amburger66's picture
LoRA fine-tune on RobotSmith task08 cutter
1cf6661 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.04,
"eval_steps": 50,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 4e-05,
"grad_norm": 14.543946266174316,
"learning_rate": 0.0,
"loss": 1.3987,
"step": 1
},
{
"epoch": 8e-05,
"grad_norm": 3.023715019226074,
"learning_rate": 2.0000000000000002e-07,
"loss": 0.2709,
"step": 2
},
{
"epoch": 0.00012,
"grad_norm": 4.022960662841797,
"learning_rate": 4.0000000000000003e-07,
"loss": 0.2763,
"step": 3
},
{
"epoch": 0.00016,
"grad_norm": 10.033286094665527,
"learning_rate": 6.000000000000001e-07,
"loss": 1.0452,
"step": 4
},
{
"epoch": 0.0002,
"grad_norm": 10.376935958862305,
"learning_rate": 8.000000000000001e-07,
"loss": 1.2463,
"step": 5
},
{
"epoch": 0.00024,
"grad_norm": 10.767428398132324,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.7617,
"step": 6
},
{
"epoch": 0.00028,
"grad_norm": 12.014700889587402,
"learning_rate": 1.2000000000000002e-06,
"loss": 0.5454,
"step": 7
},
{
"epoch": 0.00032,
"grad_norm": 6.369264125823975,
"learning_rate": 1.4000000000000001e-06,
"loss": 0.4519,
"step": 8
},
{
"epoch": 0.00036,
"grad_norm": 9.299721717834473,
"learning_rate": 1.6000000000000001e-06,
"loss": 1.0,
"step": 9
},
{
"epoch": 0.0004,
"grad_norm": 3.5960326194763184,
"learning_rate": 1.8000000000000001e-06,
"loss": 0.2663,
"step": 10
},
{
"epoch": 0.00044,
"grad_norm": 4.393957138061523,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.3945,
"step": 11
},
{
"epoch": 0.00048,
"grad_norm": 12.956979751586914,
"learning_rate": 2.2e-06,
"loss": 1.1383,
"step": 12
},
{
"epoch": 0.00052,
"grad_norm": 8.443684577941895,
"learning_rate": 2.4000000000000003e-06,
"loss": 1.3452,
"step": 13
},
{
"epoch": 0.00056,
"grad_norm": 5.837546348571777,
"learning_rate": 2.6e-06,
"loss": 0.6561,
"step": 14
},
{
"epoch": 0.0006,
"grad_norm": 12.459277153015137,
"learning_rate": 2.8000000000000003e-06,
"loss": 0.8461,
"step": 15
},
{
"epoch": 0.00064,
"grad_norm": 5.071094036102295,
"learning_rate": 3e-06,
"loss": 0.33,
"step": 16
},
{
"epoch": 0.00068,
"grad_norm": 4.945703983306885,
"learning_rate": 3.2000000000000003e-06,
"loss": 0.3767,
"step": 17
},
{
"epoch": 0.00072,
"grad_norm": 9.724709510803223,
"learning_rate": 3.4000000000000005e-06,
"loss": 1.0036,
"step": 18
},
{
"epoch": 0.00076,
"grad_norm": 7.573561668395996,
"learning_rate": 3.6000000000000003e-06,
"loss": 0.5544,
"step": 19
},
{
"epoch": 0.0008,
"grad_norm": 7.827940940856934,
"learning_rate": 3.8000000000000005e-06,
"loss": 0.5613,
"step": 20
},
{
"epoch": 0.00084,
"grad_norm": 17.20424461364746,
"learning_rate": 4.000000000000001e-06,
"loss": 0.9706,
"step": 21
},
{
"epoch": 0.00088,
"grad_norm": 10.012373924255371,
"learning_rate": 4.2000000000000004e-06,
"loss": 1.0051,
"step": 22
},
{
"epoch": 0.00092,
"grad_norm": 6.319075107574463,
"learning_rate": 4.4e-06,
"loss": 0.6423,
"step": 23
},
{
"epoch": 0.00096,
"grad_norm": 9.358919143676758,
"learning_rate": 4.600000000000001e-06,
"loss": 0.517,
"step": 24
},
{
"epoch": 0.001,
"grad_norm": 9.364446640014648,
"learning_rate": 4.800000000000001e-06,
"loss": 0.6887,
"step": 25
},
{
"epoch": 0.00104,
"grad_norm": 7.857821941375732,
"learning_rate": 5e-06,
"loss": 0.6819,
"step": 26
},
{
"epoch": 0.00108,
"grad_norm": 8.474349021911621,
"learning_rate": 5.2e-06,
"loss": 0.7982,
"step": 27
},
{
"epoch": 0.00112,
"grad_norm": 6.741444110870361,
"learning_rate": 5.400000000000001e-06,
"loss": 0.7161,
"step": 28
},
{
"epoch": 0.00116,
"grad_norm": 18.534074783325195,
"learning_rate": 5.600000000000001e-06,
"loss": 1.0627,
"step": 29
},
{
"epoch": 0.0012,
"grad_norm": 10.099328994750977,
"learning_rate": 5.8e-06,
"loss": 0.8822,
"step": 30
},
{
"epoch": 0.00124,
"grad_norm": 11.268096923828125,
"learning_rate": 6e-06,
"loss": 1.0784,
"step": 31
},
{
"epoch": 0.00128,
"grad_norm": 6.63427734375,
"learning_rate": 6.200000000000001e-06,
"loss": 0.4633,
"step": 32
},
{
"epoch": 0.00132,
"grad_norm": 6.449986457824707,
"learning_rate": 6.4000000000000006e-06,
"loss": 0.4361,
"step": 33
},
{
"epoch": 0.00136,
"grad_norm": 8.761893272399902,
"learning_rate": 6.600000000000001e-06,
"loss": 0.547,
"step": 34
},
{
"epoch": 0.0014,
"grad_norm": 6.562529563903809,
"learning_rate": 6.800000000000001e-06,
"loss": 0.8089,
"step": 35
},
{
"epoch": 0.00144,
"grad_norm": 9.770228385925293,
"learning_rate": 7e-06,
"loss": 1.1144,
"step": 36
},
{
"epoch": 0.00148,
"grad_norm": 5.148852348327637,
"learning_rate": 7.2000000000000005e-06,
"loss": 0.293,
"step": 37
},
{
"epoch": 0.00152,
"grad_norm": 4.236900329589844,
"learning_rate": 7.4e-06,
"loss": 0.262,
"step": 38
},
{
"epoch": 0.00156,
"grad_norm": 5.437747001647949,
"learning_rate": 7.600000000000001e-06,
"loss": 0.3241,
"step": 39
},
{
"epoch": 0.0016,
"grad_norm": 8.177858352661133,
"learning_rate": 7.800000000000002e-06,
"loss": 1.1006,
"step": 40
},
{
"epoch": 0.00164,
"grad_norm": 5.940142631530762,
"learning_rate": 8.000000000000001e-06,
"loss": 0.3883,
"step": 41
},
{
"epoch": 0.00168,
"grad_norm": 6.927196025848389,
"learning_rate": 8.2e-06,
"loss": 0.3768,
"step": 42
},
{
"epoch": 0.00172,
"grad_norm": 9.112464904785156,
"learning_rate": 8.400000000000001e-06,
"loss": 0.6569,
"step": 43
},
{
"epoch": 0.00176,
"grad_norm": 6.493500709533691,
"learning_rate": 8.6e-06,
"loss": 0.8258,
"step": 44
},
{
"epoch": 0.0018,
"grad_norm": 6.875064849853516,
"learning_rate": 8.8e-06,
"loss": 0.5197,
"step": 45
},
{
"epoch": 0.00184,
"grad_norm": 9.669301986694336,
"learning_rate": 9e-06,
"loss": 0.4608,
"step": 46
},
{
"epoch": 0.00188,
"grad_norm": 3.4440531730651855,
"learning_rate": 9.200000000000002e-06,
"loss": 0.4541,
"step": 47
},
{
"epoch": 0.00192,
"grad_norm": 12.573284149169922,
"learning_rate": 9.4e-06,
"loss": 0.8095,
"step": 48
},
{
"epoch": 0.00196,
"grad_norm": 7.087505340576172,
"learning_rate": 9.600000000000001e-06,
"loss": 0.48,
"step": 49
},
{
"epoch": 0.002,
"grad_norm": 9.410057067871094,
"learning_rate": 9.800000000000001e-06,
"loss": 0.6475,
"step": 50
},
{
"epoch": 0.00204,
"grad_norm": 9.903523445129395,
"learning_rate": 1e-05,
"loss": 0.3327,
"step": 51
},
{
"epoch": 0.00208,
"grad_norm": 7.001255512237549,
"learning_rate": 1.02e-05,
"loss": 0.3985,
"step": 52
},
{
"epoch": 0.00212,
"grad_norm": 4.249057292938232,
"learning_rate": 1.04e-05,
"loss": 0.4279,
"step": 53
},
{
"epoch": 0.00216,
"grad_norm": 7.542069911956787,
"learning_rate": 1.0600000000000002e-05,
"loss": 0.2604,
"step": 54
},
{
"epoch": 0.0022,
"grad_norm": 2.640810012817383,
"learning_rate": 1.0800000000000002e-05,
"loss": 0.2822,
"step": 55
},
{
"epoch": 0.00224,
"grad_norm": 7.7248921394348145,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.4523,
"step": 56
},
{
"epoch": 0.00228,
"grad_norm": 11.321695327758789,
"learning_rate": 1.1200000000000001e-05,
"loss": 0.8242,
"step": 57
},
{
"epoch": 0.00232,
"grad_norm": 6.157843589782715,
"learning_rate": 1.14e-05,
"loss": 0.2018,
"step": 58
},
{
"epoch": 0.00236,
"grad_norm": 7.961790084838867,
"learning_rate": 1.16e-05,
"loss": 0.4284,
"step": 59
},
{
"epoch": 0.0024,
"grad_norm": 3.069061756134033,
"learning_rate": 1.18e-05,
"loss": 0.3374,
"step": 60
},
{
"epoch": 0.00244,
"grad_norm": 3.2864911556243896,
"learning_rate": 1.2e-05,
"loss": 0.3036,
"step": 61
},
{
"epoch": 0.00248,
"grad_norm": 6.05866003036499,
"learning_rate": 1.22e-05,
"loss": 0.3123,
"step": 62
},
{
"epoch": 0.00252,
"grad_norm": 6.834275245666504,
"learning_rate": 1.2400000000000002e-05,
"loss": 0.4029,
"step": 63
},
{
"epoch": 0.00256,
"grad_norm": 7.419852256774902,
"learning_rate": 1.2600000000000001e-05,
"loss": 0.6642,
"step": 64
},
{
"epoch": 0.0026,
"grad_norm": 5.876731872558594,
"learning_rate": 1.2800000000000001e-05,
"loss": 0.4285,
"step": 65
},
{
"epoch": 0.00264,
"grad_norm": 5.972131252288818,
"learning_rate": 1.3000000000000001e-05,
"loss": 0.3707,
"step": 66
},
{
"epoch": 0.00268,
"grad_norm": 10.270156860351562,
"learning_rate": 1.3200000000000002e-05,
"loss": 0.806,
"step": 67
},
{
"epoch": 0.00272,
"grad_norm": 9.661238670349121,
"learning_rate": 1.3400000000000002e-05,
"loss": 0.4231,
"step": 68
},
{
"epoch": 0.00276,
"grad_norm": 7.479933738708496,
"learning_rate": 1.3600000000000002e-05,
"loss": 0.3502,
"step": 69
},
{
"epoch": 0.0028,
"grad_norm": 7.699158191680908,
"learning_rate": 1.38e-05,
"loss": 0.5334,
"step": 70
},
{
"epoch": 0.00284,
"grad_norm": 5.17159366607666,
"learning_rate": 1.4e-05,
"loss": 0.3446,
"step": 71
},
{
"epoch": 0.00288,
"grad_norm": 5.7967329025268555,
"learning_rate": 1.4200000000000001e-05,
"loss": 0.5111,
"step": 72
},
{
"epoch": 0.00292,
"grad_norm": 6.319905757904053,
"learning_rate": 1.4400000000000001e-05,
"loss": 0.5757,
"step": 73
},
{
"epoch": 0.00296,
"grad_norm": 4.1700544357299805,
"learning_rate": 1.46e-05,
"loss": 0.1925,
"step": 74
},
{
"epoch": 0.003,
"grad_norm": 8.075864791870117,
"learning_rate": 1.48e-05,
"loss": 1.1174,
"step": 75
},
{
"epoch": 0.00304,
"grad_norm": 3.858011245727539,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.378,
"step": 76
},
{
"epoch": 0.00308,
"grad_norm": 9.214083671569824,
"learning_rate": 1.5200000000000002e-05,
"loss": 0.5358,
"step": 77
},
{
"epoch": 0.00312,
"grad_norm": 6.904912948608398,
"learning_rate": 1.54e-05,
"loss": 0.6374,
"step": 78
},
{
"epoch": 0.00316,
"grad_norm": 9.454837799072266,
"learning_rate": 1.5600000000000003e-05,
"loss": 0.8326,
"step": 79
},
{
"epoch": 0.0032,
"grad_norm": 5.443726539611816,
"learning_rate": 1.58e-05,
"loss": 0.3248,
"step": 80
},
{
"epoch": 0.00324,
"grad_norm": 4.159465312957764,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.194,
"step": 81
},
{
"epoch": 0.00328,
"grad_norm": 3.700016498565674,
"learning_rate": 1.62e-05,
"loss": 0.2714,
"step": 82
},
{
"epoch": 0.00332,
"grad_norm": 5.297624588012695,
"learning_rate": 1.64e-05,
"loss": 0.4755,
"step": 83
},
{
"epoch": 0.00336,
"grad_norm": 5.727051734924316,
"learning_rate": 1.66e-05,
"loss": 0.1816,
"step": 84
},
{
"epoch": 0.0034,
"grad_norm": 3.0839762687683105,
"learning_rate": 1.6800000000000002e-05,
"loss": 0.1303,
"step": 85
},
{
"epoch": 0.00344,
"grad_norm": 7.538758754730225,
"learning_rate": 1.7e-05,
"loss": 0.5356,
"step": 86
},
{
"epoch": 0.00348,
"grad_norm": 3.444230318069458,
"learning_rate": 1.72e-05,
"loss": 0.2291,
"step": 87
},
{
"epoch": 0.00352,
"grad_norm": 3.036506414413452,
"learning_rate": 1.7400000000000003e-05,
"loss": 0.2085,
"step": 88
},
{
"epoch": 0.00356,
"grad_norm": 10.547761917114258,
"learning_rate": 1.76e-05,
"loss": 0.7026,
"step": 89
},
{
"epoch": 0.0036,
"grad_norm": 3.0479331016540527,
"learning_rate": 1.7800000000000002e-05,
"loss": 0.1748,
"step": 90
},
{
"epoch": 0.00364,
"grad_norm": 10.161134719848633,
"learning_rate": 1.8e-05,
"loss": 0.595,
"step": 91
},
{
"epoch": 0.00368,
"grad_norm": 4.264101982116699,
