826,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2516,
+      "grad_norm": 1.5815104246139526,
+      "learning_rate": 8.524400348926602e-06,
+      "loss": 0.3588,
+      "step": 3145
+    },
+    {
+      "epoch": 0.25168,
+      "grad_norm": 1.5304956436157227,
+      "learning_rate": 8.523508157812029e-06,
+      "loss": 0.3735,
+      "step": 3146
+    },
+    {
+      "epoch": 0.25176,
+      "grad_norm": 1.5469937324523926,
+      "learning_rate": 8.522615743776885e-06,
+      "loss": 0.3096,
+      "step": 3147
+    },
+    {
+      "epoch": 0.25184,
+      "grad_norm": 1.3774428367614746,
+      "learning_rate": 8.52172310687763e-06,
+      "loss": 0.248,
+      "step": 3148
+    },
+    {
+      "epoch": 0.25192,
+      "grad_norm": 1.1401811838150024,
+      "learning_rate": 8.520830247170735e-06,
+      "loss": 0.2834,
+      "step": 3149
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 1.574325680732727,
+      "learning_rate": 8.519937164712691e-06,
+      "loss": 0.3037,
+      "step": 3150
+    },
+    {
+      "epoch": 0.25208,
+      "grad_norm": 1.6470081806182861,
+      "learning_rate": 8.519043859560004e-06,
+      "loss": 0.3799,
+      "step": 3151
+    },
+    {
+      "epoch": 0.25216,
+      "grad_norm": 1.6649773120880127,
+      "learning_rate": 8.518150331769184e-06,
+      "loss": 0.3009,
+      "step": 3152
+    },
+    {
+      "epoch": 0.25224,
+      "grad_norm": 1.4246584177017212,
+      "learning_rate": 8.517256581396764e-06,
+      "loss": 0.3222,
+      "step": 3153
+    },
+    {
+      "epoch": 0.25232,
+      "grad_norm": 1.5069494247436523,
+      "learning_rate": 8.51636260849929e-06,
+      "loss": 0.3699,
+      "step": 3154
+    },
+    {
+      "epoch": 0.2524,
+      "grad_norm": 1.4292433261871338,
+      "learning_rate": 8.51546841313332e-06,
+      "loss": 0.4126,
+      "step": 3155
+    },
+    {
+      "epoch": 0.25248,
+      "grad_norm": 1.3793959617614746,
+      "learning_rate": 8.514573995355426e-06,
+      "loss": 0.2454,
+      "step": 3156
+    },
+    {
+      "epoch": 0.25256,
+      "grad_norm": 1.3602819442749023,
+      "learning_rate": 8.513679355222195e-06,
+      "loss": 0.3414,
+      "step": 3157
+    },
+    {
+      "epoch": 0.25264,
+      "grad_norm": 1.6190898418426514,
+      "learning_rate": 8.512784492790227e-06,
+      "loss": 0.3326,
+      "step": 3158
+    },
+    {
+      "epoch": 0.25272,
+      "grad_norm": 2.0170743465423584,
+      "learning_rate": 8.511889408116138e-06,
+      "loss": 0.3803,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2528,
+      "grad_norm": 1.572218418121338,
+      "learning_rate": 8.510994101256557e-06,
+      "loss": 0.3277,
+      "step": 3160
+    },
+    {
+      "epoch": 0.25288,
+      "grad_norm": 1.3365064859390259,
+      "learning_rate": 8.510098572268129e-06,
+      "loss": 0.2734,
+      "step": 3161
+    },
+    {
+      "epoch": 0.25296,
+      "grad_norm": 1.7582436800003052,
+      "learning_rate": 8.509202821207508e-06,
+      "loss": 0.4216,
+      "step": 3162
+    },
+    {
+      "epoch": 0.25304,
+      "grad_norm": 1.665613055229187,
+      "learning_rate": 8.508306848131367e-06,
+      "loss": 0.3075,
+      "step": 3163
+    },
+    {
+      "epoch": 0.25312,
+      "grad_norm": 1.382347822189331,
+      "learning_rate": 8.50741065309639e-06,
+      "loss": 0.3831,
+      "step": 3164
+    },
+    {
+      "epoch": 0.2532,
+      "grad_norm": 1.4488331079483032,
+      "learning_rate": 8.506514236159276e-06,
+      "loss": 0.353,
+      "step": 3165
+    },
+    {
+      "epoch": 0.25328,
+      "grad_norm": 1.5848164558410645,
+      "learning_rate": 8.505617597376739e-06,
+      "loss": 0.4441,
+      "step": 3166
+    },
+    {
+      "epoch": 0.25336,
+      "grad_norm": 1.5356500148773193,
+      "learning_rate": 8.504720736805507e-06,
+      "loss": 0.3197,
+      "step": 3167
+    },
+    {
+      "epoch": 0.25344,
+      "grad_norm": 1.2928869724273682,
+      "learning_rate": 8.50382365450232e-06,
+      "loss": 0.2967,
+      "step": 3168
+    },
+    {
+      "epoch": 0.25352,
+      "grad_norm": 1.418664813041687,
+      "learning_rate": 8.502926350523937e-06,
+      "loss": 0.2878,
+      "step": 3169
+    },
+    {
+      "epoch": 0.2536,
+      "grad_norm": 2.2894961833953857,
+      "learning_rate": 8.502028824927123e-06,
+      "loss": 0.5619,
+      "step": 3170
+    },
+    {
+      "epoch": 0.25368,
+      "grad_norm": 1.8666496276855469,
+      "learning_rate": 8.501131077768664e-06,
+      "loss": 0.3883,
+      "step": 3171
+    },
+    {
+      "epoch": 0.25376,
+      "grad_norm": 1.9164226055145264,
+      "learning_rate": 8.500233109105354e-06,
+      "loss": 0.3319,
+      "step": 3172
+    },
+    {
+      "epoch": 0.25384,
+      "grad_norm": 1.4399014711380005,
+      "learning_rate": 8.499334918994008e-06,
+      "loss": 0.3641,
+      "step": 3173
+    },
+    {
+      "epoch": 0.25392,
+      "grad_norm": 1.7369312047958374,
+      "learning_rate": 8.498436507491452e-06,
+      "loss": 0.3709,
+      "step": 3174
+    },
+    {
+      "epoch": 0.254,
+      "grad_norm": 1.5380709171295166,
+      "learning_rate": 8.497537874654523e-06,
+      "loss": 0.3893,
+      "step": 3175
+    },
+    {
+      "epoch": 0.25408,
+      "grad_norm": 1.6948041915893555,
+      "learning_rate": 8.496639020540074e-06,
+      "loss": 0.3261,
+      "step": 3176
+    },
+    {
+      "epoch": 0.25416,
+      "grad_norm": 1.0962127447128296,
+      "learning_rate": 8.495739945204975e-06,
+      "loss": 0.2629,
+      "step": 3177
+    },
+    {
+      "epoch": 0.25424,
+      "grad_norm": 1.5826691389083862,
+      "learning_rate": 8.494840648706107e-06,
+      "loss": 0.4263,
+      "step": 3178
+    },
+    {
+      "epoch": 0.25432,
+      "grad_norm": 1.7018965482711792,
+      "learning_rate": 8.493941131100365e-06,
+      "loss": 0.471,
+      "step": 3179
+    },
+    {
+      "epoch": 0.2544,
+      "grad_norm": 2.042285680770874,
+      "learning_rate": 8.49304139244466e-06,
+      "loss": 0.4853,
+      "step": 3180
+    },
+    {
+      "epoch": 0.25448,
+      "grad_norm": 1.3213742971420288,
+      "learning_rate": 8.492141432795911e-06,
+      "loss": 0.2682,
+      "step": 3181
+    },
+    {
+      "epoch": 0.25456,
+      "grad_norm": 1.6487033367156982,
+      "learning_rate": 8.491241252211058e-06,
+      "loss": 0.3989,
+      "step": 3182
+    },
+    {
+      "epoch": 0.25464,
+      "grad_norm": 1.4631835222244263,
+      "learning_rate": 8.490340850747055e-06,
+      "loss": 0.3634,
+      "step": 3183
+    },
+    {
+      "epoch": 0.25472,
+      "grad_norm": 1.5799041986465454,
+      "learning_rate": 8.489440228460864e-06,
+      "loss": 0.4085,
+      "step": 3184
+    },
+    {
+      "epoch": 0.2548,
+      "grad_norm": 1.6044838428497314,
+      "learning_rate": 8.488539385409464e-06,
+      "loss": 0.3399,
+      "step": 3185
+    },
+    {
+      "epoch": 0.25488,
+      "grad_norm": 1.7360122203826904,
+      "learning_rate": 8.48763832164985e-06,
+      "loss": 0.3473,
+      "step": 3186
+    },
+    {
+      "epoch": 0.25496,
+      "grad_norm": 1.6504510641098022,
+      "learning_rate": 8.486737037239032e-06,
+      "loss": 0.3314,
+      "step": 3187
+    },
+    {
+      "epoch": 0.25504,
+      "grad_norm": 1.5512864589691162,
+      "learning_rate": 8.485835532234027e-06,
+      "loss": 0.3203,
+      "step": 3188
+    },
+    {
+      "epoch": 0.25512,
+      "grad_norm": 1.515463948249817,
+      "learning_rate": 8.48493380669187e-06,
+      "loss": 0.3878,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2552,
+      "grad_norm": 1.3933261632919312,
+      "learning_rate": 8.484031860669612e-06,
+      "loss": 0.3515,
+      "step": 3190
+    },
+    {
+      "epoch": 0.25528,
+      "grad_norm": 1.280632495880127,
+      "learning_rate": 8.483129694224319e-06,
+      "loss": 0.3116,
+      "step": 3191
+    },
+    {
+      "epoch": 0.25536,
+      "grad_norm": 1.7904398441314697,
+      "learning_rate": 8.482227307413063e-06,
+      "loss": 0.4518,
+      "step": 3192
+    },
+    {
+      "epoch": 0.25544,
+      "grad_norm": 1.5730119943618774,
+      "learning_rate": 8.481324700292934e-06,
+      "loss": 0.3031,
+      "step": 3193
+    },
+    {
+      "epoch": 0.25552,
+      "grad_norm": 1.5001622438430786,
+      "learning_rate": 8.480421872921042e-06,
+      "loss": 0.2442,
+      "step": 3194
+    },
+    {
+      "epoch": 0.2556,
+      "grad_norm": 1.639999508857727,
+      "learning_rate": 8.479518825354504e-06,
+      "loss": 0.3146,
+      "step": 3195
+    },
+    {
+      "epoch": 0.25568,
+      "grad_norm": 1.664425253868103,
+      "learning_rate": 8.478615557650453e-06,
+      "loss": 0.2826,
+      "step": 3196
+    },
+    {
+      "epoch": 0.25576,
+      "grad_norm": 1.624760389328003,
+      "learning_rate": 8.477712069866033e-06,
+      "loss": 0.3324,
+      "step": 3197
+    },
+    {
+      "epoch": 0.25584,
+      "grad_norm": 1.328769564628601,
+      "learning_rate": 8.47680836205841e-06,
+      "loss": 0.296,
+      "step": 3198
+    },
+    {
+      "epoch": 0.25592,
+      "grad_norm": 1.3259910345077515,
+      "learning_rate": 8.475904434284752e-06,
+      "loss": 0.2968,
+      "step": 3199
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 1.7926405668258667,
+      "learning_rate": 8.475000286602254e-06,
+      "loss": 0.447,
+      "step": 3200
+    },
+    {
+      "epoch": 0.25608,
+      "grad_norm": 1.6081920862197876,
+      "learning_rate": 8.47409591906811e-06,
+      "loss": 0.3946,
+      "step": 3201
+    },
+    {
+      "epoch": 0.25616,
+      "grad_norm": 1.7535005807876587,
+      "learning_rate": 8.473191331739547e-06,
+      "loss": 0.4252,
+      "step": 3202
+    },
+    {
+      "epoch": 0.25624,
+      "grad_norm": 1.5326111316680908,
+      "learning_rate": 8.472286524673787e-06,
+      "loss": 0.423,
+      "step": 3203
+    },
+    {
+      "epoch": 0.25632,
+      "grad_norm": 1.3280184268951416,
+      "learning_rate": 8.471381497928079e-06,
+      "loss": 0.2994,
+      "step": 3204
+    },
+    {
+      "epoch": 0.2564,
+      "grad_norm": 1.3200703859329224,
+      "learning_rate": 8.470476251559677e-06,
+      "loss": 0.3823,
+      "step": 3205
+    },
+    {
+      "epoch": 0.25648,
+      "grad_norm": 1.8249784708023071,
+      "learning_rate": 8.469570785625856e-06,
+      "loss": 0.4093,
+      "step": 3206
+    },
+    {
+      "epoch": 0.25656,
+      "grad_norm": 1.7080583572387695,
+      "learning_rate": 8.4686651001839e-06,
+      "loss": 0.4029,
+      "step": 3207
+    },
+    {
+      "epoch": 0.25664,
+      "grad_norm": 1.537379503250122,
+      "learning_rate": 8.467759195291108e-06,
+      "loss": 0.3258,
+      "step": 3208
+    },
+    {
+      "epoch": 0.25672,
+      "grad_norm": 1.783140778541565,
+      "learning_rate": 8.466853071004797e-06,
+      "loss": 0.3672,
+      "step": 3209
+    },
+    {
+      "epoch": 0.2568,
+      "grad_norm": 1.411805272102356,
+      "learning_rate": 8.465946727382293e-06,
+      "loss": 0.2902,
+      "step": 3210
+    },
+    {
+      "epoch": 0.25688,
+      "grad_norm": 1.7587497234344482,
+      "learning_rate": 8.465040164480934e-06,
+      "loss": 0.3364,
+      "step": 3211
+    },
+    {
+      "epoch": 0.25696,
+      "grad_norm": 2.055811882019043,
+      "learning_rate": 8.46413338235808e-06,
+      "loss": 0.4274,
+      "step": 3212
+    },
+    {
+      "epoch": 0.25704,
+      "grad_norm": 1.577392339706421,
+      "learning_rate": 8.463226381071095e-06,
+      "loss": 0.2775,
+      "step": 3213
+    },
+    {
+      "epoch": 0.25712,
+      "grad_norm": 1.8780300617218018,
+      "learning_rate": 8.462319160677368e-06,
+      "loss": 0.3771,
+      "step": 3214
+    },
+    {
+      "epoch": 0.2572,
+      "grad_norm": 1.4582608938217163,
+      "learning_rate": 8.461411721234292e-06,
+      "loss": 0.4151,
+      "step": 3215
+    },
+    {
+      "epoch": 0.25728,
+      "grad_norm": 1.8028523921966553,
+      "learning_rate": 8.460504062799277e-06,
+      "loss": 0.3596,
+      "step": 3216
+    },
+    {
+      "epoch": 0.25736,
+      "grad_norm": 1.2800477743148804,
+      "learning_rate": 8.459596185429751e-06,
+      "loss": 0.2512,
+      "step": 3217
+    },
+    {
+      "epoch": 0.25744,
+      "grad_norm": 1.3028494119644165,
+      "learning_rate": 8.458688089183147e-06,
+      "loss": 0.3026,
+      "step": 3218
+    },
+    {
+      "epoch": 0.25752,
+      "grad_norm": 1.5475552082061768,
+      "learning_rate": 8.457779774116924e-06,
+      "loss": 0.2819,
+      "step": 3219
+    },
+    {
+      "epoch": 0.2576,
+      "grad_norm": 1.6385855674743652,
+      "learning_rate": 8.456871240288542e-06,
+      "loss": 0.4,
+      "step": 3220
+    },
+    {
+      "epoch": 0.25768,
+      "grad_norm": 1.2522298097610474,
+      "learning_rate": 8.455962487755484e-06,
+      "loss": 0.3563,
+      "step": 3221
+    },
+    {
+      "epoch": 0.25776,
+      "grad_norm": 1.9725730419158936,
+      "learning_rate": 8.455053516575243e-06,
+      "loss": 0.3775,
+      "step": 3222
+    },
+    {
+      "epoch": 0.25784,
+      "grad_norm": 1.417087197303772,
+      "learning_rate": 8.454144326805328e-06,
+      "loss": 0.3495,
+      "step": 3223
+    },
+    {
+      "epoch": 0.25792,
+      "grad_norm": 1.4763423204421997,
+      "learning_rate": 8.453234918503257e-06,
+      "loss": 0.3413,
+      "step": 3224
+    },
+    {
+      "epoch": 0.258,
+      "grad_norm": 1.2496310472488403,
+      "learning_rate": 8.452325291726567e-06,
+      "loss": 0.2642,
+      "step": 3225
+    },
+    {
+      "epoch": 0.25808,
+      "grad_norm": 2.0870649814605713,
+      "learning_rate": 8.45141544653281e-06,
+      "loss": 0.4066,
+      "step": 3226
+    },
+    {
+      "epoch": 0.25816,
+      "grad_norm": 1.2118961811065674,
+      "learning_rate": 8.450505382979544e-06,
+      "loss": 0.2908,
+      "step": 3227
+    },
+    {
+      "epoch": 0.25824,
+      "grad_norm": 1.6424560546875,
+      "learning_rate": 8.449595101124349e-06,
+      "loss": 0.2963,
+      "step": 3228
+    },
+    {
+      "epoch": 0.25832,
+      "grad_norm": 1.646569013595581,
+      "learning_rate": 8.448684601024812e-06,
+      "loss": 0.3907,
+      "step": 3229
+    },
+    {
+      "epoch": 0.2584,
+      "grad_norm": 1.6771202087402344,
+      "learning_rate": 8.447773882738542e-06,
+      "loss": 0.3657,
+      "step": 3230
+    },
+    {
+      "epoch": 0.25848,
+      "grad_norm": 2.0412819385528564,
+      "learning_rate": 8.446862946323154e-06,
+      "loss": 0.4464,
+      "step": 3231
+    },
+    {
+      "epoch": 0.25856,
+      "grad_norm": 1.4120683670043945,
+      "learning_rate": 8.44595179183628e-06,
+      "loss": 0.3551,
+      "step": 3232
+    },
+    {
+      "epoch": 0.25864,
+      "grad_norm": 1.4363163709640503,
+      "learning_rate": 8.445040419335569e-06,
+      "loss": 0.3243,
+      "step": 3233
+    },
+    {
+      "epoch": 0.25872,
+      "grad_norm": 1.8272966146469116,
+      "learning_rate": 8.444128828878676e-06,
+      "loss": 0.3876,
+      "step": 3234
+    },
+    {
+      "epoch": 0.2588,
+      "grad_norm": 1.5481266975402832,
+      "learning_rate": 8.443217020523275e-06,
+      "loss": 0.4025,
+      "step": 3235
+    },
+    {
+      "epoch": 0.25888,
+      "grad_norm": 1.6646552085876465,
+      "learning_rate": 8.442304994327055e-06,
+      "loss": 0.3478,
+      "step": 3236
+    },
+    {
+      "epoch": 0.25896,
+      "grad_norm": 1.440777063369751,
+      "learning_rate": 8.441392750347716e-06,
+      "loss": 0.2645,
+      "step": 3237
+    },
+    {
+      "epoch": 0.25904,
+      "grad_norm": 1.4575594663619995,
+      "learning_rate": 8.440480288642972e-06,
+      "loss": 0.3257,
+      "step": 3238
+    },
+    {
+      "epoch": 0.25912,
+      "grad_norm": 1.6355897188186646,
+      "learning_rate": 8.439567609270554e-06,
+      "loss": 0.3219,
+      "step": 3239
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 1.4267311096191406,
+      "learning_rate": 8.438654712288202e-06,
+      "loss": 0.3609,
+      "step": 3240
+    },
+    {
+      "epoch": 0.25928,
+      "grad_norm": 1.911621332168579,
+      "learning_rate": 8.43774159775367e-06,
+      "loss": 0.4765,
+      "step": 3241
+    },
+    {
+      "epoch": 0.25936,
+      "grad_norm": 2.092682123184204,
+      "learning_rate": 8.43682826572473e-06,
+      "loss": 0.4551,
+      "step": 3242
+    },
+    {
+      "epoch": 0.25944,
+      "grad_norm": 1.7436370849609375,
+      "learning_rate": 8.435914716259166e-06,
+      "loss": 0.3244,
+      "step": 3243
+    },
+    {
+      "epoch": 0.25952,
+      "grad_norm": 1.278278112411499,
+      "learning_rate": 8.435000949414775e-06,
+      "loss": 0.2589,
+      "step": 3244
+    },
+    {
+      "epoch": 0.2596,
+      "grad_norm": 1.5885848999023438,
+      "learning_rate": 8.434086965249368e-06,
+      "loss": 0.445,
+      "step": 3245
+    },
+    {
+      "epoch": 0.25968,
+      "grad_norm": 1.544396996498108,
+      "learning_rate": 8.433172763820767e-06,
+      "loss": 0.406,
+      "step": 3246
+    },
+    {
+      "epoch": 0.25976,
+      "grad_norm": 1.5454270839691162,
+      "learning_rate": 8.432258345186815e-06,
+      "loss": 0.3159,
+      "step": 3247
+    },
+    {
+      "epoch": 0.25984,
+      "grad_norm": 1.3528130054473877,
+      "learning_rate": 8.43134370940536e-06,
+      "loss": 0.2816,
+      "step": 3248
+    },
+    {
+      "epoch": 0.25992,
+      "grad_norm": 1.8098217248916626,
+      "learning_rate": 8.430428856534271e-06,
+      "loss": 0.3795,
+      "step": 3249
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2492495775222778,
+      "learning_rate": 8.429513786631428e-06,
+      "loss": 0.2786,
+      "step": 3250
+    },
+    {
+      "epoch": 0.26008,
+      "grad_norm": 1.5359301567077637,
+      "learning_rate": 8.42859849975472e-06,
+      "loss": 0.3571,
+      "step": 3251
+    },
+    {
+      "epoch": 0.26016,
+      "grad_norm": 1.8178164958953857,
+      "learning_rate": 8.427682995962058e-06,
+      "loss": 0.41,
+      "step": 3252
+    },
+    {
+      "epoch": 0.26024,
+      "grad_norm": 1.6950284242630005,
+      "learning_rate": 8.426767275311361e-06,
+      "loss": 0.2952,
+      "step": 3253
+    },
+    {
+      "epoch": 0.26032,
+      "grad_norm": 1.3474059104919434,
+      "learning_rate": 8.425851337860566e-06,
+      "loss": 0.3301,
+      "step": 3254
+    },
+    {
+      "epoch": 0.2604,
+      "grad_norm": 1.5049055814743042,
+      "learning_rate": 8.42493518366762e-06,
+      "loss": 0.3448,
+      "step": 3255
+    },
+    {
+      "epoch": 0.26048,
+      "grad_norm": 1.99711275100708,
+      "learning_rate": 8.424018812790484e-06,
+      "loss": 0.3592,
+      "step": 3256
+    },
+    {
+      "epoch": 0.26056,
+      "grad_norm": 1.4437679052352905,
+      "learning_rate": 8.423102225287135e-06,
+      "loss": 0.3078,
+      "step": 3257
+    },
+    {
+      "epoch": 0.26064,
+      "grad_norm": 2.1142630577087402,
+      "learning_rate": 8.42218542121556e-06,
+      "loss": 0.5487,
+      "step": 3258
+    },
+    {
+      "epoch": 0.26072,
+      "grad_norm": 1.680812120437622,
+      "learning_rate": 8.421268400633766e-06,
+      "loss": 0.4302,
+      "step": 3259
+    },
+    {
+      "epoch": 0.2608,
+      "grad_norm": 1.1650234460830688,
+      "learning_rate": 8.420351163599768e-06,
+      "loss": 0.2241,
+      "step": 3260
+    },
+    {
+      "epoch": 0.26088,
+      "grad_norm": 1.5910313129425049,
+      "learning_rate": 8.419433710171596e-06,
+      "loss": 0.2969,
+      "step": 3261
+    },
+    {
+      "epoch": 0.26096,
+      "grad_norm": 1.5094048976898193,
+      "learning_rate": 8.418516040407295e-06,
+      "loss": 0.4483,
+      "step": 3262
+    },
+    {
+      "epoch": 0.26104,
+      "grad_norm": 1.2878527641296387,
+      "learning_rate": 8.417598154364923e-06,
+      "loss": 0.2714,
+      "step": 3263
+    },
+    {
+      "epoch": 0.26112,
+      "grad_norm": 1.6986520290374756,
+      "learning_rate": 8.416680052102549e-06,
+      "loss": 0.388,
+      "step": 3264
+    },
+    {
+      "epoch": 0.2612,
+      "grad_norm": 1.396224021911621,
+      "learning_rate": 8.415761733678262e-06,
+      "loss": 0.3414,
+      "step": 3265
+    },
+    {
+      "epoch": 0.26128,
+      "grad_norm": 1.9210736751556396,
+      "learning_rate": 8.414843199150159e-06,
+      "loss": 0.4595,
+      "step": 3266
+    },
+    {
+      "epoch": 0.26136,
+      "grad_norm": 1.2947478294372559,
+      "learning_rate": 8.413924448576351e-06,
+      "loss": 0.2695,
+      "step": 3267
+    },
+    {
+      "epoch": 0.26144,
+      "grad_norm": 1.2264394760131836,
+      "learning_rate": 8.41300548201497e-06,
+      "loss": 0.3017,
+      "step": 3268
+    },
+    {
+      "epoch": 0.26152,
+      "grad_norm": 1.5077235698699951,
+      "learning_rate": 8.412086299524149e-06,
+      "loss": 0.3803,
+      "step": 3269
+    },
+    {
+      "epoch": 0.2616,
+      "grad_norm": 1.603265643119812,
+      "learning_rate": 8.411166901162046e-06,
+      "loss": 0.3499,
+      "step": 3270
+    },
+    {
+      "epoch": 0.26168,
+      "grad_norm": 1.7666727304458618,
+      "learning_rate": 8.410247286986827e-06,
+      "loss": 0.3375,
+      "step": 3271
+    },
+    {
+      "epoch": 0.26176,
+      "grad_norm": 1.5865931510925293,
+      "learning_rate": 8.409327457056673e-06,
+      "loss": 0.3626,
+      "step": 3272
+    },
+    {
+      "epoch": 0.26184,
+      "grad_norm": 1.4339234828948975,
+      "learning_rate": 8.408407411429777e-06,
+      "loss": 0.3711,
+      "step": 3273
+    },
+    {
+      "epoch": 0.26192,
+      "grad_norm": 1.5342092514038086,
+      "learning_rate": 8.40748715016435e-06,
+      "loss": 0.3039,
+      "step": 3274
+    },
+    {
+      "epoch": 0.262,
+      "grad_norm": 1.602797269821167,
+      "learning_rate": 8.406566673318613e-06,
+      "loss": 0.3913,
+      "step": 3275
+    },
+    {
+      "epoch": 0.26208,
+      "grad_norm": 1.554508924484253,
+      "learning_rate": 8.405645980950799e-06,
+      "loss": 0.3879,
+      "step": 3276
+    },
+    {
+      "epoch": 0.26216,
+      "grad_norm": 1.9122850894927979,
+      "learning_rate": 8.40472507311916e-06,
+      "loss": 0.4212,
+      "step": 3277
+    },
+    {
+      "epoch": 0.26224,
+      "grad_norm": 1.8808695077896118,
+      "learning_rate": 8.403803949881958e-06,
+      "loss": 0.4648,
+      "step": 3278
+    },
+    {
+      "epoch": 0.26232,
+      "grad_norm": 1.2711478471755981,
+      "learning_rate": 8.402882611297469e-06,
+      "loss": 0.2795,
+      "step": 3279
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 1.6705055236816406,
+      "learning_rate": 8.401961057423984e-06,
+      "loss": 0.4656,
+      "step": 3280
+    },
+    {
+      "epoch": 0.26248,
+      "grad_norm": 1.4070430994033813,
+      "learning_rate": 8.401039288319805e-06,
+      "loss": 0.3143,
+      "step": 3281
+    },
+    {
+      "epoch": 0.26256,
+      "grad_norm": 1.4044073820114136,
+      "learning_rate": 8.400117304043252e-06,
+      "loss": 0.2966,
+      "step": 3282
+    },
+    {
+      "epoch": 0.26264,
+      "grad_norm": 1.5478609800338745,
+      "learning_rate": 8.399195104652652e-06,
+      "loss": 0.371,
+      "step": 3283
+    },
+    {
+      "epoch": 0.26272,
+      "grad_norm": 1.9816876649856567,
+      "learning_rate": 8.398272690206353e-06,
+      "loss": 0.3998,
+      "step": 3284
+    },
+    {
+      "epoch": 0.2628,
+      "grad_norm": 1.6786158084869385,
+      "learning_rate": 8.39735006076271e-06,
+      "loss": 0.3962,
+      "step": 3285
+    },
+    {
+      "epoch": 0.26288,
+      "grad_norm": 1.7463001012802124,
+      "learning_rate": 8.396427216380096e-06,
+      "loss": 0.3922,
+      "step": 3286
+    },
+    {
+      "epoch": 0.26296,
+      "grad_norm": 1.5134419202804565,
+      "learning_rate": 8.395504157116899e-06,
+      "loss": 0.2912,
+      "step": 3287
+    },
+    {
+      "epoch": 0.26304,
+      "grad_norm": 1.6220186948776245,
+      "learning_rate": 8.394580883031512e-06,
+      "loss": 0.4388,
+      "step": 3288
+    },
+    {
+      "epoch": 0.26312,
+      "grad_norm": 2.0490458011627197,
+      "learning_rate": 8.393657394182354e-06,
+      "loss": 0.524,
+      "step": 3289
+    },
+    {
+      "epoch": 0.2632,
+      "grad_norm": 1.5475538969039917,
+      "learning_rate": 8.392733690627846e-06,
+      "loss": 0.475,
+      "step": 3290
+    },
+    {
+      "epoch": 0.26328,
+      "grad_norm": 1.4094207286834717,
+      "learning_rate": 8.391809772426429e-06,
+      "loss": 0.3468,
+      "step": 3291
+    },
+    {
+      "epoch": 0.26336,
+      "grad_norm": 1.0840815305709839,
+      "learning_rate": 8.39088563963656e-06,
+      "loss": 0.2516,
+      "step": 3292
+    },
+    {
+      "epoch": 0.26344,
+      "grad_norm": 1.739927053451538,
+      "learning_rate": 8.389961292316699e-06,
+      "loss": 0.4647,
+      "step": 3293
+    },
+    {
+      "epoch": 0.26352,
+      "grad_norm": 1.712862491607666,
+      "learning_rate": 8.389036730525331e-06,
+      "loss": 0.3649,
+      "step": 3294
+    },
+    {
+      "epoch": 0.2636,
+      "grad_norm": 1.7053585052490234,
+      "learning_rate": 8.38811195432095e-06,
+      "loss": 0.4469,
+      "step": 3295
+    },
+    {
+      "epoch": 0.26368,
+      "grad_norm": 1.1163685321807861,
+      "learning_rate": 8.38718696376206e-06,
+      "loss": 0.2379,
+      "step": 3296
+    },
+    {
+      "epoch": 0.26376,
+      "grad_norm": 1.718429446220398,
+      "learning_rate": 8.386261758907187e-06,
+      "loss": 0.4064,
+      "step": 3297
+    },
+    {
+      "epoch": 0.26384,
+      "grad_norm": 1.4358896017074585,
+      "learning_rate": 8.38533633981486e-06,
+      "loss": 0.2584,
+      "step": 3298
+    },
+    {
+      "epoch": 0.26392,
+      "grad_norm": 1.6465083360671997,
+      "learning_rate": 8.384410706543632e-06,
+      "loss": 0.3218,
+      "step": 3299
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 1.5639690160751343,
+      "learning_rate": 8.383484859152062e-06,
+      "loss": 0.4031,
+      "step": 3300
+    },
+    {
+      "epoch": 0.26408,
+      "grad_norm": 1.5496549606323242,
+      "learning_rate": 8.382558797698728e-06,
+      "loss": 0.3324,
+      "step": 3301
+    },
+    {
+      "epoch": 0.26416,
+      "grad_norm": 1.8462837934494019,
+      "learning_rate": 8.381632522242215e-06,
+      "loss": 0.3972,
+      "step": 3302
+    },
+    {
+      "epoch": 0.26424,
+      "grad_norm": 1.6575276851654053,
+      "learning_rate": 8.380706032841128e-06,
+      "loss": 0.4347,
+      "step": 3303
+    },
+    {
+      "epoch": 0.26432,
+      "grad_norm": 1.0943249464035034,
+      "learning_rate": 8.379779329554082e-06,
+      "loss": 0.2716,
+      "step": 3304
+    },
+    {
+      "epoch": 0.2644,
+      "grad_norm": 2.2363994121551514,
+      "learning_rate": 8.378852412439706e-06,
+      "loss": 0.3581,
+      "step": 3305
+    },
+    {
+      "epoch": 0.26448,
+      "grad_norm": 1.4077491760253906,
+      "learning_rate": 8.377925281556645e-06,
+      "loss": 0.2796,
+      "step": 3306
+    },
+    {
+      "epoch": 0.26456,
+      "grad_norm": 1.9200001955032349,
+      "learning_rate": 8.376997936963553e-06,
+      "loss": 0.431,
+      "step": 3307
+    },
+    {
+      "epoch": 0.26464,
+      "grad_norm": 1.444860577583313,
+      "learning_rate": 8.3760703787191e-06,
+      "loss": 0.2795,
+      "step": 3308
+    },
+    {
+      "epoch": 0.26472,
+      "grad_norm": 1.153676152229309,
+      "learning_rate": 8.375142606881973e-06,
+      "loss": 0.2588,
+      "step": 3309
+    },
+    {
+      "epoch": 0.2648,
+      "grad_norm": 2.0317533016204834,
+      "learning_rate": 8.374214621510866e-06,
+      "loss": 0.4834,
+      "step": 3310
+    },
+    {
+      "epoch": 0.26488,
+      "grad_norm": 1.2731088399887085,
+      "learning_rate": 8.373286422664489e-06,
+      "loss": 0.2232,
+      "step": 3311
+    },
+    {
+      "epoch": 0.26496,
+      "grad_norm": 1.506540298461914,
+      "learning_rate": 8.372358010401568e-06,
+      "loss": 0.3614,
+      "step": 3312
+    },
+    {
+      "epoch": 0.26504,
+      "grad_norm": 1.827217698097229,
+      "learning_rate": 8.371429384780839e-06,
+      "loss": 0.4667,
+      "step": 3313
+    },
+    {
+      "epoch": 0.26512,
+      "grad_norm": 1.9941977262496948,
+      "learning_rate": 8.370500545861053e-06,
+      "loss": 0.413,
+      "step": 3314
+    },
+    {
+      "epoch": 0.2652,
+      "grad_norm": 1.5264027118682861,
+      "learning_rate": 8.369571493700976e-06,
+      "loss": 0.3107,
+      "step": 3315
+    },
+    {
+      "epoch": 0.26528,
+      "grad_norm": 1.9555363655090332,
+      "learning_rate": 8.368642228359384e-06,
+      "loss": 0.3539,
+      "step": 3316
+    },
+    {
+      "epoch": 0.26536,
+      "grad_norm": 1.4721167087554932,
+      "learning_rate": 8.367712749895072e-06,
+      "loss": 0.3781,
+      "step": 3317
+    },
+    {
+      "epoch": 0.26544,
+      "grad_norm": 1.2694438695907593,
+      "learning_rate": 8.366783058366843e-06,
+      "loss": 0.279,
+      "step": 3318
+    },
+    {
+      "epoch": 0.26552,
+      "grad_norm": 1.8869990110397339,
+      "learning_rate": 8.365853153833511e-06,
+      "loss": 0.3966,
+      "step": 3319
+    },
+    {
+      "epoch": 0.2656,
+      "grad_norm": 1.7108945846557617,
+      "learning_rate": 8.364923036353915e-06,
+      "loss": 0.3471,
+      "step": 3320
+    },
+    {
+      "epoch": 0.26568,
+      "grad_norm": 1.5163145065307617,
+      "learning_rate": 8.3639927059869e-06,
+      "loss": 0.2747,
+      "step": 3321
+    },
+    {
+      "epoch": 0.26576,
+      "grad_norm": 1.5786588191986084,
+      "learning_rate": 8.36306216279132e-06,
+      "loss": 0.3084,
+      "step": 3322
+    },
+    {
+      "epoch": 0.26584,
+      "grad_norm": 1.5403141975402832,
+      "learning_rate": 8.362131406826049e-06,
+      "loss": 0.3665,
+      "step": 3323
+    },
+    {
+      "epoch": 0.26592,
+      "grad_norm": 1.248185634613037,
+      "learning_rate": 8.361200438149975e-06,
+      "loss": 0.2885,
+      "step": 3324
+    },
+    {
+      "epoch": 0.266,
+      "grad_norm": 1.7556192874908447,
+      "learning_rate": 8.360269256821994e-06,
+      "loss": 0.3611,
+      "step": 3325
+    },
+    {
+      "epoch": 0.26608,
+      "grad_norm": 1.7010356187820435,
+      "learning_rate": 8.359337862901023e-06,
+      "loss": 0.4007,
+      "step": 3326
+    },
+    {
+      "epoch": 0.26616,
+      "grad_norm": 1.3344570398330688,
+      "learning_rate": 8.358406256445985e-06,
+      "loss": 0.2836,
+      "step": 3327
+    },
+    {
+      "epoch": 0.26624,
+      "grad_norm": 1.2783137559890747,
+      "learning_rate": 8.357474437515819e-06,
+      "loss": 0.3394,
+      "step": 3328
+    },
+    {
+      "epoch": 0.26632,
+      "grad_norm": 1.2508137226104736,
+      "learning_rate": 8.356542406169481e-06,
+      "loss": 0.301,
+      "step": 3329
+    },
+    {
+      "epoch": 0.2664,
+      "grad_norm": 1.4016166925430298,
+      "learning_rate": 8.355610162465935e-06,
+      "loss": 0.2951,
+      "step": 3330
+    },
+    {
+      "epoch": 0.26648,
+      "grad_norm": 1.7911545038223267,
+      "learning_rate": 8.354677706464162e-06,
+      "loss": 0.4171,
+      "step": 3331
+    },
+    {
+      "epoch": 0.26656,
+      "grad_norm": 1.3079147338867188,
+      "learning_rate": 8.353745038223155e-06,
+      "loss": 0.2591,
+      "step": 3332
+    },
+    {
+      "epoch": 0.26664,
+      "grad_norm": 1.5401325225830078,
+      "learning_rate": 8.352812157801923e-06,
+      "loss": 0.3765,
+      "step": 3333
+    },
+    {
+      "epoch": 0.26672,
+      "grad_norm": 1.556061863899231,
+      "learning_rate": 8.351879065259484e-06,
+      "loss": 0.3329,
+      "step": 3334
+    },
+    {
+      "epoch": 0.2668,
+      "grad_norm": 1.2529242038726807,
+      "learning_rate": 8.35094576065487e-06,
+      "loss": 0.2728,
+      "step": 3335
+    },
+    {
+      "epoch": 0.26688,
+      "grad_norm": 1.6587908267974854,
+      "learning_rate": 8.350012244047132e-06,
+      "loss": 0.3042,
+      "step": 3336
+    },
+    {
+      "epoch": 0.26696,
+      "grad_norm": 1.5234625339508057,
+      "learning_rate": 8.349078515495327e-06,
+      "loss": 0.3707,
+      "step": 3337
+    },
+    {
+      "epoch": 0.26704,
+      "grad_norm": 1.4544942378997803,
+      "learning_rate": 8.348144575058531e-06,
+      "loss": 0.3395,
+      "step": 3338
+    },
+    {
+      "epoch": 0.26712,
+      "grad_norm": 1.9101989269256592,
+      "learning_rate": 8.347210422795831e-06,
+      "loss": 0.3822,
+      "step": 3339
+    },
+    {
+      "epoch": 0.2672,
+      "grad_norm": 1.5211460590362549,
+      "learning_rate": 8.346276058766327e-06,
+      "loss": 0.3944,
+      "step": 3340
+    },
+    {
+      "epoch": 0.26728,
+      "grad_norm": 1.3909361362457275,
+      "learning_rate": 8.345341483029134e-06,
+      "loss": 0.2667,
+      "step": 3341
+    },
+    {
+      "epoch": 0.26736,
+      "grad_norm": 1.3863193988800049,
+      "learning_rate": 8.344406695643378e-06,
+      "loss": 0.2917,
+      "step": 3342
+    },
+    {
+      "epoch": 0.26744,
+      "grad_norm": 1.6209913492202759,
+      "learning_rate": 8.343471696668202e-06,
+      "loss": 0.3255,
+      "step": 3343
+    },
+    {
+      "epoch": 0.26752,
+      "grad_norm": 1.6851791143417358,
+      "learning_rate": 8.342536486162758e-06,
+      "loss": 0.4018,
+      "step": 3344
+    },
+    {
+      "epoch": 0.2676,
+      "grad_norm": 1.5533232688903809,
+      "learning_rate": 8.341601064186215e-06,
+      "loss": 0.3628,
+      "step": 3345
+    },
+    {
+      "epoch": 0.26768,
+      "grad_norm": 1.481484055519104,
+      "learning_rate": 8.340665430797752e-06,
+      "loss": 0.3407,
+      "step": 3346
+    },
+    {
+      "epoch": 0.26776,
+      "grad_norm": 1.3775691986083984,
+      "learning_rate": 8.339729586056567e-06,
+      "loss": 0.3287,
+      "step": 3347
+    },
+    {
+      "epoch": 0.26784,
+      "grad_norm": 1.6615140438079834,
+      "learning_rate": 8.338793530021866e-06,
+      "loss": 0.4026,
+      "step": 3348
+    },
+    {
+      "epoch": 0.26792,
+      "grad_norm": 1.7780884504318237,
+      "learning_rate": 8.337857262752869e-06,
+      "loss": 0.3734,
+      "step": 3349
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 1.2541722059249878,
+      "learning_rate": 8.336920784308814e-06,
+      "loss": 0.232,
+      "step": 3350
+    },
+    {
+      "epoch": 0.26808,
+      "grad_norm": 1.472259521484375,
+      "learning_rate": 8.335984094748944e-06,
+      "loss": 0.2674,
+      "step": 3351
+    },
+    {
+      "epoch": 0.26816,
+      "grad_norm": 1.723530888557434,
+      "learning_rate": 8.335047194132522e-06,
+      "loss": 0.3773,
+      "step": 3352
+    },
+    {
+      "epoch": 0.26824,
+      "grad_norm": 1.5939034223556519,
+      "learning_rate": 8.334110082518825e-06,
+      "loss": 0.4006,
+      "step": 3353
+    },
+    {
+      "epoch": 0.26832,
+      "grad_norm": 1.7733458280563354,
+      "learning_rate": 8.333172759967137e-06,
+      "loss": 0.3455,
+      "step": 3354
+    },
+    {
+      "epoch": 0.2684,
+      "grad_norm": 1.562965750694275,
+      "learning_rate": 8.332235226536762e-06,
+      "loss": 0.2928,
+      "step": 3355
+    },
+    {
+      "epoch": 0.26848,
+      "grad_norm": 1.6739492416381836,
+      "learning_rate": 8.331297482287016e-06,
+      "loss": 0.3414,
+      "step": 3356
+    },
+    {
+      "epoch": 0.26856,
+      "grad_norm": 1.6316815614700317,
+      "learning_rate": 8.330359527277224e-06,
+      "loss": 0.3977,
+      "step": 3357
+    },
+    {
+      "epoch": 0.26864,
+      "grad_norm": 1.0808318853378296,
+      "learning_rate": 8.329421361566724e-06,
+      "loss": 0.2338,
+      "step": 3358
+    },
+    {
+      "epoch": 0.26872,
+      "grad_norm": 1.3807647228240967,
+      "learning_rate": 8.328482985214879e-06,
+      "loss": 0.246,
+      "step": 3359
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 1.602565050125122,
+      "learning_rate": 8.327544398281053e-06,
+      "loss": 0.3755,
+      "step": 3360
+    },
+    {
+      "epoch": 0.26888,
+      "grad_norm": 1.6400891542434692,
+      "learning_rate": 8.326605600824625e-06,
+      "loss": 0.3009,
+      "step": 3361
+    },
+    {
+      "epoch": 0.26896,
+      "grad_norm": 1.5679333209991455,
+      "learning_rate": 8.325666592904993e-06,
+      "loss": 0.3595,
+      "step": 3362
+    },
+    {
+      "epoch": 0.26904,
+      "grad_norm": 1.984908103942871,
+      "learning_rate": 8.324727374581563e-06,
+      "loss": 0.4163,
+      "step": 3363
+    },
+    {
+      "epoch": 0.26912,
+      "grad_norm": 1.620694875717163,
+      "learning_rate": 8.323787945913755e-06,
+      "loss": 0.3158,
+      "step": 3364
+    },
+    {
+      "epoch": 0.2692,
+      "grad_norm": 1.4381059408187866,
+      "learning_rate": 8.322848306961007e-06,
+      "loss": 0.2875,
+      "step": 3365
+    },
+    {
+      "epoch": 0.26928,
+      "grad_norm": 2.2146494388580322,
+      "learning_rate": 8.321908457782764e-06,
+      "loss": 0.419,
+      "step": 3366
+    },
+    {
+      "epoch": 0.26936,
+      "grad_norm": 1.2770941257476807,
+      "learning_rate": 8.320968398438487e-06,
+      "loss": 0.3038,
+      "step": 3367
+    },
+    {
+      "epoch": 0.26944,
+      "grad_norm": 1.5118519067764282,
+      "learning_rate": 8.320028128987653e-06,
+      "loss": 0.3433,
+      "step": 3368
+    },
+    {
+      "epoch": 0.26952,
+      "grad_norm": 1.8498775959014893,
+      "learning_rate": 8.319087649489747e-06,
+      "loss": 0.3517,
+      "step": 3369
+    },
+    {
+      "epoch": 0.2696,
+      "grad_norm": 1.4128763675689697,
+      "learning_rate": 8.318146960004273e-06,
+      "loss": 0.2989,
+      "step": 3370
+    },
+    {
+      "epoch": 0.26968,
+      "grad_norm": 1.828460931777954,
+      "learning_rate": 8.31720606059074e-06,
+      "loss": 0.4235,
+      "step": 3371
+    },
+    {
+      "epoch": 0.26976,
+      "grad_norm": 1.5870622396469116,
+      "learning_rate": 8.316264951308682e-06,
+      "loss": 0.307,
+      "step": 3372
+    },
+    {
+      "epoch": 0.26984,
+      "grad_norm": 1.6679601669311523,
+      "learning_rate": 8.315323632217636e-06,
+      "loss": 0.3352,
+      "step": 3373
+    },
+    {
+      "epoch": 0.26992,
+      "grad_norm": 1.5351213216781616,
+      "learning_rate": 8.314382103377158e-06,
+      "loss": 0.3193,
+      "step": 3374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.365196704864502,
+      "learning_rate": 8.313440364846811e-06,
+      "loss": 0.2916,
+      "step": 3375
+    },
+    {
+      "epoch": 0.27008,
+      "grad_norm": 1.6591503620147705,
+      "learning_rate": 8.312498416686183e-06,
+      "loss": 0.3194,
+      "step": 3376
+    },
+    {
+      "epoch": 0.27016,
+      "grad_norm": 1.4480193853378296,
+      "learning_rate": 8.31155625895486e-06,
+      "loss": 0.3788,
+      "step": 3377
+    },
+    {
+      "epoch": 0.27024,
+      "grad_norm": 1.819912314414978,
+      "learning_rate": 8.310613891712455e-06,
+      "loss": 0.3924,
+      "step": 3378
+    },
+    {
+      "epoch": 0.27032,
+      "grad_norm": 1.4778650999069214,
+      "learning_rate": 8.309671315018587e-06,
+      "loss": 0.4356,
+      "step": 3379
+    },
+    {
+      "epoch": 0.2704,
+      "grad_norm": 1.509982705116272,
+      "learning_rate": 8.308728528932889e-06,
+      "loss": 0.3287,
+      "step": 3380
+    },
+    {
+      "epoch": 0.27048,
+      "grad_norm": 1.2625203132629395,
+      "learning_rate": 8.307785533515007e-06,
+      "loss": 0.2483,
+      "step": 3381
+    },
+    {
+      "epoch": 0.27056,
+      "grad_norm": 1.5584619045257568,
+      "learning_rate": 8.306842328824602e-06,
+      "loss": 0.3089,
+      "step": 3382
+    },
+    {
+      "epoch": 0.27064,
+      "grad_norm": 1.7761331796646118,
+      "learning_rate": 8.305898914921348e-06,
+      "loss": 0.3591,
+      "step": 3383
+    },
+    {
+      "epoch": 0.27072,
+      "grad_norm": 1.5493628978729248,
+      "learning_rate": 8.304955291864932e-06,
+      "loss": 0.3878,
+      "step": 3384
+    },
+    {
+      "epoch": 0.2708,
+      "grad_norm": 1.6428436040878296,
+      "learning_rate": 8.304011459715052e-06,
+      "loss": 0.3434,
+      "step": 3385
+    },
+    {
+      "epoch": 0.27088,
+      "grad_norm": 1.5427204370498657,
+      "learning_rate": 8.303067418531424e-06,
+      "loss": 0.3383,
+      "step": 3386
+    },
+    {
+      "epoch": 0.27096,
+      "grad_norm": 1.6140514612197876,
+      "learning_rate": 8.302123168373771e-06,
+      "loss": 0.3603,
+      "step": 3387
+    },
+    {
+      "epoch": 0.27104,
+      "grad_norm": 1.098120927810669,
+      "learning_rate": 8.301178709301833e-06,
+      "loss": 0.2703,
+      "step": 3388
+    },
+    {
+      "epoch": 0.27112,
+      "grad_norm": 1.3397828340530396,
+      "learning_rate": 8.300234041375366e-06,
+      "loss": 0.3006,
+      "step": 3389
+    },
+    {
+      "epoch": 0.2712,
+      "grad_norm": 2.0523359775543213,
+      "learning_rate": 8.299289164654132e-06,
+      "loss": 0.321,
+      "step": 3390
+    },
+    {
+      "epoch": 0.27128,
+      "grad_norm": 1.3592901229858398,
+      "learning_rate": 8.298344079197913e-06,
+      "loss": 0.2744,
+      "step": 3391
+    },
+    {
+      "epoch": 0.27136,
+      "grad_norm": 1.4901467561721802,
+      "learning_rate": 8.297398785066501e-06,
+      "loss": 0.4224,
+      "step": 3392
+    },
+    {
+      "epoch": 0.27144,
+      "grad_norm": 1.6656169891357422,
+      "learning_rate": 8.2964532823197e-06,
+      "loss": 0.3217,
+      "step": 3393
+    },
+    {
+      "epoch": 0.27152,
+      "grad_norm": 1.3976407051086426,
+      "learning_rate": 8.29550757101733e-06,
+      "loss": 0.3364,
+      "step": 3394
+    },
+    {
+      "epoch": 0.2716,
+      "grad_norm": 1.6198853254318237,
+      "learning_rate": 8.294561651219223e-06,
+      "loss": 0.4076,
+      "step": 3395
+    },
+    {
+      "epoch": 0.27168,
+      "grad_norm": 1.8720680475234985,
+      "learning_rate": 8.293615522985224e-06,
+      "loss": 0.3663,
+      "step": 3396
+    },
+    {
+      "epoch": 0.27176,
+      "grad_norm": 1.2774734497070312,
+      "learning_rate": 8.292669186375192e-06,
+      "loss": 0.3451,
+      "step": 3397
+    },
+    {
+      "epoch": 0.27184,
+      "grad_norm": 1.308482050895691,
+      "learning_rate": 8.291722641448995e-06,
+      "loss": 0.2712,
+      "step": 3398
+    },
+    {
+      "epoch": 0.27192,
+      "grad_norm": 1.2496081590652466,
+      "learning_rate": 8.290775888266525e-06,
+      "loss": 0.2932,
+      "step": 3399
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 1.3944826126098633,
+      "learning_rate": 8.289828926887673e-06,
+      "loss": 0.2683,
+      "step": 3400
+    },
+    {
+      "epoch": 0.27208,
+      "grad_norm": 1.0699915885925293,
+      "learning_rate": 8.288881757372352e-06,
+      "loss": 0.2491,
+      "step": 3401
+    },
+    {
+      "epoch": 0.27216,
+      "grad_norm": 1.4618251323699951,
+      "learning_rate": 8.287934379780489e-06,
+      "loss": 0.3656,
+      "step": 3402
+    },
+    {
+      "epoch": 0.27224,
+      "grad_norm": 1.899168610572815,
+      "learning_rate": 8.286986794172017e-06,
+      "loss": 0.3084,
+      "step": 3403
+    },
+    {
+      "epoch": 0.27232,
+      "grad_norm": 1.8475021123886108,
+      "learning_rate": 8.286039000606889e-06,
+      "loss": 0.32,
+      "step": 3404
+    },
+    {
+      "epoch": 0.2724,
+      "grad_norm": 1.7992068529129028,
+      "learning_rate": 8.28509099914507e-06,
+      "loss": 0.3602,
+      "step": 3405
+    },
+    {
+      "epoch": 0.27248,
+      "grad_norm": 1.4878534078598022,
+      "learning_rate": 8.284142789846535e-06,
+      "loss": 0.3161,
+      "step": 3406
+    },
+    {
+      "epoch": 0.27256,
+      "grad_norm": 1.534593105316162,
+      "learning_rate": 8.283194372771274e-06,
+      "loss": 0.3459,
+      "step": 3407
+    },
+    {
+      "epoch": 0.27264,
+      "grad_norm": 1.5597976446151733,
+      "learning_rate": 8.28224574797929e-06,
+      "loss": 0.2994,
+      "step": 3408
+    },
+    {
+      "epoch": 0.27272,
+      "grad_norm": 1.2504432201385498,
+      "learning_rate": 8.281296915530602e-06,
+      "loss": 0.2811,
+      "step": 3409
+    },
+    {
+      "epoch": 0.2728,
+      "grad_norm": 1.392411231994629,
+      "learning_rate": 8.280347875485236e-06,
+      "loss": 0.2981,
+      "step": 3410
+    },
+    {
+      "epoch": 0.27288,
+      "grad_norm": 1.4544750452041626,
+      "learning_rate": 8.279398627903235e-06,
+      "loss": 0.2969,
+      "step": 3411
+    },
+    {
+      "epoch": 0.27296,
+      "grad_norm": 1.6616743803024292,
+      "learning_rate": 8.278449172844656e-06,
+      "loss": 0.3423,
+      "step": 3412
+    },
+    {
+      "epoch": 0.27304,
+      "grad_norm": 1.702860951423645,
+      "learning_rate": 8.27749951036957e-06,
+      "loss": 0.4418,
+      "step": 3413
+    },
+    {
+      "epoch": 0.27312,
+      "grad_norm": 1.229211449623108,
+      "learning_rate": 8.276549640538055e-06,
+      "loss": 0.2837,
+      "step": 3414
+    },
+    {
+      "epoch": 0.2732,
+      "grad_norm": 1.7604984045028687,
+      "learning_rate": 8.275599563410209e-06,
+      "loss": 0.3789,
+      "step": 3415
+    },
+    {
+      "epoch": 0.27328,
+      "grad_norm": 1.5034135580062866,
+      "learning_rate": 8.274649279046137e-06,
+      "loss": 0.2756,
+      "step": 3416
+    },
+    {
+      "epoch": 0.27336,
+      "grad_norm": 1.3147361278533936,
+      "learning_rate": 8.273698787505962e-06,
+      "loss": 0.2322,
+      "step": 3417
+    },
+    {
+      "epoch": 0.27344,
+      "grad_norm": 1.2072352170944214,
+      "learning_rate": 8.272748088849818e-06,
+      "loss": 0.2772,
+      "step": 3418
+    },
+    {
+      "epoch": 0.27352,
+      "grad_norm": 1.5700079202651978,
+      "learning_rate": 8.271797183137855e-06,
+      "loss": 0.3311,
+      "step": 3419
+    },
+    {
+      "epoch": 0.2736,
+      "grad_norm": 1.3132052421569824,
+      "learning_rate": 8.27084607043023e-06,
+      "loss": 0.271,
+      "step": 3420
+    },
+    {
+      "epoch": 0.27368,
+      "grad_norm": 1.235514760017395,
+      "learning_rate": 8.26989475078712e-06,
+      "loss": 0.2534,
+      "step": 3421
+    },
+    {
+      "epoch": 0.27376,
+      "grad_norm": 1.5095601081848145,
+      "learning_rate": 8.26894322426871e-06,
+      "loss": 0.294,
+      "step": 3422
+    },
+    {
+      "epoch": 0.27384,
+      "grad_norm": 1.3389074802398682,
+      "learning_rate": 8.267991490935199e-06,
+      "loss": 0.2859,
+      "step": 3423
+    },
+    {
+      "epoch": 0.27392,
+      "grad_norm": 1.3496683835983276,
+      "learning_rate": 8.267039550846802e-06,
+      "loss": 0.2212,
+      "step": 3424
+    },
+    {
+      "epoch": 0.274,
+      "grad_norm": 1.800253987312317,
+      "learning_rate": 8.266087404063743e-06,
+      "loss": 0.4047,
+      "step": 3425
+    },
+    {
+      "epoch": 0.27408,
+      "grad_norm": 1.3180961608886719,
+      "learning_rate": 8.265135050646262e-06,
+      "loss": 0.3051,
+      "step": 3426
+    },
+    {
+      "epoch": 0.27416,
+      "grad_norm": 1.7630568742752075,
+      "learning_rate": 8.264182490654614e-06,
+      "loss": 0.3703,
+      "step": 3427
+    },
+    {
+      "epoch": 0.27424,
+      "grad_norm": 1.518088459968567,
+      "learning_rate": 8.26322972414906e-06,
+      "loss": 0.3771,
+      "step": 3428
+    },
+    {
+      "epoch": 0.27432,
+      "grad_norm": 1.6900966167449951,
+      "learning_rate": 8.262276751189882e-06,
+      "loss": 0.3216,
+      "step": 3429
+    },
+    {
+      "epoch": 0.2744,
+      "grad_norm": 1.6469087600708008,
+      "learning_rate": 8.261323571837367e-06,
+      "loss": 0.3232,
+      "step": 3430
+    },
+    {
+      "epoch": 0.27448,
+      "grad_norm": 1.3444794416427612,
+      "learning_rate": 8.260370186151822e-06,
+      "loss": 0.2931,
+      "step": 3431
+    },
+    {
+      "epoch": 0.27456,
+      "grad_norm": 1.5142230987548828,
+      "learning_rate": 8.259416594193566e-06,
+      "loss": 0.3601,
+      "step": 3432
+    },
+    {
+      "epoch": 0.27464,
+      "grad_norm": 1.472582221031189,
+      "learning_rate": 8.25846279602293e-06,
+      "loss": 0.3059,
+      "step": 3433
+    },
+    {
+      "epoch": 0.27472,
+      "grad_norm": 1.7719913721084595,
+      "learning_rate": 8.257508791700253e-06,
+      "loss": 0.4288,
+      "step": 3434
+    },
+    {
+      "epoch": 0.2748,
+      "grad_norm": 1.7561125755310059,
+      "learning_rate": 8.256554581285895e-06,
+      "loss": 0.3458,
+      "step": 3435
+    },
+    {
+      "epoch": 0.27488,
+      "grad_norm": 1.25856614112854,
+      "learning_rate": 8.255600164840226e-06,
+      "loss": 0.2884,
+      "step": 3436
+    },
+    {
+      "epoch": 0.27496,
+      "grad_norm": 1.5106571912765503,
+      "learning_rate": 8.254645542423627e-06,
+      "loss": 0.314,
+      "step": 3437
+    },
+    {
+      "epoch": 0.27504,
+      "grad_norm": 1.3592396974563599,
+      "learning_rate": 8.253690714096494e-06,
+      "loss": 0.2891,
+      "step": 3438
+    },
+    {
+      "epoch": 0.27512,
+      "grad_norm": 1.3193227052688599,
+      "learning_rate": 8.252735679919238e-06,
+      "loss": 0.3048,
+      "step": 3439
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 1.72906494140625,
+      "learning_rate": 8.251780439952277e-06,
+      "loss": 0.3843,
+      "step": 3440
+    },
+    {
+      "epoch": 0.27528,
+      "grad_norm": 1.8231126070022583,
+      "learning_rate": 8.250824994256048e-06,
+      "loss": 0.4098,
+      "step": 3441
+    },
+    {
+      "epoch": 0.27536,
+      "grad_norm": 1.2340192794799805,
+      "learning_rate": 8.249869342891001e-06,
+      "loss": 0.245,
+      "step": 3442
+    },
+    {
+      "epoch": 0.27544,
+      "grad_norm": 1.667550802230835,
+      "learning_rate": 8.248913485917593e-06,
+      "loss": 0.4117,
+      "step": 3443
+    },
+    {
+      "epoch": 0.27552,
+      "grad_norm": 1.5927138328552246,
+      "learning_rate": 8.247957423396302e-06,
+      "loss": 0.4634,
+      "step": 3444
+    },
+    {
+      "epoch": 0.2756,
+      "grad_norm": 1.898964285850525,
+      "learning_rate": 8.24700115538761e-06,
+      "loss": 0.4804,
+      "step": 3445
+    },
+    {
+      "epoch": 0.27568,
+      "grad_norm": 1.5059139728546143,
+      "learning_rate": 8.246044681952022e-06,
+      "loss": 0.3067,
+      "step": 3446
+    },
+    {
+      "epoch": 0.27576,
+      "grad_norm": 1.2551331520080566,
+      "learning_rate": 8.245088003150047e-06,
+      "loss": 0.2362,
+      "step": 3447
+    },
+    {
+      "epoch": 0.27584,
+      "grad_norm": 1.5307579040527344,
+      "learning_rate": 8.244131119042211e-06,
+      "loss": 0.3315,
+      "step": 3448
+    },
+    {
+      "epoch": 0.27592,
+      "grad_norm": 1.7814452648162842,
+      "learning_rate": 8.243174029689055e-06,
+      "loss": 0.3452,
+      "step": 3449
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 1.68953275680542,
+      "learning_rate": 8.242216735151131e-06,
+      "loss": 0.3785,
+      "step": 3450
+    },
+    {
+      "epoch": 0.27608,
+      "grad_norm": 1.7366198301315308,
+      "learning_rate": 8.241259235489001e-06,
+      "loss": 0.4651,
+      "step": 3451
+    },
+    {
+      "epoch": 0.27616,
+      "grad_norm": 1.8493014574050903,
+      "learning_rate": 8.240301530763244e-06,
+      "loss": 0.3389,
+      "step": 3452
+    },
+    {
+      "epoch": 0.27624,
+      "grad_norm": 1.3720508813858032,
+      "learning_rate": 8.239343621034452e-06,
+      "loss": 0.306,
+      "step": 3453
+    },
+    {
+      "epoch": 0.27632,
+      "grad_norm": 1.694554328918457,
+      "learning_rate": 8.23838550636323e-06,
+      "loss": 0.4108,
+      "step": 3454
+    },
+    {
+      "epoch": 0.2764,
+      "grad_norm": 1.6249841451644897,
+      "learning_rate": 8.23742718681019e-06,
+      "loss": 0.3531,
+      "step": 3455
+    },
+    {
+      "epoch": 0.27648,
+      "grad_norm": 1.7144601345062256,
+      "learning_rate": 8.236468662435964e-06,
+      "loss": 0.4359,
+      "step": 3456
+    },
+    {
+      "epoch": 0.27656,
+      "grad_norm": 1.3996644020080566,
+      "learning_rate": 8.235509933301197e-06,
+      "loss": 0.3071,
+      "step": 3457
+    },
+    {
+      "epoch": 0.27664,
+      "grad_norm": 1.3151053190231323,
+      "learning_rate": 8.234550999466542e-06,
+      "loss": 0.2705,
+      "step": 3458
+    },
+    {
+      "epoch": 0.27672,
+      "grad_norm": 1.8970770835876465,
+      "learning_rate": 8.233591860992667e-06,
+      "loss": 0.4059,
+      "step": 3459
+    },
+    {
+      "epoch": 0.2768,
+      "grad_norm": 1.183557391166687,
+      "learning_rate": 8.232632517940255e-06,
+      "loss": 0.226,
+      "step": 3460
+    },
+    {
+      "epoch": 0.27688,
+      "grad_norm": 2.460853099822998,
+      "learning_rate": 8.23167297037e-06,
+      "loss": 0.5838,
+      "step": 3461
+    },
+    {
+      "epoch": 0.27696,
+      "grad_norm": 1.3229994773864746,
+      "learning_rate": 8.230713218342611e-06,
+      "loss": 0.2763,
+      "step": 3462
+    },
+    {
+      "epoch": 0.27704,
+      "grad_norm": 1.5857011079788208,
+      "learning_rate": 8.229753261918805e-06,
+      "loss": 0.3172,
+      "step": 3463
+    },
+    {
+      "epoch": 0.27712,
+      "grad_norm": 1.5672333240509033,
+      "learning_rate": 8.228793101159318e-06,
+      "loss": 0.3418,
+      "step": 3464
+    },
+    {
+      "epoch": 0.2772,
+      "grad_norm": 1.766082763671875,
+      "learning_rate": 8.227832736124895e-06,
+      "loss": 0.4143,
+      "step": 3465
+    },
+    {
+      "epoch": 0.27728,
+      "grad_norm": 1.8570321798324585,
+      "learning_rate": 8.226872166876293e-06,
+      "loss": 0.4841,
+      "step": 3466
+    },
+    {
+      "epoch": 0.27736,
+      "grad_norm": 1.8656039237976074,
+      "learning_rate": 8.225911393474289e-06,
+      "loss": 0.4092,
+      "step": 3467
+    },
+    {
+      "epoch": 0.27744,
+      "grad_norm": 1.710142731666565,
+      "learning_rate": 8.224950415979664e-06,
+      "loss": 0.3453,
+      "step": 3468
+    },
+    {
+      "epoch": 0.27752,
+      "grad_norm": 2.0013701915740967,
+      "learning_rate": 8.223989234453215e-06,
+      "loss": 0.4788,
+      "step": 3469
+    },
+    {
+      "epoch": 0.2776,
+      "grad_norm": 1.8113597631454468,
+      "learning_rate": 8.223027848955757e-06,
+      "loss": 0.4269,
+      "step": 3470
+    },
+    {
+      "epoch": 0.27768,
+      "grad_norm": 1.8451036214828491,
+      "learning_rate": 8.222066259548111e-06,
+      "loss": 0.3931,
+      "step": 3471
+    },
+    {
+      "epoch": 0.27776,
+      "grad_norm": 2.501964569091797,
+      "learning_rate": 8.221104466291112e-06,
+      "loss": 0.5903,
+      "step": 3472
+    },
+    {
+      "epoch": 0.27784,
+      "grad_norm": 1.5321495532989502,
+      "learning_rate": 8.220142469245613e-06,
+      "loss": 0.3115,
+      "step": 3473
+    },
+    {
+      "epoch": 0.27792,
+      "grad_norm": 1.721560001373291,
+      "learning_rate": 8.219180268472476e-06,
+      "loss": 0.304,
+      "step": 3474
+    },
+    {
+      "epoch": 0.278,
+      "grad_norm": 1.3853156566619873,
+      "learning_rate": 8.218217864032572e-06,
+      "loss": 0.2682,
+      "step": 3475
+    },
+    {
+      "epoch": 0.27808,
+      "grad_norm": 1.5545319318771362,
+      "learning_rate": 8.217255255986794e-06,
+      "loss": 0.3532,
+      "step": 3476
+    },
+    {
+      "epoch": 0.27816,
+      "grad_norm": 1.3426564931869507,
+      "learning_rate": 8.216292444396038e-06,
+      "loss": 0.301,
+      "step": 3477
+    },
+    {
+      "epoch": 0.27824,
+      "grad_norm": 1.5732641220092773,
+      "learning_rate": 8.215329429321224e-06,
+      "loss": 0.4052,
+      "step": 3478
+    },
+    {
+      "epoch": 0.27832,
+      "grad_norm": 1.5441874265670776,
+      "learning_rate": 8.214366210823274e-06,
+      "loss": 0.2982,
+      "step": 3479
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 1.5156362056732178,
+      "learning_rate": 8.21340278896313e-06,
+      "loss": 0.4654,
+      "step": 3480
+    },
+    {
+      "epoch": 0.27848,
+      "grad_norm": 1.3966670036315918,
+      "learning_rate": 8.212439163801743e-06,
+      "loss": 0.2895,
+      "step": 3481
+    },
+    {
+      "epoch": 0.27856,
+      "grad_norm": 1.8723714351654053,
+      "learning_rate": 8.211475335400079e-06,
+      "loss": 0.3717,
+      "step": 3482
+    },
+    {
+      "epoch": 0.27864,
+      "grad_norm": 1.6649876832962036,
+      "learning_rate": 8.210511303819116e-06,
+      "loss": 0.3856,
+      "step": 3483
+    },
+    {
+      "epoch": 0.27872,
+      "grad_norm": 1.9745384454727173,
+      "learning_rate": 8.209547069119845e-06,
+      "loss": 0.4151,
+      "step": 3484
+    },
+    {
+      "epoch": 0.2788,
+      "grad_norm": 1.885003685951233,
+      "learning_rate": 8.20858263136327e-06,
+      "loss": 0.3486,
+      "step": 3485
+    },
+    {
+      "epoch": 0.27888,
+      "grad_norm": 1.5459332466125488,
+      "learning_rate": 8.20761799061041e-06,
+      "loss": 0.2786,
+      "step": 3486
+    },
+    {
+      "epoch": 0.27896,
+      "grad_norm": 0.8227094411849976,
+      "learning_rate": 8.20665314692229e-06,
+      "loss": 0.1874,
+      "step": 3487
+    },
+    {
+      "epoch": 0.27904,
+      "grad_norm": 1.5767573118209839,
+      "learning_rate": 8.205688100359956e-06,
+      "loss": 0.3388,
+      "step": 3488
+    },
+    {
+      "epoch": 0.27912,
+      "grad_norm": 1.772367238998413,
+      "learning_rate": 8.20472285098446e-06,
+      "loss": 0.3531,
+      "step": 3489
+    },
+    {
+      "epoch": 0.2792,
+      "grad_norm": 1.7118393182754517,
+      "learning_rate": 8.203757398856875e-06,
+      "loss": 0.312,
+      "step": 3490
+    },
+    {
+      "epoch": 0.27928,
+      "grad_norm": 1.6346728801727295,
+      "learning_rate": 8.202791744038278e-06,
+      "loss": 0.3631,
+      "step": 3491
+    },
+    {
+      "epoch": 0.27936,
+      "grad_norm": 1.87623131275177,
+      "learning_rate": 8.201825886589765e-06,
+      "loss": 0.5274,
+      "step": 3492
+    },
+    {
+      "epoch": 0.27944,
+      "grad_norm": 1.3338699340820312,
+      "learning_rate": 8.20085982657244e-06,
+      "loss": 0.2704,
+      "step": 3493
+    },
+    {
+      "epoch": 0.27952,
+      "grad_norm": 1.6640794277191162,
+      "learning_rate": 8.199893564047425e-06,
+      "loss": 0.3691,
+      "step": 3494
+    },
+    {
+      "epoch": 0.2796,
+      "grad_norm": 1.3115794658660889,
+      "learning_rate": 8.19892709907585e-06,
+      "loss": 0.3016,
+      "step": 3495
+    },
+    {
+      "epoch": 0.27968,
+      "grad_norm": 1.4660612344741821,
+      "learning_rate": 8.197960431718862e-06,
+      "loss": 0.3437,
+      "step": 3496
+    },
+    {
+      "epoch": 0.27976,
+      "grad_norm": 1.5010005235671997,
+      "learning_rate": 8.196993562037618e-06,
+      "loss": 0.4072,
+      "step": 3497
+    },
+    {
+      "epoch": 0.27984,
+      "grad_norm": 1.6829041242599487,
+      "learning_rate": 8.196026490093289e-06,
+      "loss": 0.4016,
+      "step": 3498
+    },
+    {
+      "epoch": 0.27992,
+      "grad_norm": 1.4259425401687622,
+      "learning_rate": 8.195059215947057e-06,
+      "loss": 0.3122,
+      "step": 3499
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4573216438293457,
+      "learning_rate": 8.194091739660119e-06,
+      "loss": 0.3861,
+      "step": 3500
+    },
+    {
+      "epoch": 0.28008,
+      "grad_norm": 1.3331818580627441,
+      "learning_rate": 8.193124061293684e-06,
+      "loss": 0.2762,
+      "step": 3501
+    },
+    {
+      "epoch": 0.28016,
+      "grad_norm": 1.7647954225540161,
+      "learning_rate": 8.192156180908974e-06,
+      "loss": 0.3346,
+      "step": 3502
+    },
+    {
+      "epoch": 0.28024,
+      "grad_norm": 1.65705144405365,
+      "learning_rate": 8.191188098567224e-06,
+      "loss": 0.3713,
+      "step": 3503
+    },
+    {
+      "epoch": 0.28032,
+      "grad_norm": 1.42042076587677,
+      "learning_rate": 8.190219814329681e-06,
+      "loss": 0.3316,
+      "step": 3504
+    },
+    {
+      "epoch": 0.2804,
+      "grad_norm": 1.3383320569992065,
+      "learning_rate": 8.189251328257604e-06,
+      "loss": 0.3086,
+      "step": 3505
+    },
+    {
+      "epoch": 0.28048,
+      "grad_norm": 1.7345651388168335,
+      "learning_rate": 8.188282640412267e-06,
+      "loss": 0.3242,
+      "step": 3506
+    },
+    {
+      "epoch": 0.28056,
+      "grad_norm": 1.544132113456726,
+      "learning_rate": 8.187313750854956e-06,
+      "loss": 0.3949,
+      "step": 3507
+    },
+    {
+      "epoch": 0.28064,
+      "grad_norm": 1.5500211715698242,
+      "learning_rate": 8.186344659646966e-06,
+      "loss": 0.4493,
+      "step": 3508
+    },
+    {
+      "epoch": 0.28072,
+      "grad_norm": 1.221084475517273,
+      "learning_rate": 8.185375366849613e-06,
+      "loss": 0.2511,
+      "step": 3509
+    },
+    {
+      "epoch": 0.2808,
+      "grad_norm": 1.6258049011230469,
+      "learning_rate": 8.184405872524219e-06,
+      "loss": 0.3507,
+      "step": 3510
+    },
+    {
+      "epoch": 0.28088,
+      "grad_norm": 1.774895191192627,
+      "learning_rate": 8.18343617673212e-06,
+      "loss": 0.3715,
+      "step": 3511
+    },
+    {
+      "epoch": 0.28096,
+      "grad_norm": 1.593440055847168,
+      "learning_rate": 8.182466279534666e-06,
+      "loss": 0.3073,
+      "step": 3512
+    },
+    {
+      "epoch": 0.28104,
+      "grad_norm": 1.700151801109314,
+      "learning_rate": 8.18149618099322e-06,
+      "loss": 0.3785,
+      "step": 3513
+    },
+    {
+      "epoch": 0.28112,
+      "grad_norm": 1.4120454788208008,
+      "learning_rate": 8.180525881169155e-06,
+      "loss": 0.326,
+      "step": 3514
+    },
+    {
+      "epoch": 0.2812,
+      "grad_norm": 1.9900215864181519,
+      "learning_rate": 8.17955538012386e-06,
+      "loss": 0.4358,
+      "step": 3515
+    },
+    {
+      "epoch": 0.28128,
+      "grad_norm": 1.1549736261367798,
+      "learning_rate": 8.178584677918734e-06,
+      "loss": 0.2993,
+      "step": 3516
+    },
+    {
+      "epoch": 0.28136,
+      "grad_norm": 1.4273407459259033,
+      "learning_rate": 8.177613774615193e-06,
+      "loss": 0.322,
+      "step": 3517
+    },
+    {
+      "epoch": 0.28144,
+      "grad_norm": 1.7175918817520142,
+      "learning_rate": 8.17664267027466e-06,
+      "loss": 0.3311,
+      "step": 3518
+    },
+    {
+      "epoch": 0.28152,
+      "grad_norm": 1.2839045524597168,
+      "learning_rate": 8.175671364958573e-06,
+      "loss": 0.304,
+      "step": 3519
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 1.3330825567245483,
+      "learning_rate": 8.174699858728386e-06,
+      "loss": 0.3253,
+      "step": 3520
+    },
+    {
+      "epoch": 0.28168,
+      "grad_norm": 1.5143156051635742,
+      "learning_rate": 8.173728151645561e-06,
+      "loss": 0.2915,
+      "step": 3521
+    },
+    {
+      "epoch": 0.28176,
+      "grad_norm": 1.783462405204773,
+      "learning_rate": 8.172756243771575e-06,
+      "loss": 0.4024,
+      "step": 3522
+    },
+    {
+      "epoch": 0.28184,
+      "grad_norm": 1.8704419136047363,
+      "learning_rate": 8.171784135167917e-06,
+      "loss": 0.3364,
+      "step": 3523
+    },
+    {
+      "epoch": 0.28192,
+      "grad_norm": 1.4959416389465332,
+      "learning_rate": 8.17081182589609e-06,
+      "loss": 0.3278,
+      "step": 3524
+    },
+    {
+      "epoch": 0.282,
+      "grad_norm": 2.0997822284698486,
+      "learning_rate": 8.169839316017609e-06,
+      "loss": 0.4619,
+      "step": 3525
+    },
+    {
+      "epoch": 0.28208,
+      "grad_norm": 1.242011547088623,
+      "learning_rate": 8.168866605594001e-06,
+      "loss": 0.2991,
+      "step": 3526
+    },
+    {
+      "epoch": 0.28216,
+      "grad_norm": 1.39487624168396,
+      "learning_rate": 8.167893694686805e-06,
+      "loss": 0.3224,
+      "step": 3527
+    },
+    {
+      "epoch": 0.28224,
+      "grad_norm": 1.8157869577407837,
+      "learning_rate": 8.166920583357575e-06,
+      "loss": 0.3669,
+      "step": 3528
+    },
+    {
+      "epoch": 0.28232,
+      "grad_norm": 1.2828103303909302,
+      "learning_rate": 8.165947271667875e-06,
+      "loss": 0.2376,
+      "step": 3529
+    },
+    {
+      "epoch": 0.2824,
+      "grad_norm": 1.9789408445358276,
+      "learning_rate": 8.164973759679287e-06,
+      "loss": 0.4204,
+      "step": 3530
+    },
+    {
+      "epoch": 0.28248,
+      "grad_norm": 1.2703884840011597,
+      "learning_rate": 8.164000047453398e-06,
+      "loss": 0.2719,
+      "step": 3531
+    },
+    {
+      "epoch": 0.28256,
+      "grad_norm": 1.5081241130828857,
+      "learning_rate": 8.163026135051813e-06,
+      "loss": 0.3015,
+      "step": 3532
+    },
+    {
+      "epoch": 0.28264,
+      "grad_norm": 1.3452731370925903,
+      "learning_rate": 8.162052022536148e-06,