"learning_rate": 1.8200000000000002e-05,
"loss": 0.3369,
"step": 92
},
{
"epoch": 0.00372,
"grad_norm": 4.638044834136963,
"learning_rate": 1.8400000000000003e-05,
"loss": 0.4963,
"step": 93
},
{
"epoch": 0.00376,
"grad_norm": 3.6776742935180664,
"learning_rate": 1.86e-05,
"loss": 0.4082,
"step": 94
},
{
"epoch": 0.0038,
"grad_norm": 6.691776275634766,
"learning_rate": 1.88e-05,
"loss": 0.6748,
"step": 95
},
{
"epoch": 0.00384,
"grad_norm": 3.668527364730835,
"learning_rate": 1.9e-05,
"loss": 0.3195,
"step": 96
},
{
"epoch": 0.00388,
"grad_norm": 4.157833099365234,
"learning_rate": 1.9200000000000003e-05,
"loss": 0.3878,
"step": 97
},
{
"epoch": 0.00392,
"grad_norm": 4.20045280456543,
"learning_rate": 1.94e-05,
"loss": 0.325,
"step": 98
},
{
"epoch": 0.00396,
"grad_norm": 3.982357978820801,
"learning_rate": 1.9600000000000002e-05,
"loss": 0.4733,
"step": 99
},
{
"epoch": 0.004,
"grad_norm": 1.7636414766311646,
"learning_rate": 1.98e-05,
"loss": 0.1415,
"step": 100
},
{
"epoch": 0.00404,
"grad_norm": 3.102644205093384,
"learning_rate": 2e-05,
"loss": 0.2774,
"step": 101
},
{
"epoch": 0.00408,
"grad_norm": 4.943758487701416,
"learning_rate": 1.9999939076577906e-05,
"loss": 0.0937,
"step": 102
},
{
"epoch": 0.00412,
"grad_norm": 9.392162322998047,
"learning_rate": 1.9999756307053947e-05,
"loss": 0.538,
"step": 103
},
{
"epoch": 0.00416,
"grad_norm": 2.991089344024658,
"learning_rate": 1.9999451693655125e-05,
"loss": 0.2169,
"step": 104
},
{
"epoch": 0.0042,
"grad_norm": 15.384001731872559,
"learning_rate": 1.9999025240093045e-05,
"loss": 0.7037,
"step": 105
},
{
"epoch": 0.00424,
"grad_norm": 7.158446311950684,
"learning_rate": 1.9998476951563914e-05,
"loss": 0.4716,
"step": 106
},
{
"epoch": 0.00428,
"grad_norm": 8.647867202758789,
"learning_rate": 1.9997806834748455e-05,
"loss": 0.7203,
"step": 107
},
{
"epoch": 0.00432,
"grad_norm": 2.152014970779419,
"learning_rate": 1.9997014897811834e-05,
"loss": 0.2378,
"step": 108
},
{
"epoch": 0.00436,
"grad_norm": 10.478302001953125,
"learning_rate": 1.9996101150403543e-05,
"loss": 0.4447,
"step": 109
},
{
"epoch": 0.0044,
"grad_norm": 4.135096073150635,
"learning_rate": 1.9995065603657317e-05,
"loss": 0.3998,
"step": 110
},
{
"epoch": 0.00444,
"grad_norm": 8.611021995544434,
"learning_rate": 1.999390827019096e-05,
"loss": 0.6258,
"step": 111
},
{
"epoch": 0.00448,
"grad_norm": 5.728850841522217,
"learning_rate": 1.999262916410621e-05,
"loss": 0.515,
"step": 112
},
{
"epoch": 0.00452,
"grad_norm": 6.498945713043213,
"learning_rate": 1.9991228300988586e-05,
"loss": 0.4832,
"step": 113
},
{
"epoch": 0.00456,
"grad_norm": 4.493791103363037,
"learning_rate": 1.998970569790715e-05,
"loss": 0.4088,
"step": 114
},
{
"epoch": 0.0046,
"grad_norm": 1.4200173616409302,
"learning_rate": 1.9988061373414342e-05,
"loss": 0.062,
"step": 115
},
{
"epoch": 0.00464,
"grad_norm": 3.772876501083374,
"learning_rate": 1.9986295347545738e-05,
"loss": 0.2712,
"step": 116
},
{
"epoch": 0.00468,
"grad_norm": 3.594358444213867,
"learning_rate": 1.9984407641819812e-05,
"loss": 0.1541,
"step": 117
},
{
"epoch": 0.00472,
"grad_norm": 3.589097738265991,
"learning_rate": 1.9982398279237657e-05,
"loss": 0.2957,
"step": 118
},
{
"epoch": 0.00476,
"grad_norm": 9.986791610717773,
"learning_rate": 1.9980267284282718e-05,
"loss": 0.6126,
"step": 119
},
{
"epoch": 0.0048,
"grad_norm": 6.40962553024292,
"learning_rate": 1.9978014682920503e-05,
"loss": 0.4133,
"step": 120
},
{
"epoch": 0.00484,
"grad_norm": 6.308218479156494,
"learning_rate": 1.9975640502598243e-05,
"loss": 0.4272,
"step": 121
},
{
"epoch": 0.00488,
"grad_norm": 2.0763957500457764,
"learning_rate": 1.997314477224458e-05,
"loss": 0.1745,
"step": 122
},
{
"epoch": 0.00492,
"grad_norm": 1.9505949020385742,
"learning_rate": 1.9970527522269204e-05,
"loss": 0.1623,
"step": 123
},
{
"epoch": 0.00496,
"grad_norm": 7.443726062774658,
"learning_rate": 1.9967788784562474e-05,
"loss": 0.3502,
"step": 124
},
{
"epoch": 0.005,
"grad_norm": 2.5247890949249268,
"learning_rate": 1.9964928592495046e-05,
"loss": 0.1832,
"step": 125
},
{
"epoch": 0.00504,
"grad_norm": 3.5165812969207764,
"learning_rate": 1.9961946980917457e-05,
"loss": 0.2181,
"step": 126
},
{
"epoch": 0.00508,
"grad_norm": 2.217538833618164,
"learning_rate": 1.9958843986159705e-05,
"loss": 0.1518,
"step": 127
},
{
"epoch": 0.00512,
"grad_norm": 5.364406108856201,
"learning_rate": 1.99556196460308e-05,
"loss": 0.3015,
"step": 128
},
{
"epoch": 0.00516,
"grad_norm": 1.2830591201782227,
"learning_rate": 1.9952273999818312e-05,
"loss": 0.1128,
"step": 129
},
{
"epoch": 0.0052,
"grad_norm": 1.8974236249923706,
"learning_rate": 1.9948807088287884e-05,
"loss": 0.1156,
"step": 130
},
{
"epoch": 0.00524,
"grad_norm": 5.5034637451171875,
"learning_rate": 1.9945218953682736e-05,
"loss": 0.1888,
"step": 131
},
{
"epoch": 0.00528,
"grad_norm": 3.2731258869171143,
"learning_rate": 1.9941509639723155e-05,
"loss": 0.1784,
"step": 132
},
{
"epoch": 0.00532,
"grad_norm": 6.433404445648193,
"learning_rate": 1.9937679191605964e-05,
"loss": 0.2889,
"step": 133
},
{
"epoch": 0.00536,
"grad_norm": 1.0105794668197632,
"learning_rate": 1.9933727656003964e-05,
"loss": 0.0523,
"step": 134
},
{
"epoch": 0.0054,
"grad_norm": 4.031461238861084,
"learning_rate": 1.992965508106537e-05,
"loss": 0.1157,
"step": 135
},
{
"epoch": 0.00544,
"grad_norm": 6.666305065155029,
"learning_rate": 1.9925461516413224e-05,
"loss": 0.2862,
"step": 136
},
{
"epoch": 0.00548,
"grad_norm": 1.8451870679855347,
"learning_rate": 1.9921147013144782e-05,
"loss": 0.1048,
"step": 137
},
{
"epoch": 0.00552,
"grad_norm": 10.802168846130371,
"learning_rate": 1.9916711623830904e-05,
"loss": 0.3962,
"step": 138
},
{
"epoch": 0.00556,
"grad_norm": 5.268034934997559,
"learning_rate": 1.991215540251542e-05,
"loss": 0.2294,
"step": 139
},
{
"epoch": 0.0056,
"grad_norm": 4.220214366912842,
"learning_rate": 1.9907478404714438e-05,
"loss": 0.2442,
"step": 140
},
{
"epoch": 0.00564,
"grad_norm": 11.643061637878418,
"learning_rate": 1.9902680687415704e-05,
"loss": 0.227,
"step": 141
},
{
"epoch": 0.00568,
"grad_norm": 5.165961742401123,
"learning_rate": 1.989776230907789e-05,
"loss": 0.2345,
"step": 142
},
{
"epoch": 0.00572,
"grad_norm": 6.70469331741333,
"learning_rate": 1.9892723329629885e-05,
"loss": 0.2632,
"step": 143
},
{
"epoch": 0.00576,
"grad_norm": 5.160849094390869,
"learning_rate": 1.988756381047006e-05,
"loss": 0.2725,
"step": 144
},
{
"epoch": 0.0058,
"grad_norm": 22.006776809692383,
"learning_rate": 1.988228381446553e-05,
"loss": 0.7352,
"step": 145
},
{
"epoch": 0.00584,
"grad_norm": 9.433754920959473,
"learning_rate": 1.9876883405951378e-05,
"loss": 0.3162,
"step": 146
},
{
"epoch": 0.00588,
"grad_norm": 11.14606761932373,
"learning_rate": 1.987136265072988e-05,
"loss": 0.8743,
"step": 147
},
{
"epoch": 0.00592,
"grad_norm": 4.231512069702148,
"learning_rate": 1.9865721616069695e-05,
"loss": 0.2018,
"step": 148
},
{
"epoch": 0.00596,
"grad_norm": 8.273058891296387,
"learning_rate": 1.985996037070505e-05,
"loss": 0.3279,
"step": 149
},
{
"epoch": 0.006,
"grad_norm": 11.070712089538574,
"learning_rate": 1.9854078984834904e-05,
"loss": 0.3908,
"step": 150
},
{
"epoch": 0.00604,
"grad_norm": 1.3856514692306519,
"learning_rate": 1.9848077530122083e-05,
"loss": 0.0316,
"step": 151
},
{
"epoch": 0.00608,
"grad_norm": 13.312458992004395,
"learning_rate": 1.984195607969242e-05,
"loss": 0.4792,
"step": 152
},
{
"epoch": 0.00612,
"grad_norm": 6.6064887046813965,
"learning_rate": 1.983571470813386e-05,
"loss": 0.2642,
"step": 153
},
{
"epoch": 0.00616,
"grad_norm": 2.067622423171997,
"learning_rate": 1.9829353491495545e-05,
"loss": 0.1754,
"step": 154
},
{
"epoch": 0.0062,
"grad_norm": 4.129333019256592,
"learning_rate": 1.982287250728689e-05,
"loss": 0.169,
"step": 155
},
{
"epoch": 0.00624,
"grad_norm": 9.485503196716309,
"learning_rate": 1.9816271834476642e-05,
"loss": 0.5863,
"step": 156
},
{
"epoch": 0.00628,
"grad_norm": 3.2972605228424072,
"learning_rate": 1.9809551553491918e-05,
"loss": 0.2046,
"step": 157
},
{
"epoch": 0.00632,
"grad_norm": 8.819018363952637,
"learning_rate": 1.9802711746217222e-05,
"loss": 0.4437,
"step": 158
},
{
"epoch": 0.00636,
"grad_norm": 4.7125325202941895,
"learning_rate": 1.979575249599344e-05,
"loss": 0.2165,
"step": 159
},
{
"epoch": 0.0064,
"grad_norm": 3.444054126739502,
"learning_rate": 1.9788673887616852e-05,
"loss": 0.1835,
"step": 160
},
{
"epoch": 0.00644,
"grad_norm": 9.992177963256836,
"learning_rate": 1.9781476007338058e-05,
"loss": 0.5721,
"step": 161
},
{
"epoch": 0.00648,
"grad_norm": 1.2397490739822388,
"learning_rate": 1.9774158942860962e-05,
"loss": 0.1216,
"step": 162
},
{
"epoch": 0.00652,
"grad_norm": 8.95528507232666,
"learning_rate": 1.9766722783341682e-05,
"loss": 0.4782,
"step": 163
},
{
"epoch": 0.00656,
"grad_norm": 6.015171527862549,
"learning_rate": 1.9759167619387474e-05,
"loss": 0.5257,
"step": 164
},
{
"epoch": 0.0066,
"grad_norm": 8.550641059875488,
"learning_rate": 1.9751493543055634e-05,
"loss": 0.482,
"step": 165
},
{
"epoch": 0.00664,
"grad_norm": 10.4812650680542,
"learning_rate": 1.9743700647852356e-05,
"loss": 0.8337,
"step": 166
},
{
"epoch": 0.00668,
"grad_norm": 4.010517120361328,
"learning_rate": 1.9735789028731603e-05,
"loss": 0.2009,
"step": 167
},
{
"epoch": 0.00672,
"grad_norm": 12.656819343566895,
"learning_rate": 1.972775878209397e-05,
"loss": 0.7649,
"step": 168
},
{
"epoch": 0.00676,
"grad_norm": 1.799157738685608,
"learning_rate": 1.9719610005785466e-05,
"loss": 0.0878,
"step": 169
},
{
"epoch": 0.0068,
"grad_norm": 4.904173374176025,
"learning_rate": 1.971134279909636e-05,
"loss": 0.1853,
"step": 170
},
{
"epoch": 0.00684,
"grad_norm": 6.851920127868652,
"learning_rate": 1.9702957262759964e-05,
"loss": 0.2964,
"step": 171
},
{
"epoch": 0.00688,
"grad_norm": 2.041281223297119,
"learning_rate": 1.9694453498951392e-05,
"loss": 0.1472,
"step": 172
},
{
"epoch": 0.00692,
"grad_norm": 2.5453243255615234,
"learning_rate": 1.9685831611286312e-05,
"loss": 0.2008,
"step": 173
},
{
"epoch": 0.00696,
"grad_norm": 3.895217180252075,
"learning_rate": 1.9677091704819714e-05,
"loss": 0.1509,
"step": 174
},
{
"epoch": 0.007,
"grad_norm": 4.817748069763184,
"learning_rate": 1.9668233886044597e-05,
"loss": 0.2481,
"step": 175
},
{
"epoch": 0.00704,
"grad_norm": 4.640423774719238,
"learning_rate": 1.9659258262890683e-05,
"loss": 0.3179,
"step": 176
},
{
"epoch": 0.00708,
"grad_norm": 6.663148880004883,
"learning_rate": 1.9650164944723116e-05,
"loss": 0.3699,
"step": 177
},
{
"epoch": 0.00712,
"grad_norm": 9.907018661499023,
"learning_rate": 1.96409540423411e-05,
"loss": 0.3391,
"step": 178
},
{
"epoch": 0.00716,
"grad_norm": 5.532843112945557,
"learning_rate": 1.9631625667976584e-05,
"loss": 0.4711,
"step": 179
},
{
"epoch": 0.0072,
"grad_norm": 3.596653938293457,
"learning_rate": 1.9622179935292855e-05,
"loss": 0.173,
"step": 180
},
{
"epoch": 0.00724,
"grad_norm": 1.060608148574829,
"learning_rate": 1.961261695938319e-05,
"loss": 0.127,
"step": 181
},
{
"epoch": 0.00728,
"grad_norm": 2.948293685913086,
"learning_rate": 1.9602936856769432e-05,
"loss": 0.1304,
"step": 182
},
{
"epoch": 0.00732,
"grad_norm": 7.642879009246826,
"learning_rate": 1.9593139745400575e-05,
"loss": 0.4445,
"step": 183
},
{
"epoch": 0.00736,
"grad_norm": 4.727455139160156,
"learning_rate": 1.9583225744651334e-05,
"loss": 0.2587,
"step": 184
},
{
"epoch": 0.0074,
"grad_norm": 4.146674156188965,
"learning_rate": 1.9573194975320672e-05,
"loss": 0.1411,
"step": 185
},
{
"epoch": 0.00744,
"grad_norm": 3.799476385116577,
"learning_rate": 1.9563047559630356e-05,
"loss": 0.1655,
"step": 186
},
{
"epoch": 0.00748,
"grad_norm": 0.9643378853797913,
"learning_rate": 1.9552783621223437e-05,
"loss": 0.0665,
"step": 187
},
{
"epoch": 0.00752,
"grad_norm": 12.112992286682129,
"learning_rate": 1.954240328516277e-05,
"loss": 0.505,
"step": 188
},
{
"epoch": 0.00756,
"grad_norm": 5.012889862060547,
"learning_rate": 1.9531906677929472e-05,
"loss": 0.2592,
"step": 189
},
{
"epoch": 0.0076,
"grad_norm": 9.197477340698242,
"learning_rate": 1.9521293927421388e-05,
"loss": 0.3256,
"step": 190
},
{
"epoch": 0.00764,
"grad_norm": 4.14984130859375,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.1971,
"step": 191
},
{
"epoch": 0.00768,
"grad_norm": 7.69926118850708,
"learning_rate": 1.9499720515246524e-05,
"loss": 0.275,
"step": 192
},
{
"epoch": 0.00772,
"grad_norm": 1.7916607856750488,
"learning_rate": 1.9488760116444966e-05,
"loss": 0.1555,
"step": 193
},
{
"epoch": 0.00776,
"grad_norm": 10.75692081451416,
"learning_rate": 1.947768410009586e-05,
"loss": 0.4081,
"step": 194
},
{
"epoch": 0.0078,
"grad_norm": 5.5486531257629395,
"learning_rate": 1.9466492601156964e-05,
"loss": 0.3176,
"step": 195
},
{
"epoch": 0.00784,
"grad_norm": 11.867058753967285,
"learning_rate": 1.945518575599317e-05,
"loss": 0.5854,
"step": 196
},
{
"epoch": 0.00788,
"grad_norm": 2.8567678928375244,
"learning_rate": 1.944376370237481e-05,
"loss": 0.1616,
"step": 197
},
{
"epoch": 0.00792,
"grad_norm": 13.744150161743164,
"learning_rate": 1.943222657947601e-05,
"loss": 0.2823,
"step": 198
},
{
"epoch": 0.00796,
"grad_norm": 1.6545466184616089,
"learning_rate": 1.942057452787297e-05,
"loss": 0.1732,
"step": 199
},
{
"epoch": 0.008,
"grad_norm": 10.220211029052734,
"learning_rate": 1.9408807689542257e-05,
"loss": 0.4433,
"step": 200
},
{
"epoch": 0.00804,
"grad_norm": 2.839857578277588,
"learning_rate": 1.9396926207859085e-05,
"loss": 0.1988,
"step": 201
},
{
"epoch": 0.00808,
"grad_norm": 10.585383415222168,
"learning_rate": 1.938493022759556e-05,
"loss": 0.5021,
"step": 202
},
{
"epoch": 0.00812,
"grad_norm": 6.100584506988525,
"learning_rate": 1.937281989491892e-05,
"loss": 0.3638,
"step": 203
},
{
"epoch": 0.00816,
"grad_norm": 4.249466419219971,
"learning_rate": 1.9360595357389735e-05,
"loss": 0.2963,
"step": 204
},
{
"epoch": 0.0082,
"grad_norm": 1.7668002843856812,
"learning_rate": 1.9348256763960146e-05,
"loss": 0.1223,
"step": 205
},
{
"epoch": 0.00824,
"grad_norm": 1.0705002546310425,
"learning_rate": 1.9335804264972018e-05,
"loss": 0.1161,
"step": 206
},
{
"epoch": 0.00828,
"grad_norm": 1.9667918682098389,
"learning_rate": 1.9323238012155125e-05,
"loss": 0.1068,
"step": 207
},
{
"epoch": 0.00832,
"grad_norm": 5.651544570922852,
"learning_rate": 1.9310558158625286e-05,
"loss": 0.2186,
"step": 208
},
{
"epoch": 0.00836,
"grad_norm": 7.810215473175049,
"learning_rate": 1.9297764858882516e-05,
"loss": 0.3804,
"step": 209
},
{
"epoch": 0.0084,
"grad_norm": 0.9781168103218079,
"learning_rate": 1.9284858268809135e-05,
"loss": 0.1182,
"step": 210
},
{
"epoch": 0.00844,
"grad_norm": 13.060949325561523,
"learning_rate": 1.9271838545667876e-05,
"loss": 0.4201,
"step": 211
},
{
"epoch": 0.00848,
"grad_norm": 12.052680015563965,
"learning_rate": 1.925870584809995e-05,
"loss": 0.6889,
"step": 212
},
{
"epoch": 0.00852,
"grad_norm": 5.625985622406006,
"learning_rate": 1.9245460336123136e-05,
"loss": 0.7996,
"step": 213
},
{
"epoch": 0.00856,
"grad_norm": 8.364924430847168,
"learning_rate": 1.923210217112981e-05,
"loss": 0.2753,
"step": 214
},
{
"epoch": 0.0086,
"grad_norm": 8.053346633911133,
"learning_rate": 1.9218631515885007e-05,
"loss": 0.3492,
"step": 215
},
{
"epoch": 0.00864,
"grad_norm": 13.339101791381836,
"learning_rate": 1.9205048534524405e-05,
"loss": 0.3965,
"step": 216
},
{
"epoch": 0.00868,
"grad_norm": 3.560100793838501,
"learning_rate": 1.9191353392552346e-05,
"loss": 0.0683,
"step": 217
},
{
"epoch": 0.00872,
"grad_norm": 9.164094924926758,
"learning_rate": 1.9177546256839814e-05,
"loss": 0.3794,
"step": 218
},
{
"epoch": 0.00876,
"grad_norm": 4.68350076675415,
"learning_rate": 1.9163627295622397e-05,
"loss": 0.1637,
"step": 219
},
{
"epoch": 0.0088,
"grad_norm": 2.200321674346924,
"learning_rate": 1.914959667849825e-05,
"loss": 0.1489,
"step": 220
},
{
"epoch": 0.00884,
"grad_norm": 5.131304740905762,
"learning_rate": 1.913545457642601e-05,
"loss": 0.1566,
"step": 221
},
{
"epoch": 0.00888,
"grad_norm": 16.439376831054688,
"learning_rate": 1.9121201161722732e-05,
"loss": 0.4314,
"step": 222
},
{
"epoch": 0.00892,
"grad_norm": 0.6519739627838135,
"learning_rate": 1.910683660806177e-05,
"loss": 0.0739,
"step": 223
},
{
"epoch": 0.00896,
"grad_norm": 14.120660781860352,
"learning_rate": 1.9092361090470688e-05,
"loss": 0.4078,
"step": 224
},
{
"epoch": 0.009,
"grad_norm": 19.97659683227539,
"learning_rate": 1.907777478532909e-05,
"loss": 0.6373,
"step": 225
},
{
"epoch": 0.00904,
"grad_norm": 0.695967972278595,
"learning_rate": 1.9063077870366504e-05,
"loss": 0.0106,
"step": 226
},
{
"epoch": 0.00908,
"grad_norm": 1.1565967798233032,
"learning_rate": 1.9048270524660197e-05,
"loss": 0.0908,
"step": 227
},
{
"epoch": 0.00912,
"grad_norm": 4.901784896850586,
"learning_rate": 1.903335292863301e-05,
"loss": 0.199,
"step": 228
},
{
"epoch": 0.00916,
"grad_norm": 12.38315486907959,
"learning_rate": 1.901832526405114e-05,
"loss": 0.3305,
"step": 229
},
{
"epoch": 0.0092,
"grad_norm": 4.404404163360596,
"learning_rate": 1.9003187714021936e-05,
"loss": 0.23,
"step": 230
},
{
"epoch": 0.00924,
"grad_norm": 1.9097952842712402,
"learning_rate": 1.8987940462991673e-05,
"loss": 0.0975,
"step": 231
},
{
"epoch": 0.00928,
"grad_norm": 11.423251152038574,
"learning_rate": 1.8972583696743284e-05,
"loss": 0.6701,
"step": 232
},
{
"epoch": 0.00932,
"grad_norm": 5.773044109344482,
"learning_rate": 1.895711760239413e-05,
"loss": 0.2212,
"step": 233
},
{
"epoch": 0.00936,
"grad_norm": 1.8632229566574097,
"learning_rate": 1.8941542368393683e-05,
"loss": 0.1467,
"step": 234
},
{
"epoch": 0.0094,
"grad_norm": 8.34904670715332,
"learning_rate": 1.892585818452126e-05,
"loss": 0.4135,
"step": 235
},
{
"epoch": 0.00944,
"grad_norm": 1.607643485069275,
"learning_rate": 1.891006524188368e-05,
"loss": 0.1585,
"step": 236
},
{
"epoch": 0.00948,
"grad_norm": 2.586597204208374,
"learning_rate": 1.889416373291298e-05,
"loss": 0.0966,
"step": 237
},
{
"epoch": 0.00952,
"grad_norm": 10.66177749633789,
"learning_rate": 1.8878153851364013e-05,
"loss": 0.5216,
"step": 238
},
{
"epoch": 0.00956,
"grad_norm": 2.7018775939941406,
"learning_rate": 1.8862035792312148e-05,
"loss": 0.16,
"step": 239
},
{
"epoch": 0.0096,
"grad_norm": 1.1436007022857666,
"learning_rate": 1.884580975215084e-05,
"loss": 0.123,
"step": 240
},
{
"epoch": 0.00964,
"grad_norm": 11.843042373657227,
"learning_rate": 1.8829475928589272e-05,
"loss": 0.4819,
"step": 241
},
{
"epoch": 0.00968,
"grad_norm": 7.968238353729248,
"learning_rate": 1.8813034520649923e-05,
"loss": 0.2928,
"step": 242
},
{
"epoch": 0.00972,
"grad_norm": 8.552411079406738,
"learning_rate": 1.879648572866617e-05,
"loss": 0.5179,
"step": 243
},
{
"epoch": 0.00976,
"grad_norm": 15.723119735717773,
"learning_rate": 1.8779829754279806e-05,
"loss": 0.3462,
"step": 244
},
{
"epoch": 0.0098,
"grad_norm": 5.117152690887451,
"learning_rate": 1.8763066800438638e-05,
"loss": 0.1817,
"step": 245
},
{
"epoch": 0.00984,
"grad_norm": 6.2736005783081055,
"learning_rate": 1.874619707139396e-05,
"loss": 0.1897,
"step": 246
},
{
"epoch": 0.00988,
"grad_norm": 15.791886329650879,
"learning_rate": 1.8729220772698096e-05,
"loss": 0.292,
"step": 247
},
{
"epoch": 0.00992,
"grad_norm": 1.3381388187408447,
"learning_rate": 1.8712138111201898e-05,
"loss": 0.0768,
"step": 248
},
{
"epoch": 0.00996,
"grad_norm": 0.7379228472709656,
"learning_rate": 1.869494929505219e-05,
"loss": 0.1019,
"step": 249
},
{
"epoch": 0.01,
"grad_norm": 2.443669319152832,
"learning_rate": 1.8677654533689287e-05,
"loss": 0.125,
"step": 250
},
{
"epoch": 0.01004,
"grad_norm": 13.191452026367188,
"learning_rate": 1.866025403784439e-05,
"loss": 0.7511,
"step": 251
},
{
"epoch": 0.01008,
"grad_norm": 8.021163940429688,
"learning_rate": 1.864274801953705e-05,
"loss": 0.2523,
"step": 252
},
{
"epoch": 0.01012,
"grad_norm": 3.703171968460083,
"learning_rate": 1.8625136692072577e-05,
"loss": 0.1894,
"step": 253
},
{
"epoch": 0.01016,
"grad_norm": 5.219841957092285,
"learning_rate": 1.860742027003944e-05,
"loss": 0.1573,
"step": 254
},
{
"epoch": 0.0102,
"grad_norm": 7.702592849731445,
"learning_rate": 1.8589598969306646e-05,
"loss": 0.309,
"step": 255
},
{
"epoch": 0.01024,
"grad_norm": 10.890402793884277,
"learning_rate": 1.8571673007021124e-05,
"loss": 0.5104,
"step": 256
},
{
"epoch": 0.01028,
"grad_norm": 0.994693398475647,
"learning_rate": 1.855364260160507e-05,
"loss": 0.1235,
"step": 257
},
{
"epoch": 0.01032,
"grad_norm": 1.669937252998352,
"learning_rate": 1.8535507972753275e-05,
"loss": 0.0983,
"step": 258
},
{
"epoch": 0.01036,
"grad_norm": 7.680068016052246,
"learning_rate": 1.851726934143048e-05,
"loss": 0.2318,
"step": 259
},
{
"epoch": 0.0104,
"grad_norm": 7.261491775512695,
"learning_rate": 1.849892692986864e-05,
"loss": 0.3479,
"step": 260
},
{
"epoch": 0.01044,
"grad_norm": 9.842647552490234,
"learning_rate": 1.848048096156426e-05,
"loss": 0.3888,
"step": 261
},
{
"epoch": 0.01048,
"grad_norm": 7.602323055267334,
"learning_rate": 1.8461931661275642e-05,
"loss": 0.3924,
"step": 262
},
{
"epoch": 0.01052,
"grad_norm": 5.419349193572998,
"learning_rate": 1.8443279255020153e-05,
"loss": 0.2522,
"step": 263
},
{
"epoch": 0.01056,
"grad_norm": 8.029243469238281,
"learning_rate": 1.842452397007148e-05,
"loss": 0.2222,
"step": 264
},
{
"epoch": 0.0106,
"grad_norm": 3.73085880279541,
"learning_rate": 1.8405666034956842e-05,
"loss": 0.1234,
"step": 265
},
{
"epoch": 0.01064,
"grad_norm": 9.521867752075195,
"learning_rate": 1.8386705679454243e-05,
"loss": 0.2874,
"step": 266
},
{
"epoch": 0.01068,
"grad_norm": 0.8158937692642212,
"learning_rate": 1.836764313458962e-05,
"loss": 0.0548,
"step": 267
},
{
"epoch": 0.01072,
"grad_norm": 3.66943621635437,
"learning_rate": 1.8348478632634067e-05,
"loss": 0.2049,
"step": 268