+      "loss": 0.2942,
+      "step": 3533
+    },
+    {
+      "epoch": 0.28272,
+      "grad_norm": 1.6602774858474731,
+      "learning_rate": 8.16107770996803e-06,
+      "loss": 0.4434,
+      "step": 3534
+    },
+    {
+      "epoch": 0.2828,
+      "grad_norm": 2.201460361480713,
+      "learning_rate": 8.160103197409104e-06,
+      "loss": 0.3866,
+      "step": 3535
+    },
+    {
+      "epoch": 0.28288,
+      "grad_norm": 1.59469735622406,
+      "learning_rate": 8.159128484921022e-06,
+      "loss": 0.4352,
+      "step": 3536
+    },
+    {
+      "epoch": 0.28296,
+      "grad_norm": 1.86858332157135,
+      "learning_rate": 8.158153572565452e-06,
+      "loss": 0.4111,
+      "step": 3537
+    },
+    {
+      "epoch": 0.28304,
+      "grad_norm": 1.9216097593307495,
+      "learning_rate": 8.157178460404071e-06,
+      "loss": 0.4455,
+      "step": 3538
+    },
+    {
+      "epoch": 0.28312,
+      "grad_norm": 1.8076080083847046,
+      "learning_rate": 8.156203148498575e-06,
+      "loss": 0.4287,
+      "step": 3539
+    },
+    {
+      "epoch": 0.2832,
+      "grad_norm": 1.820548415184021,
+      "learning_rate": 8.155227636910665e-06,
+      "loss": 0.3744,
+      "step": 3540
+    },
+    {
+      "epoch": 0.28328,
+      "grad_norm": 1.749031662940979,
+      "learning_rate": 8.15425192570206e-06,
+      "loss": 0.4377,
+      "step": 3541
+    },
+    {
+      "epoch": 0.28336,
+      "grad_norm": 1.2342244386672974,
+      "learning_rate": 8.15327601493449e-06,
+      "loss": 0.2596,
+      "step": 3542
+    },
+    {
+      "epoch": 0.28344,
+      "grad_norm": 1.6412309408187866,
+      "learning_rate": 8.152299904669698e-06,
+      "loss": 0.344,
+      "step": 3543
+    },
+    {
+      "epoch": 0.28352,
+      "grad_norm": 1.5343546867370605,
+      "learning_rate": 8.151323594969438e-06,
+      "loss": 0.3667,
+      "step": 3544
+    },
+    {
+      "epoch": 0.2836,
+      "grad_norm": 1.3292542695999146,
+      "learning_rate": 8.150347085895479e-06,
+      "loss": 0.3124,
+      "step": 3545
+    },
+    {
+      "epoch": 0.28368,
+      "grad_norm": 1.3630868196487427,
+      "learning_rate": 8.1493703775096e-06,
+      "loss": 0.3347,
+      "step": 3546
+    },
+    {
+      "epoch": 0.28376,
+      "grad_norm": 1.5773979425430298,
+      "learning_rate": 8.148393469873596e-06,
+      "loss": 0.3451,
+      "step": 3547
+    },
+    {
+      "epoch": 0.28384,
+      "grad_norm": 1.5990095138549805,
+      "learning_rate": 8.147416363049271e-06,
+      "loss": 0.3509,
+      "step": 3548
+    },
+    {
+      "epoch": 0.28392,
+      "grad_norm": 1.6553959846496582,
+      "learning_rate": 8.146439057098446e-06,
+      "loss": 0.3553,
+      "step": 3549
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 1.6501647233963013,
+      "learning_rate": 8.145461552082948e-06,
+      "loss": 0.3922,
+      "step": 3550
+    },
+    {
+      "epoch": 0.28408,
+      "grad_norm": 1.6862366199493408,
+      "learning_rate": 8.144483848064621e-06,
+      "loss": 0.5672,
+      "step": 3551
+    },
+    {
+      "epoch": 0.28416,
+      "grad_norm": 1.5185807943344116,
+      "learning_rate": 8.143505945105325e-06,
+      "loss": 0.2501,
+      "step": 3552
+    },
+    {
+      "epoch": 0.28424,
+      "grad_norm": 1.547339916229248,
+      "learning_rate": 8.142527843266924e-06,
+      "loss": 0.3314,
+      "step": 3553
+    },
+    {
+      "epoch": 0.28432,
+      "grad_norm": 1.4977695941925049,
+      "learning_rate": 8.141549542611302e-06,
+      "loss": 0.351,
+      "step": 3554
+    },
+    {
+      "epoch": 0.2844,
+      "grad_norm": 1.4256455898284912,
+      "learning_rate": 8.140571043200354e-06,
+      "loss": 0.313,
+      "step": 3555
+    },
+    {
+      "epoch": 0.28448,
+      "grad_norm": 1.908646821975708,
+      "learning_rate": 8.139592345095982e-06,
+      "loss": 0.4029,
+      "step": 3556
+    },
+    {
+      "epoch": 0.28456,
+      "grad_norm": 1.7634451389312744,
+      "learning_rate": 8.138613448360108e-06,
+      "loss": 0.4724,
+      "step": 3557
+    },
+    {
+      "epoch": 0.28464,
+      "grad_norm": 1.4443376064300537,
+      "learning_rate": 8.137634353054664e-06,
+      "loss": 0.2718,
+      "step": 3558
+    },
+    {
+      "epoch": 0.28472,
+      "grad_norm": 1.3614927530288696,
+      "learning_rate": 8.13665505924159e-06,
+      "loss": 0.2782,
+      "step": 3559
+    },
+    {
+      "epoch": 0.2848,
+      "grad_norm": 1.582389235496521,
+      "learning_rate": 8.13567556698285e-06,
+      "loss": 0.4158,
+      "step": 3560
+    },
+    {
+      "epoch": 0.28488,
+      "grad_norm": 1.4175881147384644,
+      "learning_rate": 8.134695876340406e-06,
+      "loss": 0.3379,
+      "step": 3561
+    },
+    {
+      "epoch": 0.28496,
+      "grad_norm": 1.4460728168487549,
+      "learning_rate": 8.133715987376245e-06,
+      "loss": 0.2862,
+      "step": 3562
+    },
+    {
+      "epoch": 0.28504,
+      "grad_norm": 1.5810394287109375,
+      "learning_rate": 8.132735900152357e-06,
+      "loss": 0.288,
+      "step": 3563
+    },
+    {
+      "epoch": 0.28512,
+      "grad_norm": 1.2738404273986816,
+      "learning_rate": 8.131755614730752e-06,
+      "loss": 0.3385,
+      "step": 3564
+    },
+    {
+      "epoch": 0.2852,
+      "grad_norm": 1.656475305557251,
+      "learning_rate": 8.130775131173447e-06,
+      "loss": 0.3369,
+      "step": 3565
+    },
+    {
+      "epoch": 0.28528,
+      "grad_norm": 1.2260715961456299,
+      "learning_rate": 8.129794449542474e-06,
+      "loss": 0.2934,
+      "step": 3566
+    },
+    {
+      "epoch": 0.28536,
+      "grad_norm": 1.6491786241531372,
+      "learning_rate": 8.12881356989988e-06,
+      "loss": 0.4359,
+      "step": 3567
+    },
+    {
+      "epoch": 0.28544,
+      "grad_norm": 1.270279049873352,
+      "learning_rate": 8.127832492307722e-06,
+      "loss": 0.277,
+      "step": 3568
+    },
+    {
+      "epoch": 0.28552,
+      "grad_norm": 1.3211328983306885,
+      "learning_rate": 8.126851216828065e-06,
+      "loss": 0.2417,
+      "step": 3569
+    },
+    {
+      "epoch": 0.2856,
+      "grad_norm": 1.2138761281967163,
+      "learning_rate": 8.125869743522997e-06,
+      "loss": 0.2868,
+      "step": 3570
+    },
+    {
+      "epoch": 0.28568,
+      "grad_norm": 1.6802400350570679,
+      "learning_rate": 8.124888072454607e-06,
+      "loss": 0.3408,
+      "step": 3571
+    },
+    {
+      "epoch": 0.28576,
+      "grad_norm": 1.6839826107025146,
+      "learning_rate": 8.123906203685007e-06,
+      "loss": 0.3029,
+      "step": 3572
+    },
+    {
+      "epoch": 0.28584,
+      "grad_norm": 1.53984534740448,
+      "learning_rate": 8.122924137276311e-06,
+      "loss": 0.3264,
+      "step": 3573
+    },
+    {
+      "epoch": 0.28592,
+      "grad_norm": 1.4914592504501343,
+      "learning_rate": 8.121941873290655e-06,
+      "loss": 0.3296,
+      "step": 3574
+    },
+    {
+      "epoch": 0.286,
+      "grad_norm": 1.3114702701568604,
+      "learning_rate": 8.120959411790184e-06,
+      "loss": 0.3089,
+      "step": 3575
+    },
+    {
+      "epoch": 0.28608,
+      "grad_norm": 1.9156404733657837,
+      "learning_rate": 8.119976752837054e-06,
+      "loss": 0.3684,
+      "step": 3576
+    },
+    {
+      "epoch": 0.28616,
+      "grad_norm": 2.184542179107666,
+      "learning_rate": 8.118993896493433e-06,
+      "loss": 0.4615,
+      "step": 3577
+    },
+    {
+      "epoch": 0.28624,
+      "grad_norm": 1.3617297410964966,
+      "learning_rate": 8.118010842821504e-06,
+      "loss": 0.3079,
+      "step": 3578
+    },
+    {
+      "epoch": 0.28632,
+      "grad_norm": 1.7069450616836548,
+      "learning_rate": 8.117027591883463e-06,
+      "loss": 0.3893,
+      "step": 3579
+    },
+    {
+      "epoch": 0.2864,
+      "grad_norm": 1.5273637771606445,
+      "learning_rate": 8.116044143741517e-06,
+      "loss": 0.3337,
+      "step": 3580
+    },
+    {
+      "epoch": 0.28648,
+      "grad_norm": 1.6013727188110352,
+      "learning_rate": 8.115060498457882e-06,
+      "loss": 0.3929,
+      "step": 3581
+    },
+    {
+      "epoch": 0.28656,
+      "grad_norm": 1.5870674848556519,
+      "learning_rate": 8.114076656094794e-06,
+      "loss": 0.3977,
+      "step": 3582
+    },
+    {
+      "epoch": 0.28664,
+      "grad_norm": 1.2662407159805298,
+      "learning_rate": 8.113092616714494e-06,
+      "loss": 0.3117,
+      "step": 3583
+    },
+    {
+      "epoch": 0.28672,
+      "grad_norm": 1.407432198524475,
+      "learning_rate": 8.112108380379242e-06,
+      "loss": 0.2418,
+      "step": 3584
+    },
+    {
+      "epoch": 0.2868,
+      "grad_norm": 1.5326191186904907,
+      "learning_rate": 8.111123947151305e-06,
+      "loss": 0.3517,
+      "step": 3585
+    },
+    {
+      "epoch": 0.28688,
+      "grad_norm": 2.2068679332733154,
+      "learning_rate": 8.110139317092966e-06,
+      "loss": 0.5408,
+      "step": 3586
+    },
+    {
+      "epoch": 0.28696,
+      "grad_norm": 1.5989627838134766,
+      "learning_rate": 8.10915449026652e-06,
+      "loss": 0.4202,
+      "step": 3587
+    },
+    {
+      "epoch": 0.28704,
+      "grad_norm": 1.3212885856628418,
+      "learning_rate": 8.108169466734271e-06,
+      "loss": 0.2814,
+      "step": 3588
+    },
+    {
+      "epoch": 0.28712,
+      "grad_norm": 1.6807013750076294,
+      "learning_rate": 8.107184246558542e-06,
+      "loss": 0.3642,
+      "step": 3589
+    },
+    {
+      "epoch": 0.2872,
+      "grad_norm": 1.4167367219924927,
+      "learning_rate": 8.10619882980166e-06,
+      "loss": 0.3045,
+      "step": 3590
+    },
+    {
+      "epoch": 0.28728,
+      "grad_norm": 1.8939565420150757,
+      "learning_rate": 8.105213216525974e-06,
+      "loss": 0.3761,
+      "step": 3591
+    },
+    {
+      "epoch": 0.28736,
+      "grad_norm": 1.1385818719863892,
+      "learning_rate": 8.104227406793834e-06,
+      "loss": 0.2466,
+      "step": 3592
+    },
+    {
+      "epoch": 0.28744,
+      "grad_norm": 1.3388983011245728,
+      "learning_rate": 8.103241400667617e-06,
+      "loss": 0.3011,
+      "step": 3593
+    },
+    {
+      "epoch": 0.28752,
+      "grad_norm": 1.2075142860412598,
+      "learning_rate": 8.102255198209696e-06,
+      "loss": 0.263,
+      "step": 3594
+    },
+    {
+      "epoch": 0.2876,
+      "grad_norm": 1.6224541664123535,
+      "learning_rate": 8.101268799482472e-06,
+      "loss": 0.3047,
+      "step": 3595
+    },
+    {
+      "epoch": 0.28768,
+      "grad_norm": 1.7167912721633911,
+      "learning_rate": 8.100282204548347e-06,
+      "loss": 0.4803,
+      "step": 3596
+    },
+    {
+      "epoch": 0.28776,
+      "grad_norm": 1.5466722249984741,
+      "learning_rate": 8.09929541346974e-06,
+      "loss": 0.4716,
+      "step": 3597
+    },
+    {
+      "epoch": 0.28784,
+      "grad_norm": 1.9157578945159912,
+      "learning_rate": 8.098308426309082e-06,
+      "loss": 0.3921,
+      "step": 3598
+    },
+    {
+      "epoch": 0.28792,
+      "grad_norm": 1.7497869729995728,
+      "learning_rate": 8.097321243128817e-06,
+      "loss": 0.3694,
+      "step": 3599
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 1.7529637813568115,
+      "learning_rate": 8.096333863991402e-06,
+      "loss": 0.3621,
+      "step": 3600
+    },
+    {
+      "epoch": 0.28808,
+      "grad_norm": 1.7688055038452148,
+      "learning_rate": 8.095346288959303e-06,
+      "loss": 0.3573,
+      "step": 3601
+    },
+    {
+      "epoch": 0.28816,
+      "grad_norm": 2.178952217102051,
+      "learning_rate": 8.094358518095002e-06,
+      "loss": 0.5357,
+      "step": 3602
+    },
+    {
+      "epoch": 0.28824,
+      "grad_norm": 2.0295703411102295,
+      "learning_rate": 8.09337055146099e-06,
+      "loss": 0.4133,
+      "step": 3603
+    },
+    {
+      "epoch": 0.28832,
+      "grad_norm": 1.6052531003952026,
+      "learning_rate": 8.092382389119775e-06,
+      "loss": 0.2952,
+      "step": 3604
+    },
+    {
+      "epoch": 0.2884,
+      "grad_norm": 1.9326715469360352,
+      "learning_rate": 8.091394031133872e-06,
+      "loss": 0.3914,
+      "step": 3605
+    },
+    {
+      "epoch": 0.28848,
+      "grad_norm": 1.7421461343765259,
+      "learning_rate": 8.090405477565814e-06,
+      "loss": 0.3181,
+      "step": 3606
+    },
+    {
+      "epoch": 0.28856,
+      "grad_norm": 1.501417875289917,
+      "learning_rate": 8.089416728478141e-06,
+      "loss": 0.314,
+      "step": 3607
+    },
+    {
+      "epoch": 0.28864,
+      "grad_norm": 1.6021887063980103,
+      "learning_rate": 8.088427783933408e-06,
+      "loss": 0.3401,
+      "step": 3608
+    },
+    {
+      "epoch": 0.28872,
+      "grad_norm": 1.6619521379470825,
+      "learning_rate": 8.087438643994185e-06,
+      "loss": 0.3583,
+      "step": 3609
+    },
+    {
+      "epoch": 0.2888,
+      "grad_norm": 1.458363652229309,
+      "learning_rate": 8.086449308723048e-06,
+      "loss": 0.3985,
+      "step": 3610
+    },
+    {
+      "epoch": 0.28888,
+      "grad_norm": 1.698525309562683,
+      "learning_rate": 8.085459778182591e-06,
+      "loss": 0.3954,
+      "step": 3611
+    },
+    {
+      "epoch": 0.28896,
+      "grad_norm": 1.6027504205703735,
+      "learning_rate": 8.084470052435419e-06,
+      "loss": 0.3274,
+      "step": 3612
+    },
+    {
+      "epoch": 0.28904,
+      "grad_norm": 1.7605338096618652,
+      "learning_rate": 8.083480131544146e-06,
+      "loss": 0.3009,
+      "step": 3613
+    },
+    {
+      "epoch": 0.28912,
+      "grad_norm": 1.6151217222213745,
+      "learning_rate": 8.082490015571403e-06,
+      "loss": 0.3801,
+      "step": 3614
+    },
+    {
+      "epoch": 0.2892,
+      "grad_norm": 1.507580041885376,
+      "learning_rate": 8.08149970457983e-06,
+      "loss": 0.3096,
+      "step": 3615
+    },
+    {
+      "epoch": 0.28928,
+      "grad_norm": 1.56270432472229,
+      "learning_rate": 8.080509198632082e-06,
+      "loss": 0.2935,
+      "step": 3616
+    },
+    {
+      "epoch": 0.28936,
+      "grad_norm": 1.5426692962646484,
+      "learning_rate": 8.079518497790825e-06,
+      "loss": 0.3052,
+      "step": 3617
+    },
+    {
+      "epoch": 0.28944,
+      "grad_norm": 1.736161231994629,
+      "learning_rate": 8.078527602118735e-06,
+      "loss": 0.4002,
+      "step": 3618
+    },
+    {
+      "epoch": 0.28952,
+      "grad_norm": 1.4631191492080688,
+      "learning_rate": 8.077536511678506e-06,
+      "loss": 0.2789,
+      "step": 3619
+    },
+    {
+      "epoch": 0.2896,
+      "grad_norm": 1.389417052268982,
+      "learning_rate": 8.076545226532839e-06,
+      "loss": 0.2832,
+      "step": 3620
+    },
+    {
+      "epoch": 0.28968,
+      "grad_norm": 1.6184295415878296,
+      "learning_rate": 8.07555374674445e-06,
+      "loss": 0.3232,
+      "step": 3621
+    },
+    {
+      "epoch": 0.28976,
+      "grad_norm": 1.4686223268508911,
+      "learning_rate": 8.074562072376067e-06,
+      "loss": 0.3091,
+      "step": 3622
+    },
+    {
+      "epoch": 0.28984,
+      "grad_norm": 1.418705940246582,
+      "learning_rate": 8.073570203490428e-06,
+      "loss": 0.3433,
+      "step": 3623
+    },
+    {
+      "epoch": 0.28992,
+      "grad_norm": 2.270691394805908,
+      "learning_rate": 8.072578140150286e-06,
+      "loss": 0.5489,
+      "step": 3624
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0474798679351807,
+      "learning_rate": 8.071585882418406e-06,
+      "loss": 0.3902,
+      "step": 3625
+    },
+    {
+      "epoch": 0.29008,
+      "grad_norm": 1.4669115543365479,
+      "learning_rate": 8.070593430357565e-06,
+      "loss": 0.3705,
+      "step": 3626
+    },
+    {
+      "epoch": 0.29016,
+      "grad_norm": 1.3906588554382324,
+      "learning_rate": 8.069600784030553e-06,
+      "loss": 0.2618,
+      "step": 3627
+    },
+    {
+      "epoch": 0.29024,
+      "grad_norm": 2.14176869392395,
+      "learning_rate": 8.068607943500168e-06,
+      "loss": 0.6205,
+      "step": 3628
+    },
+    {
+      "epoch": 0.29032,
+      "grad_norm": 1.6090953350067139,
+      "learning_rate": 8.067614908829229e-06,
+      "loss": 0.3588,
+      "step": 3629
+    },
+    {
+      "epoch": 0.2904,
+      "grad_norm": 1.9249868392944336,
+      "learning_rate": 8.066621680080557e-06,
+      "loss": 0.3878,
+      "step": 3630
+    },
+    {
+      "epoch": 0.29048,
+      "grad_norm": 1.986382007598877,
+      "learning_rate": 8.065628257316993e-06,
+      "loss": 0.4614,
+      "step": 3631
+    },
+    {
+      "epoch": 0.29056,
+      "grad_norm": 1.3387824296951294,
+      "learning_rate": 8.064634640601386e-06,
+      "loss": 0.3331,
+      "step": 3632
+    },
+    {
+      "epoch": 0.29064,
+      "grad_norm": 1.2659355401992798,
+      "learning_rate": 8.0636408299966e-06,
+      "loss": 0.311,
+      "step": 3633
+    },
+    {
+      "epoch": 0.29072,
+      "grad_norm": 1.6108455657958984,
+      "learning_rate": 8.06264682556551e-06,
+      "loss": 0.4001,
+      "step": 3634
+    },
+    {
+      "epoch": 0.2908,
+      "grad_norm": 1.5894638299942017,
+      "learning_rate": 8.061652627371003e-06,
+      "loss": 0.291,
+      "step": 3635
+    },
+    {
+      "epoch": 0.29088,
+      "grad_norm": 2.1120431423187256,
+      "learning_rate": 8.060658235475978e-06,
+      "loss": 0.5541,
+      "step": 3636
+    },
+    {
+      "epoch": 0.29096,
+      "grad_norm": 1.6898366212844849,
+      "learning_rate": 8.059663649943348e-06,
+      "loss": 0.3644,
+      "step": 3637
+    },
+    {
+      "epoch": 0.29104,
+      "grad_norm": 1.6504249572753906,
+      "learning_rate": 8.058668870836035e-06,
+      "loss": 0.4423,
+      "step": 3638
+    },
+    {
+      "epoch": 0.29112,
+      "grad_norm": 1.3013478517532349,
+      "learning_rate": 8.05767389821698e-06,
+      "loss": 0.3372,
+      "step": 3639
+    },
+    {
+      "epoch": 0.2912,
+      "grad_norm": 1.45965576171875,
+      "learning_rate": 8.056678732149125e-06,
+      "loss": 0.2493,
+      "step": 3640
+    },
+    {
+      "epoch": 0.29128,
+      "grad_norm": 1.4169219732284546,
+      "learning_rate": 8.055683372695437e-06,
+      "loss": 0.3208,
+      "step": 3641
+    },
+    {
+      "epoch": 0.29136,
+      "grad_norm": 1.294084072113037,
+      "learning_rate": 8.054687819918884e-06,
+      "loss": 0.2904,
+      "step": 3642
+    },
+    {
+      "epoch": 0.29144,
+      "grad_norm": 1.3888262510299683,
+      "learning_rate": 8.053692073882456e-06,
+      "loss": 0.3211,
+      "step": 3643
+    },
+    {
+      "epoch": 0.29152,
+      "grad_norm": 1.5449509620666504,
+      "learning_rate": 8.052696134649147e-06,
+      "loss": 0.2892,
+      "step": 3644
+    },
+    {
+      "epoch": 0.2916,
+      "grad_norm": 1.308077096939087,
+      "learning_rate": 8.051700002281967e-06,
+      "loss": 0.2521,
+      "step": 3645
+    },
+    {
+      "epoch": 0.29168,
+      "grad_norm": 1.7512633800506592,
+      "learning_rate": 8.05070367684394e-06,
+      "loss": 0.3653,
+      "step": 3646
+    },
+    {
+      "epoch": 0.29176,
+      "grad_norm": 2.0131380558013916,
+      "learning_rate": 8.0497071583981e-06,
+      "loss": 0.4186,
+      "step": 3647
+    },
+    {
+      "epoch": 0.29184,
+      "grad_norm": 2.0268020629882812,
+      "learning_rate": 8.048710447007491e-06,
+      "loss": 0.4878,
+      "step": 3648
+    },
+    {
+      "epoch": 0.29192,
+      "grad_norm": 1.627658486366272,
+      "learning_rate": 8.047713542735173e-06,
+      "loss": 0.4156,
+      "step": 3649
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 1.7770787477493286,
+      "learning_rate": 8.046716445644217e-06,
+      "loss": 0.3509,
+      "step": 3650
+    },
+    {
+      "epoch": 0.29208,
+      "grad_norm": 1.3930143117904663,
+      "learning_rate": 8.045719155797708e-06,
+      "loss": 0.2563,
+      "step": 3651
+    },
+    {
+      "epoch": 0.29216,
+      "grad_norm": 1.5242277383804321,
+      "learning_rate": 8.044721673258736e-06,
+      "loss": 0.4064,
+      "step": 3652
+    },
+    {
+      "epoch": 0.29224,
+      "grad_norm": 1.174406886100769,
+      "learning_rate": 8.043723998090413e-06,
+      "loss": 0.2569,
+      "step": 3653
+    },
+    {
+      "epoch": 0.29232,
+      "grad_norm": 1.1028757095336914,
+      "learning_rate": 8.042726130355856e-06,
+      "loss": 0.2104,
+      "step": 3654
+    },
+    {
+      "epoch": 0.2924,
+      "grad_norm": 1.654245138168335,
+      "learning_rate": 8.041728070118198e-06,
+      "loss": 0.4263,
+      "step": 3655
+    },
+    {
+      "epoch": 0.29248,
+      "grad_norm": 1.5495033264160156,
+      "learning_rate": 8.040729817440584e-06,
+      "loss": 0.3549,
+      "step": 3656
+    },
+    {
+      "epoch": 0.29256,
+      "grad_norm": 1.5318132638931274,
+      "learning_rate": 8.039731372386168e-06,
+      "loss": 0.3307,
+      "step": 3657
+    },
+    {
+      "epoch": 0.29264,
+      "grad_norm": 1.3253992795944214,
+      "learning_rate": 8.038732735018118e-06,
+      "loss": 0.3224,
+      "step": 3658
+    },
+    {
+      "epoch": 0.29272,
+      "grad_norm": 1.608940601348877,
+      "learning_rate": 8.037733905399616e-06,
+      "loss": 0.3678,
+      "step": 3659
+    },
+    {
+      "epoch": 0.2928,
+      "grad_norm": 1.910239577293396,
+      "learning_rate": 8.036734883593852e-06,
+      "loss": 0.3916,
+      "step": 3660
+    },
+    {
+      "epoch": 0.29288,
+      "grad_norm": 1.2812817096710205,
+      "learning_rate": 8.035735669664037e-06,
+      "loss": 0.296,
+      "step": 3661
+    },
+    {
+      "epoch": 0.29296,
+      "grad_norm": 1.475081205368042,
+      "learning_rate": 8.034736263673382e-06,
+      "loss": 0.3323,
+      "step": 3662
+    },
+    {
+      "epoch": 0.29304,
+      "grad_norm": 1.463401436805725,
+      "learning_rate": 8.033736665685119e-06,
+      "loss": 0.3216,
+      "step": 3663
+    },
+    {
+      "epoch": 0.29312,
+      "grad_norm": 1.6229404211044312,
+      "learning_rate": 8.032736875762486e-06,
+      "loss": 0.3206,
+      "step": 3664
+    },
+    {
+      "epoch": 0.2932,
+      "grad_norm": 1.4625945091247559,
+      "learning_rate": 8.031736893968738e-06,
+      "loss": 0.3063,
+      "step": 3665
+    },
+    {
+      "epoch": 0.29328,
+      "grad_norm": 1.762115716934204,
+      "learning_rate": 8.030736720367143e-06,
+      "loss": 0.3884,
+      "step": 3666
+    },
+    {
+      "epoch": 0.29336,
+      "grad_norm": 1.6735233068466187,
+      "learning_rate": 8.029736355020975e-06,
+      "loss": 0.4038,
+      "step": 3667
+    },
+    {
+      "epoch": 0.29344,
+      "grad_norm": 1.6050176620483398,
+      "learning_rate": 8.028735797993528e-06,
+      "loss": 0.4044,
+      "step": 3668
+    },
+    {
+      "epoch": 0.29352,
+      "grad_norm": 1.3272747993469238,
+      "learning_rate": 8.027735049348099e-06,
+      "loss": 0.2959,
+      "step": 3669
+    },
+    {
+      "epoch": 0.2936,
+      "grad_norm": 1.7667698860168457,
+      "learning_rate": 8.026734109148005e-06,
+      "loss": 0.423,
+      "step": 3670
+    },
+    {
+      "epoch": 0.29368,
+      "grad_norm": 1.7235229015350342,
+      "learning_rate": 8.02573297745657e-06,
+      "loss": 0.3251,
+      "step": 3671
+    },
+    {
+      "epoch": 0.29376,
+      "grad_norm": 1.4288824796676636,
+      "learning_rate": 8.024731654337134e-06,
+      "loss": 0.3099,
+      "step": 3672
+    },
+    {
+      "epoch": 0.29384,
+      "grad_norm": 1.4889461994171143,
+      "learning_rate": 8.023730139853049e-06,
+      "loss": 0.271,
+      "step": 3673
+    },
+    {
+      "epoch": 0.29392,
+      "grad_norm": 1.5079691410064697,
+      "learning_rate": 8.022728434067675e-06,
+      "loss": 0.3027,
+      "step": 3674
+    },
+    {
+      "epoch": 0.294,
+      "grad_norm": 1.52550208568573,
+      "learning_rate": 8.021726537044385e-06,
+      "loss": 0.4014,
+      "step": 3675
+    },
+    {
+      "epoch": 0.29408,
+      "grad_norm": 1.9863523244857788,
+      "learning_rate": 8.020724448846569e-06,
+      "loss": 0.4531,
+      "step": 3676
+    },
+    {
+      "epoch": 0.29416,
+      "grad_norm": 1.5718297958374023,
+      "learning_rate": 8.019722169537624e-06,
+      "loss": 0.4036,
+      "step": 3677
+    },
+    {
+      "epoch": 0.29424,
+      "grad_norm": 1.8104145526885986,
+      "learning_rate": 8.018719699180961e-06,
+      "loss": 0.3399,
+      "step": 3678
+    },
+    {
+      "epoch": 0.29432,
+      "grad_norm": 1.6687524318695068,
+      "learning_rate": 8.017717037840005e-06,
+      "loss": 0.3931,
+      "step": 3679
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 1.5040487051010132,
+      "learning_rate": 8.016714185578189e-06,
+      "loss": 0.3671,
+      "step": 3680
+    },
+    {
+      "epoch": 0.29448,
+      "grad_norm": 1.4653910398483276,
+      "learning_rate": 8.01571114245896e-06,
+      "loss": 0.3147,
+      "step": 3681
+    },
+    {
+      "epoch": 0.29456,
+      "grad_norm": 1.5103580951690674,
+      "learning_rate": 8.014707908545776e-06,
+      "loss": 0.3386,
+      "step": 3682
+    },
+    {
+      "epoch": 0.29464,
+      "grad_norm": 1.5704234838485718,
+      "learning_rate": 8.013704483902112e-06,
+      "loss": 0.2921,
+      "step": 3683
+    },
+    {
+      "epoch": 0.29472,
+      "grad_norm": 1.7446131706237793,
+      "learning_rate": 8.012700868591449e-06,
+      "loss": 0.3766,
+      "step": 3684
+    },
+    {
+      "epoch": 0.2948,
+      "grad_norm": 1.2436984777450562,
+      "learning_rate": 8.011697062677282e-06,
+      "loss": 0.2697,
+      "step": 3685
+    },
+    {
+      "epoch": 0.29488,
+      "grad_norm": 1.496191382408142,
+      "learning_rate": 8.01069306622312e-06,
+      "loss": 0.325,
+      "step": 3686
+    },
+    {
+      "epoch": 0.29496,
+      "grad_norm": 1.426172137260437,
+      "learning_rate": 8.00968887929248e-06,
+      "loss": 0.2959,
+      "step": 3687
+    },
+    {
+      "epoch": 0.29504,
+      "grad_norm": 1.596839427947998,
+      "learning_rate": 8.008684501948895e-06,
+      "loss": 0.3525,
+      "step": 3688
+    },
+    {
+      "epoch": 0.29512,
+      "grad_norm": 1.2332483530044556,
+      "learning_rate": 8.00767993425591e-06,
+      "loss": 0.2837,
+      "step": 3689
+    },
+    {
+      "epoch": 0.2952,
+      "grad_norm": 1.593542218208313,
+      "learning_rate": 8.00667517627708e-06,
+      "loss": 0.3293,
+      "step": 3690
+    },
+    {
+      "epoch": 0.29528,
+      "grad_norm": 1.504442572593689,
+      "learning_rate": 8.00567022807597e-06,
+      "loss": 0.3502,
+      "step": 3691
+    },
+    {
+      "epoch": 0.29536,
+      "grad_norm": 1.763466477394104,
+      "learning_rate": 8.004665089716162e-06,
+      "loss": 0.3728,
+      "step": 3692
+    },
+    {
+      "epoch": 0.29544,
+      "grad_norm": 1.6120593547821045,
+      "learning_rate": 8.003659761261248e-06,
+      "loss": 0.314,
+      "step": 3693
+    },
+    {
+      "epoch": 0.29552,
+      "grad_norm": 1.3458003997802734,
+      "learning_rate": 8.00265424277483e-06,
+      "loss": 0.3373,
+      "step": 3694
+    },
+    {
+      "epoch": 0.2956,
+      "grad_norm": 1.3163591623306274,
+      "learning_rate": 8.001648534320526e-06,
+      "loss": 0.325,
+      "step": 3695
+    },
+    {
+      "epoch": 0.29568,
+      "grad_norm": 1.616263747215271,
+      "learning_rate": 8.000642635961963e-06,
+      "loss": 0.4398,
+      "step": 3696
+    },
+    {
+      "epoch": 0.29576,
+      "grad_norm": 1.4315654039382935,
+      "learning_rate": 7.99963654776278e-06,
+      "loss": 0.3828,
+      "step": 3697
+    },
+    {
+      "epoch": 0.29584,
+      "grad_norm": 1.517011284828186,
+      "learning_rate": 7.99863026978663e-06,
+      "loss": 0.3207,
+      "step": 3698
+    },
+    {
+      "epoch": 0.29592,
+      "grad_norm": 1.4280073642730713,
+      "learning_rate": 7.997623802097176e-06,
+      "loss": 0.3207,
+      "step": 3699
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 1.3090413808822632,
+      "learning_rate": 7.996617144758094e-06,
+      "loss": 0.3191,
+      "step": 3700
+    },
+    {
+      "epoch": 0.29608,
+      "grad_norm": 1.6390103101730347,
+      "learning_rate": 7.995610297833072e-06,
+      "loss": 0.3944,
+      "step": 3701
+    },
+    {
+      "epoch": 0.29616,
+      "grad_norm": 1.4760853052139282,
+      "learning_rate": 7.994603261385809e-06,
+      "loss": 0.3428,
+      "step": 3702
+    },
+    {
+      "epoch": 0.29624,
+      "grad_norm": 1.1745274066925049,
+      "learning_rate": 7.99359603548002e-06,
+      "loss": 0.2517,
+      "step": 3703
+    },
+    {
+      "epoch": 0.29632,
+      "grad_norm": 1.4406036138534546,
+      "learning_rate": 7.992588620179424e-06,
+      "loss": 0.3696,
+      "step": 3704
+    },
+    {
+      "epoch": 0.2964,
+      "grad_norm": 1.278671383857727,
+      "learning_rate": 7.99158101554776e-06,
+      "loss": 0.249,
+      "step": 3705
+    },
+    {
+      "epoch": 0.29648,
+      "grad_norm": 1.420791506767273,
+      "learning_rate": 7.990573221648775e-06,
+      "loss": 0.2905,
+      "step": 3706
+    },
+    {
+      "epoch": 0.29656,
+      "grad_norm": 1.7047336101531982,
+      "learning_rate": 7.989565238546228e-06,
+      "loss": 0.4367,
+      "step": 3707
+    },
+    {
+      "epoch": 0.29664,
+      "grad_norm": 2.036435127258301,
+      "learning_rate": 7.988557066303892e-06,
+      "loss": 0.3923,
+      "step": 3708
+    },
+    {
+      "epoch": 0.29672,
+      "grad_norm": 1.4706825017929077,
+      "learning_rate": 7.987548704985553e-06,
+      "loss": 0.429,
+      "step": 3709
+    },
+    {
+      "epoch": 0.2968,
+      "grad_norm": 1.6444703340530396,
+      "learning_rate": 7.986540154655e-06,
+      "loss": 0.3174,
+      "step": 3710
+    },
+    {
+      "epoch": 0.29688,
+      "grad_norm": 1.5745984315872192,
+      "learning_rate": 7.985531415376046e-06,
+      "loss": 0.2996,
+      "step": 3711
+    },
+    {
+      "epoch": 0.29696,
+      "grad_norm": 1.4535049200057983,
+      "learning_rate": 7.984522487212509e-06,
+      "loss": 0.3336,
+      "step": 3712
+    },
+    {
+      "epoch": 0.29704,
+      "grad_norm": 1.7702298164367676,
+      "learning_rate": 7.98351337022822e-06,
+      "loss": 0.388,
+      "step": 3713
+    },
+    {
+      "epoch": 0.29712,
+      "grad_norm": 1.5240532159805298,
+      "learning_rate": 7.982504064487022e-06,
+      "loss": 0.3296,
+      "step": 3714
+    },
+    {
+      "epoch": 0.2972,
+      "grad_norm": 1.6705554723739624,
+      "learning_rate": 7.981494570052775e-06,
+      "loss": 0.3306,
+      "step": 3715
+    },
+    {
+      "epoch": 0.29728,
+      "grad_norm": 1.8875733613967896,
+      "learning_rate": 7.980484886989338e-06,
+      "loss": 0.3438,
+      "step": 3716
+    },
+    {
+      "epoch": 0.29736,
+      "grad_norm": 1.5699726343154907,
+      "learning_rate": 7.979475015360599e-06,
+      "loss": 0.3333,
+      "step": 3717
+    },
+    {
+      "epoch": 0.29744,
+      "grad_norm": 1.580299973487854,
+      "learning_rate": 7.978464955230442e-06,
+      "loss": 0.3609,
+      "step": 3718
+    },
+    {
+      "epoch": 0.29752,
+      "grad_norm": 1.7409504652023315,
+      "learning_rate": 7.977454706662775e-06,
+      "loss": 0.3506,
+      "step": 3719
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 2.2885334491729736,
+      "learning_rate": 7.97644426972151e-06,
+      "loss": 0.4091,
+      "step": 3720
+    },
+    {
+      "epoch": 0.29768,
+      "grad_norm": 1.6000932455062866,
+      "learning_rate": 7.975433644470576e-06,
+      "loss": 0.3848,
+      "step": 3721
+    },
+    {
+      "epoch": 0.29776,
+      "grad_norm": 1.8004785776138306,
+      "learning_rate": 7.974422830973912e-06,
+      "loss": 0.3568,
+      "step": 3722
+    },
+    {
+      "epoch": 0.29784,
+      "grad_norm": 1.5973020792007446,
+      "learning_rate": 7.973411829295466e-06,
+      "loss": 0.3497,
+      "step": 3723
+    },
+    {
+      "epoch": 0.29792,
+      "grad_norm": 1.4873243570327759,
+      "learning_rate": 7.972400639499204e-06,
+      "loss": 0.3358,
+      "step": 3724
+    },
+    {
+      "epoch": 0.298,
+      "grad_norm": 1.824233055114746,
+      "learning_rate": 7.971389261649099e-06,
+      "loss": 0.4382,
+      "step": 3725
+    },
+    {
+      "epoch": 0.29808,
+      "grad_norm": 1.824907660484314,
+      "learning_rate": 7.970377695809138e-06,
+      "loss": 0.4416,
+      "step": 3726
+    },
+    {
+      "epoch": 0.29816,
+      "grad_norm": 1.4309464693069458,
+      "learning_rate": 7.96936594204332e-06,
+      "loss": 0.2923,
+      "step": 3727
+    },
+    {
+      "epoch": 0.29824,
+      "grad_norm": 1.8320714235305786,
+      "learning_rate": 7.968354000415652e-06,
+      "loss": 0.4268,
+      "step": 3728
+    },
+    {
+      "epoch": 0.29832,
+      "grad_norm": 1.6846635341644287,
+      "learning_rate": 7.967341870990159e-06,
+      "loss": 0.297,
+      "step": 3729
+    },
+    {
+      "epoch": 0.2984,
+      "grad_norm": 1.6068602800369263,
+      "learning_rate": 7.966329553830876e-06,
+      "loss": 0.3604,
+      "step": 3730
+    },
+    {
+      "epoch": 0.29848,
+      "grad_norm": 1.7175391912460327,
+      "learning_rate": 7.965317049001847e-06,
+      "loss": 0.394,
+      "step": 3731
+    },
+    {
+      "epoch": 0.29856,
+      "grad_norm": 1.2361955642700195,
+      "learning_rate": 7.96430435656713e-06,
+      "loss": 0.2636,
+      "step": 3732
+    },
+    {
+      "epoch": 0.29864,
+      "grad_norm": 1.6340537071228027,
+      "learning_rate": 7.963291476590795e-06,
+      "loss": 0.3437,
+      "step": 3733
+    },
+    {
+      "epoch": 0.29872,
+      "grad_norm": 1.4597103595733643,
+      "learning_rate": 7.962278409136924e-06,
+      "loss": 0.307,
+      "step": 3734
+    },
+    {
+      "epoch": 0.2988,
+      "grad_norm": 1.6404805183410645,
+      "learning_rate": 7.961265154269608e-06,
+      "loss": 0.5112,
+      "step": 3735
+    },
+    {
+      "epoch": 0.29888,
+      "grad_norm": 1.6717686653137207,
+      "learning_rate": 7.960251712052955e-06,
+      "loss": 0.3567,
+      "step": 3736
+    },
+    {
+      "epoch": 0.29896,
+      "grad_norm": 1.874589204788208,
+      "learning_rate": 7.959238082551081e-06,
+      "loss": 0.4107,
+      "step": 3737
+    },
+    {
+      "epoch": 0.29904,
+      "grad_norm": 1.2089720964431763,
+      "learning_rate": 7.958224265828118e-06,
+      "loss": 0.3063,
+      "step": 3738
+    },
+    {
+      "epoch": 0.29912,
+      "grad_norm": 1.2900773286819458,
+      "learning_rate": 7.957210261948201e-06,
+      "loss": 0.2518,
+      "step": 3739
+    },
+    {
+      "epoch": 0.2992,
+      "grad_norm": 1.737890362739563,
+      "learning_rate": 7.956196070975485e-06,
+      "loss": 0.3588,
+      "step": 3740
+    },
+    {
+      "epoch": 0.29928,
+      "grad_norm": 1.681058406829834,
+      "learning_rate": 7.955181692974138e-06,
+      "loss": 0.3226,
+      "step": 3741
+    },
+    {
+      "epoch": 0.29936,
+      "grad_norm": 1.6634056568145752,
+      "learning_rate": 7.954167128008332e-06,
+      "loss": 0.3443,
+      "step": 3742
+    },
+    {
+      "epoch": 0.29944,
+      "grad_norm": 1.8775358200073242,
+      "learning_rate": 7.953152376142255e-06,
+      "loss": 0.3255,
+      "step": 3743
+    },
+    {
+      "epoch": 0.29952,
+      "grad_norm": 1.728934645652771,
+      "learning_rate": 7.95213743744011e-06,
+      "loss": 0.3183,
+      "step": 3744
+    },
+    {
+      "epoch": 0.2996,
+      "grad_norm": 1.514907717704773,
+      "learning_rate": 7.951122311966108e-06,
+      "loss": 0.3914,
+      "step": 3745
+    },
+    {
+      "epoch": 0.29968,
+      "grad_norm": 1.5943530797958374,
+      "learning_rate": 7.95010699978447e-06,
+      "loss": 0.349,
+      "step": 3746
+    },
+    {
+      "epoch": 0.29976,
+      "grad_norm": 1.815433382987976,
+      "learning_rate": 7.949091500959434e-06,
+      "loss": 0.4143,
+      "step": 3747
+    },
+    {
+      "epoch": 0.29984,
+      "grad_norm": 1.7097547054290771,
+      "learning_rate": 7.948075815555246e-06,
+      "loss": 0.3884,
+      "step": 3748
+    },
+    {
+      "epoch": 0.29992,
+      "grad_norm": 1.619554877281189,
+      "learning_rate": 7.947059943636166e-06,
+      "loss": 0.3996,
+      "step": 3749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.394123911857605,
+      "learning_rate": 7.946043885266465e-06,
+      "loss": 0.2721,
+      "step": 3750
+    },
+    {
+      "epoch": 0.30008,
+      "grad_norm": 1.3874683380126953,
+      "learning_rate": 7.945027640510423e-06,
+      "loss": 0.3412,
+      "step": 3751
+    },
+    {
+      "epoch": 0.30016,
+      "grad_norm": 1.5768663883209229,
+      "learning_rate": 7.944011209432336e-06,
+      "loss": 0.3391,
+      "step": 3752
+    },
+    {
+      "epoch": 0.30024,
+      "grad_norm": 1.4615979194641113,
+      "learning_rate": 7.942994592096513e-06,
+      "loss": 0.2966,
+      "step": 3753
+    },
+    {
+      "epoch": 0.30032,
+      "grad_norm": 1.3654316663742065,
+      "learning_rate": 7.941977788567267e-06,
+      "loss": 0.3807,
+      "step": 3754
+    },
+    {
+      "epoch": 0.3004,
+      "grad_norm": 1.8147246837615967,
+      "learning_rate": 7.940960798908933e-06,
+      "loss": 0.3741,
+      "step": 3755
+    },
+    {
+      "epoch": 0.30048,
+      "grad_norm": 1.5056862831115723,
+      "learning_rate": 7.939943623185847e-06,
+      "loss": 0.3309,
+      "step": 3756
+    },
+    {
+      "epoch": 0.30056,
+      "grad_norm": 1.4294978380203247,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.3607,
+      "step": 3757
+    },
+    {
+      "epoch": 0.30064,
+      "grad_norm": 1.4605019092559814,
+      "learning_rate": 7.937908713802855e-06,
+      "loss": 0.3732,
+      "step": 3758
+    },
+    {
+      "epoch": 0.30072,
+      "grad_norm": 1.7275772094726562,
+      "learning_rate": 7.936890980271688e-06,
+      "loss": 0.3738,
+      "step": 3759
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 1.480759620666504,
+      "learning_rate": 7.935873060933257e-06,
+      "loss": 0.3517,
+      "step": 3760
+    },
+    {
+      "epoch": 0.30088,
+      "grad_norm": 1.655974268913269,
+      "learning_rate": 7.934854955851961e-06,
+      "loss": 0.3499,
+      "step": 3761
+    },
+    {
+      "epoch": 0.30096,
+      "grad_norm": 1.35233736038208,
+      "learning_rate": 7.933836665092212e-06,
+      "loss": 0.3247,
+      "step": 3762
+    },
+    {
+      "epoch": 0.30104,
+      "grad_norm": 1.2412519454956055,
+      "learning_rate": 7.932818188718433e-06,
+      "loss": 0.2889,
+      "step": 3763
+    },
+    {
+      "epoch": 0.30112,
+      "grad_norm": 1.5485812425613403,
+      "learning_rate": 7.931799526795062e-06,
+      "loss": 0.3616,
+      "step": 3764
+    },
+    {
+      "epoch": 0.3012,
+      "grad_norm": 1.448938250541687,
+      "learning_rate": 7.930780679386542e-06,
+      "loss": 0.3128,
+      "step": 3765
+    },
+    {
+      "epoch": 0.30128,
+      "grad_norm": 1.4666680097579956,
+      "learning_rate": 7.929761646557337e-06,
+      "loss": 0.3414,
+      "step": 3766
+    },
+    {
+      "epoch": 0.30136,
+      "grad_norm": 1.6032980680465698,
+      "learning_rate": 7.928742428371916e-06,
+      "loss": 0.3667,
+      "step": 3767
+    },
+    {
+      "epoch": 0.30144,
+      "grad_norm": 1.5190998315811157,
+      "learning_rate": 7.92772302489476e-06,
+      "loss": 0.3274,
+      "step": 3768
+    },
+    {
+      "epoch": 0.30152,
+      "grad_norm": 1.6384193897247314,
+      "learning_rate": 7.926703436190363e-06,
+      "loss": 0.371,
+      "step": 3769
+    },
+    {
+      "epoch": 0.3016,
+      "grad_norm": 1.7404835224151611,
+      "learning_rate": 7.925683662323235e-06,
+      "loss": 0.3464,
+      "step": 3770
+    },
+    {
+      "epoch": 0.30168,
+      "grad_norm": 1.719647765159607,
+      "learning_rate": 7.924663703357892e-06,
+      "loss": 0.371,
+      "step": 3771
+    },
+    {
+      "epoch": 0.30176,
+      "grad_norm": 1.545730471611023,
+      "learning_rate": 7.923643559358858e-06,
+      "loss": 0.3684,
+      "step": 3772
+    },
+    {
+      "epoch": 0.30184,
+      "grad_norm": 1.717699646949768,
+      "learning_rate": 7.922623230390682e-06,
+      "loss": 0.3934,
+      "step": 3773
+    },
+    {
+      "epoch": 0.30192,
+      "grad_norm": 1.8281782865524292,
+      "learning_rate": 7.921602716517914e-06,
+      "loss": 0.508,
+      "step": 3774
+    },
+    {
+      "epoch": 0.302,
+      "grad_norm": 2.0022242069244385,
+      "learning_rate": 7.920582017805114e-06,
+      "loss": 0.36,
+      "step": 3775
+    },
+    {
+      "epoch": 0.30208,
+      "grad_norm": 1.4404866695404053,
+      "learning_rate": 7.919561134316865e-06,
+      "loss": 0.2932,
+      "step": 3776
+    },
+    {
+      "epoch": 0.30216,
+      "grad_norm": 1.687625527381897,
+      "learning_rate": 7.918540066117752e-06,
+      "loss": 0.3467,
+      "step": 3777
+    },
+    {
+      "epoch": 0.30224,
+      "grad_norm": 1.2726261615753174,
+      "learning_rate": 7.917518813272373e-06,
+      "loss": 0.3044,
+      "step": 3778
+    },
+    {
+      "epoch": 0.30232,
+      "grad_norm": 2.043890953063965,
+      "learning_rate": 7.916497375845342e-06,
+      "loss": 0.3605,
+      "step": 3779
+    },
+    {
+      "epoch": 0.3024,
+      "grad_norm": 1.251383900642395,
+      "learning_rate": 7.91547575390128e-06,
+      "loss": 0.3175,
+      "step": 3780
+    },
+    {
+      "epoch": 0.30248,
+      "grad_norm": 1.544699788093567,
+      "learning_rate": 7.914453947504822e-06,
+      "loss": 0.3783,
+      "step": 3781
+    },
+    {
+      "epoch": 0.30256,
+      "grad_norm": 1.4141426086425781,
+      "learning_rate": 7.913431956720615e-06,
+      "loss": 0.26,
+      "step": 3782
+    },
+    {
+      "epoch": 0.30264,
+      "grad_norm": 1.7382965087890625,
+      "learning_rate": 7.912409781613317e-06,
+      "loss": 0.3664,
+      "step": 3783
+    },
+    {
+      "epoch": 0.30272,
+      "grad_norm": 1.1553996801376343,
+      "learning_rate": 7.911387422247596e-06,
+      "loss": 0.2621,
+      "step": 3784
+    },
+    {
+      "epoch": 0.3028,
+      "grad_norm": 1.534834623336792,
+      "learning_rate": 7.910364878688135e-06,
+      "loss": 0.3252,
+      "step": 3785
+    },
+    {
+      "epoch": 0.30288,
+      "grad_norm": 1.4503859281539917,
+      "learning_rate": 7.909342150999626e-06,
+      "loss": 0.3371,
+      "step": 3786
+    },
+    {
+      "epoch": 0.30296,
+      "grad_norm": 1.7136962413787842,
+      "learning_rate": 7.908319239246774e-06,
+      "loss": 0.3998,
+      "step": 3787
+    },
+    {
+      "epoch": 0.30304,
+      "grad_norm": 1.6705702543258667,
+      "learning_rate": 7.907296143494293e-06,
+      "loss": 0.4058,
+      "step": 3788
+    },
+    {
+      "epoch": 0.30312,
+      "grad_norm": 1.4927536249160767,
+      "learning_rate": 7.906272863806916e-06,
+      "loss": 0.3728,
+      "step": 3789
+    },
+    {
+      "epoch": 0.3032,
+      "grad_norm": 1.1272435188293457,
+      "learning_rate": 7.905249400249377e-06,
+      "loss": 0.2479,
+      "step": 3790
+    },
+    {
+      "epoch": 0.30328,
+      "grad_norm": 1.3360072374343872,
+      "learning_rate": 7.904225752886433e-06,
+      "loss": 0.269,
+      "step": 3791
+    },
+    {
+      "epoch": 0.30336,
+      "grad_norm": 1.445589542388916,
+      "learning_rate": 7.90320192178284e-06,
+      "loss": 0.311,
+      "step": 3792
+    },
+    {
+      "epoch": 0.30344,
+      "grad_norm": 2.228759527206421,
+      "learning_rate": 7.902177907003375e-06,
+      "loss": 0.4801,
+      "step": 3793
+    },
+    {
+      "epoch": 0.30352,
+      "grad_norm": 1.5473653078079224,
+      "learning_rate": 7.901153708612825e-06,
+      "loss": 0.2992,
+      "step": 3794
+    },
+    {
+      "epoch": 0.3036,
+      "grad_norm": 1.85487699508667,
+      "learning_rate": 7.90012932667599e-06,
+      "loss": 0.3764,
+      "step": 3795
+    },
+    {
+      "epoch": 0.30368,
+      "grad_norm": 1.6702008247375488,
+      "learning_rate": 7.899104761257673e-06,
+      "loss": 0.3616,
+      "step": 3796
+    },
+    {
+      "epoch": 0.30376,
+      "grad_norm": 1.6418166160583496,
+      "learning_rate": 7.8980800124227e-06,
+      "loss": 0.3333,
+      "step": 3797
+    },
+    {
+      "epoch": 0.30384,
+      "grad_norm": 1.5101503133773804,
+      "learning_rate": 7.8970550802359e-06,
+      "loss": 0.3233,
+      "step": 3798
+    },
+    {
+      "epoch": 0.30392,
+      "grad_norm": 1.378827452659607,
+      "learning_rate": 7.896029964762119e-06,
+      "loss": 0.2467,
+      "step": 3799
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 1.2998595237731934,
+      "learning_rate": 7.895004666066214e-06,
+      "loss": 0.2874,
+      "step": 3800
+    },
+    {
+      "epoch": 0.30408,
+      "grad_norm": 1.455960750579834,
+      "learning_rate": 7.893979184213049e-06,
+      "loss": 0.2777,
+      "step": 3801
+    },
+    {
+      "epoch": 0.30416,
+      "grad_norm": 1.5600316524505615,
+      "learning_rate": 7.892953519267506e-06,
+      "loss": 0.3268,
+      "step": 3802
+    },
+    {
+      "epoch": 0.30424,
+      "grad_norm": 1.37201726436615,
+      "learning_rate": 7.891927671294472e-06,
+      "loss": 0.3666,
+      "step": 3803
+    },
+    {
+      "epoch": 0.30432,
+      "grad_norm": 1.4936875104904175,
+      "learning_rate": 7.890901640358852e-06,
+      "loss": 0.2798,
+      "step": 3804
+    },
+    {
+      "epoch": 0.3044,
+      "grad_norm": 1.624596357345581,
+      "learning_rate": 7.889875426525557e-06,
+      "loss": 0.4087,
+      "step": 3805
+    },
+    {
+      "epoch": 0.30448,
+      "grad_norm": 2.078230381011963,
+      "learning_rate": 7.888849029859513e-06,
+      "loss": 0.4298,
+      "step": 3806
+    },
+    {
+      "epoch": 0.30456,
+      "grad_norm": 1.4350990056991577,
+      "learning_rate": 7.887822450425658e-06,
+      "loss": 0.3688,
+      "step": 3807
+    },
+    {
+      "epoch": 0.30464,
+      "grad_norm": 1.392849087715149,
+      "learning_rate": 7.886795688288937e-06,
+      "loss": 0.3813,
+      "step": 3808
+    },
+    {
+      "epoch": 0.30472,
+      "grad_norm": 1.645775556564331,
+      "learning_rate": 7.885768743514316e-06,
+      "loss": 0.3152,
+      "step": 3809
+    },
+    {
+      "epoch": 0.3048,
+      "grad_norm": 1.6383919715881348,
+      "learning_rate": 7.88474161616676e-06,
+      "loss": 0.3558,
+      "step": 3810
+    },
+    {
+      "epoch": 0.30488,
+      "grad_norm": 2.151557445526123,
+      "learning_rate": 7.883714306311255e-06,
+      "loss": 0.4004,
+      "step": 3811
+    },
+    {
+      "epoch": 0.30496,
+      "grad_norm": 1.6432582139968872,
+      "learning_rate": 7.882686814012792e-06,
+      "loss": 0.4162,
+      "step": 3812
+    },
+    {
+      "epoch": 0.30504,
+      "grad_norm": 1.6074105501174927,
+      "learning_rate": 7.881659139336383e-06,
+      "loss": 0.3813,
+      "step": 3813
+    },
+    {
+      "epoch": 0.30512,
+      "grad_norm": 1.7226243019104004,
+      "learning_rate": 7.880631282347042e-06,
+      "loss": 0.446,
+      "step": 3814
+    },
+    {
+      "epoch": 0.3052,
+      "grad_norm": 1.3685539960861206,
+      "learning_rate": 7.879603243109799e-06,
+      "loss": 0.2664,
+      "step": 3815
+    },
+    {
+      "epoch": 0.30528,
+      "grad_norm": 2.002601146697998,
+      "learning_rate": 7.878575021689693e-06,
+      "loss": 0.402,
+      "step": 3816
+    },
+    {
+      "epoch": 0.30536,
+      "grad_norm": 1.7490203380584717,
+      "learning_rate": 7.877546618151776e-06,
+      "loss": 0.3748,
+      "step": 3817
+    },
+    {
+      "epoch": 0.30544,
+      "grad_norm": 1.197702169418335,
+      "learning_rate": 7.876518032561113e-06,
+      "loss": 0.2688,
+      "step": 3818
+    },
+    {
+      "epoch": 0.30552,
+      "grad_norm": 1.7372758388519287,
+      "learning_rate": 7.875489264982781e-06,
+      "loss": 0.377,
+      "step": 3819
+    },
+    {
+      "epoch": 0.3056,
+      "grad_norm": 1.6473854780197144,
+      "learning_rate": 7.874460315481863e-06,
+      "loss": 0.3239,
+      "step": 3820
+    },
+    {
+      "epoch": 0.30568,
+      "grad_norm": 1.2986738681793213,
+      "learning_rate": 7.87343118412346e-06,
+      "loss": 0.3873,
+      "step": 3821
+    },
+    {
+      "epoch": 0.30576,
+      "grad_norm": 2.0069668292999268,
+      "learning_rate": 7.872401870972679e-06,
+      "loss": 0.3118,
+      "step": 3822
+    },
+    {
+      "epoch": 0.30584,
+      "grad_norm": 1.1595829725265503,
+      "learning_rate": 7.871372376094642e-06,
+      "loss": 0.3185,
+      "step": 3823
+    },
+    {
+      "epoch": 0.30592,
+      "grad_norm": 1.917394995689392,
+      "learning_rate": 7.870342699554484e-06,
+      "loss": 0.4492,
+      "step": 3824
+    },
+    {
+      "epoch": 0.306,
+      "grad_norm": 1.6679039001464844,
+      "learning_rate": 7.869312841417346e-06,
+      "loss": 0.4018,
+      "step": 3825
+    },
+    {
+      "epoch": 0.30608,
+      "grad_norm": 1.2022053003311157,
+      "learning_rate": 7.868282801748389e-06,
+      "loss": 0.2811,
+      "step": 3826
+    },
+    {
+      "epoch": 0.30616,
+      "grad_norm": 1.3668828010559082,
+      "learning_rate": 7.867252580612775e-06,
+      "loss": 0.2921,
+      "step": 3827
+    },
+    {
+      "epoch": 0.30624,
+      "grad_norm": 1.7087875604629517,
+      "learning_rate": 7.866222178075681e-06,
+      "loss": 0.3836,
+      "step": 3828
+    },
+    {
+      "epoch": 0.30632,
+      "grad_norm": 1.1265833377838135,
+      "learning_rate": 7.865191594202302e-06,
+      "loss": 0.2809,
+      "step": 3829
+    },
+    {
+      "epoch": 0.3064,
+      "grad_norm": 1.4156330823898315,
+      "learning_rate": 7.864160829057838e-06,
+      "loss": 0.3044,
+      "step": 3830
+    },
+    {
+      "epoch": 0.30648,
+      "grad_norm": 1.5269126892089844,
+      "learning_rate": 7.863129882707504e-06,
+      "loss": 0.3253,
+      "step": 3831
+    },
+    {
+      "epoch": 0.30656,
+      "grad_norm": 1.6441112756729126,
+      "learning_rate": 7.862098755216519e-06,
+      "loss": 0.3304,
+      "step": 3832
+    },
+    {
+      "epoch": 0.30664,
+      "grad_norm": 1.804513931274414,
+      "learning_rate": 7.861067446650124e-06,
+      "loss": 0.3199,
+      "step": 3833
+    },
+    {
+      "epoch": 0.30672,
+      "grad_norm": 1.4178650379180908,
+      "learning_rate": 7.860035957073566e-06,
+      "loss": 0.326,
+      "step": 3834
+    },
+    {
+      "epoch": 0.3068,
+      "grad_norm": 1.4408081769943237,
+      "learning_rate": 7.8590042865521e-06,
+      "loss": 0.3487,
+      "step": 3835
+    },
+    {
+      "epoch": 0.30688,
+      "grad_norm": 1.7407159805297852,
+      "learning_rate": 7.857972435151e-06,
+      "loss": 0.3687,
+      "step": 3836
+    },
+    {
+      "epoch": 0.30696,
+      "grad_norm": 1.395830512046814,
+      "learning_rate": 7.856940402935547e-06,
+      "loss": 0.2618,
+      "step": 3837
+    },
+    {
+      "epoch": 0.30704,
+      "grad_norm": 2.0533602237701416,
+      "learning_rate": 7.855908189971036e-06,
+      "loss": 0.3489,
+      "step": 3838
+    },
+    {
+      "epoch": 0.30712,
+      "grad_norm": 1.6558175086975098,
+      "learning_rate": 7.854875796322767e-06,
+      "loss": 0.3476,
+      "step": 3839
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 1.524119257926941,
+      "learning_rate": 7.853843222056059e-06,
+      "loss": 0.4397,
+      "step": 3840
+    },
+    {
+      "epoch": 0.30728,
+      "grad_norm": 1.5205790996551514,
+      "learning_rate": 7.85281046723624e-06,
+      "loss": 0.3255,
+      "step": 3841
+    },
+    {
+      "epoch": 0.30736,
+      "grad_norm": 1.547203779220581,
+      "learning_rate": 7.851777531928648e-06,
+      "loss": 0.369,
+      "step": 3842
+    },
+    {
+      "epoch": 0.30744,
+      "grad_norm": 1.3715342283248901,
+      "learning_rate": 7.850744416198635e-06,
+      "loss": 0.2742,
+      "step": 3843
+    },
+    {
+      "epoch": 0.30752,
+      "grad_norm": 1.7399414777755737,
+      "learning_rate": 7.849711120111561e-06,
+      "loss": 0.3737,
+      "step": 3844
+    },
+    {
+      "epoch": 0.3076,
+      "grad_norm": 1.67813241481781,
+      "learning_rate": 7.848677643732799e-06,
+      "loss": 0.3005,
+      "step": 3845
+    },
+    {
+      "epoch": 0.30768,
+      "grad_norm": 1.495720386505127,
+      "learning_rate": 7.847643987127733e-06,
+      "loss": 0.3383,
+      "step": 3846
+    },
+    {
+      "epoch": 0.30776,
+      "grad_norm": 1.943218469619751,
+      "learning_rate": 7.846610150361761e-06,
+      "loss": 0.4522,
+      "step": 3847
+    },
+    {
+      "epoch": 0.30784,
+      "grad_norm": 2.0549914836883545,
+      "learning_rate": 7.84557613350029e-06,
+      "loss": 0.4285,
+      "step": 3848
+    },
+    {
+      "epoch": 0.30792,
+      "grad_norm": 2.2861602306365967,
+      "learning_rate": 7.844541936608738e-06,
+      "loss": 0.5222,
+      "step": 3849
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 1.814911961555481,
+      "learning_rate": 7.843507559752537e-06,
+      "loss": 0.4032,
+      "step": 3850
+    },
+    {
+      "epoch": 0.30808,
+      "grad_norm": 1.2270601987838745,
+      "learning_rate": 7.842473002997127e-06,
+      "loss": 0.2968,
+      "step": 3851
+    },
+    {
+      "epoch": 0.30816,
+      "grad_norm": 1.692413091659546,
+      "learning_rate": 7.841438266407959e-06,
+      "loss": 0.4398,
+      "step": 3852
+    },
+    {
+      "epoch": 0.30824,
+      "grad_norm": 1.7804385423660278,
+      "learning_rate": 7.8404033500505e-06,
+      "loss": 0.4585,
+      "step": 3853
+    },
+    {
+      "epoch": 0.30832,
+      "grad_norm": 1.3115049600601196,
+      "learning_rate": 7.839368253990224e-06,
+      "loss": 0.2864,
+      "step": 3854
+    },
+    {
+      "epoch": 0.3084,
+      "grad_norm": 1.490980625152588,
+      "learning_rate": 7.838332978292622e-06,
+      "loss": 0.3292,
+      "step": 3855
+    },
+    {
+      "epoch": 0.30848,
+      "grad_norm": 1.78223717212677,
+      "learning_rate": 7.837297523023189e-06,
+      "loss": 0.4908,
+      "step": 3856
+    },
+    {
+      "epoch": 0.30856,
+      "grad_norm": 1.4501593112945557,
+      "learning_rate": 7.836261888247434e-06,
+      "loss": 0.3176,
+      "step": 3857
+    },
+    {
+      "epoch": 0.30864,
+      "grad_norm": 1.7654818296432495,
+      "learning_rate": 7.83522607403088e-06,
+      "loss": 0.3421,
+      "step": 3858
+    },
+    {
+      "epoch": 0.30872,
+      "grad_norm": 1.4273486137390137,
+      "learning_rate": 7.834190080439058e-06,
+      "loss": 0.3251,
+      "step": 3859
+    },
+    {
+      "epoch": 0.3088,
+      "grad_norm": 1.3740286827087402,
+      "learning_rate": 7.833153907537515e-06,
+      "loss": 0.3186,
+      "step": 3860
+    },
+    {
+      "epoch": 0.30888,
+      "grad_norm": 1.4216959476470947,
+      "learning_rate": 7.832117555391803e-06,
+      "loss": 0.3298,
+      "step": 3861
+    },
+    {
+      "epoch": 0.30896,
+      "grad_norm": 1.8324718475341797,
+      "learning_rate": 7.831081024067489e-06,
+      "loss": 0.4795,
+      "step": 3862
+    },
+    {
+      "epoch": 0.30904,
+      "grad_norm": 1.479018211364746,
+      "learning_rate": 7.83004431363015e-06,
+      "loss": 0.32,
+      "step": 3863
+    },
+    {
+      "epoch": 0.30912,
+      "grad_norm": 1.3187556266784668,
+      "learning_rate": 7.829007424145379e-06,
+      "loss": 0.2529,
+      "step": 3864
+    },
+    {
+      "epoch": 0.3092,
+      "grad_norm": 1.7161333560943604,
+      "learning_rate": 7.827970355678774e-06,
+      "loss": 0.3463,
+      "step": 3865
+    },
+    {
+      "epoch": 0.30928,
+      "grad_norm": 1.6725596189498901,
+      "learning_rate": 7.826933108295947e-06,
+      "loss": 0.4048,
+      "step": 3866
+    },
+    {
+      "epoch": 0.30936,
+      "grad_norm": 1.5978217124938965,
+      "learning_rate": 7.82589568206252e-06,
+      "loss": 0.333,
+      "step": 3867
+    },
+    {
+      "epoch": 0.30944,
+      "grad_norm": 1.6118669509887695,
+      "learning_rate": 7.82485807704413e-06,
+      "loss": 0.4502,
+      "step": 3868
+    },
+    {
+      "epoch": 0.30952,
+      "grad_norm": 1.6804618835449219,
+      "learning_rate": 7.823820293306421e-06,
+      "loss": 0.2939,
+      "step": 3869
+    },
+    {
+      "epoch": 0.3096,
+      "grad_norm": 1.3810745477676392,
+      "learning_rate": 7.82278233091505e-06,
+      "loss": 0.2853,
+      "step": 3870
+    },
+    {
+      "epoch": 0.30968,
+      "grad_norm": 1.4669585227966309,
+      "learning_rate": 7.821744189935687e-06,
+      "loss": 0.3352,
+      "step": 3871
+    },
+    {
+      "epoch": 0.30976,
+      "grad_norm": 1.5521767139434814,
+      "learning_rate": 7.82070587043401e-06,
+      "loss": 0.3828,
+      "step": 3872
+    },
+    {
+      "epoch": 0.30984,
+      "grad_norm": 2.0385186672210693,
+      "learning_rate": 7.81966737247571e-06,
+      "loss": 0.4688,
+      "step": 3873
+    },
+    {
+      "epoch": 0.30992,
+      "grad_norm": 1.3120101690292358,
+      "learning_rate": 7.81862869612649e-06,
+      "loss": 0.2976,
+      "step": 3874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5607187747955322,
+      "learning_rate": 7.817589841452065e-06,
+      "loss": 0.2853,
+      "step": 3875
+    },
+    {
+      "epoch": 0.31008,
+      "grad_norm": 1.4162628650665283,
+      "learning_rate": 7.816550808518158e-06,
+      "loss": 0.3474,
+      "step": 3876
+    },
+    {
+      "epoch": 0.31016,
+      "grad_norm": 1.9220107793807983,
+      "learning_rate": 7.815511597390506e-06,
+      "loss": 0.3522,
+      "step": 3877
+    },
+    {
+      "epoch": 0.31024,
+      "grad_norm": 1.6877249479293823,
+      "learning_rate": 7.814472208134855e-06,
+      "loss": 0.331,
+      "step": 3878
+    },
+    {
+      "epoch": 0.31032,
+      "grad_norm": 1.6581977605819702,
+      "learning_rate": 7.813432640816965e-06,
+      "loss": 0.4292,
+      "step": 3879
+    },
+    {
+      "epoch": 0.3104,
+      "grad_norm": 1.860425591468811,
+      "learning_rate": 7.812392895502605e-06,
+      "loss": 0.4297,
+      "step": 3880
+    },
+    {
+      "epoch": 0.31048,
+      "grad_norm": 1.4564399719238281,
+      "learning_rate": 7.811352972257558e-06,
+      "loss": 0.3113,
+      "step": 3881
+    },
+    {
+      "epoch": 0.31056,
+      "grad_norm": 1.661933183670044,
+      "learning_rate": 7.810312871147614e-06,
+      "loss": 0.381,
+      "step": 3882
+    },
+    {
+      "epoch": 0.31064,
+      "grad_norm": 1.430387258529663,
+      "learning_rate": 7.809272592238579e-06,
+      "loss": 0.2592,
+      "step": 3883
+    },
+    {
+      "epoch": 0.31072,
+      "grad_norm": 1.3695735931396484,
+      "learning_rate": 7.808232135596267e-06,
+      "loss": 0.3324,
+      "step": 3884
+    },
+    {
+      "epoch": 0.3108,
+      "grad_norm": 1.8727529048919678,
+      "learning_rate": 7.807191501286505e-06,
+      "loss": 0.4287,
+      "step": 3885
+    },
+    {
+      "epoch": 0.31088,
+      "grad_norm": 1.6684083938598633,
+      "learning_rate": 7.806150689375126e-06,
+      "loss": 0.3635,
+      "step": 3886
+    },
+    {
+      "epoch": 0.31096,
+      "grad_norm": 1.6127012968063354,
+      "learning_rate": 7.805109699927985e-06,
+      "loss": 0.3655,
+      "step": 3887
+    },
+    {
+      "epoch": 0.31104,
+      "grad_norm": 1.7159547805786133,
+      "learning_rate": 7.804068533010938e-06,
+      "loss": 0.3361,
+      "step": 3888
+    },
+    {
+      "epoch": 0.31112,
+      "grad_norm": 1.4053112268447876,
+      "learning_rate": 7.803027188689859e-06,
+      "loss": 0.3538,
+      "step": 3889
+    },
+    {
+      "epoch": 0.3112,
+      "grad_norm": 1.524722695350647,
+      "learning_rate": 7.801985667030628e-06,
+      "loss": 0.4273,
+      "step": 3890
+    },
+    {
+      "epoch": 0.31128,
+      "grad_norm": 1.4059796333312988,
+      "learning_rate": 7.800943968099139e-06,
+      "loss": 0.3329,
+      "step": 3891
+    },
+    {
+      "epoch": 0.31136,
+      "grad_norm": 1.8556325435638428,
+      "learning_rate": 7.799902091961298e-06,
+      "loss": 0.4473,
+      "step": 3892
+    },
+    {
+      "epoch": 0.31144,
+      "grad_norm": 1.5466563701629639,
+      "learning_rate": 7.798860038683019e-06,
+      "loss": 0.3742,
+      "step": 3893
+    },
+    {
+      "epoch": 0.31152,
+      "grad_norm": 1.2450904846191406,
+      "learning_rate": 7.79781780833023e-06,
+      "loss": 0.2928,
+      "step": 3894
+    },
+    {
+      "epoch": 0.3116,
+      "grad_norm": 1.7872172594070435,
+      "learning_rate": 7.796775400968871e-06,
+      "loss": 0.415,
+      "step": 3895
+    },
+    {
+      "epoch": 0.31168,
+      "grad_norm": 1.6194071769714355,
+      "learning_rate": 7.79573281666489e-06,
+      "loss": 0.3123,
+      "step": 3896
+    },
+    {
+      "epoch": 0.31176,
+      "grad_norm": 1.4827814102172852,
+      "learning_rate": 7.794690055484249e-06,
+      "loss": 0.2888,
+      "step": 3897
+    },
+    {
+      "epoch": 0.31184,
+      "grad_norm": 1.2528879642486572,
+      "learning_rate": 7.793647117492919e-06,
+      "loss": 0.2737,
+      "step": 3898
+    },
+    {
+      "epoch": 0.31192,
+      "grad_norm": 1.405361294746399,
+      "learning_rate": 7.792604002756882e-06,
+      "loss": 0.3119,
+      "step": 3899
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 1.7106093168258667,
+      "learning_rate": 7.791560711342134e-06,
+      "loss": 0.3068,
+      "step": 3900
+    },
+    {
+      "epoch": 0.31208,
+      "grad_norm": 2.097712755203247,
+      "learning_rate": 7.790517243314682e-06,
+      "loss": 0.5649,
+      "step": 3901
+    },
+    {
+      "epoch": 0.31216,
+      "grad_norm": 1.1007894277572632,
+      "learning_rate": 7.78947359874054e-06,
+      "loss": 0.2437,
+      "step": 3902
+    },
+    {
+      "epoch": 0.31224,
+      "grad_norm": 1.5791645050048828,
+      "learning_rate": 7.788429777685737e-06,
+      "loss": 0.3363,
+      "step": 3903
+    },
+    {
+      "epoch": 0.31232,
+      "grad_norm": 1.4127541780471802,
+      "learning_rate": 7.787385780216313e-06,
+      "loss": 0.3199,
+      "step": 3904
+    },
+    {
+      "epoch": 0.3124,
+      "grad_norm": 1.4177106618881226,
+      "learning_rate": 7.786341606398318e-06,
+      "loss": 0.3293,
+      "step": 3905
+    },
+    {
+      "epoch": 0.31248,
+      "grad_norm": 1.3515725135803223,
+      "learning_rate": 7.785297256297811e-06,
+      "loss": 0.2774,
+      "step": 3906
+    },
+    {
+      "epoch": 0.31256,
+      "grad_norm": 1.4087157249450684,
+      "learning_rate": 7.784252729980866e-06,
+      "loss": 0.3314,
+      "step": 3907
+    },
+    {
+      "epoch": 0.31264,
+      "grad_norm": 1.7299154996871948,
+      "learning_rate": 7.783208027513569e-06,
+      "loss": 0.3618,
+      "step": 3908
+    },
+    {
+      "epoch": 0.31272,
+      "grad_norm": 1.4551969766616821,
+      "learning_rate": 7.78216314896201e-06,
+      "loss": 0.3251,
+      "step": 3909
+    },
+    {
+      "epoch": 0.3128,
+      "grad_norm": 1.9065160751342773,
+      "learning_rate": 7.7811180943923e-06,
+      "loss": 0.5049,
+      "step": 3910
+    },
+    {
+      "epoch": 0.31288,
+      "grad_norm": 1.8001066446304321,
+      "learning_rate": 7.780072863870551e-06,
+      "loss": 0.4066,
+      "step": 3911
+    },
+    {
+      "epoch": 0.31296,
+      "grad_norm": 1.3444501161575317,
+      "learning_rate": 7.779027457462896e-06,
+      "loss": 0.3117,
+      "step": 3912
+    },
+    {
+      "epoch": 0.31304,
+      "grad_norm": 1.6955708265304565,
+      "learning_rate": 7.777981875235472e-06,
+      "loss": 0.4009,
+      "step": 3913
+    },
+    {
+      "epoch": 0.31312,
+      "grad_norm": 2.015012264251709,
+      "learning_rate": 7.776936117254432e-06,
+      "loss": 0.4429,
+      "step": 3914
+    },
+    {
+      "epoch": 0.3132,
+      "grad_norm": 1.5816140174865723,
+      "learning_rate": 7.775890183585932e-06,
+      "loss": 0.3805,
+      "step": 3915
+    },
+    {
+      "epoch": 0.31328,
+      "grad_norm": 1.2359122037887573,
+      "learning_rate": 7.77484407429615e-06,
+      "loss": 0.2843,
+      "step": 3916
+    },
+    {
+      "epoch": 0.31336,
+      "grad_norm": 1.453830361366272,
+      "learning_rate": 7.77379778945127e-06,
+      "loss": 0.3051,
+      "step": 3917
+    },
+    {
+      "epoch": 0.31344,
+      "grad_norm": 1.4686434268951416,
+      "learning_rate": 7.772751329117481e-06,
+      "loss": 0.3015,
+      "step": 3918
+    },
+    {
+      "epoch": 0.31352,
+      "grad_norm": 1.454747200012207,
+      "learning_rate": 7.771704693360996e-06,
+      "loss": 0.3953,
+      "step": 3919
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 1.7359517812728882,
+      "learning_rate": 7.77065788224803e-06,
+      "loss": 0.2675,
+      "step": 3920
+    },
+    {
+      "epoch": 0.31368,
+      "grad_norm": 1.285766839981079,
+      "learning_rate": 7.769610895844808e-06,
+      "loss": 0.2614,
+      "step": 3921
+    },
+    {
+      "epoch": 0.31376,
+      "grad_norm": 1.5808966159820557,
+      "learning_rate": 7.768563734217572e-06,
+      "loss": 0.2836,