},
{
"epoch": 0.01076,
"grad_norm": 8.897522926330566,
"learning_rate": 1.8329212407100996e-05,
"loss": 0.4369,
"step": 269
},
{
"epoch": 0.0108,
"grad_norm": 10.526529312133789,
"learning_rate": 1.8309844692743283e-05,
"loss": 0.1816,
"step": 270
},
{
"epoch": 0.01084,
"grad_norm": 6.618228912353516,
"learning_rate": 1.8290375725550417e-05,
"loss": 0.1797,
"step": 271
},
{
"epoch": 0.01088,
"grad_norm": 4.8069047927856445,
"learning_rate": 1.827080574274562e-05,
"loss": 0.1868,
"step": 272
},
{
"epoch": 0.01092,
"grad_norm": 1.0820038318634033,
"learning_rate": 1.8251134982782952e-05,
"loss": 0.0999,
"step": 273
},
{
"epoch": 0.01096,
"grad_norm": 3.5766992568969727,
"learning_rate": 1.8231363685344422e-05,
"loss": 0.1693,
"step": 274
},
{
"epoch": 0.011,
"grad_norm": 2.414207935333252,
"learning_rate": 1.821149209133704e-05,
"loss": 0.137,
"step": 275
},
{
"epoch": 0.01104,
"grad_norm": 1.2485153675079346,
"learning_rate": 1.819152044288992e-05,
"loss": 0.1115,
"step": 276
},
{
"epoch": 0.01108,
"grad_norm": 8.366353988647461,
"learning_rate": 1.8171448983351284e-05,
"loss": 0.3453,
"step": 277
},
{
"epoch": 0.01112,
"grad_norm": 2.9404959678649902,
"learning_rate": 1.815127795728554e-05,
"loss": 0.1714,
"step": 278
},
{
"epoch": 0.01116,
"grad_norm": 4.233757972717285,
"learning_rate": 1.8131007610470278e-05,
"loss": 0.1122,
"step": 279
},
{
"epoch": 0.0112,
"grad_norm": 4.970396995544434,
"learning_rate": 1.8110638189893267e-05,
"loss": 0.1663,
"step": 280
},
{
"epoch": 0.01124,
"grad_norm": 6.489607334136963,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.2407,
"step": 281
},
{
"epoch": 0.01128,
"grad_norm": 10.362878799438477,
"learning_rate": 1.806960312143802e-05,
"loss": 0.3367,
"step": 282
},
{
"epoch": 0.01132,
"grad_norm": 2.0283446311950684,
"learning_rate": 1.804893797355914e-05,
"loss": 0.2153,
"step": 283
},
{
"epoch": 0.01136,
"grad_norm": 2.664823532104492,
"learning_rate": 1.8028174751911147e-05,
"loss": 0.1209,
"step": 284
},
{
"epoch": 0.0114,
"grad_norm": 2.6452689170837402,
"learning_rate": 1.8007313709487334e-05,
"loss": 0.0532,
"step": 285
},
{
"epoch": 0.01144,
"grad_norm": 9.645456314086914,
"learning_rate": 1.798635510047293e-05,
"loss": 0.4621,
"step": 286
},
{
"epoch": 0.01148,
"grad_norm": 2.127562999725342,
"learning_rate": 1.7965299180241963e-05,
"loss": 0.041,
"step": 287
},
{
"epoch": 0.01152,
"grad_norm": 9.280315399169922,
"learning_rate": 1.7944146205354182e-05,
"loss": 0.3096,
"step": 288
},
{
"epoch": 0.01156,
"grad_norm": 1.2085825204849243,
"learning_rate": 1.792289643355191e-05,
"loss": 0.0812,
"step": 289
},
{
"epoch": 0.0116,
"grad_norm": 1.6799724102020264,
"learning_rate": 1.7901550123756906e-05,
"loss": 0.1082,
"step": 290
},
{
"epoch": 0.01164,
"grad_norm": 11.399935722351074,
"learning_rate": 1.788010753606722e-05,
"loss": 0.3543,
"step": 291
},
{
"epoch": 0.01168,
"grad_norm": 5.221828937530518,
"learning_rate": 1.785856893175402e-05,
"loss": 0.1749,
"step": 292
},
{
"epoch": 0.01172,
"grad_norm": 2.5803136825561523,
"learning_rate": 1.78369345732584e-05,
"loss": 0.1306,
"step": 293
},
{
"epoch": 0.01176,
"grad_norm": 1.390418291091919,
"learning_rate": 1.781520472418819e-05,
"loss": 0.1334,
"step": 294
},
{
"epoch": 0.0118,
"grad_norm": 6.414823055267334,
"learning_rate": 1.7793379649314743e-05,
"loss": 0.267,
"step": 295
},
{
"epoch": 0.01184,
"grad_norm": 0.8822182416915894,
"learning_rate": 1.777145961456971e-05,
"loss": 0.085,
"step": 296
},
{
"epoch": 0.01188,
"grad_norm": 10.217781066894531,
"learning_rate": 1.7749444887041797e-05,
"loss": 0.2932,
"step": 297
},
{
"epoch": 0.01192,
"grad_norm": 7.361355304718018,
"learning_rate": 1.7727335734973512e-05,
"loss": 0.2422,
"step": 298
},
{
"epoch": 0.01196,
"grad_norm": 3.531399965286255,
"learning_rate": 1.7705132427757895e-05,
"loss": 0.1664,
"step": 299
},
{
"epoch": 0.012,
"grad_norm": 6.05151891708374,
"learning_rate": 1.7682835235935236e-05,
"loss": 0.1624,
"step": 300
},
{
"epoch": 0.01204,
"grad_norm": 14.604363441467285,
"learning_rate": 1.766044443118978e-05,
"loss": 0.5158,
"step": 301
},
{
"epoch": 0.01208,
"grad_norm": 0.9803701639175415,
"learning_rate": 1.7637960286346423e-05,
"loss": 0.098,
"step": 302
},
{
"epoch": 0.01212,
"grad_norm": 1.744943618774414,
"learning_rate": 1.761538307536737e-05,
"loss": 0.147,
"step": 303
},
{
"epoch": 0.01216,
"grad_norm": 7.417971134185791,
"learning_rate": 1.759271307334881e-05,
"loss": 0.314,
"step": 304
},
{
"epoch": 0.0122,
"grad_norm": 0.9366175532341003,
"learning_rate": 1.7569950556517566e-05,
"loss": 0.0922,
"step": 305
},
{
"epoch": 0.01224,
"grad_norm": 9.111652374267578,
"learning_rate": 1.7547095802227723e-05,
"loss": 0.7093,
"step": 306
},
{
"epoch": 0.01228,
"grad_norm": 6.51871919631958,
"learning_rate": 1.7524149088957244e-05,
"loss": 0.2305,
"step": 307
},
{
"epoch": 0.01232,
"grad_norm": 2.127795696258545,
"learning_rate": 1.7501110696304598e-05,
"loss": 0.087,
"step": 308
},
{
"epoch": 0.01236,
"grad_norm": 4.677096843719482,
"learning_rate": 1.747798090498532e-05,
"loss": 0.137,
"step": 309
},
{
"epoch": 0.0124,
"grad_norm": 5.362130165100098,
"learning_rate": 1.7454759996828622e-05,
"loss": 0.1796,
"step": 310
},
{
"epoch": 0.01244,
"grad_norm": 7.157973766326904,
"learning_rate": 1.7431448254773943e-05,
"loss": 0.3007,
"step": 311
},
{
"epoch": 0.01248,
"grad_norm": 3.481816530227661,
"learning_rate": 1.74080459628675e-05,
"loss": 0.1828,
"step": 312
},
{
"epoch": 0.01252,
"grad_norm": 5.90020227432251,
"learning_rate": 1.7384553406258842e-05,
"loss": 0.2356,
"step": 313
},
{
"epoch": 0.01256,
"grad_norm": 3.9650144577026367,
"learning_rate": 1.7360970871197347e-05,
"loss": 0.1994,
"step": 314
},
{
"epoch": 0.0126,
"grad_norm": 10.77530288696289,
"learning_rate": 1.7337298645028764e-05,
"loss": 0.2847,
"step": 315
},
{
"epoch": 0.01264,
"grad_norm": 2.7759039402008057,
"learning_rate": 1.7313537016191706e-05,
"loss": 0.1115,
"step": 316
},
{
"epoch": 0.01268,
"grad_norm": 1.2038787603378296,
"learning_rate": 1.7289686274214116e-05,
"loss": 0.1048,
"step": 317
},
{
"epoch": 0.01272,
"grad_norm": 4.866477966308594,
"learning_rate": 1.7265746709709762e-05,
"loss": 0.2001,
"step": 318
},
{
"epoch": 0.01276,
"grad_norm": 1.526968002319336,
"learning_rate": 1.7241718614374678e-05,
"loss": 0.0881,
"step": 319
},
{
"epoch": 0.0128,
"grad_norm": 9.805798530578613,
"learning_rate": 1.7217602280983622e-05,
"loss": 0.2875,
"step": 320
},
{
"epoch": 0.01284,
"grad_norm": 10.69047737121582,
"learning_rate": 1.7193398003386514e-05,
"loss": 0.64,
"step": 321
},
{
"epoch": 0.01288,
"grad_norm": 5.783357620239258,
"learning_rate": 1.716910607650483e-05,
"loss": 0.1505,
"step": 322
},
{
"epoch": 0.01292,
"grad_norm": 0.5916599631309509,
"learning_rate": 1.7144726796328034e-05,
"loss": 0.01,
"step": 323
},
{
"epoch": 0.01296,
"grad_norm": 2.425915002822876,
"learning_rate": 1.712026045990997e-05,
"loss": 0.1067,
"step": 324
},
{
"epoch": 0.013,
"grad_norm": 8.923468589782715,
"learning_rate": 1.709570736536521e-05,
"loss": 0.3188,
"step": 325
},
{
"epoch": 0.01304,
"grad_norm": 1.3892550468444824,
"learning_rate": 1.7071067811865477e-05,
"loss": 0.076,
"step": 326
},
{
"epoch": 0.01308,
"grad_norm": 8.633707046508789,
"learning_rate": 1.7046342099635948e-05,
"loss": 0.2557,
"step": 327
},
{
"epoch": 0.01312,
"grad_norm": 1.0339399576187134,
"learning_rate": 1.7021530529951627e-05,
"loss": 0.1276,
"step": 328
},
{
"epoch": 0.01316,
"grad_norm": 2.390277624130249,
"learning_rate": 1.6996633405133656e-05,
"loss": 0.0351,
"step": 329
},
{
"epoch": 0.0132,
"grad_norm": 10.007804870605469,
"learning_rate": 1.697165102854565e-05,
"loss": 0.3109,
"step": 330
},
{
"epoch": 0.01324,
"grad_norm": 12.343931198120117,
"learning_rate": 1.6946583704589973e-05,
"loss": 0.3273,
"step": 331
},
{
"epoch": 0.01328,
"grad_norm": 7.776202201843262,
"learning_rate": 1.692143173870407e-05,
"loss": 0.2345,
"step": 332
},
{
"epoch": 0.01332,
"grad_norm": 1.7889080047607422,
"learning_rate": 1.68961954373567e-05,
"loss": 0.0373,
"step": 333
},
{
"epoch": 0.01336,
"grad_norm": 15.659445762634277,
"learning_rate": 1.6870875108044233e-05,
"loss": 0.4619,
"step": 334
},
{
"epoch": 0.0134,
"grad_norm": 5.062111854553223,
"learning_rate": 1.684547105928689e-05,
"loss": 0.154,
"step": 335
},
{
"epoch": 0.01344,
"grad_norm": 9.29782485961914,
"learning_rate": 1.6819983600624986e-05,
"loss": 0.3113,
"step": 336
},
{
"epoch": 0.01348,
"grad_norm": 3.840139627456665,
"learning_rate": 1.6794413042615168e-05,
"loss": 0.0555,
"step": 337
},
{
"epoch": 0.01352,
"grad_norm": 16.228111267089844,
"learning_rate": 1.6768759696826608e-05,
"loss": 0.3978,
"step": 338
},
{
"epoch": 0.01356,
"grad_norm": 6.077329158782959,
"learning_rate": 1.6743023875837233e-05,
"loss": 0.224,
"step": 339
},
{
"epoch": 0.0136,
"grad_norm": 3.7727584838867188,
"learning_rate": 1.6717205893229904e-05,
"loss": 0.1967,
"step": 340
},
{
"epoch": 0.01364,
"grad_norm": 7.737804889678955,
"learning_rate": 1.6691306063588583e-05,
"loss": 0.2176,
"step": 341
},
{
"epoch": 0.01368,
"grad_norm": 6.864624500274658,
"learning_rate": 1.6665324702494524e-05,
"loss": 0.5521,
"step": 342
},
{
"epoch": 0.01372,
"grad_norm": 0.9121661186218262,
"learning_rate": 1.6639262126522417e-05,
"loss": 0.0974,
"step": 343
},
{
"epoch": 0.01376,
"grad_norm": 1.549866795539856,
"learning_rate": 1.661311865323652e-05,
"loss": 0.0327,
"step": 344
},
{
"epoch": 0.0138,
"grad_norm": 4.400731086730957,
"learning_rate": 1.6586894601186804e-05,
"loss": 0.2095,
"step": 345
},
{
"epoch": 0.01384,
"grad_norm": 10.884912490844727,
"learning_rate": 1.6560590289905074e-05,
"loss": 0.2897,
"step": 346
},
{
"epoch": 0.01388,
"grad_norm": 10.11740493774414,
"learning_rate": 1.6534206039901057e-05,
"loss": 0.3878,
"step": 347
},
{
"epoch": 0.01392,
"grad_norm": 0.5388306379318237,
"learning_rate": 1.650774217265851e-05,
"loss": 0.0376,
"step": 348
},
{
"epoch": 0.01396,
"grad_norm": 1.1815996170043945,
"learning_rate": 1.6481199010631312e-05,
"loss": 0.1069,
"step": 349
},
{
"epoch": 0.014,
"grad_norm": 5.323544025421143,
"learning_rate": 1.645457687723951e-05,
"loss": 0.2179,
"step": 350
},
{
"epoch": 0.01404,
"grad_norm": 0.40559855103492737,
"learning_rate": 1.6427876096865394e-05,
"loss": 0.0109,
"step": 351
},
{
"epoch": 0.01408,
"grad_norm": 4.135653972625732,
"learning_rate": 1.6401096994849558e-05,
"loss": 0.1731,
"step": 352
},
{
"epoch": 0.01412,
"grad_norm": 1.5208616256713867,
"learning_rate": 1.63742398974869e-05,
"loss": 0.0669,
"step": 353
},
{
"epoch": 0.01416,
"grad_norm": 3.293498992919922,
"learning_rate": 1.6347305132022677e-05,
"loss": 0.1614,
"step": 354
},
{
"epoch": 0.0142,
"grad_norm": 6.616631984710693,
"learning_rate": 1.632029302664851e-05,
"loss": 0.3651,
"step": 355
},
{
"epoch": 0.01424,
"grad_norm": 0.7986364960670471,
"learning_rate": 1.6293203910498375e-05,
"loss": 0.0951,
"step": 356
},
{
"epoch": 0.01428,
"grad_norm": 10.48706340789795,
"learning_rate": 1.6266038113644605e-05,
"loss": 0.2277,
"step": 357
},
{
"epoch": 0.01432,
"grad_norm": 5.115804195404053,
"learning_rate": 1.6238795967093865e-05,
"loss": 0.1061,
"step": 358
},
{
"epoch": 0.01436,
"grad_norm": 2.124567985534668,
"learning_rate": 1.6211477802783105e-05,
"loss": 0.1295,
"step": 359
},
{
"epoch": 0.0144,
"grad_norm": 7.559396743774414,
"learning_rate": 1.6184083953575543e-05,
"loss": 0.2108,
"step": 360
},
{
"epoch": 0.01444,
"grad_norm": 1.1018240451812744,
"learning_rate": 1.6156614753256583e-05,
"loss": 0.1014,
"step": 361
},
{
"epoch": 0.01448,
"grad_norm": 3.3138890266418457,
"learning_rate": 1.6129070536529767e-05,
"loss": 0.1083,
"step": 362
},
{
"epoch": 0.01452,
"grad_norm": 12.400237083435059,
"learning_rate": 1.610145163901268e-05,
"loss": 0.5254,
"step": 363
},
{
"epoch": 0.01456,
"grad_norm": 0.5973766446113586,
"learning_rate": 1.607375839723287e-05,
"loss": 0.0861,
"step": 364
},
{
"epoch": 0.0146,
"grad_norm": 24.465190887451172,
"learning_rate": 1.6045991148623752e-05,
"loss": 0.6478,
"step": 365
},
{
"epoch": 0.01464,
"grad_norm": 10.307867050170898,
"learning_rate": 1.6018150231520486e-05,
"loss": 0.2596,
"step": 366
},
{
"epoch": 0.01468,
"grad_norm": 1.6973850727081299,
"learning_rate": 1.599023598515586e-05,
"loss": 0.0152,
"step": 367
},
{
"epoch": 0.01472,
"grad_norm": 0.9010747671127319,
"learning_rate": 1.5962248749656158e-05,
"loss": 0.1107,
"step": 368
},
{
"epoch": 0.01476,
"grad_norm": 5.204710006713867,
"learning_rate": 1.5934188866037017e-05,
"loss": 0.2175,
"step": 369
},
{
"epoch": 0.0148,
"grad_norm": 6.111196517944336,
"learning_rate": 1.5906056676199256e-05,
"loss": 0.0492,
"step": 370
},
{
"epoch": 0.01484,
"grad_norm": 0.7997010350227356,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.114,
"step": 371
},
{
"epoch": 0.01488,
"grad_norm": 8.590513229370117,
"learning_rate": 1.584957674987216e-05,
"loss": 0.2807,
"step": 372
},
{
"epoch": 0.01492,
"grad_norm": 1.363950252532959,
"learning_rate": 1.5821229701572897e-05,
"loss": 0.1094,
"step": 373
},
{
"epoch": 0.01496,
"grad_norm": 6.824077129364014,
"learning_rate": 1.5792811723426787e-05,
"loss": 0.2313,
"step": 374
},
{
"epoch": 0.015,
"grad_norm": 6.9105730056762695,
"learning_rate": 1.5764323161697933e-05,
"loss": 0.4,
"step": 375
},
{
"epoch": 0.01504,
"grad_norm": 1.814996361732483,
"learning_rate": 1.573576436351046e-05,
"loss": 0.0708,
"step": 376
},
{
"epoch": 0.01508,
"grad_norm": 10.81425666809082,
"learning_rate": 1.570713567684432e-05,
"loss": 0.299,
"step": 377
},
{
"epoch": 0.01512,
"grad_norm": 1.391947627067566,
"learning_rate": 1.5678437450531014e-05,
"loss": 0.0955,
"step": 378
},
{
"epoch": 0.01516,
"grad_norm": 2.4521093368530273,
"learning_rate": 1.564967003424938e-05,
"loss": 0.085,
"step": 379
},
{
"epoch": 0.0152,
"grad_norm": 18.016063690185547,
"learning_rate": 1.5620833778521306e-05,
"loss": 0.5673,
"step": 380
},
{
"epoch": 0.01524,
"grad_norm": 0.2768113613128662,
"learning_rate": 1.5591929034707468e-05,
"loss": 0.0033,
"step": 381
},
{
"epoch": 0.01528,
"grad_norm": 3.321561574935913,
"learning_rate": 1.556295615500305e-05,
"loss": 0.2416,
"step": 382
},
{
"epoch": 0.01532,
"grad_norm": 8.045482635498047,
"learning_rate": 1.553391549243344e-05,
"loss": 0.2358,
"step": 383
},
{
"epoch": 0.01536,
"grad_norm": 2.1974477767944336,
"learning_rate": 1.5504807400849957e-05,
"loss": 0.0652,
"step": 384
},
{
"epoch": 0.0154,
"grad_norm": 0.9510599970817566,
"learning_rate": 1.5475632234925505e-05,
"loss": 0.0642,
"step": 385
},
{
"epoch": 0.01544,
"grad_norm": 1.2250057458877563,
"learning_rate": 1.5446390350150272e-05,
"loss": 0.0709,
"step": 386
},
{
"epoch": 0.01548,
"grad_norm": 6.398176193237305,
"learning_rate": 1.54170821028274e-05,
"loss": 0.2132,
"step": 387
},
{
"epoch": 0.01552,
"grad_norm": 11.493470191955566,
"learning_rate": 1.5387707850068633e-05,
"loss": 0.2124,
"step": 388
},
{
"epoch": 0.01556,
"grad_norm": 1.0480350255966187,
"learning_rate": 1.5358267949789968e-05,
"loss": 0.0957,
"step": 389
},
{
"epoch": 0.0156,
"grad_norm": 0.7153192758560181,
"learning_rate": 1.53287627607073e-05,
"loss": 0.069,
"step": 390
},
{
"epoch": 0.01564,
"grad_norm": 5.019766330718994,
"learning_rate": 1.529919264233205e-05,
"loss": 0.1526,
"step": 391
},
{
"epoch": 0.01568,
"grad_norm": 0.6558953523635864,
"learning_rate": 1.5269557954966777e-05,
"loss": 0.0693,
"step": 392
},
{
"epoch": 0.01572,
"grad_norm": 0.387051522731781,
"learning_rate": 1.5239859059700794e-05,
"loss": 0.008,
"step": 393
},
{
"epoch": 0.01576,
"grad_norm": 2.104628562927246,
"learning_rate": 1.5210096318405768e-05,
"loss": 0.1135,
"step": 394
},
{
"epoch": 0.0158,
"grad_norm": 0.6971196532249451,
"learning_rate": 1.5180270093731305e-05,
"loss": 0.0825,
"step": 395
},
{
"epoch": 0.01584,
"grad_norm": 1.813346028327942,
"learning_rate": 1.5150380749100545e-05,
"loss": 0.1208,
"step": 396
},
{
"epoch": 0.01588,
"grad_norm": 1.9763466119766235,
"learning_rate": 1.5120428648705716e-05,
"loss": 0.1479,
"step": 397
},
{
"epoch": 0.01592,
"grad_norm": 1.2226390838623047,
"learning_rate": 1.5090414157503715e-05,
"loss": 0.0836,
"step": 398
},
{
"epoch": 0.01596,
"grad_norm": 0.650795578956604,
"learning_rate": 1.5060337641211637e-05,
"loss": 0.092,
"step": 399
},
{
"epoch": 0.016,
"grad_norm": 1.016252040863037,
"learning_rate": 1.5030199466302354e-05,
"loss": 0.098,
"step": 400
},
{
"epoch": 0.01604,
"grad_norm": 2.108137369155884,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.1077,
"step": 401
},
{
"epoch": 0.01608,
"grad_norm": 2.788259506225586,
"learning_rate": 1.4969739610275556e-05,
"loss": 0.139,
"step": 402
},
{
"epoch": 0.01612,
"grad_norm": 1.8917183876037598,
"learning_rate": 1.493941866584231e-05,
"loss": 0.1193,
"step": 403
},
{
"epoch": 0.01616,
"grad_norm": 2.2906570434570312,
"learning_rate": 1.490903753615141e-05,
"loss": 0.1176,
"step": 404
},
{
"epoch": 0.0162,
"grad_norm": 1.0551007986068726,
"learning_rate": 1.4878596591387329e-05,
"loss": 0.1186,
"step": 405
},
{
"epoch": 0.01624,
"grad_norm": 2.6187283992767334,
"learning_rate": 1.4848096202463373e-05,
"loss": 0.1058,
"step": 406
},
{
"epoch": 0.01628,
"grad_norm": 7.166258335113525,
"learning_rate": 1.4817536741017153e-05,
"loss": 0.2088,
"step": 407
},
{
"epoch": 0.01632,
"grad_norm": 17.303178787231445,
"learning_rate": 1.478691857940607e-05,
"loss": 0.5358,
"step": 408
},
{
"epoch": 0.01636,
"grad_norm": 1.0750049352645874,
"learning_rate": 1.4756242090702756e-05,
"loss": 0.0824,
"step": 409
},
{
"epoch": 0.0164,
"grad_norm": 3.905895233154297,
"learning_rate": 1.4725507648690542e-05,
"loss": 0.1465,
"step": 410
},
{
"epoch": 0.01644,
"grad_norm": 0.7625576853752136,
"learning_rate": 1.469471562785891e-05,
"loss": 0.0654,
"step": 411
},
{
"epoch": 0.01648,
"grad_norm": 17.50716781616211,
"learning_rate": 1.4663866403398915e-05,
"loss": 0.8332,
"step": 412
},
{
"epoch": 0.01652,
"grad_norm": 5.220765590667725,
"learning_rate": 1.463296035119862e-05,
"loss": 0.1106,
"step": 413
},
{
"epoch": 0.01656,
"grad_norm": 2.130483388900757,
"learning_rate": 1.4601997847838518e-05,
"loss": 0.1326,
"step": 414
},
{
"epoch": 0.0166,
"grad_norm": 0.3927195072174072,
"learning_rate": 1.4570979270586944e-05,
"loss": 0.0046,
"step": 415
},
{
"epoch": 0.01664,
"grad_norm": 9.210716247558594,
"learning_rate": 1.4539904997395468e-05,
"loss": 0.2094,
"step": 416
},
{
"epoch": 0.01668,
"grad_norm": 8.319063186645508,
"learning_rate": 1.4508775406894308e-05,
"loss": 0.2885,
"step": 417
},
{
"epoch": 0.01672,
"grad_norm": 6.176471710205078,
"learning_rate": 1.4477590878387697e-05,
"loss": 0.1594,
"step": 418
},
{
"epoch": 0.01676,
"grad_norm": 6.869743347167969,
"learning_rate": 1.4446351791849276e-05,
"loss": 0.1992,
"step": 419
},
{
"epoch": 0.0168,
"grad_norm": 12.559111595153809,
"learning_rate": 1.4415058527917454e-05,
"loss": 0.3767,
"step": 420
},
{
"epoch": 0.01684,
"grad_norm": 7.452301502227783,
"learning_rate": 1.4383711467890776e-05,
"loss": 0.1535,
"step": 421
},
{
"epoch": 0.01688,
"grad_norm": 1.5470104217529297,
"learning_rate": 1.4352310993723277e-05,
"loss": 0.1071,
"step": 422
},
{
"epoch": 0.01692,
"grad_norm": 3.176043748855591,
"learning_rate": 1.4320857488019826e-05,
"loss": 0.154,
"step": 423
},
{
"epoch": 0.01696,
"grad_norm": 1.6062177419662476,
"learning_rate": 1.4289351334031461e-05,
"loss": 0.097,
"step": 424
},
{
"epoch": 0.017,
"grad_norm": 4.27922248840332,
"learning_rate": 1.4257792915650728e-05,
"loss": 0.1833,
"step": 425
},
{
"epoch": 0.01704,
"grad_norm": 7.684791564941406,
"learning_rate": 1.4226182617406996e-05,
"loss": 0.2762,
"step": 426
},
{
"epoch": 0.01708,
"grad_norm": 4.815121650695801,
"learning_rate": 1.4194520824461773e-05,
"loss": 0.1371,
"step": 427
},
{
"epoch": 0.01712,
"grad_norm": 2.0773727893829346,
"learning_rate": 1.4162807922604014e-05,
"loss": 0.1093,
"step": 428
},
{
"epoch": 0.01716,
"grad_norm": 5.84626579284668,
"learning_rate": 1.413104429824542e-05,
"loss": 0.1837,
"step": 429
},
{
"epoch": 0.0172,
"grad_norm": 7.951850414276123,
"learning_rate": 1.4099230338415728e-05,
"loss": 0.3683,
"step": 430
},
{
"epoch": 0.01724,
"grad_norm": 10.412501335144043,
"learning_rate": 1.4067366430758004e-05,
"loss": 0.4045,
"step": 431
},
{
"epoch": 0.01728,
"grad_norm": 5.612797737121582,
"learning_rate": 1.4035452963523903e-05,
"loss": 0.2604,
"step": 432
},
{
"epoch": 0.01732,
"grad_norm": 0.8029801249504089,
"learning_rate": 1.4003490325568953e-05,
"loss": 0.0923,
"step": 433
},
{
"epoch": 0.01736,
"grad_norm": 10.546273231506348,
"learning_rate": 1.3971478906347806e-05,
"loss": 0.2647,
"step": 434
},
{
"epoch": 0.0174,
"grad_norm": 0.4125298261642456,
"learning_rate": 1.3939419095909513e-05,
"loss": 0.0645,
"step": 435
},
{
"epoch": 0.01744,
"grad_norm": 7.981992721557617,
"learning_rate": 1.3907311284892737e-05,
"loss": 0.2246,
"step": 436
},
{
"epoch": 0.01748,
"grad_norm": 10.765192985534668,
"learning_rate": 1.3875155864521031e-05,
"loss": 0.333,
"step": 437
},
{
"epoch": 0.01752,
"grad_norm": 0.6964925527572632,
"learning_rate": 1.3842953226598036e-05,
"loss": 0.0757,
"step": 438
},
{
"epoch": 0.01756,
"grad_norm": 10.305777549743652,
"learning_rate": 1.3810703763502744e-05,
"loss": 0.4982,
"step": 439
},
{
"epoch": 0.0176,
"grad_norm": 5.353335380554199,
"learning_rate": 1.3778407868184674e-05,
"loss": 0.1668,
"step": 440
},
{
"epoch": 0.01764,
"grad_norm": 1.4502415657043457,
"learning_rate": 1.3746065934159123e-05,
"loss": 0.1328,
"step": 441
},
{
"epoch": 0.01768,
"grad_norm": 1.9912700653076172,
"learning_rate": 1.371367835550235e-05,
"loss": 0.1016,
"step": 442
},
{
"epoch": 0.01772,
"grad_norm": 1.4893964529037476,
"learning_rate": 1.3681245526846782e-05,
"loss": 0.1186,
"step": 443
},
{
"epoch": 0.01776,
"grad_norm": 0.9857932925224304,