+      "step": 3922
+    },
+    {
+      "epoch": 0.31384,
+      "grad_norm": 1.6945921182632446,
+      "learning_rate": 7.767516397432574e-06,
+      "loss": 0.3945,
+      "step": 3923
+    },
+    {
+      "epoch": 0.31392,
+      "grad_norm": 1.5754956007003784,
+      "learning_rate": 7.766468885556072e-06,
+      "loss": 0.3035,
+      "step": 3924
+    },
+    {
+      "epoch": 0.314,
+      "grad_norm": 1.4004344940185547,
+      "learning_rate": 7.76542119865434e-06,
+      "loss": 0.3592,
+      "step": 3925
+    },
+    {
+      "epoch": 0.31408,
+      "grad_norm": 1.5656622648239136,
+      "learning_rate": 7.764373336793662e-06,
+      "loss": 0.3183,
+      "step": 3926
+    },
+    {
+      "epoch": 0.31416,
+      "grad_norm": 1.9068760871887207,
+      "learning_rate": 7.763325300040333e-06,
+      "loss": 0.3495,
+      "step": 3927
+    },
+    {
+      "epoch": 0.31424,
+      "grad_norm": 1.5390478372573853,
+      "learning_rate": 7.762277088460659e-06,
+      "loss": 0.3439,
+      "step": 3928
+    },
+    {
+      "epoch": 0.31432,
+      "grad_norm": 1.4874063730239868,
+      "learning_rate": 7.761228702120956e-06,
+      "loss": 0.3257,
+      "step": 3929
+    },
+    {
+      "epoch": 0.3144,
+      "grad_norm": 1.8362594842910767,
+      "learning_rate": 7.760180141087552e-06,
+      "loss": 0.5189,
+      "step": 3930
+    },
+    {
+      "epoch": 0.31448,
+      "grad_norm": 1.7211016416549683,
+      "learning_rate": 7.759131405426785e-06,
+      "loss": 0.3418,
+      "step": 3931
+    },
+    {
+      "epoch": 0.31456,
+      "grad_norm": 1.237994909286499,
+      "learning_rate": 7.758082495205007e-06,
+      "loss": 0.2734,
+      "step": 3932
+    },
+    {
+      "epoch": 0.31464,
+      "grad_norm": 1.7199347019195557,
+      "learning_rate": 7.757033410488577e-06,
+      "loss": 0.4034,
+      "step": 3933
+    },
+    {
+      "epoch": 0.31472,
+      "grad_norm": 1.5101882219314575,
+      "learning_rate": 7.75598415134387e-06,
+      "loss": 0.288,
+      "step": 3934
+    },
+    {
+      "epoch": 0.3148,
+      "grad_norm": 1.4008091688156128,
+      "learning_rate": 7.754934717837267e-06,
+      "loss": 0.3252,
+      "step": 3935
+    },
+    {
+      "epoch": 0.31488,
+      "grad_norm": 1.382588505744934,
+      "learning_rate": 7.753885110035161e-06,
+      "loss": 0.2844,
+      "step": 3936
+    },
+    {
+      "epoch": 0.31496,
+      "grad_norm": 1.5880348682403564,
+      "learning_rate": 7.75283532800396e-06,
+      "loss": 0.2899,
+      "step": 3937
+    },
+    {
+      "epoch": 0.31504,
+      "grad_norm": 1.820833444595337,
+      "learning_rate": 7.751785371810079e-06,
+      "loss": 0.4055,
+      "step": 3938
+    },
+    {
+      "epoch": 0.31512,
+      "grad_norm": 1.5325983762741089,
+      "learning_rate": 7.750735241519945e-06,
+      "loss": 0.3185,
+      "step": 3939
+    },
+    {
+      "epoch": 0.3152,
+      "grad_norm": 1.4334882497787476,
+      "learning_rate": 7.749684937199995e-06,
+      "loss": 0.3512,
+      "step": 3940
+    },
+    {
+      "epoch": 0.31528,
+      "grad_norm": 1.317023754119873,
+      "learning_rate": 7.74863445891668e-06,
+      "loss": 0.2794,
+      "step": 3941
+    },
+    {
+      "epoch": 0.31536,
+      "grad_norm": 1.646425724029541,
+      "learning_rate": 7.74758380673646e-06,
+      "loss": 0.4987,
+      "step": 3942
+    },
+    {
+      "epoch": 0.31544,
+      "grad_norm": 1.4610693454742432,
+      "learning_rate": 7.746532980725806e-06,
+      "loss": 0.3181,
+      "step": 3943
+    },
+    {
+      "epoch": 0.31552,
+      "grad_norm": 1.675247311592102,
+      "learning_rate": 7.7454819809512e-06,
+      "loss": 0.3994,
+      "step": 3944
+    },
+    {
+      "epoch": 0.3156,
+      "grad_norm": 1.453763723373413,
+      "learning_rate": 7.744430807479136e-06,
+      "loss": 0.3001,
+      "step": 3945
+    },
+    {
+      "epoch": 0.31568,
+      "grad_norm": 1.418670415878296,
+      "learning_rate": 7.743379460376117e-06,
+      "loss": 0.2949,
+      "step": 3946
+    },
+    {
+      "epoch": 0.31576,
+      "grad_norm": 1.5007295608520508,
+      "learning_rate": 7.74232793970866e-06,
+      "loss": 0.4157,
+      "step": 3947
+    },
+    {
+      "epoch": 0.31584,
+      "grad_norm": 1.3415838479995728,
+      "learning_rate": 7.741276245543287e-06,
+      "loss": 0.3165,
+      "step": 3948
+    },
+    {
+      "epoch": 0.31592,
+      "grad_norm": 1.6637834310531616,
+      "learning_rate": 7.740224377946543e-06,
+      "loss": 0.419,
+      "step": 3949
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 1.9902657270431519,
+      "learning_rate": 7.739172336984969e-06,
+      "loss": 0.3941,
+      "step": 3950
+    },
+    {
+      "epoch": 0.31608,
+      "grad_norm": 1.3266291618347168,
+      "learning_rate": 7.738120122725126e-06,
+      "loss": 0.3171,
+      "step": 3951
+    },
+    {
+      "epoch": 0.31616,
+      "grad_norm": 1.462837815284729,
+      "learning_rate": 7.737067735233586e-06,
+      "loss": 0.3602,
+      "step": 3952
+    },
+    {
+      "epoch": 0.31624,
+      "grad_norm": 1.3161622285842896,
+      "learning_rate": 7.736015174576927e-06,
+      "loss": 0.2974,
+      "step": 3953
+    },
+    {
+      "epoch": 0.31632,
+      "grad_norm": 1.4802780151367188,
+      "learning_rate": 7.734962440821742e-06,
+      "loss": 0.3562,
+      "step": 3954
+    },
+    {
+      "epoch": 0.3164,
+      "grad_norm": 1.5523854494094849,
+      "learning_rate": 7.733909534034638e-06,
+      "loss": 0.3115,
+      "step": 3955
+    },
+    {
+      "epoch": 0.31648,
+      "grad_norm": 1.815574288368225,
+      "learning_rate": 7.732856454282223e-06,
+      "loss": 0.3293,
+      "step": 3956
+    },
+    {
+      "epoch": 0.31656,
+      "grad_norm": 2.250521183013916,
+      "learning_rate": 7.731803201631125e-06,
+      "loss": 0.597,
+      "step": 3957
+    },
+    {
+      "epoch": 0.31664,
+      "grad_norm": 1.4423999786376953,
+      "learning_rate": 7.730749776147978e-06,
+      "loss": 0.3607,
+      "step": 3958
+    },
+    {
+      "epoch": 0.31672,
+      "grad_norm": 1.7350084781646729,
+      "learning_rate": 7.729696177899432e-06,
+      "loss": 0.4186,
+      "step": 3959
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 1.1580049991607666,
+      "learning_rate": 7.72864240695214e-06,
+      "loss": 0.2433,
+      "step": 3960
+    },
+    {
+      "epoch": 0.31688,
+      "grad_norm": 1.309934139251709,
+      "learning_rate": 7.727588463372775e-06,
+      "loss": 0.2844,
+      "step": 3961
+    },
+    {
+      "epoch": 0.31696,
+      "grad_norm": 1.4602081775665283,
+      "learning_rate": 7.726534347228013e-06,
+      "loss": 0.4136,
+      "step": 3962
+    },
+    {
+      "epoch": 0.31704,
+      "grad_norm": 1.3389019966125488,
+      "learning_rate": 7.725480058584547e-06,
+      "loss": 0.2797,
+      "step": 3963
+    },
+    {
+      "epoch": 0.31712,
+      "grad_norm": 1.8739509582519531,
+      "learning_rate": 7.724425597509077e-06,
+      "loss": 0.4236,
+      "step": 3964
+    },
+    {
+      "epoch": 0.3172,
+      "grad_norm": 0.9077721834182739,
+      "learning_rate": 7.723370964068315e-06,
+      "loss": 0.2111,
+      "step": 3965
+    },
+    {
+      "epoch": 0.31728,
+      "grad_norm": 1.4916582107543945,
+      "learning_rate": 7.722316158328983e-06,
+      "loss": 0.3819,
+      "step": 3966
+    },
+    {
+      "epoch": 0.31736,
+      "grad_norm": 1.4052239656448364,
+      "learning_rate": 7.72126118035782e-06,
+      "loss": 0.2928,
+      "step": 3967
+    },
+    {
+      "epoch": 0.31744,
+      "grad_norm": 1.968377947807312,
+      "learning_rate": 7.720206030221567e-06,
+      "loss": 0.3799,
+      "step": 3968
+    },
+    {
+      "epoch": 0.31752,
+      "grad_norm": 1.677701473236084,
+      "learning_rate": 7.719150707986977e-06,
+      "loss": 0.4268,
+      "step": 3969
+    },
+    {
+      "epoch": 0.3176,
+      "grad_norm": 1.2914515733718872,
+      "learning_rate": 7.718095213720822e-06,
+      "loss": 0.2356,
+      "step": 3970
+    },
+    {
+      "epoch": 0.31768,
+      "grad_norm": 1.7517569065093994,
+      "learning_rate": 7.717039547489877e-06,
+      "loss": 0.3818,
+      "step": 3971
+    },
+    {
+      "epoch": 0.31776,
+      "grad_norm": 1.708135962486267,
+      "learning_rate": 7.715983709360932e-06,
+      "loss": 0.3412,
+      "step": 3972
+    },
+    {
+      "epoch": 0.31784,
+      "grad_norm": 1.5727461576461792,
+      "learning_rate": 7.714927699400785e-06,
+      "loss": 0.3977,
+      "step": 3973
+    },
+    {
+      "epoch": 0.31792,
+      "grad_norm": 1.4590837955474854,
+      "learning_rate": 7.713871517676247e-06,
+      "loss": 0.3593,
+      "step": 3974
+    },
+    {
+      "epoch": 0.318,
+      "grad_norm": 1.291216492652893,
+      "learning_rate": 7.712815164254138e-06,
+      "loss": 0.2984,
+      "step": 3975
+    },
+    {
+      "epoch": 0.31808,
+      "grad_norm": 2.102529525756836,
+      "learning_rate": 7.71175863920129e-06,
+      "loss": 0.392,
+      "step": 3976
+    },
+    {
+      "epoch": 0.31816,
+      "grad_norm": 2.003671169281006,
+      "learning_rate": 7.710701942584549e-06,
+      "loss": 0.3816,
+      "step": 3977
+    },
+    {
+      "epoch": 0.31824,
+      "grad_norm": 1.6712678670883179,
+      "learning_rate": 7.709645074470764e-06,
+      "loss": 0.3481,
+      "step": 3978
+    },
+    {
+      "epoch": 0.31832,
+      "grad_norm": 1.2910213470458984,
+      "learning_rate": 7.708588034926802e-06,
+      "loss": 0.2783,
+      "step": 3979
+    },
+    {
+      "epoch": 0.3184,
+      "grad_norm": 2.094609260559082,
+      "learning_rate": 7.70753082401954e-06,
+      "loss": 0.6013,
+      "step": 3980
+    },
+    {
+      "epoch": 0.31848,
+      "grad_norm": 1.9006894826889038,
+      "learning_rate": 7.70647344181586e-06,
+      "loss": 0.3544,
+      "step": 3981
+    },
+    {
+      "epoch": 0.31856,
+      "grad_norm": 1.320239543914795,
+      "learning_rate": 7.705415888382662e-06,
+      "loss": 0.3105,
+      "step": 3982
+    },
+    {
+      "epoch": 0.31864,
+      "grad_norm": 1.9996472597122192,
+      "learning_rate": 7.704358163786852e-06,
+      "loss": 0.4681,
+      "step": 3983
+    },
+    {
+      "epoch": 0.31872,
+      "grad_norm": 1.3153249025344849,
+      "learning_rate": 7.70330026809535e-06,
+      "loss": 0.2759,
+      "step": 3984
+    },
+    {
+      "epoch": 0.3188,
+      "grad_norm": 1.4559813737869263,
+      "learning_rate": 7.702242201375087e-06,
+      "loss": 0.3541,
+      "step": 3985
+    },
+    {
+      "epoch": 0.31888,
+      "grad_norm": 1.5201846361160278,
+      "learning_rate": 7.701183963693e-06,
+      "loss": 0.323,
+      "step": 3986
+    },
+    {
+      "epoch": 0.31896,
+      "grad_norm": 1.1571747064590454,
+      "learning_rate": 7.700125555116043e-06,
+      "loss": 0.2244,
+      "step": 3987
+    },
+    {
+      "epoch": 0.31904,
+      "grad_norm": 1.2063324451446533,
+      "learning_rate": 7.699066975711176e-06,
+      "loss": 0.3063,
+      "step": 3988
+    },
+    {
+      "epoch": 0.31912,
+      "grad_norm": 1.5614107847213745,
+      "learning_rate": 7.698008225545372e-06,
+      "loss": 0.3284,
+      "step": 3989
+    },
+    {
+      "epoch": 0.3192,
+      "grad_norm": 1.713163137435913,
+      "learning_rate": 7.696949304685616e-06,
+      "loss": 0.384,
+      "step": 3990
+    },
+    {
+      "epoch": 0.31928,
+      "grad_norm": 1.4818403720855713,
+      "learning_rate": 7.695890213198902e-06,
+      "loss": 0.3769,
+      "step": 3991
+    },
+    {
+      "epoch": 0.31936,
+      "grad_norm": 1.8636175394058228,
+      "learning_rate": 7.694830951152233e-06,
+      "loss": 0.3745,
+      "step": 3992
+    },
+    {
+      "epoch": 0.31944,
+      "grad_norm": 1.5398733615875244,
+      "learning_rate": 7.693771518612627e-06,
+      "loss": 0.3747,
+      "step": 3993
+    },
+    {
+      "epoch": 0.31952,
+      "grad_norm": 1.2448610067367554,
+      "learning_rate": 7.69271191564711e-06,
+      "loss": 0.2528,
+      "step": 3994
+    },
+    {
+      "epoch": 0.3196,
+      "grad_norm": 1.3927911520004272,
+      "learning_rate": 7.691652142322723e-06,
+      "loss": 0.3104,
+      "step": 3995
+    },
+    {
+      "epoch": 0.31968,
+      "grad_norm": 1.3788466453552246,
+      "learning_rate": 7.690592198706508e-06,
+      "loss": 0.3017,
+      "step": 3996
+    },
+    {
+      "epoch": 0.31976,
+      "grad_norm": 1.8893440961837769,
+      "learning_rate": 7.68953208486553e-06,
+      "loss": 0.4398,
+      "step": 3997
+    },
+    {
+      "epoch": 0.31984,
+      "grad_norm": 1.4432144165039062,
+      "learning_rate": 7.688471800866855e-06,
+      "loss": 0.3526,
+      "step": 3998
+    },
+    {
+      "epoch": 0.31992,
+      "grad_norm": 1.7225422859191895,
+      "learning_rate": 7.687411346777564e-06,
+      "loss": 0.3249,
+      "step": 3999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5686019659042358,
+      "learning_rate": 7.686350722664751e-06,
+      "loss": 0.2928,
+      "step": 4000
+    },
+    {
+      "epoch": 0.32008,
+      "grad_norm": 1.5652544498443604,
+      "learning_rate": 7.685289928595514e-06,
+      "loss": 0.3355,
+      "step": 4001
+    },
+    {
+      "epoch": 0.32016,
+      "grad_norm": 1.6727375984191895,
+      "learning_rate": 7.68422896463697e-06,
+      "loss": 0.3482,
+      "step": 4002
+    },
+    {
+      "epoch": 0.32024,
+      "grad_norm": 1.4307200908660889,
+      "learning_rate": 7.68316783085624e-06,
+      "loss": 0.34,
+      "step": 4003
+    },
+    {
+      "epoch": 0.32032,
+      "grad_norm": 1.5332558155059814,
+      "learning_rate": 7.682106527320458e-06,
+      "loss": 0.3796,
+      "step": 4004
+    },
+    {
+      "epoch": 0.3204,
+      "grad_norm": 1.9234296083450317,
+      "learning_rate": 7.681045054096773e-06,
+      "loss": 0.3732,
+      "step": 4005
+    },
+    {
+      "epoch": 0.32048,
+      "grad_norm": 1.8026024103164673,
+      "learning_rate": 7.679983411252336e-06,
+      "loss": 0.3543,
+      "step": 4006
+    },
+    {
+      "epoch": 0.32056,
+      "grad_norm": 1.5684106349945068,
+      "learning_rate": 7.678921598854316e-06,
+      "loss": 0.3274,
+      "step": 4007
+    },
+    {
+      "epoch": 0.32064,
+      "grad_norm": 1.4080421924591064,
+      "learning_rate": 7.677859616969892e-06,
+      "loss": 0.346,
+      "step": 4008
+    },
+    {
+      "epoch": 0.32072,
+      "grad_norm": 1.667383074760437,
+      "learning_rate": 7.67679746566625e-06,
+      "loss": 0.3846,
+      "step": 4009
+    },
+    {
+      "epoch": 0.3208,
+      "grad_norm": 1.5758278369903564,
+      "learning_rate": 7.67573514501059e-06,
+      "loss": 0.396,
+      "step": 4010
+    },
+    {
+      "epoch": 0.32088,
+      "grad_norm": 1.7792264223098755,
+      "learning_rate": 7.674672655070117e-06,
+      "loss": 0.441,
+      "step": 4011
+    },
+    {
+      "epoch": 0.32096,
+      "grad_norm": 1.3458983898162842,
+      "learning_rate": 7.673609995912056e-06,
+      "loss": 0.2971,
+      "step": 4012
+    },
+    {
+      "epoch": 0.32104,
+      "grad_norm": 1.343664526939392,
+      "learning_rate": 7.672547167603638e-06,
+      "loss": 0.3028,
+      "step": 4013
+    },
+    {
+      "epoch": 0.32112,
+      "grad_norm": 1.9842946529388428,
+      "learning_rate": 7.6714841702121e-06,
+      "loss": 0.3945,
+      "step": 4014
+    },
+    {
+      "epoch": 0.3212,
+      "grad_norm": 1.2746963500976562,
+      "learning_rate": 7.6704210038047e-06,
+      "loss": 0.2742,
+      "step": 4015
+    },
+    {
+      "epoch": 0.32128,
+      "grad_norm": 1.8831450939178467,
+      "learning_rate": 7.669357668448695e-06,
+      "loss": 0.3977,
+      "step": 4016
+    },
+    {
+      "epoch": 0.32136,
+      "grad_norm": 1.4678828716278076,
+      "learning_rate": 7.668294164211365e-06,
+      "loss": 0.3188,
+      "step": 4017
+    },
+    {
+      "epoch": 0.32144,
+      "grad_norm": 2.1258130073547363,
+      "learning_rate": 7.667230491159992e-06,
+      "loss": 0.3754,
+      "step": 4018
+    },
+    {
+      "epoch": 0.32152,
+      "grad_norm": 2.3288698196411133,
+      "learning_rate": 7.666166649361868e-06,
+      "loss": 0.6207,
+      "step": 4019
+    },
+    {
+      "epoch": 0.3216,
+      "grad_norm": 1.5426217317581177,
+      "learning_rate": 7.665102638884303e-06,
+      "loss": 0.331,
+      "step": 4020
+    },
+    {
+      "epoch": 0.32168,
+      "grad_norm": 1.523314356803894,
+      "learning_rate": 7.664038459794612e-06,
+      "loss": 0.3148,
+      "step": 4021
+    },
+    {
+      "epoch": 0.32176,
+      "grad_norm": 1.3950161933898926,
+      "learning_rate": 7.66297411216012e-06,
+      "loss": 0.2575,
+      "step": 4022
+    },
+    {
+      "epoch": 0.32184,
+      "grad_norm": 1.7832293510437012,
+      "learning_rate": 7.661909596048167e-06,
+      "loss": 0.3661,
+      "step": 4023
+    },
+    {
+      "epoch": 0.32192,
+      "grad_norm": 1.4251043796539307,
+      "learning_rate": 7.6608449115261e-06,
+      "loss": 0.3476,
+      "step": 4024
+    },
+    {
+      "epoch": 0.322,
+      "grad_norm": 1.7455947399139404,
+      "learning_rate": 7.659780058661278e-06,
+      "loss": 0.4794,
+      "step": 4025
+    },
+    {
+      "epoch": 0.32208,
+      "grad_norm": 1.5899649858474731,
+      "learning_rate": 7.658715037521074e-06,
+      "loss": 0.3369,
+      "step": 4026
+    },
+    {
+      "epoch": 0.32216,
+      "grad_norm": 1.148187279701233,
+      "learning_rate": 7.657649848172864e-06,
+      "loss": 0.2667,
+      "step": 4027
+    },
+    {
+      "epoch": 0.32224,
+      "grad_norm": 1.3705857992172241,
+      "learning_rate": 7.65658449068404e-06,
+      "loss": 0.2696,
+      "step": 4028
+    },
+    {
+      "epoch": 0.32232,
+      "grad_norm": 1.8258371353149414,
+      "learning_rate": 7.655518965122004e-06,
+      "loss": 0.3487,
+      "step": 4029
+    },
+    {
+      "epoch": 0.3224,
+      "grad_norm": 1.5610127449035645,
+      "learning_rate": 7.65445327155417e-06,
+      "loss": 0.353,
+      "step": 4030
+    },
+    {
+      "epoch": 0.32248,
+      "grad_norm": 1.566694974899292,
+      "learning_rate": 7.653387410047958e-06,
+      "loss": 0.3971,
+      "step": 4031
+    },
+    {
+      "epoch": 0.32256,
+      "grad_norm": 1.6996960639953613,
+      "learning_rate": 7.652321380670805e-06,
+      "loss": 0.3876,
+      "step": 4032
+    },
+    {
+      "epoch": 0.32264,
+      "grad_norm": 1.460684061050415,
+      "learning_rate": 7.65125518349015e-06,
+      "loss": 0.3302,
+      "step": 4033
+    },
+    {
+      "epoch": 0.32272,
+      "grad_norm": 1.5468209981918335,
+      "learning_rate": 7.650188818573452e-06,
+      "loss": 0.3631,
+      "step": 4034
+    },
+    {
+      "epoch": 0.3228,
+      "grad_norm": 1.5066543817520142,
+      "learning_rate": 7.649122285988175e-06,
+      "loss": 0.3756,
+      "step": 4035
+    },
+    {
+      "epoch": 0.32288,
+      "grad_norm": 1.4689096212387085,
+      "learning_rate": 7.648055585801795e-06,
+      "loss": 0.3488,
+      "step": 4036
+    },
+    {
+      "epoch": 0.32296,
+      "grad_norm": 1.2508280277252197,
+      "learning_rate": 7.646988718081799e-06,
+      "loss": 0.2692,
+      "step": 4037
+    },
+    {
+      "epoch": 0.32304,
+      "grad_norm": 1.4631690979003906,
+      "learning_rate": 7.645921682895684e-06,
+      "loss": 0.3593,
+      "step": 4038
+    },
+    {
+      "epoch": 0.32312,
+      "grad_norm": 1.4271583557128906,
+      "learning_rate": 7.644854480310955e-06,
+      "loss": 0.3081,
+      "step": 4039
+    },
+    {
+      "epoch": 0.3232,
+      "grad_norm": 1.613905906677246,
+      "learning_rate": 7.643787110395135e-06,
+      "loss": 0.4458,
+      "step": 4040
+    },
+    {
+      "epoch": 0.32328,
+      "grad_norm": 1.3938994407653809,
+      "learning_rate": 7.642719573215748e-06,
+      "loss": 0.3038,
+      "step": 4041
+    },
+    {
+      "epoch": 0.32336,
+      "grad_norm": 1.4477261304855347,
+      "learning_rate": 7.641651868840338e-06,
+      "loss": 0.3112,
+      "step": 4042
+    },
+    {
+      "epoch": 0.32344,
+      "grad_norm": 1.692185401916504,
+      "learning_rate": 7.640583997336455e-06,
+      "loss": 0.4446,
+      "step": 4043
+    },
+    {
+      "epoch": 0.32352,
+      "grad_norm": 1.4064701795578003,
+      "learning_rate": 7.639515958771656e-06,
+      "loss": 0.2536,
+      "step": 4044
+    },
+    {
+      "epoch": 0.3236,
+      "grad_norm": 1.7881546020507812,
+      "learning_rate": 7.638447753213516e-06,
+      "loss": 0.4048,
+      "step": 4045
+    },
+    {
+      "epoch": 0.32368,
+      "grad_norm": 1.3405356407165527,
+      "learning_rate": 7.637379380729612e-06,
+      "loss": 0.2965,
+      "step": 4046
+    },
+    {
+      "epoch": 0.32376,
+      "grad_norm": 1.182118535041809,
+      "learning_rate": 7.636310841387541e-06,
+      "loss": 0.2477,
+      "step": 4047
+    },
+    {
+      "epoch": 0.32384,
+      "grad_norm": 1.4148246049880981,
+      "learning_rate": 7.635242135254903e-06,
+      "loss": 0.2628,
+      "step": 4048
+    },
+    {
+      "epoch": 0.32392,
+      "grad_norm": 1.527912974357605,
+      "learning_rate": 7.634173262399315e-06,
+      "loss": 0.3746,
+      "step": 4049
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 1.4282273054122925,
+      "learning_rate": 7.633104222888396e-06,
+      "loss": 0.2963,
+      "step": 4050
+    },
+    {
+      "epoch": 0.32408,
+      "grad_norm": 1.5053621530532837,
+      "learning_rate": 7.632035016789785e-06,
+      "loss": 0.3858,
+      "step": 4051
+    },
+    {
+      "epoch": 0.32416,
+      "grad_norm": 1.633727788925171,
+      "learning_rate": 7.630965644171124e-06,
+      "loss": 0.4096,
+      "step": 4052
+    },
+    {
+      "epoch": 0.32424,
+      "grad_norm": 1.6081035137176514,
+      "learning_rate": 7.62989610510007e-06,
+      "loss": 0.296,
+      "step": 4053
+    },
+    {
+      "epoch": 0.32432,
+      "grad_norm": 1.2334622144699097,
+      "learning_rate": 7.628826399644292e-06,
+      "loss": 0.2918,
+      "step": 4054
+    },
+    {
+      "epoch": 0.3244,
+      "grad_norm": 1.6431307792663574,
+      "learning_rate": 7.627756527871459e-06,
+      "loss": 0.3395,
+      "step": 4055
+    },
+    {
+      "epoch": 0.32448,
+      "grad_norm": 1.6120516061782837,
+      "learning_rate": 7.626686489849266e-06,
+      "loss": 0.3045,
+      "step": 4056
+    },
+    {
+      "epoch": 0.32456,
+      "grad_norm": 1.4079047441482544,
+      "learning_rate": 7.625616285645408e-06,
+      "loss": 0.316,
+      "step": 4057
+    },
+    {
+      "epoch": 0.32464,
+      "grad_norm": 1.4125021696090698,
+      "learning_rate": 7.624545915327593e-06,
+      "loss": 0.3104,
+      "step": 4058
+    },
+    {
+      "epoch": 0.32472,
+      "grad_norm": 1.4613399505615234,
+      "learning_rate": 7.6234753789635374e-06,
+      "loss": 0.4579,
+      "step": 4059
+    },
+    {
+      "epoch": 0.3248,
+      "grad_norm": 1.8771039247512817,
+      "learning_rate": 7.622404676620974e-06,
+      "loss": 0.3648,
+      "step": 4060
+    },
+    {
+      "epoch": 0.32488,
+      "grad_norm": 1.4824306964874268,
+      "learning_rate": 7.621333808367643e-06,
+      "loss": 0.2957,
+      "step": 4061
+    },
+    {
+      "epoch": 0.32496,
+      "grad_norm": 1.4490859508514404,
+      "learning_rate": 7.62026277427129e-06,
+      "loss": 0.2975,
+      "step": 4062
+    },
+    {
+      "epoch": 0.32504,
+      "grad_norm": 1.4416332244873047,
+      "learning_rate": 7.61919157439968e-06,
+      "loss": 0.4024,
+      "step": 4063
+    },
+    {
+      "epoch": 0.32512,
+      "grad_norm": 1.0941718816757202,
+      "learning_rate": 7.618120208820583e-06,
+      "loss": 0.2395,
+      "step": 4064
+    },
+    {
+      "epoch": 0.3252,
+      "grad_norm": 1.8790380954742432,
+      "learning_rate": 7.617048677601781e-06,
+      "loss": 0.4126,
+      "step": 4065
+    },
+    {
+      "epoch": 0.32528,
+      "grad_norm": 1.4821059703826904,
+      "learning_rate": 7.615976980811067e-06,
+      "loss": 0.3245,
+      "step": 4066
+    },
+    {
+      "epoch": 0.32536,
+      "grad_norm": 1.9037108421325684,
+      "learning_rate": 7.61490511851624e-06,
+      "loss": 0.4843,
+      "step": 4067
+    },
+    {
+      "epoch": 0.32544,
+      "grad_norm": 1.4991943836212158,
+      "learning_rate": 7.613833090785117e-06,
+      "loss": 0.3384,
+      "step": 4068
+    },
+    {
+      "epoch": 0.32552,
+      "grad_norm": 1.4513674974441528,
+      "learning_rate": 7.612760897685519e-06,
+      "loss": 0.3904,
+      "step": 4069
+    },
+    {
+      "epoch": 0.3256,
+      "grad_norm": 1.277267575263977,
+      "learning_rate": 7.611688539285283e-06,
+      "loss": 0.2926,
+      "step": 4070
+    },
+    {
+      "epoch": 0.32568,
+      "grad_norm": 1.3457417488098145,
+      "learning_rate": 7.61061601565225e-06,
+      "loss": 0.2733,
+      "step": 4071
+    },
+    {
+      "epoch": 0.32576,
+      "grad_norm": 1.9008210897445679,
+      "learning_rate": 7.609543326854278e-06,
+      "loss": 0.4293,
+      "step": 4072
+    },
+    {
+      "epoch": 0.32584,
+      "grad_norm": 1.4903104305267334,
+      "learning_rate": 7.608470472959233e-06,
+      "loss": 0.3002,
+      "step": 4073
+    },
+    {
+      "epoch": 0.32592,
+      "grad_norm": 1.4304206371307373,
+      "learning_rate": 7.607397454034986e-06,
+      "loss": 0.3073,
+      "step": 4074
+    },
+    {
+      "epoch": 0.326,
+      "grad_norm": 1.7637519836425781,
+      "learning_rate": 7.606324270149428e-06,
+      "loss": 0.3062,
+      "step": 4075
+    },
+    {
+      "epoch": 0.32608,
+      "grad_norm": 1.7573155164718628,
+      "learning_rate": 7.605250921370454e-06,
+      "loss": 0.3944,
+      "step": 4076
+    },
+    {
+      "epoch": 0.32616,
+      "grad_norm": 1.3438637256622314,
+      "learning_rate": 7.604177407765972e-06,
+      "loss": 0.2591,
+      "step": 4077
+    },
+    {
+      "epoch": 0.32624,
+      "grad_norm": 1.9403959512710571,
+      "learning_rate": 7.603103729403898e-06,
+      "loss": 0.5459,
+      "step": 4078
+    },
+    {
+      "epoch": 0.32632,
+      "grad_norm": 1.421299934387207,
+      "learning_rate": 7.602029886352163e-06,
+      "loss": 0.3475,
+      "step": 4079
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 1.635377287864685,
+      "learning_rate": 7.600955878678702e-06,
+      "loss": 0.3548,
+      "step": 4080
+    },
+    {
+      "epoch": 0.32648,
+      "grad_norm": 1.75666344165802,
+      "learning_rate": 7.599881706451464e-06,
+      "loss": 0.5445,
+      "step": 4081
+    },
+    {
+      "epoch": 0.32656,
+      "grad_norm": 1.7399910688400269,
+      "learning_rate": 7.598807369738411e-06,
+      "loss": 0.3803,
+      "step": 4082
+    },
+    {
+      "epoch": 0.32664,
+      "grad_norm": 1.6762953996658325,
+      "learning_rate": 7.597732868607511e-06,
+      "loss": 0.3633,
+      "step": 4083
+    },
+    {
+      "epoch": 0.32672,
+      "grad_norm": 1.6950074434280396,
+      "learning_rate": 7.5966582031267455e-06,
+      "loss": 0.4613,
+      "step": 4084
+    },
+    {
+      "epoch": 0.3268,
+      "grad_norm": 1.7901536226272583,
+      "learning_rate": 7.595583373364104e-06,
+      "loss": 0.3687,
+      "step": 4085
+    },
+    {
+      "epoch": 0.32688,
+      "grad_norm": 1.0144829750061035,
+      "learning_rate": 7.594508379387586e-06,
+      "loss": 0.2249,
+      "step": 4086
+    },
+    {
+      "epoch": 0.32696,
+      "grad_norm": 1.360308051109314,
+      "learning_rate": 7.593433221265205e-06,
+      "loss": 0.3309,
+      "step": 4087
+    },
+    {
+      "epoch": 0.32704,
+      "grad_norm": 1.6082754135131836,
+      "learning_rate": 7.59235789906498e-06,
+      "loss": 0.3511,
+      "step": 4088
+    },
+    {
+      "epoch": 0.32712,
+      "grad_norm": 1.4917949438095093,
+      "learning_rate": 7.591282412854947e-06,
+      "loss": 0.3855,
+      "step": 4089
+    },
+    {
+      "epoch": 0.3272,
+      "grad_norm": 1.895923376083374,
+      "learning_rate": 7.590206762703145e-06,
+      "loss": 0.3869,
+      "step": 4090
+    },
+    {
+      "epoch": 0.32728,
+      "grad_norm": 1.37894606590271,
+      "learning_rate": 7.58913094867763e-06,
+      "loss": 0.3502,
+      "step": 4091
+    },
+    {
+      "epoch": 0.32736,
+      "grad_norm": 1.5004284381866455,
+      "learning_rate": 7.588054970846461e-06,
+      "loss": 0.3698,
+      "step": 4092
+    },
+    {
+      "epoch": 0.32744,
+      "grad_norm": 1.6978641748428345,
+      "learning_rate": 7.586978829277714e-06,
+      "loss": 0.3473,
+      "step": 4093
+    },
+    {
+      "epoch": 0.32752,
+      "grad_norm": 1.7193430662155151,
+      "learning_rate": 7.5859025240394725e-06,
+      "loss": 0.3368,
+      "step": 4094
+    },
+    {
+      "epoch": 0.3276,
+      "grad_norm": 1.4797018766403198,
+      "learning_rate": 7.584826055199831e-06,
+      "loss": 0.2682,
+      "step": 4095
+    },
+    {
+      "epoch": 0.32768,
+      "grad_norm": 1.4176084995269775,
+      "learning_rate": 7.583749422826894e-06,
+      "loss": 0.2922,
+      "step": 4096
+    },
+    {
+      "epoch": 0.32776,
+      "grad_norm": 1.4341297149658203,
+      "learning_rate": 7.582672626988776e-06,
+      "loss": 0.3214,
+      "step": 4097
+    },
+    {
+      "epoch": 0.32784,
+      "grad_norm": 1.48208749294281,
+      "learning_rate": 7.581595667753603e-06,
+      "loss": 0.3318,
+      "step": 4098
+    },
+    {
+      "epoch": 0.32792,
+      "grad_norm": 1.4464733600616455,
+      "learning_rate": 7.580518545189509e-06,
+      "loss": 0.358,
+      "step": 4099
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 1.890705943107605,
+      "learning_rate": 7.579441259364643e-06,
+      "loss": 0.5347,
+      "step": 4100
+    },
+    {
+      "epoch": 0.32808,
+      "grad_norm": 1.4273359775543213,
+      "learning_rate": 7.578363810347158e-06,
+      "loss": 0.3098,
+      "step": 4101
+    },
+    {
+      "epoch": 0.32816,
+      "grad_norm": 1.5816519260406494,
+      "learning_rate": 7.577286198205223e-06,
+      "loss": 0.3446,
+      "step": 4102
+    },
+    {
+      "epoch": 0.32824,
+      "grad_norm": 1.684738278388977,
+      "learning_rate": 7.576208423007012e-06,
+      "loss": 0.4043,
+      "step": 4103
+    },
+    {
+      "epoch": 0.32832,
+      "grad_norm": 1.6417845487594604,
+      "learning_rate": 7.575130484820715e-06,
+      "loss": 0.3435,
+      "step": 4104
+    },
+    {
+      "epoch": 0.3284,
+      "grad_norm": 1.5224345922470093,
+      "learning_rate": 7.57405238371453e-06,
+      "loss": 0.4164,
+      "step": 4105
+    },
+    {
+      "epoch": 0.32848,
+      "grad_norm": 1.3996219635009766,
+      "learning_rate": 7.5729741197566604e-06,
+      "loss": 0.3818,
+      "step": 4106
+    },
+    {
+      "epoch": 0.32856,
+      "grad_norm": 1.8475956916809082,
+      "learning_rate": 7.571895693015329e-06,
+      "loss": 0.5135,
+      "step": 4107
+    },
+    {
+      "epoch": 0.32864,
+      "grad_norm": 2.0243968963623047,
+      "learning_rate": 7.570817103558764e-06,
+      "loss": 0.3888,
+      "step": 4108
+    },
+    {
+      "epoch": 0.32872,
+      "grad_norm": 2.109097480773926,
+      "learning_rate": 7.569738351455203e-06,
+      "loss": 0.3905,
+      "step": 4109
+    },
+    {
+      "epoch": 0.3288,
+      "grad_norm": 1.4075511693954468,
+      "learning_rate": 7.568659436772892e-06,
+      "loss": 0.2856,
+      "step": 4110
+    },
+    {
+      "epoch": 0.32888,
+      "grad_norm": 1.384484052658081,
+      "learning_rate": 7.567580359580095e-06,
+      "loss": 0.3404,
+      "step": 4111
+    },
+    {
+      "epoch": 0.32896,
+      "grad_norm": 1.7616000175476074,
+      "learning_rate": 7.56650111994508e-06,
+      "loss": 0.3693,
+      "step": 4112
+    },
+    {
+      "epoch": 0.32904,
+      "grad_norm": 1.8043757677078247,
+      "learning_rate": 7.565421717936127e-06,
+      "loss": 0.3948,
+      "step": 4113
+    },
+    {
+      "epoch": 0.32912,
+      "grad_norm": 1.4485535621643066,
+      "learning_rate": 7.564342153621525e-06,
+      "loss": 0.3289,
+      "step": 4114
+    },
+    {
+      "epoch": 0.3292,
+      "grad_norm": 1.7356315851211548,
+      "learning_rate": 7.5632624270695755e-06,
+      "loss": 0.3701,
+      "step": 4115
+    },
+    {
+      "epoch": 0.32928,
+      "grad_norm": 1.4794209003448486,
+      "learning_rate": 7.562182538348589e-06,
+      "loss": 0.3695,
+      "step": 4116
+    },
+    {
+      "epoch": 0.32936,
+      "grad_norm": 1.806711196899414,
+      "learning_rate": 7.5611024875268856e-06,
+      "loss": 0.3913,
+      "step": 4117
+    },
+    {
+      "epoch": 0.32944,
+      "grad_norm": 1.9311702251434326,
+      "learning_rate": 7.560022274672798e-06,
+      "loss": 0.3523,
+      "step": 4118
+    },
+    {
+      "epoch": 0.32952,
+      "grad_norm": 1.8792706727981567,
+      "learning_rate": 7.5589418998546675e-06,
+      "loss": 0.477,
+      "step": 4119
+    },
+    {
+      "epoch": 0.3296,
+      "grad_norm": 1.8583472967147827,
+      "learning_rate": 7.5578613631408444e-06,
+      "loss": 0.4641,
+      "step": 4120
+    },
+    {
+      "epoch": 0.32968,
+      "grad_norm": 1.8453302383422852,
+      "learning_rate": 7.556780664599692e-06,
+      "loss": 0.365,
+      "step": 4121
+    },
+    {
+      "epoch": 0.32976,
+      "grad_norm": 1.3581293821334839,
+      "learning_rate": 7.555699804299581e-06,
+      "loss": 0.3245,
+      "step": 4122
+    },
+    {
+      "epoch": 0.32984,
+      "grad_norm": 1.490319848060608,
+      "learning_rate": 7.554618782308897e-06,
+      "loss": 0.2991,
+      "step": 4123
+    },
+    {
+      "epoch": 0.32992,
+      "grad_norm": 1.4145665168762207,
+      "learning_rate": 7.553537598696028e-06,
+      "loss": 0.2676,
+      "step": 4124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0299646854400635,
+      "learning_rate": 7.552456253529381e-06,
+      "loss": 0.3604,
+      "step": 4125
+    },
+    {
+      "epoch": 0.33008,
+      "grad_norm": 1.8027198314666748,
+      "learning_rate": 7.551374746877367e-06,
+      "loss": 0.5369,
+      "step": 4126
+    },
+    {
+      "epoch": 0.33016,
+      "grad_norm": 1.3814693689346313,
+      "learning_rate": 7.550293078808408e-06,
+      "loss": 0.3013,
+      "step": 4127
+    },
+    {
+      "epoch": 0.33024,
+      "grad_norm": 1.5172748565673828,
+      "learning_rate": 7.549211249390942e-06,
+      "loss": 0.3157,
+      "step": 4128
+    },
+    {
+      "epoch": 0.33032,
+      "grad_norm": 1.5251926183700562,
+      "learning_rate": 7.548129258693407e-06,
+      "loss": 0.3171,
+      "step": 4129
+    },
+    {
+      "epoch": 0.3304,
+      "grad_norm": 1.377307415008545,
+      "learning_rate": 7.547047106784262e-06,
+      "loss": 0.3191,
+      "step": 4130
+    },
+    {
+      "epoch": 0.33048,
+      "grad_norm": 1.4914178848266602,
+      "learning_rate": 7.545964793731968e-06,
+      "loss": 0.3543,
+      "step": 4131
+    },
+    {
+      "epoch": 0.33056,
+      "grad_norm": 1.3784854412078857,
+      "learning_rate": 7.544882319605e-06,
+      "loss": 0.2985,
+      "step": 4132
+    },
+    {
+      "epoch": 0.33064,
+      "grad_norm": 1.4411064386367798,
+      "learning_rate": 7.543799684471845e-06,
+      "loss": 0.3399,
+      "step": 4133
+    },
+    {
+      "epoch": 0.33072,
+      "grad_norm": 0.9017268419265747,
+      "learning_rate": 7.542716888400994e-06,
+      "loss": 0.2339,
+      "step": 4134
+    },
+    {
+      "epoch": 0.3308,
+      "grad_norm": 1.7791081666946411,
+      "learning_rate": 7.541633931460953e-06,
+      "loss": 0.4546,
+      "step": 4135
+    },
+    {
+      "epoch": 0.33088,
+      "grad_norm": 1.6857120990753174,
+      "learning_rate": 7.540550813720238e-06,
+      "loss": 0.3708,
+      "step": 4136
+    },
+    {
+      "epoch": 0.33096,
+      "grad_norm": 1.3614606857299805,
+      "learning_rate": 7.5394675352473735e-06,
+      "loss": 0.2817,
+      "step": 4137
+    },
+    {
+      "epoch": 0.33104,
+      "grad_norm": 1.4521223306655884,
+      "learning_rate": 7.538384096110896e-06,
+      "loss": 0.3394,
+      "step": 4138
+    },
+    {
+      "epoch": 0.33112,
+      "grad_norm": 1.4563080072402954,
+      "learning_rate": 7.5373004963793496e-06,
+      "loss": 0.2831,
+      "step": 4139
+    },
+    {
+      "epoch": 0.3312,
+      "grad_norm": 1.2740604877471924,
+      "learning_rate": 7.53621673612129e-06,
+      "loss": 0.3314,
+      "step": 4140
+    },
+    {
+      "epoch": 0.33128,
+      "grad_norm": 1.6157991886138916,
+      "learning_rate": 7.5351328154052835e-06,
+      "loss": 0.334,
+      "step": 4141
+    },
+    {
+      "epoch": 0.33136,
+      "grad_norm": 1.7817760705947876,
+      "learning_rate": 7.534048734299908e-06,
+      "loss": 0.3722,
+      "step": 4142
+    },
+    {
+      "epoch": 0.33144,
+      "grad_norm": 1.3499414920806885,
+      "learning_rate": 7.532964492873747e-06,
+      "loss": 0.2966,
+      "step": 4143
+    },
+    {
+      "epoch": 0.33152,
+      "grad_norm": 1.4766731262207031,
+      "learning_rate": 7.531880091195398e-06,
+      "loss": 0.3477,
+      "step": 4144
+    },
+    {
+      "epoch": 0.3316,
+      "grad_norm": 1.642026424407959,
+      "learning_rate": 7.530795529333468e-06,
+      "loss": 0.3435,
+      "step": 4145
+    },
+    {
+      "epoch": 0.33168,
+      "grad_norm": 1.4514563083648682,
+      "learning_rate": 7.529710807356572e-06,
+      "loss": 0.2747,
+      "step": 4146
+    },
+    {
+      "epoch": 0.33176,
+      "grad_norm": 1.4751369953155518,
+      "learning_rate": 7.528625925333337e-06,
+      "loss": 0.3201,
+      "step": 4147
+    },
+    {
+      "epoch": 0.33184,
+      "grad_norm": 1.3058736324310303,
+      "learning_rate": 7.5275408833324025e-06,
+      "loss": 0.2439,
+      "step": 4148
+    },
+    {
+      "epoch": 0.33192,
+      "grad_norm": 1.7958042621612549,
+      "learning_rate": 7.526455681422413e-06,
+      "loss": 0.344,
+      "step": 4149
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 1.462531328201294,
+      "learning_rate": 7.525370319672025e-06,
+      "loss": 0.2951,
+      "step": 4150
+    },
+    {
+      "epoch": 0.33208,
+      "grad_norm": 1.6958011388778687,
+      "learning_rate": 7.5242847981499064e-06,
+      "loss": 0.4044,
+      "step": 4151
+    },
+    {
+      "epoch": 0.33216,
+      "grad_norm": 1.6132330894470215,
+      "learning_rate": 7.5231991169247355e-06,
+      "loss": 0.5282,
+      "step": 4152
+    },
+    {
+      "epoch": 0.33224,
+      "grad_norm": 1.6550413370132446,
+      "learning_rate": 7.522113276065199e-06,
+      "loss": 0.3572,
+      "step": 4153
+    },
+    {
+      "epoch": 0.33232,
+      "grad_norm": 1.4192287921905518,
+      "learning_rate": 7.521027275639996e-06,
+      "loss": 0.3088,
+      "step": 4154
+    },
+    {
+      "epoch": 0.3324,
+      "grad_norm": 1.506937861442566,
+      "learning_rate": 7.5199411157178316e-06,
+      "loss": 0.2994,
+      "step": 4155
+    },
+    {
+      "epoch": 0.33248,
+      "grad_norm": 1.411502480506897,
+      "learning_rate": 7.518854796367424e-06,
+      "loss": 0.2824,
+      "step": 4156
+    },
+    {
+      "epoch": 0.33256,
+      "grad_norm": 1.7876211404800415,
+      "learning_rate": 7.517768317657503e-06,
+      "loss": 0.3639,
+      "step": 4157
+    },
+    {
+      "epoch": 0.33264,
+      "grad_norm": 1.318710207939148,
+      "learning_rate": 7.516681679656804e-06,
+      "loss": 0.2662,
+      "step": 4158
+    },
+    {
+      "epoch": 0.33272,
+      "grad_norm": 1.4276552200317383,
+      "learning_rate": 7.515594882434076e-06,
+      "loss": 0.3084,
+      "step": 4159
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 1.7345210313796997,
+      "learning_rate": 7.514507926058077e-06,
+      "loss": 0.3099,
+      "step": 4160
+    },
+    {
+      "epoch": 0.33288,
+      "grad_norm": 1.3311537504196167,
+      "learning_rate": 7.513420810597576e-06,
+      "loss": 0.2696,
+      "step": 4161
+    },
+    {
+      "epoch": 0.33296,
+      "grad_norm": 1.3819355964660645,
+      "learning_rate": 7.512333536121349e-06,
+      "loss": 0.3543,
+      "step": 4162
+    },
+    {
+      "epoch": 0.33304,
+      "grad_norm": 1.6310116052627563,
+      "learning_rate": 7.5112461026981855e-06,
+      "loss": 0.385,
+      "step": 4163
+    },
+    {
+      "epoch": 0.33312,
+      "grad_norm": 1.8169513940811157,
+      "learning_rate": 7.510158510396883e-06,
+      "loss": 0.4667,
+      "step": 4164
+    },
+    {
+      "epoch": 0.3332,
+      "grad_norm": 1.5222045183181763,
+      "learning_rate": 7.509070759286252e-06,
+      "loss": 0.3944,
+      "step": 4165
+    },
+    {
+      "epoch": 0.33328,
+      "grad_norm": 1.5854097604751587,
+      "learning_rate": 7.507982849435109e-06,
+      "loss": 0.3542,
+      "step": 4166
+    },
+    {
+      "epoch": 0.33336,
+      "grad_norm": 1.7982919216156006,
+      "learning_rate": 7.5068947809122835e-06,
+      "loss": 0.4572,
+      "step": 4167
+    },
+    {
+      "epoch": 0.33344,
+      "grad_norm": 1.4023102521896362,
+      "learning_rate": 7.505806553786613e-06,
+      "loss": 0.2961,
+      "step": 4168
+    },
+    {
+      "epoch": 0.33352,
+      "grad_norm": 1.5655568838119507,
+      "learning_rate": 7.504718168126947e-06,
+      "loss": 0.451,
+      "step": 4169
+    },
+    {
+      "epoch": 0.3336,
+      "grad_norm": 1.6367048025131226,
+      "learning_rate": 7.5036296240021425e-06,
+      "loss": 0.3058,
+      "step": 4170
+    },
+    {
+      "epoch": 0.33368,
+      "grad_norm": 1.4759256839752197,
+      "learning_rate": 7.5025409214810676e-06,
+      "loss": 0.3068,
+      "step": 4171
+    },
+    {
+      "epoch": 0.33376,
+      "grad_norm": 2.158902406692505,
+      "learning_rate": 7.501452060632603e-06,
+      "loss": 0.4489,
+      "step": 4172
+    },
+    {
+      "epoch": 0.33384,
+      "grad_norm": 1.8630950450897217,
+      "learning_rate": 7.500363041525638e-06,
+      "loss": 0.4119,
+      "step": 4173
+    },
+    {
+      "epoch": 0.33392,
+      "grad_norm": 1.9762685298919678,
+      "learning_rate": 7.499273864229069e-06,
+      "loss": 0.491,
+      "step": 4174
+    },
+    {
+      "epoch": 0.334,
+      "grad_norm": 1.4855585098266602,
+      "learning_rate": 7.4981845288118026e-06,
+      "loss": 0.2987,
+      "step": 4175
+    },
+    {
+      "epoch": 0.33408,
+      "grad_norm": 1.9851583242416382,
+      "learning_rate": 7.497095035342762e-06,
+      "loss": 0.3656,
+      "step": 4176
+    },
+    {
+      "epoch": 0.33416,
+      "grad_norm": 1.7855724096298218,
+      "learning_rate": 7.496005383890874e-06,
+      "loss": 0.344,
+      "step": 4177
+    },
+    {
+      "epoch": 0.33424,
+      "grad_norm": 1.5612376928329468,
+      "learning_rate": 7.4949155745250765e-06,
+      "loss": 0.3669,
+      "step": 4178
+    },
+    {
+      "epoch": 0.33432,
+      "grad_norm": 1.6685304641723633,
+      "learning_rate": 7.493825607314319e-06,
+      "loss": 0.4009,
+      "step": 4179
+    },
+    {
+      "epoch": 0.3344,
+      "grad_norm": 1.3716298341751099,
+      "learning_rate": 7.49273548232756e-06,
+      "loss": 0.3536,
+      "step": 4180
+    },
+    {
+      "epoch": 0.33448,
+      "grad_norm": 1.7887216806411743,
+      "learning_rate": 7.491645199633768e-06,
+      "loss": 0.4492,
+      "step": 4181
+    },
+    {
+      "epoch": 0.33456,
+      "grad_norm": 1.4562711715698242,
+      "learning_rate": 7.4905547593019205e-06,
+      "loss": 0.3379,
+      "step": 4182
+    },
+    {
+      "epoch": 0.33464,
+      "grad_norm": 1.4248229265213013,
+      "learning_rate": 7.4894641614010065e-06,
+      "loss": 0.3323,
+      "step": 4183
+    },
+    {
+      "epoch": 0.33472,
+      "grad_norm": 1.4600697755813599,
+      "learning_rate": 7.488373406000024e-06,
+      "loss": 0.3882,
+      "step": 4184
+    },
+    {
+      "epoch": 0.3348,
+      "grad_norm": 1.9589154720306396,
+      "learning_rate": 7.487282493167985e-06,
+      "loss": 0.4016,
+      "step": 4185
+    },
+    {
+      "epoch": 0.33488,
+      "grad_norm": 1.8087719678878784,
+      "learning_rate": 7.486191422973903e-06,
+      "loss": 0.3524,
+      "step": 4186
+    },
+    {
+      "epoch": 0.33496,
+      "grad_norm": 1.154066801071167,
+      "learning_rate": 7.485100195486808e-06,
+      "loss": 0.2329,
+      "step": 4187
+    },
+    {
+      "epoch": 0.33504,
+      "grad_norm": 2.5276176929473877,
+      "learning_rate": 7.48400881077574e-06,
+      "loss": 0.5058,
+      "step": 4188
+    },
+    {
+      "epoch": 0.33512,
+      "grad_norm": 2.2821571826934814,
+      "learning_rate": 7.482917268909746e-06,
+      "loss": 0.5157,
+      "step": 4189
+    },
+    {
+      "epoch": 0.3352,
+      "grad_norm": 1.4835360050201416,
+      "learning_rate": 7.481825569957886e-06,
+      "loss": 0.3223,
+      "step": 4190
+    },
+    {
+      "epoch": 0.33528,
+      "grad_norm": 1.7583729028701782,
+      "learning_rate": 7.4807337139892235e-06,
+      "loss": 0.3243,
+      "step": 4191
+    },
+    {
+      "epoch": 0.33536,
+      "grad_norm": 1.958134412765503,
+      "learning_rate": 7.479641701072841e-06,
+      "loss": 0.448,
+      "step": 4192
+    },
+    {
+      "epoch": 0.33544,
+      "grad_norm": 1.5461760759353638,
+      "learning_rate": 7.478549531277824e-06,
+      "loss": 0.2942,
+      "step": 4193
+    },
+    {
+      "epoch": 0.33552,
+      "grad_norm": 1.5328489542007446,
+      "learning_rate": 7.477457204673272e-06,
+      "loss": 0.327,
+      "step": 4194
+    },
+    {
+      "epoch": 0.3356,
+      "grad_norm": 1.610932469367981,
+      "learning_rate": 7.476364721328292e-06,
+      "loss": 0.3925,
+      "step": 4195
+    },
+    {
+      "epoch": 0.33568,
+      "grad_norm": 1.6235414743423462,
+      "learning_rate": 7.475272081312003e-06,
+      "loss": 0.3026,
+      "step": 4196
+    },
+    {
+      "epoch": 0.33576,
+      "grad_norm": 1.615761399269104,
+      "learning_rate": 7.4741792846935304e-06,
+      "loss": 0.3267,
+      "step": 4197
+    },
+    {
+      "epoch": 0.33584,
+      "grad_norm": 1.9006683826446533,
+      "learning_rate": 7.473086331542013e-06,
+      "loss": 0.3666,
+      "step": 4198
+    },
+    {
+      "epoch": 0.33592,
+      "grad_norm": 1.3763328790664673,
+      "learning_rate": 7.471993221926599e-06,
+      "loss": 0.2783,
+      "step": 4199
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 1.5629264116287231,
+      "learning_rate": 7.4708999559164455e-06,
+      "loss": 0.3361,
+      "step": 4200
+    },
+    {
+      "epoch": 0.33608,
+      "grad_norm": 1.5894604921340942,
+      "learning_rate": 7.469806533580719e-06,
+      "loss": 0.3866,
+      "step": 4201
+    },
+    {
+      "epoch": 0.33616,
+      "grad_norm": 1.707032322883606,
+      "learning_rate": 7.468712954988597e-06,
+      "loss": 0.337,
+      "step": 4202
+    },
+    {
+      "epoch": 0.33624,
+      "grad_norm": 1.5144277811050415,
+      "learning_rate": 7.467619220209268e-06,
+      "loss": 0.4724,
+      "step": 4203
+    },
+    {
+      "epoch": 0.33632,
+      "grad_norm": 1.9514338970184326,
+      "learning_rate": 7.466525329311927e-06,
+      "loss": 0.4198,
+      "step": 4204
+    },
+    {
+      "epoch": 0.3364,
+      "grad_norm": 1.3086645603179932,
+      "learning_rate": 7.4654312823657805e-06,
+      "loss": 0.2661,
+      "step": 4205
+    },
+    {
+      "epoch": 0.33648,
+      "grad_norm": 1.8274202346801758,
+      "learning_rate": 7.464337079440046e-06,
+      "loss": 0.3858,
+      "step": 4206
+    },
+    {
+      "epoch": 0.33656,
+      "grad_norm": 1.772868037223816,
+      "learning_rate": 7.4632427206039505e-06,
+      "loss": 0.3487,
+      "step": 4207
+    },
+    {
+      "epoch": 0.33664,
+      "grad_norm": 1.3562657833099365,
+      "learning_rate": 7.4621482059267315e-06,
+      "loss": 0.3073,
+      "step": 4208
+    },
+    {
+      "epoch": 0.33672,
+      "grad_norm": 1.3614493608474731,
+      "learning_rate": 7.461053535477632e-06,
+      "loss": 0.3029,
+      "step": 4209
+    },
+    {
+      "epoch": 0.3368,
+      "grad_norm": 1.4911876916885376,
+      "learning_rate": 7.459958709325911e-06,
+      "loss": 0.4744,
+      "step": 4210
+    },
+    {
+      "epoch": 0.33688,
+      "grad_norm": 1.3329766988754272,
+      "learning_rate": 7.458863727540832e-06,
+      "loss": 0.2758,
+      "step": 4211
+    },
+    {
+      "epoch": 0.33696,
+      "grad_norm": 1.798316478729248,
+      "learning_rate": 7.457768590191674e-06,
+      "loss": 0.3735,
+      "step": 4212
+    },
+    {
+      "epoch": 0.33704,
+      "grad_norm": 1.8065177202224731,
+      "learning_rate": 7.456673297347721e-06,
+      "loss": 0.3828,
+      "step": 4213
+    },
+    {
+      "epoch": 0.33712,
+      "grad_norm": 1.4247829914093018,
+      "learning_rate": 7.455577849078269e-06,
+      "loss": 0.3492,
+      "step": 4214
+    },
+    {
+      "epoch": 0.3372,
+      "grad_norm": 1.6128746271133423,
+      "learning_rate": 7.454482245452621e-06,
+      "loss": 0.4182,
+      "step": 4215
+    },
+    {
+      "epoch": 0.33728,
+      "grad_norm": 1.8242870569229126,
+      "learning_rate": 7.453386486540095e-06,
+      "loss": 0.4306,
+      "step": 4216
+    },
+    {
+      "epoch": 0.33736,
+      "grad_norm": 1.6618531942367554,
+      "learning_rate": 7.4522905724100155e-06,
+      "loss": 0.3258,
+      "step": 4217
+    },
+    {
+      "epoch": 0.33744,
+      "grad_norm": 2.2641119956970215,
+      "learning_rate": 7.451194503131715e-06,
+      "loss": 0.4345,
+      "step": 4218
+    },
+    {
+      "epoch": 0.33752,
+      "grad_norm": 1.4390738010406494,
+      "learning_rate": 7.450098278774542e-06,
+      "loss": 0.2435,
+      "step": 4219
+    },
+    {
+      "epoch": 0.3376,
+      "grad_norm": 1.6688294410705566,
+      "learning_rate": 7.449001899407849e-06,
+      "loss": 0.3547,
+      "step": 4220
+    },
+    {
+      "epoch": 0.33768,
+      "grad_norm": 1.3295515775680542,
+      "learning_rate": 7.447905365101e-06,
+      "loss": 0.3372,
+      "step": 4221
+    },
+    {
+      "epoch": 0.33776,
+      "grad_norm": 1.7421692609786987,
+      "learning_rate": 7.446808675923369e-06,
+      "loss": 0.417,
+      "step": 4222
+    },
+    {
+      "epoch": 0.33784,
+      "grad_norm": 1.7241042852401733,
+      "learning_rate": 7.44571183194434e-06,
+      "loss": 0.436,
+      "step": 4223
+    },
+    {
+      "epoch": 0.33792,
+      "grad_norm": 1.7911804914474487,
+      "learning_rate": 7.4446148332333065e-06,
+      "loss": 0.4371,
+      "step": 4224
+    },
+    {
+      "epoch": 0.338,
+      "grad_norm": 1.1841015815734863,
+      "learning_rate": 7.4435176798596735e-06,
+      "loss": 0.3404,
+      "step": 4225
+    },
+    {
+      "epoch": 0.33808,
+      "grad_norm": 2.037875175476074,
+      "learning_rate": 7.442420371892852e-06,
+      "loss": 0.373,
+      "step": 4226
+    },
+    {
+      "epoch": 0.33816,
+      "grad_norm": 1.385729432106018,
+      "learning_rate": 7.4413229094022655e-06,
+      "loss": 0.297,
+      "step": 4227
+    },
+    {
+      "epoch": 0.33824,
+      "grad_norm": 1.7231978178024292,
+      "learning_rate": 7.440225292457348e-06,
+      "loss": 0.4303,
+      "step": 4228
+    },
+    {
+      "epoch": 0.33832,
+      "grad_norm": 1.3325543403625488,
+      "learning_rate": 7.43912752112754e-06,
+      "loss": 0.243,
+      "step": 4229
+    },
+    {
+      "epoch": 0.3384,
+      "grad_norm": 1.3618645668029785,
+      "learning_rate": 7.438029595482297e-06,
+      "loss": 0.3506,
+      "step": 4230
+    },
+    {
+      "epoch": 0.33848,
+      "grad_norm": 1.3418787717819214,
+      "learning_rate": 7.436931515591077e-06,
+      "loss": 0.2405,
+      "step": 4231
+    },
+    {
+      "epoch": 0.33856,
+      "grad_norm": 1.4859660863876343,
+      "learning_rate": 7.435833281523356e-06,
+      "loss": 0.3964,
+      "step": 4232
+    },
+    {
+      "epoch": 0.33864,
+      "grad_norm": 1.6270725727081299,
+      "learning_rate": 7.434734893348612e-06,
+      "loss": 0.3465,
+      "step": 4233
+    },
+    {
+      "epoch": 0.33872,
+      "grad_norm": 1.3810713291168213,
+      "learning_rate": 7.433636351136338e-06,
+      "loss": 0.3105,
+      "step": 4234
+    },
+    {
+      "epoch": 0.3388,
+      "grad_norm": 1.595778465270996,
+      "learning_rate": 7.432537654956036e-06,
+      "loss": 0.3078,
+      "step": 4235
+    },
+    {
+      "epoch": 0.33888,
+      "grad_norm": 1.770795464515686,
+      "learning_rate": 7.431438804877216e-06,
+      "loss": 0.3117,
+      "step": 4236
+    },
+    {
+      "epoch": 0.33896,
+      "grad_norm": 1.5661720037460327,
+      "learning_rate": 7.4303398009694e-06,
+      "loss": 0.3066,
+      "step": 4237
+    },
+    {
+      "epoch": 0.33904,
+      "grad_norm": 1.6826071739196777,
+      "learning_rate": 7.429240643302114e-06,
+      "loss": 0.2934,
+      "step": 4238
+    },
+    {
+      "epoch": 0.33912,
+      "grad_norm": 1.8159326314926147,
+      "learning_rate": 7.428141331944901e-06,
+      "loss": 0.3794,
+      "step": 4239
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 1.6127220392227173,
+      "learning_rate": 7.42704186696731e-06,
+      "loss": 0.3804,
+      "step": 4240
+    },
+    {
+      "epoch": 0.33928,
+      "grad_norm": 1.844524621963501,
+      "learning_rate": 7.425942248438902e-06,
+      "loss": 0.4147,
+      "step": 4241
+    },
+    {
+      "epoch": 0.33936,
+      "grad_norm": 1.4185341596603394,
+      "learning_rate": 7.424842476429246e-06,
+      "loss": 0.247,
+      "step": 4242
+    },
+    {
+      "epoch": 0.33944,
+      "grad_norm": 1.1865193843841553,
+      "learning_rate": 7.423742551007919e-06,
+      "loss": 0.2286,
+      "step": 4243
+    },
+    {
+      "epoch": 0.33952,
+      "grad_norm": 1.284876823425293,
+      "learning_rate": 7.422642472244511e-06,
+      "loss": 0.2601,
+      "step": 4244
+    },
+    {
+      "epoch": 0.3396,
+      "grad_norm": 1.4666966199874878,
+      "learning_rate": 7.42154224020862e-06,
+      "loss": 0.315,
+      "step": 4245
+    },
+    {
+      "epoch": 0.33968,
+      "grad_norm": 1.2044739723205566,
+      "learning_rate": 7.420441854969853e-06,
+      "loss": 0.2234,
+      "step": 4246
+    },
+    {
+      "epoch": 0.33976,
+      "grad_norm": 1.548977255821228,
+      "learning_rate": 7.41934131659783e-06,
+      "loss": 0.3675,
+      "step": 4247
+    },
+    {
+      "epoch": 0.33984,
+      "grad_norm": 1.2386624813079834,
+      "learning_rate": 7.418240625162178e-06,
+      "loss": 0.2938,
+      "step": 4248
+    },
+    {
+      "epoch": 0.33992,
+      "grad_norm": 1.4964457750320435,
+      "learning_rate": 7.4171397807325314e-06,
+      "loss": 0.3568,
+      "step": 4249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.543899416923523,
+      "learning_rate": 7.416038783378539e-06,
+      "loss": 0.4038,
+      "step": 4250
+    },
+    {
+      "epoch": 0.34008,
+      "grad_norm": 1.7202891111373901,
+      "learning_rate": 7.414937633169857e-06,
+      "loss": 0.4456,
+      "step": 4251
+    },
+    {
+      "epoch": 0.34016,
+      "grad_norm": 1.3532027006149292,
+      "learning_rate": 7.413836330176149e-06,
+      "loss": 0.2768,
+      "step": 4252
+    },
+    {
+      "epoch": 0.34024,
+      "grad_norm": 1.1683356761932373,
+      "learning_rate": 7.412734874467096e-06,
+      "loss": 0.3269,
+      "step": 4253
+    },
+    {
+      "epoch": 0.34032,
+      "grad_norm": 1.4227070808410645,
+      "learning_rate": 7.411633266112379e-06,
+      "loss": 0.3285,
+      "step": 4254
+    },
+    {
+      "epoch": 0.3404,
+      "grad_norm": 1.7242900133132935,
+      "learning_rate": 7.410531505181697e-06,
+      "loss": 0.4367,
+      "step": 4255
+    },
+    {
+      "epoch": 0.34048,
+      "grad_norm": 1.3391542434692383,
+      "learning_rate": 7.40942959174475e-06,
+      "loss": 0.3332,
+      "step": 4256
+    },
+    {
+      "epoch": 0.34056,
+      "grad_norm": 1.4209461212158203,
+      "learning_rate": 7.408327525871255e-06,
+      "loss": 0.3338,
+      "step": 4257
+    },
+    {
+      "epoch": 0.34064,
+      "grad_norm": 2.0780012607574463,
+      "learning_rate": 7.4072253076309375e-06,
+      "loss": 0.51,
+      "step": 4258
+    },
+    {
+      "epoch": 0.34072,
+      "grad_norm": 2.3870983123779297,
+      "learning_rate": 7.4061229370935275e-06,
+      "loss": 0.4668,
+      "step": 4259
+    },
+    {
+      "epoch": 0.3408,
+      "grad_norm": 1.4779009819030762,
+      "learning_rate": 7.405020414328771e-06,
+      "loss": 0.2749,
+      "step": 4260
+    },
+    {
+      "epoch": 0.34088,
+      "grad_norm": 1.7060835361480713,
+      "learning_rate": 7.4039177394064196e-06,
+      "loss": 0.3663,
+      "step": 4261
+    },
+    {
+      "epoch": 0.34096,
+      "grad_norm": 1.5579965114593506,
+      "learning_rate": 7.4028149123962365e-06,
+      "loss": 0.3155,
+      "step": 4262
+    },
+    {
+      "epoch": 0.34104,
+      "grad_norm": 1.6776325702667236,
+      "learning_rate": 7.401711933367994e-06,
+      "loss": 0.3946,
+      "step": 4263
+    },
+    {
+      "epoch": 0.34112,
+      "grad_norm": 2.1484310626983643,
+      "learning_rate": 7.4006088023914735e-06,
+      "loss": 0.541,
+      "step": 4264
+    },
+    {
+      "epoch": 0.3412,
+      "grad_norm": 1.8193962574005127,
+      "learning_rate": 7.399505519536465e-06,
+      "loss": 0.4295,
+      "step": 4265
+    },
+    {
+      "epoch": 0.34128,
+      "grad_norm": 1.7744461297988892,
+      "learning_rate": 7.398402084872773e-06,
+      "loss": 0.3818,
+      "step": 4266
+    },
+    {
+      "epoch": 0.34136,
+      "grad_norm": 1.5246503353118896,
+      "learning_rate": 7.397298498470206e-06,
+      "loss": 0.3992,
+      "step": 4267
+    },
+    {
+      "epoch": 0.34144,
+      "grad_norm": 1.3743784427642822,
+      "learning_rate": 7.396194760398584e-06,
+      "loss": 0.3579,
+      "step": 4268
+    },
+    {
+      "epoch": 0.34152,
+      "grad_norm": 1.4627759456634521,
+      "learning_rate": 7.395090870727737e-06,
+      "loss": 0.4135,
+      "step": 4269
+    },
+    {
+      "epoch": 0.3416,
+      "grad_norm": 1.4635417461395264,
+      "learning_rate": 7.393986829527506e-06,
+      "loss": 0.3232,
+      "step": 4270
+    },
+    {
+      "epoch": 0.34168,
+      "grad_norm": 1.804682731628418,
+      "learning_rate": 7.392882636867738e-06,
+      "loss": 0.3609,
+      "step": 4271
+    },
+    {
+      "epoch": 0.34176,
+      "grad_norm": 1.6113015413284302,
+      "learning_rate": 7.39177829281829e-06,
+      "loss": 0.319,
+      "step": 4272
+    },
+    {
+      "epoch": 0.34184,
+      "grad_norm": 1.5073604583740234,
+      "learning_rate": 7.390673797449035e-06,
+      "loss": 0.3688,
+      "step": 4273
+    },
+    {
+      "epoch": 0.34192,
+      "grad_norm": 1.9971095323562622,
+      "learning_rate": 7.389569150829847e-06,
+      "loss": 0.4176,
+      "step": 4274
+    },
+    {
+      "epoch": 0.342,
+      "grad_norm": 1.400302767753601,
+      "learning_rate": 7.3884643530306146e-06,
+      "loss": 0.281,
+      "step": 4275
+    },
+    {
+      "epoch": 0.34208,
+      "grad_norm": 1.5128846168518066,
+      "learning_rate": 7.387359404121234e-06,
+      "loss": 0.3501,
+      "step": 4276
+    },
+    {
+      "epoch": 0.34216,
+      "grad_norm": 1.645851969718933,
+      "learning_rate": 7.386254304171612e-06,
+      "loss": 0.3803,
+      "step": 4277
+    },
+    {
+      "epoch": 0.34224,
+      "grad_norm": 1.6219562292099,
+      "learning_rate": 7.385149053251664e-06,
+      "loss": 0.3261,
+      "step": 4278
+    },
+    {
+      "epoch": 0.34232,
+      "grad_norm": 2.3219873905181885,
+      "learning_rate": 7.3840436514313185e-06,
+      "loss": 0.5242,
+      "step": 4279
+    },
+    {
+      "epoch": 0.3424,
+      "grad_norm": 1.3534795045852661,
+      "learning_rate": 7.382938098780505e-06,
+      "loss": 0.2412,
+      "step": 4280
+    },
+    {
+      "epoch": 0.34248,
+      "grad_norm": 1.458430290222168,
+      "learning_rate": 7.381832395369175e-06,
+      "loss": 0.303,
+      "step": 4281
+    },
+    {
+      "epoch": 0.34256,
+      "grad_norm": 1.8676531314849854,
+      "learning_rate": 7.380726541267276e-06,
+      "loss": 0.4078,
+      "step": 4282
+    },
+    {
+      "epoch": 0.34264,
+      "grad_norm": 1.8521391153335571,
+      "learning_rate": 7.379620536544776e-06,
+      "loss": 0.5314,
+      "step": 4283
+    },
+    {
+      "epoch": 0.34272,
+      "grad_norm": 1.4564507007598877,
+      "learning_rate": 7.378514381271646e-06,
+      "loss": 0.3131,
+      "step": 4284
+    },
+    {
+      "epoch": 0.3428,
+      "grad_norm": 1.7293351888656616,
+      "learning_rate": 7.377408075517871e-06,
+      "loss": 0.3745,
+      "step": 4285
+    },
+    {
+      "epoch": 0.34288,
+      "grad_norm": 1.8894083499908447,
+      "learning_rate": 7.376301619353441e-06,
+      "loss": 0.4519,
+      "step": 4286
+    },
+    {
+      "epoch": 0.34296,
+      "grad_norm": 1.4705125093460083,
+      "learning_rate": 7.375195012848359e-06,
+      "loss": 0.3943,
+      "step": 4287
+    },
+    {
+      "epoch": 0.34304,
+      "grad_norm": 1.75259530544281,
+      "learning_rate": 7.374088256072635e-06,
+      "loss": 0.3682,
+      "step": 4288
+    },
+    {
+      "epoch": 0.34312,
+      "grad_norm": 1.3996787071228027,
+      "learning_rate": 7.372981349096291e-06,
+      "loss": 0.2633,
+      "step": 4289
+    },
+    {
+      "epoch": 0.3432,
+      "grad_norm": 1.3005949258804321,
+      "learning_rate": 7.371874291989358e-06,
+      "loss": 0.2874,
+      "step": 4290
+    },
+    {
+      "epoch": 0.34328,
+      "grad_norm": 1.8466405868530273,
+      "learning_rate": 7.370767084821875e-06,
+      "loss": 0.4616,
+      "step": 4291
+    },
+    {
+      "epoch": 0.34336,
+      "grad_norm": 1.9775362014770508,
+      "learning_rate": 7.369659727663889e-06,
+      "loss": 0.4652,
+      "step": 4292
+    },
+    {
+      "epoch": 0.34344,
+      "grad_norm": 1.816942811012268,
+      "learning_rate": 7.368552220585464e-06,
+      "loss": 0.5346,
+      "step": 4293
+    },
+    {
+      "epoch": 0.34352,
+      "grad_norm": 1.3692759275436401,
+      "learning_rate": 7.367444563656663e-06,
+      "loss": 0.4167,
+      "step": 4294
+    },
+    {
+      "epoch": 0.3436,
+      "grad_norm": 1.5143382549285889,
+      "learning_rate": 7.3663367569475665e-06,
+      "loss": 0.2954,
+      "step": 4295
+    },
+    {
+      "epoch": 0.34368,
+      "grad_norm": 1.7048492431640625,
+      "learning_rate": 7.365228800528263e-06,
+      "loss": 0.3539,
+      "step": 4296
+    },
+    {
+      "epoch": 0.34376,
+      "grad_norm": 1.409371256828308,
+      "learning_rate": 7.364120694468845e-06,
+      "loss": 0.3441,
+      "step": 4297
+    },
+    {
+      "epoch": 0.34384,
+      "grad_norm": 1.3780807256698608,
+      "learning_rate": 7.363012438839421e-06,
+      "loss": 0.3102,
+      "step": 4298
+    },
+    {
+      "epoch": 0.34392,
+      "grad_norm": 1.3144261837005615,
+      "learning_rate": 7.361904033710108e-06,
+      "loss": 0.2998,
+      "step": 4299
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 1.5273244380950928,
+      "learning_rate": 7.360795479151029e-06,
+      "loss": 0.3879,
+      "step": 4300
+    },
+    {
+      "epoch": 0.34408,
+      "grad_norm": 1.3805782794952393,
+      "learning_rate": 7.359686775232318e-06,
+      "loss": 0.3264,
+      "step": 4301
+    },
+    {
+      "epoch": 0.34416,
+      "grad_norm": 1.8906997442245483,
+      "learning_rate": 7.358577922024123e-06,
+      "loss": 0.496,
+      "step": 4302
+    },
+    {
+      "epoch": 0.34424,
+      "grad_norm": 1.627760410308838,
+      "learning_rate": 7.357468919596593e-06,
+      "loss": 0.3229,
+      "step": 4303
+    },
+    {
+      "epoch": 0.34432,
+      "grad_norm": 1.7278337478637695,
+      "learning_rate": 7.356359768019894e-06,
+      "loss": 0.4291,
+      "step": 4304
+    },
+    {
+      "epoch": 0.3444,
+      "grad_norm": 1.5394865274429321,
+      "learning_rate": 7.355250467364196e-06,
+      "loss": 0.321,
+      "step": 4305
+    },
+    {
+      "epoch": 0.34448,
+      "grad_norm": 1.5800862312316895,
+      "learning_rate": 7.354141017699681e-06,
+      "loss": 0.3424,
+      "step": 4306
+    },
+    {
+      "epoch": 0.34456,
+      "grad_norm": 1.81803297996521,
+      "learning_rate": 7.35303141909654e-06,
+      "loss": 0.3574,
+      "step": 4307
+    },
+    {
+      "epoch": 0.34464,
+      "grad_norm": 1.6725924015045166,
+      "learning_rate": 7.351921671624977e-06,
+      "loss": 0.3973,
+      "step": 4308
+    },
+    {
+      "epoch": 0.34472,
+      "grad_norm": 1.7673554420471191,
+      "learning_rate": 7.350811775355197e-06,
+      "loss": 0.4279,
+      "step": 4309
+    },
+    {
+      "epoch": 0.3448,
+      "grad_norm": 1.319742202758789,
+      "learning_rate": 7.349701730357424e-06,
+      "loss": 0.2885,
+      "step": 4310
+    },
+    {
+      "epoch": 0.34488,
+      "grad_norm": 1.64664888381958,
+      "learning_rate": 7.348591536701882e-06,
+      "loss": 0.4351,
+      "step": 4311
+    },
+    {
+      "epoch": 0.34496,
+      "grad_norm": 2.0317161083221436,
+      "learning_rate": 7.347481194458813e-06,
+      "loss": 0.4843,
+      "step": 4312
+    },
+    {
+      "epoch": 0.34504,
+      "grad_norm": 1.2819757461547852,
+      "learning_rate": 7.346370703698464e-06,
+      "loss": 0.2645,
+      "step": 4313
+    },
+    {
+      "epoch": 0.34512,
+      "grad_norm": 1.667986512184143,
+      "learning_rate": 7.345260064491092e-06,
+      "loss": 0.4948,
+      "step": 4314
+    },
+    {
+      "epoch": 0.3452,
+      "grad_norm": 1.6327108144760132,
+      "learning_rate": 7.344149276906962e-06,
+      "loss": 0.3498,
+      "step": 4315
+    },
+    {
+      "epoch": 0.34528,
+      "grad_norm": 1.8304697275161743,
+      "learning_rate": 7.343038341016352e-06,
+      "loss": 0.364,
+      "step": 4316
+    },
+    {
+      "epoch": 0.34536,
+      "grad_norm": 1.6535539627075195,
+      "learning_rate": 7.341927256889545e-06,
+      "loss": 0.3881,
+      "step": 4317
+    },
+    {
+      "epoch": 0.34544,
+      "grad_norm": 1.561083197593689,
+      "learning_rate": 7.340816024596838e-06,
+      "loss": 0.4237,
+      "step": 4318
+    },
+    {
+      "epoch": 0.34552,
+      "grad_norm": 1.6609004735946655,
+      "learning_rate": 7.339704644208531e-06,
+      "loss": 0.3515,
+      "step": 4319
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 1.5211528539657593,
+      "learning_rate": 7.338593115794942e-06,
+      "loss": 0.3339,
+      "step": 4320
+    },
+    {
+      "epoch": 0.34568,
+      "grad_norm": 1.4612761735916138,
+      "learning_rate": 7.33748143942639e-06,
+      "loss": 0.3212,
+      "step": 4321
+    },
+    {
+      "epoch": 0.34576,
+      "grad_norm": 1.503321647644043,
+      "learning_rate": 7.336369615173209e-06,
+      "loss": 0.4281,
+      "step": 4322
+    },
+    {
+      "epoch": 0.34584,
+      "grad_norm": 1.4206557273864746,
+      "learning_rate": 7.33525764310574e-06,
+      "loss": 0.309,
+      "step": 4323
+    },
+    {
+      "epoch": 0.34592,
+      "grad_norm": 1.1924537420272827,
+      "learning_rate": 7.334145523294334e-06,
+      "loss": 0.2613,
+      "step": 4324
+    },
+    {
+      "epoch": 0.346,
+      "grad_norm": 1.3378185033798218,
+      "learning_rate": 7.333033255809351e-06,
+      "loss": 0.2799,
+      "step": 4325
+    },
+    {
+      "epoch": 0.34608,
+      "grad_norm": 1.678286075592041,
+      "learning_rate": 7.33192084072116e-06,
+      "loss": 0.3864,
+      "step": 4326
+    },
+    {
+      "epoch": 0.34616,
+      "grad_norm": 1.57291841506958,
+      "learning_rate": 7.330808278100141e-06,
+      "loss": 0.3265,
+      "step": 4327
+    },
+    {
+      "epoch": 0.34624,
+      "grad_norm": 1.6815236806869507,
+      "learning_rate": 7.329695568016679e-06,
+      "loss": 0.3757,
+      "step": 4328
+    },
+    {
+      "epoch": 0.34632,
+      "grad_norm": 1.3270570039749146,
+      "learning_rate": 7.328582710541174e-06,
+      "loss": 0.2978,
+      "step": 4329
+    },
+    {
+      "epoch": 0.3464,
+      "grad_norm": 1.3988102674484253,
+      "learning_rate": 7.327469705744034e-06,
+      "loss": 0.3347,
+      "step": 4330
+    },
+    {
+      "epoch": 0.34648,
+      "grad_norm": 1.7103873491287231,
+      "learning_rate": 7.3263565536956725e-06,
+      "loss": 0.3344,
+      "step": 4331
+    },
+    {
+      "epoch": 0.34656,
+      "grad_norm": 1.768531322479248,
+      "learning_rate": 7.325243254466516e-06,
+      "loss": 0.3403,
+      "step": 4332
+    },
+    {
+      "epoch": 0.34664,
+      "grad_norm": 1.3608916997909546,
+      "learning_rate": 7.3241298081269995e-06,
+      "loss": 0.2652,
+      "step": 4333
+    },
+    {
+      "epoch": 0.34672,
+      "grad_norm": 2.1497552394866943,
+      "learning_rate": 7.323016214747566e-06,
+      "loss": 0.404,
+      "step": 4334
+    },
+    {
+      "epoch": 0.3468,
+      "grad_norm": 1.8790044784545898,
+      "learning_rate": 7.321902474398669e-06,
+      "loss": 0.3277,
+      "step": 4335
+    },
+    {
+      "epoch": 0.34688,
+      "grad_norm": 1.679563045501709,
+      "learning_rate": 7.3207885871507715e-06,
+      "loss": 0.3758,
+      "step": 4336
+    },
+    {
+      "epoch": 0.34696,
+      "grad_norm": 1.669385552406311,
+      "learning_rate": 7.319674553074347e-06,
+      "loss": 0.3434,
+      "step": 4337
+    },
+    {
+      "epoch": 0.34704,
+      "grad_norm": 1.5998573303222656,
+      "learning_rate": 7.318560372239873e-06,
+      "loss": 0.4371,
+      "step": 4338
+    },
+    {
+      "epoch": 0.34712,
+      "grad_norm": 1.4684016704559326,
+      "learning_rate": 7.317446044717845e-06,
+      "loss": 0.2922,
+      "step": 4339
+    },
+    {
+      "epoch": 0.3472,
+      "grad_norm": 1.7351469993591309,
+      "learning_rate": 7.316331570578759e-06,
+      "loss": 0.41,
+      "step": 4340
+    },
+    {
+      "epoch": 0.34728,
+      "grad_norm": 1.467858910560608,
+      "learning_rate": 7.315216949893124e-06,
+      "loss": 0.353,
+      "step": 4341
+    },
+    {
+      "epoch": 0.34736,
+      "grad_norm": 1.418001651763916,
+      "learning_rate": 7.314102182731459e-06,
+      "loss": 0.3429,
+      "step": 4342
+    },
+    {
+      "epoch": 0.34744,
+      "grad_norm": 1.5263274908065796,
+      "learning_rate": 7.3129872691642935e-06,
+      "loss": 0.4126,
+      "step": 4343
+    },
+    {
+      "epoch": 0.34752,
+      "grad_norm": 1.2809735536575317,
+      "learning_rate": 7.3118722092621615e-06,
+      "loss": 0.2515,
+      "step": 4344
+    },
+    {
+      "epoch": 0.3476,
+      "grad_norm": 1.377831220626831,
+      "learning_rate": 7.31075700309561e-06,
+      "loss": 0.2804,
+      "step": 4345
+    },
+    {
+      "epoch": 0.34768,
+      "grad_norm": 1.5023661851882935,
+      "learning_rate": 7.309641650735195e-06,
+      "loss": 0.3874,
+      "step": 4346
+    },
+    {
+      "epoch": 0.34776,
+      "grad_norm": 1.7414848804473877,
+      "learning_rate": 7.308526152251482e-06,
+      "loss": 0.346,
+      "step": 4347
+    },
+    {
+      "epoch": 0.34784,
+      "grad_norm": 1.4180262088775635,
+      "learning_rate": 7.307410507715044e-06,
+      "loss": 0.3632,
+      "step": 4348
+    },
+    {
+      "epoch": 0.34792,
+      "grad_norm": 1.8054182529449463,
+      "learning_rate": 7.3062947171964626e-06,
+      "loss": 0.3541,
+      "step": 4349
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 1.9470112323760986,
+      "learning_rate": 7.305178780766332e-06,
+      "loss": 0.3118,
+      "step": 4350
+    },
+    {
+      "epoch": 0.34808,
+      "grad_norm": 1.3836066722869873,
+      "learning_rate": 7.304062698495253e-06,
+      "loss": 0.2609,
+      "step": 4351
+    },
+    {
+      "epoch": 0.34816,
+      "grad_norm": 1.3901145458221436,
+      "learning_rate": 7.302946470453835e-06,
+      "loss": 0.3282,
+      "step": 4352
+    },
+    {
+      "epoch": 0.34824,
+      "grad_norm": 1.5342923402786255,
+      "learning_rate": 7.301830096712701e-06,
+      "loss": 0.284,
+      "step": 4353
+    },
+    {
+      "epoch": 0.34832,
+      "grad_norm": 1.7975738048553467,
+      "learning_rate": 7.300713577342477e-06,
+      "loss": 0.415,
+      "step": 4354
+    },
+    {
+      "epoch": 0.3484,
+      "grad_norm": 1.4693024158477783,
+      "learning_rate": 7.2995969124138054e-06,
+      "loss": 0.3352,
+      "step": 4355
+    },
+    {
+      "epoch": 0.34848,
+      "grad_norm": 1.8647490739822388,
+      "learning_rate": 7.298480101997331e-06,
+      "loss": 0.399,
+      "step": 4356
+    },
+    {
+      "epoch": 0.34856,
+      "grad_norm": 1.3949613571166992,
+      "learning_rate": 7.297363146163711e-06,
+      "loss": 0.4118,
+      "step": 4357
+    },
+    {
+      "epoch": 0.34864,
+      "grad_norm": 1.5804004669189453,
+      "learning_rate": 7.296246044983611e-06,
+      "loss": 0.3258,
+      "step": 4358
+    },
+    {
+      "epoch": 0.34872,
+      "grad_norm": 1.418825626373291,
+      "learning_rate": 7.295128798527708e-06,
+      "loss": 0.324,
+      "step": 4359
+    },
+    {
+      "epoch": 0.3488,
+      "grad_norm": 1.6800353527069092,
+      "learning_rate": 7.294011406866686e-06,
+      "loss": 0.4247,
+      "step": 4360
+    },
+    {
+      "epoch": 0.34888,
+      "grad_norm": 1.2685750722885132,
+      "learning_rate": 7.292893870071238e-06,
+      "loss": 0.2687,
+      "step": 4361
+    },
+    {
+      "epoch": 0.34896,
+      "grad_norm": 1.8051155805587769,
+      "learning_rate": 7.2917761882120655e-06,
+      "loss": 0.4082,
+      "step": 4362
+    },
+    {
+      "epoch": 0.34904,
+      "grad_norm": 1.8455771207809448,
+      "learning_rate": 7.290658361359883e-06,
+      "loss": 0.3075,
+      "step": 4363
+    },
+    {
+      "epoch": 0.34912,
+      "grad_norm": 1.6396703720092773,
+      "learning_rate": 7.28954038958541e-06,
+      "loss": 0.4184,
+      "step": 4364
+    },
+    {
+      "epoch": 0.3492,
+      "grad_norm": 1.6618815660476685,
+      "learning_rate": 7.2884222729593765e-06,
+      "loss": 0.3217,
+      "step": 4365
+    },
+    {
+      "epoch": 0.34928,
+      "grad_norm": 1.307723045349121,
+      "learning_rate": 7.287304011552524e-06,
+      "loss": 0.3596,
+      "step": 4366
+    },
+    {
+      "epoch": 0.34936,
+      "grad_norm": 1.5229804515838623,
+      "learning_rate": 7.2861856054356e-06,
+      "loss": 0.3199,
+      "step": 4367
+    },
+    {
+      "epoch": 0.34944,
+      "grad_norm": 1.6636910438537598,
+      "learning_rate": 7.285067054679362e-06,
+      "loss": 0.2866,
+      "step": 4368
+    },
+    {
+      "epoch": 0.34952,
+      "grad_norm": 2.0462241172790527,
+      "learning_rate": 7.283948359354578e-06,
+      "loss": 0.4956,
+      "step": 4369
+    },
+    {
+      "epoch": 0.3496,
+      "grad_norm": 1.820526361465454,
+      "learning_rate": 7.282829519532022e-06,
+      "loss": 0.369,
+      "step": 4370
+    },
+    {
+      "epoch": 0.34968,
+      "grad_norm": 1.3398878574371338,
+      "learning_rate": 7.281710535282482e-06,
+      "loss": 0.3695,
+      "step": 4371
+    },
+    {
+      "epoch": 0.34976,
+      "grad_norm": 1.5608426332473755,
+      "learning_rate": 7.280591406676751e-06,
+      "loss": 0.34,
+      "step": 4372
+    },
+    {
+      "epoch": 0.34984,
+      "grad_norm": 1.7027915716171265,
+      "learning_rate": 7.279472133785633e-06,
+      "loss": 0.3301,
+      "step": 4373
+    },
+    {
+      "epoch": 0.34992,
+      "grad_norm": 1.7262037992477417,
+      "learning_rate": 7.278352716679939e-06,
+      "loss": 0.3066,
+      "step": 4374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5613858699798584,
+      "learning_rate": 7.277233155430492e-06,
+      "loss": 0.3567,
+      "step": 4375
+    },
+    {
+      "epoch": 0.35008,
+      "grad_norm": 1.6144390106201172,
+      "learning_rate": 7.2761134501081246e-06,
+      "loss": 0.4078,
+      "step": 4376
+    },
+    {
+      "epoch": 0.35016,
+      "grad_norm": 1.7558294534683228,
+      "learning_rate": 7.274993600783673e-06,
+      "loss": 0.4121,
+      "step": 4377
+    },
+    {
+      "epoch": 0.35024,
+      "grad_norm": 1.5915411710739136,
+      "learning_rate": 7.27387360752799e-06,
+      "loss": 0.3196,
+      "step": 4378
+    },
+    {
+      "epoch": 0.35032,
+      "grad_norm": 2.133772850036621,
+      "learning_rate": 7.272753470411931e-06,
+      "loss": 0.4273,
+      "step": 4379
+    },
+    {
+      "epoch": 0.3504,
+      "grad_norm": 1.4743467569351196,
+      "learning_rate": 7.271633189506366e-06,
+      "loss": 0.2437,
+      "step": 4380
+    },
+    {
+      "epoch": 0.35048,
+      "grad_norm": 1.521664023399353,
+      "learning_rate": 7.270512764882168e-06,
+      "loss": 0.3741,
+      "step": 4381
+    },
+    {
+      "epoch": 0.35056,
+      "grad_norm": 1.5323858261108398,
+      "learning_rate": 7.269392196610226e-06,
+      "loss": 0.3979,
+      "step": 4382
+    },
+    {
+      "epoch": 0.35064,
+      "grad_norm": 1.616807222366333,
+      "learning_rate": 7.268271484761433e-06,
+      "loss": 0.3299,
+      "step": 4383
+    },
+    {
+      "epoch": 0.35072,
+      "grad_norm": 1.7666263580322266,
+      "learning_rate": 7.267150629406694e-06,
+      "loss": 0.4004,
+      "step": 4384
+    },
+    {
+      "epoch": 0.3508,
+      "grad_norm": 1.719724178314209,
+      "learning_rate": 7.266029630616918e-06,
+      "loss": 0.4081,
+      "step": 4385
+    },
+    {
+      "epoch": 0.35088,
+      "grad_norm": 1.9632675647735596,
+      "learning_rate": 7.2649084884630305e-06,
+      "loss": 0.3993,
+      "step": 4386
+    },
+    {
+      "epoch": 0.35096,
+      "grad_norm": 1.2466901540756226,
+      "learning_rate": 7.2637872030159616e-06,
+      "loss": 0.2621,
+      "step": 4387
+    },
+    {
+      "epoch": 0.35104,
+      "grad_norm": 1.5172184705734253,
+      "learning_rate": 7.262665774346651e-06,
+      "loss": 0.3656,
+      "step": 4388
+    },
+    {
+      "epoch": 0.35112,
+      "grad_norm": 1.544595718383789,
+      "learning_rate": 7.261544202526047e-06,
+      "loss": 0.4009,
+      "step": 4389
+    },
+    {
+      "epoch": 0.3512,
+      "grad_norm": 1.762475609779358,
+      "learning_rate": 7.260422487625109e-06,
+      "loss": 0.4339,
+      "step": 4390
+    },
+    {
+      "epoch": 0.35128,
+      "grad_norm": 1.3777724504470825,
+      "learning_rate": 7.259300629714805e-06,
+      "loss": 0.3057,
+      "step": 4391
+    },
+    {
+      "epoch": 0.35136,
+      "grad_norm": 1.9480410814285278,
+      "learning_rate": 7.258178628866108e-06,
+      "loss": 0.493,
+      "step": 4392
+    },
+    {
+      "epoch": 0.35144,
+      "grad_norm": 1.656561255455017,
+      "learning_rate": 7.257056485150004e-06,
+      "loss": 0.5685,
+      "step": 4393
+    },
+    {
+      "epoch": 0.35152,
+      "grad_norm": 1.3406410217285156,
+      "learning_rate": 7.255934198637492e-06,
+      "loss": 0.321,
+      "step": 4394
+    },
+    {
+      "epoch": 0.3516,
+      "grad_norm": 1.745924472808838,
+      "learning_rate": 7.254811769399569e-06,
+      "loss": 0.4519,
+      "step": 4395
+    },
+    {
+      "epoch": 0.35168,
+      "grad_norm": 1.6247973442077637,
+      "learning_rate": 7.253689197507252e-06,
+      "loss": 0.3926,
+      "step": 4396
+    },
+    {
+      "epoch": 0.35176,
+      "grad_norm": 1.585526943206787,
+      "learning_rate": 7.252566483031558e-06,
+      "loss": 0.32,
+      "step": 4397
+    },
+    {
+      "epoch": 0.35184,
+      "grad_norm": 1.8269659280776978,
+      "learning_rate": 7.251443626043521e-06,
+      "loss": 0.442,
+      "step": 4398
+    },
+    {
+      "epoch": 0.35192,
+      "grad_norm": 2.0274364948272705,
+      "learning_rate": 7.250320626614178e-06,
+      "loss": 0.3676,
+      "step": 4399
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 1.7257592678070068,
+      "learning_rate": 7.249197484814579e-06,
+      "loss": 0.4199,
+      "step": 4400
+    },
+    {
+      "epoch": 0.35208,
+      "grad_norm": 2.1232810020446777,
+      "learning_rate": 7.2480742007157815e-06,
+      "loss": 0.3919,
+      "step": 4401
+    },
+    {
+      "epoch": 0.35216,
+      "grad_norm": 1.9855226278305054,
+      "learning_rate": 7.246950774388851e-06,
+      "loss": 0.5002,
+      "step": 4402
+    },
+    {
+      "epoch": 0.35224,
+      "grad_norm": 1.150065541267395,
+      "learning_rate": 7.245827205904864e-06,
+      "loss": 0.2116,
+      "step": 4403
+    },
+    {
+      "epoch": 0.35232,
+      "grad_norm": 1.5630650520324707,
+      "learning_rate": 7.244703495334904e-06,
+      "loss": 0.2761,
+      "step": 4404
+    },
+    {
+      "epoch": 0.3524,
+      "grad_norm": 1.6066936254501343,
+      "learning_rate": 7.243579642750064e-06,
+      "loss": 0.3575,
+      "step": 4405
+    },
+    {
+      "epoch": 0.35248,
+      "grad_norm": 1.361370325088501,
+      "learning_rate": 7.242455648221447e-06,
+      "loss": 0.2968,
+      "step": 4406
+    },
+    {
+      "epoch": 0.35256,
+      "grad_norm": 1.8183329105377197,
+      "learning_rate": 7.241331511820165e-06,
+      "loss": 0.3925,
+      "step": 4407
+    },
+    {
+      "epoch": 0.35264,
+      "grad_norm": 1.5168720483779907,
+      "learning_rate": 7.240207233617338e-06,
+      "loss": 0.3869,
+      "step": 4408
+    },
+    {
+      "epoch": 0.35272,
+      "grad_norm": 1.3194420337677002,
+      "learning_rate": 7.239082813684095e-06,
+      "loss": 0.2716,
+      "step": 4409
+    },
+    {
+      "epoch": 0.3528,
+      "grad_norm": 1.8594125509262085,
+      "learning_rate": 7.237958252091573e-06,
+      "loss": 0.4105,
+      "step": 4410
+    },
+    {
+      "epoch": 0.35288,
+      "grad_norm": 1.588093638420105,
+      "learning_rate": 7.236833548910922e-06,
+      "loss": 0.3534,
+      "step": 4411
+    },
+    {
+      "epoch": 0.35296,
+      "grad_norm": 1.6373640298843384,
+      "learning_rate": 7.235708704213297e-06,
+      "loss": 0.3045,
+      "step": 4412
+    },
+    {
+      "epoch": 0.35304,
+      "grad_norm": 2.0101304054260254,
+      "learning_rate": 7.234583718069862e-06,
+      "loss": 0.4869,
+      "step": 4413
+    },
+    {
+      "epoch": 0.35312,
+      "grad_norm": 1.357629656791687,
+      "learning_rate": 7.233458590551793e-06,
+      "loss": 0.2507,
+      "step": 4414
+    },
+    {
+      "epoch": 0.3532,
+      "grad_norm": 1.851575255393982,
+      "learning_rate": 7.232333321730271e-06,
+      "loss": 0.3451,
+      "step": 4415
+    },
+    {
+      "epoch": 0.35328,
+      "grad_norm": 1.6210441589355469,
+      "learning_rate": 7.2312079116764895e-06,
+      "loss": 0.3416,
+      "step": 4416
+    },
+    {
+      "epoch": 0.35336,
+      "grad_norm": 1.2965114116668701,
+      "learning_rate": 7.230082360461651e-06,
+      "loss": 0.2877,
+      "step": 4417
+    },
+    {
+      "epoch": 0.35344,
+      "grad_norm": 1.6795660257339478,
+      "learning_rate": 7.228956668156961e-06,
+      "loss": 0.4307,
+      "step": 4418
+    },
+    {
+      "epoch": 0.35352,
+      "grad_norm": 1.4039353132247925,
+      "learning_rate": 7.2278308348336425e-06,
+      "loss": 0.3429,
+      "step": 4419
+    },
+    {
+      "epoch": 0.3536,
+      "grad_norm": 1.2655251026153564,
+      "learning_rate": 7.226704860562921e-06,
+      "loss": 0.3212,
+      "step": 4420
+    },
+    {
+      "epoch": 0.35368,
+      "grad_norm": 1.6746938228607178,
+      "learning_rate": 7.225578745416033e-06,
+      "loss": 0.3959,
+      "step": 4421
+    },
+    {
+      "epoch": 0.35376,
+      "grad_norm": 2.1023738384246826,
+      "learning_rate": 7.224452489464224e-06,
+      "loss": 0.3883,
+      "step": 4422
+    },
+    {
+      "epoch": 0.35384,
+      "grad_norm": 1.2368510961532593,
+      "learning_rate": 7.2233260927787495e-06,
+      "loss": 0.241,
+      "step": 4423
+    },
+    {
+      "epoch": 0.35392,
+      "grad_norm": 1.6504780054092407,
+      "learning_rate": 7.222199555430872e-06,
+      "loss": 0.5443,
+      "step": 4424
+    },
+    {
+      "epoch": 0.354,
+      "grad_norm": 1.4802947044372559,
+      "learning_rate": 7.221072877491866e-06,
+      "loss": 0.3644,
+      "step": 4425
+    },
+    {
+      "epoch": 0.35408,
+      "grad_norm": 1.553727626800537,
+      "learning_rate": 7.219946059033009e-06,
+      "loss": 0.3166,
+      "step": 4426
+    },
+    {
+      "epoch": 0.35416,
+      "grad_norm": 1.4172799587249756,
+      "learning_rate": 7.2188191001255935e-06,
+      "loss": 0.3096,
+      "step": 4427
+    },
+    {
+      "epoch": 0.35424,
+      "grad_norm": 1.5616695880889893,
+      "learning_rate": 7.2176920008409175e-06,
+      "loss": 0.3423,
+      "step": 4428
+    },
+    {
+      "epoch": 0.35432,
+      "grad_norm": 1.895857810974121,
+      "learning_rate": 7.216564761250289e-06,
+      "loss": 0.3599,
+      "step": 4429
+    },
+    {
+      "epoch": 0.3544,
+      "grad_norm": 1.2682089805603027,
+      "learning_rate": 7.2154373814250246e-06,
+      "loss": 0.2516,
+      "step": 4430
+    },
+    {
+      "epoch": 0.35448,
+      "grad_norm": 1.800615906715393,
+      "learning_rate": 7.2143098614364504e-06,
+      "loss": 0.5255,
+      "step": 4431
+    },
+    {
+      "epoch": 0.35456,
+      "grad_norm": 1.2873491048812866,
+      "learning_rate": 7.2131822013559e-06,
+      "loss": 0.2427,
+      "step": 4432
+    },
+    {
+      "epoch": 0.35464,
+      "grad_norm": 1.6039749383926392,
+      "learning_rate": 7.212054401254718e-06,
+      "loss": 0.3177,
+      "step": 4433
+    },
+    {
+      "epoch": 0.35472,
+      "grad_norm": 1.0338603258132935,
+      "learning_rate": 7.210926461204254e-06,
+      "loss": 0.2171,
+      "step": 4434
+    },
+    {
+      "epoch": 0.3548,
+      "grad_norm": 1.7069857120513916,
+      "learning_rate": 7.209798381275871e-06,
+      "loss": 0.3947,
+      "step": 4435
+    },
+    {
+      "epoch": 0.35488,
+      "grad_norm": 1.2922322750091553,
+      "learning_rate": 7.208670161540938e-06,
+      "loss": 0.2809,
+      "step": 4436
+    },
+    {
+      "epoch": 0.35496,
+      "grad_norm": 1.5950067043304443,
+      "learning_rate": 7.207541802070836e-06,
+      "loss": 0.3408,
+      "step": 4437
+    },
+    {
+      "epoch": 0.35504,
+      "grad_norm": 2.009056806564331,
+      "learning_rate": 7.206413302936948e-06,
+      "loss": 0.423,
+      "step": 4438
+    },
+    {
+      "epoch": 0.35512,
+      "grad_norm": 1.1245336532592773,
+      "learning_rate": 7.2052846642106754e-06,
+      "loss": 0.2455,
+      "step": 4439
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 1.768202781677246,
+      "learning_rate": 7.204155885963421e-06,
+      "loss": 0.3644,
+      "step": 4440
+    },
+    {
+      "epoch": 0.35528,
+      "grad_norm": 1.4672619104385376,
+      "learning_rate": 7.203026968266598e-06,
+      "loss": 0.4002,
+      "step": 4441
+    },
+    {
+      "epoch": 0.35536,
+      "grad_norm": 1.898700475692749,
+      "learning_rate": 7.201897911191629e-06,
+      "loss": 0.6915,
+      "step": 4442
+    },
+    {
+      "epoch": 0.35544,
+      "grad_norm": 1.4047620296478271,
+      "learning_rate": 7.200768714809949e-06,
+      "loss": 0.2622,
+      "step": 4443
+    },
+    {
+      "epoch": 0.35552,
+      "grad_norm": 1.597065806388855,
+      "learning_rate": 7.199639379192994e-06,
+      "loss": 0.3354,
+      "step": 4444
+    },
+    {
+      "epoch": 0.3556,
+      "grad_norm": 1.5272079706192017,
+      "learning_rate": 7.198509904412216e-06,
+      "loss": 0.3772,
+      "step": 4445
+    },
+    {
+      "epoch": 0.35568,
+      "grad_norm": 1.6030325889587402,
+      "learning_rate": 7.197380290539073e-06,
+      "loss": 0.4348,
+      "step": 4446
+    },
+    {
+      "epoch": 0.35576,
+      "grad_norm": 1.7476811408996582,
+      "learning_rate": 7.1962505376450305e-06,
+      "loss": 0.3492,
+      "step": 4447
+    },
+    {
+      "epoch": 0.35584,
+      "grad_norm": 1.6694729328155518,
+      "learning_rate": 7.195120645801567e-06,
+      "loss": 0.2928,
+      "step": 4448
+    },
+    {
+      "epoch": 0.35592,
+      "grad_norm": 1.4422017335891724,
+      "learning_rate": 7.193990615080165e-06,
+      "loss": 0.3645,
+      "step": 4449
+    },
+    {
+      "epoch": 0.356,
+      "grad_norm": 1.421806812286377,
+      "learning_rate": 7.192860445552317e-06,
+      "loss": 0.3291,
+      "step": 4450
+    },
+    {
+      "epoch": 0.35608,
+      "grad_norm": 1.5557352304458618,
+      "learning_rate": 7.1917301372895265e-06,
+      "loss": 0.321,
+      "step": 4451
+    },
+    {
+      "epoch": 0.35616,
+      "grad_norm": 1.34674870967865,
+      "learning_rate": 7.190599690363303e-06,
+      "loss": 0.3187,
+      "step": 4452
+    },
+    {
+      "epoch": 0.35624,
+      "grad_norm": 1.7821531295776367,
+      "learning_rate": 7.189469104845167e-06,
+      "loss": 0.3548,
+      "step": 4453
+    },
+    {
+      "epoch": 0.35632,
+      "grad_norm": 1.5039072036743164,
+      "learning_rate": 7.1883383808066474e-06,
+      "loss": 0.3209,
+      "step": 4454
+    },
+    {
+      "epoch": 0.3564,
+      "grad_norm": 1.5421476364135742,
+      "learning_rate": 7.187207518319281e-06,
+      "loss": 0.3517,
+      "step": 4455
+    },
+    {
+      "epoch": 0.35648,
+      "grad_norm": 1.7966395616531372,
+      "learning_rate": 7.186076517454612e-06,
+      "loss": 0.5422,
+      "step": 4456
+    },
+    {
+      "epoch": 0.35656,
+      "grad_norm": 1.0441017150878906,
+      "learning_rate": 7.184945378284196e-06,
+      "loss": 0.2174,
+      "step": 4457
+    },
+    {
+      "epoch": 0.35664,
+      "grad_norm": 1.6243857145309448,
+      "learning_rate": 7.1838141008795985e-06,
+      "loss": 0.3314,
+      "step": 4458
+    },
+    {
+      "epoch": 0.35672,
+      "grad_norm": 1.5501540899276733,
+      "learning_rate": 7.182682685312389e-06,
+      "loss": 0.3994,
+      "step": 4459
+    },
+    {
+      "epoch": 0.3568,
+      "grad_norm": 1.3322809934616089,
+      "learning_rate": 7.181551131654149e-06,
+      "loss": 0.2615,
+      "step": 4460
+    },
+    {
+      "epoch": 0.35688,
+      "grad_norm": 1.8182624578475952,
+      "learning_rate": 7.1804194399764695e-06,
+      "loss": 0.4124,
+      "step": 4461
+    },
+    {
+      "epoch": 0.35696,
+      "grad_norm": 1.7917581796646118,
+      "learning_rate": 7.179287610350947e-06,
+      "loss": 0.3101,
+      "step": 4462
+    },
+    {
+      "epoch": 0.35704,
+      "grad_norm": 1.4548834562301636,
+      "learning_rate": 7.1781556428491895e-06,
+      "loss": 0.3088,
+      "step": 4463
+    },
+    {
+      "epoch": 0.35712,
+      "grad_norm": 1.8527675867080688,
+      "learning_rate": 7.177023537542812e-06,
+      "loss": 0.4651,
+      "step": 4464
+    },
+    {
+      "epoch": 0.3572,
+      "grad_norm": 2.1842334270477295,
+      "learning_rate": 7.17589129450344e-06,
+      "loss": 0.4734,
+      "step": 4465
+    },
+    {
+      "epoch": 0.35728,
+      "grad_norm": 1.522463321685791,
+      "learning_rate": 7.174758913802707e-06,
+      "loss": 0.3419,
+      "step": 4466
+    },
+    {
+      "epoch": 0.35736,
+      "grad_norm": 1.5129618644714355,
+      "learning_rate": 7.173626395512253e-06,
+      "loss": 0.3018,
+      "step": 4467
+    },
+    {
+      "epoch": 0.35744,
+      "grad_norm": 1.5700559616088867,
+      "learning_rate": 7.172493739703731e-06,
+      "loss": 0.3354,
+      "step": 4468
+    },
+    {
+      "epoch": 0.35752,
+      "grad_norm": 1.571109414100647,
+      "learning_rate": 7.171360946448799e-06,
+      "loss": 0.4211,
+      "step": 4469
+    },
+    {
+      "epoch": 0.3576,
+      "grad_norm": 1.8297759294509888,
+      "learning_rate": 7.170228015819125e-06,
+      "loss": 0.4264,
+      "step": 4470
+    },
+    {
+      "epoch": 0.35768,
+      "grad_norm": 1.6647297143936157,
+      "learning_rate": 7.169094947886386e-06,
+      "loss": 0.3417,
+      "step": 4471
+    },
+    {
+      "epoch": 0.35776,
+      "grad_norm": 1.950027585029602,
+      "learning_rate": 7.167961742722268e-06,
+      "loss": 0.3798,
+      "step": 4472
+    },
+    {
+      "epoch": 0.35784,
+      "grad_norm": 1.2567858695983887,
+      "learning_rate": 7.166828400398465e-06,
+      "loss": 0.2779,
+      "step": 4473
+    },
+    {
+      "epoch": 0.35792,
+      "grad_norm": 1.0137466192245483,
+      "learning_rate": 7.165694920986679e-06,
+      "loss": 0.2356,
+      "step": 4474
+    },
+    {
+      "epoch": 0.358,
+      "grad_norm": 1.2423343658447266,
+      "learning_rate": 7.16456130455862e-06,
+      "loss": 0.2513,
+      "step": 4475
+    },
+    {
+      "epoch": 0.35808,
+      "grad_norm": 1.8165005445480347,
+      "learning_rate": 7.163427551186012e-06,
+      "loss": 0.3987,
+      "step": 4476
+    },
+    {
+      "epoch": 0.35816,
+      "grad_norm": 1.666602373123169,
+      "learning_rate": 7.1622936609405804e-06,
+      "loss": 0.3845,
+      "step": 4477
+    },
+    {
+      "epoch": 0.35824,
+      "grad_norm": 1.850458025932312,
+      "learning_rate": 7.161159633894065e-06,
+      "loss": 0.3629,
+      "step": 4478
+    },
+    {
+      "epoch": 0.35832,
+      "grad_norm": 1.5257556438446045,
+      "learning_rate": 7.1600254701182106e-06,
+      "loss": 0.4075,
+      "step": 4479
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 2.071976661682129,
+      "learning_rate": 7.158891169684772e-06,
+      "loss": 0.394,
+      "step": 4480
+    },
+    {
+      "epoch": 0.35848,
+      "grad_norm": 1.8309593200683594,
+      "learning_rate": 7.157756732665512e-06,
+      "loss": 0.3949,
+      "step": 4481
+    },
+    {
+      "epoch": 0.35856,
+      "grad_norm": 1.468177080154419,
+      "learning_rate": 7.156622159132204e-06,
+      "loss": 0.3528,
+      "step": 4482
+    },
+    {
+      "epoch": 0.35864,
+      "grad_norm": 1.4701149463653564,
+      "learning_rate": 7.1554874491566274e-06,
+      "loss": 0.334,
+      "step": 4483
+    },
+    {
+      "epoch": 0.35872,
+      "grad_norm": 1.1661218404769897,
+      "learning_rate": 7.1543526028105735e-06,
+      "loss": 0.2672,
+      "step": 4484
+    },
+    {
+      "epoch": 0.3588,
+      "grad_norm": 1.2732799053192139,
+      "learning_rate": 7.153217620165838e-06,
+      "loss": 0.2915,
+      "step": 4485
+    },
+    {
+      "epoch": 0.35888,
+      "grad_norm": 1.7382800579071045,
+      "learning_rate": 7.15208250129423e-06,
+      "loss": 0.296,
+      "step": 4486
+    },
+    {
+      "epoch": 0.35896,
+      "grad_norm": 1.537358283996582,
+      "learning_rate": 7.150947246267561e-06,
+      "loss": 0.3903,
+      "step": 4487
+    },
+    {
+      "epoch": 0.35904,
+      "grad_norm": 1.4270464181900024,
+      "learning_rate": 7.1498118551576574e-06,
+      "loss": 0.3617,
+      "step": 4488
+    },
+    {
+      "epoch": 0.35912,
+      "grad_norm": 1.366553544998169,
+      "learning_rate": 7.148676328036352e-06,
+      "loss": 0.3077,
+      "step": 4489
+    },
+    {
+      "epoch": 0.3592,
+      "grad_norm": 1.4373116493225098,
+      "learning_rate": 7.1475406649754845e-06,
+      "loss": 0.3871,
+      "step": 4490
+    },
+    {
+      "epoch": 0.35928,
+      "grad_norm": 1.6480387449264526,
+      "learning_rate": 7.146404866046903e-06,
+      "loss": 0.3344,
+      "step": 4491
+    },
+    {
+      "epoch": 0.35936,
+      "grad_norm": 1.5638269186019897,
+      "learning_rate": 7.145268931322469e-06,
+      "loss": 0.301,
+      "step": 4492
+    },
+    {
+      "epoch": 0.35944,
+      "grad_norm": 1.2544904947280884,
+      "learning_rate": 7.144132860874047e-06,
+      "loss": 0.2363,
+      "step": 4493
+    },
+    {
+      "epoch": 0.35952,
+      "grad_norm": 1.566571593284607,
+      "learning_rate": 7.142996654773514e-06,
+      "loss": 0.3634,
+      "step": 4494
+    },
+    {
+      "epoch": 0.3596,
+      "grad_norm": 2.4747540950775146,
+      "learning_rate": 7.141860313092753e-06,
+      "loss": 0.5304,
+      "step": 4495
+    },
+    {
+      "epoch": 0.35968,
+      "grad_norm": 1.7263084650039673,
+      "learning_rate": 7.1407238359036565e-06,
+      "loss": 0.3384,
+      "step": 4496
+    },
+    {
+      "epoch": 0.35976,
+      "grad_norm": 1.6196907758712769,
+      "learning_rate": 7.139587223278127e-06,
+      "loss": 0.3003,
+      "step": 4497
+    },
+    {
+      "epoch": 0.35984,
+      "grad_norm": 0.9107118844985962,
+      "learning_rate": 7.138450475288072e-06,
+      "loss": 0.163,
+      "step": 4498
+    },
+    {
+      "epoch": 0.35992,
+      "grad_norm": 1.3455191850662231,
+      "learning_rate": 7.1373135920054104e-06,
+      "loss": 0.308,
+      "step": 4499
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6801081895828247,
+      "learning_rate": 7.1361765735020695e-06,
+      "loss": 0.3765,
+      "step": 4500
+    },
+    {
+      "epoch": 0.36008,
+      "grad_norm": 1.4999953508377075,
+      "learning_rate": 7.135039419849984e-06,
+      "loss": 0.2996,
+      "step": 4501
+    },
+    {
+      "epoch": 0.36016,
+      "grad_norm": 1.4781824350357056,
+      "learning_rate": 7.1339021311211e-06,
+      "loss": 0.3048,
+      "step": 4502
+    },
+    {
+      "epoch": 0.36024,
+      "grad_norm": 1.7782272100448608,
+      "learning_rate": 7.1327647073873665e-06,
+      "loss": 0.4415,
+      "step": 4503
+    },
+    {
+      "epoch": 0.36032,
+      "grad_norm": 1.6102408170700073,
+      "learning_rate": 7.131627148720746e-06,
+      "loss": 0.3111,
+      "step": 4504
+    },
+    {
+      "epoch": 0.3604,
+      "grad_norm": 1.8156652450561523,
+      "learning_rate": 7.130489455193208e-06,
+      "loss": 0.2951,
+      "step": 4505
+    },
+    {
+      "epoch": 0.36048,
+      "grad_norm": 1.3658769130706787,
+      "learning_rate": 7.129351626876733e-06,
+      "loss": 0.348,
+      "step": 4506
+    },
+    {
+      "epoch": 0.36056,
+      "grad_norm": 1.5961403846740723,
+      "learning_rate": 7.128213663843304e-06,
+      "loss": 0.3197,
+      "step": 4507
+    },
+    {
+      "epoch": 0.36064,
+      "grad_norm": 1.7644686698913574,
+      "learning_rate": 7.127075566164919e-06,
+      "loss": 0.3424,
+      "step": 4508
+    },
+    {
+      "epoch": 0.36072,
+      "grad_norm": 1.4091618061065674,
+      "learning_rate": 7.125937333913577e-06,
+      "loss": 0.3399,
+      "step": 4509
+    },
+    {
+      "epoch": 0.3608,
+      "grad_norm": 2.0317022800445557,
+      "learning_rate": 7.124798967161296e-06,
+      "loss": 0.4621,
+      "step": 4510
+    },
+    {
+      "epoch": 0.36088,
+      "grad_norm": 2.1083295345306396,
+      "learning_rate": 7.123660465980093e-06,
+      "loss": 0.4287,
+      "step": 4511
+    },
+    {
+      "epoch": 0.36096,
+      "grad_norm": 2.015735626220703,
+      "learning_rate": 7.122521830441998e-06,
+      "loss": 0.4459,
+      "step": 4512
+    },
+    {
+      "epoch": 0.36104,
+      "grad_norm": 1.581146001815796,
+      "learning_rate": 7.121383060619048e-06,
+      "loss": 0.3427,
+      "step": 4513
+    },
+    {
+      "epoch": 0.36112,
+      "grad_norm": 1.9534862041473389,
+      "learning_rate": 7.120244156583291e-06,
+      "loss": 0.4336,
+      "step": 4514
+    },
+    {
+      "epoch": 0.3612,
+      "grad_norm": 1.0503820180892944,
+      "learning_rate": 7.11910511840678e-06,
+      "loss": 0.2304,
+      "step": 4515
+    },
+    {
+      "epoch": 0.36128,
+      "grad_norm": 1.4276210069656372,
+      "learning_rate": 7.11796594616158e-06,
+      "loss": 0.3487,
+      "step": 4516
+    },
+    {
+      "epoch": 0.36136,
+      "grad_norm": 1.9091681241989136,
+      "learning_rate": 7.116826639919761e-06,
+      "loss": 0.4407,
+      "step": 4517
+    },
+    {
+      "epoch": 0.36144,
+      "grad_norm": 1.8984267711639404,
+      "learning_rate": 7.115687199753403e-06,
+      "loss": 0.4895,
+      "step": 4518
+    },
+    {
+      "epoch": 0.36152,
+      "grad_norm": 1.8324373960494995,
+      "learning_rate": 7.114547625734593e-06,
+      "loss": 0.384,
+      "step": 4519
+    },
+    {
+      "epoch": 0.3616,
+      "grad_norm": 1.296120524406433,
+      "learning_rate": 7.113407917935433e-06,
+      "loss": 0.2754,
+      "step": 4520
+    },
+    {
+      "epoch": 0.36168,
+      "grad_norm": 1.7863622903823853,
+      "learning_rate": 7.112268076428025e-06,
+      "loss": 0.3302,
+      "step": 4521
+    },
+    {
+      "epoch": 0.36176,
+      "grad_norm": 2.1770036220550537,
+      "learning_rate": 7.1111281012844825e-06,
+      "loss": 0.3736,
+      "step": 4522
+    },
+    {
+      "epoch": 0.36184,
+      "grad_norm": 1.410852074623108,
+      "learning_rate": 7.109987992576929e-06,
+      "loss": 0.3158,
+      "step": 4523
+    },
+    {
+      "epoch": 0.36192,
+      "grad_norm": 1.6189733743667603,
+      "learning_rate": 7.108847750377494e-06,
+      "loss": 0.4493,
+      "step": 4524
+    },
+    {
+      "epoch": 0.362,
+      "grad_norm": 1.5762578248977661,
+      "learning_rate": 7.107707374758321e-06,
+      "loss": 0.3384,
+      "step": 4525
+    },
+    {
+      "epoch": 0.36208,
+      "grad_norm": 1.3969612121582031,
+      "learning_rate": 7.106566865791553e-06,
+      "loss": 0.3198,
+      "step": 4526
+    },
+    {
+      "epoch": 0.36216,
+      "grad_norm": 1.4169987440109253,
+      "learning_rate": 7.105426223549349e-06,
+      "loss": 0.3192,
+      "step": 4527
+    },
+    {
+      "epoch": 0.36224,
+      "grad_norm": 1.601337194442749,
+      "learning_rate": 7.104285448103871e-06,
+      "loss": 0.4275,
+      "step": 4528
+    },
+    {
+      "epoch": 0.36232,
+      "grad_norm": 1.4484046697616577,
+      "learning_rate": 7.103144539527295e-06,
+      "loss": 0.2617,
+      "step": 4529
+    },
+    {
+      "epoch": 0.3624,
+      "grad_norm": 1.1574383974075317,
+      "learning_rate": 7.1020034978918006e-06,
+      "loss": 0.248,
+      "step": 4530
+    },
+    {
+      "epoch": 0.36248,
+      "grad_norm": 1.2772170305252075,
+      "learning_rate": 7.100862323269579e-06,
+      "loss": 0.2806,
+      "step": 4531
+    },
+    {
+      "epoch": 0.36256,
+      "grad_norm": 1.5458264350891113,
+      "learning_rate": 7.099721015732828e-06,
+      "loss": 0.4107,
+      "step": 4532
+    },
+    {
+      "epoch": 0.36264,
+      "grad_norm": 1.3658666610717773,
+      "learning_rate": 7.0985795753537525e-06,
+      "loss": 0.3861,
+      "step": 4533
+    },
+    {
+      "epoch": 0.36272,
+      "grad_norm": 1.4563086032867432,
+      "learning_rate": 7.097438002204568e-06,
+      "loss": 0.3225,
+      "step": 4534
+    },
+    {
+      "epoch": 0.3628,
+      "grad_norm": 1.7949572801589966,
+      "learning_rate": 7.096296296357502e-06,
+      "loss": 0.5492,
+      "step": 4535
+    },
+    {
+      "epoch": 0.36288,
+      "grad_norm": 1.9033316373825073,
+      "learning_rate": 7.095154457884782e-06,
+      "loss": 0.3523,
+      "step": 4536
+    },
+    {
+      "epoch": 0.36296,
+      "grad_norm": 1.7623441219329834,
+      "learning_rate": 7.094012486858652e-06,
+      "loss": 0.4402,
+      "step": 4537
+    },
+    {
+      "epoch": 0.36304,
+      "grad_norm": 1.6129101514816284,
+      "learning_rate": 7.092870383351355e-06,
+      "loss": 0.4225,
+      "step": 4538
+    },
+    {
+      "epoch": 0.36312,
+      "grad_norm": 1.721638798713684,
+      "learning_rate": 7.091728147435154e-06,
+      "loss": 0.3564,
+      "step": 4539
+    },
+    {
+      "epoch": 0.3632,
+      "grad_norm": 1.588778018951416,
+      "learning_rate": 7.090585779182311e-06,
+      "loss": 0.3925,
+      "step": 4540
+    },
+    {
+      "epoch": 0.36328,
+      "grad_norm": 1.32707941532135,
+      "learning_rate": 7.089443278665102e-06,
+      "loss": 0.324,
+      "step": 4541
+    },
+    {
+      "epoch": 0.36336,
+      "grad_norm": 2.0091614723205566,
+      "learning_rate": 7.088300645955808e-06,
+      "loss": 0.405,
+      "step": 4542
+    },
+    {
+      "epoch": 0.36344,
+      "grad_norm": 1.9140995740890503,
+      "learning_rate": 7.087157881126719e-06,
+      "loss": 0.4189,
+      "step": 4543
+    },
+    {
+      "epoch": 0.36352,
+      "grad_norm": 1.6873313188552856,
+      "learning_rate": 7.0860149842501345e-06,
+      "loss": 0.2981,
+      "step": 4544
+    },
+    {
+      "epoch": 0.3636,
+      "grad_norm": 1.7330468893051147,
+      "learning_rate": 7.084871955398361e-06,
+      "loss": 0.4079,
+      "step": 4545
+    },
+    {
+      "epoch": 0.36368,
+      "grad_norm": 1.319456934928894,
+      "learning_rate": 7.083728794643716e-06,
+      "loss": 0.2838,
+      "step": 4546
+    },
+    {
+      "epoch": 0.36376,
+      "grad_norm": 1.4083654880523682,
+      "learning_rate": 7.082585502058522e-06,
+      "loss": 0.2926,
+      "step": 4547
+    },
+    {
+      "epoch": 0.36384,
+      "grad_norm": 1.7500334978103638,
+      "learning_rate": 7.0814420777151115e-06,
+      "loss": 0.4215,
+      "step": 4548
+    },
+    {
+      "epoch": 0.36392,
+      "grad_norm": 2.052196979522705,
+      "learning_rate": 7.080298521685826e-06,
+      "loss": 0.4126,
+      "step": 4549
+    },
+    {
+      "epoch": 0.364,
+      "grad_norm": 1.5984829664230347,
+      "learning_rate": 7.0791548340430125e-06,
+      "loss": 0.3908,
+      "step": 4550
+    },
+    {
+      "epoch": 0.36408,
+      "grad_norm": 1.5750199556350708,
+      "learning_rate": 7.0780110148590305e-06,
+      "loss": 0.4188,
+      "step": 4551
+    },
+    {
+      "epoch": 0.36416,
+      "grad_norm": 1.4217820167541504,
+      "learning_rate": 7.076867064206244e-06,
+      "loss": 0.2855,
+      "step": 4552
+    },
+    {
+      "epoch": 0.36424,
+      "grad_norm": 1.2456365823745728,
+      "learning_rate": 7.0757229821570285e-06,
+      "loss": 0.3137,
+      "step": 4553
+    },
+    {
+      "epoch": 0.36432,
+      "grad_norm": 1.511091947555542,
+      "learning_rate": 7.074578768783764e-06,
+      "loss": 0.3978,
+      "step": 4554
+    },
+    {
+      "epoch": 0.3644,
+      "grad_norm": 1.2580100297927856,
+      "learning_rate": 7.073434424158845e-06,
+      "loss": 0.2876,
+      "step": 4555
+    },
+    {
+      "epoch": 0.36448,
+      "grad_norm": 1.751238465309143,
+      "learning_rate": 7.072289948354665e-06,
+      "loss": 0.3838,
+      "step": 4556
+    },
+    {
+      "epoch": 0.36456,
+      "grad_norm": 1.7836511135101318,
+      "learning_rate": 7.071145341443635e-06,
+      "loss": 0.418,
+      "step": 4557
+    },
+    {
+      "epoch": 0.36464,
+      "grad_norm": 1.676069974899292,
+      "learning_rate": 7.070000603498169e-06,
+      "loss": 0.4017,
+      "step": 4558
+    },
+    {
+      "epoch": 0.36472,
+      "grad_norm": 1.334571123123169,
+      "learning_rate": 7.06885573459069e-06,
+      "loss": 0.2776,
+      "step": 4559
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 1.4708949327468872,
+      "learning_rate": 7.067710734793631e-06,
+      "loss": 0.2896,
+      "step": 4560
+    },
+    {
+      "epoch": 0.36488,
+      "grad_norm": 1.422624945640564,
+      "learning_rate": 7.066565604179435e-06,
+      "loss": 0.3168,
+      "step": 4561
+    },
+    {
+      "epoch": 0.36496,
+      "grad_norm": 1.2501122951507568,
+      "learning_rate": 7.065420342820546e-06,
+      "loss": 0.2806,
+      "step": 4562
+    },
+    {
+      "epoch": 0.36504,
+      "grad_norm": 1.5056577920913696,
+      "learning_rate": 7.064274950789424e-06,
+      "loss": 0.2764,
+      "step": 4563
+    },
+    {
+      "epoch": 0.36512,
+      "grad_norm": 1.5189259052276611,
+      "learning_rate": 7.063129428158533e-06,
+      "loss": 0.3423,
+      "step": 4564
+    },
+    {
+      "epoch": 0.3652,
+      "grad_norm": 1.5131878852844238,
+      "learning_rate": 7.061983775000345e-06,
+      "loss": 0.3619,
+      "step": 4565
+    },
+    {
+      "epoch": 0.36528,
+      "grad_norm": 1.3569074869155884,
+      "learning_rate": 7.0608379913873444e-06,
+      "loss": 0.2639,
+      "step": 4566
+    },
+    {
+      "epoch": 0.36536,
+      "grad_norm": 1.1969329118728638,
+      "learning_rate": 7.059692077392018e-06,
+      "loss": 0.2341,
+      "step": 4567
+    },
+    {
+      "epoch": 0.36544,
+      "grad_norm": 1.7154649496078491,
+      "learning_rate": 7.058546033086867e-06,
+      "loss": 0.4684,
+      "step": 4568
+    },
+    {
+      "epoch": 0.36552,
+      "grad_norm": 1.4121015071868896,
+      "learning_rate": 7.057399858544396e-06,
+      "loss": 0.3091,
+      "step": 4569
+    },
+    {
+      "epoch": 0.3656,
+      "grad_norm": 1.7438020706176758,
+      "learning_rate": 7.056253553837119e-06,
+      "loss": 0.4294,
+      "step": 4570
+    },
+    {
+      "epoch": 0.36568,
+      "grad_norm": 1.392918586730957,
+      "learning_rate": 7.0551071190375605e-06,
+      "loss": 0.3318,
+      "step": 4571
+    },
+    {
+      "epoch": 0.36576,
+      "grad_norm": 1.6059353351593018,
+      "learning_rate": 7.053960554218252e-06,
+      "loss": 0.3442,
+      "step": 4572
+    },
+    {
+      "epoch": 0.36584,
+      "grad_norm": 1.400192379951477,
+      "learning_rate": 7.052813859451731e-06,
+      "loss": 0.3341,
+      "step": 4573
+    },
+    {
+      "epoch": 0.36592,
+      "grad_norm": 1.4998594522476196,
+      "learning_rate": 7.0516670348105455e-06,
+      "loss": 0.365,
+      "step": 4574
+    },
+    {
+      "epoch": 0.366,
+      "grad_norm": 1.7953191995620728,
+      "learning_rate": 7.050520080367254e-06,
+      "loss": 0.3027,
+      "step": 4575
+    },
+    {
+      "epoch": 0.36608,
+      "grad_norm": 2.084156036376953,
+      "learning_rate": 7.049372996194415e-06,
+      "loss": 0.43,
+      "step": 4576
+    },
+    {
+      "epoch": 0.36616,
+      "grad_norm": 1.0851385593414307,
+      "learning_rate": 7.048225782364605e-06,
+      "loss": 0.244,
+      "step": 4577
+    },
+    {
+      "epoch": 0.36624,
+      "grad_norm": 1.446227788925171,
+      "learning_rate": 7.047078438950403e-06,
+      "loss": 0.3054,
+      "step": 4578
+    },
+    {
+      "epoch": 0.36632,
+      "grad_norm": 1.2726980447769165,
+      "learning_rate": 7.0459309660243976e-06,
+      "loss": 0.292,
+      "step": 4579
+    },
+    {
+      "epoch": 0.3664,
+      "grad_norm": 1.6056383848190308,
+      "learning_rate": 7.044783363659185e-06,
+      "loss": 0.3092,
+      "step": 4580
+    },
+    {
+      "epoch": 0.36648,
+      "grad_norm": 1.5209200382232666,
+      "learning_rate": 7.043635631927372e-06,
+      "loss": 0.3111,
+      "step": 4581
+    },
+    {
+      "epoch": 0.36656,
+      "grad_norm": 1.5532723665237427,
+      "learning_rate": 7.04248777090157e-06,
+      "loss": 0.339,
+      "step": 4582
+    },
+    {
+      "epoch": 0.36664,
+      "grad_norm": 2.0113229751586914,
+      "learning_rate": 7.041339780654401e-06,
+      "loss": 0.3921,
+      "step": 4583
+    },
+    {
+      "epoch": 0.36672,
+      "grad_norm": 1.3075275421142578,
+      "learning_rate": 7.040191661258495e-06,
+      "loss": 0.346,
+      "step": 4584
+    },
+    {
+      "epoch": 0.3668,
+      "grad_norm": 1.2967735528945923,
+      "learning_rate": 7.039043412786489e-06,
+      "loss": 0.3008,
+      "step": 4585
+    },
+    {
+      "epoch": 0.36688,
+      "grad_norm": 1.4599136114120483,
+      "learning_rate": 7.037895035311029e-06,
+      "loss": 0.337,
+      "step": 4586
+    },
+    {
+      "epoch": 0.36696,
+      "grad_norm": 1.6987855434417725,
+      "learning_rate": 7.0367465289047685e-06,
+      "loss": 0.3539,
+      "step": 4587
+    },
+    {
+      "epoch": 0.36704,
+      "grad_norm": 1.5327343940734863,
+      "learning_rate": 7.03559789364037e-06,
+      "loss": 0.4788,
+      "step": 4588
+    },
+    {
+      "epoch": 0.36712,
+      "grad_norm": 1.7550090551376343,
+      "learning_rate": 7.034449129590504e-06,
+      "loss": 0.3794,
+      "step": 4589
+    },
+    {
+      "epoch": 0.3672,
+      "grad_norm": 1.623400092124939,
+      "learning_rate": 7.033300236827849e-06,
+      "loss": 0.323,
+      "step": 4590
+    },
+    {
+      "epoch": 0.36728,
+      "grad_norm": 1.9743831157684326,
+      "learning_rate": 7.032151215425092e-06,
+      "loss": 0.3837,
+      "step": 4591
+    },
+    {
+      "epoch": 0.36736,
+      "grad_norm": 1.4057821035385132,
+      "learning_rate": 7.0310020654549264e-06,
+      "loss": 0.289,
+      "step": 4592
+    },
+    {
+      "epoch": 0.36744,
+      "grad_norm": 1.7358758449554443,
+      "learning_rate": 7.029852786990056e-06,
+      "loss": 0.3107,
+      "step": 4593
+    },
+    {
+      "epoch": 0.36752,
+      "grad_norm": 1.3099263906478882,
+      "learning_rate": 7.028703380103192e-06,
+      "loss": 0.2482,
+      "step": 4594
+    },
+    {
+      "epoch": 0.3676,
+      "grad_norm": 1.3868745565414429,
+      "learning_rate": 7.027553844867052e-06,
+      "loss": 0.3824,
+      "step": 4595
+    },
+    {
+      "epoch": 0.36768,
+      "grad_norm": 1.7561724185943604,
+      "learning_rate": 7.0264041813543655e-06,
+      "loss": 0.3757,
+      "step": 4596
+    },
+    {
+      "epoch": 0.36776,
+      "grad_norm": 1.4879529476165771,
+      "learning_rate": 7.0252543896378666e-06,
+      "loss": 0.3778,
+      "step": 4597
+    },
+    {
+      "epoch": 0.36784,
+      "grad_norm": 1.437186360359192,
+      "learning_rate": 7.024104469790301e-06,
+      "loss": 0.3313,
+      "step": 4598
+    },
+    {
+      "epoch": 0.36792,
+      "grad_norm": 1.691285252571106,
+      "learning_rate": 7.022954421884415e-06,
+      "loss": 0.4338,
+      "step": 4599
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 1.7200967073440552,
+      "learning_rate": 7.021804245992973e-06,
+      "loss": 0.3275,
+      "step": 4600
+    },
+    {
+      "epoch": 0.36808,
+      "grad_norm": 1.6157788038253784,
+      "learning_rate": 7.020653942188741e-06,
+      "loss": 0.4735,
+      "step": 4601
+    },
+    {
+      "epoch": 0.36816,
+      "grad_norm": 1.5696213245391846,
+      "learning_rate": 7.019503510544496e-06,
+      "loss": 0.3177,
+      "step": 4602
+    },
+    {
+      "epoch": 0.36824,
+      "grad_norm": 1.3805010318756104,
+      "learning_rate": 7.01835295113302e-06,
+      "loss": 0.2877,
+      "step": 4603
+    },
+    {
+      "epoch": 0.36832,
+      "grad_norm": 2.140430212020874,
+      "learning_rate": 7.017202264027108e-06,
+      "loss": 0.5055,
+      "step": 4604
+    },
+    {
+      "epoch": 0.3684,
+      "grad_norm": 1.4156843423843384,
+      "learning_rate": 7.016051449299556e-06,
+      "loss": 0.347,
+      "step": 4605
+    },
+    {
+      "epoch": 0.36848,
+      "grad_norm": 1.9364955425262451,
+      "learning_rate": 7.014900507023175e-06,
+      "loss": 0.3778,
+      "step": 4606
+    },
+    {
+      "epoch": 0.36856,
+      "grad_norm": 1.7083888053894043,
+      "learning_rate": 7.013749437270781e-06,
+      "loss": 0.3028,
+      "step": 4607
+    },
+    {
+      "epoch": 0.36864,
+      "grad_norm": 1.3670393228530884,
+      "learning_rate": 7.012598240115201e-06,
+      "loss": 0.2461,
+      "step": 4608
+    },
+    {
+      "epoch": 0.36872,
+      "grad_norm": 1.3490629196166992,
+      "learning_rate": 7.011446915629261e-06,
+      "loss": 0.2549,
+      "step": 4609
+    },
+    {
+      "epoch": 0.3688,
+      "grad_norm": 1.7474075555801392,
+      "learning_rate": 7.0102954638858065e-06,
+      "loss": 0.4072,
+      "step": 4610
+    },
+    {
+      "epoch": 0.36888,
+      "grad_norm": 1.9252272844314575,
+      "learning_rate": 7.009143884957684e-06,
+      "loss": 0.3497,
+      "step": 4611
+    },
+    {
+      "epoch": 0.36896,
+      "grad_norm": 1.2210733890533447,
+      "learning_rate": 7.007992178917751e-06,
+      "loss": 0.2395,
+      "step": 4612
+    },
+    {
+      "epoch": 0.36904,
+      "grad_norm": 1.4923193454742432,
+      "learning_rate": 7.0068403458388715e-06,
+      "loss": 0.3063,
+      "step": 4613
+    },
+    {
+      "epoch": 0.36912,
+      "grad_norm": 1.8423126935958862,
+      "learning_rate": 7.0056883857939174e-06,
+      "loss": 0.5195,
+      "step": 4614
+    },
+    {
+      "epoch": 0.3692,
+      "grad_norm": 1.8047568798065186,
+      "learning_rate": 7.004536298855771e-06,
+      "loss": 0.4048,
+      "step": 4615
+    },
+    {
+      "epoch": 0.36928,
+      "grad_norm": 1.594499111175537,
+      "learning_rate": 7.003384085097319e-06,
+      "loss": 0.3086,
+      "step": 4616
+    },
+    {
+      "epoch": 0.36936,
+      "grad_norm": 1.4673727750778198,
+      "learning_rate": 7.00223174459146e-06,
+      "loss": 0.2774,
+      "step": 4617
+    },
+    {
+      "epoch": 0.36944,
+      "grad_norm": 1.760596752166748,
+      "learning_rate": 7.001079277411098e-06,
+      "loss": 0.3534,
+      "step": 4618
+    },
+    {
+      "epoch": 0.36952,
+      "grad_norm": 1.7721569538116455,
+      "learning_rate": 6.9999266836291464e-06,
+      "loss": 0.3507,
+      "step": 4619
+    },
+    {
+      "epoch": 0.3696,
+      "grad_norm": 1.4745303392410278,
+      "learning_rate": 6.9987739633185245e-06,
+      "loss": 0.35,
+      "step": 4620
+    },
+    {
+      "epoch": 0.36968,
+      "grad_norm": 1.7846648693084717,
+      "learning_rate": 6.9976211165521635e-06,
+      "loss": 0.4137,
+      "step": 4621
+    },
+    {
+      "epoch": 0.36976,
+      "grad_norm": 1.6463090181350708,
+      "learning_rate": 6.996468143402997e-06,
+      "loss": 0.3458,
+      "step": 4622
+    },
+    {
+      "epoch": 0.36984,
+      "grad_norm": 1.7248306274414062,
+      "learning_rate": 6.995315043943971e-06,
+      "loss": 0.3225,
+      "step": 4623
+    },
+    {
+      "epoch": 0.36992,
+      "grad_norm": 2.039926528930664,
+      "learning_rate": 6.9941618182480384e-06,
+      "loss": 0.4484,
+      "step": 4624
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9294219017028809,
+      "learning_rate": 6.993008466388161e-06,
+      "loss": 0.407,
+      "step": 4625
+    },
+    {
+      "epoch": 0.37008,
+      "grad_norm": 1.5425759553909302,
+      "learning_rate": 6.991854988437307e-06,
+      "loss": 0.3614,
+      "step": 4626
+    },
+    {
+      "epoch": 0.37016,
+      "grad_norm": 1.3891255855560303,
+      "learning_rate": 6.990701384468451e-06,
+      "loss": 0.2851,
+      "step": 4627
+    },
+    {
+      "epoch": 0.37024,
+      "grad_norm": 2.027620792388916,
+      "learning_rate": 6.9895476545545804e-06,
+      "loss": 0.3724,
+      "step": 4628
+    },
+    {
+      "epoch": 0.37032,
+      "grad_norm": 1.5156015157699585,
+      "learning_rate": 6.988393798768685e-06,
+      "loss": 0.3557,
+      "step": 4629
+    },
+    {
+      "epoch": 0.3704,
+      "grad_norm": 1.4785630702972412,
+      "learning_rate": 6.987239817183769e-06,
+      "loss": 0.3771,
+      "step": 4630
+    },
+    {
+      "epoch": 0.37048,
+      "grad_norm": 1.4804576635360718,
+      "learning_rate": 6.98608570987284e-06,
+      "loss": 0.4071,
+      "step": 4631
+    },
+    {
+      "epoch": 0.37056,
+      "grad_norm": 1.5395426750183105,
+      "learning_rate": 6.984931476908911e-06,
+      "loss": 0.337,
+      "step": 4632
+    },
+    {
+      "epoch": 0.37064,
+      "grad_norm": 1.7918102741241455,
+      "learning_rate": 6.983777118365011e-06,
+      "loss": 0.4346,
+      "step": 4633
+    },
+    {
+      "epoch": 0.37072,
+      "grad_norm": 1.6034590005874634,
+      "learning_rate": 6.982622634314171e-06,
+      "loss": 0.3478,
+      "step": 4634
+    },
+    {
+      "epoch": 0.3708,
+      "grad_norm": 1.6067657470703125,
+      "learning_rate": 6.981468024829428e-06,
+      "loss": 0.4161,
+      "step": 4635
+    },
+    {
+      "epoch": 0.37088,
+      "grad_norm": 1.1507148742675781,
+      "learning_rate": 6.980313289983836e-06,
+      "loss": 0.2877,
+      "step": 4636
+    },
+    {
+      "epoch": 0.37096,
+      "grad_norm": 1.345964789390564,
+      "learning_rate": 6.979158429850448e-06,
+      "loss": 0.2516,
+      "step": 4637
+    },
+    {
+      "epoch": 0.37104,
+      "grad_norm": 1.6395801305770874,
+      "learning_rate": 6.978003444502326e-06,
+      "loss": 0.4643,
+      "step": 4638
+    },
+    {
+      "epoch": 0.37112,
+      "grad_norm": 1.3134948015213013,
+      "learning_rate": 6.976848334012546e-06,
+      "loss": 0.3187,
+      "step": 4639
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 1.9457036256790161,
+      "learning_rate": 6.975693098454186e-06,
+      "loss": 0.4266,
+      "step": 4640
+    },
+    {
+      "epoch": 0.37128,
+      "grad_norm": 1.6113479137420654,
+      "learning_rate": 6.974537737900336e-06,
+      "loss": 0.3117,
+      "step": 4641
+    },
+    {
+      "epoch": 0.37136,
+      "grad_norm": 1.1665359735488892,
+      "learning_rate": 6.973382252424088e-06,
+      "loss": 0.2507,
+      "step": 4642
+    },
+    {
+      "epoch": 0.37144,
+      "grad_norm": 1.7264834642410278,
+      "learning_rate": 6.9722266420985495e-06,
+      "loss": 0.4146,
+      "step": 4643
+    },
+    {
+      "epoch": 0.37152,
+      "grad_norm": 2.008244276046753,
+      "learning_rate": 6.97107090699683e-06,
+      "loss": 0.3615,
+      "step": 4644
+    },
+    {
+      "epoch": 0.3716,
+      "grad_norm": 1.2290860414505005,
+      "learning_rate": 6.969915047192049e-06,
+      "loss": 0.2695,
+      "step": 4645
+    },
+    {
+      "epoch": 0.37168,
+      "grad_norm": 1.5274767875671387,
+      "learning_rate": 6.968759062757334e-06,
+      "loss": 0.281,
+      "step": 4646
+    },
+    {
+      "epoch": 0.37176,
+      "grad_norm": 1.9756278991699219,
+      "learning_rate": 6.967602953765821e-06,
+      "loss": 0.3648,
+      "step": 4647
+    },
+    {
+      "epoch": 0.37184,
+      "grad_norm": 1.4363765716552734,
+      "learning_rate": 6.966446720290652e-06,
+      "loss": 0.3549,
+      "step": 4648
+    },
+    {
+      "epoch": 0.37192,
+      "grad_norm": 1.5073497295379639,
+      "learning_rate": 6.9652903624049804e-06,
+      "loss": 0.3297,
+      "step": 4649
+    },
+    {
+      "epoch": 0.372,
+      "grad_norm": 1.4804567098617554,
+      "learning_rate": 6.964133880181962e-06,
+      "loss": 0.3181,
+      "step": 4650
+    },
+    {
+      "epoch": 0.37208,
+      "grad_norm": 1.5861263275146484,
+      "learning_rate": 6.962977273694765e-06,
+      "loss": 0.3307,
+      "step": 4651
+    },
+    {
+      "epoch": 0.37216,
+      "grad_norm": 2.007474422454834,
+      "learning_rate": 6.961820543016565e-06,
+      "loss": 0.3815,
+      "step": 4652
+    },
+    {
+      "epoch": 0.37224,
+      "grad_norm": 1.6723278760910034,
+      "learning_rate": 6.960663688220543e-06,
+      "loss": 0.3828,
+      "step": 4653
+    },
+    {
+      "epoch": 0.37232,
+      "grad_norm": 1.481465458869934,
+      "learning_rate": 6.959506709379891e-06,
+      "loss": 0.2932,
+      "step": 4654
+    },
+    {
+      "epoch": 0.3724,
+      "grad_norm": 1.6673827171325684,
+      "learning_rate": 6.958349606567806e-06,
+      "loss": 0.3636,
+      "step": 4655
+    },
+    {
+      "epoch": 0.37248,
+      "grad_norm": 2.048527956008911,
+      "learning_rate": 6.957192379857493e-06,
+      "loss": 0.45,
+      "step": 4656
+    },
+    {
+      "epoch": 0.37256,
+      "grad_norm": 1.8853121995925903,
+      "learning_rate": 6.9560350293221655e-06,
+      "loss": 0.3628,
+      "step": 4657
+    },
+    {
+      "epoch": 0.37264,
+      "grad_norm": 1.4432258605957031,
+      "learning_rate": 6.954877555035049e-06,
+      "loss": 0.2615,
+      "step": 4658
+    },
+    {
+      "epoch": 0.37272,
+      "grad_norm": 1.38565993309021,
+      "learning_rate": 6.953719957069369e-06,
+      "loss": 0.2794,
+      "step": 4659
+    },
+    {
+      "epoch": 0.3728,
+      "grad_norm": 2.1929290294647217,
+      "learning_rate": 6.952562235498366e-06,
+      "loss": 0.5944,
+      "step": 4660
+    },
+    {
+      "epoch": 0.37288,
+      "grad_norm": 1.4748060703277588,
+      "learning_rate": 6.951404390395282e-06,
+      "loss": 0.2881,
+      "step": 4661
+    },
+    {
+      "epoch": 0.37296,
+      "grad_norm": 1.4496852159500122,
+      "learning_rate": 6.950246421833373e-06,
+      "loss": 0.2684,
+      "step": 4662
+    },
+    {
+      "epoch": 0.37304,
+      "grad_norm": 1.4460296630859375,
+      "learning_rate": 6.949088329885898e-06,
+      "loss": 0.2797,
+      "step": 4663
+    },
+    {
+      "epoch": 0.37312,
+      "grad_norm": 1.7262259721755981,
+      "learning_rate": 6.947930114626125e-06,
+      "loss": 0.3871,
+      "step": 4664
+    },
+    {
+      "epoch": 0.3732,
+      "grad_norm": 1.6757460832595825,
+      "learning_rate": 6.946771776127334e-06,
+      "loss": 0.2729,
+      "step": 4665
+    },
+    {
+      "epoch": 0.37328,
+      "grad_norm": 1.40213143825531,
+      "learning_rate": 6.945613314462804e-06,
+      "loss": 0.3466,
+      "step": 4666
+    },
+    {
+      "epoch": 0.37336,
+      "grad_norm": 2.0440454483032227,
+      "learning_rate": 6.9444547297058315e-06,
+      "loss": 0.4531,
+      "step": 4667
+    },
+    {
+      "epoch": 0.37344,
+      "grad_norm": 1.950875163078308,
+      "learning_rate": 6.943296021929713e-06,
+      "loss": 0.4574,
+      "step": 4668
+    },
+    {
+      "epoch": 0.37352,
+      "grad_norm": 1.5116314888000488,
+      "learning_rate": 6.9421371912077585e-06,
+      "loss": 0.4164,
+      "step": 4669
+    },
+    {
+      "epoch": 0.3736,
+      "grad_norm": 2.183765172958374,
+      "learning_rate": 6.9409782376132805e-06,
+      "loss": 0.5813,
+      "step": 4670
+    },
+    {
+      "epoch": 0.37368,
+      "grad_norm": 1.54366934299469,
+      "learning_rate": 6.9398191612196055e-06,
+      "loss": 0.3999,
+      "step": 4671
+    },
+    {
+      "epoch": 0.37376,
+      "grad_norm": 1.3263498544692993,
+      "learning_rate": 6.938659962100064e-06,
+      "loss": 0.2936,
+      "step": 4672
+    },
+    {
+      "epoch": 0.37384,
+      "grad_norm": 1.3171329498291016,
+      "learning_rate": 6.937500640327992e-06,
+      "loss": 0.3129,
+      "step": 4673
+    },
+    {
+      "epoch": 0.37392,
+      "grad_norm": 1.4824219942092896,
+      "learning_rate": 6.936341195976737e-06,
+      "loss": 0.2798,
+      "step": 4674
+    },
+    {
+      "epoch": 0.374,
+      "grad_norm": 1.6193792819976807,
+      "learning_rate": 6.935181629119654e-06,
+      "loss": 0.3232,
+      "step": 4675
+    },
+    {
+      "epoch": 0.37408,
+      "grad_norm": 2.0140795707702637,
+      "learning_rate": 6.934021939830104e-06,
+      "loss": 0.4507,
+      "step": 4676
+    },
+    {
+      "epoch": 0.37416,
+      "grad_norm": 1.3184332847595215,
+      "learning_rate": 6.932862128181459e-06,
+      "loss": 0.3604,
+      "step": 4677
+    },
+    {
+      "epoch": 0.37424,
+      "grad_norm": 1.5860306024551392,
+      "learning_rate": 6.931702194247094e-06,
+      "loss": 0.3256,
+      "step": 4678
+    },
+    {
+      "epoch": 0.37432,
+      "grad_norm": 1.9520553350448608,
+      "learning_rate": 6.930542138100393e-06,
+      "loss": 0.416,
+      "step": 4679
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 1.1498091220855713,
+      "learning_rate": 6.929381959814751e-06,
+      "loss": 0.3029,
+      "step": 4680
+    },
+    {
+      "epoch": 0.37448,
+      "grad_norm": 1.5657634735107422,
+      "learning_rate": 6.928221659463568e-06,
+      "loss": 0.3745,
+      "step": 4681
+    },
+    {
+      "epoch": 0.37456,
+      "grad_norm": 1.403543472290039,
+      "learning_rate": 6.9270612371202524e-06,
+      "loss": 0.3247,
+      "step": 4682
+    },
+    {
+      "epoch": 0.37464,
+      "grad_norm": 1.869012475013733,
+      "learning_rate": 6.925900692858222e-06,
+      "loss": 0.3587,
+      "step": 4683
+    },
+    {
+      "epoch": 0.37472,
+      "grad_norm": 1.9943881034851074,
+      "learning_rate": 6.924740026750898e-06,
+      "loss": 0.4173,
+      "step": 4684
+    },
+    {
+      "epoch": 0.3748,
+      "grad_norm": 1.7604013681411743,
+      "learning_rate": 6.923579238871712e-06,
+      "loss": 0.3363,
+      "step": 4685
+    },
+    {
+      "epoch": 0.37488,
+      "grad_norm": 1.7223265171051025,
+      "learning_rate": 6.922418329294104e-06,
+      "loss": 0.3656,
+      "step": 4686
+    },
+    {
+      "epoch": 0.37496,
+      "grad_norm": 1.5790249109268188,
+      "learning_rate": 6.921257298091522e-06,
+      "loss": 0.3046,
+      "step": 4687
+    },
+    {
+      "epoch": 0.37504,
+      "grad_norm": 1.5063668489456177,
+      "learning_rate": 6.920096145337418e-06,
+      "loss": 0.2886,
+      "step": 4688
+    },
+    {
+      "epoch": 0.37512,
+      "grad_norm": 1.7410809993743896,
+      "learning_rate": 6.9189348711052565e-06,
+      "loss": 0.3991,
+      "step": 4689
+    },
+    {
+      "epoch": 0.3752,
+      "grad_norm": 1.309981346130371,
+      "learning_rate": 6.9177734754685055e-06,
+      "loss": 0.2905,
+      "step": 4690
+    },
+    {
+      "epoch": 0.37528,
+      "grad_norm": 1.7682570219039917,
+      "learning_rate": 6.916611958500644e-06,
+      "loss": 0.3659,
+      "step": 4691
+    },
+    {
+      "epoch": 0.37536,
+      "grad_norm": 1.6350557804107666,
+      "learning_rate": 6.9154503202751564e-06,
+      "loss": 0.3403,
+      "step": 4692
+    },
+    {
+      "epoch": 0.37544,
+      "grad_norm": 1.6600843667984009,
+      "learning_rate": 6.914288560865536e-06,
+      "loss": 0.3286,
+      "step": 4693
+    },
+    {
+      "epoch": 0.37552,
+      "grad_norm": 1.66802179813385,
+      "learning_rate": 6.913126680345285e-06,
+      "loss": 0.4507,
+      "step": 4694
+    },
+    {
+      "epoch": 0.3756,
+      "grad_norm": 2.0192372798919678,
+      "learning_rate": 6.911964678787908e-06,
+      "loss": 0.459,
+      "step": 4695
+    },
+    {
+      "epoch": 0.37568,
+      "grad_norm": 1.459499478340149,
+      "learning_rate": 6.910802556266927e-06,
+      "loss": 0.279,
+      "step": 4696
+    },
+    {
+      "epoch": 0.37576,
+      "grad_norm": 2.2443490028381348,
+      "learning_rate": 6.909640312855859e-06,
+      "loss": 0.3926,
+      "step": 4697
+    },
+    {
+      "epoch": 0.37584,
+      "grad_norm": 1.5917470455169678,
+      "learning_rate": 6.9084779486282385e-06,
+      "loss": 0.3271,
+      "step": 4698
+    },
+    {
+      "epoch": 0.37592,
+      "grad_norm": 1.4331306219100952,