"learning_rate": 1.3648767843376196e-05,
"loss": 0.0893,
"step": 444
},
{
"epoch": 0.0178,
"grad_norm": 5.2783522605896,
"learning_rate": 1.3616245700820922e-05,
"loss": 0.2204,
"step": 445
},
{
"epoch": 0.01784,
"grad_norm": 2.701751470565796,
"learning_rate": 1.3583679495453e-05,
"loss": 0.0914,
"step": 446
},
{
"epoch": 0.01788,
"grad_norm": 0.8524318933486938,
"learning_rate": 1.3551069624081372e-05,
"loss": 0.0984,
"step": 447
},
{
"epoch": 0.01792,
"grad_norm": 3.844679355621338,
"learning_rate": 1.3518416484047018e-05,
"loss": 0.1487,
"step": 448
},
{
"epoch": 0.01796,
"grad_norm": 10.225765228271484,
"learning_rate": 1.3485720473218153e-05,
"loss": 0.3349,
"step": 449
},
{
"epoch": 0.018,
"grad_norm": 6.975287914276123,
"learning_rate": 1.3452981989985347e-05,
"loss": 0.2389,
"step": 450
},
{
"epoch": 0.01804,
"grad_norm": 1.5389938354492188,
"learning_rate": 1.342020143325669e-05,
"loss": 0.102,
"step": 451
},
{
"epoch": 0.01808,
"grad_norm": 8.041338920593262,
"learning_rate": 1.3387379202452917e-05,
"loss": 0.2631,
"step": 452
},
{
"epoch": 0.01812,
"grad_norm": 5.118985652923584,
"learning_rate": 1.3354515697502552e-05,
"loss": 0.1999,
"step": 453
},
{
"epoch": 0.01816,
"grad_norm": 1.8870747089385986,
"learning_rate": 1.3321611318837033e-05,
"loss": 0.021,
"step": 454
},
{
"epoch": 0.0182,
"grad_norm": 1.1476842164993286,
"learning_rate": 1.3288666467385834e-05,
"loss": 0.0722,
"step": 455
},
{
"epoch": 0.01824,
"grad_norm": 1.8196552991867065,
"learning_rate": 1.3255681544571568e-05,
"loss": 0.1269,
"step": 456
},
{
"epoch": 0.01828,
"grad_norm": 6.326663494110107,
"learning_rate": 1.3222656952305113e-05,
"loss": 0.2738,
"step": 457
},
{
"epoch": 0.01832,
"grad_norm": 7.299874782562256,
"learning_rate": 1.3189593092980701e-05,
"loss": 0.2479,
"step": 458
},
{
"epoch": 0.01836,
"grad_norm": 5.693652629852295,
"learning_rate": 1.3156490369471026e-05,
"loss": 0.2037,
"step": 459
},
{
"epoch": 0.0184,
"grad_norm": 1.2171533107757568,
"learning_rate": 1.3123349185122328e-05,
"loss": 0.0698,
"step": 460
},
{
"epoch": 0.01844,
"grad_norm": 1.0542670488357544,
"learning_rate": 1.3090169943749475e-05,
"loss": 0.018,
"step": 461
},
{
"epoch": 0.01848,
"grad_norm": 10.163932800292969,
"learning_rate": 1.3056953049631059e-05,
"loss": 0.3857,
"step": 462
},
{
"epoch": 0.01852,
"grad_norm": 1.0132485628128052,
"learning_rate": 1.3023698907504447e-05,
"loss": 0.1171,
"step": 463
},
{
"epoch": 0.01856,
"grad_norm": 5.476101398468018,
"learning_rate": 1.2990407922560869e-05,
"loss": 0.1351,
"step": 464
},
{
"epoch": 0.0186,
"grad_norm": 9.09190845489502,
"learning_rate": 1.2957080500440469e-05,
"loss": 0.4134,
"step": 465
},
{
"epoch": 0.01864,
"grad_norm": 3.677666425704956,
"learning_rate": 1.2923717047227368e-05,
"loss": 0.3247,
"step": 466
},
{
"epoch": 0.01868,
"grad_norm": 8.35246467590332,
"learning_rate": 1.2890317969444716e-05,
"loss": 0.2693,
"step": 467
},
{
"epoch": 0.01872,
"grad_norm": 1.9546464681625366,
"learning_rate": 1.2856883674049736e-05,
"loss": 0.1272,
"step": 468
},
{
"epoch": 0.01876,
"grad_norm": 1.7502423524856567,
"learning_rate": 1.2823414568428767e-05,
"loss": 0.1115,
"step": 469
},
{
"epoch": 0.0188,
"grad_norm": 1.2106819152832031,
"learning_rate": 1.2789911060392295e-05,
"loss": 0.0206,
"step": 470
},
{
"epoch": 0.01884,
"grad_norm": 17.259103775024414,
"learning_rate": 1.2756373558169992e-05,
"loss": 0.6014,
"step": 471
},
{
"epoch": 0.01888,
"grad_norm": 2.2297418117523193,
"learning_rate": 1.2722802470405744e-05,
"loss": 0.1324,
"step": 472
},
{
"epoch": 0.01892,
"grad_norm": 0.8471062779426575,
"learning_rate": 1.2689198206152657e-05,
"loss": 0.0748,
"step": 473
},
{
"epoch": 0.01896,
"grad_norm": 1.8467243909835815,
"learning_rate": 1.265556117486809e-05,
"loss": 0.1331,
"step": 474
},
{
"epoch": 0.019,
"grad_norm": 3.141777276992798,
"learning_rate": 1.2621891786408648e-05,
"loss": 0.1546,
"step": 475
},
{
"epoch": 0.01904,
"grad_norm": 0.970522403717041,
"learning_rate": 1.2588190451025209e-05,
"loss": 0.0567,
"step": 476
},
{
"epoch": 0.01908,
"grad_norm": 2.7901699542999268,
"learning_rate": 1.2554457579357906e-05,
"loss": 0.1133,
"step": 477
},
{
"epoch": 0.01912,
"grad_norm": 2.7912991046905518,
"learning_rate": 1.252069358243114e-05,
"loss": 0.1258,
"step": 478
},
{
"epoch": 0.01916,
"grad_norm": 4.593515396118164,
"learning_rate": 1.2486898871648552e-05,
"loss": 0.1626,
"step": 479
},
{
"epoch": 0.0192,
"grad_norm": 3.334031343460083,
"learning_rate": 1.2453073858788027e-05,
"loss": 0.1112,
"step": 480
},
{
"epoch": 0.01924,
"grad_norm": 1.3391278982162476,
"learning_rate": 1.2419218955996677e-05,
"loss": 0.1023,
"step": 481
},
{
"epoch": 0.01928,
"grad_norm": 1.4753631353378296,
"learning_rate": 1.238533457578581e-05,
"loss": 0.1277,
"step": 482
},
{
"epoch": 0.01932,
"grad_norm": 7.0725998878479,
"learning_rate": 1.23514211310259e-05,
"loss": 0.2071,
"step": 483
},
{
"epoch": 0.01936,
"grad_norm": 1.921910285949707,
"learning_rate": 1.2317479034941572e-05,
"loss": 0.0584,
"step": 484
},
{
"epoch": 0.0194,
"grad_norm": 8.189485549926758,
"learning_rate": 1.2283508701106559e-05,
"loss": 0.2548,
"step": 485
},
{
"epoch": 0.01944,
"grad_norm": 11.063411712646484,
"learning_rate": 1.2249510543438652e-05,
"loss": 0.2414,
"step": 486
},
{
"epoch": 0.01948,
"grad_norm": 0.6383995413780212,
"learning_rate": 1.2215484976194675e-05,
"loss": 0.0638,
"step": 487
},
{
"epoch": 0.01952,
"grad_norm": 2.7015726566314697,
"learning_rate": 1.2181432413965428e-05,
"loss": 0.06,
"step": 488
},
{
"epoch": 0.01956,
"grad_norm": 3.063474655151367,
"learning_rate": 1.2147353271670634e-05,
"loss": 0.1085,
"step": 489
},
{
"epoch": 0.0196,
"grad_norm": 4.147983074188232,
"learning_rate": 1.211324796455389e-05,
"loss": 0.1581,
"step": 490
},
{
"epoch": 0.01964,
"grad_norm": 9.769529342651367,
"learning_rate": 1.2079116908177592e-05,
"loss": 0.266,
"step": 491
},
{
"epoch": 0.01968,
"grad_norm": 1.1441335678100586,
"learning_rate": 1.2044960518417902e-05,
"loss": 0.0852,
"step": 492
},
{
"epoch": 0.01972,
"grad_norm": 1.9526094198226929,
"learning_rate": 1.2010779211459649e-05,
"loss": 0.096,
"step": 493
},
{
"epoch": 0.01976,
"grad_norm": 0.789183497428894,
"learning_rate": 1.1976573403791263e-05,
"loss": 0.0588,
"step": 494
},
{
"epoch": 0.0198,
"grad_norm": 8.374926567077637,
"learning_rate": 1.194234351219972e-05,
"loss": 0.1863,
"step": 495
},
{
"epoch": 0.01984,
"grad_norm": 0.7244264483451843,
"learning_rate": 1.190808995376545e-05,
"loss": 0.0689,
"step": 496
},
{
"epoch": 0.01988,
"grad_norm": 5.050563812255859,
"learning_rate": 1.187381314585725e-05,
"loss": 0.1605,
"step": 497
},
{
"epoch": 0.01992,
"grad_norm": 0.8633529543876648,
"learning_rate": 1.1839513506127202e-05,
"loss": 0.0991,
"step": 498
},
{
"epoch": 0.01996,
"grad_norm": 6.967101097106934,
"learning_rate": 1.1805191452505602e-05,
"loss": 0.2348,
"step": 499
},
{
"epoch": 0.02,
"grad_norm": 7.428819179534912,
"learning_rate": 1.1770847403195836e-05,
"loss": 0.1639,
"step": 500
},
{
"epoch": 0.02004,
"grad_norm": 11.58919620513916,
"learning_rate": 1.1736481776669307e-05,
"loss": 0.3558,
"step": 501
},
{
"epoch": 0.02008,
"grad_norm": 8.670886993408203,
"learning_rate": 1.1702094991660326e-05,
"loss": 0.1081,
"step": 502
},
{
"epoch": 0.02012,
"grad_norm": 7.145753860473633,
"learning_rate": 1.1667687467161025e-05,
"loss": 0.2203,
"step": 503
},
{
"epoch": 0.02016,
"grad_norm": 3.4680211544036865,
"learning_rate": 1.1633259622416224e-05,
"loss": 0.1054,
"step": 504
},
{
"epoch": 0.0202,
"grad_norm": 1.2078346014022827,
"learning_rate": 1.159881187691835e-05,
"loss": 0.0897,
"step": 505
},
{
"epoch": 0.02024,
"grad_norm": 4.303440570831299,
"learning_rate": 1.156434465040231e-05,
"loss": 0.0711,
"step": 506
},
{
"epoch": 0.02028,
"grad_norm": 1.6246788501739502,
"learning_rate": 1.1529858362840383e-05,
"loss": 0.1108,
"step": 507
},
{
"epoch": 0.02032,
"grad_norm": 2.3163769245147705,
"learning_rate": 1.1495353434437098e-05,
"loss": 0.1022,
"step": 508
},
{
"epoch": 0.02036,
"grad_norm": 2.608320951461792,
"learning_rate": 1.1460830285624119e-05,
"loss": 0.1068,
"step": 509
},
{
"epoch": 0.0204,
"grad_norm": 5.529253005981445,
"learning_rate": 1.1426289337055119e-05,
"loss": 0.1649,
"step": 510
},
{
"epoch": 0.02044,
"grad_norm": 0.5982191562652588,
"learning_rate": 1.1391731009600655e-05,
"loss": 0.0557,
"step": 511
},
{
"epoch": 0.02048,
"grad_norm": 3.7937324047088623,
"learning_rate": 1.1357155724343046e-05,
"loss": 0.118,
"step": 512
},
{
"epoch": 0.02052,
"grad_norm": 5.356758117675781,
"learning_rate": 1.1322563902571227e-05,
"loss": 0.1571,
"step": 513
},
{
"epoch": 0.02056,
"grad_norm": 1.3954918384552002,
"learning_rate": 1.128795596577563e-05,
"loss": 0.0908,
"step": 514
},
{
"epoch": 0.0206,
"grad_norm": 12.87922191619873,
"learning_rate": 1.1253332335643043e-05,
"loss": 0.5037,
"step": 515
},
{
"epoch": 0.02064,
"grad_norm": 1.6985728740692139,
"learning_rate": 1.1218693434051475e-05,
"loss": 0.1146,
"step": 516
},
{
"epoch": 0.02068,
"grad_norm": 0.6133851408958435,
"learning_rate": 1.1184039683065014e-05,
"loss": 0.0748,
"step": 517
},
{
"epoch": 0.02072,
"grad_norm": 9.497859001159668,
"learning_rate": 1.1149371504928667e-05,
"loss": 0.2467,
"step": 518
},
{
"epoch": 0.02076,
"grad_norm": 4.493142127990723,
"learning_rate": 1.1114689322063255e-05,
"loss": 0.1343,
"step": 519
},
{
"epoch": 0.0208,
"grad_norm": 7.272453308105469,
"learning_rate": 1.1079993557060228e-05,
"loss": 0.1963,
"step": 520
},
{
"epoch": 0.02084,
"grad_norm": 2.106896162033081,
"learning_rate": 1.1045284632676535e-05,
"loss": 0.1459,
"step": 521
},
{
"epoch": 0.02088,
"grad_norm": 1.226660132408142,
"learning_rate": 1.1010562971829464e-05,
"loss": 0.0992,
"step": 522
},
{
"epoch": 0.02092,
"grad_norm": 4.346235275268555,
"learning_rate": 1.0975828997591496e-05,
"loss": 0.1298,
"step": 523
},
{
"epoch": 0.02096,
"grad_norm": 19.246997833251953,
"learning_rate": 1.0941083133185146e-05,
"loss": 0.568,
"step": 524
},
{
"epoch": 0.021,
"grad_norm": 1.2629976272583008,
"learning_rate": 1.0906325801977804e-05,
"loss": 0.1072,
"step": 525
},
{
"epoch": 0.02104,
"grad_norm": 1.5009398460388184,
"learning_rate": 1.0871557427476585e-05,
"loss": 0.0948,
"step": 526
},
{
"epoch": 0.02108,
"grad_norm": 1.3292146921157837,
"learning_rate": 1.083677843332316e-05,
"loss": 0.1274,
"step": 527
},
{
"epoch": 0.02112,
"grad_norm": 1.411028504371643,
"learning_rate": 1.0801989243288588e-05,
"loss": 0.0569,
"step": 528
},
{
"epoch": 0.02116,
"grad_norm": 14.293712615966797,
"learning_rate": 1.0767190281268187e-05,
"loss": 0.3627,
"step": 529
},
{
"epoch": 0.0212,
"grad_norm": 1.669136881828308,
"learning_rate": 1.0732381971276318e-05,
"loss": 0.1217,
"step": 530
},
{
"epoch": 0.02124,
"grad_norm": 4.228145122528076,
"learning_rate": 1.0697564737441254e-05,
"loss": 0.1459,
"step": 531
},
{
"epoch": 0.02128,