+      "learning_rate": 6.907315463657603e-06,
+      "loss": 0.2767,
+      "step": 4699
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 1.7702385187149048,
+      "learning_rate": 6.906152858017502e-06,
+      "loss": 0.3622,
+      "step": 4700
+    },
+    {
+      "epoch": 0.37608,
+      "grad_norm": 1.6627930402755737,
+      "learning_rate": 6.904990131781486e-06,
+      "loss": 0.4196,
+      "step": 4701
+    },
+    {
+      "epoch": 0.37616,
+      "grad_norm": 1.6581454277038574,
+      "learning_rate": 6.90382728502312e-06,
+      "loss": 0.3285,
+      "step": 4702
+    },
+    {
+      "epoch": 0.37624,
+      "grad_norm": 1.6215664148330688,
+      "learning_rate": 6.9026643178159714e-06,
+      "loss": 0.3781,
+      "step": 4703
+    },
+    {
+      "epoch": 0.37632,
+      "grad_norm": 1.4504485130310059,
+      "learning_rate": 6.901501230233617e-06,
+      "loss": 0.3214,
+      "step": 4704
+    },
+    {
+      "epoch": 0.3764,
+      "grad_norm": 1.4753299951553345,
+      "learning_rate": 6.900338022349643e-06,
+      "loss": 0.3267,
+      "step": 4705
+    },
+    {
+      "epoch": 0.37648,
+      "grad_norm": 1.9508014917373657,
+      "learning_rate": 6.89917469423764e-06,
+      "loss": 0.4347,
+      "step": 4706
+    },
+    {
+      "epoch": 0.37656,
+      "grad_norm": 1.515520453453064,
+      "learning_rate": 6.89801124597121e-06,
+      "loss": 0.3053,
+      "step": 4707
+    },
+    {
+      "epoch": 0.37664,
+      "grad_norm": 1.8075672388076782,
+      "learning_rate": 6.896847677623959e-06,
+      "loss": 0.3736,
+      "step": 4708
+    },
+    {
+      "epoch": 0.37672,
+      "grad_norm": 1.3629133701324463,
+      "learning_rate": 6.8956839892695015e-06,
+      "loss": 0.2908,
+      "step": 4709
+    },
+    {
+      "epoch": 0.3768,
+      "grad_norm": 1.7496662139892578,
+      "learning_rate": 6.894520180981461e-06,
+      "loss": 0.3834,
+      "step": 4710
+    },
+    {
+      "epoch": 0.37688,
+      "grad_norm": 2.0693397521972656,
+      "learning_rate": 6.893356252833469e-06,
+      "loss": 0.3402,
+      "step": 4711
+    },
+    {
+      "epoch": 0.37696,
+      "grad_norm": 1.662286400794983,
+      "learning_rate": 6.892192204899161e-06,
+      "loss": 0.3915,
+      "step": 4712
+    },
+    {
+      "epoch": 0.37704,
+      "grad_norm": 1.633707880973816,
+      "learning_rate": 6.8910280372521834e-06,
+      "loss": 0.2893,
+      "step": 4713
+    },
+    {
+      "epoch": 0.37712,
+      "grad_norm": 1.2424566745758057,
+      "learning_rate": 6.8898637499661906e-06,
+      "loss": 0.2871,
+      "step": 4714
+    },
+    {
+      "epoch": 0.3772,
+      "grad_norm": 1.8725136518478394,
+      "learning_rate": 6.8886993431148395e-06,
+      "loss": 0.406,
+      "step": 4715
+    },
+    {
+      "epoch": 0.37728,
+      "grad_norm": 1.4558591842651367,
+      "learning_rate": 6.887534816771802e-06,
+      "loss": 0.2686,
+      "step": 4716
+    },
+    {
+      "epoch": 0.37736,
+      "grad_norm": 1.2151341438293457,
+      "learning_rate": 6.886370171010752e-06,
+      "loss": 0.2585,
+      "step": 4717
+    },
+    {
+      "epoch": 0.37744,
+      "grad_norm": 1.3010770082473755,
+      "learning_rate": 6.885205405905373e-06,
+      "loss": 0.2596,
+      "step": 4718
+    },
+    {
+      "epoch": 0.37752,
+      "grad_norm": 1.303667664527893,
+      "learning_rate": 6.884040521529356e-06,
+      "loss": 0.2937,
+      "step": 4719
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 1.5519078969955444,
+      "learning_rate": 6.8828755179564e-06,
+      "loss": 0.3064,
+      "step": 4720
+    },
+    {
+      "epoch": 0.37768,
+      "grad_norm": 2.096662759780884,
+      "learning_rate": 6.88171039526021e-06,
+      "loss": 0.469,
+      "step": 4721
+    },
+    {
+      "epoch": 0.37776,
+      "grad_norm": 1.5880454778671265,
+      "learning_rate": 6.880545153514498e-06,
+      "loss": 0.3606,
+      "step": 4722
+    },
+    {
+      "epoch": 0.37784,
+      "grad_norm": 1.351523756980896,
+      "learning_rate": 6.879379792792988e-06,
+      "loss": 0.3292,
+      "step": 4723
+    },
+    {
+      "epoch": 0.37792,
+      "grad_norm": 1.5579664707183838,
+      "learning_rate": 6.878214313169407e-06,
+      "loss": 0.3451,
+      "step": 4724
+    },
+    {
+      "epoch": 0.378,
+      "grad_norm": 1.906211018562317,
+      "learning_rate": 6.87704871471749e-06,
+      "loss": 0.3471,
+      "step": 4725
+    },
+    {
+      "epoch": 0.37808,
+      "grad_norm": 1.8630839586257935,
+      "learning_rate": 6.875882997510982e-06,
+      "loss": 0.4373,
+      "step": 4726
+    },
+    {
+      "epoch": 0.37816,
+      "grad_norm": 1.5359853506088257,
+      "learning_rate": 6.874717161623633e-06,
+      "loss": 0.323,
+      "step": 4727
+    },
+    {
+      "epoch": 0.37824,
+      "grad_norm": 1.6257649660110474,
+      "learning_rate": 6.8735512071292024e-06,
+      "loss": 0.4539,
+      "step": 4728
+    },
+    {
+      "epoch": 0.37832,
+      "grad_norm": 1.5281004905700684,
+      "learning_rate": 6.872385134101454e-06,
+      "loss": 0.3276,
+      "step": 4729
+    },
+    {
+      "epoch": 0.3784,
+      "grad_norm": 1.4861382246017456,
+      "learning_rate": 6.871218942614165e-06,
+      "loss": 0.398,
+      "step": 4730
+    },
+    {
+      "epoch": 0.37848,
+      "grad_norm": 1.907596230506897,
+      "learning_rate": 6.8700526327411155e-06,
+      "loss": 0.3705,
+      "step": 4731
+    },
+    {
+      "epoch": 0.37856,
+      "grad_norm": 1.6404950618743896,
+      "learning_rate": 6.868886204556092e-06,
+      "loss": 0.4075,
+      "step": 4732
+    },
+    {
+      "epoch": 0.37864,
+      "grad_norm": 1.4109077453613281,
+      "learning_rate": 6.867719658132892e-06,
+      "loss": 0.3178,
+      "step": 4733
+    },
+    {
+      "epoch": 0.37872,
+      "grad_norm": 1.863938808441162,
+      "learning_rate": 6.866552993545319e-06,
+      "loss": 0.3065,
+      "step": 4734
+    },
+    {
+      "epoch": 0.3788,
+      "grad_norm": 2.2239177227020264,
+      "learning_rate": 6.865386210867182e-06,
+      "loss": 0.4868,
+      "step": 4735
+    },
+    {
+      "epoch": 0.37888,
+      "grad_norm": 1.7756961584091187,
+      "learning_rate": 6.864219310172302e-06,
+      "loss": 0.3413,
+      "step": 4736
+    },
+    {
+      "epoch": 0.37896,
+      "grad_norm": 1.6238456964492798,
+      "learning_rate": 6.863052291534505e-06,
+      "loss": 0.3501,
+      "step": 4737
+    },
+    {
+      "epoch": 0.37904,
+      "grad_norm": 1.5405563116073608,
+      "learning_rate": 6.8618851550276225e-06,
+      "loss": 0.3405,
+      "step": 4738
+    },
+    {
+      "epoch": 0.37912,
+      "grad_norm": 1.663956642150879,
+      "learning_rate": 6.860717900725495e-06,
+      "loss": 0.378,
+      "step": 4739
+    },
+    {
+      "epoch": 0.3792,
+      "grad_norm": 1.5309809446334839,
+      "learning_rate": 6.859550528701972e-06,
+      "loss": 0.3202,
+      "step": 4740
+    },
+    {
+      "epoch": 0.37928,
+      "grad_norm": 1.7055078744888306,
+      "learning_rate": 6.858383039030911e-06,
+      "loss": 0.3591,
+      "step": 4741
+    },
+    {
+      "epoch": 0.37936,
+      "grad_norm": 1.452616572380066,
+      "learning_rate": 6.857215431786172e-06,
+      "loss": 0.3181,
+      "step": 4742
+    },
+    {
+      "epoch": 0.37944,
+      "grad_norm": 1.454076886177063,
+      "learning_rate": 6.856047707041628e-06,
+      "loss": 0.319,
+      "step": 4743
+    },
+    {
+      "epoch": 0.37952,
+      "grad_norm": 1.9659279584884644,
+      "learning_rate": 6.854879864871155e-06,
+      "loss": 0.4883,
+      "step": 4744
+    },
+    {
+      "epoch": 0.3796,
+      "grad_norm": 1.0691478252410889,
+      "learning_rate": 6.85371190534864e-06,
+      "loss": 0.2573,
+      "step": 4745
+    },
+    {
+      "epoch": 0.37968,
+      "grad_norm": 1.682841420173645,
+      "learning_rate": 6.8525438285479755e-06,
+      "loss": 0.2741,
+      "step": 4746
+    },
+    {
+      "epoch": 0.37976,
+      "grad_norm": 1.7029143571853638,
+      "learning_rate": 6.85137563454306e-06,
+      "loss": 0.3921,
+      "step": 4747
+    },
+    {
+      "epoch": 0.37984,
+      "grad_norm": 1.5727744102478027,
+      "learning_rate": 6.850207323407803e-06,
+      "loss": 0.3508,
+      "step": 4748
+    },
+    {
+      "epoch": 0.37992,
+      "grad_norm": 1.4125081300735474,
+      "learning_rate": 6.8490388952161215e-06,
+      "loss": 0.3447,
+      "step": 4749
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9592466354370117,
+      "learning_rate": 6.847870350041934e-06,
+      "loss": 0.3249,
+      "step": 4750
+    },
+    {
+      "epoch": 0.38008,
+      "grad_norm": 1.4628976583480835,
+      "learning_rate": 6.846701687959173e-06,
+      "loss": 0.2875,
+      "step": 4751
+    },
+    {
+      "epoch": 0.38016,
+      "grad_norm": 1.5076208114624023,
+      "learning_rate": 6.845532909041775e-06,
+      "loss": 0.3145,
+      "step": 4752
+    },
+    {
+      "epoch": 0.38024,
+      "grad_norm": 1.3270679712295532,
+      "learning_rate": 6.8443640133636845e-06,
+      "loss": 0.3381,
+      "step": 4753
+    },
+    {
+      "epoch": 0.38032,
+      "grad_norm": 2.057631731033325,
+      "learning_rate": 6.8431950009988565e-06,
+      "loss": 0.3674,
+      "step": 4754
+    },
+    {
+      "epoch": 0.3804,
+      "grad_norm": 1.4664708375930786,
+      "learning_rate": 6.842025872021246e-06,
+      "loss": 0.3211,
+      "step": 4755
+    },
+    {
+      "epoch": 0.38048,
+      "grad_norm": 1.819962739944458,
+      "learning_rate": 6.840856626504822e-06,
+      "loss": 0.3592,
+      "step": 4756
+    },
+    {
+      "epoch": 0.38056,
+      "grad_norm": 1.3360856771469116,
+      "learning_rate": 6.8396872645235615e-06,
+      "loss": 0.2608,
+      "step": 4757
+    },
+    {
+      "epoch": 0.38064,
+      "grad_norm": 1.8079757690429688,
+      "learning_rate": 6.838517786151441e-06,
+      "loss": 0.4004,
+      "step": 4758
+    },
+    {
+      "epoch": 0.38072,
+      "grad_norm": 1.4537521600723267,
+      "learning_rate": 6.837348191462452e-06,
+      "loss": 0.288,
+      "step": 4759
+    },
+    {
+      "epoch": 0.3808,
+      "grad_norm": 1.3475890159606934,
+      "learning_rate": 6.8361784805305905e-06,
+      "loss": 0.2607,
+      "step": 4760
+    },
+    {
+      "epoch": 0.38088,
+      "grad_norm": 1.749821424484253,
+      "learning_rate": 6.835008653429862e-06,
+      "loss": 0.3755,
+      "step": 4761
+    },
+    {
+      "epoch": 0.38096,
+      "grad_norm": 1.7868049144744873,
+      "learning_rate": 6.833838710234274e-06,
+      "loss": 0.5124,
+      "step": 4762
+    },
+    {
+      "epoch": 0.38104,
+      "grad_norm": 1.7680143117904663,
+      "learning_rate": 6.8326686510178475e-06,
+      "loss": 0.3669,
+      "step": 4763
+    },
+    {
+      "epoch": 0.38112,
+      "grad_norm": 1.4955719709396362,
+      "learning_rate": 6.831498475854607e-06,
+      "loss": 0.3469,
+      "step": 4764
+    },
+    {
+      "epoch": 0.3812,
+      "grad_norm": 1.3218600749969482,
+      "learning_rate": 6.8303281848185864e-06,
+      "loss": 0.3065,
+      "step": 4765
+    },
+    {
+      "epoch": 0.38128,
+      "grad_norm": 1.1174854040145874,
+      "learning_rate": 6.829157777983828e-06,
+      "loss": 0.2463,
+      "step": 4766
+    },
+    {
+      "epoch": 0.38136,
+      "grad_norm": 1.5425974130630493,
+      "learning_rate": 6.827987255424375e-06,
+      "loss": 0.3862,
+      "step": 4767
+    },
+    {
+      "epoch": 0.38144,
+      "grad_norm": 1.8264150619506836,
+      "learning_rate": 6.826816617214287e-06,
+      "loss": 0.4402,
+      "step": 4768
+    },
+    {
+      "epoch": 0.38152,
+      "grad_norm": 1.5018960237503052,
+      "learning_rate": 6.8256458634276216e-06,
+      "loss": 0.3495,
+      "step": 4769
+    },
+    {
+      "epoch": 0.3816,
+      "grad_norm": 1.9828064441680908,
+      "learning_rate": 6.824474994138453e-06,
+      "loss": 0.4644,
+      "step": 4770
+    },
+    {
+      "epoch": 0.38168,
+      "grad_norm": 1.6333398818969727,
+      "learning_rate": 6.823304009420855e-06,
+      "loss": 0.3376,
+      "step": 4771
+    },
+    {
+      "epoch": 0.38176,
+      "grad_norm": 1.5310996770858765,
+      "learning_rate": 6.822132909348913e-06,
+      "loss": 0.3445,
+      "step": 4772
+    },
+    {
+      "epoch": 0.38184,
+      "grad_norm": 1.3783681392669678,
+      "learning_rate": 6.820961693996719e-06,
+      "loss": 0.2656,
+      "step": 4773
+    },
+    {
+      "epoch": 0.38192,
+      "grad_norm": 1.7104060649871826,
+      "learning_rate": 6.819790363438372e-06,
+      "loss": 0.3171,
+      "step": 4774
+    },
+    {
+      "epoch": 0.382,
+      "grad_norm": 1.5256901979446411,
+      "learning_rate": 6.818618917747977e-06,
+      "loss": 0.3218,
+      "step": 4775
+    },
+    {
+      "epoch": 0.38208,
+      "grad_norm": 1.4606126546859741,
+      "learning_rate": 6.817447356999649e-06,
+      "loss": 0.2958,
+      "step": 4776
+    },
+    {
+      "epoch": 0.38216,
+      "grad_norm": 1.5362343788146973,
+      "learning_rate": 6.816275681267507e-06,
+      "loss": 0.3253,
+      "step": 4777
+    },
+    {
+      "epoch": 0.38224,
+      "grad_norm": 1.2135709524154663,
+      "learning_rate": 6.815103890625682e-06,
+      "loss": 0.2384,
+      "step": 4778
+    },
+    {
+      "epoch": 0.38232,
+      "grad_norm": 1.9567755460739136,
+      "learning_rate": 6.813931985148306e-06,
+      "loss": 0.4436,
+      "step": 4779
+    },
+    {
+      "epoch": 0.3824,
+      "grad_norm": 1.4235215187072754,
+      "learning_rate": 6.812759964909522e-06,
+      "loss": 0.2772,
+      "step": 4780
+    },
+    {
+      "epoch": 0.38248,
+      "grad_norm": 1.557592749595642,
+      "learning_rate": 6.811587829983481e-06,
+      "loss": 0.3733,
+      "step": 4781
+    },
+    {
+      "epoch": 0.38256,
+      "grad_norm": 1.7264912128448486,
+      "learning_rate": 6.810415580444339e-06,
+      "loss": 0.3979,
+      "step": 4782
+    },
+    {
+      "epoch": 0.38264,
+      "grad_norm": 1.3333731889724731,
+      "learning_rate": 6.809243216366261e-06,
+      "loss": 0.3936,
+      "step": 4783
+    },
+    {
+      "epoch": 0.38272,
+      "grad_norm": 1.5328093767166138,
+      "learning_rate": 6.80807073782342e-06,
+      "loss": 0.3305,
+      "step": 4784
+    },
+    {
+      "epoch": 0.3828,
+      "grad_norm": 1.6207122802734375,
+      "learning_rate": 6.806898144889992e-06,
+      "loss": 0.3319,
+      "step": 4785
+    },
+    {
+      "epoch": 0.38288,
+      "grad_norm": 1.7416425943374634,
+      "learning_rate": 6.8057254376401635e-06,
+      "loss": 0.4054,
+      "step": 4786
+    },
+    {
+      "epoch": 0.38296,
+      "grad_norm": 1.8004319667816162,
+      "learning_rate": 6.80455261614813e-06,
+      "loss": 0.4405,
+      "step": 4787
+    },
+    {
+      "epoch": 0.38304,
+      "grad_norm": 1.8042232990264893,
+      "learning_rate": 6.803379680488089e-06,
+      "loss": 0.3759,
+      "step": 4788
+    },
+    {
+      "epoch": 0.38312,
+      "grad_norm": 1.7607706785202026,
+      "learning_rate": 6.802206630734252e-06,
+      "loss": 0.4345,
+      "step": 4789
+    },
+    {
+      "epoch": 0.3832,
+      "grad_norm": 1.388512372970581,
+      "learning_rate": 6.80103346696083e-06,
+      "loss": 0.2903,
+      "step": 4790
+    },
+    {
+      "epoch": 0.38328,
+      "grad_norm": 1.2507061958312988,
+      "learning_rate": 6.799860189242049e-06,
+      "loss": 0.2915,
+      "step": 4791
+    },
+    {
+      "epoch": 0.38336,
+      "grad_norm": 1.4583854675292969,
+      "learning_rate": 6.798686797652134e-06,
+      "loss": 0.3479,
+      "step": 4792
+    },
+    {
+      "epoch": 0.38344,
+      "grad_norm": 1.3498902320861816,
+      "learning_rate": 6.797513292265323e-06,
+      "loss": 0.3524,
+      "step": 4793
+    },
+    {
+      "epoch": 0.38352,
+      "grad_norm": 1.3097642660140991,
+      "learning_rate": 6.7963396731558615e-06,
+      "loss": 0.2726,
+      "step": 4794
+    },
+    {
+      "epoch": 0.3836,
+      "grad_norm": 1.6635525226593018,
+      "learning_rate": 6.795165940398e-06,
+      "loss": 0.3982,
+      "step": 4795
+    },
+    {
+      "epoch": 0.38368,
+      "grad_norm": 1.2881489992141724,
+      "learning_rate": 6.7939920940659966e-06,
+      "loss": 0.2503,
+      "step": 4796
+    },
+    {
+      "epoch": 0.38376,
+      "grad_norm": 1.7574576139450073,
+      "learning_rate": 6.792818134234115e-06,
+      "loss": 0.4502,
+      "step": 4797
+    },
+    {
+      "epoch": 0.38384,
+      "grad_norm": 1.5125421285629272,
+      "learning_rate": 6.791644060976629e-06,
+      "loss": 0.401,
+      "step": 4798
+    },
+    {
+      "epoch": 0.38392,
+      "grad_norm": 1.5928010940551758,
+      "learning_rate": 6.7904698743678175e-06,
+      "loss": 0.3318,
+      "step": 4799
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 1.3006750345230103,
+      "learning_rate": 6.789295574481969e-06,
+      "loss": 0.2972,
+      "step": 4800
+    },
+    {
+      "epoch": 0.38408,
+      "grad_norm": 1.9647408723831177,
+      "learning_rate": 6.788121161393376e-06,
+      "loss": 0.4477,
+      "step": 4801
+    },
+    {
+      "epoch": 0.38416,
+      "grad_norm": 1.5676745176315308,
+      "learning_rate": 6.7869466351763405e-06,
+      "loss": 0.3331,
+      "step": 4802
+    },
+    {
+      "epoch": 0.38424,
+      "grad_norm": 1.4142674207687378,
+      "learning_rate": 6.78577199590517e-06,
+      "loss": 0.2803,
+      "step": 4803
+    },
+    {
+      "epoch": 0.38432,
+      "grad_norm": 1.6045714616775513,
+      "learning_rate": 6.784597243654182e-06,
+      "loss": 0.3822,
+      "step": 4804
+    },
+    {
+      "epoch": 0.3844,
+      "grad_norm": 1.3804314136505127,
+      "learning_rate": 6.783422378497696e-06,
+      "loss": 0.362,
+      "step": 4805
+    },
+    {
+      "epoch": 0.38448,
+      "grad_norm": 1.6979622840881348,
+      "learning_rate": 6.7822474005100435e-06,
+      "loss": 0.413,
+      "step": 4806
+    },
+    {
+      "epoch": 0.38456,
+      "grad_norm": 1.5267022848129272,
+      "learning_rate": 6.781072309765563e-06,
+      "loss": 0.3208,
+      "step": 4807
+    },
+    {
+      "epoch": 0.38464,
+      "grad_norm": 1.3634998798370361,
+      "learning_rate": 6.779897106338595e-06,
+      "loss": 0.2869,
+      "step": 4808
+    },
+    {
+      "epoch": 0.38472,
+      "grad_norm": 1.6127568483352661,
+      "learning_rate": 6.778721790303494e-06,
+      "loss": 0.3086,
+      "step": 4809
+    },
+    {
+      "epoch": 0.3848,
+      "grad_norm": 1.515779733657837,
+      "learning_rate": 6.777546361734616e-06,
+      "loss": 0.3614,
+      "step": 4810
+    },
+    {
+      "epoch": 0.38488,
+      "grad_norm": 1.6880332231521606,
+      "learning_rate": 6.7763708207063286e-06,
+      "loss": 0.3785,
+      "step": 4811
+    },
+    {
+      "epoch": 0.38496,
+      "grad_norm": 1.5529166460037231,
+      "learning_rate": 6.775195167293003e-06,
+      "loss": 0.3199,
+      "step": 4812
+    },
+    {
+      "epoch": 0.38504,
+      "grad_norm": 1.7941169738769531,
+      "learning_rate": 6.7740194015690175e-06,
+      "loss": 0.4966,
+      "step": 4813
+    },
+    {
+      "epoch": 0.38512,
+      "grad_norm": 1.9801031351089478,
+      "learning_rate": 6.772843523608762e-06,
+      "loss": 0.3773,
+      "step": 4814
+    },
+    {
+      "epoch": 0.3852,
+      "grad_norm": 1.4930901527404785,
+      "learning_rate": 6.771667533486628e-06,
+      "loss": 0.3426,
+      "step": 4815
+    },
+    {
+      "epoch": 0.38528,
+      "grad_norm": 1.4776076078414917,
+      "learning_rate": 6.770491431277017e-06,
+      "loss": 0.3437,
+      "step": 4816
+    },
+    {
+      "epoch": 0.38536,
+      "grad_norm": 1.2407801151275635,
+      "learning_rate": 6.7693152170543365e-06,
+      "loss": 0.2532,
+      "step": 4817
+    },
+    {
+      "epoch": 0.38544,
+      "grad_norm": 1.7486261129379272,
+      "learning_rate": 6.768138890893004e-06,
+      "loss": 0.353,
+      "step": 4818
+    },
+    {
+      "epoch": 0.38552,
+      "grad_norm": 2.0305092334747314,
+      "learning_rate": 6.766962452867439e-06,
+      "loss": 0.4327,
+      "step": 4819
+    },
+    {
+      "epoch": 0.3856,
+      "grad_norm": 1.3625537157058716,
+      "learning_rate": 6.765785903052072e-06,
+      "loss": 0.2888,
+      "step": 4820
+    },
+    {
+      "epoch": 0.38568,
+      "grad_norm": 1.9757248163223267,
+      "learning_rate": 6.764609241521339e-06,
+      "loss": 0.4204,
+      "step": 4821
+    },
+    {
+      "epoch": 0.38576,
+      "grad_norm": 1.4327483177185059,
+      "learning_rate": 6.763432468349684e-06,
+      "loss": 0.297,
+      "step": 4822
+    },
+    {
+      "epoch": 0.38584,
+      "grad_norm": 1.9168870449066162,
+      "learning_rate": 6.762255583611557e-06,
+      "loss": 0.435,
+      "step": 4823
+    },
+    {
+      "epoch": 0.38592,
+      "grad_norm": 1.9956098794937134,
+      "learning_rate": 6.761078587381416e-06,
+      "loss": 0.3668,
+      "step": 4824
+    },
+    {
+      "epoch": 0.386,
+      "grad_norm": 1.5235965251922607,
+      "learning_rate": 6.759901479733727e-06,
+      "loss": 0.3124,
+      "step": 4825
+    },
+    {
+      "epoch": 0.38608,
+      "grad_norm": 1.2701911926269531,
+      "learning_rate": 6.7587242607429585e-06,
+      "loss": 0.2314,
+      "step": 4826
+    },
+    {
+      "epoch": 0.38616,
+      "grad_norm": 2.225844144821167,
+      "learning_rate": 6.7575469304835905e-06,
+      "loss": 0.467,
+      "step": 4827
+    },
+    {
+      "epoch": 0.38624,
+      "grad_norm": 1.8088279962539673,
+      "learning_rate": 6.756369489030109e-06,
+      "loss": 0.4096,
+      "step": 4828
+    },
+    {
+      "epoch": 0.38632,
+      "grad_norm": 1.5828858613967896,
+      "learning_rate": 6.755191936457006e-06,
+      "loss": 0.3111,
+      "step": 4829
+    },
+    {
+      "epoch": 0.3864,
+      "grad_norm": 2.1753506660461426,
+      "learning_rate": 6.754014272838782e-06,
+      "loss": 0.3957,
+      "step": 4830
+    },
+    {
+      "epoch": 0.38648,
+      "grad_norm": 1.830126166343689,
+      "learning_rate": 6.752836498249946e-06,
+      "loss": 0.3539,
+      "step": 4831
+    },
+    {
+      "epoch": 0.38656,
+      "grad_norm": 1.4874694347381592,
+      "learning_rate": 6.751658612765008e-06,
+      "loss": 0.3003,
+      "step": 4832
+    },
+    {
+      "epoch": 0.38664,
+      "grad_norm": 1.2147022485733032,
+      "learning_rate": 6.750480616458492e-06,
+      "loss": 0.2779,
+      "step": 4833
+    },
+    {
+      "epoch": 0.38672,
+      "grad_norm": 1.595517873764038,
+      "learning_rate": 6.749302509404924e-06,
+      "loss": 0.3439,
+      "step": 4834
+    },
+    {
+      "epoch": 0.3868,
+      "grad_norm": 1.3618996143341064,
+      "learning_rate": 6.748124291678839e-06,
+      "loss": 0.3388,
+      "step": 4835
+    },
+    {
+      "epoch": 0.38688,
+      "grad_norm": 1.43161940574646,
+      "learning_rate": 6.74694596335478e-06,
+      "loss": 0.2954,
+      "step": 4836
+    },
+    {
+      "epoch": 0.38696,
+      "grad_norm": 1.5240548849105835,
+      "learning_rate": 6.745767524507296e-06,
+      "loss": 0.2675,
+      "step": 4837
+    },
+    {
+      "epoch": 0.38704,
+      "grad_norm": 1.5682686567306519,
+      "learning_rate": 6.744588975210939e-06,
+      "loss": 0.3627,
+      "step": 4838
+    },
+    {
+      "epoch": 0.38712,
+      "grad_norm": 1.8606730699539185,
+      "learning_rate": 6.743410315540277e-06,
+      "loss": 0.3254,
+      "step": 4839
+    },
+    {
+      "epoch": 0.3872,
+      "grad_norm": 1.5660977363586426,
+      "learning_rate": 6.742231545569879e-06,
+      "loss": 0.4426,
+      "step": 4840
+    },
+    {
+      "epoch": 0.38728,
+      "grad_norm": 1.6271204948425293,
+      "learning_rate": 6.741052665374318e-06,
+      "loss": 0.3316,
+      "step": 4841
+    },
+    {
+      "epoch": 0.38736,
+      "grad_norm": 1.4664791822433472,
+      "learning_rate": 6.739873675028182e-06,
+      "loss": 0.3391,
+      "step": 4842
+    },
+    {
+      "epoch": 0.38744,
+      "grad_norm": 1.467913269996643,
+      "learning_rate": 6.738694574606059e-06,
+      "loss": 0.2995,
+      "step": 4843
+    },
+    {
+      "epoch": 0.38752,
+      "grad_norm": 1.3580753803253174,
+      "learning_rate": 6.73751536418255e-06,
+      "loss": 0.282,
+      "step": 4844
+    },
+    {
+      "epoch": 0.3876,
+      "grad_norm": 1.404686689376831,
+      "learning_rate": 6.736336043832255e-06,
+      "loss": 0.3058,
+      "step": 4845
+    },
+    {
+      "epoch": 0.38768,
+      "grad_norm": 1.6138893365859985,
+      "learning_rate": 6.73515661362979e-06,
+      "loss": 0.3093,
+      "step": 4846
+    },
+    {
+      "epoch": 0.38776,
+      "grad_norm": 1.5155889987945557,
+      "learning_rate": 6.733977073649774e-06,
+      "loss": 0.2914,
+      "step": 4847
+    },
+    {
+      "epoch": 0.38784,
+      "grad_norm": 1.9643943309783936,
+      "learning_rate": 6.732797423966828e-06,
+      "loss": 0.3926,
+      "step": 4848
+    },
+    {
+      "epoch": 0.38792,
+      "grad_norm": 1.6930899620056152,
+      "learning_rate": 6.731617664655586e-06,
+      "loss": 0.4257,
+      "step": 4849
+    },
+    {
+      "epoch": 0.388,
+      "grad_norm": 1.435917854309082,
+      "learning_rate": 6.73043779579069e-06,
+      "loss": 0.3335,
+      "step": 4850
+    },
+    {
+      "epoch": 0.38808,
+      "grad_norm": 1.7660568952560425,
+      "learning_rate": 6.7292578174467835e-06,
+      "loss": 0.3879,
+      "step": 4851
+    },
+    {
+      "epoch": 0.38816,
+      "grad_norm": 1.6072431802749634,
+      "learning_rate": 6.728077729698521e-06,
+      "loss": 0.3515,
+      "step": 4852
+    },
+    {
+      "epoch": 0.38824,
+      "grad_norm": 1.5983455181121826,
+      "learning_rate": 6.726897532620564e-06,
+      "loss": 0.3655,
+      "step": 4853
+    },
+    {
+      "epoch": 0.38832,
+      "grad_norm": 1.5852428674697876,
+      "learning_rate": 6.725717226287578e-06,
+      "loss": 0.3414,
+      "step": 4854
+    },
+    {
+      "epoch": 0.3884,
+      "grad_norm": 1.3663932085037231,
+      "learning_rate": 6.724536810774237e-06,
+      "loss": 0.2862,
+      "step": 4855
+    },
+    {
+      "epoch": 0.38848,
+      "grad_norm": 1.508460521697998,
+      "learning_rate": 6.723356286155223e-06,
+      "loss": 0.3473,
+      "step": 4856
+    },
+    {
+      "epoch": 0.38856,
+      "grad_norm": 2.041231870651245,
+      "learning_rate": 6.722175652505222e-06,
+      "loss": 0.3408,
+      "step": 4857
+    },
+    {
+      "epoch": 0.38864,
+      "grad_norm": 1.6656372547149658,
+      "learning_rate": 6.72099490989893e-06,
+      "loss": 0.3285,
+      "step": 4858
+    },
+    {
+      "epoch": 0.38872,
+      "grad_norm": 1.4684637784957886,
+      "learning_rate": 6.719814058411049e-06,
+      "loss": 0.3015,
+      "step": 4859
+    },
+    {
+      "epoch": 0.3888,
+      "grad_norm": 1.4412018060684204,
+      "learning_rate": 6.718633098116288e-06,
+      "loss": 0.3504,
+      "step": 4860
+    },
+    {
+      "epoch": 0.38888,
+      "grad_norm": 1.2994134426116943,
+      "learning_rate": 6.7174520290893594e-06,
+      "loss": 0.3174,
+      "step": 4861
+    },
+    {
+      "epoch": 0.38896,
+      "grad_norm": 2.318960428237915,
+      "learning_rate": 6.716270851404989e-06,
+      "loss": 0.5163,
+      "step": 4862
+    },
+    {
+      "epoch": 0.38904,
+      "grad_norm": 1.9144383668899536,
+      "learning_rate": 6.715089565137904e-06,
+      "loss": 0.3362,
+      "step": 4863
+    },
+    {
+      "epoch": 0.38912,
+      "grad_norm": 1.4348660707473755,
+      "learning_rate": 6.713908170362841e-06,
+      "loss": 0.383,
+      "step": 4864
+    },
+    {
+      "epoch": 0.3892,
+      "grad_norm": 1.6656911373138428,
+      "learning_rate": 6.712726667154542e-06,
+      "loss": 0.3147,
+      "step": 4865
+    },
+    {
+      "epoch": 0.38928,
+      "grad_norm": 1.9064844846725464,
+      "learning_rate": 6.711545055587759e-06,
+      "loss": 0.5062,
+      "step": 4866
+    },
+    {
+      "epoch": 0.38936,
+      "grad_norm": 1.5904337167739868,
+      "learning_rate": 6.710363335737246e-06,
+      "loss": 0.3469,
+      "step": 4867
+    },
+    {
+      "epoch": 0.38944,
+      "grad_norm": 1.8707817792892456,
+      "learning_rate": 6.709181507677769e-06,
+      "loss": 0.3365,
+      "step": 4868
+    },
+    {
+      "epoch": 0.38952,
+      "grad_norm": 1.9314450025558472,
+      "learning_rate": 6.7079995714840965e-06,
+      "loss": 0.3962,
+      "step": 4869
+    },
+    {
+      "epoch": 0.3896,
+      "grad_norm": 1.5333346128463745,
+      "learning_rate": 6.706817527231006e-06,
+      "loss": 0.304,
+      "step": 4870
+    },
+    {
+      "epoch": 0.38968,
+      "grad_norm": 1.4997177124023438,
+      "learning_rate": 6.70563537499328e-06,
+      "loss": 0.3568,
+      "step": 4871
+    },
+    {
+      "epoch": 0.38976,
+      "grad_norm": 2.006237745285034,
+      "learning_rate": 6.704453114845712e-06,
+      "loss": 0.4563,
+      "step": 4872
+    },
+    {
+      "epoch": 0.38984,
+      "grad_norm": 2.0449886322021484,
+      "learning_rate": 6.7032707468631e-06,
+      "loss": 0.4209,
+      "step": 4873
+    },
+    {
+      "epoch": 0.38992,
+      "grad_norm": 1.588484525680542,
+      "learning_rate": 6.702088271120245e-06,
+      "loss": 0.3102,
+      "step": 4874
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.401896595954895,
+      "learning_rate": 6.700905687691961e-06,
+      "loss": 0.3116,
+      "step": 4875
+    },
+    {
+      "epoch": 0.39008,
+      "grad_norm": 1.8815094232559204,
+      "learning_rate": 6.699722996653065e-06,
+      "loss": 0.481,
+      "step": 4876
+    },
+    {
+      "epoch": 0.39016,
+      "grad_norm": 1.3775348663330078,
+      "learning_rate": 6.6985401980783826e-06,
+      "loss": 0.2989,
+      "step": 4877
+    },
+    {
+      "epoch": 0.39024,
+      "grad_norm": 1.5223547220230103,
+      "learning_rate": 6.697357292042746e-06,
+      "loss": 0.3186,
+      "step": 4878
+    },
+    {
+      "epoch": 0.39032,
+      "grad_norm": 1.3824423551559448,
+      "learning_rate": 6.6961742786209925e-06,
+      "loss": 0.2951,
+      "step": 4879
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 1.2650786638259888,
+      "learning_rate": 6.694991157887967e-06,
+      "loss": 0.2914,
+      "step": 4880
+    },
+    {
+      "epoch": 0.39048,
+      "grad_norm": 1.40303373336792,
+      "learning_rate": 6.693807929918523e-06,
+      "loss": 0.3228,
+      "step": 4881
+    },
+    {
+      "epoch": 0.39056,
+      "grad_norm": 1.308904767036438,
+      "learning_rate": 6.692624594787519e-06,
+      "loss": 0.248,
+      "step": 4882
+    },
+    {
+      "epoch": 0.39064,
+      "grad_norm": 1.8516181707382202,
+      "learning_rate": 6.69144115256982e-06,
+      "loss": 0.4687,
+      "step": 4883
+    },
+    {
+      "epoch": 0.39072,
+      "grad_norm": 1.5992012023925781,
+      "learning_rate": 6.6902576033403e-06,
+      "loss": 0.363,
+      "step": 4884
+    },
+    {
+      "epoch": 0.3908,
+      "grad_norm": 2.104926109313965,
+      "learning_rate": 6.689073947173835e-06,
+      "loss": 0.3872,
+      "step": 4885
+    },
+    {
+      "epoch": 0.39088,
+      "grad_norm": 1.384028434753418,
+      "learning_rate": 6.6878901841453135e-06,
+      "loss": 0.3222,
+      "step": 4886
+    },
+    {
+      "epoch": 0.39096,
+      "grad_norm": 1.4685722589492798,
+      "learning_rate": 6.6867063143296285e-06,
+      "loss": 0.3079,
+      "step": 4887
+    },
+    {
+      "epoch": 0.39104,
+      "grad_norm": 1.832414984703064,
+      "learning_rate": 6.685522337801678e-06,
+      "loss": 0.3503,
+      "step": 4888
+    },
+    {
+      "epoch": 0.39112,
+      "grad_norm": 1.4985992908477783,
+      "learning_rate": 6.68433825463637e-06,
+      "loss": 0.343,
+      "step": 4889
+    },
+    {
+      "epoch": 0.3912,
+      "grad_norm": 1.4093996286392212,
+      "learning_rate": 6.6831540649086165e-06,
+      "loss": 0.3069,
+      "step": 4890
+    },
+    {
+      "epoch": 0.39128,
+      "grad_norm": 1.5814567804336548,
+      "learning_rate": 6.681969768693336e-06,
+      "loss": 0.3293,
+      "step": 4891
+    },
+    {
+      "epoch": 0.39136,
+      "grad_norm": 1.4069743156433105,
+      "learning_rate": 6.680785366065458e-06,
+      "loss": 0.3107,
+      "step": 4892
+    },
+    {
+      "epoch": 0.39144,
+      "grad_norm": 1.08096182346344,
+      "learning_rate": 6.679600857099913e-06,
+      "loss": 0.2361,
+      "step": 4893
+    },
+    {
+      "epoch": 0.39152,
+      "grad_norm": 1.1872847080230713,
+      "learning_rate": 6.67841624187164e-06,
+      "loss": 0.3018,
+      "step": 4894
+    },
+    {
+      "epoch": 0.3916,
+      "grad_norm": 1.4051975011825562,
+      "learning_rate": 6.6772315204555896e-06,
+      "loss": 0.3002,
+      "step": 4895
+    },
+    {
+      "epoch": 0.39168,
+      "grad_norm": 1.3737307786941528,
+      "learning_rate": 6.676046692926712e-06,
+      "loss": 0.2727,
+      "step": 4896
+    },
+    {
+      "epoch": 0.39176,
+      "grad_norm": 1.3034765720367432,
+      "learning_rate": 6.674861759359967e-06,
+      "loss": 0.3119,
+      "step": 4897
+    },
+    {
+      "epoch": 0.39184,
+      "grad_norm": 1.5516210794448853,
+      "learning_rate": 6.6736767198303245e-06,
+      "loss": 0.2893,
+      "step": 4898
+    },
+    {
+      "epoch": 0.39192,
+      "grad_norm": 1.2979193925857544,
+      "learning_rate": 6.672491574412755e-06,
+      "loss": 0.2884,
+      "step": 4899
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 1.7181899547576904,
+      "learning_rate": 6.671306323182239e-06,
+      "loss": 0.3425,
+      "step": 4900
+    },
+    {
+      "epoch": 0.39208,
+      "grad_norm": 1.7885940074920654,
+      "learning_rate": 6.6701209662137665e-06,
+      "loss": 0.3883,
+      "step": 4901
+    },
+    {
+      "epoch": 0.39216,
+      "grad_norm": 2.0114247798919678,
+      "learning_rate": 6.668935503582328e-06,
+      "loss": 0.347,
+      "step": 4902
+    },
+    {
+      "epoch": 0.39224,
+      "grad_norm": 1.6095882654190063,
+      "learning_rate": 6.667749935362922e-06,
+      "loss": 0.3343,
+      "step": 4903
+    },
+    {
+      "epoch": 0.39232,
+      "grad_norm": 1.5267553329467773,
+      "learning_rate": 6.6665642616305615e-06,
+      "loss": 0.3258,
+      "step": 4904
+    },
+    {
+      "epoch": 0.3924,
+      "grad_norm": 1.5813835859298706,
+      "learning_rate": 6.665378482460254e-06,
+      "loss": 0.3566,
+      "step": 4905
+    },
+    {
+      "epoch": 0.39248,
+      "grad_norm": 1.3659838438034058,
+      "learning_rate": 6.664192597927024e-06,
+      "loss": 0.3428,
+      "step": 4906
+    },
+    {
+      "epoch": 0.39256,
+      "grad_norm": 1.4735257625579834,
+      "learning_rate": 6.663006608105897e-06,
+      "loss": 0.3298,
+      "step": 4907
+    },
+    {
+      "epoch": 0.39264,
+      "grad_norm": 1.72593092918396,
+      "learning_rate": 6.6618205130719034e-06,
+      "loss": 0.3582,
+      "step": 4908
+    },
+    {
+      "epoch": 0.39272,
+      "grad_norm": 1.6992379426956177,
+      "learning_rate": 6.660634312900089e-06,
+      "loss": 0.3377,
+      "step": 4909
+    },
+    {
+      "epoch": 0.3928,
+      "grad_norm": 1.6187288761138916,
+      "learning_rate": 6.659448007665496e-06,
+      "loss": 0.2841,
+      "step": 4910
+    },
+    {
+      "epoch": 0.39288,
+      "grad_norm": 1.6214263439178467,
+      "learning_rate": 6.658261597443182e-06,
+      "loss": 0.4277,
+      "step": 4911
+    },
+    {
+      "epoch": 0.39296,
+      "grad_norm": 1.5356061458587646,
+      "learning_rate": 6.657075082308204e-06,
+      "loss": 0.3734,
+      "step": 4912
+    },
+    {
+      "epoch": 0.39304,
+      "grad_norm": 1.5965813398361206,
+      "learning_rate": 6.655888462335631e-06,
+      "loss": 0.3108,
+      "step": 4913
+    },
+    {
+      "epoch": 0.39312,
+      "grad_norm": 1.5381505489349365,
+      "learning_rate": 6.654701737600536e-06,
+      "loss": 0.3088,
+      "step": 4914
+    },
+    {
+      "epoch": 0.3932,
+      "grad_norm": 1.7315890789031982,
+      "learning_rate": 6.653514908177999e-06,
+      "loss": 0.3554,
+      "step": 4915
+    },
+    {
+      "epoch": 0.39328,
+      "grad_norm": 1.5875256061553955,
+      "learning_rate": 6.652327974143106e-06,
+      "loss": 0.355,
+      "step": 4916
+    },
+    {
+      "epoch": 0.39336,
+      "grad_norm": 1.443566083908081,
+      "learning_rate": 6.65114093557095e-06,
+      "loss": 0.2864,
+      "step": 4917
+    },
+    {
+      "epoch": 0.39344,
+      "grad_norm": 1.1734968423843384,
+      "learning_rate": 6.649953792536632e-06,
+      "loss": 0.2207,
+      "step": 4918
+    },
+    {
+      "epoch": 0.39352,
+      "grad_norm": 1.580723524093628,
+      "learning_rate": 6.648766545115258e-06,
+      "loss": 0.3254,
+      "step": 4919
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 1.456050157546997,
+      "learning_rate": 6.647579193381942e-06,
+      "loss": 0.3356,
+      "step": 4920
+    },
+    {
+      "epoch": 0.39368,
+      "grad_norm": 0.9805746078491211,
+      "learning_rate": 6.646391737411803e-06,
+      "loss": 0.1802,
+      "step": 4921
+    },
+    {
+      "epoch": 0.39376,
+      "grad_norm": 1.4817471504211426,
+      "learning_rate": 6.645204177279968e-06,
+      "loss": 0.4042,
+      "step": 4922
+    },
+    {
+      "epoch": 0.39384,
+      "grad_norm": 1.3607851266860962,
+      "learning_rate": 6.644016513061569e-06,
+      "loss": 0.3145,
+      "step": 4923
+    },
+    {
+      "epoch": 0.39392,
+      "grad_norm": 1.8702746629714966,
+      "learning_rate": 6.642828744831746e-06,
+      "loss": 0.3977,
+      "step": 4924
+    },
+    {
+      "epoch": 0.394,
+      "grad_norm": 1.4964871406555176,
+      "learning_rate": 6.641640872665647e-06,
+      "loss": 0.3943,
+      "step": 4925
+    },
+    {
+      "epoch": 0.39408,
+      "grad_norm": 1.3008499145507812,
+      "learning_rate": 6.640452896638421e-06,
+      "loss": 0.3038,
+      "step": 4926
+    },
+    {
+      "epoch": 0.39416,
+      "grad_norm": 1.3478739261627197,
+      "learning_rate": 6.639264816825231e-06,
+      "loss": 0.3404,
+      "step": 4927
+    },
+    {
+      "epoch": 0.39424,
+      "grad_norm": 1.33133864402771,
+      "learning_rate": 6.638076633301239e-06,
+      "loss": 0.3305,
+      "step": 4928
+    },
+    {
+      "epoch": 0.39432,
+      "grad_norm": 1.773242712020874,
+      "learning_rate": 6.63688834614162e-06,
+      "loss": 0.3097,
+      "step": 4929
+    },
+    {
+      "epoch": 0.3944,
+      "grad_norm": 1.4814815521240234,
+      "learning_rate": 6.635699955421553e-06,
+      "loss": 0.3365,
+      "step": 4930
+    },
+    {
+      "epoch": 0.39448,
+      "grad_norm": 1.6367100477218628,
+      "learning_rate": 6.6345114612162235e-06,
+      "loss": 0.319,
+      "step": 4931
+    },
+    {
+      "epoch": 0.39456,
+      "grad_norm": 1.5399671792984009,
+      "learning_rate": 6.633322863600822e-06,
+      "loss": 0.2835,
+      "step": 4932
+    },
+    {
+      "epoch": 0.39464,
+      "grad_norm": 1.9397109746932983,
+      "learning_rate": 6.632134162650547e-06,
+      "loss": 0.3802,
+      "step": 4933
+    },
+    {
+      "epoch": 0.39472,
+      "grad_norm": 1.1391005516052246,
+      "learning_rate": 6.630945358440606e-06,
+      "loss": 0.2647,
+      "step": 4934
+    },
+    {
+      "epoch": 0.3948,
+      "grad_norm": 1.9419512748718262,
+      "learning_rate": 6.6297564510462075e-06,
+      "loss": 0.3465,
+      "step": 4935
+    },
+    {
+      "epoch": 0.39488,
+      "grad_norm": 1.6960084438323975,
+      "learning_rate": 6.6285674405425746e-06,
+      "loss": 0.3128,
+      "step": 4936
+    },
+    {
+      "epoch": 0.39496,
+      "grad_norm": 1.3205006122589111,
+      "learning_rate": 6.627378327004927e-06,
+      "loss": 0.2925,
+      "step": 4937
+    },
+    {
+      "epoch": 0.39504,
+      "grad_norm": 1.6510692834854126,
+      "learning_rate": 6.626189110508498e-06,
+      "loss": 0.3364,
+      "step": 4938
+    },
+    {
+      "epoch": 0.39512,
+      "grad_norm": 1.390600323677063,
+      "learning_rate": 6.624999791128527e-06,
+      "loss": 0.2794,
+      "step": 4939
+    },
+    {
+      "epoch": 0.3952,
+      "grad_norm": 1.7261474132537842,
+      "learning_rate": 6.623810368940254e-06,
+      "loss": 0.3829,
+      "step": 4940
+    },
+    {
+      "epoch": 0.39528,
+      "grad_norm": 1.63364839553833,
+      "learning_rate": 6.6226208440189334e-06,
+      "loss": 0.3381,
+      "step": 4941
+    },
+    {
+      "epoch": 0.39536,
+      "grad_norm": 2.4538915157318115,
+      "learning_rate": 6.621431216439822e-06,
+      "loss": 0.4066,
+      "step": 4942
+    },
+    {
+      "epoch": 0.39544,
+      "grad_norm": 1.6599690914154053,
+      "learning_rate": 6.620241486278181e-06,
+      "loss": 0.3646,
+      "step": 4943
+    },
+    {
+      "epoch": 0.39552,
+      "grad_norm": 1.3066409826278687,
+      "learning_rate": 6.619051653609284e-06,
+      "loss": 0.2685,
+      "step": 4944
+    },
+    {
+      "epoch": 0.3956,
+      "grad_norm": 1.3815221786499023,
+      "learning_rate": 6.6178617185084045e-06,
+      "loss": 0.2759,
+      "step": 4945
+    },
+    {
+      "epoch": 0.39568,
+      "grad_norm": 2.038173198699951,
+      "learning_rate": 6.616671681050829e-06,
+      "loss": 0.3673,
+      "step": 4946
+    },
+    {
+      "epoch": 0.39576,
+      "grad_norm": 1.204803466796875,
+      "learning_rate": 6.615481541311846e-06,
+      "loss": 0.2506,
+      "step": 4947
+    },
+    {
+      "epoch": 0.39584,
+      "grad_norm": 1.624482274055481,
+      "learning_rate": 6.614291299366751e-06,
+      "loss": 0.3633,
+      "step": 4948
+    },
+    {
+      "epoch": 0.39592,
+      "grad_norm": 1.5271668434143066,
+      "learning_rate": 6.613100955290845e-06,
+      "loss": 0.4242,
+      "step": 4949
+    },
+    {
+      "epoch": 0.396,
+      "grad_norm": 1.756809949874878,
+      "learning_rate": 6.611910509159443e-06,
+      "loss": 0.5438,
+      "step": 4950
+    },
+    {
+      "epoch": 0.39608,
+      "grad_norm": 1.434124231338501,
+      "learning_rate": 6.610719961047853e-06,
+      "loss": 0.2511,
+      "step": 4951
+    },
+    {
+      "epoch": 0.39616,
+      "grad_norm": 1.770772933959961,
+      "learning_rate": 6.609529311031402e-06,
+      "loss": 0.5268,
+      "step": 4952
+    },
+    {
+      "epoch": 0.39624,
+      "grad_norm": 1.617964267730713,
+      "learning_rate": 6.608338559185417e-06,
+      "loss": 0.3534,
+      "step": 4953
+    },
+    {
+      "epoch": 0.39632,
+      "grad_norm": 1.4315916299819946,
+      "learning_rate": 6.607147705585233e-06,
+      "loss": 0.2874,
+      "step": 4954
+    },
+    {
+      "epoch": 0.3964,
+      "grad_norm": 1.1115981340408325,
+      "learning_rate": 6.6059567503061905e-06,
+      "loss": 0.2664,
+      "step": 4955
+    },
+    {
+      "epoch": 0.39648,
+      "grad_norm": 1.2988982200622559,
+      "learning_rate": 6.604765693423637e-06,
+      "loss": 0.2593,
+      "step": 4956
+    },
+    {
+      "epoch": 0.39656,
+      "grad_norm": 1.7615128755569458,
+      "learning_rate": 6.6035745350129285e-06,
+      "loss": 0.4727,
+      "step": 4957
+    },
+    {
+      "epoch": 0.39664,
+      "grad_norm": 1.4541040658950806,
+      "learning_rate": 6.602383275149425e-06,
+      "loss": 0.3868,
+      "step": 4958
+    },
+    {
+      "epoch": 0.39672,
+      "grad_norm": 1.7334989309310913,
+      "learning_rate": 6.601191913908495e-06,
+      "loss": 0.3599,
+      "step": 4959
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 2.142695188522339,
+      "learning_rate": 6.600000451365508e-06,
+      "loss": 0.4301,
+      "step": 4960
+    },
+    {
+      "epoch": 0.39688,
+      "grad_norm": 1.5532326698303223,
+      "learning_rate": 6.598808887595847e-06,
+      "loss": 0.3237,
+      "step": 4961
+    },
+    {
+      "epoch": 0.39696,
+      "grad_norm": 1.5868544578552246,
+      "learning_rate": 6.5976172226748965e-06,
+      "loss": 0.3802,
+      "step": 4962
+    },
+    {
+      "epoch": 0.39704,
+      "grad_norm": 1.2728503942489624,
+      "learning_rate": 6.596425456678051e-06,
+      "loss": 0.3255,
+      "step": 4963
+    },
+    {
+      "epoch": 0.39712,
+      "grad_norm": 1.2911931276321411,
+      "learning_rate": 6.595233589680708e-06,
+      "loss": 0.2739,
+      "step": 4964
+    },
+    {
+      "epoch": 0.3972,
+      "grad_norm": 1.98240065574646,
+      "learning_rate": 6.594041621758272e-06,
+      "loss": 0.4007,
+      "step": 4965
+    },
+    {
+      "epoch": 0.39728,
+      "grad_norm": 1.2459379434585571,
+      "learning_rate": 6.592849552986159e-06,
+      "loss": 0.2915,
+      "step": 4966
+    },
+    {
+      "epoch": 0.39736,
+      "grad_norm": 1.5010486841201782,
+      "learning_rate": 6.591657383439782e-06,
+      "loss": 0.3646,
+      "step": 4967
+    },
+    {
+      "epoch": 0.39744,
+      "grad_norm": 1.6176382303237915,
+      "learning_rate": 6.590465113194569e-06,
+      "loss": 0.2956,
+      "step": 4968
+    },
+    {
+      "epoch": 0.39752,
+      "grad_norm": 1.6453156471252441,
+      "learning_rate": 6.58927274232595e-06,
+      "loss": 0.3022,
+      "step": 4969
+    },
+    {
+      "epoch": 0.3976,
+      "grad_norm": 1.58073091506958,
+      "learning_rate": 6.588080270909363e-06,
+      "loss": 0.3601,
+      "step": 4970
+    },
+    {
+      "epoch": 0.39768,
+      "grad_norm": 1.4092754125595093,
+      "learning_rate": 6.586887699020252e-06,
+      "loss": 0.3454,
+      "step": 4971
+    },
+    {
+      "epoch": 0.39776,
+      "grad_norm": 1.4815930128097534,
+      "learning_rate": 6.585695026734065e-06,
+      "loss": 0.3053,
+      "step": 4972
+    },
+    {
+      "epoch": 0.39784,
+      "grad_norm": 1.828636884689331,
+      "learning_rate": 6.584502254126258e-06,
+      "loss": 0.3526,
+      "step": 4973
+    },
+    {
+      "epoch": 0.39792,
+      "grad_norm": 1.1700973510742188,
+      "learning_rate": 6.583309381272296e-06,
+      "loss": 0.2726,
+      "step": 4974
+    },
+    {
+      "epoch": 0.398,
+      "grad_norm": 1.965246319770813,
+      "learning_rate": 6.582116408247647e-06,
+      "loss": 0.3759,
+      "step": 4975
+    },
+    {
+      "epoch": 0.39808,
+      "grad_norm": 1.8582028150558472,
+      "learning_rate": 6.580923335127787e-06,
+      "loss": 0.3591,
+      "step": 4976
+    },
+    {
+      "epoch": 0.39816,
+      "grad_norm": 1.556747317314148,
+      "learning_rate": 6.579730161988197e-06,
+      "loss": 0.401,
+      "step": 4977
+    },
+    {
+      "epoch": 0.39824,
+      "grad_norm": 1.0947948694229126,
+      "learning_rate": 6.578536888904367e-06,
+      "loss": 0.201,
+      "step": 4978
+    },
+    {
+      "epoch": 0.39832,
+      "grad_norm": 1.6996358633041382,
+      "learning_rate": 6.577343515951787e-06,
+      "loss": 0.3871,
+      "step": 4979
+    },
+    {
+      "epoch": 0.3984,
+      "grad_norm": 1.6404669284820557,
+      "learning_rate": 6.576150043205962e-06,
+      "loss": 0.3518,
+      "step": 4980
+    },
+    {
+      "epoch": 0.39848,
+      "grad_norm": 1.4954818487167358,
+      "learning_rate": 6.574956470742398e-06,
+      "loss": 0.2871,
+      "step": 4981
+    },
+    {
+      "epoch": 0.39856,
+      "grad_norm": 1.6011344194412231,
+      "learning_rate": 6.573762798636608e-06,
+      "loss": 0.3557,
+      "step": 4982
+    },
+    {
+      "epoch": 0.39864,
+      "grad_norm": 1.716171383857727,
+      "learning_rate": 6.57256902696411e-06,
+      "loss": 0.3298,
+      "step": 4983
+    },
+    {
+      "epoch": 0.39872,
+      "grad_norm": 1.5395945310592651,
+      "learning_rate": 6.571375155800434e-06,
+      "loss": 0.4424,
+      "step": 4984
+    },
+    {
+      "epoch": 0.3988,
+      "grad_norm": 1.4372494220733643,
+      "learning_rate": 6.570181185221109e-06,
+      "loss": 0.3702,
+      "step": 4985
+    },
+    {
+      "epoch": 0.39888,
+      "grad_norm": 1.6280698776245117,
+      "learning_rate": 6.5689871153016726e-06,
+      "loss": 0.3452,
+      "step": 4986
+    },
+    {
+      "epoch": 0.39896,
+      "grad_norm": 1.3518006801605225,
+      "learning_rate": 6.567792946117673e-06,
+      "loss": 0.2486,
+      "step": 4987
+    },
+    {
+      "epoch": 0.39904,
+      "grad_norm": 1.4701176881790161,
+      "learning_rate": 6.566598677744658e-06,
+      "loss": 0.3986,
+      "step": 4988
+    },
+    {
+      "epoch": 0.39912,
+      "grad_norm": 1.6428401470184326,
+      "learning_rate": 6.565404310258189e-06,
+      "loss": 0.3209,
+      "step": 4989
+    },
+    {
+      "epoch": 0.3992,
+      "grad_norm": 1.4474351406097412,
+      "learning_rate": 6.5642098437338245e-06,
+      "loss": 0.2907,
+      "step": 4990
+    },
+    {
+      "epoch": 0.39928,
+      "grad_norm": 1.8276543617248535,
+      "learning_rate": 6.563015278247138e-06,
+      "loss": 0.4284,
+      "step": 4991
+    },
+    {
+      "epoch": 0.39936,
+      "grad_norm": 1.533747673034668,
+      "learning_rate": 6.561820613873704e-06,
+      "loss": 0.3259,
+      "step": 4992
+    },
+    {
+      "epoch": 0.39944,
+      "grad_norm": 1.8883445262908936,
+      "learning_rate": 6.560625850689106e-06,
+      "loss": 0.3844,
+      "step": 4993
+    },
+    {
+      "epoch": 0.39952,
+      "grad_norm": 1.6395256519317627,
+      "learning_rate": 6.559430988768934e-06,
+      "loss": 0.348,
+      "step": 4994
+    },
+    {
+      "epoch": 0.3996,
+      "grad_norm": 1.8606520891189575,
+      "learning_rate": 6.55823602818878e-06,
+      "loss": 0.353,
+      "step": 4995
+    },
+    {
+      "epoch": 0.39968,
+      "grad_norm": 1.542927861213684,
+      "learning_rate": 6.557040969024246e-06,
+      "loss": 0.3696,
+      "step": 4996
+    },
+    {
+      "epoch": 0.39976,
+      "grad_norm": 1.8894641399383545,
+      "learning_rate": 6.55584581135094e-06,
+      "loss": 0.3741,
+      "step": 4997
+    },
+    {
+      "epoch": 0.39984,
+      "grad_norm": 1.6438285112380981,
+      "learning_rate": 6.554650555244476e-06,
+      "loss": 0.3712,
+      "step": 4998
+    },
+    {
+      "epoch": 0.39992,
+      "grad_norm": 1.3953157663345337,
+      "learning_rate": 6.5534552007804715e-06,
+      "loss": 0.3146,
+      "step": 4999
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.405379295349121,
+      "learning_rate": 6.5522597480345564e-06,
+      "loss": 0.6259,
+      "step": 5000
+    },
+    {
+      "epoch": 0.40008,
+      "grad_norm": 1.4833146333694458,
+      "learning_rate": 6.551064197082361e-06,
+      "loss": 0.3241,
+      "step": 5001
+    },
+    {
+      "epoch": 0.40016,
+      "grad_norm": 1.4547537565231323,
+      "learning_rate": 6.549868547999523e-06,
+      "loss": 0.3425,
+      "step": 5002
+    },
+    {
+      "epoch": 0.40024,
+      "grad_norm": 1.2077940702438354,
+      "learning_rate": 6.548672800861686e-06,
+      "loss": 0.2656,
+      "step": 5003
+    },
+    {
+      "epoch": 0.40032,
+      "grad_norm": 1.5668079853057861,
+      "learning_rate": 6.547476955744505e-06,
+      "loss": 0.3011,
+      "step": 5004
+    },
+    {
+      "epoch": 0.4004,
+      "grad_norm": 1.4270055294036865,
+      "learning_rate": 6.546281012723634e-06,
+      "loss": 0.2762,
+      "step": 5005
+    },
+    {
+      "epoch": 0.40048,
+      "grad_norm": 1.215239405632019,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.2474,
+      "step": 5006
+    },
+    {
+      "epoch": 0.40056,
+      "grad_norm": 1.7257269620895386,
+      "learning_rate": 6.543888833273486e-06,
+      "loss": 0.3579,
+      "step": 5007
+    },
+    {
+      "epoch": 0.40064,
+      "grad_norm": 1.6582391262054443,
+      "learning_rate": 6.5426925969955524e-06,
+      "loss": 0.3379,
+      "step": 5008
+    },
+    {
+      "epoch": 0.40072,
+      "grad_norm": 1.4435465335845947,
+      "learning_rate": 6.54149626311662e-06,
+      "loss": 0.3572,
+      "step": 5009
+    },
+    {
+      "epoch": 0.4008,
+      "grad_norm": 1.3516596555709839,
+      "learning_rate": 6.5402998317123765e-06,
+      "loss": 0.3044,
+      "step": 5010
+    },
+    {
+      "epoch": 0.40088,
+      "grad_norm": 1.3053381443023682,
+      "learning_rate": 6.539103302858517e-06,
+      "loss": 0.2901,
+      "step": 5011
+    },
+    {
+      "epoch": 0.40096,
+      "grad_norm": 1.8313065767288208,
+      "learning_rate": 6.537906676630741e-06,
+      "loss": 0.369,
+      "step": 5012
+    },
+    {
+      "epoch": 0.40104,
+      "grad_norm": 1.388274908065796,
+      "learning_rate": 6.536709953104756e-06,
+      "loss": 0.2789,
+      "step": 5013
+    },
+    {
+      "epoch": 0.40112,
+      "grad_norm": 2.075113296508789,
+      "learning_rate": 6.535513132356275e-06,
+      "loss": 0.5173,
+      "step": 5014
+    },
+    {
+      "epoch": 0.4012,
+      "grad_norm": 1.500080943107605,
+      "learning_rate": 6.534316214461014e-06,
+      "loss": 0.4011,
+      "step": 5015
+    },
+    {
+      "epoch": 0.40128,
+      "grad_norm": 1.7447412014007568,
+      "learning_rate": 6.533119199494702e-06,
+      "loss": 0.3899,
+      "step": 5016
+    },
+    {
+      "epoch": 0.40136,
+      "grad_norm": 1.5773173570632935,
+      "learning_rate": 6.531922087533067e-06,
+      "loss": 0.2934,
+      "step": 5017
+    },
+    {
+      "epoch": 0.40144,
+      "grad_norm": 1.36077880859375,
+      "learning_rate": 6.530724878651846e-06,
+      "loss": 0.2573,
+      "step": 5018
+    },
+    {
+      "epoch": 0.40152,
+      "grad_norm": 1.796042799949646,
+      "learning_rate": 6.5295275729267874e-06,
+      "loss": 0.3084,
+      "step": 5019
+    },
+    {
+      "epoch": 0.4016,
+      "grad_norm": 1.5335108041763306,
+      "learning_rate": 6.528330170433634e-06,
+      "loss": 0.3073,
+      "step": 5020
+    },
+    {
+      "epoch": 0.40168,
+      "grad_norm": 1.5497459173202515,
+      "learning_rate": 6.527132671248145e-06,
+      "loss": 0.3406,
+      "step": 5021
+    },
+    {
+      "epoch": 0.40176,
+      "grad_norm": 1.802089810371399,
+      "learning_rate": 6.525935075446081e-06,
+      "loss": 0.5411,
+      "step": 5022
+    },
+    {
+      "epoch": 0.40184,
+      "grad_norm": 1.59895658493042,
+      "learning_rate": 6.524737383103212e-06,
+      "loss": 0.3232,
+      "step": 5023
+    },
+    {
+      "epoch": 0.40192,
+      "grad_norm": 1.7338372468948364,
+      "learning_rate": 6.523539594295309e-06,
+      "loss": 0.3667,
+      "step": 5024
+    },
+    {
+      "epoch": 0.402,
+      "grad_norm": 1.4909459352493286,
+      "learning_rate": 6.5223417090981565e-06,
+      "loss": 0.3944,
+      "step": 5025
+    },
+    {
+      "epoch": 0.40208,
+      "grad_norm": 1.567449927330017,
+      "learning_rate": 6.521143727587536e-06,
+      "loss": 0.3354,
+      "step": 5026
+    },
+    {
+      "epoch": 0.40216,
+      "grad_norm": 1.2927497625350952,
+      "learning_rate": 6.519945649839241e-06,
+      "loss": 0.2782,
+      "step": 5027
+    },
+    {
+      "epoch": 0.40224,
+      "grad_norm": 1.6907589435577393,
+      "learning_rate": 6.518747475929073e-06,
+      "loss": 0.4177,
+      "step": 5028
+    },
+    {
+      "epoch": 0.40232,
+      "grad_norm": 1.5473442077636719,
+      "learning_rate": 6.517549205932832e-06,
+      "loss": 0.4238,
+      "step": 5029
+    },
+    {
+      "epoch": 0.4024,
+      "grad_norm": 1.1968157291412354,
+      "learning_rate": 6.51635083992633e-06,
+      "loss": 0.2689,
+      "step": 5030
+    },
+    {
+      "epoch": 0.40248,
+      "grad_norm": 1.7409216165542603,
+      "learning_rate": 6.515152377985385e-06,
+      "loss": 0.5057,
+      "step": 5031
+    },
+    {
+      "epoch": 0.40256,
+      "grad_norm": 1.6909297704696655,
+      "learning_rate": 6.513953820185819e-06,
+      "loss": 0.3212,
+      "step": 5032
+    },
+    {
+      "epoch": 0.40264,
+      "grad_norm": 1.6705437898635864,
+      "learning_rate": 6.512755166603459e-06,
+      "loss": 0.4307,
+      "step": 5033
+    },
+    {
+      "epoch": 0.40272,
+      "grad_norm": 1.2356233596801758,
+      "learning_rate": 6.511556417314142e-06,
+      "loss": 0.2553,
+      "step": 5034
+    },
+    {
+      "epoch": 0.4028,
+      "grad_norm": 1.6176491975784302,
+      "learning_rate": 6.510357572393709e-06,
+      "loss": 0.3523,
+      "step": 5035
+    },
+    {
+      "epoch": 0.40288,
+      "grad_norm": 1.7396360635757446,
+      "learning_rate": 6.509158631918006e-06,
+      "loss": 0.4734,
+      "step": 5036
+    },
+    {
+      "epoch": 0.40296,
+      "grad_norm": 1.2196110486984253,
+      "learning_rate": 6.507959595962885e-06,
+      "loss": 0.3238,
+      "step": 5037
+    },
+    {
+      "epoch": 0.40304,
+      "grad_norm": 1.266950249671936,
+      "learning_rate": 6.506760464604206e-06,
+      "loss": 0.2835,
+      "step": 5038
+    },
+    {
+      "epoch": 0.40312,
+      "grad_norm": 1.3458009958267212,
+      "learning_rate": 6.5055612379178355e-06,
+      "loss": 0.278,
+      "step": 5039
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 1.686802864074707,
+      "learning_rate": 6.504361915979643e-06,
+      "loss": 0.391,
+      "step": 5040
+    },
+    {
+      "epoch": 0.40328,
+      "grad_norm": 1.483590841293335,
+      "learning_rate": 6.503162498865504e-06,
+      "loss": 0.3368,
+      "step": 5041
+    },
+    {
+      "epoch": 0.40336,
+      "grad_norm": 1.6863470077514648,
+      "learning_rate": 6.501962986651305e-06,
+      "loss": 0.3795,
+      "step": 5042
+    },
+    {
+      "epoch": 0.40344,
+      "grad_norm": 1.6084065437316895,
+      "learning_rate": 6.500763379412932e-06,
+      "loss": 0.4195,
+      "step": 5043
+    },
+    {
+      "epoch": 0.40352,
+      "grad_norm": 1.6228187084197998,
+      "learning_rate": 6.499563677226281e-06,
+      "loss": 0.4213,
+      "step": 5044
+    },
+    {
+      "epoch": 0.4036,
+      "grad_norm": 1.7429604530334473,
+      "learning_rate": 6.498363880167256e-06,
+      "loss": 0.3857,
+      "step": 5045
+    },
+    {
+      "epoch": 0.40368,
+      "grad_norm": 0.9011700749397278,
+      "learning_rate": 6.49716398831176e-06,
+      "loss": 0.2047,
+      "step": 5046
+    },
+    {
+      "epoch": 0.40376,
+      "grad_norm": 1.3937594890594482,
+      "learning_rate": 6.4959640017357086e-06,
+      "loss": 0.331,
+      "step": 5047
+    },
+    {
+      "epoch": 0.40384,
+      "grad_norm": 1.7457072734832764,
+      "learning_rate": 6.494763920515021e-06,
+      "loss": 0.3449,
+      "step": 5048
+    },
+    {
+      "epoch": 0.40392,
+      "grad_norm": 1.6541334390640259,
+      "learning_rate": 6.493563744725621e-06,
+      "loss": 0.3061,
+      "step": 5049
+    },
+    {
+      "epoch": 0.404,
+      "grad_norm": 1.375489592552185,
+      "learning_rate": 6.49236347444344e-06,
+      "loss": 0.3041,
+      "step": 5050
+    },
+    {
+      "epoch": 0.40408,
+      "grad_norm": 1.7582788467407227,
+      "learning_rate": 6.491163109744416e-06,
+      "loss": 0.4728,
+      "step": 5051
+    },
+    {
+      "epoch": 0.40416,
+      "grad_norm": 1.3825554847717285,
+      "learning_rate": 6.489962650704491e-06,
+      "loss": 0.3534,
+      "step": 5052
+    },
+    {
+      "epoch": 0.40424,
+      "grad_norm": 1.3173959255218506,
+      "learning_rate": 6.4887620973996145e-06,
+      "loss": 0.2669,
+      "step": 5053
+    },
+    {
+      "epoch": 0.40432,
+      "grad_norm": 1.8339534997940063,
+      "learning_rate": 6.487561449905744e-06,
+      "loss": 0.3635,
+      "step": 5054
+    },
+    {
+      "epoch": 0.4044,
+      "grad_norm": 1.564780831336975,
+      "learning_rate": 6.4863607082988345e-06,
+      "loss": 0.3529,
+      "step": 5055
+    },
+    {
+      "epoch": 0.40448,
+      "grad_norm": 1.4042036533355713,
+      "learning_rate": 6.485159872654858e-06,
+      "loss": 0.2709,
+      "step": 5056
+    },
+    {
+      "epoch": 0.40456,
+      "grad_norm": 1.776026725769043,
+      "learning_rate": 6.483958943049785e-06,
+      "loss": 0.4495,
+      "step": 5057
+    },
+    {
+      "epoch": 0.40464,
+      "grad_norm": 1.5129555463790894,
+      "learning_rate": 6.482757919559594e-06,
+      "loss": 0.3174,
+      "step": 5058
+    },
+    {
+      "epoch": 0.40472,
+      "grad_norm": 1.493586540222168,
+      "learning_rate": 6.481556802260273e-06,
+      "loss": 0.3843,
+      "step": 5059
+    },
+    {
+      "epoch": 0.4048,
+      "grad_norm": 1.5436426401138306,
+      "learning_rate": 6.4803555912278106e-06,
+      "loss": 0.3938,
+      "step": 5060
+    },
+    {
+      "epoch": 0.40488,
+      "grad_norm": 1.4713767766952515,
+      "learning_rate": 6.4791542865382015e-06,
+      "loss": 0.3512,
+      "step": 5061
+    },
+    {
+      "epoch": 0.40496,
+      "grad_norm": 1.3419086933135986,
+      "learning_rate": 6.477952888267451e-06,
+      "loss": 0.3038,
+      "step": 5062
+    },
+    {
+      "epoch": 0.40504,
+      "grad_norm": 1.6907343864440918,
+      "learning_rate": 6.476751396491566e-06,
+      "loss": 0.3195,
+      "step": 5063
+    },
+    {
+      "epoch": 0.40512,
+      "grad_norm": 1.6713168621063232,
+      "learning_rate": 6.475549811286561e-06,
+      "loss": 0.3558,
+      "step": 5064
+    },
+    {
+      "epoch": 0.4052,
+      "grad_norm": 1.596295714378357,
+      "learning_rate": 6.474348132728457e-06,
+      "loss": 0.3346,
+      "step": 5065
+    },
+    {
+      "epoch": 0.40528,
+      "grad_norm": 1.508131980895996,
+      "learning_rate": 6.473146360893281e-06,
+      "loss": 0.341,
+      "step": 5066
+    },
+    {
+      "epoch": 0.40536,
+      "grad_norm": 1.456239104270935,
+      "learning_rate": 6.4719444958570635e-06,
+      "loss": 0.3222,
+      "step": 5067
+    },
+    {
+      "epoch": 0.40544,
+      "grad_norm": 1.7773851156234741,
+      "learning_rate": 6.470742537695842e-06,
+      "loss": 0.4088,
+      "step": 5068
+    },
+    {
+      "epoch": 0.40552,
+      "grad_norm": 1.5959703922271729,
+      "learning_rate": 6.469540486485662e-06,
+      "loss": 0.3296,
+      "step": 5069
+    },
+    {
+      "epoch": 0.4056,
+      "grad_norm": 1.6706092357635498,
+      "learning_rate": 6.468338342302571e-06,
+      "loss": 0.3701,
+      "step": 5070
+    },
+    {
+      "epoch": 0.40568,
+      "grad_norm": 1.721103549003601,
+      "learning_rate": 6.467136105222627e-06,
+      "loss": 0.3396,
+      "step": 5071
+    },
+    {
+      "epoch": 0.40576,
+      "grad_norm": 1.8070697784423828,
+      "learning_rate": 6.465933775321891e-06,
+      "loss": 0.3382,
+      "step": 5072
+    },
+    {
+      "epoch": 0.40584,
+      "grad_norm": 1.7369276285171509,
+      "learning_rate": 6.46473135267643e-06,
+      "loss": 0.3338,
+      "step": 5073
+    },
+    {
+      "epoch": 0.40592,
+      "grad_norm": 1.6247727870941162,
+      "learning_rate": 6.463528837362319e-06,
+      "loss": 0.3429,
+      "step": 5074
+    },
+    {
+      "epoch": 0.406,
+      "grad_norm": 1.6333701610565186,
+      "learning_rate": 6.4623262294556335e-06,
+      "loss": 0.4485,
+      "step": 5075
+    },
+    {
+      "epoch": 0.40608,
+      "grad_norm": 1.7924059629440308,
+      "learning_rate": 6.46112352903246e-06,
+      "loss": 0.2964,
+      "step": 5076
+    },
+    {
+      "epoch": 0.40616,
+      "grad_norm": 1.6652660369873047,
+      "learning_rate": 6.4599207361688895e-06,
+      "loss": 0.389,
+      "step": 5077
+    },
+    {
+      "epoch": 0.40624,
+      "grad_norm": 1.426188349723816,
+      "learning_rate": 6.458717850941021e-06,
+      "loss": 0.286,
+      "step": 5078
+    },
+    {
+      "epoch": 0.40632,
+      "grad_norm": 1.797730565071106,
+      "learning_rate": 6.4575148734249535e-06,
+      "loss": 0.4004,
+      "step": 5079
+    },
+    {
+      "epoch": 0.4064,
+      "grad_norm": 1.5713727474212646,
+      "learning_rate": 6.456311803696797e-06,
+      "loss": 0.3227,
+      "step": 5080
+    },
+    {
+      "epoch": 0.40648,
+      "grad_norm": 1.5011001825332642,
+      "learning_rate": 6.455108641832666e-06,
+      "loss": 0.3584,
+      "step": 5081
+    },
+    {
+      "epoch": 0.40656,
+      "grad_norm": 1.5323615074157715,
+      "learning_rate": 6.453905387908679e-06,
+      "loss": 0.3302,
+      "step": 5082
+    },
+    {
+      "epoch": 0.40664,
+      "grad_norm": 1.2687554359436035,
+      "learning_rate": 6.452702042000964e-06,
+      "loss": 0.2704,
+      "step": 5083
+    },
+    {
+      "epoch": 0.40672,
+      "grad_norm": 1.6853008270263672,
+      "learning_rate": 6.4514986041856506e-06,
+      "loss": 0.3229,
+      "step": 5084
+    },
+    {
+      "epoch": 0.4068,
+      "grad_norm": 1.6804182529449463,
+      "learning_rate": 6.450295074538879e-06,
+      "loss": 0.3343,
+      "step": 5085
+    },
+    {
+      "epoch": 0.40688,
+      "grad_norm": 1.2748759984970093,
+      "learning_rate": 6.449091453136789e-06,
+      "loss": 0.2909,
+      "step": 5086
+    },
+    {
+      "epoch": 0.40696,
+      "grad_norm": 1.752908706665039,
+      "learning_rate": 6.447887740055532e-06,
+      "loss": 0.3041,
+      "step": 5087
+    },
+    {
+      "epoch": 0.40704,
+      "grad_norm": 1.6320358514785767,
+      "learning_rate": 6.446683935371262e-06,
+      "loss": 0.3439,
+      "step": 5088
+    },
+    {
+      "epoch": 0.40712,
+      "grad_norm": 1.5598769187927246,
+      "learning_rate": 6.445480039160141e-06,
+      "loss": 0.3354,
+      "step": 5089
+    },
+    {
+      "epoch": 0.4072,
+      "grad_norm": 1.6841180324554443,
+      "learning_rate": 6.444276051498334e-06,
+      "loss": 0.3556,
+      "step": 5090
+    },
+    {
+      "epoch": 0.40728,
+      "grad_norm": 1.4227954149246216,
+      "learning_rate": 6.443071972462013e-06,
+      "loss": 0.331,
+      "step": 5091
+    },
+    {
+      "epoch": 0.40736,
+      "grad_norm": 1.5023025274276733,
+      "learning_rate": 6.441867802127357e-06,
+      "loss": 0.3008,
+      "step": 5092
+    },
+    {
+      "epoch": 0.40744,
+      "grad_norm": 1.3716212511062622,
+      "learning_rate": 6.44066354057055e-06,
+      "loss": 0.313,
+      "step": 5093
+    },
+    {
+      "epoch": 0.40752,
+      "grad_norm": 1.4322949647903442,
+      "learning_rate": 6.43945918786778e-06,
+      "loss": 0.3185,
+      "step": 5094
+    },
+    {
+      "epoch": 0.4076,
+      "grad_norm": 1.3301920890808105,
+      "learning_rate": 6.438254744095247e-06,
+      "loss": 0.2555,
+      "step": 5095
+    },
+    {
+      "epoch": 0.40768,
+      "grad_norm": 1.4562873840332031,
+      "learning_rate": 6.437050209329147e-06,
+      "loss": 0.2832,
+      "step": 5096
+    },
+    {
+      "epoch": 0.40776,
+      "grad_norm": 2.004117965698242,
+      "learning_rate": 6.435845583645688e-06,
+      "loss": 0.3892,
+      "step": 5097
+    },
+    {
+      "epoch": 0.40784,
+      "grad_norm": 1.5688518285751343,
+      "learning_rate": 6.434640867121084e-06,
+      "loss": 0.372,
+      "step": 5098
+    },
+    {
+      "epoch": 0.40792,
+      "grad_norm": 1.229386806488037,
+      "learning_rate": 6.433436059831552e-06,
+      "loss": 0.2306,
+      "step": 5099
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 1.6038435697555542,
+      "learning_rate": 6.432231161853317e-06,
+      "loss": 0.3113,
+      "step": 5100
+    },
+    {
+      "epoch": 0.40808,
+      "grad_norm": 1.4195326566696167,
+      "learning_rate": 6.431026173262609e-06,
+      "loss": 0.3563,
+      "step": 5101
+    },
+    {
+      "epoch": 0.40816,
+      "grad_norm": 1.3468478918075562,
+      "learning_rate": 6.429821094135663e-06,
+      "loss": 0.3255,
+      "step": 5102
+    },
+    {
+      "epoch": 0.40824,
+      "grad_norm": 1.5173962116241455,
+      "learning_rate": 6.42861592454872e-06,
+      "loss": 0.3577,
+      "step": 5103
+    },
+    {
+      "epoch": 0.40832,
+      "grad_norm": 1.4830057621002197,
+      "learning_rate": 6.427410664578029e-06,
+      "loss": 0.3339,
+      "step": 5104
+    },
+    {
+      "epoch": 0.4084,
+      "grad_norm": 1.4850959777832031,
+      "learning_rate": 6.42620531429984e-06,
+      "loss": 0.3405,
+      "step": 5105
+    },
+    {
+      "epoch": 0.40848,
+      "grad_norm": 1.410702109336853,
+      "learning_rate": 6.424999873790414e-06,
+      "loss": 0.3099,
+      "step": 5106
+    },
+    {
+      "epoch": 0.40856,
+      "grad_norm": 1.793792486190796,
+      "learning_rate": 6.423794343126013e-06,
+      "loss": 0.3515,
+      "step": 5107
+    },
+    {
+      "epoch": 0.40864,
+      "grad_norm": 1.6600618362426758,
+      "learning_rate": 6.422588722382909e-06,
+      "loss": 0.2998,
+      "step": 5108
+    },
+    {
+      "epoch": 0.40872,
+      "grad_norm": 1.887887954711914,
+      "learning_rate": 6.4213830116373766e-06,
+      "loss": 0.3906,
+      "step": 5109
+    },
+    {
+      "epoch": 0.4088,
+      "grad_norm": 1.5626739263534546,
+      "learning_rate": 6.4201772109656956e-06,
+      "loss": 0.3316,
+      "step": 5110
+    },
+    {
+      "epoch": 0.40888,
+      "grad_norm": 1.23286771774292,
+      "learning_rate": 6.418971320444155e-06,
+      "loss": 0.3091,
+      "step": 5111
+    },
+    {
+      "epoch": 0.40896,
+      "grad_norm": 1.5620896816253662,
+      "learning_rate": 6.417765340149047e-06,
+      "loss": 0.3231,
+      "step": 5112
+    },
+    {
+      "epoch": 0.40904,
+      "grad_norm": 1.6344822645187378,
+      "learning_rate": 6.41655927015667e-06,
+      "loss": 0.4167,
+      "step": 5113
+    },
+    {
+      "epoch": 0.40912,
+      "grad_norm": 2.126511335372925,
+      "learning_rate": 6.4153531105433276e-06,
+      "loss": 0.4323,
+      "step": 5114
+    },
+    {
+      "epoch": 0.4092,
+      "grad_norm": 1.6435238122940063,
+      "learning_rate": 6.4141468613853295e-06,
+      "loss": 0.3263,
+      "step": 5115
+    },
+    {
+      "epoch": 0.40928,
+      "grad_norm": 1.6907066106796265,
+      "learning_rate": 6.412940522758992e-06,
+      "loss": 0.3561,
+      "step": 5116
+    },
+    {
+      "epoch": 0.40936,
+      "grad_norm": 1.7472286224365234,
+      "learning_rate": 6.411734094740634e-06,
+      "loss": 0.375,
+      "step": 5117
+    },
+    {
+      "epoch": 0.40944,
+      "grad_norm": 1.2598026990890503,
+      "learning_rate": 6.4105275774065846e-06,
+      "loss": 0.2562,
+      "step": 5118
+    },
+    {
+      "epoch": 0.40952,
+      "grad_norm": 1.2201982736587524,
+      "learning_rate": 6.409320970833176e-06,
+      "loss": 0.271,
+      "step": 5119
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 2.0412731170654297,
+      "learning_rate": 6.408114275096743e-06,
+      "loss": 0.3964,
+      "step": 5120
+    },
+    {
+      "epoch": 0.40968,
+      "grad_norm": 1.8842438459396362,
+      "learning_rate": 6.406907490273633e-06,
+      "loss": 0.427,
+      "step": 5121
+    },
+    {
+      "epoch": 0.40976,
+      "grad_norm": 1.1666141748428345,
+      "learning_rate": 6.405700616440191e-06,
+      "loss": 0.2608,
+      "step": 5122
+    },
+    {
+      "epoch": 0.40984,
+      "grad_norm": 1.613979697227478,
+      "learning_rate": 6.404493653672776e-06,
+      "loss": 0.34,
+      "step": 5123
+    },
+    {
+      "epoch": 0.40992,
+      "grad_norm": 1.5906524658203125,
+      "learning_rate": 6.403286602047748e-06,
+      "loss": 0.4057,
+      "step": 5124
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3950166702270508,
+      "learning_rate": 6.4020794616414725e-06,
+      "loss": 0.3133,
+      "step": 5125
+    },
+    {
+      "epoch": 0.41008,
+      "grad_norm": 1.232621431350708,
+      "learning_rate": 6.40087223253032e-06,
+      "loss": 0.2786,
+      "step": 5126
+    },
+    {
+      "epoch": 0.41016,
+      "grad_norm": 1.6018669605255127,
+      "learning_rate": 6.3996649147906675e-06,
+      "loss": 0.2908,
+      "step": 5127
+    },
+    {
+      "epoch": 0.41024,
+      "grad_norm": 1.5088248252868652,
+      "learning_rate": 6.3984575084989e-06,
+      "loss": 0.2802,
+      "step": 5128
+    },
+    {
+      "epoch": 0.41032,
+      "grad_norm": 1.2472470998764038,
+      "learning_rate": 6.3972500137314066e-06,
+      "loss": 0.26,
+      "step": 5129
+    },
+    {
+      "epoch": 0.4104,
+      "grad_norm": 1.7943204641342163,
+      "learning_rate": 6.396042430564577e-06,
+      "loss": 0.3852,
+      "step": 5130
+    },
+    {
+      "epoch": 0.41048,
+      "grad_norm": 1.9341444969177246,
+      "learning_rate": 6.394834759074817e-06,
+      "loss": 0.392,
+      "step": 5131
+    },
+    {
+      "epoch": 0.41056,
+      "grad_norm": 1.7144229412078857,
+      "learning_rate": 6.393626999338527e-06,
+      "loss": 0.3879,
+      "step": 5132
+    },
+    {
+      "epoch": 0.41064,
+      "grad_norm": 1.494651198387146,
+      "learning_rate": 6.392419151432121e-06,
+      "loss": 0.3137,
+      "step": 5133
+    },
+    {
+      "epoch": 0.41072,
+      "grad_norm": 1.3737579584121704,
+      "learning_rate": 6.3912112154320135e-06,
+      "loss": 0.3458,
+      "step": 5134
+    },
+    {
+      "epoch": 0.4108,
+      "grad_norm": 1.907779574394226,
+      "learning_rate": 6.390003191414627e-06,
+      "loss": 0.368,
+      "step": 5135
+    },
+    {
+      "epoch": 0.41088,
+      "grad_norm": 1.4840914011001587,
+      "learning_rate": 6.388795079456392e-06,
+      "loss": 0.3191,
+      "step": 5136
+    },
+    {
+      "epoch": 0.41096,
+      "grad_norm": 1.8090542554855347,
+      "learning_rate": 6.387586879633736e-06,
+      "loss": 0.366,
+      "step": 5137
+    },
+    {
+      "epoch": 0.41104,
+      "grad_norm": 1.704310417175293,
+      "learning_rate": 6.386378592023103e-06,
+      "loss": 0.3767,
+      "step": 5138
+    },
+    {
+      "epoch": 0.41112,
+      "grad_norm": 1.956494927406311,
+      "learning_rate": 6.385170216700934e-06,
+      "loss": 0.3635,
+      "step": 5139
+    },
+    {
+      "epoch": 0.4112,
+      "grad_norm": 1.283545732498169,
+      "learning_rate": 6.383961753743681e-06,
+      "loss": 0.2738,
+      "step": 5140
+    },
+    {
+      "epoch": 0.41128,
+      "grad_norm": 1.40552818775177,
+      "learning_rate": 6.382753203227799e-06,
+      "loss": 0.3511,
+      "step": 5141
+    },
+    {
+      "epoch": 0.41136,
+      "grad_norm": 1.738274335861206,
+      "learning_rate": 6.381544565229749e-06,
+      "loss": 0.4043,
+      "step": 5142
+    },
+    {
+      "epoch": 0.41144,
+      "grad_norm": 1.6897915601730347,
+      "learning_rate": 6.380335839825996e-06,
+      "loss": 0.3846,
+      "step": 5143
+    },
+    {
+      "epoch": 0.41152,
+      "grad_norm": 1.5609641075134277,
+      "learning_rate": 6.379127027093013e-06,
+      "loss": 0.3149,
+      "step": 5144
+    },
+    {
+      "epoch": 0.4116,
+      "grad_norm": 1.7952966690063477,
+      "learning_rate": 6.377918127107277e-06,
+      "loss": 0.3799,
+      "step": 5145
+    },
+    {
+      "epoch": 0.41168,
+      "grad_norm": 1.743720293045044,
+      "learning_rate": 6.376709139945273e-06,
+      "loss": 0.3468,
+      "step": 5146
+    },
+    {
+      "epoch": 0.41176,
+      "grad_norm": 1.60451078414917,
+      "learning_rate": 6.375500065683486e-06,
+      "loss": 0.3882,
+      "step": 5147
+    },
+    {
+      "epoch": 0.41184,
+      "grad_norm": 1.7727147340774536,
+      "learning_rate": 6.374290904398416e-06,
+      "loss": 0.367,
+      "step": 5148
+    },
+    {
+      "epoch": 0.41192,
+      "grad_norm": 1.4645249843597412,
+      "learning_rate": 6.373081656166557e-06,
+      "loss": 0.3483,
+      "step": 5149
+    },
+    {
+      "epoch": 0.412,
+      "grad_norm": 1.3234875202178955,
+      "learning_rate": 6.371872321064414e-06,
+      "loss": 0.2924,
+      "step": 5150
+    },
+    {
+      "epoch": 0.41208,
+      "grad_norm": 1.859861969947815,
+      "learning_rate": 6.370662899168501e-06,
+      "loss": 0.3757,
+      "step": 5151
+    },
+    {
+      "epoch": 0.41216,
+      "grad_norm": 1.227710485458374,
+      "learning_rate": 6.3694533905553346e-06,
+      "loss": 0.3016,
+      "step": 5152
+    },
+    {
+      "epoch": 0.41224,
+      "grad_norm": 1.643816351890564,
+      "learning_rate": 6.368243795301432e-06,
+      "loss": 0.3616,
+      "step": 5153
+    },
+    {
+      "epoch": 0.41232,
+      "grad_norm": 1.3771861791610718,
+      "learning_rate": 6.3670341134833235e-06,
+      "loss": 0.3275,
+      "step": 5154
+    },
+    {
+      "epoch": 0.4124,
+      "grad_norm": 1.796872854232788,
+      "learning_rate": 6.3658243451775404e-06,
+      "loss": 0.3235,
+      "step": 5155
+    },
+    {
+      "epoch": 0.41248,
+      "grad_norm": 1.5238722562789917,
+      "learning_rate": 6.3646144904606214e-06,
+      "loss": 0.3444,
+      "step": 5156
+    },
+    {
+      "epoch": 0.41256,
+      "grad_norm": 1.4715383052825928,
+      "learning_rate": 6.363404549409109e-06,
+      "loss": 0.3818,
+      "step": 5157
+    },
+    {
+      "epoch": 0.41264,
+      "grad_norm": 1.8105343580245972,
+      "learning_rate": 6.362194522099553e-06,
+      "loss": 0.4152,
+      "step": 5158
+    },
+    {
+      "epoch": 0.41272,
+      "grad_norm": 1.5102286338806152,
+      "learning_rate": 6.360984408608507e-06,
+      "loss": 0.3434,
+      "step": 5159
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 1.4143092632293701,
+      "learning_rate": 6.359774209012532e-06,
+      "loss": 0.3415,
+      "step": 5160
+    },
+    {
+      "epoch": 0.41288,
+      "grad_norm": 1.791418433189392,
+      "learning_rate": 6.358563923388194e-06,
+      "loss": 0.3569,
+      "step": 5161
+    },
+    {
+      "epoch": 0.41296,
+      "grad_norm": 1.825252890586853,
+      "learning_rate": 6.35735355181206e-06,
+      "loss": 0.3656,
+      "step": 5162
+    },
+    {
+      "epoch": 0.41304,
+      "grad_norm": 1.6053016185760498,
+      "learning_rate": 6.3561430943607105e-06,
+      "loss": 0.3853,
+      "step": 5163
+    },
+    {
+      "epoch": 0.41312,
+      "grad_norm": 1.2607011795043945,
+      "learning_rate": 6.354932551110724e-06,
+      "loss": 0.259,
+      "step": 5164
+    },
+    {
+      "epoch": 0.4132,
+      "grad_norm": 1.842266321182251,
+      "learning_rate": 6.3537219221386885e-06,
+      "loss": 0.4299,
+      "step": 5165
+    },
+    {
+      "epoch": 0.41328,
+      "grad_norm": 1.4322572946548462,
+      "learning_rate": 6.352511207521197e-06,
+      "loss": 0.3413,
+      "step": 5166
+    },
+    {
+      "epoch": 0.41336,
+      "grad_norm": 1.5515720844268799,
+      "learning_rate": 6.3513004073348465e-06,
+      "loss": 0.2727,
+      "step": 5167
+    },
+    {
+      "epoch": 0.41344,
+      "grad_norm": 1.2557002305984497,
+      "learning_rate": 6.35008952165624e-06,
+      "loss": 0.3055,
+      "step": 5168
+    },
+    {
+      "epoch": 0.41352,
+      "grad_norm": 1.6650270223617554,
+      "learning_rate": 6.3488785505619875e-06,
+      "loss": 0.3785,
+      "step": 5169
+    },
+    {
+      "epoch": 0.4136,
+      "grad_norm": 1.5979880094528198,
+      "learning_rate": 6.347667494128702e-06,
+      "loss": 0.3962,
+      "step": 5170
+    },
+    {
+      "epoch": 0.41368,
+      "grad_norm": 2.0301644802093506,
+      "learning_rate": 6.3464563524330034e-06,
+      "loss": 0.3584,
+      "step": 5171
+    },
+    {
+      "epoch": 0.41376,
+      "grad_norm": 1.6554670333862305,
+      "learning_rate": 6.345245125551518e-06,
+      "loss": 0.3181,
+      "step": 5172
+    },
+    {
+      "epoch": 0.41384,
+      "grad_norm": 1.6080145835876465,
+      "learning_rate": 6.344033813560875e-06,
+      "loss": 0.5606,
+      "step": 5173
+    },
+    {
+      "epoch": 0.41392,
+      "grad_norm": 1.3876444101333618,
+      "learning_rate": 6.342822416537708e-06,
+      "loss": 0.4102,
+      "step": 5174
+    },
+    {
+      "epoch": 0.414,
+      "grad_norm": 1.4370923042297363,
+      "learning_rate": 6.341610934558662e-06,
+      "loss": 0.3051,
+      "step": 5175
+    },
+    {
+      "epoch": 0.41408,
+      "grad_norm": 1.7648351192474365,
+      "learning_rate": 6.340399367700379e-06,
+      "loss": 0.4381,
+      "step": 5176
+    },
+    {
+      "epoch": 0.41416,
+      "grad_norm": 1.2672674655914307,
+      "learning_rate": 6.339187716039514e-06,
+      "loss": 0.2934,
+      "step": 5177
+    },
+    {
+      "epoch": 0.41424,
+      "grad_norm": 1.697166085243225,
+      "learning_rate": 6.337975979652723e-06,
+      "loss": 0.3332,
+      "step": 5178
+    },
+    {
+      "epoch": 0.41432,
+      "grad_norm": 1.230234980583191,
+      "learning_rate": 6.336764158616669e-06,
+      "loss": 0.302,
+      "step": 5179
+    },
+    {
+      "epoch": 0.4144,
+      "grad_norm": 1.0796769857406616,
+      "learning_rate": 6.335552253008018e-06,
+      "loss": 0.2222,
+      "step": 5180
+    },
+    {
+      "epoch": 0.41448,
+      "grad_norm": 1.8272526264190674,
+      "learning_rate": 6.3343402629034456e-06,
+      "loss": 0.4759,
+      "step": 5181
+    },
+    {
+      "epoch": 0.41456,
+      "grad_norm": 2.0101819038391113,
+      "learning_rate": 6.333128188379629e-06,
+      "loss": 0.4156,
+      "step": 5182
+    },
+    {
+      "epoch": 0.41464,
+      "grad_norm": 1.8950108289718628,
+      "learning_rate": 6.3319160295132544e-06,
+      "loss": 0.3522,
+      "step": 5183
+    },
+    {
+      "epoch": 0.41472,
+      "grad_norm": 1.547398328781128,
+      "learning_rate": 6.3307037863810075e-06,
+      "loss": 0.274,
+      "step": 5184
+    },
+    {
+      "epoch": 0.4148,
+      "grad_norm": 1.70745849609375,
+      "learning_rate": 6.329491459059584e-06,
+      "loss": 0.3314,
+      "step": 5185
+    },
+    {
+      "epoch": 0.41488,
+      "grad_norm": 1.835158348083496,
+      "learning_rate": 6.328279047625687e-06,
+      "loss": 0.4241,
+      "step": 5186
+    },
+    {
+      "epoch": 0.41496,
+      "grad_norm": 1.7435204982757568,
+      "learning_rate": 6.327066552156018e-06,
+      "loss": 0.4203,
+      "step": 5187
+    },
+    {
+      "epoch": 0.41504,
+      "grad_norm": 1.393307089805603,
+      "learning_rate": 6.325853972727288e-06,
+      "loss": 0.2993,
+      "step": 5188
+    },
+    {
+      "epoch": 0.41512,
+      "grad_norm": 1.5650931596755981,
+      "learning_rate": 6.324641309416215e-06,
+      "loss": 0.3302,
+      "step": 5189
+    },
+    {
+      "epoch": 0.4152,
+      "grad_norm": 1.2271121740341187,
+      "learning_rate": 6.323428562299516e-06,
+      "loss": 0.2451,
+      "step": 5190
+    },
+    {
+      "epoch": 0.41528,
+      "grad_norm": 1.5217052698135376,
+      "learning_rate": 6.322215731453922e-06,
+      "loss": 0.3315,
+      "step": 5191
+    },
+    {
+      "epoch": 0.41536,
+      "grad_norm": 1.3272700309753418,
+      "learning_rate": 6.321002816956162e-06,
+      "loss": 0.273,
+      "step": 5192
+    },
+    {
+      "epoch": 0.41544,
+      "grad_norm": 1.313028335571289,
+      "learning_rate": 6.3197898188829736e-06,
+      "loss": 0.2515,
+      "step": 5193
+    },
+    {
+      "epoch": 0.41552,
+      "grad_norm": 1.6728700399398804,
+      "learning_rate": 6.3185767373111e-06,
+      "loss": 0.4031,
+      "step": 5194
+    },
+    {
+      "epoch": 0.4156,
+      "grad_norm": 1.6574065685272217,
+      "learning_rate": 6.317363572317289e-06,
+      "loss": 0.3371,
+      "step": 5195
+    },
+    {
+      "epoch": 0.41568,
+      "grad_norm": 1.8870322704315186,
+      "learning_rate": 6.316150323978291e-06,
+      "loss": 0.3752,
+      "step": 5196
+    },
+    {
+      "epoch": 0.41576,
+      "grad_norm": 1.766973853111267,
+      "learning_rate": 6.3149369923708656e-06,
+      "loss": 0.3786,
+      "step": 5197
+    },
+    {
+      "epoch": 0.41584,
+      "grad_norm": 1.1528550386428833,
+      "learning_rate": 6.313723577571779e-06,
+      "loss": 0.2809,
+      "step": 5198
+    },
+    {
+      "epoch": 0.41592,
+      "grad_norm": 1.3488599061965942,
+      "learning_rate": 6.3125100796577956e-06,
+      "loss": 0.3158,
+      "step": 5199
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 1.634043574333191,
+      "learning_rate": 6.311296498705691e-06,
+      "loss": 0.3833,
+      "step": 5200
+    },
+    {
+      "epoch": 0.41608,
+      "grad_norm": 1.544848084449768,
+      "learning_rate": 6.310082834792246e-06,
+      "loss": 0.3798,
+      "step": 5201
+    },
+    {
+      "epoch": 0.41616,
+      "grad_norm": 1.647979736328125,
+      "learning_rate": 6.308869087994243e-06,
+      "loss": 0.3992,
+      "step": 5202
+    },
+    {
+      "epoch": 0.41624,
+      "grad_norm": 1.9953075647354126,
+      "learning_rate": 6.307655258388471e-06,
+      "loss": 0.4337,
+      "step": 5203
+    },
+    {
+      "epoch": 0.41632,
+      "grad_norm": 1.6946043968200684,
+      "learning_rate": 6.306441346051727e-06,
+      "loss": 0.4,
+      "step": 5204
+    },
+    {
+      "epoch": 0.4164,
+      "grad_norm": 1.4132634401321411,
+      "learning_rate": 6.305227351060809e-06,
+      "loss": 0.2904,
+      "step": 5205
+    },
+    {
+      "epoch": 0.41648,
+      "grad_norm": 1.3155604600906372,
+      "learning_rate": 6.304013273492526e-06,
+      "loss": 0.2809,
+      "step": 5206
+    },
+    {
+      "epoch": 0.41656,
+      "grad_norm": 1.562078595161438,
+      "learning_rate": 6.302799113423686e-06,
+      "loss": 0.323,
+      "step": 5207
+    },
+    {
+      "epoch": 0.41664,
+      "grad_norm": 1.5895940065383911,
+      "learning_rate": 6.3015848709311055e-06,
+      "loss": 0.332,
+      "step": 5208
+    },
+    {
+      "epoch": 0.41672,
+      "grad_norm": 1.8904573917388916,
+      "learning_rate": 6.300370546091605e-06,
+      "loss": 0.4672,
+      "step": 5209
+    },
+    {
+      "epoch": 0.4168,
+      "grad_norm": 1.3617215156555176,
+      "learning_rate": 6.299156138982011e-06,
+      "loss": 0.2794,
+      "step": 5210
+    },
+    {
+      "epoch": 0.41688,
+      "grad_norm": 1.2635133266448975,
+      "learning_rate": 6.2979416496791545e-06,
+      "loss": 0.2584,
+      "step": 5211
+    },
+    {
+      "epoch": 0.41696,
+      "grad_norm": 1.8867225646972656,
+      "learning_rate": 6.2967270782598735e-06,
+      "loss": 0.4578,
+      "step": 5212
+    },
+    {
+      "epoch": 0.41704,
+      "grad_norm": 1.5454792976379395,
+      "learning_rate": 6.295512424801009e-06,
+      "loss": 0.3402,
+      "step": 5213
+    },
+    {
+      "epoch": 0.41712,
+      "grad_norm": 1.5802242755889893,
+      "learning_rate": 6.2942976893794085e-06,
+      "loss": 0.2926,
+      "step": 5214
+    },
+    {
+      "epoch": 0.4172,
+      "grad_norm": 1.3554202318191528,
+      "learning_rate": 6.293082872071923e-06,
+      "loss": 0.3325,
+      "step": 5215
+    },
+    {
+      "epoch": 0.41728,
+      "grad_norm": 1.8882520198822021,
+      "learning_rate": 6.291867972955411e-06,
+      "loss": 0.463,
+      "step": 5216
+    },
+    {
+      "epoch": 0.41736,
+      "grad_norm": 1.6655519008636475,
+      "learning_rate": 6.290652992106735e-06,
+      "loss": 0.3663,
+      "step": 5217
+    },
+    {
+      "epoch": 0.41744,
+      "grad_norm": 1.787269115447998,
+      "learning_rate": 6.289437929602763e-06,
+      "loss": 0.358,
+      "step": 5218
+    },
+    {
+      "epoch": 0.41752,
+      "grad_norm": 1.5246529579162598,
+      "learning_rate": 6.288222785520368e-06,
+      "loss": 0.3243,
+      "step": 5219
+    },
+    {
+      "epoch": 0.4176,
+      "grad_norm": 1.75812828540802,
+      "learning_rate": 6.287007559936426e-06,
+      "loss": 0.3917,
+      "step": 5220
+    },
+    {
+      "epoch": 0.41768,
+      "grad_norm": 1.2533934116363525,
+      "learning_rate": 6.285792252927826e-06,
+      "loss": 0.3793,
+      "step": 5221
+    },
+    {
+      "epoch": 0.41776,
+      "grad_norm": 1.364107608795166,
+      "learning_rate": 6.284576864571449e-06,
+      "loss": 0.3192,
+      "step": 5222
+    },
+    {
+      "epoch": 0.41784,
+      "grad_norm": 1.3123117685317993,
+      "learning_rate": 6.283361394944193e-06,
+      "loss": 0.3421,
+      "step": 5223
+    },
+    {
+      "epoch": 0.41792,
+      "grad_norm": 1.1852989196777344,
+      "learning_rate": 6.282145844122956e-06,
+      "loss": 0.2568,
+      "step": 5224
+    },
+    {
+      "epoch": 0.418,
+      "grad_norm": 1.9244329929351807,
+      "learning_rate": 6.2809302121846415e-06,
+      "loss": 0.5014,
+      "step": 5225
+    },
+    {
+      "epoch": 0.41808,
+      "grad_norm": 1.4049652814865112,
+      "learning_rate": 6.279714499206157e-06,
+      "loss": 0.3108,
+      "step": 5226
+    },
+    {
+      "epoch": 0.41816,
+      "grad_norm": 1.9336098432540894,
+      "learning_rate": 6.27849870526442e-06,
+      "loss": 0.3474,
+      "step": 5227
+    },
+    {
+      "epoch": 0.41824,
+      "grad_norm": 1.9458836317062378,
+      "learning_rate": 6.277282830436346e-06,
+      "loss": 0.3939,
+      "step": 5228
+    },
+    {
+      "epoch": 0.41832,
+      "grad_norm": 2.153730869293213,
+      "learning_rate": 6.276066874798862e-06,
+      "loss": 0.4109,
+      "step": 5229
+    },
+    {
+      "epoch": 0.4184,
+      "grad_norm": 1.5371249914169312,
+      "learning_rate": 6.274850838428896e-06,
+      "loss": 0.3472,
+      "step": 5230
+    },
+    {
+      "epoch": 0.41848,
+      "grad_norm": 1.760744571685791,
+      "learning_rate": 6.273634721403385e-06,
+      "loss": 0.3729,
+      "step": 5231
+    },
+    {
+      "epoch": 0.41856,
+      "grad_norm": 1.7278648614883423,
+      "learning_rate": 6.272418523799266e-06,
+      "loss": 0.3745,
+      "step": 5232
+    },
+    {
+      "epoch": 0.41864,
+      "grad_norm": 1.7429898977279663,
+      "learning_rate": 6.271202245693484e-06,
+      "loss": 0.3522,
+      "step": 5233
+    },
+    {
+      "epoch": 0.41872,
+      "grad_norm": 1.5041425228118896,
+      "learning_rate": 6.269985887162988e-06,
+      "loss": 0.3114,
+      "step": 5234
+    },
+    {
+      "epoch": 0.4188,
+      "grad_norm": 1.4007189273834229,
+      "learning_rate": 6.268769448284736e-06,
+      "loss": 0.2983,
+      "step": 5235
+    },
+    {
+      "epoch": 0.41888,
+      "grad_norm": 1.4600435495376587,
+      "learning_rate": 6.267552929135688e-06,
+      "loss": 0.3146,
+      "step": 5236
+    },
+    {
+      "epoch": 0.41896,
+      "grad_norm": 1.7335883378982544,
+      "learning_rate": 6.266336329792804e-06,
+      "loss": 0.5255,
+      "step": 5237
+    },
+    {
+      "epoch": 0.41904,
+      "grad_norm": 1.3453007936477661,
+      "learning_rate": 6.265119650333059e-06,
+      "loss": 0.3394,
+      "step": 5238
+    },
+    {
+      "epoch": 0.41912,
+      "grad_norm": 1.1924803256988525,
+      "learning_rate": 6.263902890833427e-06,
+      "loss": 0.2937,
+      "step": 5239
+    },
+    {
+      "epoch": 0.4192,
+      "grad_norm": 1.4555060863494873,
+      "learning_rate": 6.2626860513708875e-06,
+      "loss": 0.2826,
+      "step": 5240
+    },
+    {
+      "epoch": 0.41928,
+      "grad_norm": 1.6543127298355103,
+      "learning_rate": 6.261469132022426e-06,
+      "loss": 0.3925,
+      "step": 5241
+    },
+    {
+      "epoch": 0.41936,
+      "grad_norm": 1.2936607599258423,
+      "learning_rate": 6.260252132865035e-06,
+      "loss": 0.2698,
+      "step": 5242
+    },
+    {
+      "epoch": 0.41944,
+      "grad_norm": 1.5900158882141113,
+      "learning_rate": 6.259035053975708e-06,
+      "loss": 0.35,
+      "step": 5243
+    },
+    {
+      "epoch": 0.41952,
+      "grad_norm": 1.5317350625991821,
+      "learning_rate": 6.257817895431446e-06,
+      "loss": 0.3286,
+      "step": 5244
+    },
+    {
+      "epoch": 0.4196,
+      "grad_norm": 1.4321229457855225,
+      "learning_rate": 6.256600657309254e-06,
+      "loss": 0.2899,
+      "step": 5245
+    },
+    {
+      "epoch": 0.41968,
+      "grad_norm": 1.6110750436782837,
+      "learning_rate": 6.255383339686143e-06,
+      "loss": 0.3403,
+      "step": 5246
+    },
+    {
+      "epoch": 0.41976,
+      "grad_norm": 1.6724129915237427,
+      "learning_rate": 6.254165942639128e-06,
+      "loss": 0.3894,
+      "step": 5247
+    },
+    {
+      "epoch": 0.41984,
+      "grad_norm": 1.6472164392471313,
+      "learning_rate": 6.252948466245232e-06,
+      "loss": 0.4126,
+      "step": 5248
+    },
+    {
+      "epoch": 0.41992,
+      "grad_norm": 1.4509929418563843,
+      "learning_rate": 6.251730910581478e-06,
+      "loss": 0.3197,
+      "step": 5249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9968012571334839,
+      "learning_rate": 6.250513275724896e-06,
+      "loss": 0.3536,
+      "step": 5250
+    },
+    {
+      "epoch": 0.42008,
+      "grad_norm": 1.2318665981292725,
+      "learning_rate": 6.249295561752525e-06,
+      "loss": 0.322,
+      "step": 5251
+    },
+    {
+      "epoch": 0.42016,
+      "grad_norm": 1.3581112623214722,
+      "learning_rate": 6.248077768741404e-06,
+      "loss": 0.2863,
+      "step": 5252
+    },
+    {
+      "epoch": 0.42024,
+      "grad_norm": 1.4493217468261719,
+      "learning_rate": 6.246859896768579e-06,
+      "loss": 0.2835,
+      "step": 5253
+    },
+    {
+      "epoch": 0.42032,
+      "grad_norm": 1.253295660018921,
+      "learning_rate": 6.245641945911099e-06,
+      "loss": 0.2689,
+      "step": 5254
+    },
+    {
+      "epoch": 0.4204,
+      "grad_norm": 1.5289180278778076,
+      "learning_rate": 6.244423916246023e-06,
+      "loss": 0.3176,
+      "step": 5255
+    },
+    {
+      "epoch": 0.42048,
+      "grad_norm": 1.4782484769821167,
+      "learning_rate": 6.243205807850408e-06,
+      "loss": 0.3583,
+      "step": 5256
+    },
+    {
+      "epoch": 0.42056,
+      "grad_norm": 1.5398263931274414,
+      "learning_rate": 6.241987620801322e-06,
+      "loss": 0.3898,
+      "step": 5257
+    },
+    {
+      "epoch": 0.42064,
+      "grad_norm": 1.4873570203781128,
+      "learning_rate": 6.240769355175834e-06,
+      "loss": 0.3556,
+      "step": 5258
+    },
+    {
+      "epoch": 0.42072,
+      "grad_norm": 1.5727860927581787,
+      "learning_rate": 6.239551011051021e-06,
+      "loss": 0.2808,
+      "step": 5259
+    },
+    {
+      "epoch": 0.4208,
+      "grad_norm": 1.371330738067627,
+      "learning_rate": 6.2383325885039635e-06,
+      "loss": 0.2745,
+      "step": 5260
+    },
+    {
+      "epoch": 0.42088,
+      "grad_norm": 1.5367326736450195,
+      "learning_rate": 6.237114087611747e-06,
+      "loss": 0.4118,
+      "step": 5261
+    },
+    {
+      "epoch": 0.42096,
+      "grad_norm": 1.6745377779006958,
+      "learning_rate": 6.23589550845146e-06,
+      "loss": 0.3761,
+      "step": 5262
+    },
+    {
+      "epoch": 0.42104,
+      "grad_norm": 1.4240877628326416,
+      "learning_rate": 6.234676851100201e-06,
+      "loss": 0.3344,
+      "step": 5263
+    },
+    {
+      "epoch": 0.42112,
+      "grad_norm": 1.6593492031097412,
+      "learning_rate": 6.233458115635067e-06,
+      "loss": 0.4413,
+      "step": 5264
+    },
+    {
+      "epoch": 0.4212,
+      "grad_norm": 1.6215040683746338,
+      "learning_rate": 6.232239302133167e-06,
+      "loss": 0.4039,
+      "step": 5265
+    },
+    {
+      "epoch": 0.42128,
+      "grad_norm": 1.453888177871704,
+      "learning_rate": 6.23102041067161e-06,
+      "loss": 0.253,
+      "step": 5266
+    },
+    {
+      "epoch": 0.42136,
+      "grad_norm": 1.473840355873108,
+      "learning_rate": 6.22980144132751e-06,
+      "loss": 0.2827,
+      "step": 5267
+    },
+    {
+      "epoch": 0.42144,
+      "grad_norm": 2.1219794750213623,
+      "learning_rate": 6.2285823941779864e-06,
+      "loss": 0.374,
+      "step": 5268
+    },
+    {
+      "epoch": 0.42152,
+      "grad_norm": 1.2442418336868286,
+      "learning_rate": 6.227363269300166e-06,
+      "loss": 0.2711,
+      "step": 5269
+    },
+    {
+      "epoch": 0.4216,
+      "grad_norm": 1.3598564863204956,
+      "learning_rate": 6.226144066771179e-06,
+      "loss": 0.2622,
+      "step": 5270
+    },
+    {
+      "epoch": 0.42168,
+      "grad_norm": 2.1796743869781494,
+      "learning_rate": 6.224924786668161e-06,
+      "loss": 0.4318,
+      "step": 5271
+    },
+    {
+      "epoch": 0.42176,
+      "grad_norm": 1.6060525178909302,
+      "learning_rate": 6.2237054290682475e-06,
+      "loss": 0.2812,
+      "step": 5272
+    },
+    {
+      "epoch": 0.42184,
+      "grad_norm": 1.5547593832015991,
+      "learning_rate": 6.2224859940485874e-06,
+      "loss": 0.3616,
+      "step": 5273
+    },
+    {
+      "epoch": 0.42192,
+      "grad_norm": 1.3605103492736816,
+      "learning_rate": 6.221266481686328e-06,
+      "loss": 0.2399,
+      "step": 5274
+    },
+    {
+      "epoch": 0.422,
+      "grad_norm": 1.3548734188079834,
+      "learning_rate": 6.220046892058626e-06,
+      "loss": 0.2625,
+      "step": 5275
+    },
+    {
+      "epoch": 0.42208,
+      "grad_norm": 1.6121068000793457,
+      "learning_rate": 6.218827225242638e-06,
+      "loss": 0.3836,
+      "step": 5276
+    },
+    {
+      "epoch": 0.42216,
+      "grad_norm": 1.4999924898147583,
+      "learning_rate": 6.217607481315531e-06,
+      "loss": 0.3994,
+      "step": 5277
+    },
+    {
+      "epoch": 0.42224,
+      "grad_norm": 1.484944462776184,
+      "learning_rate": 6.216387660354472e-06,
+      "loss": 0.3187,
+      "step": 5278
+    },
+    {
+      "epoch": 0.42232,
+      "grad_norm": 1.4199094772338867,
+      "learning_rate": 6.215167762436637e-06,
+      "loss": 0.3337,
+      "step": 5279
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 1.7457900047302246,
+      "learning_rate": 6.213947787639203e-06,
+      "loss": 0.3496,
+      "step": 5280
+    },
+    {
+      "epoch": 0.42248,
+      "grad_norm": 1.5127125978469849,
+      "learning_rate": 6.212727736039354e-06,
+      "loss": 0.3331,
+      "step": 5281
+    },
+    {
+      "epoch": 0.42256,
+      "grad_norm": 1.618870735168457,
+      "learning_rate": 6.211507607714277e-06,
+      "loss": 0.2917,
+      "step": 5282
+    },
+    {
+      "epoch": 0.42264,
+      "grad_norm": 1.5503782033920288,
+      "learning_rate": 6.210287402741171e-06,
+      "loss": 0.3699,
+      "step": 5283
+    },
+    {
+      "epoch": 0.42272,
+      "grad_norm": 1.6964318752288818,
+      "learning_rate": 6.209067121197228e-06,
+      "loss": 0.3575,
+      "step": 5284
+    },
+    {
+      "epoch": 0.4228,
+      "grad_norm": 1.616135835647583,
+      "learning_rate": 6.207846763159655e-06,
+      "loss": 0.3734,
+      "step": 5285
+    },
+    {
+      "epoch": 0.42288,
+      "grad_norm": 1.5970280170440674,
+      "learning_rate": 6.206626328705659e-06,
+      "loss": 0.3787,
+      "step": 5286
+    },
+    {
+      "epoch": 0.42296,
+      "grad_norm": 1.7362499237060547,
+      "learning_rate": 6.205405817912452e-06,
+      "loss": 0.3442,
+      "step": 5287
+    },
+    {
+      "epoch": 0.42304,
+      "grad_norm": 1.6331170797348022,
+      "learning_rate": 6.204185230857252e-06,
+      "loss": 0.3461,
+      "step": 5288
+    },
+    {
+      "epoch": 0.42312,
+      "grad_norm": 2.0485804080963135,
+      "learning_rate": 6.202964567617283e-06,
+      "loss": 0.3591,
+      "step": 5289
+    },
+    {
+      "epoch": 0.4232,
+      "grad_norm": 1.9766350984573364,
+      "learning_rate": 6.20174382826977e-06,
+      "loss": 0.4748,
+      "step": 5290
+    },
+    {
+      "epoch": 0.42328,
+      "grad_norm": 1.2820550203323364,
+      "learning_rate": 6.200523012891945e-06,
+      "loss": 0.2704,
+      "step": 5291
+    },
+    {
+      "epoch": 0.42336,
+      "grad_norm": 1.3562285900115967,
+      "learning_rate": 6.199302121561048e-06,
+      "loss": 0.329,
+      "step": 5292
+    },
+    {
+      "epoch": 0.42344,
+      "grad_norm": 1.444642186164856,
+      "learning_rate": 6.198081154354317e-06,
+      "loss": 0.4026,
+      "step": 5293
+    },
+    {
+      "epoch": 0.42352,
+      "grad_norm": 1.574292778968811,
+      "learning_rate": 6.196860111349001e-06,
+      "loss": 0.299,
+      "step": 5294
+    },
+    {
+      "epoch": 0.4236,
+      "grad_norm": 1.5647331476211548,
+      "learning_rate": 6.19563899262235e-06,
+      "loss": 0.2892,
+      "step": 5295
+    },
+    {
+      "epoch": 0.42368,
+      "grad_norm": 1.810318112373352,
+      "learning_rate": 6.194417798251622e-06,
+      "loss": 0.3837,
+      "step": 5296
+    },
+    {
+      "epoch": 0.42376,
+      "grad_norm": 1.7629601955413818,
+      "learning_rate": 6.193196528314073e-06,
+      "loss": 0.3893,
+      "step": 5297
+    },
+    {
+      "epoch": 0.42384,
+      "grad_norm": 1.5911134481430054,
+      "learning_rate": 6.191975182886976e-06,
+      "loss": 0.3948,
+      "step": 5298
+    },
+    {
+      "epoch": 0.42392,
+      "grad_norm": 1.661783218383789,
+      "learning_rate": 6.1907537620475955e-06,
+      "loss": 0.3691,
+      "step": 5299
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 1.1115741729736328,
+      "learning_rate": 6.189532265873209e-06,
+      "loss": 0.2339,
+      "step": 5300
+    },
+    {
+      "epoch": 0.42408,
+      "grad_norm": 1.4616998434066772,
+      "learning_rate": 6.188310694441097e-06,
+      "loss": 0.2663,
+      "step": 5301
+    },
+    {
+      "epoch": 0.42416,
+      "grad_norm": 1.6125479936599731,
+      "learning_rate": 6.187089047828542e-06,
+      "loss": 0.3089,
+      "step": 5302
+    },
+    {
+      "epoch": 0.42424,
+      "grad_norm": 1.8839313983917236,
+      "learning_rate": 6.1858673261128364e-06,
+      "loss": 0.4514,
+      "step": 5303
+    },
+    {
+      "epoch": 0.42432,
+      "grad_norm": 1.655595302581787,
+      "learning_rate": 6.184645529371272e-06,
+      "loss": 0.3448,
+      "step": 5304
+    },
+    {
+      "epoch": 0.4244,
+      "grad_norm": 1.311193585395813,
+      "learning_rate": 6.183423657681149e-06,
+      "loss": 0.2564,
+      "step": 5305
+    },
+    {
+      "epoch": 0.42448,
+      "grad_norm": 2.025151491165161,
+      "learning_rate": 6.182201711119771e-06,
+      "loss": 0.5696,
+      "step": 5306
+    },
+    {
+      "epoch": 0.42456,
+      "grad_norm": 1.321874737739563,
+      "learning_rate": 6.180979689764447e-06,
+      "loss": 0.3092,
+      "step": 5307
+    },
+    {
+      "epoch": 0.42464,
+      "grad_norm": 2.123602867126465,
+      "learning_rate": 6.179757593692488e-06,
+      "loss": 0.4035,
+      "step": 5308
+    },
+    {
+      "epoch": 0.42472,
+      "grad_norm": 1.5167906284332275,
+      "learning_rate": 6.178535422981216e-06,
+      "loss": 0.3193,
+      "step": 5309
+    },
+    {
+      "epoch": 0.4248,
+      "grad_norm": 1.7157925367355347,
+      "learning_rate": 6.17731317770795e-06,
+      "loss": 0.3725,
+      "step": 5310
+    },
+    {
+      "epoch": 0.42488,
+      "grad_norm": 1.4427510499954224,
+      "learning_rate": 6.176090857950018e-06,
+      "loss": 0.2961,
+      "step": 5311
+    },
+    {
+      "epoch": 0.42496,
+      "grad_norm": 1.3995834589004517,
+      "learning_rate": 6.174868463784752e-06,
+      "loss": 0.3033,
+      "step": 5312
+    },
+    {
+      "epoch": 0.42504,
+      "grad_norm": 1.511171817779541,
+      "learning_rate": 6.173645995289491e-06,
+      "loss": 0.3218,
+      "step": 5313
+    },
+    {
+      "epoch": 0.42512,
+      "grad_norm": 1.7538864612579346,
+      "learning_rate": 6.172423452541574e-06,
+      "loss": 0.4702,
+      "step": 5314
+    },
+    {
+      "epoch": 0.4252,
+      "grad_norm": 1.4485112428665161,
+      "learning_rate": 6.1712008356183485e-06,
+      "loss": 0.2855,
+      "step": 5315
+    },
+    {
+      "epoch": 0.42528,
+      "grad_norm": 1.475102186203003,
+      "learning_rate": 6.169978144597164e-06,
+      "loss": 0.3669,
+      "step": 5316
+    },
+    {
+      "epoch": 0.42536,
+      "grad_norm": 1.569522500038147,
+      "learning_rate": 6.168755379555378e-06,
+      "loss": 0.3393,
+      "step": 5317
+    },
+    {
+      "epoch": 0.42544,
+      "grad_norm": 1.7826415300369263,
+      "learning_rate": 6.167532540570351e-06,
+      "loss": 0.3331,
+      "step": 5318
+    },
+    {
+      "epoch": 0.42552,
+      "grad_norm": 1.3612468242645264,
+      "learning_rate": 6.166309627719444e-06,
+      "loss": 0.3009,
+      "step": 5319
+    },
+    {
+      "epoch": 0.4256,
+      "grad_norm": 1.4943021535873413,
+      "learning_rate": 6.165086641080032e-06,
+      "loss": 0.3387,
+      "step": 5320
+    },
+    {
+      "epoch": 0.42568,
+      "grad_norm": 1.629370927810669,
+      "learning_rate": 6.163863580729484e-06,
+      "loss": 0.3885,
+      "step": 5321
+    },
+    {
+      "epoch": 0.42576,
+      "grad_norm": 2.004427433013916,
+      "learning_rate": 6.162640446745184e-06,
+      "loss": 0.4166,
+      "step": 5322
+    },
+    {
+      "epoch": 0.42584,
+      "grad_norm": 1.5527446269989014,
+      "learning_rate": 6.161417239204512e-06,
+      "loss": 0.3396,
+      "step": 5323
+    },
+    {
+      "epoch": 0.42592,
+      "grad_norm": 1.4790587425231934,
+      "learning_rate": 6.160193958184858e-06,
+      "loss": 0.3218,
+      "step": 5324
+    },
+    {
+      "epoch": 0.426,
+      "grad_norm": 1.5748834609985352,
+      "learning_rate": 6.158970603763615e-06,
+      "loss": 0.3968,
+      "step": 5325
+    },
+    {
+      "epoch": 0.42608,
+      "grad_norm": 1.7657493352890015,
+      "learning_rate": 6.157747176018177e-06,
+      "loss": 0.3378,
+      "step": 5326
+    },
+    {
+      "epoch": 0.42616,
+      "grad_norm": 2.0443902015686035,
+      "learning_rate": 6.15652367502595e-06,
+      "loss": 0.4723,
+      "step": 5327
+    },
+    {
+      "epoch": 0.42624,
+      "grad_norm": 1.6636515855789185,
+      "learning_rate": 6.155300100864341e-06,
+      "loss": 0.3116,
+      "step": 5328
+    },
+    {
+      "epoch": 0.42632,
+      "grad_norm": 1.5545495748519897,
+      "learning_rate": 6.154076453610759e-06,
+      "loss": 0.3614,
+      "step": 5329
+    },
+    {
+      "epoch": 0.4264,
+      "grad_norm": 1.6116411685943604,
+      "learning_rate": 6.152852733342623e-06,
+      "loss": 0.3714,
+      "step": 5330
+    },
+    {
+      "epoch": 0.42648,
+      "grad_norm": 1.3033878803253174,
+      "learning_rate": 6.151628940137351e-06,
+      "loss": 0.3168,
+      "step": 5331
+    },
+    {
+      "epoch": 0.42656,
+      "grad_norm": 1.3937110900878906,
+      "learning_rate": 6.150405074072369e-06,
+      "loss": 0.274,
+      "step": 5332
+    },
+    {
+      "epoch": 0.42664,
+      "grad_norm": 1.9616461992263794,
+      "learning_rate": 6.1491811352251085e-06,
+      "loss": 0.4114,
+      "step": 5333
+    },
+    {
+      "epoch": 0.42672,
+      "grad_norm": 1.5841161012649536,
+      "learning_rate": 6.1479571236730005e-06,
+      "loss": 0.3684,
+      "step": 5334
+    },
+    {
+      "epoch": 0.4268,
+      "grad_norm": 1.403443455696106,
+      "learning_rate": 6.146733039493487e-06,
+      "loss": 0.2633,
+      "step": 5335
+    },
+    {
+      "epoch": 0.42688,
+      "grad_norm": 1.5101227760314941,
+      "learning_rate": 6.145508882764013e-06,
+      "loss": 0.2751,
+      "step": 5336
+    },
+    {
+      "epoch": 0.42696,
+      "grad_norm": 1.6489311456680298,
+      "learning_rate": 6.144284653562024e-06,
+      "loss": 0.3991,
+      "step": 5337
+    },
+    {
+      "epoch": 0.42704,
+      "grad_norm": 1.270580530166626,
+      "learning_rate": 6.143060351964973e-06,
+      "loss": 0.2707,
+      "step": 5338
+    },
+    {
+      "epoch": 0.42712,
+      "grad_norm": 1.7063907384872437,
+      "learning_rate": 6.141835978050318e-06,
+      "loss": 0.5227,
+      "step": 5339
+    },
+    {
+      "epoch": 0.4272,
+      "grad_norm": 1.2827900648117065,
+      "learning_rate": 6.140611531895522e-06,
+      "loss": 0.3149,
+      "step": 5340
+    },
+    {
+      "epoch": 0.42728,
+      "grad_norm": 1.45085608959198,
+      "learning_rate": 6.139387013578051e-06,
+      "loss": 0.3797,
+      "step": 5341
+    },
+    {
+      "epoch": 0.42736,
+      "grad_norm": 1.335211992263794,
+      "learning_rate": 6.138162423175375e-06,
+      "loss": 0.3144,
+      "step": 5342
+    },
+    {
+      "epoch": 0.42744,
+      "grad_norm": 1.3709321022033691,
+      "learning_rate": 6.136937760764972e-06,
+      "loss": 0.2786,
+      "step": 5343
+    },
+    {
+      "epoch": 0.42752,
+      "grad_norm": 1.9491933584213257,
+      "learning_rate": 6.13571302642432e-06,
+      "loss": 0.3513,
+      "step": 5344
+    },
+    {
+      "epoch": 0.4276,
+      "grad_norm": 1.6201810836791992,
+      "learning_rate": 6.1344882202309075e-06,
+      "loss": 0.3616,
+      "step": 5345
+    },
+    {
+      "epoch": 0.42768,
+      "grad_norm": 1.6798819303512573,
+      "learning_rate": 6.133263342262219e-06,
+      "loss": 0.3304,
+      "step": 5346
+    },
+    {
+      "epoch": 0.42776,
+      "grad_norm": 1.485134243965149,
+      "learning_rate": 6.132038392595751e-06,
+      "loss": 0.2942,
+      "step": 5347
+    },
+    {
+      "epoch": 0.42784,
+      "grad_norm": 1.5289043188095093,
+      "learning_rate": 6.130813371309002e-06,
+      "loss": 0.311,
+      "step": 5348
+    },
+    {
+      "epoch": 0.42792,
+      "grad_norm": 1.708020567893982,
+      "learning_rate": 6.129588278479475e-06,
+      "loss": 0.3565,
+      "step": 5349
+    },
+    {
+      "epoch": 0.428,
+      "grad_norm": 1.249941349029541,
+      "learning_rate": 6.1283631141846755e-06,
+      "loss": 0.239,
+      "step": 5350
+    },
+    {
+      "epoch": 0.42808,
+      "grad_norm": 1.3397772312164307,
+      "learning_rate": 6.127137878502118e-06,
+      "loss": 0.3618,
+      "step": 5351
+    },
+    {
+      "epoch": 0.42816,
+      "grad_norm": 1.3979480266571045,
+      "learning_rate": 6.125912571509319e-06,
+      "loss": 0.2997,
+      "step": 5352
+    },
+    {
+      "epoch": 0.42824,
+      "grad_norm": 1.456049919128418,
+      "learning_rate": 6.124687193283799e-06,
+      "loss": 0.315,
+      "step": 5353
+    },
+    {
+      "epoch": 0.42832,
+      "grad_norm": 1.965668797492981,
+      "learning_rate": 6.123461743903084e-06,
+      "loss": 0.4174,
+      "step": 5354
+    },
+    {
+      "epoch": 0.4284,
+      "grad_norm": 1.4012924432754517,
+      "learning_rate": 6.122236223444703e-06,
+      "loss": 0.325,
+      "step": 5355
+    },
+    {
+      "epoch": 0.42848,
+      "grad_norm": 1.5004931688308716,
+      "learning_rate": 6.121010631986192e-06,
+      "loss": 0.3237,
+      "step": 5356
+    },
+    {
+      "epoch": 0.42856,
+      "grad_norm": 1.6268855333328247,
+      "learning_rate": 6.119784969605088e-06,
+      "loss": 0.4094,
+      "step": 5357
+    },
+    {
+      "epoch": 0.42864,
+      "grad_norm": 1.6038322448730469,
+      "learning_rate": 6.1185592363789355e-06,
+      "loss": 0.2773,
+      "step": 5358
+    },
+    {
+      "epoch": 0.42872,
+      "grad_norm": 1.811691403388977,
+      "learning_rate": 6.117333432385283e-06,
+      "loss": 0.3442,
+      "step": 5359
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 1.5288585424423218,
+      "learning_rate": 6.116107557701685e-06,
+      "loss": 0.3593,
+      "step": 5360
+    },
+    {
+      "epoch": 0.42888,
+      "grad_norm": 1.7061458826065063,
+      "learning_rate": 6.114881612405694e-06,
+      "loss": 0.4021,
+      "step": 5361
+    },
+    {
+      "epoch": 0.42896,
+      "grad_norm": 1.6404433250427246,
+      "learning_rate": 6.1136555965748735e-06,
+      "loss": 0.3261,
+      "step": 5362
+    },
+    {
+      "epoch": 0.42904,
+      "grad_norm": 1.6574904918670654,
+      "learning_rate": 6.11242951028679e-06,
+      "loss": 0.4901,
+      "step": 5363
+    },
+    {
+      "epoch": 0.42912,
+      "grad_norm": 1.2963271141052246,
+      "learning_rate": 6.111203353619014e-06,
+      "loss": 0.2512,
+      "step": 5364
+    },
+    {
+      "epoch": 0.4292,
+      "grad_norm": 1.7801556587219238,
+      "learning_rate": 6.109977126649121e-06,
+      "loss": 0.3984,
+      "step": 5365
+    },
+    {
+      "epoch": 0.42928,
+      "grad_norm": 1.7426446676254272,
+      "learning_rate": 6.108750829454688e-06,
+      "loss": 0.3063,
+      "step": 5366
+    },
+    {
+      "epoch": 0.42936,
+      "grad_norm": 1.6115596294403076,
+      "learning_rate": 6.1075244621133e-06,
+      "loss": 0.3511,
+      "step": 5367
+    },
+    {
+      "epoch": 0.42944,
+      "grad_norm": 1.2986208200454712,
+      "learning_rate": 6.106298024702546e-06,
+      "loss": 0.3206,
+      "step": 5368
+    },
+    {
+      "epoch": 0.42952,
+      "grad_norm": 1.9597961902618408,
+      "learning_rate": 6.105071517300017e-06,
+      "loss": 0.3735,
+      "step": 5369
+    },
+    {
+      "epoch": 0.4296,
+      "grad_norm": 1.4968289136886597,
+      "learning_rate": 6.10384493998331e-06,
+      "loss": 0.2758,
+      "step": 5370
+    },
+    {
+      "epoch": 0.42968,
+      "grad_norm": 1.4992220401763916,
+      "learning_rate": 6.102618292830029e-06,
+      "loss": 0.3457,
+      "step": 5371
+    },
+    {
+      "epoch": 0.42976,
+      "grad_norm": 1.5327579975128174,
+      "learning_rate": 6.1013915759177765e-06,
+      "loss": 0.3693,
+      "step": 5372
+    },
+    {
+      "epoch": 0.42984,
+      "grad_norm": 1.6249293088912964,
+      "learning_rate": 6.1001647893241634e-06,
+      "loss": 0.3094,
+      "step": 5373
+    },
+    {
+      "epoch": 0.42992,
+      "grad_norm": 1.7367304563522339,
+      "learning_rate": 6.098937933126806e-06,
+      "loss": 0.341,
+      "step": 5374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3340009450912476,
+      "learning_rate": 6.097711007403323e-06,
+      "loss": 0.279,
+      "step": 5375
+    },
+    {
+      "epoch": 0.43008,
+      "grad_norm": 1.6486116647720337,
+      "learning_rate": 6.096484012231337e-06,
+      "loss": 0.3304,
+      "step": 5376
+    },
+    {
+      "epoch": 0.43016,
+      "grad_norm": 1.1886416673660278,
+      "learning_rate": 6.095256947688478e-06,
+      "loss": 0.2639,
+      "step": 5377
+    },
+    {
+      "epoch": 0.43024,
+      "grad_norm": 1.6678515672683716,
+      "learning_rate": 6.094029813852376e-06,
+      "loss": 0.3143,
+      "step": 5378
+    },
+    {
+      "epoch": 0.43032,
+      "grad_norm": 1.7223871946334839,
+      "learning_rate": 6.0928026108006675e-06,
+      "loss": 0.4146,
+      "step": 5379
+    },
+    {
+      "epoch": 0.4304,
+      "grad_norm": 1.9803104400634766,
+      "learning_rate": 6.091575338610994e-06,
+      "loss": 0.4194,
+      "step": 5380
+    },
+    {
+      "epoch": 0.43048,
+      "grad_norm": 1.469686508178711,
+      "learning_rate": 6.090347997361002e-06,
+      "loss": 0.3625,
+      "step": 5381
+    },
+    {
+      "epoch": 0.43056,
+      "grad_norm": 1.2683942317962646,
+      "learning_rate": 6.089120587128341e-06,
+      "loss": 0.258,
+      "step": 5382
+    },
+    {
+      "epoch": 0.43064,
+      "grad_norm": 1.4345426559448242,
+      "learning_rate": 6.087893107990665e-06,
+      "loss": 0.3373,
+      "step": 5383
+    },
+    {
+      "epoch": 0.43072,
+      "grad_norm": 1.7910370826721191,
+      "learning_rate": 6.0866655600256305e-06,
+      "loss": 0.4067,
+      "step": 5384
+    },
+    {
+      "epoch": 0.4308,
+      "grad_norm": 1.5200092792510986,
+      "learning_rate": 6.085437943310902e-06,
+      "loss": 0.2964,
+      "step": 5385
+    },
+    {
+      "epoch": 0.43088,
+      "grad_norm": 1.9939285516738892,
+      "learning_rate": 6.084210257924148e-06,
+      "loss": 0.3602,
+      "step": 5386
+    },
+    {
+      "epoch": 0.43096,
+      "grad_norm": 1.6643867492675781,
+      "learning_rate": 6.082982503943038e-06,
+      "loss": 0.3337,
+      "step": 5387
+    },
+    {
+      "epoch": 0.43104,
+      "grad_norm": 1.9779936075210571,
+      "learning_rate": 6.081754681445249e-06,
+      "loss": 0.3836,
+      "step": 5388
+    },
+    {
+      "epoch": 0.43112,
+      "grad_norm": 1.9725874662399292,
+      "learning_rate": 6.080526790508461e-06,
+      "loss": 0.3404,
+      "step": 5389
+    },
+    {
+      "epoch": 0.4312,
+      "grad_norm": 1.6242984533309937,
+      "learning_rate": 6.079298831210357e-06,
+      "loss": 0.3975,
+      "step": 5390
+    },
+    {
+      "epoch": 0.43128,
+      "grad_norm": 1.8685696125030518,
+      "learning_rate": 6.078070803628629e-06,
+      "loss": 0.3465,
+      "step": 5391
+    },
+    {
+      "epoch": 0.43136,
+      "grad_norm": 1.4240777492523193,
+      "learning_rate": 6.076842707840969e-06,
+      "loss": 0.2717,
+      "step": 5392
+    },
+    {
+      "epoch": 0.43144,
+      "grad_norm": 1.6460999250411987,
+      "learning_rate": 6.0756145439250725e-06,
+      "loss": 0.4144,
+      "step": 5393
+    },
+    {
+      "epoch": 0.43152,
+      "grad_norm": 1.7656724452972412,
+      "learning_rate": 6.074386311958643e-06,
+      "loss": 0.383,
+      "step": 5394
+    },
+    {
+      "epoch": 0.4316,
+      "grad_norm": 1.7070986032485962,
+      "learning_rate": 6.073158012019388e-06,
+      "loss": 0.3203,
+      "step": 5395
+    },
+    {
+      "epoch": 0.43168,
+      "grad_norm": 2.166088581085205,
+      "learning_rate": 6.071929644185014e-06,
+      "loss": 0.454,
+      "step": 5396
+    },
+    {
+      "epoch": 0.43176,
+      "grad_norm": 1.7778912782669067,
+      "learning_rate": 6.07070120853324e-06,
+      "loss": 0.3748,
+      "step": 5397
+    },
+    {
+      "epoch": 0.43184,
+      "grad_norm": 1.4228088855743408,
+      "learning_rate": 6.069472705141781e-06,
+      "loss": 0.3252,
+      "step": 5398
+    },
+    {
+      "epoch": 0.43192,
+      "grad_norm": 1.510135293006897,
+      "learning_rate": 6.068244134088363e-06,
+      "loss": 0.4119,
+      "step": 5399
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 1.453439712524414,
+      "learning_rate": 6.067015495450715e-06,
+      "loss": 0.2788,
+      "step": 5400
+    },
+    {
+      "epoch": 0.43208,
+      "grad_norm": 1.4725075960159302,
+      "learning_rate": 6.065786789306566e-06,
+      "loss": 0.3727,
+      "step": 5401
+    },
+    {
+      "epoch": 0.43216,
+      "grad_norm": 1.9456934928894043,
+      "learning_rate": 6.064558015733653e-06,
+      "loss": 0.3712,
+      "step": 5402
+    },
+    {
+      "epoch": 0.43224,
+      "grad_norm": 1.5379313230514526,
+      "learning_rate": 6.063329174809715e-06,
+      "loss": 0.3203,
+      "step": 5403
+    },
+    {
+      "epoch": 0.43232,
+      "grad_norm": 1.5046453475952148,
+      "learning_rate": 6.0621002666124995e-06,
+      "loss": 0.3101,
+      "step": 5404
+    },
+    {
+      "epoch": 0.4324,
+      "grad_norm": 1.9868191480636597,
+      "learning_rate": 6.060871291219753e-06,
+      "loss": 0.3739,
+      "step": 5405
+    },
+    {
+      "epoch": 0.43248,
+      "grad_norm": 1.5251084566116333,
+      "learning_rate": 6.0596422487092295e-06,
+      "loss": 0.322,
+      "step": 5406
+    },
+    {
+      "epoch": 0.43256,
+      "grad_norm": 1.3554657697677612,
+      "learning_rate": 6.058413139158687e-06,
+      "loss": 0.3143,
+      "step": 5407
+    },
+    {
+      "epoch": 0.43264,
+      "grad_norm": 1.592686414718628,
+      "learning_rate": 6.0571839626458875e-06,
+      "loss": 0.3187,
+      "step": 5408
+    },
+    {
+      "epoch": 0.43272,
+      "grad_norm": 1.3215630054473877,
+      "learning_rate": 6.055954719248595e-06,
+      "loss": 0.308,
+      "step": 5409
+    },
+    {
+      "epoch": 0.4328,
+      "grad_norm": 1.2852091789245605,
+      "learning_rate": 6.054725409044579e-06,
+      "loss": 0.2711,
+      "step": 5410
+    },
+    {
+      "epoch": 0.43288,
+      "grad_norm": 1.8097467422485352,
+      "learning_rate": 6.0534960321116175e-06,
+      "loss": 0.3427,
+      "step": 5411
+    },
+    {
+      "epoch": 0.43296,
+      "grad_norm": 1.3292269706726074,
+      "learning_rate": 6.052266588527488e-06,
+      "loss": 0.3537,
+      "step": 5412
+    },
+    {
+      "epoch": 0.43304,
+      "grad_norm": 1.6236307621002197,
+      "learning_rate": 6.051037078369972e-06,
+      "loss": 0.4301,
+      "step": 5413
+    },
+    {
+      "epoch": 0.43312,
+      "grad_norm": 1.4823275804519653,
+      "learning_rate": 6.049807501716856e-06,
+      "loss": 0.3199,
+      "step": 5414
+    },
+    {
+      "epoch": 0.4332,
+      "grad_norm": 1.5742340087890625,
+      "learning_rate": 6.048577858645932e-06,
+      "loss": 0.3312,
+      "step": 5415
+    },
+    {
+      "epoch": 0.43328,
+      "grad_norm": 1.4714792966842651,
+      "learning_rate": 6.047348149234995e-06,
+      "loss": 0.3228,
+      "step": 5416
+    },
+    {
+      "epoch": 0.43336,
+      "grad_norm": 1.3467466831207275,
+      "learning_rate": 6.046118373561845e-06,
+      "loss": 0.3045,
+      "step": 5417
+    },
+    {
+      "epoch": 0.43344,
+      "grad_norm": 1.4545137882232666,
+      "learning_rate": 6.044888531704287e-06,
+      "loss": 0.3308,
+      "step": 5418
+    },
+    {
+      "epoch": 0.43352,
+      "grad_norm": 1.3145484924316406,
+      "learning_rate": 6.043658623740127e-06,
+      "loss": 0.2933,
+      "step": 5419
+    },
+    {
+      "epoch": 0.4336,
+      "grad_norm": 1.6407479047775269,
+      "learning_rate": 6.042428649747177e-06,
+      "loss": 0.3418,
+      "step": 5420
+    },
+    {
+      "epoch": 0.43368,
+      "grad_norm": 1.6189351081848145,
+      "learning_rate": 6.041198609803256e-06,
+      "loss": 0.3938,
+      "step": 5421
+    },
+    {
+      "epoch": 0.43376,
+      "grad_norm": 1.1933389902114868,
+      "learning_rate": 6.039968503986182e-06,
+      "loss": 0.2499,
+      "step": 5422
+    },
+    {
+      "epoch": 0.43384,
+      "grad_norm": 1.6682744026184082,
+      "learning_rate": 6.038738332373781e-06,
+      "loss": 0.3783,
+      "step": 5423
+    },
+    {
+      "epoch": 0.43392,
+      "grad_norm": 1.450366497039795,
+      "learning_rate": 6.037508095043881e-06,
+      "loss": 0.3095,
+      "step": 5424
+    },
+    {
+      "epoch": 0.434,
+      "grad_norm": 1.6315852403640747,
+      "learning_rate": 6.036277792074316e-06,
+      "loss": 0.3228,
+      "step": 5425
+    },
+    {
+      "epoch": 0.43408,
+      "grad_norm": 1.3595662117004395,
+      "learning_rate": 6.035047423542922e-06,
+      "loss": 0.2532,
+      "step": 5426
+    },
+    {
+      "epoch": 0.43416,
+      "grad_norm": 1.3640127182006836,
+      "learning_rate": 6.033816989527541e-06,
+      "loss": 0.3054,
+      "step": 5427
+    },
+    {
+      "epoch": 0.43424,
+      "grad_norm": 1.3128430843353271,
+      "learning_rate": 6.032586490106018e-06,
+      "loss": 0.2737,
+      "step": 5428
+    },
+    {
+      "epoch": 0.43432,
+      "grad_norm": 1.4516959190368652,
+      "learning_rate": 6.0313559253562016e-06,
+      "loss": 0.3047,
+      "step": 5429
+    },
+    {
+      "epoch": 0.4344,
+      "grad_norm": 1.5541858673095703,
+      "learning_rate": 6.030125295355949e-06,
+      "loss": 0.2977,
+      "step": 5430
+    },
+    {
+      "epoch": 0.43448,
+      "grad_norm": 1.3899632692337036,
+      "learning_rate": 6.028894600183114e-06,
+      "loss": 0.2981,
+      "step": 5431
+    },
+    {
+      "epoch": 0.43456,
+      "grad_norm": 1.6935724020004272,
+      "learning_rate": 6.027663839915561e-06,
+      "loss": 0.4806,
+      "step": 5432
+    },
+    {
+      "epoch": 0.43464,
+      "grad_norm": 1.4013090133666992,
+      "learning_rate": 6.026433014631155e-06,
+      "loss": 0.3428,
+      "step": 5433
+    },
+    {
+      "epoch": 0.43472,
+      "grad_norm": 1.6702803373336792,
+      "learning_rate": 6.025202124407766e-06,
+      "loss": 0.3447,
+      "step": 5434
+    },
+    {
+      "epoch": 0.4348,
+      "grad_norm": 1.4864176511764526,
+      "learning_rate": 6.023971169323272e-06,
+      "loss": 0.3252,
+      "step": 5435
+    },
+    {
+      "epoch": 0.43488,
+      "grad_norm": 1.4419050216674805,
+      "learning_rate": 6.022740149455547e-06,
+      "loss": 0.314,
+      "step": 5436
+    },
+    {
+      "epoch": 0.43496,
+      "grad_norm": 1.489224910736084,
+      "learning_rate": 6.021509064882473e-06,
+      "loss": 0.353,
+      "step": 5437
+    },
+    {
+      "epoch": 0.43504,
+      "grad_norm": 1.4233548641204834,
+      "learning_rate": 6.0202779156819405e-06,
+      "loss": 0.3279,
+      "step": 5438
+    },
+    {
+      "epoch": 0.43512,
+      "grad_norm": 1.631506323814392,
+      "learning_rate": 6.019046701931836e-06,
+      "loss": 0.4391,
+      "step": 5439
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 1.7712832689285278,
+      "learning_rate": 6.0178154237100575e-06,
+      "loss": 0.3618,
+      "step": 5440
+    },
+    {
+      "epoch": 0.43528,
+      "grad_norm": 1.745080828666687,
+      "learning_rate": 6.016584081094503e-06,
+      "loss": 0.5306,
+      "step": 5441
+    },
+    {
+      "epoch": 0.43536,
+      "grad_norm": 1.2281311750411987,
+      "learning_rate": 6.015352674163075e-06,
+      "loss": 0.2491,
+      "step": 5442
+    },
+    {
+      "epoch": 0.43544,
+      "grad_norm": 1.725820779800415,
+      "learning_rate": 6.014121202993682e-06,
+      "loss": 0.3775,
+      "step": 5443
+    },
+    {
+      "epoch": 0.43552,
+      "grad_norm": 1.3258579969406128,
+      "learning_rate": 6.012889667664231e-06,
+      "loss": 0.2745,
+      "step": 5444
+    },
+    {
+      "epoch": 0.4356,
+      "grad_norm": 1.2049994468688965,
+      "learning_rate": 6.0116580682526415e-06,
+      "loss": 0.2574,
+      "step": 5445
+    },
+    {
+      "epoch": 0.43568,
+      "grad_norm": 1.1984772682189941,
+      "learning_rate": 6.010426404836831e-06,
+      "loss": 0.283,
+      "step": 5446
+    },
+    {
+      "epoch": 0.43576,
+      "grad_norm": 1.4736144542694092,
+      "learning_rate": 6.009194677494723e-06,
+      "loss": 0.3,
+      "step": 5447
+    },
+    {
+      "epoch": 0.43584,
+      "grad_norm": 1.5378073453903198,
+      "learning_rate": 6.007962886304245e-06,
+      "loss": 0.3541,
+      "step": 5448
+    },
+    {
+      "epoch": 0.43592,
+      "grad_norm": 1.6598979234695435,
+      "learning_rate": 6.006731031343327e-06,
+      "loss": 0.3801,
+      "step": 5449
+    },
+    {
+      "epoch": 0.436,
+      "grad_norm": 1.6594270467758179,
+      "learning_rate": 6.0054991126899055e-06,
+      "loss": 0.4225,
+      "step": 5450
+    },
+    {
+      "epoch": 0.43608,
+      "grad_norm": 1.6767255067825317,
+      "learning_rate": 6.004267130421918e-06,
+      "loss": 0.4542,
+      "step": 5451
+    },
+    {
+      "epoch": 0.43616,
+      "grad_norm": 1.594008207321167,
+      "learning_rate": 6.003035084617311e-06,
+      "loss": 0.2978,
+      "step": 5452
+    },
+    {
+      "epoch": 0.43624,
+      "grad_norm": 1.3313404321670532,
+      "learning_rate": 6.0018029753540295e-06,
+      "loss": 0.2959,
+      "step": 5453
+    },
+    {
+      "epoch": 0.43632,
+      "grad_norm": 1.4482040405273438,
+      "learning_rate": 6.0005708027100274e-06,
+      "loss": 0.4336,
+      "step": 5454
+    },
+    {
+      "epoch": 0.4364,
+      "grad_norm": 1.7941350936889648,
+      "learning_rate": 5.999338566763258e-06,
+      "loss": 0.3423,
+      "step": 5455
+    },
+    {
+      "epoch": 0.43648,
+      "grad_norm": 1.9843556880950928,
+      "learning_rate": 5.998106267591679e-06,
+      "loss": 0.4683,
+      "step": 5456
+    },
+    {
+      "epoch": 0.43656,
+      "grad_norm": 1.8331342935562134,
+      "learning_rate": 5.996873905273259e-06,
+      "loss": 0.4164,
+      "step": 5457
+    },
+    {
+      "epoch": 0.43664,
+      "grad_norm": 1.5630244016647339,
+      "learning_rate": 5.995641479885962e-06,
+      "loss": 0.3497,
+      "step": 5458
+    },
+    {
+      "epoch": 0.43672,
+      "grad_norm": 1.7044670581817627,
+      "learning_rate": 5.99440899150776e-06,
+      "loss": 0.4218,
+      "step": 5459
+    },
+    {
+      "epoch": 0.4368,
+      "grad_norm": 2.0335018634796143,
+      "learning_rate": 5.993176440216627e-06,
+      "loss": 0.3786,
+      "step": 5460
+    },
+    {
+      "epoch": 0.43688,
+      "grad_norm": 1.4986952543258667,
+      "learning_rate": 5.991943826090545e-06,
+      "loss": 0.3274,
+      "step": 5461
+    },
+    {
+      "epoch": 0.43696,
+      "grad_norm": 1.511153221130371,
+      "learning_rate": 5.990711149207496e-06,
+      "loss": 0.3809,
+      "step": 5462
+    },
+    {
+      "epoch": 0.43704,
+      "grad_norm": 1.593832015991211,
+      "learning_rate": 5.989478409645466e-06,
+      "loss": 0.346,
+      "step": 5463
+    },
+    {
+      "epoch": 0.43712,
+      "grad_norm": 1.7036159038543701,
+      "learning_rate": 5.988245607482449e-06,
+      "loss": 0.4385,
+      "step": 5464
+    },
+    {
+      "epoch": 0.4372,
+      "grad_norm": 1.4725526571273804,
+      "learning_rate": 5.987012742796441e-06,
+      "loss": 0.3309,
+      "step": 5465
+    },
+    {
+      "epoch": 0.43728,
+      "grad_norm": 1.5867393016815186,
+      "learning_rate": 5.985779815665436e-06,
+      "loss": 0.3505,
+      "step": 5466
+    },
+    {
+      "epoch": 0.43736,
+      "grad_norm": 2.0825116634368896,
+      "learning_rate": 5.9845468261674435e-06,
+      "loss": 0.4112,
+      "step": 5467
+    },
+    {
+      "epoch": 0.43744,
+      "grad_norm": 1.8679453134536743,
+      "learning_rate": 5.9833137743804645e-06,
+      "loss": 0.4366,
+      "step": 5468
+    },
+    {
+      "epoch": 0.43752,
+      "grad_norm": 1.5901966094970703,
+      "learning_rate": 5.982080660382516e-06,
+      "loss": 0.3359,
+      "step": 5469
+    },
+    {
+      "epoch": 0.4376,
+      "grad_norm": 1.1409281492233276,
+      "learning_rate": 5.98084748425161e-06,
+      "loss": 0.2324,
+      "step": 5470
+    },
+    {
+      "epoch": 0.43768,
+      "grad_norm": 1.6136082410812378,
+      "learning_rate": 5.979614246065765e-06,
+      "loss": 0.3601,
+      "step": 5471
+    },
+    {
+      "epoch": 0.43776,
+      "grad_norm": 1.8404332399368286,
+      "learning_rate": 5.978380945903004e-06,
+      "loss": 0.4764,
+      "step": 5472
+    },
+    {
+      "epoch": 0.43784,
+      "grad_norm": 1.9501445293426514,
+      "learning_rate": 5.977147583841354e-06,
+      "loss": 0.3829,
+      "step": 5473
+    },
+    {
+      "epoch": 0.43792,
+      "grad_norm": 1.4439702033996582,
+      "learning_rate": 5.975914159958846e-06,
+      "loss": 0.317,
+      "step": 5474
+    },
+    {
+      "epoch": 0.438,
+      "grad_norm": 1.3018183708190918,
+      "learning_rate": 5.974680674333514e-06,
+      "loss": 0.2696,