"grad_norm": 1.0226049423217773,
"learning_rate": 1.0662739004000005e-05,
"loss": 0.0954,
"step": 532
},
{
"epoch": 0.02132,
"grad_norm": 2.543148994445801,
"learning_rate": 1.0627905195293135e-05,
"loss": 0.1074,
"step": 533
},
{
"epoch": 0.02136,
"grad_norm": 10.598540306091309,
"learning_rate": 1.0593063735759619e-05,
"loss": 0.3346,
"step": 534
},
{
"epoch": 0.0214,
"grad_norm": 0.8847026228904724,
"learning_rate": 1.055821504993164e-05,
"loss": 0.0749,
"step": 535
},
{
"epoch": 0.02144,
"grad_norm": 10.834789276123047,
"learning_rate": 1.0523359562429441e-05,
"loss": 0.3152,
"step": 536
},
{
"epoch": 0.02148,
"grad_norm": 1.7861707210540771,
"learning_rate": 1.0488497697956134e-05,
"loss": 0.1677,
"step": 537
},
{
"epoch": 0.02152,
"grad_norm": 8.30764389038086,
"learning_rate": 1.0453629881292537e-05,
"loss": 0.214,
"step": 538
},
{
"epoch": 0.02156,
"grad_norm": 0.793691873550415,
"learning_rate": 1.0418756537291996e-05,
"loss": 0.0881,
"step": 539
},
{
"epoch": 0.0216,
"grad_norm": 0.6629102826118469,
"learning_rate": 1.03838780908752e-05,
"loss": 0.0828,
"step": 540
},
{
"epoch": 0.02164,
"grad_norm": 9.965836524963379,
"learning_rate": 1.0348994967025012e-05,
"loss": 0.3388,
"step": 541
},
{
"epoch": 0.02168,
"grad_norm": 6.322588920593262,
"learning_rate": 1.0314107590781284e-05,
"loss": 0.1622,
"step": 542
},
{
"epoch": 0.02172,
"grad_norm": 2.410224199295044,
"learning_rate": 1.0279216387235691e-05,
"loss": 0.1012,
"step": 543
},
{
"epoch": 0.02176,
"grad_norm": 2.0198872089385986,
"learning_rate": 1.0244321781526533e-05,
"loss": 0.1105,
"step": 544
},
{
"epoch": 0.0218,
"grad_norm": 0.9930577278137207,
"learning_rate": 1.0209424198833571e-05,
"loss": 0.0915,
"step": 545
},
{
"epoch": 0.02184,
"grad_norm": 9.126015663146973,
"learning_rate": 1.0174524064372837e-05,
"loss": 0.2565,
"step": 546
},
{
"epoch": 0.02188,
"grad_norm": 7.834664821624756,
"learning_rate": 1.0139621803391454e-05,
"loss": 0.1523,
"step": 547
},
{
"epoch": 0.02192,
"grad_norm": 4.842357635498047,
"learning_rate": 1.010471784116246e-05,
"loss": 0.1699,
"step": 548
},
{
"epoch": 0.02196,
"grad_norm": 2.422339916229248,
"learning_rate": 1.0069812602979617e-05,
"loss": 0.0967,
"step": 549
},
{
"epoch": 0.022,
"grad_norm": 1.126199722290039,
"learning_rate": 1.0034906514152239e-05,
"loss": 0.0913,
"step": 550
},
{
"epoch": 0.02204,
"grad_norm": 10.77258586883545,
"learning_rate": 1e-05,
"loss": 0.401,
"step": 551
},
{
"epoch": 0.02208,
"grad_norm": 10.26835823059082,
"learning_rate": 9.965093485847766e-06,
"loss": 0.6585,
"step": 552
},
{
"epoch": 0.02212,
"grad_norm": 1.4840925931930542,
"learning_rate": 9.930187397020385e-06,
"loss": 0.084,
"step": 553
},
{
"epoch": 0.02216,
"grad_norm": 9.270743370056152,
"learning_rate": 9.895282158837545e-06,
"loss": 0.2642,
"step": 554
},
{
"epoch": 0.0222,
"grad_norm": 0.8641135096549988,
"learning_rate": 9.860378196608549e-06,
"loss": 0.092,
"step": 555
},
{
"epoch": 0.02224,
"grad_norm": 6.451079845428467,
"learning_rate": 9.825475935627165e-06,
"loss": 0.1601,
"step": 556
},
{
"epoch": 0.02228,
"grad_norm": 1.1541681289672852,
"learning_rate": 9.790575801166432e-06,
"loss": 0.1084,
"step": 557
},
{
"epoch": 0.02232,
"grad_norm": 5.990635871887207,
"learning_rate": 9.75567821847347e-06,
"loss": 0.1798,
"step": 558
},
{
"epoch": 0.02236,
"grad_norm": 0.6577791571617126,
"learning_rate": 9.720783612764314e-06,
"loss": 0.0601,
"step": 559
},
{
"epoch": 0.0224,
"grad_norm": 1.2313590049743652,
"learning_rate": 9.685892409218718e-06,
"loss": 0.0915,
"step": 560
},
{
"epoch": 0.02244,
"grad_norm": 6.3493266105651855,
"learning_rate": 9.651005032974994e-06,
"loss": 0.1923,
"step": 561
},
{
"epoch": 0.02248,
"grad_norm": 2.7375197410583496,
"learning_rate": 9.616121909124801e-06,
"loss": 0.1315,
"step": 562
},
{
"epoch": 0.02252,
"grad_norm": 8.959554672241211,
"learning_rate": 9.581243462708007e-06,
"loss": 0.2059,
"step": 563
},
{
"epoch": 0.02256,
"grad_norm": 3.0078227519989014,
"learning_rate": 9.546370118707463e-06,
"loss": 0.1254,
"step": 564
},
{
"epoch": 0.0226,
"grad_norm": 6.305171489715576,
"learning_rate": 9.511502302043867e-06,
"loss": 0.1747,
"step": 565
},
{
"epoch": 0.02264,
"grad_norm": 1.763507604598999,
"learning_rate": 9.476640437570562e-06,
"loss": 0.022,
"step": 566
},
{
"epoch": 0.02268,
"grad_norm": 8.843292236328125,
"learning_rate": 9.441784950068362e-06,
"loss": 0.2302,
"step": 567
},
{
"epoch": 0.02272,
"grad_norm": 1.6799709796905518,
"learning_rate": 9.406936264240386e-06,
"loss": 0.0959,
"step": 568
},
{
"epoch": 0.02276,
"grad_norm": 3.8081953525543213,
"learning_rate": 9.372094804706867e-06,
"loss": 0.0505,
"step": 569
},
{
"epoch": 0.0228,
"grad_norm": 11.039603233337402,
"learning_rate": 9.337260996000002e-06,
"loss": 0.2455,
"step": 570
},
{
"epoch": 0.02284,
"grad_norm": 2.4005353450775146,
"learning_rate": 9.302435262558748e-06,
"loss": 0.0818,
"step": 571
},
{
"epoch": 0.02288,
"grad_norm": 0.7494300007820129,
"learning_rate": 9.267618028723687e-06,
"loss": 0.0079,
"step": 572
},
{
"epoch": 0.02292,
"grad_norm": 15.082864761352539,
"learning_rate": 9.232809718731815e-06,
"loss": 0.416,
"step": 573
},
{
"epoch": 0.02296,
"grad_norm": 1.07305908203125,
"learning_rate": 9.198010756711413e-06,
"loss": 0.0868,
"step": 574
},
{
"epoch": 0.023,
"grad_norm": 26.359376907348633,
"learning_rate": 9.163221566676847e-06,
"loss": 0.5973,
"step": 575
},
{
"epoch": 0.02304,
"grad_norm": 1.1738792657852173,
"learning_rate": 9.128442572523418e-06,
"loss": 0.0617,
"step": 576
},
{
"epoch": 0.02308,
"grad_norm": 1.177330732345581,
"learning_rate": 9.093674198022201e-06,
"loss": 0.0946,
"step": 577
},
{
"epoch": 0.02312,
"grad_norm": 12.364805221557617,
"learning_rate": 9.058916866814857e-06,
"loss": 0.338,
"step": 578
},
{
"epoch": 0.02316,
"grad_norm": 5.758530139923096,
"learning_rate": 9.024171002408507e-06,
"loss": 0.142,
"step": 579
},
{
"epoch": 0.0232,
"grad_norm": 9.649211883544922,
"learning_rate": 8.989437028170537e-06,
"loss": 0.2821,
"step": 580
},
{
"epoch": 0.02324,
"grad_norm": 0.7843397855758667,
"learning_rate": 8.954715367323468e-06,
"loss": 0.0734,
"step": 581
},
{
"epoch": 0.02328,
"grad_norm": 1.0275856256484985,
"learning_rate": 8.920006442939772e-06,
"loss": 0.0663,
"step": 582
},
{
"epoch": 0.02332,
"grad_norm": 0.9548624157905579,
"learning_rate": 8.885310677936746e-06,
"loss": 0.0447,
"step": 583
},
{
"epoch": 0.02336,
"grad_norm": 0.07963267713785172,
"learning_rate": 8.850628495071336e-06,
"loss": 0.0009,
"step": 584
},
{
"epoch": 0.0234,
"grad_norm": 7.675271034240723,
"learning_rate": 8.815960316934991e-06,
"loss": 0.1785,
"step": 585
},
{
"epoch": 0.02344,
"grad_norm": 1.1505829095840454,
"learning_rate": 8.781306565948528e-06,
"loss": 0.1221,
"step": 586
},
{
"epoch": 0.02348,
"grad_norm": 1.3688298463821411,
"learning_rate": 8.746667664356957e-06,
"loss": 0.0481,
"step": 587
},
{
"epoch": 0.02352,
"grad_norm": 1.3733991384506226,
"learning_rate": 8.712044034224374e-06,
"loss": 0.0888,
"step": 588
},
{
"epoch": 0.02356,
"grad_norm": 3.525146961212158,
"learning_rate": 8.677436097428775e-06,
"loss": 0.1311,
"step": 589
},
{
"epoch": 0.0236,
"grad_norm": 3.220660448074341,
"learning_rate": 8.642844275656957e-06,
"loss": 0.1336,
"step": 590
},
{
"epoch": 0.02364,
"grad_norm": 2.6903154850006104,
"learning_rate": 8.60826899039935e-06,
"loss": 0.1081,
"step": 591
},
{
"epoch": 0.02368,
"grad_norm": 5.824352264404297,
"learning_rate": 8.573710662944884e-06,
"loss": 0.1411,
"step": 592
},
{
"epoch": 0.02372,
"grad_norm": 2.6695704460144043,
"learning_rate": 8.539169714375885e-06,
"loss": 0.0731,
"step": 593
},
{
"epoch": 0.02376,
"grad_norm": 0.5055465698242188,
"learning_rate": 8.504646565562907e-06,
"loss": 0.0865,
"step": 594
},
{
"epoch": 0.0238,
"grad_norm": 8.769765853881836,
"learning_rate": 8.47014163715962e-06,
"loss": 0.1465,
"step": 595
},
{
"epoch": 0.02384,
"grad_norm": 0.3937555253505707,
"learning_rate": 8.43565534959769e-06,
"loss": 0.0095,
"step": 596
},
{
"epoch": 0.02388,
"grad_norm": 0.722723126411438,
"learning_rate": 8.401188123081653e-06,
"loss": 0.0803,
"step": 597
},
{
"epoch": 0.02392,
"grad_norm": 2.0351316928863525,
"learning_rate": 8.366740377583781e-06,
"loss": 0.0818,
"step": 598
},
{
"epoch": 0.02396,
"grad_norm": 4.883152484893799,
"learning_rate": 8.332312532838978e-06,
"loss": 0.1746,
"step": 599
},
{
"epoch": 0.024,
"grad_norm": 0.7112212181091309,
"learning_rate": 8.297905008339677e-06,
"loss": 0.0585,
"step": 600
},
{
"epoch": 0.02404,
"grad_norm": 13.582132339477539,
"learning_rate": 8.263518223330698e-06,
"loss": 0.4968,
"step": 601
},
{
"epoch": 0.02408,
"grad_norm": 1.8729467391967773,
"learning_rate": 8.22915259680417e-06,
"loss": 0.0765,
"step": 602
},
{
"epoch": 0.02412,
"grad_norm": 8.067035675048828,
"learning_rate": 8.194808547494401e-06,
"loss": 0.1511,
"step": 603
},
{
"epoch": 0.02416,
"grad_norm": 1.1380573511123657,
"learning_rate": 8.1604864938728e-06,
"loss": 0.0719,
"step": 604
},
{
"epoch": 0.0242,
"grad_norm": 1.4015597105026245,
"learning_rate": 8.126186854142752e-06,
"loss": 0.0895,
"step": 605
},
{
"epoch": 0.02424,
"grad_norm": 12.339815139770508,
"learning_rate": 8.091910046234552e-06,
"loss": 0.3199,
"step": 606
},
{
"epoch": 0.02428,
"grad_norm": 10.4027738571167,
"learning_rate": 8.057656487800283e-06,
"loss": 0.1667,
"step": 607
},
{
"epoch": 0.02432,
"grad_norm": 2.0488264560699463,
"learning_rate": 8.023426596208739e-06,
"loss": 0.1091,
"step": 608
},
{
"epoch": 0.02436,
"grad_norm": 0.10466651618480682,
"learning_rate": 7.989220788540356e-06,
"loss": 0.0019,
"step": 609
},
{
"epoch": 0.0244,
"grad_norm": 4.709284782409668,
"learning_rate": 7.955039481582098e-06,
"loss": 0.0946,
"step": 610
},
{
"epoch": 0.02444,
"grad_norm": 1.8861703872680664,
"learning_rate": 7.92088309182241e-06,
"loss": 0.0787,
"step": 611
},
{
"epoch": 0.02448,
"grad_norm": 9.586947441101074,
"learning_rate": 7.886752035446116e-06,
"loss": 0.2047,
"step": 612
},
{
"epoch": 0.02452,
"grad_norm": 1.8871467113494873,
"learning_rate": 7.852646728329368e-06,
"loss": 0.0991,
"step": 613
},
{
"epoch": 0.02456,
"grad_norm": 7.058659076690674,
"learning_rate": 7.818567586034578e-06,
"loss": 0.1542,
"step": 614
},
{
"epoch": 0.0246,
"grad_norm": 7.522183418273926,
"learning_rate": 7.784515023805328e-06,
"loss": 0.0948,
"step": 615
},
{
"epoch": 0.02464,
"grad_norm": 0.7540088891983032,
"learning_rate": 7.750489456561351e-06,
"loss": 0.0869,
"step": 616
},
{
"epoch": 0.02468,
"grad_norm": 0.9133477210998535,
"learning_rate": 7.716491298893443e-06,
"loss": 0.0843,
"step": 617
},
{
"epoch": 0.02472,
"grad_norm": 14.849818229675293,
"learning_rate": 7.68252096505843e-06,
"loss": 0.43,
"step": 618
},
{
"epoch": 0.02476,
"grad_norm": 2.335676908493042,
"learning_rate": 7.6485788689741e-06,
"loss": 0.1183,
"step": 619
},
{
"epoch": 0.0248,
"grad_norm": 0.8231529593467712,