+      "step": 5475
+    },
+    {
+      "epoch": 0.43808,
+      "grad_norm": 1.4886512756347656,
+      "learning_rate": 5.973447127043398e-06,
+      "loss": 0.3454,
+      "step": 5476
+    },
+    {
+      "epoch": 0.43816,
+      "grad_norm": 1.5890697240829468,
+      "learning_rate": 5.97221351816654e-06,
+      "loss": 0.3576,
+      "step": 5477
+    },
+    {
+      "epoch": 0.43824,
+      "grad_norm": 1.8298115730285645,
+      "learning_rate": 5.970979847780984e-06,
+      "loss": 0.3515,
+      "step": 5478
+    },
+    {
+      "epoch": 0.43832,
+      "grad_norm": 1.5328552722930908,
+      "learning_rate": 5.969746115964783e-06,
+      "loss": 0.3744,
+      "step": 5479
+    },
+    {
+      "epoch": 0.4384,
+      "grad_norm": 1.7203729152679443,
+      "learning_rate": 5.968512322795991e-06,
+      "loss": 0.3269,
+      "step": 5480
+    },
+    {
+      "epoch": 0.43848,
+      "grad_norm": 1.672613501548767,
+      "learning_rate": 5.967278468352663e-06,
+      "loss": 0.3997,
+      "step": 5481
+    },
+    {
+      "epoch": 0.43856,
+      "grad_norm": 1.2868820428848267,
+      "learning_rate": 5.966044552712864e-06,
+      "loss": 0.3223,
+      "step": 5482
+    },
+    {
+      "epoch": 0.43864,
+      "grad_norm": 1.2581580877304077,
+      "learning_rate": 5.9648105759546595e-06,
+      "loss": 0.2698,
+      "step": 5483
+    },
+    {
+      "epoch": 0.43872,
+      "grad_norm": 1.465779185295105,
+      "learning_rate": 5.963576538156116e-06,
+      "loss": 0.3777,
+      "step": 5484
+    },
+    {
+      "epoch": 0.4388,
+      "grad_norm": 1.3545647859573364,
+      "learning_rate": 5.96234243939531e-06,
+      "loss": 0.3381,
+      "step": 5485
+    },
+    {
+      "epoch": 0.43888,
+      "grad_norm": 1.3060506582260132,
+      "learning_rate": 5.9611082797503175e-06,
+      "loss": 0.3209,
+      "step": 5486
+    },
+    {
+      "epoch": 0.43896,
+      "grad_norm": 1.4939861297607422,
+      "learning_rate": 5.95987405929922e-06,
+      "loss": 0.3916,
+      "step": 5487
+    },
+    {
+      "epoch": 0.43904,
+      "grad_norm": 1.9790840148925781,
+      "learning_rate": 5.9586397781201034e-06,
+      "loss": 0.516,
+      "step": 5488
+    },
+    {
+      "epoch": 0.43912,
+      "grad_norm": 1.3512998819351196,
+      "learning_rate": 5.957405436291055e-06,
+      "loss": 0.2739,
+      "step": 5489
+    },
+    {
+      "epoch": 0.4392,
+      "grad_norm": 1.5846513509750366,
+      "learning_rate": 5.956171033890168e-06,
+      "loss": 0.4435,
+      "step": 5490
+    },
+    {
+      "epoch": 0.43928,
+      "grad_norm": 1.4433051347732544,
+      "learning_rate": 5.95493657099554e-06,
+      "loss": 0.3871,
+      "step": 5491
+    },
+    {
+      "epoch": 0.43936,
+      "grad_norm": 1.4104619026184082,
+      "learning_rate": 5.953702047685271e-06,
+      "loss": 0.2823,
+      "step": 5492
+    },
+    {
+      "epoch": 0.43944,
+      "grad_norm": 1.4222887754440308,
+      "learning_rate": 5.952467464037462e-06,
+      "loss": 0.3227,
+      "step": 5493
+    },
+    {
+      "epoch": 0.43952,
+      "grad_norm": 1.3719890117645264,
+      "learning_rate": 5.951232820130224e-06,
+      "loss": 0.295,
+      "step": 5494
+    },
+    {
+      "epoch": 0.4396,
+      "grad_norm": 1.7587943077087402,
+      "learning_rate": 5.949998116041671e-06,
+      "loss": 0.4271,
+      "step": 5495
+    },
+    {
+      "epoch": 0.43968,
+      "grad_norm": 1.418924331665039,
+      "learning_rate": 5.948763351849913e-06,
+      "loss": 0.2773,
+      "step": 5496
+    },
+    {
+      "epoch": 0.43976,
+      "grad_norm": 1.8153401613235474,
+      "learning_rate": 5.947528527633073e-06,
+      "loss": 0.3063,
+      "step": 5497
+    },
+    {
+      "epoch": 0.43984,
+      "grad_norm": 1.4398388862609863,
+      "learning_rate": 5.946293643469274e-06,
+      "loss": 0.3319,
+      "step": 5498
+    },
+    {
+      "epoch": 0.43992,
+      "grad_norm": 1.4004582166671753,
+      "learning_rate": 5.945058699436641e-06,
+      "loss": 0.327,
+      "step": 5499
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4153037071228027,
+      "learning_rate": 5.943823695613308e-06,
+      "loss": 0.2834,
+      "step": 5500
+    },
+    {
+      "epoch": 0.44008,
+      "grad_norm": 1.4712837934494019,
+      "learning_rate": 5.9425886320774086e-06,
+      "loss": 0.2809,
+      "step": 5501
+    },
+    {
+      "epoch": 0.44016,
+      "grad_norm": 1.3245508670806885,
+      "learning_rate": 5.941353508907078e-06,
+      "loss": 0.224,
+      "step": 5502
+    },
+    {
+      "epoch": 0.44024,
+      "grad_norm": 1.4912713766098022,
+      "learning_rate": 5.940118326180463e-06,
+      "loss": 0.4387,
+      "step": 5503
+    },
+    {
+      "epoch": 0.44032,
+      "grad_norm": 1.4831585884094238,
+      "learning_rate": 5.938883083975706e-06,
+      "loss": 0.303,
+      "step": 5504
+    },
+    {
+      "epoch": 0.4404,
+      "grad_norm": 0.9887499809265137,
+      "learning_rate": 5.937647782370957e-06,
+      "loss": 0.2525,
+      "step": 5505
+    },
+    {
+      "epoch": 0.44048,
+      "grad_norm": 1.6899783611297607,
+      "learning_rate": 5.936412421444372e-06,
+      "loss": 0.3949,
+      "step": 5506
+    },
+    {
+      "epoch": 0.44056,
+      "grad_norm": 1.812455415725708,
+      "learning_rate": 5.935177001274105e-06,
+      "loss": 0.4223,
+      "step": 5507
+    },
+    {
+      "epoch": 0.44064,
+      "grad_norm": 1.5728967189788818,
+      "learning_rate": 5.933941521938318e-06,
+      "loss": 0.3647,
+      "step": 5508
+    },
+    {
+      "epoch": 0.44072,
+      "grad_norm": 1.8846977949142456,
+      "learning_rate": 5.932705983515176e-06,
+      "loss": 0.4741,
+      "step": 5509
+    },
+    {
+      "epoch": 0.4408,
+      "grad_norm": 1.6265419721603394,
+      "learning_rate": 5.931470386082847e-06,
+      "loss": 0.3013,
+      "step": 5510
+    },
+    {
+      "epoch": 0.44088,
+      "grad_norm": 1.5003933906555176,
+      "learning_rate": 5.930234729719504e-06,
+      "loss": 0.2989,
+      "step": 5511
+    },
+    {
+      "epoch": 0.44096,
+      "grad_norm": 1.764898657798767,
+      "learning_rate": 5.9289990145033226e-06,
+      "loss": 0.3484,
+      "step": 5512
+    },
+    {
+      "epoch": 0.44104,
+      "grad_norm": 1.7658780813217163,
+      "learning_rate": 5.927763240512482e-06,
+      "loss": 0.4016,
+      "step": 5513
+    },
+    {
+      "epoch": 0.44112,
+      "grad_norm": 1.515289545059204,
+      "learning_rate": 5.926527407825164e-06,
+      "loss": 0.3122,
+      "step": 5514
+    },
+    {
+      "epoch": 0.4412,
+      "grad_norm": 1.3999195098876953,
+      "learning_rate": 5.92529151651956e-06,
+      "loss": 0.4573,
+      "step": 5515
+    },
+    {
+      "epoch": 0.44128,
+      "grad_norm": 2.083272695541382,
+      "learning_rate": 5.924055566673855e-06,
+      "loss": 0.3821,
+      "step": 5516
+    },
+    {
+      "epoch": 0.44136,
+      "grad_norm": 1.596103310585022,
+      "learning_rate": 5.922819558366247e-06,
+      "loss": 0.4768,
+      "step": 5517
+    },
+    {
+      "epoch": 0.44144,
+      "grad_norm": 1.2892342805862427,
+      "learning_rate": 5.921583491674935e-06,
+      "loss": 0.2956,
+      "step": 5518
+    },
+    {
+      "epoch": 0.44152,
+      "grad_norm": 1.8127224445343018,
+      "learning_rate": 5.920347366678117e-06,
+      "loss": 0.3626,
+      "step": 5519
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 1.7415847778320312,
+      "learning_rate": 5.9191111834540006e-06,
+      "loss": 0.3858,
+      "step": 5520
+    },
+    {
+      "epoch": 0.44168,
+      "grad_norm": 1.5482152700424194,
+      "learning_rate": 5.917874942080796e-06,
+      "loss": 0.3114,
+      "step": 5521
+    },
+    {
+      "epoch": 0.44176,
+      "grad_norm": 1.4196959733963013,
+      "learning_rate": 5.916638642636714e-06,
+      "loss": 0.368,
+      "step": 5522
+    },
+    {
+      "epoch": 0.44184,
+      "grad_norm": 1.5246591567993164,
+      "learning_rate": 5.9154022851999725e-06,
+      "loss": 0.3554,
+      "step": 5523
+    },
+    {
+      "epoch": 0.44192,
+      "grad_norm": 1.4558786153793335,
+      "learning_rate": 5.914165869848793e-06,
+      "loss": 0.3025,
+      "step": 5524
+    },
+    {
+      "epoch": 0.442,
+      "grad_norm": 1.1743532419204712,
+      "learning_rate": 5.912929396661396e-06,
+      "loss": 0.2267,
+      "step": 5525
+    },
+    {
+      "epoch": 0.44208,
+      "grad_norm": 1.816625952720642,
+      "learning_rate": 5.911692865716011e-06,
+      "loss": 0.4033,
+      "step": 5526
+    },
+    {
+      "epoch": 0.44216,
+      "grad_norm": 1.3450242280960083,
+      "learning_rate": 5.910456277090869e-06,
+      "loss": 0.3432,
+      "step": 5527
+    },
+    {
+      "epoch": 0.44224,
+      "grad_norm": 1.6333425045013428,
+      "learning_rate": 5.909219630864204e-06,
+      "loss": 0.3124,
+      "step": 5528
+    },
+    {
+      "epoch": 0.44232,
+      "grad_norm": 1.9529705047607422,
+      "learning_rate": 5.907982927114257e-06,
+      "loss": 0.4574,
+      "step": 5529
+    },
+    {
+      "epoch": 0.4424,
+      "grad_norm": 1.9013431072235107,
+      "learning_rate": 5.906746165919267e-06,
+      "loss": 0.4198,
+      "step": 5530
+    },
+    {
+      "epoch": 0.44248,
+      "grad_norm": 1.8581068515777588,
+      "learning_rate": 5.905509347357481e-06,
+      "loss": 0.4414,
+      "step": 5531
+    },
+    {
+      "epoch": 0.44256,
+      "grad_norm": 1.7523456811904907,
+      "learning_rate": 5.904272471507148e-06,
+      "loss": 0.3963,
+      "step": 5532
+    },
+    {
+      "epoch": 0.44264,
+      "grad_norm": 1.8803638219833374,
+      "learning_rate": 5.903035538446524e-06,
+      "loss": 0.3608,
+      "step": 5533
+    },
+    {
+      "epoch": 0.44272,
+      "grad_norm": 1.630553126335144,
+      "learning_rate": 5.901798548253859e-06,
+      "loss": 0.3344,
+      "step": 5534
+    },
+    {
+      "epoch": 0.4428,
+      "grad_norm": 1.7994118928909302,
+      "learning_rate": 5.90056150100742e-06,
+      "loss": 0.4549,
+      "step": 5535
+    },
+    {
+      "epoch": 0.44288,
+      "grad_norm": 1.4933981895446777,
+      "learning_rate": 5.8993243967854685e-06,
+      "loss": 0.3576,
+      "step": 5536
+    },
+    {
+      "epoch": 0.44296,
+      "grad_norm": 1.6949869394302368,
+      "learning_rate": 5.898087235666271e-06,
+      "loss": 0.3546,
+      "step": 5537
+    },
+    {
+      "epoch": 0.44304,
+      "grad_norm": 1.6014376878738403,
+      "learning_rate": 5.8968500177281e-06,
+      "loss": 0.3171,
+      "step": 5538
+    },
+    {
+      "epoch": 0.44312,
+      "grad_norm": 1.835667371749878,
+      "learning_rate": 5.895612743049227e-06,
+      "loss": 0.509,
+      "step": 5539
+    },
+    {
+      "epoch": 0.4432,
+      "grad_norm": 1.977283000946045,
+      "learning_rate": 5.894375411707933e-06,
+      "loss": 0.4484,
+      "step": 5540
+    },
+    {
+      "epoch": 0.44328,
+      "grad_norm": 1.4197109937667847,
+      "learning_rate": 5.8931380237825e-06,
+      "loss": 0.2965,
+      "step": 5541
+    },
+    {
+      "epoch": 0.44336,
+      "grad_norm": 1.751121997833252,
+      "learning_rate": 5.891900579351213e-06,
+      "loss": 0.367,
+      "step": 5542
+    },
+    {
+      "epoch": 0.44344,
+      "grad_norm": 1.6477973461151123,
+      "learning_rate": 5.89066307849236e-06,
+      "loss": 0.3525,
+      "step": 5543
+    },
+    {
+      "epoch": 0.44352,
+      "grad_norm": 1.5611004829406738,
+      "learning_rate": 5.889425521284234e-06,
+      "loss": 0.3347,
+      "step": 5544
+    },
+    {
+      "epoch": 0.4436,
+      "grad_norm": 1.138242244720459,
+      "learning_rate": 5.888187907805132e-06,
+      "loss": 0.252,
+      "step": 5545
+    },
+    {
+      "epoch": 0.44368,
+      "grad_norm": 1.4641177654266357,
+      "learning_rate": 5.8869502381333525e-06,
+      "loss": 0.3063,
+      "step": 5546
+    },
+    {
+      "epoch": 0.44376,
+      "grad_norm": 1.6272233724594116,
+      "learning_rate": 5.8857125123472e-06,
+      "loss": 0.2902,
+      "step": 5547
+    },
+    {
+      "epoch": 0.44384,
+      "grad_norm": 1.4607881307601929,
+      "learning_rate": 5.88447473052498e-06,
+      "loss": 0.2488,
+      "step": 5548
+    },
+    {
+      "epoch": 0.44392,
+      "grad_norm": 1.5880279541015625,
+      "learning_rate": 5.883236892745003e-06,
+      "loss": 0.426,
+      "step": 5549
+    },
+    {
+      "epoch": 0.444,
+      "grad_norm": 1.4784557819366455,
+      "learning_rate": 5.881998999085583e-06,
+      "loss": 0.2709,
+      "step": 5550
+    },
+    {
+      "epoch": 0.44408,
+      "grad_norm": 1.4526431560516357,
+      "learning_rate": 5.880761049625038e-06,
+      "loss": 0.3239,
+      "step": 5551
+    },
+    {
+      "epoch": 0.44416,
+      "grad_norm": 1.838352918624878,
+      "learning_rate": 5.879523044441687e-06,
+      "loss": 0.3932,
+      "step": 5552
+    },
+    {
+      "epoch": 0.44424,
+      "grad_norm": 1.5688480138778687,
+      "learning_rate": 5.878284983613858e-06,
+      "loss": 0.2671,
+      "step": 5553
+    },
+    {
+      "epoch": 0.44432,
+      "grad_norm": 1.8419432640075684,
+      "learning_rate": 5.877046867219876e-06,
+      "loss": 0.3733,
+      "step": 5554
+    },
+    {
+      "epoch": 0.4444,
+      "grad_norm": 2.1512904167175293,
+      "learning_rate": 5.8758086953380725e-06,
+      "loss": 0.419,
+      "step": 5555
+    },
+    {
+      "epoch": 0.44448,
+      "grad_norm": 1.3534338474273682,
+      "learning_rate": 5.874570468046784e-06,
+      "loss": 0.3388,
+      "step": 5556
+    },
+    {
+      "epoch": 0.44456,
+      "grad_norm": 2.1905033588409424,
+      "learning_rate": 5.873332185424348e-06,
+      "loss": 0.4197,
+      "step": 5557
+    },
+    {
+      "epoch": 0.44464,
+      "grad_norm": 1.410273790359497,
+      "learning_rate": 5.872093847549106e-06,
+      "loss": 0.3875,
+      "step": 5558
+    },
+    {
+      "epoch": 0.44472,
+      "grad_norm": 1.7642848491668701,
+      "learning_rate": 5.870855454499407e-06,
+      "loss": 0.4297,
+      "step": 5559
+    },
+    {
+      "epoch": 0.4448,
+      "grad_norm": 1.5598673820495605,
+      "learning_rate": 5.869617006353596e-06,
+      "loss": 0.3774,
+      "step": 5560
+    },
+    {
+      "epoch": 0.44488,
+      "grad_norm": 1.347758412361145,
+      "learning_rate": 5.868378503190027e-06,
+      "loss": 0.2839,
+      "step": 5561
+    },
+    {
+      "epoch": 0.44496,
+      "grad_norm": 1.6148842573165894,
+      "learning_rate": 5.8671399450870535e-06,
+      "loss": 0.3108,
+      "step": 5562
+    },
+    {
+      "epoch": 0.44504,
+      "grad_norm": 1.48881196975708,
+      "learning_rate": 5.8659013321230385e-06,
+      "loss": 0.3588,
+      "step": 5563
+    },
+    {
+      "epoch": 0.44512,
+      "grad_norm": 1.3120616674423218,
+      "learning_rate": 5.8646626643763435e-06,
+      "loss": 0.2724,
+      "step": 5564
+    },
+    {
+      "epoch": 0.4452,
+      "grad_norm": 1.1921863555908203,
+      "learning_rate": 5.863423941925337e-06,
+      "loss": 0.2622,
+      "step": 5565
+    },
+    {
+      "epoch": 0.44528,
+      "grad_norm": 1.8236253261566162,
+      "learning_rate": 5.862185164848384e-06,
+      "loss": 0.3556,
+      "step": 5566
+    },
+    {
+      "epoch": 0.44536,
+      "grad_norm": 1.6157597303390503,
+      "learning_rate": 5.860946333223862e-06,
+      "loss": 0.369,
+      "step": 5567
+    },
+    {
+      "epoch": 0.44544,
+      "grad_norm": 1.5578926801681519,
+      "learning_rate": 5.859707447130144e-06,
+      "loss": 0.3256,
+      "step": 5568
+    },
+    {
+      "epoch": 0.44552,
+      "grad_norm": 1.4830211400985718,
+      "learning_rate": 5.858468506645613e-06,
+      "loss": 0.3329,
+      "step": 5569
+    },
+    {
+      "epoch": 0.4456,
+      "grad_norm": 1.290466547012329,
+      "learning_rate": 5.857229511848655e-06,
+      "loss": 0.2373,
+      "step": 5570
+    },
+    {
+      "epoch": 0.44568,
+      "grad_norm": 1.6431125402450562,
+      "learning_rate": 5.855990462817651e-06,
+      "loss": 0.3732,
+      "step": 5571
+    },
+    {
+      "epoch": 0.44576,
+      "grad_norm": 1.8566932678222656,
+      "learning_rate": 5.854751359630997e-06,
+      "loss": 0.4304,
+      "step": 5572
+    },
+    {
+      "epoch": 0.44584,
+      "grad_norm": 1.466186285018921,
+      "learning_rate": 5.853512202367083e-06,
+      "loss": 0.328,
+      "step": 5573
+    },
+    {
+      "epoch": 0.44592,
+      "grad_norm": 1.5933393239974976,
+      "learning_rate": 5.852272991104308e-06,
+      "loss": 0.3883,
+      "step": 5574
+    },
+    {
+      "epoch": 0.446,
+      "grad_norm": 1.850595474243164,
+      "learning_rate": 5.851033725921073e-06,
+      "loss": 0.486,
+      "step": 5575
+    },
+    {
+      "epoch": 0.44608,
+      "grad_norm": 1.4042195081710815,
+      "learning_rate": 5.84979440689578e-06,
+      "loss": 0.3212,
+      "step": 5576
+    },
+    {
+      "epoch": 0.44616,
+      "grad_norm": 1.733971118927002,
+      "learning_rate": 5.848555034106841e-06,
+      "loss": 0.3999,
+      "step": 5577
+    },
+    {
+      "epoch": 0.44624,
+      "grad_norm": 1.5726109743118286,
+      "learning_rate": 5.847315607632662e-06,
+      "loss": 0.3346,
+      "step": 5578
+    },
+    {
+      "epoch": 0.44632,
+      "grad_norm": 1.2969645261764526,
+      "learning_rate": 5.846076127551661e-06,
+      "loss": 0.2933,
+      "step": 5579
+    },
+    {
+      "epoch": 0.4464,
+      "grad_norm": 1.2956920862197876,
+      "learning_rate": 5.8448365939422534e-06,
+      "loss": 0.275,
+      "step": 5580
+    },
+    {
+      "epoch": 0.44648,
+      "grad_norm": 1.5254887342453003,
+      "learning_rate": 5.8435970068828605e-06,
+      "loss": 0.3264,
+      "step": 5581
+    },
+    {
+      "epoch": 0.44656,
+      "grad_norm": 1.3373383283615112,
+      "learning_rate": 5.842357366451911e-06,
+      "loss": 0.2972,
+      "step": 5582
+    },
+    {
+      "epoch": 0.44664,
+      "grad_norm": 1.5913358926773071,
+      "learning_rate": 5.841117672727827e-06,
+      "loss": 0.5288,
+      "step": 5583
+    },
+    {
+      "epoch": 0.44672,
+      "grad_norm": 1.6409987211227417,
+      "learning_rate": 5.839877925789043e-06,
+      "loss": 0.3628,
+      "step": 5584
+    },
+    {
+      "epoch": 0.4468,
+      "grad_norm": 1.7899576425552368,
+      "learning_rate": 5.8386381257139925e-06,
+      "loss": 0.3251,
+      "step": 5585
+    },
+    {
+      "epoch": 0.44688,
+      "grad_norm": 1.5527466535568237,
+      "learning_rate": 5.837398272581114e-06,
+      "loss": 0.3213,
+      "step": 5586
+    },
+    {
+      "epoch": 0.44696,
+      "grad_norm": 1.662186622619629,
+      "learning_rate": 5.836158366468848e-06,
+      "loss": 0.3146,
+      "step": 5587
+    },
+    {
+      "epoch": 0.44704,
+      "grad_norm": 1.686964750289917,
+      "learning_rate": 5.8349184074556396e-06,
+      "loss": 0.3395,
+      "step": 5588
+    },
+    {
+      "epoch": 0.44712,
+      "grad_norm": 1.8355437517166138,
+      "learning_rate": 5.833678395619939e-06,
+      "loss": 0.449,
+      "step": 5589
+    },
+    {
+      "epoch": 0.4472,
+      "grad_norm": 1.7132534980773926,
+      "learning_rate": 5.832438331040196e-06,
+      "loss": 0.3185,
+      "step": 5590
+    },
+    {
+      "epoch": 0.44728,
+      "grad_norm": 1.7874171733856201,
+      "learning_rate": 5.831198213794863e-06,
+      "loss": 0.4094,
+      "step": 5591
+    },
+    {
+      "epoch": 0.44736,
+      "grad_norm": 1.4985543489456177,
+      "learning_rate": 5.829958043962402e-06,
+      "loss": 0.4014,
+      "step": 5592
+    },
+    {
+      "epoch": 0.44744,
+      "grad_norm": 1.901485562324524,
+      "learning_rate": 5.828717821621272e-06,
+      "loss": 0.3675,
+      "step": 5593
+    },
+    {
+      "epoch": 0.44752,
+      "grad_norm": 1.6370835304260254,
+      "learning_rate": 5.827477546849938e-06,
+      "loss": 0.3236,
+      "step": 5594
+    },
+    {
+      "epoch": 0.4476,
+      "grad_norm": 1.7742931842803955,
+      "learning_rate": 5.826237219726869e-06,
+      "loss": 0.4238,
+      "step": 5595
+    },
+    {
+      "epoch": 0.44768,
+      "grad_norm": 1.4299365282058716,
+      "learning_rate": 5.824996840330536e-06,
+      "loss": 0.3525,
+      "step": 5596
+    },
+    {
+      "epoch": 0.44776,
+      "grad_norm": 1.4459290504455566,
+      "learning_rate": 5.823756408739412e-06,
+      "loss": 0.2749,
+      "step": 5597
+    },
+    {
+      "epoch": 0.44784,
+      "grad_norm": 1.5706748962402344,
+      "learning_rate": 5.822515925031977e-06,
+      "loss": 0.3489,
+      "step": 5598
+    },
+    {
+      "epoch": 0.44792,
+      "grad_norm": 0.996229350566864,
+      "learning_rate": 5.821275389286711e-06,
+      "loss": 0.2767,
+      "step": 5599
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 1.8221417665481567,
+      "learning_rate": 5.820034801582101e-06,
+      "loss": 0.3905,
+      "step": 5600
+    },
+    {
+      "epoch": 0.44808,
+      "grad_norm": 1.9977222681045532,
+      "learning_rate": 5.818794161996631e-06,
+      "loss": 0.3864,
+      "step": 5601
+    },
+    {
+      "epoch": 0.44816,
+      "grad_norm": 2.152876615524292,
+      "learning_rate": 5.817553470608795e-06,
+      "loss": 0.5186,
+      "step": 5602
+    },
+    {
+      "epoch": 0.44824,
+      "grad_norm": 1.5512678623199463,
+      "learning_rate": 5.816312727497085e-06,
+      "loss": 0.2871,
+      "step": 5603
+    },
+    {
+      "epoch": 0.44832,
+      "grad_norm": 1.9361984729766846,
+      "learning_rate": 5.815071932740002e-06,
+      "loss": 0.3828,
+      "step": 5604
+    },
+    {
+      "epoch": 0.4484,
+      "grad_norm": 1.641486644744873,
+      "learning_rate": 5.813831086416044e-06,
+      "loss": 0.4174,
+      "step": 5605
+    },
+    {
+      "epoch": 0.44848,
+      "grad_norm": 1.4502854347229004,
+      "learning_rate": 5.812590188603718e-06,
+      "loss": 0.2888,
+      "step": 5606
+    },
+    {
+      "epoch": 0.44856,
+      "grad_norm": 1.4937154054641724,
+      "learning_rate": 5.811349239381528e-06,
+      "loss": 0.3603,
+      "step": 5607
+    },
+    {
+      "epoch": 0.44864,
+      "grad_norm": 1.1726678609848022,
+      "learning_rate": 5.810108238827986e-06,
+      "loss": 0.2616,
+      "step": 5608
+    },
+    {
+      "epoch": 0.44872,
+      "grad_norm": 1.500806212425232,
+      "learning_rate": 5.808867187021607e-06,
+      "loss": 0.2688,
+      "step": 5609
+    },
+    {
+      "epoch": 0.4488,
+      "grad_norm": 1.740702509880066,
+      "learning_rate": 5.8076260840409086e-06,
+      "loss": 0.4273,
+      "step": 5610
+    },
+    {
+      "epoch": 0.44888,
+      "grad_norm": 1.5013270378112793,
+      "learning_rate": 5.806384929964408e-06,
+      "loss": 0.3114,
+      "step": 5611
+    },
+    {
+      "epoch": 0.44896,
+      "grad_norm": 1.7903553247451782,
+      "learning_rate": 5.805143724870633e-06,
+      "loss": 0.3147,
+      "step": 5612
+    },
+    {
+      "epoch": 0.44904,
+      "grad_norm": 1.476843237876892,
+      "learning_rate": 5.8039024688381074e-06,
+      "loss": 0.3131,
+      "step": 5613
+    },
+    {
+      "epoch": 0.44912,
+      "grad_norm": 1.402813196182251,
+      "learning_rate": 5.802661161945363e-06,
+      "loss": 0.303,
+      "step": 5614
+    },
+    {
+      "epoch": 0.4492,
+      "grad_norm": 1.5428776741027832,
+      "learning_rate": 5.801419804270932e-06,
+      "loss": 0.3801,
+      "step": 5615
+    },
+    {
+      "epoch": 0.44928,
+      "grad_norm": 1.5787391662597656,
+      "learning_rate": 5.800178395893353e-06,
+      "loss": 0.3183,
+      "step": 5616
+    },
+    {
+      "epoch": 0.44936,
+      "grad_norm": 1.7391003370285034,
+      "learning_rate": 5.798936936891163e-06,
+      "loss": 0.3259,
+      "step": 5617
+    },
+    {
+      "epoch": 0.44944,
+      "grad_norm": 1.6685212850570679,
+      "learning_rate": 5.797695427342908e-06,
+      "loss": 0.3274,
+      "step": 5618
+    },
+    {
+      "epoch": 0.44952,
+      "grad_norm": 1.4090831279754639,
+      "learning_rate": 5.79645386732713e-06,
+      "loss": 0.3116,
+      "step": 5619
+    },
+    {
+      "epoch": 0.4496,
+      "grad_norm": 1.5553460121154785,
+      "learning_rate": 5.795212256922382e-06,
+      "loss": 0.3605,
+      "step": 5620
+    },
+    {
+      "epoch": 0.44968,
+      "grad_norm": 1.7685738801956177,
+      "learning_rate": 5.793970596207214e-06,
+      "loss": 0.3015,
+      "step": 5621
+    },
+    {
+      "epoch": 0.44976,
+      "grad_norm": 1.7890232801437378,
+      "learning_rate": 5.792728885260184e-06,
+      "loss": 0.3894,
+      "step": 5622
+    },
+    {
+      "epoch": 0.44984,
+      "grad_norm": 2.025926113128662,
+      "learning_rate": 5.79148712415985e-06,
+      "loss": 0.3836,
+      "step": 5623
+    },
+    {
+      "epoch": 0.44992,
+      "grad_norm": 1.6190468072891235,
+      "learning_rate": 5.790245312984775e-06,
+      "loss": 0.3422,
+      "step": 5624
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7557851076126099,
+      "learning_rate": 5.789003451813522e-06,
+      "loss": 0.3468,
+      "step": 5625
+    },
+    {
+      "epoch": 0.45008,
+      "grad_norm": 1.489891529083252,
+      "learning_rate": 5.78776154072466e-06,
+      "loss": 0.3636,
+      "step": 5626
+    },
+    {
+      "epoch": 0.45016,
+      "grad_norm": 1.5357917547225952,
+      "learning_rate": 5.786519579796764e-06,
+      "loss": 0.3056,
+      "step": 5627
+    },
+    {
+      "epoch": 0.45024,
+      "grad_norm": 1.792183756828308,
+      "learning_rate": 5.785277569108403e-06,
+      "loss": 0.3856,
+      "step": 5628
+    },
+    {
+      "epoch": 0.45032,
+      "grad_norm": 1.6786377429962158,
+      "learning_rate": 5.7840355087381575e-06,
+      "loss": 0.4319,
+      "step": 5629
+    },
+    {
+      "epoch": 0.4504,
+      "grad_norm": 1.8204642534255981,
+      "learning_rate": 5.7827933987646115e-06,
+      "loss": 0.3844,
+      "step": 5630
+    },
+    {
+      "epoch": 0.45048,
+      "grad_norm": 1.8311585187911987,
+      "learning_rate": 5.781551239266344e-06,
+      "loss": 0.5054,
+      "step": 5631
+    },
+    {
+      "epoch": 0.45056,
+      "grad_norm": 1.9697678089141846,
+      "learning_rate": 5.780309030321945e-06,
+      "loss": 0.359,
+      "step": 5632
+    },
+    {
+      "epoch": 0.45064,
+      "grad_norm": 1.9044475555419922,
+      "learning_rate": 5.779066772010005e-06,
+      "loss": 0.4049,
+      "step": 5633
+    },
+    {
+      "epoch": 0.45072,
+      "grad_norm": 1.8143421411514282,
+      "learning_rate": 5.777824464409117e-06,
+      "loss": 0.4335,
+      "step": 5634
+    },
+    {
+      "epoch": 0.4508,
+      "grad_norm": 1.4178447723388672,
+      "learning_rate": 5.776582107597877e-06,
+      "loss": 0.3715,
+      "step": 5635
+    },
+    {
+      "epoch": 0.45088,
+      "grad_norm": 1.598649024963379,
+      "learning_rate": 5.775339701654887e-06,
+      "loss": 0.3177,
+      "step": 5636
+    },
+    {
+      "epoch": 0.45096,
+      "grad_norm": 1.538147211074829,
+      "learning_rate": 5.7740972466587476e-06,
+      "loss": 0.3332,
+      "step": 5637
+    },
+    {
+      "epoch": 0.45104,
+      "grad_norm": 1.4716747999191284,
+      "learning_rate": 5.772854742688066e-06,
+      "loss": 0.4715,
+      "step": 5638
+    },
+    {
+      "epoch": 0.45112,
+      "grad_norm": 1.5471874475479126,
+      "learning_rate": 5.771612189821451e-06,
+      "loss": 0.3747,
+      "step": 5639
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 1.5069724321365356,
+      "learning_rate": 5.770369588137513e-06,
+      "loss": 0.2939,
+      "step": 5640
+    },
+    {
+      "epoch": 0.45128,
+      "grad_norm": 1.0864536762237549,
+      "learning_rate": 5.76912693771487e-06,
+      "loss": 0.215,
+      "step": 5641
+    },
+    {
+      "epoch": 0.45136,
+      "grad_norm": 1.2346502542495728,
+      "learning_rate": 5.76788423863214e-06,
+      "loss": 0.2422,
+      "step": 5642
+    },
+    {
+      "epoch": 0.45144,
+      "grad_norm": 1.443285346031189,
+      "learning_rate": 5.766641490967942e-06,
+      "loss": 0.26,
+      "step": 5643
+    },
+    {
+      "epoch": 0.45152,
+      "grad_norm": 1.8851925134658813,
+      "learning_rate": 5.765398694800902e-06,
+      "loss": 0.4111,
+      "step": 5644
+    },
+    {
+      "epoch": 0.4516,
+      "grad_norm": 1.4499890804290771,
+      "learning_rate": 5.764155850209649e-06,
+      "loss": 0.2916,
+      "step": 5645
+    },
+    {
+      "epoch": 0.45168,
+      "grad_norm": 1.421256422996521,
+      "learning_rate": 5.7629129572728105e-06,
+      "loss": 0.3533,
+      "step": 5646
+    },
+    {
+      "epoch": 0.45176,
+      "grad_norm": 1.4521507024765015,
+      "learning_rate": 5.761670016069025e-06,
+      "loss": 0.3407,
+      "step": 5647
+    },
+    {
+      "epoch": 0.45184,
+      "grad_norm": 1.6597949266433716,
+      "learning_rate": 5.760427026676923e-06,
+      "loss": 0.3523,
+      "step": 5648
+    },
+    {
+      "epoch": 0.45192,
+      "grad_norm": 1.5322909355163574,
+      "learning_rate": 5.759183989175148e-06,
+      "loss": 0.2657,
+      "step": 5649
+    },
+    {
+      "epoch": 0.452,
+      "grad_norm": 1.5093351602554321,
+      "learning_rate": 5.7579409036423426e-06,
+      "loss": 0.3397,
+      "step": 5650
+    },
+    {
+      "epoch": 0.45208,
+      "grad_norm": 1.5804753303527832,
+      "learning_rate": 5.756697770157152e-06,
+      "loss": 0.3578,
+      "step": 5651
+    },
+    {
+      "epoch": 0.45216,
+      "grad_norm": 1.694562315940857,
+      "learning_rate": 5.755454588798226e-06,
+      "loss": 0.3719,
+      "step": 5652
+    },
+    {
+      "epoch": 0.45224,
+      "grad_norm": 1.9576027393341064,
+      "learning_rate": 5.754211359644217e-06,
+      "loss": 0.3491,
+      "step": 5653
+    },
+    {
+      "epoch": 0.45232,
+      "grad_norm": 1.5842691659927368,
+      "learning_rate": 5.752968082773778e-06,
+      "loss": 0.3311,
+      "step": 5654
+    },
+    {
+      "epoch": 0.4524,
+      "grad_norm": 1.2589999437332153,
+      "learning_rate": 5.751724758265567e-06,
+      "loss": 0.2414,
+      "step": 5655
+    },
+    {
+      "epoch": 0.45248,
+      "grad_norm": 2.144315719604492,
+      "learning_rate": 5.750481386198246e-06,
+      "loss": 0.4948,
+      "step": 5656
+    },
+    {
+      "epoch": 0.45256,
+      "grad_norm": 1.865082859992981,
+      "learning_rate": 5.749237966650478e-06,
+      "loss": 0.4131,
+      "step": 5657
+    },
+    {
+      "epoch": 0.45264,
+      "grad_norm": 1.8475942611694336,
+      "learning_rate": 5.747994499700932e-06,
+      "loss": 0.3729,
+      "step": 5658
+    },
+    {
+      "epoch": 0.45272,
+      "grad_norm": 1.8844106197357178,
+      "learning_rate": 5.746750985428278e-06,
+      "loss": 0.5364,
+      "step": 5659
+    },
+    {
+      "epoch": 0.4528,
+      "grad_norm": 1.6912811994552612,
+      "learning_rate": 5.745507423911185e-06,
+      "loss": 0.3392,
+      "step": 5660
+    },
+    {
+      "epoch": 0.45288,
+      "grad_norm": 1.4193228483200073,
+      "learning_rate": 5.744263815228334e-06,
+      "loss": 0.3969,
+      "step": 5661
+    },
+    {
+      "epoch": 0.45296,
+      "grad_norm": 1.7651104927062988,
+      "learning_rate": 5.743020159458401e-06,
+      "loss": 0.4356,
+      "step": 5662
+    },
+    {
+      "epoch": 0.45304,
+      "grad_norm": 1.2144150733947754,
+      "learning_rate": 5.741776456680068e-06,
+      "loss": 0.2408,
+      "step": 5663
+    },
+    {
+      "epoch": 0.45312,
+      "grad_norm": 1.4597735404968262,
+      "learning_rate": 5.740532706972022e-06,
+      "loss": 0.3017,
+      "step": 5664
+    },
+    {
+      "epoch": 0.4532,
+      "grad_norm": 1.3759361505508423,
+      "learning_rate": 5.739288910412949e-06,
+      "loss": 0.2679,
+      "step": 5665
+    },
+    {
+      "epoch": 0.45328,
+      "grad_norm": 1.6209099292755127,
+      "learning_rate": 5.738045067081539e-06,
+      "loss": 0.2948,
+      "step": 5666
+    },
+    {
+      "epoch": 0.45336,
+      "grad_norm": 1.576300859451294,
+      "learning_rate": 5.736801177056488e-06,
+      "loss": 0.3133,
+      "step": 5667
+    },
+    {
+      "epoch": 0.45344,
+      "grad_norm": 1.6057082414627075,
+      "learning_rate": 5.735557240416492e-06,
+      "loss": 0.3194,
+      "step": 5668
+    },
+    {
+      "epoch": 0.45352,
+      "grad_norm": 1.371180772781372,
+      "learning_rate": 5.73431325724025e-06,
+      "loss": 0.2877,
+      "step": 5669
+    },
+    {
+      "epoch": 0.4536,
+      "grad_norm": 1.5031378269195557,
+      "learning_rate": 5.733069227606466e-06,
+      "loss": 0.3562,
+      "step": 5670
+    },
+    {
+      "epoch": 0.45368,
+      "grad_norm": 2.041217803955078,
+      "learning_rate": 5.731825151593845e-06,
+      "loss": 0.4971,
+      "step": 5671
+    },
+    {
+      "epoch": 0.45376,
+      "grad_norm": 1.57797372341156,
+      "learning_rate": 5.730581029281095e-06,
+      "loss": 0.3594,
+      "step": 5672
+    },
+    {
+      "epoch": 0.45384,
+      "grad_norm": 1.516638159751892,
+      "learning_rate": 5.729336860746928e-06,
+      "loss": 0.3172,
+      "step": 5673
+    },
+    {
+      "epoch": 0.45392,
+      "grad_norm": 1.504821538925171,
+      "learning_rate": 5.728092646070058e-06,
+      "loss": 0.3502,
+      "step": 5674
+    },
+    {
+      "epoch": 0.454,
+      "grad_norm": 1.4294888973236084,
+      "learning_rate": 5.726848385329202e-06,
+      "loss": 0.3397,
+      "step": 5675
+    },
+    {
+      "epoch": 0.45408,
+      "grad_norm": 1.1908982992172241,
+      "learning_rate": 5.725604078603081e-06,
+      "loss": 0.2893,
+      "step": 5676
+    },
+    {
+      "epoch": 0.45416,
+      "grad_norm": 1.6021041870117188,
+      "learning_rate": 5.724359725970419e-06,
+      "loss": 0.3478,
+      "step": 5677
+    },
+    {
+      "epoch": 0.45424,
+      "grad_norm": 1.542633056640625,
+      "learning_rate": 5.72311532750994e-06,
+      "loss": 0.3476,
+      "step": 5678
+    },
+    {
+      "epoch": 0.45432,
+      "grad_norm": 1.6992759704589844,
+      "learning_rate": 5.721870883300374e-06,
+      "loss": 0.3492,
+      "step": 5679
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 1.4652698040008545,
+      "learning_rate": 5.720626393420451e-06,
+      "loss": 0.3238,
+      "step": 5680
+    },
+    {
+      "epoch": 0.45448,
+      "grad_norm": 1.3540560007095337,
+      "learning_rate": 5.719381857948908e-06,
+      "loss": 0.3257,
+      "step": 5681
+    },
+    {
+      "epoch": 0.45456,
+      "grad_norm": 1.3179233074188232,
+      "learning_rate": 5.718137276964481e-06,
+      "loss": 0.3374,
+      "step": 5682
+    },
+    {
+      "epoch": 0.45464,
+      "grad_norm": 1.2918938398361206,
+      "learning_rate": 5.716892650545914e-06,
+      "loss": 0.2643,
+      "step": 5683
+    },
+    {
+      "epoch": 0.45472,
+      "grad_norm": 1.5984219312667847,
+      "learning_rate": 5.715647978771946e-06,
+      "loss": 0.3463,
+      "step": 5684
+    },
+    {
+      "epoch": 0.4548,
+      "grad_norm": 1.7894304990768433,
+      "learning_rate": 5.714403261721327e-06,
+      "loss": 0.4156,
+      "step": 5685
+    },
+    {
+      "epoch": 0.45488,
+      "grad_norm": 1.436471939086914,
+      "learning_rate": 5.713158499472802e-06,
+      "loss": 0.3837,
+      "step": 5686
+    },
+    {
+      "epoch": 0.45496,
+      "grad_norm": 1.6427680253982544,
+      "learning_rate": 5.711913692105126e-06,
+      "loss": 0.3569,
+      "step": 5687
+    },
+    {
+      "epoch": 0.45504,
+      "grad_norm": 1.8029788732528687,
+      "learning_rate": 5.710668839697051e-06,
+      "loss": 0.3681,
+      "step": 5688
+    },
+    {
+      "epoch": 0.45512,
+      "grad_norm": 1.501164197921753,
+      "learning_rate": 5.709423942327339e-06,
+      "loss": 0.3378,
+      "step": 5689
+    },
+    {
+      "epoch": 0.4552,
+      "grad_norm": 1.6958562135696411,
+      "learning_rate": 5.708179000074746e-06,
+      "loss": 0.3265,
+      "step": 5690
+    },
+    {
+      "epoch": 0.45528,
+      "grad_norm": 1.5554603338241577,
+      "learning_rate": 5.7069340130180375e-06,
+      "loss": 0.3386,
+      "step": 5691
+    },
+    {
+      "epoch": 0.45536,
+      "grad_norm": 1.462551474571228,
+      "learning_rate": 5.705688981235979e-06,
+      "loss": 0.2857,
+      "step": 5692
+    },
+    {
+      "epoch": 0.45544,
+      "grad_norm": 1.7861984968185425,
+      "learning_rate": 5.704443904807341e-06,
+      "loss": 0.426,
+      "step": 5693
+    },
+    {
+      "epoch": 0.45552,
+      "grad_norm": 2.0433483123779297,
+      "learning_rate": 5.7031987838108945e-06,
+      "loss": 0.3355,
+      "step": 5694
+    },
+    {
+      "epoch": 0.4556,
+      "grad_norm": 1.1957581043243408,
+      "learning_rate": 5.701953618325413e-06,
+      "loss": 0.2442,
+      "step": 5695
+    },
+    {
+      "epoch": 0.45568,
+      "grad_norm": 1.7323002815246582,
+      "learning_rate": 5.700708408429676e-06,
+      "loss": 0.3109,
+      "step": 5696
+    },
+    {
+      "epoch": 0.45576,
+      "grad_norm": 1.2197766304016113,
+      "learning_rate": 5.699463154202461e-06,
+      "loss": 0.3155,
+      "step": 5697
+    },
+    {
+      "epoch": 0.45584,
+      "grad_norm": 1.9804476499557495,
+      "learning_rate": 5.698217855722553e-06,
+      "loss": 0.5005,
+      "step": 5698
+    },
+    {
+      "epoch": 0.45592,
+      "grad_norm": 1.695799708366394,
+      "learning_rate": 5.696972513068738e-06,
+      "loss": 0.3383,
+      "step": 5699
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 1.70815110206604,
+      "learning_rate": 5.695727126319805e-06,
+      "loss": 0.4038,
+      "step": 5700
+    },
+    {
+      "epoch": 0.45608,
+      "grad_norm": 1.5041000843048096,
+      "learning_rate": 5.694481695554542e-06,
+      "loss": 0.3908,
+      "step": 5701
+    },
+    {
+      "epoch": 0.45616,
+      "grad_norm": 1.1593267917633057,
+      "learning_rate": 5.693236220851748e-06,
+      "loss": 0.2464,
+      "step": 5702
+    },
+    {
+      "epoch": 0.45624,
+      "grad_norm": 1.5389819145202637,
+      "learning_rate": 5.691990702290217e-06,
+      "loss": 0.3828,
+      "step": 5703
+    },
+    {
+      "epoch": 0.45632,
+      "grad_norm": 1.3315058946609497,
+      "learning_rate": 5.69074513994875e-06,
+      "loss": 0.3166,
+      "step": 5704
+    },
+    {
+      "epoch": 0.4564,
+      "grad_norm": 1.4812344312667847,
+      "learning_rate": 5.6894995339061484e-06,
+      "loss": 0.294,
+      "step": 5705
+    },
+    {
+      "epoch": 0.45648,
+      "grad_norm": 2.301971197128296,
+      "learning_rate": 5.688253884241221e-06,
+      "loss": 0.4218,
+      "step": 5706
+    },
+    {
+      "epoch": 0.45656,
+      "grad_norm": 1.6880017518997192,
+      "learning_rate": 5.687008191032771e-06,
+      "loss": 0.3879,
+      "step": 5707
+    },
+    {
+      "epoch": 0.45664,
+      "grad_norm": 2.1264901161193848,
+      "learning_rate": 5.685762454359612e-06,
+      "loss": 0.4719,
+      "step": 5708
+    },
+    {
+      "epoch": 0.45672,
+      "grad_norm": 1.2993032932281494,
+      "learning_rate": 5.684516674300557e-06,
+      "loss": 0.2644,
+      "step": 5709
+    },
+    {
+      "epoch": 0.4568,
+      "grad_norm": 1.6621084213256836,
+      "learning_rate": 5.6832708509344215e-06,
+      "loss": 0.3838,
+      "step": 5710
+    },
+    {
+      "epoch": 0.45688,
+      "grad_norm": 1.6229028701782227,
+      "learning_rate": 5.682024984340027e-06,
+      "loss": 0.2956,
+      "step": 5711
+    },
+    {
+      "epoch": 0.45696,
+      "grad_norm": 1.6341001987457275,
+      "learning_rate": 5.6807790745961935e-06,
+      "loss": 0.4009,
+      "step": 5712
+    },
+    {
+      "epoch": 0.45704,
+      "grad_norm": 1.3154834508895874,
+      "learning_rate": 5.679533121781745e-06,
+      "loss": 0.2954,
+      "step": 5713
+    },
+    {
+      "epoch": 0.45712,
+      "grad_norm": 1.2968782186508179,
+      "learning_rate": 5.67828712597551e-06,
+      "loss": 0.2647,
+      "step": 5714
+    },
+    {
+      "epoch": 0.4572,
+      "grad_norm": 1.9382835626602173,
+      "learning_rate": 5.677041087256319e-06,
+      "loss": 0.4006,
+      "step": 5715
+    },
+    {
+      "epoch": 0.45728,
+      "grad_norm": 1.3526051044464111,
+      "learning_rate": 5.675795005703002e-06,
+      "loss": 0.3288,
+      "step": 5716
+    },
+    {
+      "epoch": 0.45736,
+      "grad_norm": 1.6821001768112183,
+      "learning_rate": 5.674548881394398e-06,
+      "loss": 0.3787,
+      "step": 5717
+    },
+    {
+      "epoch": 0.45744,
+      "grad_norm": 1.6728947162628174,
+      "learning_rate": 5.673302714409342e-06,
+      "loss": 0.3524,
+      "step": 5718
+    },
+    {
+      "epoch": 0.45752,
+      "grad_norm": 1.9203122854232788,
+      "learning_rate": 5.672056504826677e-06,
+      "loss": 0.4147,
+      "step": 5719
+    },
+    {
+      "epoch": 0.4576,
+      "grad_norm": 1.767820954322815,
+      "learning_rate": 5.670810252725246e-06,
+      "loss": 0.5169,
+      "step": 5720
+    },
+    {
+      "epoch": 0.45768,
+      "grad_norm": 1.085750937461853,
+      "learning_rate": 5.669563958183893e-06,
+      "loss": 0.2131,
+      "step": 5721
+    },
+    {
+      "epoch": 0.45776,
+      "grad_norm": 2.1359403133392334,
+      "learning_rate": 5.668317621281471e-06,
+      "loss": 0.4206,
+      "step": 5722
+    },
+    {
+      "epoch": 0.45784,
+      "grad_norm": 1.5051878690719604,
+      "learning_rate": 5.667071242096828e-06,
+      "loss": 0.2954,
+      "step": 5723
+    },
+    {
+      "epoch": 0.45792,
+      "grad_norm": 1.2976740598678589,
+      "learning_rate": 5.66582482070882e-06,
+      "loss": 0.284,
+      "step": 5724
+    },
+    {
+      "epoch": 0.458,
+      "grad_norm": 1.684295415878296,
+      "learning_rate": 5.664578357196303e-06,
+      "loss": 0.3567,
+      "step": 5725
+    },
+    {
+      "epoch": 0.45808,
+      "grad_norm": 1.7063944339752197,
+      "learning_rate": 5.663331851638136e-06,
+      "loss": 0.4559,
+      "step": 5726
+    },
+    {
+      "epoch": 0.45816,
+      "grad_norm": 1.91475510597229,
+      "learning_rate": 5.662085304113184e-06,
+      "loss": 0.4579,
+      "step": 5727
+    },
+    {
+      "epoch": 0.45824,
+      "grad_norm": 1.6786874532699585,
+      "learning_rate": 5.66083871470031e-06,
+      "loss": 0.3606,
+      "step": 5728
+    },
+    {
+      "epoch": 0.45832,
+      "grad_norm": 2.066380023956299,
+      "learning_rate": 5.6595920834783815e-06,
+      "loss": 0.3455,
+      "step": 5729
+    },
+    {
+      "epoch": 0.4584,
+      "grad_norm": 1.214521884918213,
+      "learning_rate": 5.658345410526269e-06,
+      "loss": 0.2417,
+      "step": 5730
+    },
+    {
+      "epoch": 0.45848,
+      "grad_norm": 1.282861351966858,
+      "learning_rate": 5.657098695922845e-06,
+      "loss": 0.3091,
+      "step": 5731
+    },
+    {
+      "epoch": 0.45856,
+      "grad_norm": 1.3990386724472046,
+      "learning_rate": 5.655851939746985e-06,
+      "loss": 0.2745,
+      "step": 5732
+    },
+    {
+      "epoch": 0.45864,
+      "grad_norm": 2.0859076976776123,
+      "learning_rate": 5.654605142077567e-06,
+      "loss": 0.3879,
+      "step": 5733
+    },
+    {
+      "epoch": 0.45872,
+      "grad_norm": 1.3720123767852783,
+      "learning_rate": 5.653358302993473e-06,
+      "loss": 0.2961,
+      "step": 5734
+    },
+    {
+      "epoch": 0.4588,
+      "grad_norm": 1.6038662195205688,
+      "learning_rate": 5.652111422573584e-06,
+      "loss": 0.3159,
+      "step": 5735
+    },
+    {
+      "epoch": 0.45888,
+      "grad_norm": 1.583141565322876,
+      "learning_rate": 5.6508645008967885e-06,
+      "loss": 0.3831,
+      "step": 5736
+    },
+    {
+      "epoch": 0.45896,
+      "grad_norm": 1.398667573928833,
+      "learning_rate": 5.649617538041973e-06,
+      "loss": 0.3121,
+      "step": 5737
+    },
+    {
+      "epoch": 0.45904,
+      "grad_norm": 1.727538824081421,
+      "learning_rate": 5.6483705340880305e-06,
+      "loss": 0.3834,
+      "step": 5738
+    },
+    {
+      "epoch": 0.45912,
+      "grad_norm": 1.9524167776107788,
+      "learning_rate": 5.647123489113852e-06,
+      "loss": 0.4467,
+      "step": 5739
+    },
+    {
+      "epoch": 0.4592,
+      "grad_norm": 1.4175621271133423,
+      "learning_rate": 5.645876403198337e-06,
+      "loss": 0.3502,
+      "step": 5740
+    },
+    {
+      "epoch": 0.45928,
+      "grad_norm": 1.418363094329834,
+      "learning_rate": 5.6446292764203825e-06,
+      "loss": 0.3052,
+      "step": 5741
+    },
+    {
+      "epoch": 0.45936,
+      "grad_norm": 1.6092485189437866,
+      "learning_rate": 5.643382108858891e-06,
+      "loss": 0.3851,
+      "step": 5742
+    },
+    {
+      "epoch": 0.45944,
+      "grad_norm": 1.4313188791275024,
+      "learning_rate": 5.642134900592766e-06,
+      "loss": 0.3024,
+      "step": 5743
+    },
+    {
+      "epoch": 0.45952,
+      "grad_norm": 1.4228692054748535,
+      "learning_rate": 5.6408876517009145e-06,
+      "loss": 0.2845,
+      "step": 5744
+    },
+    {
+      "epoch": 0.4596,
+      "grad_norm": 1.565037488937378,
+      "learning_rate": 5.6396403622622455e-06,
+      "loss": 0.3583,
+      "step": 5745
+    },
+    {
+      "epoch": 0.45968,
+      "grad_norm": 1.750694751739502,
+      "learning_rate": 5.638393032355671e-06,
+      "loss": 0.3198,
+      "step": 5746
+    },
+    {
+      "epoch": 0.45976,
+      "grad_norm": 1.5013595819473267,
+      "learning_rate": 5.637145662060106e-06,
+      "loss": 0.3254,
+      "step": 5747
+    },
+    {
+      "epoch": 0.45984,
+      "grad_norm": 1.337208867073059,
+      "learning_rate": 5.635898251454467e-06,
+      "loss": 0.2618,
+      "step": 5748
+    },
+    {
+      "epoch": 0.45992,
+      "grad_norm": 1.4910343885421753,
+      "learning_rate": 5.634650800617672e-06,
+      "loss": 0.2893,
+      "step": 5749
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4709289073944092,
+      "learning_rate": 5.633403309628645e-06,
+      "loss": 0.3244,
+      "step": 5750
+    },
+    {
+      "epoch": 0.46008,
+      "grad_norm": 1.6060997247695923,
+      "learning_rate": 5.6321557785663105e-06,
+      "loss": 0.3631,
+      "step": 5751
+    },
+    {
+      "epoch": 0.46016,
+      "grad_norm": 1.4144808053970337,
+      "learning_rate": 5.630908207509596e-06,
+      "loss": 0.3068,
+      "step": 5752
+    },
+    {
+      "epoch": 0.46024,
+      "grad_norm": 1.6381995677947998,
+      "learning_rate": 5.62966059653743e-06,
+      "loss": 0.3946,
+      "step": 5753
+    },
+    {
+      "epoch": 0.46032,
+      "grad_norm": 1.3315595388412476,
+      "learning_rate": 5.628412945728743e-06,
+      "loss": 0.2966,
+      "step": 5754
+    },
+    {
+      "epoch": 0.4604,
+      "grad_norm": 1.3688490390777588,
+      "learning_rate": 5.627165255162472e-06,
+      "loss": 0.2812,
+      "step": 5755
+    },
+    {
+      "epoch": 0.46048,
+      "grad_norm": 1.2457916736602783,
+      "learning_rate": 5.625917524917555e-06,
+      "loss": 0.316,
+      "step": 5756
+    },
+    {
+      "epoch": 0.46056,
+      "grad_norm": 1.8323296308517456,
+      "learning_rate": 5.624669755072929e-06,
+      "loss": 0.426,
+      "step": 5757
+    },
+    {
+      "epoch": 0.46064,
+      "grad_norm": 1.5934995412826538,
+      "learning_rate": 5.623421945707538e-06,
+      "loss": 0.3284,
+      "step": 5758
+    },
+    {
+      "epoch": 0.46072,
+      "grad_norm": 1.4549036026000977,
+      "learning_rate": 5.622174096900328e-06,
+      "loss": 0.3153,
+      "step": 5759
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 1.9752590656280518,
+      "learning_rate": 5.620926208730244e-06,
+      "loss": 0.3554,
+      "step": 5760
+    },
+    {
+      "epoch": 0.46088,
+      "grad_norm": 1.4583925008773804,
+      "learning_rate": 5.619678281276235e-06,
+      "loss": 0.2659,
+      "step": 5761
+    },
+    {
+      "epoch": 0.46096,
+      "grad_norm": 1.7075728178024292,
+      "learning_rate": 5.618430314617256e-06,
+      "loss": 0.3486,
+      "step": 5762
+    },
+    {
+      "epoch": 0.46104,
+      "grad_norm": 1.5881019830703735,
+      "learning_rate": 5.617182308832261e-06,
+      "loss": 0.3726,
+      "step": 5763
+    },
+    {
+      "epoch": 0.46112,
+      "grad_norm": 1.6844027042388916,
+      "learning_rate": 5.615934264000205e-06,
+      "loss": 0.2897,
+      "step": 5764
+    },
+    {
+      "epoch": 0.4612,
+      "grad_norm": 1.464428186416626,
+      "learning_rate": 5.614686180200051e-06,
+      "loss": 0.2831,
+      "step": 5765
+    },
+    {
+      "epoch": 0.46128,
+      "grad_norm": 1.572912573814392,
+      "learning_rate": 5.613438057510757e-06,
+      "loss": 0.3686,
+      "step": 5766
+    },
+    {
+      "epoch": 0.46136,
+      "grad_norm": 1.6168537139892578,
+      "learning_rate": 5.61218989601129e-06,
+      "loss": 0.4234,
+      "step": 5767
+    },
+    {
+      "epoch": 0.46144,
+      "grad_norm": 1.732604742050171,
+      "learning_rate": 5.610941695780616e-06,
+      "loss": 0.3026,
+      "step": 5768
+    },
+    {
+      "epoch": 0.46152,
+      "grad_norm": 1.795530915260315,
+      "learning_rate": 5.6096934568977065e-06,
+      "loss": 0.4124,
+      "step": 5769
+    },
+    {
+      "epoch": 0.4616,
+      "grad_norm": 1.4683985710144043,
+      "learning_rate": 5.60844517944153e-06,
+      "loss": 0.3594,
+      "step": 5770
+    },
+    {
+      "epoch": 0.46168,
+      "grad_norm": 1.6536790132522583,
+      "learning_rate": 5.607196863491067e-06,
+      "loss": 0.3532,
+      "step": 5771
+    },
+    {
+      "epoch": 0.46176,
+      "grad_norm": 1.6116361618041992,
+      "learning_rate": 5.605948509125288e-06,
+      "loss": 0.3776,
+      "step": 5772
+    },
+    {
+      "epoch": 0.46184,
+      "grad_norm": 1.5489414930343628,
+      "learning_rate": 5.604700116423173e-06,
+      "loss": 0.3541,
+      "step": 5773
+    },
+    {
+      "epoch": 0.46192,
+      "grad_norm": 1.51923668384552,
+      "learning_rate": 5.603451685463706e-06,
+      "loss": 0.3029,
+      "step": 5774
+    },
+    {
+      "epoch": 0.462,
+      "grad_norm": 1.9836581945419312,
+      "learning_rate": 5.60220321632587e-06,
+      "loss": 0.3891,
+      "step": 5775
+    },
+    {
+      "epoch": 0.46208,
+      "grad_norm": 1.7087833881378174,
+      "learning_rate": 5.600954709088651e-06,
+      "loss": 0.4867,
+      "step": 5776
+    },
+    {
+      "epoch": 0.46216,
+      "grad_norm": 1.7288364171981812,
+      "learning_rate": 5.5997061638310405e-06,
+      "loss": 0.321,
+      "step": 5777
+    },
+    {
+      "epoch": 0.46224,
+      "grad_norm": 1.787791132926941,
+      "learning_rate": 5.598457580632025e-06,
+      "loss": 0.4304,
+      "step": 5778
+    },
+    {
+      "epoch": 0.46232,
+      "grad_norm": 1.1242058277130127,
+      "learning_rate": 5.597208959570602e-06,
+      "loss": 0.3531,
+      "step": 5779
+    },
+    {
+      "epoch": 0.4624,
+      "grad_norm": 1.8064368963241577,
+      "learning_rate": 5.595960300725765e-06,
+      "loss": 0.3596,
+      "step": 5780
+    },
+    {
+      "epoch": 0.46248,
+      "grad_norm": 2.188750982284546,
+      "learning_rate": 5.594711604176515e-06,
+      "loss": 0.3838,
+      "step": 5781
+    },
+    {
+      "epoch": 0.46256,
+      "grad_norm": 1.5916962623596191,
+      "learning_rate": 5.593462870001851e-06,
+      "loss": 0.3726,
+      "step": 5782
+    },
+    {
+      "epoch": 0.46264,
+      "grad_norm": 1.3215919733047485,
+      "learning_rate": 5.592214098280778e-06,
+      "loss": 0.3084,
+      "step": 5783
+    },
+    {
+      "epoch": 0.46272,
+      "grad_norm": 1.3940409421920776,
+      "learning_rate": 5.5909652890923004e-06,
+      "loss": 0.3357,
+      "step": 5784
+    },
+    {
+      "epoch": 0.4628,
+      "grad_norm": 1.4418216943740845,
+      "learning_rate": 5.589716442515426e-06,
+      "loss": 0.2619,
+      "step": 5785
+    },
+    {
+      "epoch": 0.46288,
+      "grad_norm": 1.2304273843765259,
+      "learning_rate": 5.588467558629167e-06,
+      "loss": 0.2796,
+      "step": 5786
+    },
+    {
+      "epoch": 0.46296,
+      "grad_norm": 1.5511798858642578,
+      "learning_rate": 5.587218637512532e-06,
+      "loss": 0.2769,
+      "step": 5787
+    },
+    {
+      "epoch": 0.46304,
+      "grad_norm": 1.0966295003890991,
+      "learning_rate": 5.58596967924454e-06,
+      "loss": 0.2145,
+      "step": 5788
+    },
+    {
+      "epoch": 0.46312,
+      "grad_norm": 1.4968533515930176,
+      "learning_rate": 5.5847206839042075e-06,
+      "loss": 0.3548,
+      "step": 5789
+    },
+    {
+      "epoch": 0.4632,
+      "grad_norm": 1.6240606307983398,
+      "learning_rate": 5.5834716515705535e-06,
+      "loss": 0.3444,
+      "step": 5790
+    },
+    {
+      "epoch": 0.46328,
+      "grad_norm": 1.2611035108566284,
+      "learning_rate": 5.5822225823226e-06,
+      "loss": 0.2815,
+      "step": 5791
+    },
+    {
+      "epoch": 0.46336,
+      "grad_norm": 1.897449016571045,
+      "learning_rate": 5.580973476239371e-06,
+      "loss": 0.3436,
+      "step": 5792
+    },
+    {
+      "epoch": 0.46344,
+      "grad_norm": 1.5465292930603027,
+      "learning_rate": 5.5797243333998955e-06,
+      "loss": 0.293,
+      "step": 5793
+    },
+    {
+      "epoch": 0.46352,
+      "grad_norm": 1.9730859994888306,
+      "learning_rate": 5.578475153883201e-06,
+      "loss": 0.3797,
+      "step": 5794
+    },
+    {
+      "epoch": 0.4636,
+      "grad_norm": 1.8337812423706055,
+      "learning_rate": 5.577225937768319e-06,
+      "loss": 0.331,
+      "step": 5795
+    },
+    {
+      "epoch": 0.46368,
+      "grad_norm": 1.6676514148712158,
+      "learning_rate": 5.575976685134282e-06,
+      "loss": 0.3331,
+      "step": 5796
+    },
+    {
+      "epoch": 0.46376,
+      "grad_norm": 1.2623851299285889,
+      "learning_rate": 5.57472739606013e-06,
+      "loss": 0.2177,
+      "step": 5797
+    },
+    {
+      "epoch": 0.46384,
+      "grad_norm": 1.3866863250732422,
+      "learning_rate": 5.573478070624897e-06,
+      "loss": 0.3358,
+      "step": 5798
+    },
+    {
+      "epoch": 0.46392,
+      "grad_norm": 1.5590800046920776,
+      "learning_rate": 5.572228708907626e-06,
+      "loss": 0.3113,
+      "step": 5799
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 1.5001630783081055,
+      "learning_rate": 5.570979310987359e-06,
+      "loss": 0.3252,
+      "step": 5800
+    },
+    {
+      "epoch": 0.46408,
+      "grad_norm": 1.6427888870239258,
+      "learning_rate": 5.5697298769431404e-06,
+      "loss": 0.4096,
+      "step": 5801
+    },
+    {
+      "epoch": 0.46416,
+      "grad_norm": 1.886500358581543,
+      "learning_rate": 5.56848040685402e-06,
+      "loss": 0.3298,
+      "step": 5802
+    },
+    {
+      "epoch": 0.46424,
+      "grad_norm": 1.4350028038024902,
+      "learning_rate": 5.567230900799046e-06,
+      "loss": 0.3214,
+      "step": 5803
+    },
+    {
+      "epoch": 0.46432,
+      "grad_norm": 1.796977162361145,
+      "learning_rate": 5.565981358857271e-06,
+      "loss": 0.3826,
+      "step": 5804
+    },
+    {
+      "epoch": 0.4644,
+      "grad_norm": 1.937929391860962,
+      "learning_rate": 5.564731781107749e-06,
+      "loss": 0.5146,
+      "step": 5805
+    },
+    {
+      "epoch": 0.46448,
+      "grad_norm": 1.6140251159667969,
+      "learning_rate": 5.563482167629537e-06,
+      "loss": 0.3653,
+      "step": 5806
+    },
+    {
+      "epoch": 0.46456,
+      "grad_norm": 1.971123218536377,
+      "learning_rate": 5.562232518501694e-06,
+      "loss": 0.481,
+      "step": 5807
+    },
+    {
+      "epoch": 0.46464,
+      "grad_norm": 1.7507970333099365,
+      "learning_rate": 5.56098283380328e-06,
+      "loss": 0.3149,
+      "step": 5808
+    },
+    {
+      "epoch": 0.46472,
+      "grad_norm": 1.2038670778274536,
+      "learning_rate": 5.55973311361336e-06,
+      "loss": 0.2867,
+      "step": 5809
+    },
+    {
+      "epoch": 0.4648,
+      "grad_norm": 1.8903367519378662,
+      "learning_rate": 5.558483358010999e-06,
+      "loss": 0.3769,
+      "step": 5810
+    },
+    {
+      "epoch": 0.46488,
+      "grad_norm": 1.3014912605285645,
+      "learning_rate": 5.557233567075263e-06,
+      "loss": 0.3411,
+      "step": 5811
+    },
+    {
+      "epoch": 0.46496,
+      "grad_norm": 1.7084940671920776,
+      "learning_rate": 5.555983740885225e-06,
+      "loss": 0.3719,
+      "step": 5812
+    },
+    {
+      "epoch": 0.46504,
+      "grad_norm": 1.7625243663787842,
+      "learning_rate": 5.554733879519956e-06,
+      "loss": 0.4562,
+      "step": 5813
+    },
+    {
+      "epoch": 0.46512,
+      "grad_norm": 1.159744381904602,
+      "learning_rate": 5.553483983058531e-06,
+      "loss": 0.2261,
+      "step": 5814
+    },
+    {
+      "epoch": 0.4652,
+      "grad_norm": 1.9187002182006836,
+      "learning_rate": 5.552234051580024e-06,
+      "loss": 0.3823,
+      "step": 5815
+    },
+    {
+      "epoch": 0.46528,
+      "grad_norm": 1.6461594104766846,
+      "learning_rate": 5.5509840851635185e-06,
+      "loss": 0.3652,
+      "step": 5816
+    },
+    {
+      "epoch": 0.46536,
+      "grad_norm": 1.612166166305542,
+      "learning_rate": 5.549734083888093e-06,
+      "loss": 0.3041,
+      "step": 5817
+    },
+    {
+      "epoch": 0.46544,
+      "grad_norm": 1.890860915184021,
+      "learning_rate": 5.548484047832833e-06,
+      "loss": 0.5575,
+      "step": 5818
+    },
+    {
+      "epoch": 0.46552,
+      "grad_norm": 1.3313186168670654,
+      "learning_rate": 5.54723397707682e-06,
+      "loss": 0.2978,
+      "step": 5819
+    },
+    {
+      "epoch": 0.4656,
+      "grad_norm": 1.85710608959198,
+      "learning_rate": 5.5459838716991465e-06,
+      "loss": 0.3179,
+      "step": 5820
+    },
+    {
+      "epoch": 0.46568,
+      "grad_norm": 1.5691063404083252,
+      "learning_rate": 5.5447337317788986e-06,
+      "loss": 0.3356,
+      "step": 5821
+    },
+    {
+      "epoch": 0.46576,
+      "grad_norm": 1.6753313541412354,
+      "learning_rate": 5.5434835573951704e-06,
+      "loss": 0.2848,
+      "step": 5822
+    },
+    {
+      "epoch": 0.46584,
+      "grad_norm": 1.4183037281036377,
+      "learning_rate": 5.542233348627056e-06,
+      "loss": 0.3081,
+      "step": 5823
+    },
+    {
+      "epoch": 0.46592,
+      "grad_norm": 1.2803384065628052,
+      "learning_rate": 5.540983105553654e-06,
+      "loss": 0.3479,
+      "step": 5824
+    },
+    {
+      "epoch": 0.466,
+      "grad_norm": 1.3356515169143677,
+      "learning_rate": 5.539732828254059e-06,
+      "loss": 0.2828,
+      "step": 5825
+    },
+    {
+      "epoch": 0.46608,
+      "grad_norm": 1.4406170845031738,
+      "learning_rate": 5.538482516807374e-06,
+      "loss": 0.3388,
+      "step": 5826
+    },
+    {
+      "epoch": 0.46616,
+      "grad_norm": 1.526328682899475,
+      "learning_rate": 5.537232171292702e-06,
+      "loss": 0.312,
+      "step": 5827
+    },
+    {
+      "epoch": 0.46624,
+      "grad_norm": 1.2586652040481567,
+      "learning_rate": 5.535981791789148e-06,
+      "loss": 0.2932,
+      "step": 5828
+    },
+    {
+      "epoch": 0.46632,
+      "grad_norm": 1.7249115705490112,
+      "learning_rate": 5.534731378375819e-06,
+      "loss": 0.489,
+      "step": 5829
+    },
+    {
+      "epoch": 0.4664,
+      "grad_norm": 1.484268069267273,
+      "learning_rate": 5.533480931131827e-06,
+      "loss": 0.2965,
+      "step": 5830
+    },
+    {
+      "epoch": 0.46648,
+      "grad_norm": 1.7964249849319458,
+      "learning_rate": 5.53223045013628e-06,
+      "loss": 0.4045,
+      "step": 5831
+    },
+    {
+      "epoch": 0.46656,
+      "grad_norm": 1.7476415634155273,
+      "learning_rate": 5.530979935468294e-06,
+      "loss": 0.3451,
+      "step": 5832
+    },
+    {
+      "epoch": 0.46664,
+      "grad_norm": 1.3235749006271362,
+      "learning_rate": 5.529729387206983e-06,
+      "loss": 0.2304,
+      "step": 5833
+    },
+    {
+      "epoch": 0.46672,
+      "grad_norm": 1.5181303024291992,
+      "learning_rate": 5.5284788054314665e-06,
+      "loss": 0.314,
+      "step": 5834
+    },
+    {
+      "epoch": 0.4668,
+      "grad_norm": 1.5556458234786987,
+      "learning_rate": 5.527228190220866e-06,
+      "loss": 0.3762,
+      "step": 5835
+    },
+    {
+      "epoch": 0.46688,
+      "grad_norm": 1.3270909786224365,
+      "learning_rate": 5.525977541654299e-06,
+      "loss": 0.2586,
+      "step": 5836
+    },
+    {
+      "epoch": 0.46696,
+      "grad_norm": 1.6928744316101074,
+      "learning_rate": 5.524726859810895e-06,
+      "loss": 0.3365,
+      "step": 5837
+    },
+    {
+      "epoch": 0.46704,
+      "grad_norm": 1.3896623849868774,
+      "learning_rate": 5.523476144769777e-06,
+      "loss": 0.3052,
+      "step": 5838
+    },
+    {
+      "epoch": 0.46712,
+      "grad_norm": 1.6852587461471558,
+      "learning_rate": 5.522225396610076e-06,
+      "loss": 0.3974,
+      "step": 5839
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 1.4615007638931274,
+      "learning_rate": 5.520974615410921e-06,
+      "loss": 0.3015,
+      "step": 5840
+    },
+    {
+      "epoch": 0.46728,
+      "grad_norm": 1.2389006614685059,
+      "learning_rate": 5.519723801251445e-06,
+      "loss": 0.2909,
+      "step": 5841
+    },
+    {
+      "epoch": 0.46736,
+      "grad_norm": 1.441849946975708,
+      "learning_rate": 5.518472954210784e-06,
+      "loss": 0.3162,
+      "step": 5842
+    },
+    {
+      "epoch": 0.46744,
+      "grad_norm": 1.5538463592529297,
+      "learning_rate": 5.5172220743680745e-06,
+      "loss": 0.3241,
+      "step": 5843
+    },
+    {
+      "epoch": 0.46752,
+      "grad_norm": 1.311487078666687,
+      "learning_rate": 5.515971161802454e-06,
+      "loss": 0.2759,
+      "step": 5844
+    },
+    {
+      "epoch": 0.4676,
+      "grad_norm": 1.836043119430542,
+      "learning_rate": 5.514720216593063e-06,
+      "loss": 0.4221,
+      "step": 5845
+    },
+    {
+      "epoch": 0.46768,
+      "grad_norm": 1.3168381452560425,
+      "learning_rate": 5.513469238819048e-06,
+      "loss": 0.3044,
+      "step": 5846
+    },
+    {
+      "epoch": 0.46776,
+      "grad_norm": 1.2070047855377197,
+      "learning_rate": 5.5122182285595525e-06,
+      "loss": 0.229,
+      "step": 5847
+    },
+    {
+      "epoch": 0.46784,
+      "grad_norm": 1.5772783756256104,
+      "learning_rate": 5.510967185893723e-06,
+      "loss": 0.3556,
+      "step": 5848
+    },
+    {
+      "epoch": 0.46792,
+      "grad_norm": 1.5532734394073486,
+      "learning_rate": 5.509716110900709e-06,
+      "loss": 0.3562,
+      "step": 5849
+    },
+    {
+      "epoch": 0.468,
+      "grad_norm": 1.7133153676986694,
+      "learning_rate": 5.508465003659663e-06,
+      "loss": 0.3732,
+      "step": 5850
+    },
+    {
+      "epoch": 0.46808,
+      "grad_norm": 1.6906065940856934,
+      "learning_rate": 5.5072138642497365e-06,
+      "loss": 0.398,
+      "step": 5851
+    },
+    {
+      "epoch": 0.46816,
+      "grad_norm": 1.8008275032043457,
+      "learning_rate": 5.505962692750087e-06,
+      "loss": 0.3578,
+      "step": 5852
+    },
+    {
+      "epoch": 0.46824,
+      "grad_norm": 2.004988193511963,
+      "learning_rate": 5.504711489239871e-06,
+      "loss": 0.4561,
+      "step": 5853
+    },
+    {
+      "epoch": 0.46832,
+      "grad_norm": 1.8562557697296143,
+      "learning_rate": 5.5034602537982485e-06,
+      "loss": 0.5832,
+      "step": 5854
+    },
+    {
+      "epoch": 0.4684,
+      "grad_norm": 1.4720324277877808,
+      "learning_rate": 5.502208986504378e-06,
+      "loss": 0.2837,
+      "step": 5855
+    },
+    {
+      "epoch": 0.46848,
+      "grad_norm": 1.579901099205017,
+      "learning_rate": 5.500957687437427e-06,
+      "loss": 0.2924,
+      "step": 5856
+    },
+    {
+      "epoch": 0.46856,
+      "grad_norm": 1.6447190046310425,
+      "learning_rate": 5.499706356676559e-06,
+      "loss": 0.325,
+      "step": 5857
+    },
+    {
+      "epoch": 0.46864,
+      "grad_norm": 1.4578676223754883,
+      "learning_rate": 5.498454994300941e-06,
+      "loss": 0.341,
+      "step": 5858
+    },
+    {
+      "epoch": 0.46872,
+      "grad_norm": 1.5523934364318848,
+      "learning_rate": 5.497203600389744e-06,
+      "loss": 0.4472,
+      "step": 5859
+    },
+    {
+      "epoch": 0.4688,
+      "grad_norm": 1.6504297256469727,
+      "learning_rate": 5.495952175022139e-06,
+      "loss": 0.3244,
+      "step": 5860
+    },
+    {
+      "epoch": 0.46888,
+      "grad_norm": 1.4491195678710938,
+      "learning_rate": 5.4947007182772995e-06,
+      "loss": 0.358,
+      "step": 5861
+    },
+    {
+      "epoch": 0.46896,
+      "grad_norm": 1.6385431289672852,
+      "learning_rate": 5.4934492302344e-06,
+      "loss": 0.3625,
+      "step": 5862
+    },
+    {
+      "epoch": 0.46904,
+      "grad_norm": 1.3259345293045044,
+      "learning_rate": 5.492197710972618e-06,
+      "loss": 0.3199,
+      "step": 5863
+    },
+    {
+      "epoch": 0.46912,
+      "grad_norm": 1.4403702020645142,
+      "learning_rate": 5.4909461605711365e-06,
+      "loss": 0.2822,
+      "step": 5864
+    },
+    {
+      "epoch": 0.4692,
+      "grad_norm": 1.74454665184021,
+      "learning_rate": 5.489694579109133e-06,
+      "loss": 0.3237,
+      "step": 5865
+    },
+    {
+      "epoch": 0.46928,
+      "grad_norm": 1.565183162689209,
+      "learning_rate": 5.4884429666657925e-06,
+      "loss": 0.3899,
+      "step": 5866
+    },
+    {
+      "epoch": 0.46936,
+      "grad_norm": 1.3134307861328125,
+      "learning_rate": 5.487191323320298e-06,
+      "loss": 0.421,
+      "step": 5867
+    },
+    {
+      "epoch": 0.46944,
+      "grad_norm": 1.5535221099853516,
+      "learning_rate": 5.48593964915184e-06,
+      "loss": 0.403,
+      "step": 5868
+    },
+    {
+      "epoch": 0.46952,
+      "grad_norm": 1.4464749097824097,
+      "learning_rate": 5.484687944239605e-06,
+      "loss": 0.3067,
+      "step": 5869
+    },
+    {
+      "epoch": 0.4696,
+      "grad_norm": 1.3258053064346313,
+      "learning_rate": 5.483436208662787e-06,
+      "loss": 0.3042,
+      "step": 5870
+    },
+    {
+      "epoch": 0.46968,
+      "grad_norm": 1.434024453163147,
+      "learning_rate": 5.482184442500578e-06,
+      "loss": 0.3709,
+      "step": 5871
+    },
+    {
+      "epoch": 0.46976,
+      "grad_norm": 1.493878960609436,
+      "learning_rate": 5.480932645832171e-06,
+      "loss": 0.2702,
+      "step": 5872
+    },
+    {
+      "epoch": 0.46984,
+      "grad_norm": 1.5920265913009644,
+      "learning_rate": 5.479680818736765e-06,
+      "loss": 0.3356,
+      "step": 5873
+    },
+    {
+      "epoch": 0.46992,
+      "grad_norm": 1.51193106174469,
+      "learning_rate": 5.478428961293559e-06,
+      "loss": 0.3876,
+      "step": 5874
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8197968006134033,
+      "learning_rate": 5.477177073581754e-06,
+      "loss": 0.4746,
+      "step": 5875
+    },
+    {
+      "epoch": 0.47008,
+      "grad_norm": 1.4704205989837646,
+      "learning_rate": 5.475925155680552e-06,
+      "loss": 0.2828,
+      "step": 5876
+    },
+    {
+      "epoch": 0.47016,
+      "grad_norm": 1.455899953842163,
+      "learning_rate": 5.474673207669159e-06,
+      "loss": 0.2777,
+      "step": 5877
+    },
+    {
+      "epoch": 0.47024,
+      "grad_norm": 1.9199295043945312,
+      "learning_rate": 5.473421229626779e-06,
+      "loss": 0.4936,
+      "step": 5878
+    },
+    {
+      "epoch": 0.47032,
+      "grad_norm": 1.642419457435608,
+      "learning_rate": 5.472169221632622e-06,
+      "loss": 0.4494,
+      "step": 5879
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 1.428835391998291,
+      "learning_rate": 5.470917183765898e-06,
+      "loss": 0.3226,
+      "step": 5880
+    },
+    {
+      "epoch": 0.47048,
+      "grad_norm": 1.3991670608520508,
+      "learning_rate": 5.4696651161058186e-06,
+      "loss": 0.2963,
+      "step": 5881
+    },
+    {
+      "epoch": 0.47056,
+      "grad_norm": 1.5357342958450317,
+      "learning_rate": 5.468413018731601e-06,
+      "loss": 0.3023,
+      "step": 5882
+    },
+    {
+      "epoch": 0.47064,
+      "grad_norm": 1.5561589002609253,
+      "learning_rate": 5.467160891722459e-06,
+      "loss": 0.3432,
+      "step": 5883
+    },
+    {
+      "epoch": 0.47072,
+      "grad_norm": 1.3665988445281982,
+      "learning_rate": 5.465908735157608e-06,
+      "loss": 0.333,
+      "step": 5884
+    },
+    {
+      "epoch": 0.4708,
+      "grad_norm": 1.4518331289291382,
+      "learning_rate": 5.4646565491162716e-06,
+      "loss": 0.3156,
+      "step": 5885
+    },
+    {
+      "epoch": 0.47088,
+      "grad_norm": 1.7487683296203613,
+      "learning_rate": 5.46340433367767e-06,
+      "loss": 0.3997,
+      "step": 5886
+    },
+    {
+      "epoch": 0.47096,
+      "grad_norm": 1.7349579334259033,
+      "learning_rate": 5.462152088921028e-06,
+      "loss": 0.3941,
+      "step": 5887
+    },
+    {
+      "epoch": 0.47104,
+      "grad_norm": 1.3524267673492432,
+      "learning_rate": 5.460899814925567e-06,
+      "loss": 0.3129,
+      "step": 5888
+    },
+    {
+      "epoch": 0.47112,
+      "grad_norm": 1.5749634504318237,
+      "learning_rate": 5.459647511770521e-06,
+      "loss": 0.4254,
+      "step": 5889
+    },
+    {
+      "epoch": 0.4712,
+      "grad_norm": 1.641700267791748,
+      "learning_rate": 5.458395179535112e-06,
+      "loss": 0.3265,
+      "step": 5890
+    },
+    {
+      "epoch": 0.47128,
+      "grad_norm": 1.3452178239822388,
+      "learning_rate": 5.457142818298573e-06,
+      "loss": 0.3459,
+      "step": 5891
+    },
+    {
+      "epoch": 0.47136,
+      "grad_norm": 1.5965007543563843,
+      "learning_rate": 5.455890428140139e-06,
+      "loss": 0.3374,
+      "step": 5892
+    },
+    {
+      "epoch": 0.47144,
+      "grad_norm": 1.6195091009140015,
+      "learning_rate": 5.4546380091390425e-06,
+      "loss": 0.4122,
+      "step": 5893
+    },
+    {
+      "epoch": 0.47152,
+      "grad_norm": 2.059089422225952,
+      "learning_rate": 5.453385561374521e-06,
+      "loss": 0.5191,
+      "step": 5894
+    },
+    {
+      "epoch": 0.4716,
+      "grad_norm": 1.6009624004364014,
+      "learning_rate": 5.452133084925812e-06,
+      "loss": 0.3699,
+      "step": 5895
+    },
+    {
+      "epoch": 0.47168,
+      "grad_norm": 1.5404019355773926,
+      "learning_rate": 5.450880579872156e-06,
+      "loss": 0.347,
+      "step": 5896
+    },
+    {
+      "epoch": 0.47176,
+      "grad_norm": 2.4010934829711914,
+      "learning_rate": 5.449628046292792e-06,
+      "loss": 0.4262,
+      "step": 5897
+    },
+    {
+      "epoch": 0.47184,
+      "grad_norm": 1.5734963417053223,
+      "learning_rate": 5.448375484266968e-06,
+      "loss": 0.4524,
+      "step": 5898
+    },
+    {
+      "epoch": 0.47192,
+      "grad_norm": 1.7515249252319336,
+      "learning_rate": 5.447122893873927e-06,
+      "loss": 0.4706,
+      "step": 5899
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 1.6091927289962769,
+      "learning_rate": 5.445870275192918e-06,
+      "loss": 0.3039,
+      "step": 5900
+    },
+    {
+      "epoch": 0.47208,
+      "grad_norm": 1.3763673305511475,
+      "learning_rate": 5.444617628303187e-06,
+      "loss": 0.2805,
+      "step": 5901
+    },
+    {
+      "epoch": 0.47216,
+      "grad_norm": 1.7855626344680786,
+      "learning_rate": 5.443364953283986e-06,
+      "loss": 0.3412,
+      "step": 5902
+    },
+    {
+      "epoch": 0.47224,
+      "grad_norm": 1.6489816904067993,
+      "learning_rate": 5.442112250214569e-06,
+      "loss": 0.3524,
+      "step": 5903
+    },
+    {
+      "epoch": 0.47232,
+      "grad_norm": 1.6601721048355103,
+      "learning_rate": 5.440859519174187e-06,
+      "loss": 0.3718,
+      "step": 5904
+    },
+    {
+      "epoch": 0.4724,
+      "grad_norm": 1.6896872520446777,
+      "learning_rate": 5.4396067602421e-06,
+      "loss": 0.3397,
+      "step": 5905
+    },
+    {
+      "epoch": 0.47248,
+      "grad_norm": 1.254325270652771,
+      "learning_rate": 5.438353973497565e-06,
+      "loss": 0.2802,
+      "step": 5906
+    },
+    {
+      "epoch": 0.47256,
+      "grad_norm": 2.1470906734466553,
+      "learning_rate": 5.43710115901984e-06,
+      "loss": 0.5072,
+      "step": 5907
+    },
+    {
+      "epoch": 0.47264,
+      "grad_norm": 1.403336763381958,
+      "learning_rate": 5.435848316888187e-06,
+      "loss": 0.2941,
+      "step": 5908
+    },
+    {
+      "epoch": 0.47272,
+      "grad_norm": 1.7086148262023926,
+      "learning_rate": 5.4345954471818695e-06,
+      "loss": 0.3538,
+      "step": 5909
+    },
+    {
+      "epoch": 0.4728,
+      "grad_norm": 1.3657201528549194,
+      "learning_rate": 5.433342549980153e-06,
+      "loss": 0.2622,
+      "step": 5910
+    },
+    {
+      "epoch": 0.47288,
+      "grad_norm": 1.6097228527069092,
+      "learning_rate": 5.432089625362302e-06,
+      "loss": 0.4418,
+      "step": 5911
+    },
+    {
+      "epoch": 0.47296,
+      "grad_norm": 1.664275050163269,
+      "learning_rate": 5.430836673407588e-06,
+      "loss": 0.3672,
+      "step": 5912
+    },
+    {
+      "epoch": 0.47304,
+      "grad_norm": 1.4426720142364502,
+      "learning_rate": 5.4295836941952775e-06,
+      "loss": 0.33,
+      "step": 5913
+    },
+    {
+      "epoch": 0.47312,
+      "grad_norm": 1.6172068119049072,
+      "learning_rate": 5.428330687804643e-06,
+      "loss": 0.4129,
+      "step": 5914
+    },
+    {
+      "epoch": 0.4732,
+      "grad_norm": 1.322485327720642,
+      "learning_rate": 5.427077654314961e-06,
+      "loss": 0.3158,
+      "step": 5915
+    },
+    {
+      "epoch": 0.47328,
+      "grad_norm": 1.6107004880905151,
+      "learning_rate": 5.425824593805505e-06,
+      "loss": 0.3416,
+      "step": 5916
+    },
+    {
+      "epoch": 0.47336,
+      "grad_norm": 2.201343297958374,
+      "learning_rate": 5.424571506355552e-06,
+      "loss": 0.3547,
+      "step": 5917
+    },
+    {
+      "epoch": 0.47344,
+      "grad_norm": 1.490373969078064,
+      "learning_rate": 5.42331839204438e-06,
+      "loss": 0.3098,
+      "step": 5918
+    },
+    {
+      "epoch": 0.47352,
+      "grad_norm": 1.3307541608810425,
+      "learning_rate": 5.422065250951268e-06,
+      "loss": 0.3176,
+      "step": 5919
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 1.459407925605774,
+      "learning_rate": 5.420812083155502e-06,
+      "loss": 0.3212,
+      "step": 5920
+    },
+    {
+      "epoch": 0.47368,
+      "grad_norm": 1.4235657453536987,
+      "learning_rate": 5.4195588887363635e-06,
+      "loss": 0.3285,
+      "step": 5921
+    },
+    {
+      "epoch": 0.47376,
+      "grad_norm": 1.2670882940292358,
+      "learning_rate": 5.4183056677731376e-06,
+      "loss": 0.3147,
+      "step": 5922
+    },
+    {
+      "epoch": 0.47384,
+      "grad_norm": 1.553990125656128,
+      "learning_rate": 5.41705242034511e-06,
+      "loss": 0.4067,
+      "step": 5923
+    },
+    {
+      "epoch": 0.47392,
+      "grad_norm": 1.3905086517333984,
+      "learning_rate": 5.415799146531574e-06,
+      "loss": 0.3246,
+      "step": 5924
+    },
+    {
+      "epoch": 0.474,
+      "grad_norm": 1.6242258548736572,
+      "learning_rate": 5.414545846411815e-06,
+      "loss": 0.3597,
+      "step": 5925
+    },
+    {
+      "epoch": 0.47408,
+      "grad_norm": 2.0094828605651855,
+      "learning_rate": 5.413292520065129e-06,
+      "loss": 0.3506,
+      "step": 5926
+    },
+    {
+      "epoch": 0.47416,
+      "grad_norm": 1.0113637447357178,
+      "learning_rate": 5.4120391675708065e-06,
+      "loss": 0.2126,
+      "step": 5927
+    },
+    {
+      "epoch": 0.47424,
+      "grad_norm": 1.8587594032287598,
+      "learning_rate": 5.410785789008145e-06,
+      "loss": 0.4143,
+      "step": 5928
+    },
+    {
+      "epoch": 0.47432,
+      "grad_norm": 1.5328631401062012,
+      "learning_rate": 5.4095323844564425e-06,
+      "loss": 0.3364,
+      "step": 5929
+    },
+    {
+      "epoch": 0.4744,
+      "grad_norm": 1.8271329402923584,
+      "learning_rate": 5.408278953994996e-06,
+      "loss": 0.3848,
+      "step": 5930
+    },
+    {
+      "epoch": 0.47448,
+      "grad_norm": 1.4636245965957642,
+      "learning_rate": 5.407025497703105e-06,
+      "loss": 0.3291,
+      "step": 5931
+    },
+    {
+      "epoch": 0.47456,
+      "grad_norm": 1.4129557609558105,
+      "learning_rate": 5.405772015660072e-06,
+      "loss": 0.3842,
+      "step": 5932
+    },
+    {
+      "epoch": 0.47464,
+      "grad_norm": 1.7334128618240356,
+      "learning_rate": 5.404518507945205e-06,
+      "loss": 0.4864,
+      "step": 5933
+    },
+    {
+      "epoch": 0.47472,
+      "grad_norm": 1.5284255743026733,
+      "learning_rate": 5.403264974637802e-06,
+      "loss": 0.3454,
+      "step": 5934
+    },
+    {
+      "epoch": 0.4748,
+      "grad_norm": 1.4403399229049683,
+      "learning_rate": 5.402011415817176e-06,
+      "loss": 0.29,
+      "step": 5935
+    },
+    {
+      "epoch": 0.47488,
+      "grad_norm": 1.4352822303771973,
+      "learning_rate": 5.400757831562631e-06,
+      "loss": 0.2549,
+      "step": 5936
+    },
+    {
+      "epoch": 0.47496,
+      "grad_norm": 1.5623631477355957,
+      "learning_rate": 5.399504221953478e-06,
+      "loss": 0.2864,
+      "step": 5937
+    },
+    {
+      "epoch": 0.47504,
+      "grad_norm": 1.2767705917358398,
+      "learning_rate": 5.3982505870690316e-06,
+      "loss": 0.289,
+      "step": 5938
+    },
+    {
+      "epoch": 0.47512,
+      "grad_norm": 1.3673750162124634,
+      "learning_rate": 5.396996926988601e-06,
+      "loss": 0.2931,
+      "step": 5939
+    },
+    {
+      "epoch": 0.4752,
+      "grad_norm": 1.759554386138916,
+      "learning_rate": 5.395743241791504e-06,
+      "loss": 0.3399,
+      "step": 5940
+    },
+    {
+      "epoch": 0.47528,
+      "grad_norm": 1.6224939823150635,
+      "learning_rate": 5.394489531557059e-06,
+      "loss": 0.3372,
+      "step": 5941
+    },
+    {
+      "epoch": 0.47536,
+      "grad_norm": 1.5486100912094116,
+      "learning_rate": 5.393235796364578e-06,
+      "loss": 0.2582,
+      "step": 5942
+    },
+    {
+      "epoch": 0.47544,
+      "grad_norm": 0.9767140746116638,
+      "learning_rate": 5.391982036293385e-06,
+      "loss": 0.1933,
+      "step": 5943
+    },
+    {
+      "epoch": 0.47552,
+      "grad_norm": 1.7568838596343994,
+      "learning_rate": 5.390728251422801e-06,
+      "loss": 0.3333,
+      "step": 5944
+    },
+    {
+      "epoch": 0.4756,
+      "grad_norm": 1.5353167057037354,
+      "learning_rate": 5.389474441832148e-06,
+      "loss": 0.3076,
+      "step": 5945
+    },
+    {
+      "epoch": 0.47568,
+      "grad_norm": 1.9737414121627808,
+      "learning_rate": 5.388220607600748e-06,
+      "loss": 0.434,
+      "step": 5946
+    },
+    {
+      "epoch": 0.47576,
+      "grad_norm": 1.3293046951293945,
+      "learning_rate": 5.386966748807932e-06,
+      "loss": 0.2834,
+      "step": 5947
+    },
+    {
+      "epoch": 0.47584,
+      "grad_norm": 1.4846619367599487,
+      "learning_rate": 5.385712865533023e-06,
+      "loss": 0.3274,
+      "step": 5948
+    },
+    {
+      "epoch": 0.47592,
+      "grad_norm": 1.7403323650360107,
+      "learning_rate": 5.384458957855351e-06,
+      "loss": 0.3113,
+      "step": 5949
+    },
+    {
+      "epoch": 0.476,
+      "grad_norm": 1.7294260263442993,
+      "learning_rate": 5.383205025854248e-06,
+      "loss": 0.3651,
+      "step": 5950
+    },
+    {
+      "epoch": 0.47608,
+      "grad_norm": 1.7438633441925049,
+      "learning_rate": 5.381951069609045e-06,
+      "loss": 0.3918,
+      "step": 5951
+    },
+    {
+      "epoch": 0.47616,
+      "grad_norm": 1.3631834983825684,
+      "learning_rate": 5.380697089199075e-06,
+      "loss": 0.2842,
+      "step": 5952
+    },
+    {
+      "epoch": 0.47624,
+      "grad_norm": 1.3905704021453857,
+      "learning_rate": 5.379443084703676e-06,
+      "loss": 0.2753,
+      "step": 5953
+    },
+    {
+      "epoch": 0.47632,
+      "grad_norm": 1.5892817974090576,
+      "learning_rate": 5.378189056202181e-06,
+      "loss": 0.3348,
+      "step": 5954
+    },
+    {
+      "epoch": 0.4764,
+      "grad_norm": 1.856491208076477,
+      "learning_rate": 5.376935003773931e-06,
+      "loss": 0.3874,
+      "step": 5955
+    },
+    {
+      "epoch": 0.47648,
+      "grad_norm": 1.2700897455215454,
+      "learning_rate": 5.375680927498265e-06,
+      "loss": 0.309,
+      "step": 5956
+    },
+    {
+      "epoch": 0.47656,
+      "grad_norm": 1.3322601318359375,
+      "learning_rate": 5.374426827454522e-06,
+      "loss": 0.2654,
+      "step": 5957
+    },
+    {
+      "epoch": 0.47664,
+      "grad_norm": 2.0289957523345947,
+      "learning_rate": 5.373172703722046e-06,
+      "loss": 0.5895,
+      "step": 5958
+    },
+    {
+      "epoch": 0.47672,
+      "grad_norm": 1.688299298286438,
+      "learning_rate": 5.371918556380185e-06,
+      "loss": 0.4384,
+      "step": 5959
+    },
+    {
+      "epoch": 0.4768,
+      "grad_norm": 1.5574098825454712,
+      "learning_rate": 5.370664385508278e-06,
+      "loss": 0.3618,
+      "step": 5960
+    },
+    {
+      "epoch": 0.47688,
+      "grad_norm": 1.470866322517395,
+      "learning_rate": 5.369410191185676e-06,
+      "loss": 0.3197,
+      "step": 5961
+    },
+    {
+      "epoch": 0.47696,
+      "grad_norm": 1.5503671169281006,
+      "learning_rate": 5.368155973491729e-06,
+      "loss": 0.4128,
+      "step": 5962
+    },
+    {
+      "epoch": 0.47704,
+      "grad_norm": 1.393202304840088,
+      "learning_rate": 5.366901732505784e-06,
+      "loss": 0.2566,
+      "step": 5963
+    },
+    {
+      "epoch": 0.47712,
+      "grad_norm": 1.478087306022644,
+      "learning_rate": 5.365647468307193e-06,
+      "loss": 0.3087,
+      "step": 5964
+    },
+    {
+      "epoch": 0.4772,
+      "grad_norm": 1.4363526105880737,
+      "learning_rate": 5.364393180975314e-06,
+      "loss": 0.2757,
+      "step": 5965
+    },
+    {
+      "epoch": 0.47728,
+      "grad_norm": 1.8411908149719238,
+      "learning_rate": 5.363138870589495e-06,
+      "loss": 0.4493,
+      "step": 5966
+    },
+    {
+      "epoch": 0.47736,
+      "grad_norm": 1.8688772916793823,
+      "learning_rate": 5.361884537229095e-06,
+      "loss": 0.3322,
+      "step": 5967
+    },
+    {
+      "epoch": 0.47744,
+      "grad_norm": 1.6204429864883423,
+      "learning_rate": 5.360630180973472e-06,
+      "loss": 0.321,
+      "step": 5968
+    },
+    {
+      "epoch": 0.47752,
+      "grad_norm": 1.8783400058746338,
+      "learning_rate": 5.359375801901982e-06,
+      "loss": 0.3558,
+      "step": 5969
+    },
+    {
+      "epoch": 0.4776,
+      "grad_norm": 1.60300874710083,
+      "learning_rate": 5.358121400093989e-06,
+      "loss": 0.3528,
+      "step": 5970
+    },
+    {
+      "epoch": 0.47768,
+      "grad_norm": 1.5977694988250732,
+      "learning_rate": 5.356866975628854e-06,
+      "loss": 0.3363,
+      "step": 5971
+    },
+    {
+      "epoch": 0.47776,
+      "grad_norm": 1.698292851448059,
+      "learning_rate": 5.355612528585938e-06,
+      "loss": 0.391,
+      "step": 5972
+    },
+    {
+      "epoch": 0.47784,
+      "grad_norm": 1.7773863077163696,
+      "learning_rate": 5.354358059044608e-06,
+      "loss": 0.3306,
+      "step": 5973
+    },
+    {
+      "epoch": 0.47792,
+      "grad_norm": 1.5922447443008423,
+      "learning_rate": 5.353103567084229e-06,
+      "loss": 0.3131,
+      "step": 5974
+    },
+    {
+      "epoch": 0.478,
+      "grad_norm": 1.8159767389297485,
+      "learning_rate": 5.3518490527841685e-06,
+      "loss": 0.3942,
+      "step": 5975
+    },
+    {
+      "epoch": 0.47808,
+      "grad_norm": 1.657160997390747,
+      "learning_rate": 5.350594516223797e-06,
+      "loss": 0.4472,
+      "step": 5976
+    },
+    {
+      "epoch": 0.47816,
+      "grad_norm": 2.0365302562713623,
+      "learning_rate": 5.3493399574824824e-06,
+      "loss": 0.3914,
+      "step": 5977
+    },
+    {
+      "epoch": 0.47824,
+      "grad_norm": 1.715842843055725,
+      "learning_rate": 5.348085376639598e-06,
+      "loss": 0.3958,
+      "step": 5978
+    },
+    {
+      "epoch": 0.47832,
+      "grad_norm": 1.8368576765060425,
+      "learning_rate": 5.3468307737745175e-06,
+      "loss": 0.3646,
+      "step": 5979
+    },
+    {
+      "epoch": 0.4784,
+      "grad_norm": 1.8690496683120728,
+      "learning_rate": 5.345576148966612e-06,
+      "loss": 0.5069,
+      "step": 5980
+    },
+    {
+      "epoch": 0.47848,
+      "grad_norm": 1.9646300077438354,
+      "learning_rate": 5.344321502295262e-06,
+      "loss": 0.4976,
+      "step": 5981
+    },
+    {
+      "epoch": 0.47856,
+      "grad_norm": 1.7964569330215454,
+      "learning_rate": 5.3430668338398425e-06,
+      "loss": 0.3761,
+      "step": 5982
+    },
+    {
+      "epoch": 0.47864,
+      "grad_norm": 1.9966002702713013,
+      "learning_rate": 5.341812143679732e-06,
+      "loss": 0.4334,
+      "step": 5983
+    },
+    {
+      "epoch": 0.47872,
+      "grad_norm": 2.4749326705932617,
+      "learning_rate": 5.34055743189431e-06,
+      "loss": 0.5564,
+      "step": 5984
+    },
+    {
+      "epoch": 0.4788,
+      "grad_norm": 1.9291296005249023,
+      "learning_rate": 5.339302698562959e-06,
+      "loss": 0.3668,
+      "step": 5985
+    },
+    {
+      "epoch": 0.47888,
+      "grad_norm": 1.5976835489273071,
+      "learning_rate": 5.3380479437650625e-06,
+      "loss": 0.3629,
+      "step": 5986
+    },
+    {
+      "epoch": 0.47896,
+      "grad_norm": 1.5634150505065918,
+      "learning_rate": 5.336793167580002e-06,
+      "loss": 0.3366,
+      "step": 5987
+    },
+    {
+      "epoch": 0.47904,
+      "grad_norm": 1.9057620763778687,
+      "learning_rate": 5.3355383700871665e-06,
+      "loss": 0.3585,
+      "step": 5988
+    },
+    {
+      "epoch": 0.47912,
+      "grad_norm": 1.6130526065826416,
+      "learning_rate": 5.334283551365941e-06,
+      "loss": 0.4087,
+      "step": 5989
+    },
+    {
+      "epoch": 0.4792,
+      "grad_norm": 2.0664384365081787,
+      "learning_rate": 5.333028711495713e-06,
+      "loss": 0.5229,
+      "step": 5990
+    },
+    {
+      "epoch": 0.47928,
+      "grad_norm": 1.7554843425750732,
+      "learning_rate": 5.331773850555874e-06,
+      "loss": 0.3727,
+      "step": 5991
+    },
+    {
+      "epoch": 0.47936,
+      "grad_norm": 1.2213190793991089,
+      "learning_rate": 5.330518968625812e-06,
+      "loss": 0.2522,
+      "step": 5992
+    },
+    {
+      "epoch": 0.47944,
+      "grad_norm": 1.3879023790359497,
+      "learning_rate": 5.32926406578492e-06,
+      "loss": 0.3652,
+      "step": 5993
+    },
+    {
+      "epoch": 0.47952,
+      "grad_norm": 1.7158325910568237,
+      "learning_rate": 5.328009142112594e-06,
+      "loss": 0.3132,
+      "step": 5994
+    },
+    {
+      "epoch": 0.4796,
+      "grad_norm": 1.8352590799331665,
+      "learning_rate": 5.326754197688227e-06,
+      "loss": 0.3482,
+      "step": 5995
+    },
+    {
+      "epoch": 0.47968,
+      "grad_norm": 1.52081298828125,
+      "learning_rate": 5.3254992325912144e-06,
+      "loss": 0.309,
+      "step": 5996
+    },
+    {
+      "epoch": 0.47976,
+      "grad_norm": 1.6611655950546265,
+      "learning_rate": 5.324244246900955e-06,
+      "loss": 0.4123,
+      "step": 5997
+    },
+    {
+      "epoch": 0.47984,
+      "grad_norm": 1.4265227317810059,
+      "learning_rate": 5.322989240696846e-06,
+      "loss": 0.2749,
+      "step": 5998
+    },
+    {
+      "epoch": 0.47992,
+      "grad_norm": 1.5053348541259766,
+      "learning_rate": 5.3217342140582895e-06,
+      "loss": 0.268,
+      "step": 5999
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3345094919204712,
+      "learning_rate": 5.3204791670646875e-06,
+      "loss": 0.2847,
+      "step": 6000
+    },
+    {
+      "epoch": 0.48008,
+      "grad_norm": 1.3959380388259888,
+      "learning_rate": 5.319224099795438e-06,
+      "loss": 0.321,
+      "step": 6001
+    },
+    {
+      "epoch": 0.48016,
+      "grad_norm": 1.5935245752334595,
+      "learning_rate": 5.31796901232995e-06,
+      "loss": 0.3813,
+      "step": 6002
+    },
+    {
+      "epoch": 0.48024,
+      "grad_norm": 1.4641132354736328,
+      "learning_rate": 5.316713904747626e-06,
+      "loss": 0.3184,
+      "step": 6003
+    },
+    {
+      "epoch": 0.48032,
+      "grad_norm": 1.3873530626296997,
+      "learning_rate": 5.315458777127872e-06,
+      "loss": 0.2754,
+      "step": 6004
+    },
+    {
+      "epoch": 0.4804,
+      "grad_norm": 1.330491065979004,
+      "learning_rate": 5.3142036295500965e-06,
+      "loss": 0.2641,
+      "step": 6005
+    },
+    {
+      "epoch": 0.48048,
+      "grad_norm": 1.4904730319976807,
+      "learning_rate": 5.31294846209371e-06,
+      "loss": 0.3074,
+      "step": 6006
+    },
+    {
+      "epoch": 0.48056,
+      "grad_norm": 1.3027527332305908,
+      "learning_rate": 5.311693274838121e-06,
+      "loss": 0.3014,
+      "step": 6007
+    },
+    {
+      "epoch": 0.48064,
+      "grad_norm": 1.5734492540359497,
+      "learning_rate": 5.310438067862741e-06,
+      "loss": 0.2831,
+      "step": 6008
+    },
+    {
+      "epoch": 0.48072,
+      "grad_norm": 2.1448333263397217,
+      "learning_rate": 5.309182841246984e-06,
+      "loss": 0.4077,
+      "step": 6009
+    },
+    {
+      "epoch": 0.4808,
+      "grad_norm": 1.4222567081451416,
+      "learning_rate": 5.307927595070261e-06,
+      "loss": 0.3582,
+      "step": 6010
+    },
+    {
+      "epoch": 0.48088,
+      "grad_norm": 1.7736330032348633,
+      "learning_rate": 5.306672329411993e-06,
+      "loss": 0.4902,
+      "step": 6011
+    },
+    {
+      "epoch": 0.48096,
+      "grad_norm": 1.4863662719726562,
+      "learning_rate": 5.3054170443515895e-06,
+      "loss": 0.3672,
+      "step": 6012
+    },
+    {
+      "epoch": 0.48104,
+      "grad_norm": 1.2911968231201172,
+      "learning_rate": 5.304161739968474e-06,
+      "loss": 0.3233,
+      "step": 6013
+    },
+    {
+      "epoch": 0.48112,
+      "grad_norm": 1.451702356338501,
+      "learning_rate": 5.302906416342062e-06,
+      "loss": 0.2513,
+      "step": 6014
+    },
+    {
+      "epoch": 0.4812,
+      "grad_norm": 1.4997286796569824,
+      "learning_rate": 5.301651073551774e-06,
+      "loss": 0.2994,
+      "step": 6015
+    },
+    {
+      "epoch": 0.48128,
+      "grad_norm": 1.796286702156067,
+      "learning_rate": 5.300395711677032e-06,
+      "loss": 0.3123,
+      "step": 6016
+    },
+    {
+      "epoch": 0.48136,
+      "grad_norm": 1.705010175704956,
+      "learning_rate": 5.299140330797258e-06,
+      "loss": 0.4163,
+      "step": 6017
+    },
+    {
+      "epoch": 0.48144,
+      "grad_norm": 1.3299634456634521,
+      "learning_rate": 5.297884930991878e-06,
+      "loss": 0.2821,
+      "step": 6018
+    },
+    {
+      "epoch": 0.48152,
+      "grad_norm": 1.7674593925476074,
+      "learning_rate": 5.296629512340313e-06,
+      "loss": 0.3192,
+      "step": 6019
+    },
+    {
+      "epoch": 0.4816,
+      "grad_norm": 1.500632405281067,
+      "learning_rate": 5.295374074921993e-06,
+      "loss": 0.4209,
+      "step": 6020
+    },
+    {
+      "epoch": 0.48168,
+      "grad_norm": 1.5047935247421265,
+      "learning_rate": 5.294118618816342e-06,
+      "loss": 0.344,
+      "step": 6021
+    },
+    {
+      "epoch": 0.48176,
+      "grad_norm": 1.670544147491455,
+      "learning_rate": 5.292863144102791e-06,
+      "loss": 0.3467,
+      "step": 6022
+    },
+    {
+      "epoch": 0.48184,
+      "grad_norm": 1.3187230825424194,
+      "learning_rate": 5.291607650860769e-06,
+      "loss": 0.3187,
+      "step": 6023
+    },
+    {
+      "epoch": 0.48192,
+      "grad_norm": 1.355542778968811,
+      "learning_rate": 5.290352139169708e-06,
+      "loss": 0.2727,
+      "step": 6024
+    },
+    {
+      "epoch": 0.482,
+      "grad_norm": 1.7600772380828857,
+      "learning_rate": 5.289096609109037e-06,
+      "loss": 0.5256,
+      "step": 6025
+    },
+    {
+      "epoch": 0.48208,
+      "grad_norm": 1.4076069593429565,
+      "learning_rate": 5.287841060758191e-06,
+      "loss": 0.2691,
+      "step": 6026
+    },
+    {
+      "epoch": 0.48216,
+      "grad_norm": 1.5772290229797363,
+      "learning_rate": 5.2865854941966036e-06,
+      "loss": 0.3356,
+      "step": 6027
+    },
+    {
+      "epoch": 0.48224,
+      "grad_norm": 1.2514276504516602,
+      "learning_rate": 5.285329909503711e-06,
+      "loss": 0.2623,
+      "step": 6028
+    },
+    {
+      "epoch": 0.48232,
+      "grad_norm": 1.582513689994812,
+      "learning_rate": 5.284074306758951e-06,
+      "loss": 0.3788,
+      "step": 6029
+    },
+    {
+      "epoch": 0.4824,
+      "grad_norm": 1.5977832078933716,
+      "learning_rate": 5.2828186860417594e-06,
+      "loss": 0.3679,
+      "step": 6030
+    },
+    {
+      "epoch": 0.48248,
+      "grad_norm": 2.7796149253845215,
+      "learning_rate": 5.281563047431576e-06,
+      "loss": 3.9853,
+      "step": 6031
+    },
+    {
+      "epoch": 0.48256,
+      "grad_norm": 1.570940613746643,
+      "learning_rate": 5.2803073910078405e-06,
+      "loss": 0.3789,
+      "step": 6032
+    },
+    {
+      "epoch": 0.48264,
+      "grad_norm": 1.6044830083847046,
+      "learning_rate": 5.279051716849993e-06,
+      "loss": 0.3946,
+      "step": 6033
+    },
+    {
+      "epoch": 0.48272,
+      "grad_norm": 1.9162310361862183,
+      "learning_rate": 5.27779602503748e-06,
+      "loss": 0.3879,
+      "step": 6034
+    },
+    {
+      "epoch": 0.4828,
+      "grad_norm": 1.536901831626892,
+      "learning_rate": 5.27654031564974e-06,
+      "loss": 0.3266,
+      "step": 6035
+    },
+    {
+      "epoch": 0.48288,
+      "grad_norm": 1.6972392797470093,
+      "learning_rate": 5.275284588766221e-06,
+      "loss": 0.3758,
+      "step": 6036
+    },
+    {
+      "epoch": 0.48296,
+      "grad_norm": 1.5778000354766846,
+      "learning_rate": 5.274028844466366e-06,
+      "loss": 0.3237,
+      "step": 6037
+    },
+    {
+      "epoch": 0.48304,
+      "grad_norm": 1.4701905250549316,
+      "learning_rate": 5.272773082829623e-06,
+      "loss": 0.3285,
+      "step": 6038
+    },
+    {
+      "epoch": 0.48312,
+      "grad_norm": 1.5418689250946045,
+      "learning_rate": 5.2715173039354395e-06,
+      "loss": 0.3325,
+      "step": 6039
+    },
+    {
+      "epoch": 0.4832,
+      "grad_norm": 1.3240424394607544,
+      "learning_rate": 5.270261507863265e-06,
+      "loss": 0.2852,
+      "step": 6040
+    },
+    {
+      "epoch": 0.48328,
+      "grad_norm": 1.316301941871643,
+      "learning_rate": 5.26900569469255e-06,
+      "loss": 0.2481,
+      "step": 6041
+    },
+    {
+      "epoch": 0.48336,
+      "grad_norm": 1.2276591062545776,
+      "learning_rate": 5.267749864502744e-06,
+      "loss": 0.3035,
+      "step": 6042
+    },
+    {
+      "epoch": 0.48344,
+      "grad_norm": 1.5969600677490234,
+      "learning_rate": 5.266494017373299e-06,
+      "loss": 0.3646,
+      "step": 6043
+    },
+    {
+      "epoch": 0.48352,
+      "grad_norm": 1.2858132123947144,
+      "learning_rate": 5.265238153383669e-06,
+      "loss": 0.2527,
+      "step": 6044
+    },
+    {
+      "epoch": 0.4836,
+      "grad_norm": 1.9083600044250488,
+      "learning_rate": 5.26398227261331e-06,
+      "loss": 0.3858,
+      "step": 6045
+    },
+    {
+      "epoch": 0.48368,
+      "grad_norm": 2.1194543838500977,
+      "learning_rate": 5.2627263751416765e-06,
+      "loss": 0.4669,
+      "step": 6046
+    },
+    {
+      "epoch": 0.48376,
+      "grad_norm": 1.6560252904891968,
+      "learning_rate": 5.261470461048225e-06,
+      "loss": 0.3027,
+      "step": 6047
+    },
+    {
+      "epoch": 0.48384,
+      "grad_norm": 1.7517454624176025,
+      "learning_rate": 5.260214530412409e-06,
+      "loss": 0.4387,
+      "step": 6048
+    },
+    {
+      "epoch": 0.48392,
+      "grad_norm": 1.775425910949707,
+      "learning_rate": 5.258958583313692e-06,
+      "loss": 0.2734,
+      "step": 6049
+    },
+    {
+      "epoch": 0.484,
+      "grad_norm": 1.6235383749008179,
+      "learning_rate": 5.257702619831531e-06,
+      "loss": 0.2836,
+      "step": 6050
+    },
+    {
+      "epoch": 0.48408,
+      "grad_norm": 1.6095234155654907,
+      "learning_rate": 5.25644664004539e-06,
+      "loss": 0.3566,
+      "step": 6051
+    },
+    {
+      "epoch": 0.48416,
+      "grad_norm": 2.052558183670044,
+      "learning_rate": 5.2551906440347254e-06,
+      "loss": 0.5749,
+      "step": 6052
+    },
+    {
+      "epoch": 0.48424,
+      "grad_norm": 1.8320833444595337,
+      "learning_rate": 5.253934631879005e-06,
+      "loss": 0.3727,
+      "step": 6053
+    },
+    {
+      "epoch": 0.48432,
+      "grad_norm": 1.698598027229309,
+      "learning_rate": 5.252678603657689e-06,
+      "loss": 0.3987,
+      "step": 6054
+    },
+    {
+      "epoch": 0.4844,
+      "grad_norm": 2.1110999584198,
+      "learning_rate": 5.251422559450243e-06,
+      "loss": 0.4145,
+      "step": 6055
+    },
+    {
+      "epoch": 0.48448,
+      "grad_norm": 1.7855218648910522,
+      "learning_rate": 5.250166499336132e-06,
+      "loss": 0.3626,
+      "step": 6056
+    },
+    {
+      "epoch": 0.48456,
+      "grad_norm": 1.150992512702942,
+      "learning_rate": 5.248910423394827e-06,
+      "loss": 0.2392,
+      "step": 6057
+    },
+    {
+      "epoch": 0.48464,
+      "grad_norm": 1.548820972442627,
+      "learning_rate": 5.2476543317057896e-06,
+      "loss": 0.2712,
+      "step": 6058
+    },
+    {
+      "epoch": 0.48472,
+      "grad_norm": 1.7660967111587524,
+      "learning_rate": 5.246398224348492e-06,
+      "loss": 0.4438,
+      "step": 6059
+    },
+    {
+      "epoch": 0.4848,
+      "grad_norm": 1.4774236679077148,
+      "learning_rate": 5.245142101402403e-06,
+      "loss": 0.3322,
+      "step": 6060
+    },
+    {
+      "epoch": 0.48488,
+      "grad_norm": 1.747956395149231,
+      "learning_rate": 5.2438859629469926e-06,
+      "loss": 0.3714,
+      "step": 6061
+    },
+    {
+      "epoch": 0.48496,
+      "grad_norm": 1.422593116760254,
+      "learning_rate": 5.242629809061735e-06,
+      "loss": 0.3844,
+      "step": 6062
+    },
+    {
+      "epoch": 0.48504,
+      "grad_norm": 1.4882540702819824,
+      "learning_rate": 5.2413736398260995e-06,
+      "loss": 0.3285,
+      "step": 6063
+    },
+    {
+      "epoch": 0.48512,
+      "grad_norm": 1.318860650062561,
+      "learning_rate": 5.240117455319562e-06,
+      "loss": 0.2693,
+      "step": 6064
+    },
+    {
+      "epoch": 0.4852,
+      "grad_norm": 1.7626543045043945,
+      "learning_rate": 5.238861255621598e-06,
+      "loss": 0.3453,
+      "step": 6065
+    },
+    {
+      "epoch": 0.48528,
+      "grad_norm": 1.5399316549301147,
+      "learning_rate": 5.2376050408116805e-06,
+      "loss": 0.2871,
+      "step": 6066
+    },
+    {
+      "epoch": 0.48536,
+      "grad_norm": 1.6021755933761597,
+      "learning_rate": 5.236348810969287e-06,
+      "loss": 0.2933,
+      "step": 6067
+    },
+    {
+      "epoch": 0.48544,
+      "grad_norm": 1.620920181274414,
+      "learning_rate": 5.235092566173896e-06,
+      "loss": 0.3498,
+      "step": 6068
+    },
+    {
+      "epoch": 0.48552,
+      "grad_norm": 1.4818662405014038,
+      "learning_rate": 5.233836306504983e-06,
+      "loss": 0.2698,
+      "step": 6069
+    },
+    {
+      "epoch": 0.4856,
+      "grad_norm": 1.618439793586731,
+      "learning_rate": 5.232580032042032e-06,
+      "loss": 0.4304,
+      "step": 6070
+    },
+    {
+      "epoch": 0.48568,
+      "grad_norm": 1.531455636024475,
+      "learning_rate": 5.231323742864519e-06,
+      "loss": 0.3385,
+      "step": 6071
+    },
+    {
+      "epoch": 0.48576,
+      "grad_norm": 1.6968480348587036,
+      "learning_rate": 5.230067439051927e-06,
+      "loss": 0.33,
+      "step": 6072
+    },
+    {
+      "epoch": 0.48584,
+      "grad_norm": 1.7031455039978027,
+      "learning_rate": 5.228811120683738e-06,
+      "loss": 0.3422,
+      "step": 6073
+    },
+    {
+      "epoch": 0.48592,
+      "grad_norm": 1.5051372051239014,
+      "learning_rate": 5.2275547878394335e-06,
+      "loss": 0.2661,
+      "step": 6074
+    },
+    {
+      "epoch": 0.486,
+      "grad_norm": 1.9860918521881104,
+      "learning_rate": 5.2262984405985005e-06,
+      "loss": 0.3608,
+      "step": 6075
+    },
+    {
+      "epoch": 0.48608,
+      "grad_norm": 1.6986693143844604,
+      "learning_rate": 5.225042079040424e-06,
+      "loss": 0.335,
+      "step": 6076
+    },
+    {
+      "epoch": 0.48616,
+      "grad_norm": 1.692690372467041,
+      "learning_rate": 5.223785703244685e-06,
+      "loss": 0.4128,
+      "step": 6077
+    },
+    {
+      "epoch": 0.48624,
+      "grad_norm": 1.331215739250183,
+      "learning_rate": 5.222529313290774e-06,
+      "loss": 0.3168,
+      "step": 6078
+    },
+    {
+      "epoch": 0.48632,
+      "grad_norm": 1.5098655223846436,
+      "learning_rate": 5.221272909258178e-06,
+      "loss": 0.2776,
+      "step": 6079
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 1.5736565589904785,
+      "learning_rate": 5.220016491226387e-06,
+      "loss": 0.4146,
+      "step": 6080
+    },
+    {
+      "epoch": 0.48648,
+      "grad_norm": 1.3182066679000854,
+      "learning_rate": 5.2187600592748876e-06,
+      "loss": 0.2653,
+      "step": 6081
+    },
+    {
+      "epoch": 0.48656,
+      "grad_norm": 1.3879201412200928,
+      "learning_rate": 5.217503613483172e-06,
+      "loss": 0.3127,
+      "step": 6082
+    },
+    {
+      "epoch": 0.48664,
+      "grad_norm": 1.3078670501708984,
+      "learning_rate": 5.21624715393073e-06,
+      "loss": 0.2701,
+      "step": 6083
+    },
+    {
+      "epoch": 0.48672,
+      "grad_norm": 1.890555739402771,
+      "learning_rate": 5.214990680697054e-06,
+      "loss": 0.5105,
+      "step": 6084
+    },
+    {
+      "epoch": 0.4868,
+      "grad_norm": 1.693420171737671,
+      "learning_rate": 5.213734193861637e-06,
+      "loss": 0.3455,
+      "step": 6085
+    },
+    {
+      "epoch": 0.48688,
+      "grad_norm": 1.6514363288879395,
+      "learning_rate": 5.212477693503973e-06,
+      "loss": 0.4083,
+      "step": 6086
+    },
+    {
+      "epoch": 0.48696,
+      "grad_norm": 1.5142364501953125,
+      "learning_rate": 5.211221179703555e-06,
+      "loss": 0.3556,
+      "step": 6087
+    },
+    {
+      "epoch": 0.48704,
+      "grad_norm": 1.2250992059707642,
+      "learning_rate": 5.209964652539882e-06,
+      "loss": 0.2917,
+      "step": 6088
+    },
+    {
+      "epoch": 0.48712,
+      "grad_norm": 1.4357571601867676,
+      "learning_rate": 5.2087081120924464e-06,
+      "loss": 0.2784,
+      "step": 6089
+    },
+    {
+      "epoch": 0.4872,
+      "grad_norm": 1.5746808052062988,
+      "learning_rate": 5.207451558440747e-06,
+      "loss": 0.3645,
+      "step": 6090
+    },
+    {
+      "epoch": 0.48728,
+      "grad_norm": 1.1087597608566284,
+      "learning_rate": 5.206194991664283e-06,
+      "loss": 0.2236,
+      "step": 6091
+    },
+    {
+      "epoch": 0.48736,
+      "grad_norm": 1.6777054071426392,
+      "learning_rate": 5.204938411842551e-06,
+      "loss": 0.4101,
+      "step": 6092
+    },
+    {
+      "epoch": 0.48744,
+      "grad_norm": 2.015167474746704,
+      "learning_rate": 5.2036818190550496e-06,
+      "loss": 0.4252,
+      "step": 6093
+    },
+    {
+      "epoch": 0.48752,
+      "grad_norm": 1.6957621574401855,
+      "learning_rate": 5.202425213381284e-06,
+      "loss": 0.4532,
+      "step": 6094
+    },
+    {
+      "epoch": 0.4876,
+      "grad_norm": 1.689834475517273,
+      "learning_rate": 5.2011685949007506e-06,
+      "loss": 0.322,
+      "step": 6095
+    },
+    {
+      "epoch": 0.48768,
+      "grad_norm": 1.2407318353652954,
+      "learning_rate": 5.199911963692953e-06,
+      "loss": 0.2784,
+      "step": 6096
+    },
+    {
+      "epoch": 0.48776,
+      "grad_norm": 0.9591389298439026,
+      "learning_rate": 5.198655319837395e-06,
+      "loss": 0.232,
+      "step": 6097
+    },
+    {
+      "epoch": 0.48784,
+      "grad_norm": 1.498170018196106,
+      "learning_rate": 5.197398663413579e-06,
+      "loss": 0.2981,
+      "step": 6098
+    },
+    {
+      "epoch": 0.48792,
+      "grad_norm": 1.4004147052764893,
+      "learning_rate": 5.1961419945010104e-06,
+      "loss": 0.3348,
+      "step": 6099
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 1.3157734870910645,
+      "learning_rate": 5.194885313179195e-06,
+      "loss": 0.3471,
+      "step": 6100
+    },
+    {
+      "epoch": 0.48808,
+      "grad_norm": 1.5406330823898315,
+      "learning_rate": 5.1936286195276374e-06,
+      "loss": 0.334,
+      "step": 6101
+    },
+    {
+      "epoch": 0.48816,
+      "grad_norm": 1.4737725257873535,
+      "learning_rate": 5.192371913625845e-06,
+      "loss": 0.3255,
+      "step": 6102
+    },
+    {
+      "epoch": 0.48824,
+      "grad_norm": 1.704572319984436,
+      "learning_rate": 5.1911151955533254e-06,
+      "loss": 0.3709,
+      "step": 6103
+    },
+    {
+      "epoch": 0.48832,
+      "grad_norm": 1.5798331499099731,
+      "learning_rate": 5.1898584653895865e-06,
+      "loss": 0.2797,
+      "step": 6104
+    },
+    {
+      "epoch": 0.4884,
+      "grad_norm": 1.6223543882369995,
+      "learning_rate": 5.188601723214139e-06,
+      "loss": 0.3452,
+      "step": 6105
+    },
+    {
+      "epoch": 0.48848,
+      "grad_norm": 1.5775635242462158,
+      "learning_rate": 5.18734496910649e-06,
+      "loss": 0.3373,
+      "step": 6106
+    },
+    {
+      "epoch": 0.48856,
+      "grad_norm": 1.9005051851272583,
+      "learning_rate": 5.186088203146152e-06,
+      "loss": 0.4212,
+      "step": 6107
+    },
+    {
+      "epoch": 0.48864,
+      "grad_norm": 1.3695884943008423,
+      "learning_rate": 5.184831425412636e-06,
+      "loss": 0.3052,
+      "step": 6108
+    },
+    {
+      "epoch": 0.48872,
+      "grad_norm": 1.3192788362503052,
+      "learning_rate": 5.1835746359854544e-06,
+      "loss": 0.2665,
+      "step": 6109
+    },
+    {
+      "epoch": 0.4888,
+      "grad_norm": 1.645479679107666,
+      "learning_rate": 5.1823178349441195e-06,
+      "loss": 0.3848,
+      "step": 6110
+    },
+    {
+      "epoch": 0.48888,
+      "grad_norm": 1.4701766967773438,
+      "learning_rate": 5.181061022368145e-06,
+      "loss": 0.2934,
+      "step": 6111
+    },
+    {
+      "epoch": 0.48896,
+      "grad_norm": 1.2284507751464844,
+      "learning_rate": 5.179804198337046e-06,
+      "loss": 0.2826,
+      "step": 6112
+    },
+    {
+      "epoch": 0.48904,
+      "grad_norm": 1.8432642221450806,
+      "learning_rate": 5.178547362930337e-06,
+      "loss": 0.354,
+      "step": 6113
+    },
+    {
+      "epoch": 0.48912,
+      "grad_norm": 1.703628420829773,
+      "learning_rate": 5.1772905162275345e-06,
+      "loss": 0.3431,
+      "step": 6114
+    },
+    {
+      "epoch": 0.4892,
+      "grad_norm": 1.5386126041412354,
+      "learning_rate": 5.176033658308154e-06,
+      "loss": 0.3512,
+      "step": 6115
+    },
+    {
+      "epoch": 0.48928,
+      "grad_norm": 1.7287826538085938,
+      "learning_rate": 5.174776789251712e-06,
+      "loss": 0.3485,
+      "step": 6116
+    },
+    {
+      "epoch": 0.48936,
+      "grad_norm": 1.4497462511062622,
+      "learning_rate": 5.173519909137728e-06,
+      "loss": 0.3208,
+      "step": 6117
+    },
+    {
+      "epoch": 0.48944,
+      "grad_norm": 1.2991448640823364,
+      "learning_rate": 5.17226301804572e-06,
+      "loss": 0.2982,
+      "step": 6118
+    },
+    {
+      "epoch": 0.48952,
+      "grad_norm": 1.4704984426498413,
+      "learning_rate": 5.171006116055206e-06,
+      "loss": 0.3264,
+      "step": 6119
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 1.2503539323806763,
+      "learning_rate": 5.169749203245709e-06,
+      "loss": 0.2779,
+      "step": 6120
+    },
+    {
+      "epoch": 0.48968,
+      "grad_norm": 1.3474243879318237,
+      "learning_rate": 5.168492279696747e-06,
+      "loss": 0.3001,
+      "step": 6121
+    },
+    {
+      "epoch": 0.48976,
+      "grad_norm": 1.846341609954834,
+      "learning_rate": 5.167235345487841e-06,
+      "loss": 0.3778,
+      "step": 6122
+    },
+    {
+      "epoch": 0.48984,
+      "grad_norm": 1.60556960105896,
+      "learning_rate": 5.165978400698516e-06,
+      "loss": 0.3083,
+      "step": 6123
+    },
+    {
+      "epoch": 0.48992,
+      "grad_norm": 1.5098539590835571,
+      "learning_rate": 5.164721445408292e-06,
+      "loss": 0.2916,
+      "step": 6124
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.64169180393219,
+      "learning_rate": 5.163464479696694e-06,
+      "loss": 0.3603,
+      "step": 6125
+    },
+    {
+      "epoch": 0.49008,
+      "grad_norm": 1.6956416368484497,
+      "learning_rate": 5.162207503643246e-06,
+      "loss": 0.3837,
+      "step": 6126
+    },
+    {
+      "epoch": 0.49016,
+      "grad_norm": 1.6618326902389526,
+      "learning_rate": 5.160950517327471e-06,
+      "loss": 0.3823,
+      "step": 6127
+    },
+    {
+      "epoch": 0.49024,
+      "grad_norm": 1.7037644386291504,
+      "learning_rate": 5.159693520828895e-06,
+      "loss": 0.3085,
+      "step": 6128
+    },
+    {
+      "epoch": 0.49032,
+      "grad_norm": 1.144739031791687,
+      "learning_rate": 5.158436514227045e-06,
+      "loss": 0.2081,
+      "step": 6129
+    },
+    {
+      "epoch": 0.4904,
+      "grad_norm": 1.2798271179199219,
+      "learning_rate": 5.157179497601447e-06,
+      "loss": 0.3009,
+      "step": 6130
+    },
+    {
+      "epoch": 0.49048,
+      "grad_norm": 1.7405157089233398,
+      "learning_rate": 5.155922471031627e-06,
+      "loss": 0.3543,
+      "step": 6131
+    },
+    {
+      "epoch": 0.49056,
+      "grad_norm": 1.0767748355865479,
+      "learning_rate": 5.154665434597115e-06,
+      "loss": 0.2166,
+      "step": 6132
+    },
+    {
+      "epoch": 0.49064,
+      "grad_norm": 1.536720633506775,
+      "learning_rate": 5.153408388377438e-06,
+      "loss": 0.2829,
+      "step": 6133
+    },
+    {
+      "epoch": 0.49072,
+      "grad_norm": 1.578690767288208,
+      "learning_rate": 5.152151332452125e-06,
+      "loss": 0.3479,
+      "step": 6134
+    },
+    {
+      "epoch": 0.4908,
+      "grad_norm": 1.888026475906372,
+      "learning_rate": 5.150894266900708e-06,
+      "loss": 0.3414,
+      "step": 6135
+    },
+    {
+      "epoch": 0.49088,
+      "grad_norm": 1.6138415336608887,
+      "learning_rate": 5.149637191802714e-06,
+      "loss": 0.2913,
+      "step": 6136
+    },
+    {
+      "epoch": 0.49096,
+      "grad_norm": 1.5494178533554077,
+      "learning_rate": 5.148380107237677e-06,
+      "loss": 0.3035,
+      "step": 6137
+    },
+    {
+      "epoch": 0.49104,
+      "grad_norm": 1.4410408735275269,
+      "learning_rate": 5.1471230132851254e-06,
+      "loss": 0.3202,
+      "step": 6138
+    },
+    {
+      "epoch": 0.49112,
+      "grad_norm": 1.510621428489685,
+      "learning_rate": 5.145865910024595e-06,
+      "loss": 0.2863,
+      "step": 6139
+    },
+    {
+      "epoch": 0.4912,
+      "grad_norm": 1.8136560916900635,
+      "learning_rate": 5.144608797535614e-06,
+      "loss": 0.3635,
+      "step": 6140
+    },
+    {
+      "epoch": 0.49128,
+      "grad_norm": 1.6779876947402954,
+      "learning_rate": 5.143351675897721e-06,
+      "loss": 0.4096,
+      "step": 6141
+    },
+    {
+      "epoch": 0.49136,
+      "grad_norm": 1.5982199907302856,
+      "learning_rate": 5.1420945451904455e-06,
+      "loss": 0.2868,
+      "step": 6142
+    },
+    {
+      "epoch": 0.49144,
+      "grad_norm": 1.6804651021957397,
+      "learning_rate": 5.140837405493324e-06,
+      "loss": 0.3119,
+      "step": 6143
+    },
+    {
+      "epoch": 0.49152,
+      "grad_norm": 1.7489296197891235,
+      "learning_rate": 5.13958025688589e-06,
+      "loss": 0.302,
+      "step": 6144
+    },
+    {
+      "epoch": 0.4916,
+      "grad_norm": 1.693847417831421,
+      "learning_rate": 5.138323099447681e-06,
+      "loss": 0.3248,
+      "step": 6145
+    },
+    {
+      "epoch": 0.49168,
+      "grad_norm": 1.360862374305725,
+      "learning_rate": 5.137065933258233e-06,
+      "loss": 0.2943,
+      "step": 6146
+    },
+    {
+      "epoch": 0.49176,
+      "grad_norm": 2.244718313217163,
+      "learning_rate": 5.135808758397082e-06,
+      "loss": 0.4029,
+      "step": 6147
+    },
+    {
+      "epoch": 0.49184,
+      "grad_norm": 1.749489426612854,
+      "learning_rate": 5.134551574943765e-06,
+      "loss": 0.3362,
+      "step": 6148
+    },
+    {
+      "epoch": 0.49192,
+      "grad_norm": 1.4736480712890625,
+      "learning_rate": 5.1332943829778205e-06,
+      "loss": 0.2781,
+      "step": 6149
+    },
+    {
+      "epoch": 0.492,
+      "grad_norm": 1.5586931705474854,
+      "learning_rate": 5.132037182578785e-06,
+      "loss": 0.2709,
+      "step": 6150
+    },
+    {
+      "epoch": 0.49208,
+      "grad_norm": 1.2036563158035278,
+      "learning_rate": 5.130779973826199e-06,
+      "loss": 0.2957,
+      "step": 6151
+    },
+    {
+      "epoch": 0.49216,
+      "grad_norm": 2.098465919494629,
+      "learning_rate": 5.129522756799602e-06,
+      "loss": 0.4698,
+      "step": 6152
+    },
+    {
+      "epoch": 0.49224,
+      "grad_norm": 1.3323612213134766,
+      "learning_rate": 5.128265531578535e-06,
+      "loss": 0.3015,
+      "step": 6153
+    },
+    {
+      "epoch": 0.49232,
+      "grad_norm": 1.4440571069717407,
+      "learning_rate": 5.127008298242535e-06,
+      "loss": 0.3037,
+      "step": 6154
+    },
+    {
+      "epoch": 0.4924,
+      "grad_norm": 1.412387728691101,
+      "learning_rate": 5.125751056871146e-06,
+      "loss": 0.3701,
+      "step": 6155
+    },
+    {
+      "epoch": 0.49248,
+      "grad_norm": 1.841917634010315,
+      "learning_rate": 5.124493807543908e-06,
+      "loss": 0.3041,
+      "step": 6156
+    },
+    {
+      "epoch": 0.49256,
+      "grad_norm": 1.612427830696106,
+      "learning_rate": 5.1232365503403626e-06,
+      "loss": 0.3198,
+      "step": 6157
+    },
+    {
+      "epoch": 0.49264,
+      "grad_norm": 1.3332172632217407,
+      "learning_rate": 5.1219792853400545e-06,
+      "loss": 0.3406,
+      "step": 6158
+    },
+    {
+      "epoch": 0.49272,
+      "grad_norm": 1.876379370689392,
+      "learning_rate": 5.120722012622524e-06,
+      "loss": 0.4073,
+      "step": 6159
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 1.5823307037353516,
+      "learning_rate": 5.119464732267317e-06,
+      "loss": 0.3825,
+      "step": 6160
+    },
+    {
+      "epoch": 0.49288,
+      "grad_norm": 1.216632604598999,
+      "learning_rate": 5.118207444353975e-06,
+      "loss": 0.3464,
+      "step": 6161
+    },
+    {
+      "epoch": 0.49296,
+      "grad_norm": 2.3204121589660645,
+      "learning_rate": 5.116950148962043e-06,
+      "loss": 0.3724,
+      "step": 6162
+    },
+    {
+      "epoch": 0.49304,
+      "grad_norm": 1.6256400346755981,
+      "learning_rate": 5.115692846171067e-06,
+      "loss": 0.3355,
+      "step": 6163
+    },
+    {
+      "epoch": 0.49312,
+      "grad_norm": 1.4591851234436035,
+      "learning_rate": 5.114435536060593e-06,
+      "loss": 0.309,
+      "step": 6164
+    },
+    {
+      "epoch": 0.4932,
+      "grad_norm": 2.0748255252838135,
+      "learning_rate": 5.113178218710164e-06,
+      "loss": 0.3612,
+      "step": 6165
+    },
+    {
+      "epoch": 0.49328,
+      "grad_norm": 1.8025262355804443,
+      "learning_rate": 5.11192089419933e-06,
+      "loss": 0.4716,
+      "step": 6166
+    },
+    {
+      "epoch": 0.49336,
+      "grad_norm": 1.4925312995910645,
+      "learning_rate": 5.110663562607632e-06,
+      "loss": 0.3646,
+      "step": 6167
+    },
+    {
+      "epoch": 0.49344,
+      "grad_norm": 1.710847020149231,
+      "learning_rate": 5.109406224014623e-06,
+      "loss": 0.353,
+      "step": 6168
+    },
+    {
+      "epoch": 0.49352,
+      "grad_norm": 1.3324249982833862,
+      "learning_rate": 5.108148878499847e-06,
+      "loss": 0.2612,
+      "step": 6169
+    },
+    {
+      "epoch": 0.4936,
+      "grad_norm": 1.7299015522003174,
+      "learning_rate": 5.106891526142854e-06,
+      "loss": 0.3321,
+      "step": 6170
+    },
+    {
+      "epoch": 0.49368,
+      "grad_norm": 1.7776432037353516,
+      "learning_rate": 5.105634167023193e-06,
+      "loss": 0.3346,
+      "step": 6171
+    },
+    {
+      "epoch": 0.49376,
+      "grad_norm": 1.4523450136184692,
+      "learning_rate": 5.104376801220411e-06,
+      "loss": 0.2788,
+      "step": 6172
+    },
+    {
+      "epoch": 0.49384,
+      "grad_norm": 1.5704206228256226,
+      "learning_rate": 5.103119428814057e-06,
+      "loss": 0.3746,
+      "step": 6173
+    },
+    {
+      "epoch": 0.49392,
+      "grad_norm": 1.2726726531982422,
+      "learning_rate": 5.1018620498836825e-06,
+      "loss": 0.2869,
+      "step": 6174
+    },
+    {
+      "epoch": 0.494,
+      "grad_norm": 1.7216542959213257,
+      "learning_rate": 5.1006046645088355e-06,
+      "loss": 0.4106,
+      "step": 6175
+    },
+    {
+      "epoch": 0.49408,
+      "grad_norm": 1.4835233688354492,
+      "learning_rate": 5.09934727276907e-06,
+      "loss": 0.3307,
+      "step": 6176
+    },
+    {
+      "epoch": 0.49416,
+      "grad_norm": 1.483788251876831,
+      "learning_rate": 5.0980898747439345e-06,
+      "loss": 0.3311,
+      "step": 6177
+    },
+    {
+      "epoch": 0.49424,
+      "grad_norm": 1.5280448198318481,
+      "learning_rate": 5.096832470512981e-06,
+      "loss": 0.3275,
+      "step": 6178
+    },
+    {
+      "epoch": 0.49432,
+      "grad_norm": 1.6087018251419067,
+      "learning_rate": 5.095575060155761e-06,
+      "loss": 0.3893,
+      "step": 6179
+    },
+    {
+      "epoch": 0.4944,
+      "grad_norm": 1.2583895921707153,
+      "learning_rate": 5.094317643751825e-06,
+      "loss": 0.269,
+      "step": 6180
+    },
+    {
+      "epoch": 0.49448,
+      "grad_norm": 2.5767970085144043,
+      "learning_rate": 5.09306022138073e-06,
+      "loss": 0.3843,
+      "step": 6181
+    },
+    {
+      "epoch": 0.49456,
+      "grad_norm": 1.3988746404647827,
+      "learning_rate": 5.091802793122025e-06,
+      "loss": 0.2755,
+      "step": 6182
+    },
+    {
+      "epoch": 0.49464,
+      "grad_norm": 1.3515316247940063,
+      "learning_rate": 5.090545359055264e-06,
+      "loss": 0.2335,
+      "step": 6183
+    },
+    {
+      "epoch": 0.49472,
+      "grad_norm": 1.4173052310943604,
+      "learning_rate": 5.089287919260002e-06,
+      "loss": 0.3242,
+      "step": 6184
+    },
+    {
+      "epoch": 0.4948,
+      "grad_norm": 1.7955999374389648,
+      "learning_rate": 5.088030473815791e-06,
+      "loss": 0.4096,
+      "step": 6185
+    },
+    {
+      "epoch": 0.49488,
+      "grad_norm": 1.374130129814148,
+      "learning_rate": 5.0867730228021875e-06,
+      "loss": 0.2919,
+      "step": 6186
+    },
+    {
+      "epoch": 0.49496,
+      "grad_norm": 1.3074638843536377,
+      "learning_rate": 5.085515566298744e-06,
+      "loss": 0.2317,
+      "step": 6187
+    },
+    {
+      "epoch": 0.49504,
+      "grad_norm": 1.6454929113388062,
+      "learning_rate": 5.084258104385018e-06,
+      "loss": 0.286,
+      "step": 6188
+    },
+    {
+      "epoch": 0.49512,
+      "grad_norm": 1.3493096828460693,
+      "learning_rate": 5.0830006371405625e-06,
+      "loss": 0.2989,
+      "step": 6189
+    },
+    {
+      "epoch": 0.4952,
+      "grad_norm": 1.6204330921173096,
+      "learning_rate": 5.081743164644935e-06,
+      "loss": 0.2795,
+      "step": 6190
+    },
+    {
+      "epoch": 0.49528,
+      "grad_norm": 1.340806245803833,
+      "learning_rate": 5.080485686977691e-06,
+      "loss": 0.2935,
+      "step": 6191
+    },
+    {
+      "epoch": 0.49536,
+      "grad_norm": 1.6534948348999023,
+      "learning_rate": 5.079228204218387e-06,
+      "loss": 0.3131,
+      "step": 6192
+    },
+    {
+      "epoch": 0.49544,
+      "grad_norm": 1.6619195938110352,
+      "learning_rate": 5.07797071644658e-06,
+      "loss": 0.4199,
+      "step": 6193
+    },
+    {
+      "epoch": 0.49552,
+      "grad_norm": 1.7682374715805054,
+      "learning_rate": 5.0767132237418275e-06,
+      "loss": 0.3768,
+      "step": 6194
+    },
+    {
+      "epoch": 0.4956,
+      "grad_norm": 1.9795269966125488,
+      "learning_rate": 5.075455726183685e-06,
+      "loss": 0.5178,
+      "step": 6195
+    },
+    {
+      "epoch": 0.49568,
+      "grad_norm": 1.5913708209991455,
+      "learning_rate": 5.07419822385171e-06,
+      "loss": 0.2969,
+      "step": 6196
+    },
+    {
+      "epoch": 0.49576,
+      "grad_norm": 1.3208245038986206,
+      "learning_rate": 5.072940716825462e-06,
+      "loss": 0.2773,
+      "step": 6197
+    },
+    {
+      "epoch": 0.49584,
+      "grad_norm": 1.503617286682129,
+      "learning_rate": 5.071683205184499e-06,
+      "loss": 0.2821,
+      "step": 6198
+    },
+    {
+      "epoch": 0.49592,
+      "grad_norm": 1.469290852546692,
+      "learning_rate": 5.070425689008381e-06,
+      "loss": 0.289,
+      "step": 6199
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 1.5494263172149658,
+      "learning_rate": 5.069168168376664e-06,
+      "loss": 0.3019,
+      "step": 6200
+    },
+    {
+      "epoch": 0.49608,
+      "grad_norm": 1.5420727729797363,
+      "learning_rate": 5.067910643368908e-06,
+      "loss": 0.2645,
+      "step": 6201
+    },
+    {
+      "epoch": 0.49616,
+      "grad_norm": 1.399019718170166,
+      "learning_rate": 5.066653114064674e-06,
+      "loss": 0.3092,
+      "step": 6202
+    },
+    {
+      "epoch": 0.49624,
+      "grad_norm": 1.9305436611175537,
+      "learning_rate": 5.065395580543519e-06,
+      "loss": 0.3418,
+      "step": 6203
+    },
+    {
+      "epoch": 0.49632,
+      "grad_norm": 1.197814702987671,
+      "learning_rate": 5.064138042885006e-06,
+      "loss": 0.2597,
+      "step": 6204
+    },
+    {
+      "epoch": 0.4964,
+      "grad_norm": 1.5163662433624268,
+      "learning_rate": 5.062880501168693e-06,
+      "loss": 0.3871,
+      "step": 6205
+    },
+    {
+      "epoch": 0.49648,
+      "grad_norm": 1.3596129417419434,
+      "learning_rate": 5.061622955474142e-06,
+      "loss": 0.3607,
+      "step": 6206
+    },
+    {
+      "epoch": 0.49656,
+      "grad_norm": 2.0122718811035156,
+      "learning_rate": 5.060365405880911e-06,
+      "loss": 0.3757,
+      "step": 6207
+    },
+    {
+      "epoch": 0.49664,
+      "grad_norm": 1.1948397159576416,
+      "learning_rate": 5.059107852468565e-06,
+      "loss": 0.256,
+      "step": 6208
+    },
+    {
+      "epoch": 0.49672,
+      "grad_norm": 1.5326647758483887,
+      "learning_rate": 5.057850295316661e-06,
+      "loss": 0.4097,
+      "step": 6209
+    },
+    {
+      "epoch": 0.4968,
+      "grad_norm": 1.3815491199493408,
+      "learning_rate": 5.0565927345047614e-06,
+      "loss": 0.3258,
+      "step": 6210
+    },
+    {
+      "epoch": 0.49688,
+      "grad_norm": 1.5262930393218994,
+      "learning_rate": 5.055335170112432e-06,
+      "loss": 0.3474,
+      "step": 6211
+    },
+    {
+      "epoch": 0.49696,
+      "grad_norm": 1.2492114305496216,
+      "learning_rate": 5.054077602219229e-06,
+      "loss": 0.2956,
+      "step": 6212
+    },
+    {
+      "epoch": 0.49704,
+      "grad_norm": 1.3622100353240967,
+      "learning_rate": 5.052820030904719e-06,
+      "loss": 0.3019,
+      "step": 6213
+    },
+    {
+      "epoch": 0.49712,
+      "grad_norm": 1.563354730606079,
+      "learning_rate": 5.051562456248461e-06,
+      "loss": 0.4146,
+      "step": 6214
+    },
+    {
+      "epoch": 0.4972,
+      "grad_norm": 1.7303005456924438,
+      "learning_rate": 5.05030487833002e-06,
+      "loss": 0.3324,
+      "step": 6215
+    },
+    {
+      "epoch": 0.49728,
+      "grad_norm": 1.239866852760315,
+      "learning_rate": 5.049047297228956e-06,
+      "loss": 0.2878,
+      "step": 6216
+    },
+    {
+      "epoch": 0.49736,
+      "grad_norm": 1.492857813835144,
+      "learning_rate": 5.047789713024836e-06,
+      "loss": 0.3036,
+      "step": 6217
+    },
+    {
+      "epoch": 0.49744,
+      "grad_norm": 1.4982048273086548,
+      "learning_rate": 5.046532125797219e-06,
+      "loss": 0.3185,
+      "step": 6218
+    },
+    {
+      "epoch": 0.49752,
+      "grad_norm": 1.7222771644592285,
+      "learning_rate": 5.0452745356256705e-06,
+      "loss": 0.4005,
+      "step": 6219
+    },
+    {
+      "epoch": 0.4976,
+      "grad_norm": 1.1927220821380615,
+      "learning_rate": 5.044016942589754e-06,
+      "loss": 0.2655,
+      "step": 6220
+    },
+    {
+      "epoch": 0.49768,
+      "grad_norm": 1.6021721363067627,
+      "learning_rate": 5.042759346769031e-06,
+      "loss": 0.2886,
+      "step": 6221
+    },
+    {
+      "epoch": 0.49776,
+      "grad_norm": 2.103205442428589,
+      "learning_rate": 5.041501748243069e-06,
+      "loss": 0.5297,
+      "step": 6222
+    },
+    {
+      "epoch": 0.49784,
+      "grad_norm": 1.3879965543746948,
+      "learning_rate": 5.040244147091431e-06,
+      "loss": 0.3109,
+      "step": 6223
+    },
+    {
+      "epoch": 0.49792,
+      "grad_norm": 2.045581817626953,
+      "learning_rate": 5.0389865433936776e-06,
+      "loss": 0.3605,
+      "step": 6224
+    },
+    {
+      "epoch": 0.498,
+      "grad_norm": 1.3637545108795166,
+      "learning_rate": 5.037728937229378e-06,
+      "loss": 0.2776,
+      "step": 6225
+    },
+    {
+      "epoch": 0.49808,
+      "grad_norm": 1.311774492263794,
+      "learning_rate": 5.036471328678095e-06,
+      "loss": 0.2672,
+      "step": 6226
+    },
+    {
+      "epoch": 0.49816,
+      "grad_norm": 1.8853634595870972,
+      "learning_rate": 5.035213717819393e-06,
+      "loss": 0.3784,
+      "step": 6227
+    },
+    {
+      "epoch": 0.49824,
+      "grad_norm": 1.301746129989624,
+      "learning_rate": 5.033956104732836e-06,
+      "loss": 0.2905,
+      "step": 6228
+    },
+    {
+      "epoch": 0.49832,
+      "grad_norm": 1.6232069730758667,
+      "learning_rate": 5.032698489497991e-06,
+      "loss": 0.3362,
+      "step": 6229
+    },
+    {
+      "epoch": 0.4984,
+      "grad_norm": 1.5022000074386597,
+      "learning_rate": 5.031440872194422e-06,
+      "loss": 0.2796,
+      "step": 6230
+    },
+    {
+      "epoch": 0.49848,
+      "grad_norm": 1.7945005893707275,
+      "learning_rate": 5.030183252901694e-06,
+      "loss": 0.3596,
+      "step": 6231
+    },
+    {
+      "epoch": 0.49856,
+      "grad_norm": 1.5525250434875488,
+      "learning_rate": 5.02892563169937e-06,
+      "loss": 0.343,
+      "step": 6232
+    },
+    {
+      "epoch": 0.49864,
+      "grad_norm": 1.631269931793213,
+      "learning_rate": 5.027668008667022e-06,
+      "loss": 0.3799,
+      "step": 6233
+    },
+    {
+      "epoch": 0.49872,
+      "grad_norm": 1.4517936706542969,
+      "learning_rate": 5.026410383884209e-06,
+      "loss": 0.338,
+      "step": 6234
+    },
+    {
+      "epoch": 0.4988,
+      "grad_norm": 1.387960433959961,
+      "learning_rate": 5.025152757430501e-06,
+      "loss": 0.3515,
+      "step": 6235
+    },
+    {
+      "epoch": 0.49888,
+      "grad_norm": 1.4274933338165283,
+      "learning_rate": 5.023895129385461e-06,
+      "loss": 0.3152,
+      "step": 6236
+    },
+    {
+      "epoch": 0.49896,
+      "grad_norm": 1.2925071716308594,
+      "learning_rate": 5.022637499828656e-06,
+      "loss": 0.3087,
+      "step": 6237
+    },
+    {
+      "epoch": 0.49904,
+      "grad_norm": 1.6201573610305786,
+      "learning_rate": 5.021379868839655e-06,
+      "loss": 0.3988,
+      "step": 6238
+    },
+    {
+      "epoch": 0.49912,
+      "grad_norm": 1.1447267532348633,
+      "learning_rate": 5.020122236498018e-06,
+      "loss": 0.281,
+      "step": 6239
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 1.618818998336792,
+      "learning_rate": 5.018864602883315e-06,
+      "loss": 0.34,
+      "step": 6240
+    },
+    {
+      "epoch": 0.49928,
+      "grad_norm": 1.4714045524597168,
+      "learning_rate": 5.017606968075113e-06,
+      "loss": 0.3352,
+      "step": 6241
+    },
+    {
+      "epoch": 0.49936,
+      "grad_norm": 1.8967351913452148,
+      "learning_rate": 5.016349332152975e-06,
+      "loss": 0.4835,
+      "step": 6242
+    },
+    {
+      "epoch": 0.49944,
+      "grad_norm": 1.4806861877441406,
+      "learning_rate": 5.0150916951964715e-06,
+      "loss": 0.3297,
+      "step": 6243
+    },
+    {
+      "epoch": 0.49952,
+      "grad_norm": 1.4988765716552734,
+      "learning_rate": 5.013834057285165e-06,
+      "loss": 0.3896,
+      "step": 6244
+    },
+    {
+      "epoch": 0.4996,
+      "grad_norm": 1.3121827840805054,
+      "learning_rate": 5.012576418498626e-06,
+      "loss": 0.3397,
+      "step": 6245
+    },
+    {
+      "epoch": 0.49968,
+      "grad_norm": 1.7312182188034058,
+      "learning_rate": 5.0113187789164176e-06,
+      "loss": 0.3462,
+      "step": 6246
+    },
+    {
+      "epoch": 0.49976,
+      "grad_norm": 1.460113763809204,
+      "learning_rate": 5.010061138618109e-06,
+      "loss": 0.2435,
+      "step": 6247
+    },
+    {
+      "epoch": 0.49984,
+      "grad_norm": 1.681500792503357,
+      "learning_rate": 5.008803497683266e-06,
+      "loss": 0.452,
+      "step": 6248
+    },
+    {
+      "epoch": 0.49992,
+      "grad_norm": 1.528382658958435,
+      "learning_rate": 5.007545856191453e-06,
+      "loss": 0.3297,
+      "step": 6249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7061470746994019,
+      "learning_rate": 5.006288214222242e-06,
+      "loss": 0.4542,
+      "step": 6250
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 12500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1250,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.2449718975467684e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}