"learning_rate": 7.6146654242141935e-06,
"loss": 0.0619,
"step": 620
},
{
"epoch": 0.02484,
"grad_norm": 8.771549224853516,
"learning_rate": 7.580781044003324e-06,
"loss": 0.2373,
"step": 621
},
{
"epoch": 0.02488,
"grad_norm": 4.117128372192383,
"learning_rate": 7.546926141211975e-06,
"loss": 0.1831,
"step": 622
},
{
"epoch": 0.02492,
"grad_norm": 6.745905876159668,
"learning_rate": 7.513101128351454e-06,
"loss": 0.1943,
"step": 623
},
{
"epoch": 0.02496,
"grad_norm": 1.0095088481903076,
"learning_rate": 7.4793064175688635e-06,
"loss": 0.0843,
"step": 624
},
{
"epoch": 0.025,
"grad_norm": 13.766512870788574,
"learning_rate": 7.445542420642097e-06,
"loss": 0.4766,
"step": 625
},
{
"epoch": 0.02504,
"grad_norm": 5.015426158905029,
"learning_rate": 7.411809548974792e-06,
"loss": 0.1318,
"step": 626
},
{
"epoch": 0.02508,
"grad_norm": 0.746637761592865,
"learning_rate": 7.378108213591355e-06,
"loss": 0.0791,
"step": 627
},
{
"epoch": 0.02512,
"grad_norm": 10.200055122375488,
"learning_rate": 7.344438825131912e-06,
"loss": 0.1492,
"step": 628
},
{
"epoch": 0.02516,
"grad_norm": 1.3007488250732422,
"learning_rate": 7.310801793847344e-06,
"loss": 0.1039,
"step": 629
},
{
"epoch": 0.0252,
"grad_norm": 0.6161604523658752,
"learning_rate": 7.277197529594257e-06,
"loss": 0.0367,
"step": 630
},
{
"epoch": 0.02524,
"grad_norm": 1.0233659744262695,
"learning_rate": 7.243626441830009e-06,
"loss": 0.0744,
"step": 631
},
{
"epoch": 0.02528,
"grad_norm": 0.6330849528312683,
"learning_rate": 7.210088939607709e-06,
"loss": 0.0546,
"step": 632
},
{
"epoch": 0.02532,
"grad_norm": 1.658908486366272,
"learning_rate": 7.176585431571235e-06,
"loss": 0.0921,
"step": 633
},
{
"epoch": 0.02536,
"grad_norm": 3.3533036708831787,
"learning_rate": 7.143116325950266e-06,
"loss": 0.1334,
"step": 634
},
{
"epoch": 0.0254,
"grad_norm": 1.7121665477752686,
"learning_rate": 7.109682030555283e-06,
"loss": 0.0234,
"step": 635
},
{
"epoch": 0.02544,
"grad_norm": 0.5985904932022095,
"learning_rate": 7.076282952772634e-06,
"loss": 0.0289,
"step": 636
},
{
"epoch": 0.02548,
"grad_norm": 15.734526634216309,
"learning_rate": 7.042919499559538e-06,
"loss": 0.1382,
"step": 637
},
{
"epoch": 0.02552,
"grad_norm": 3.260596513748169,
"learning_rate": 7.009592077439135e-06,
"loss": 0.1029,
"step": 638
},
{
"epoch": 0.02556,
"grad_norm": 14.168946266174316,
"learning_rate": 6.976301092495556e-06,
"loss": 0.3536,
"step": 639
},
{
"epoch": 0.0256,
"grad_norm": 0.07166524231433868,
"learning_rate": 6.943046950368944e-06,
"loss": 0.0014,
"step": 640
},
{
"epoch": 0.02564,
"grad_norm": 5.099658966064453,
"learning_rate": 6.909830056250527e-06,
"loss": 0.1522,
"step": 641
},
{
"epoch": 0.02568,
"grad_norm": 9.89889144897461,
"learning_rate": 6.876650814877675e-06,
"loss": 0.1485,
"step": 642
},
{
"epoch": 0.02572,
"grad_norm": 1.1847699880599976,
"learning_rate": 6.843509630528977e-06,
"loss": 0.1075,
"step": 643
},
{
"epoch": 0.02576,
"grad_norm": 3.0199272632598877,
"learning_rate": 6.8104069070193e-06,
"loss": 0.1185,
"step": 644
},
{
"epoch": 0.0258,
"grad_norm": 5.844971656799316,
"learning_rate": 6.777343047694891e-06,
"loss": 0.098,
"step": 645
},
{
"epoch": 0.02584,
"grad_norm": 0.6271488666534424,
"learning_rate": 6.744318455428436e-06,
"loss": 0.0681,
"step": 646
},
{
"epoch": 0.02588,
"grad_norm": 0.0692223459482193,
"learning_rate": 6.711333532614168e-06,
"loss": 0.001,
"step": 647
},
{
"epoch": 0.02592,
"grad_norm": 5.7579803466796875,
"learning_rate": 6.67838868116297e-06,
"loss": 0.2011,
"step": 648
},
{
"epoch": 0.02596,
"grad_norm": 0.9805623888969421,
"learning_rate": 6.645484302497452e-06,
"loss": 0.0754,
"step": 649
},
{
"epoch": 0.026,
"grad_norm": 2.4046757221221924,
"learning_rate": 6.612620797547087e-06,
"loss": 0.1013,
"step": 650
},
{
"epoch": 0.02604,
"grad_norm": 1.556180715560913,
"learning_rate": 6.579798566743314e-06,
"loss": 0.0877,
"step": 651
},
{
"epoch": 0.02608,
"grad_norm": 1.316867470741272,
"learning_rate": 6.547018010014654e-06,
"loss": 0.1152,
"step": 652
},
{
"epoch": 0.02612,
"grad_norm": 8.158949851989746,
"learning_rate": 6.5142795267818505e-06,
"loss": 0.1959,
"step": 653
},
{
"epoch": 0.02616,
"grad_norm": 0.13539642095565796,
"learning_rate": 6.481583515952983e-06,
"loss": 0.0022,
"step": 654
},
{
"epoch": 0.0262,
"grad_norm": 0.39811989665031433,
"learning_rate": 6.448930375918632e-06,
"loss": 0.0048,
"step": 655
},
{
"epoch": 0.02624,
"grad_norm": 1.1902718544006348,
"learning_rate": 6.4163205045469975e-06,
"loss": 0.1139,
"step": 656
},
{
"epoch": 0.02628,
"grad_norm": 17.662755966186523,
"learning_rate": 6.383754299179079e-06,
"loss": 0.7298,
"step": 657
},
{
"epoch": 0.02632,
"grad_norm": 4.676215648651123,
"learning_rate": 6.351232156623803e-06,
"loss": 0.0418,
"step": 658
},
{
"epoch": 0.02636,
"grad_norm": 16.79064178466797,
"learning_rate": 6.318754473153221e-06,
"loss": 0.3631,
"step": 659
},
{
"epoch": 0.0264,
"grad_norm": 0.9962823987007141,
"learning_rate": 6.286321644497655e-06,
"loss": 0.1097,
"step": 660
},
{
"epoch": 0.02644,
"grad_norm": 3.6475906372070312,
"learning_rate": 6.25393406584088e-06,
"loss": 0.1277,
"step": 661
},
{
"epoch": 0.02648,
"grad_norm": 1.1611474752426147,
"learning_rate": 6.22159213181533e-06,
"loss": 0.0698,
"step": 662
},
{
"epoch": 0.02652,
"grad_norm": 8.814127922058105,
"learning_rate": 6.18929623649726e-06,
"loss": 0.158,
"step": 663
},
{
"epoch": 0.02656,
"grad_norm": 17.962182998657227,
"learning_rate": 6.157046773401964e-06,
"loss": 0.4137,
"step": 664
},
{
"epoch": 0.0266,
"grad_norm": 1.3109703063964844,
"learning_rate": 6.124844135478971e-06,
"loss": 0.0932,
"step": 665
},
{
"epoch": 0.02664,
"grad_norm": 8.629732131958008,
"learning_rate": 6.092688715107265e-06,
"loss": 0.1725,
"step": 666
},
{
"epoch": 0.02668,
"grad_norm": 1.6746913194656372,
"learning_rate": 6.06058090409049e-06,
"loss": 0.113,
"step": 667
},
{
"epoch": 0.02672,
"grad_norm": 2.0330495834350586,
"learning_rate": 6.028521093652195e-06,
"loss": 0.0938,
"step": 668
},
{
"epoch": 0.02676,
"grad_norm": 2.7967946529388428,
"learning_rate": 5.996509674431053e-06,
"loss": 0.1244,
"step": 669
},
{
"epoch": 0.0268,
"grad_norm": 14.413750648498535,
"learning_rate": 5.9645470364761e-06,
"loss": 0.2191,
"step": 670
},
{
"epoch": 0.02684,
"grad_norm": 1.1546025276184082,
"learning_rate": 5.932633569242e-06,
"loss": 0.1152,
"step": 671
},
{
"epoch": 0.02688,
"grad_norm": 4.537411689758301,
"learning_rate": 5.900769661584273e-06,
"loss": 0.0362,
"step": 672
},
{
"epoch": 0.02692,
"grad_norm": 0.9595037698745728,
"learning_rate": 5.868955701754584e-06,
"loss": 0.0735,
"step": 673
},
{
"epoch": 0.02696,
"grad_norm": 1.7835566997528076,
"learning_rate": 5.83719207739599e-06,
"loss": 0.1286,
"step": 674
},
{
"epoch": 0.027,
"grad_norm": 1.942439317703247,
"learning_rate": 5.8054791755382286e-06,
"loss": 0.0588,
"step": 675
},
{
"epoch": 0.02704,
"grad_norm": 2.122523069381714,
"learning_rate": 5.773817382593008e-06,
"loss": 0.1106,
"step": 676
},
{
"epoch": 0.02708,
"grad_norm": 4.370386600494385,
"learning_rate": 5.742207084349274e-06,
"loss": 0.1554,
"step": 677
},
{
"epoch": 0.02712,
"grad_norm": 3.195181131362915,
"learning_rate": 5.710648665968543e-06,
"loss": 0.1112,
"step": 678
},
{
"epoch": 0.02716,
"grad_norm": 2.4643239974975586,
"learning_rate": 5.679142511980176e-06,
"loss": 0.0518,
"step": 679
},
{
"epoch": 0.0272,
"grad_norm": 0.5306792855262756,
"learning_rate": 5.647689006276727e-06,
"loss": 0.0515,
"step": 680
},
{
"epoch": 0.02724,
"grad_norm": 13.169659614562988,
"learning_rate": 5.616288532109225e-06,
"loss": 0.2524,
"step": 681
},
{
"epoch": 0.02728,
"grad_norm": 0.46876150369644165,
"learning_rate": 5.584941472082549e-06,
"loss": 0.0587,
"step": 682
},
{
"epoch": 0.02732,
"grad_norm": 7.073272705078125,
"learning_rate": 5.553648208150728e-06,
"loss": 0.1459,
"step": 683
},
{
"epoch": 0.02736,
"grad_norm": 3.0721280574798584,
"learning_rate": 5.522409121612304e-06,
"loss": 0.1116,
"step": 684
},
{
"epoch": 0.0274,
"grad_norm": 13.192761421203613,
"learning_rate": 5.491224593105695e-06,
"loss": 0.2268,
"step": 685
},
{
"epoch": 0.02744,
"grad_norm": 0.04046627879142761,
"learning_rate": 5.460095002604533e-06,
"loss": 0.0009,
"step": 686
},
{
"epoch": 0.02748,
"grad_norm": 1.0833042860031128,
"learning_rate": 5.429020729413062e-06,
"loss": 0.049,
"step": 687
},
{
"epoch": 0.02752,
"grad_norm": 4.5854573249816895,
"learning_rate": 5.398002152161484e-06,
"loss": 0.117,
"step": 688
},
{
"epoch": 0.02756,
"grad_norm": 11.034661293029785,
"learning_rate": 5.367039648801386e-06,
"loss": 0.438,
"step": 689
},
{
"epoch": 0.0276,
"grad_norm": 1.4639097452163696,
"learning_rate": 5.336133596601089e-06,
"loss": 0.0895,
"step": 690
},
{
"epoch": 0.02764,
"grad_norm": 21.207754135131836,
"learning_rate": 5.305284372141095e-06,
"loss": 0.4302,
"step": 691
},
{
"epoch": 0.02768,
"grad_norm": 3.981537342071533,
"learning_rate": 5.274492351309462e-06,
"loss": 0.0905,
"step": 692
},
{
"epoch": 0.02772,
"grad_norm": 0.7105590105056763,
"learning_rate": 5.243757909297247e-06,
"loss": 0.0829,
"step": 693
},
{
"epoch": 0.02776,
"grad_norm": 19.251842498779297,
"learning_rate": 5.213081420593933e-06,
"loss": 0.2563,
"step": 694
},
{
"epoch": 0.0278,
"grad_norm": 1.9615482091903687,
"learning_rate": 5.1824632589828465e-06,
"loss": 0.1217,
"step": 695
},
{
"epoch": 0.02784,
"grad_norm": 3.6855382919311523,
"learning_rate": 5.151903797536631e-06,
"loss": 0.1215,
"step": 696
},
{
"epoch": 0.02788,
"grad_norm": 1.6666749715805054,
"learning_rate": 5.121403408612672e-06,
"loss": 0.0829,
"step": 697
},
{
"epoch": 0.02792,
"grad_norm": 1.03498113155365,
"learning_rate": 5.090962463848592e-06,
"loss": 0.067,
"step": 698
},
{
"epoch": 0.02796,
"grad_norm": 1.6069234609603882,
"learning_rate": 5.060581334157693e-06,
"loss": 0.0304,
"step": 699
},
{
"epoch": 0.028,
"grad_norm": 0.9400602579116821,
"learning_rate": 5.030260389724447e-06,
"loss": 0.0878,
"step": 700
},
{
"epoch": 0.02804,
"grad_norm": 1.3280009031295776,
"learning_rate": 5.000000000000003e-06,
"loss": 0.1502,
"step": 701
},
{
"epoch": 0.02808,
"grad_norm": 0.6366593241691589,
"learning_rate": 4.96980053369765e-06,
"loss": 0.085,
"step": 702
},
{
"epoch": 0.02812,
"grad_norm": 0.1440792828798294,
"learning_rate": 4.939662358788364e-06,
"loss": 0.0009,
"step": 703
},
{
"epoch": 0.02816,
"grad_norm": 0.3521063029766083,
"learning_rate": 4.909585842496287e-06,
"loss": 0.0365,
"step": 704
},
{
"epoch": 0.0282,
"grad_norm": 1.2491483688354492,
"learning_rate": 4.879571351294287e-06,
"loss": 0.118,
"step": 705
},
{
"epoch": 0.02824,
"grad_norm": 12.680964469909668,
"learning_rate": 4.849619250899458e-06,
"loss": 0.3829,
"step": 706
},
{
"epoch": 0.02828,
"grad_norm": 8.474669456481934,
"learning_rate": 4.8197299062687e-06,
"loss": 0.1051,
"step": 707
},
{
"epoch": 0.02832,
"grad_norm": 1.310383677482605,
"learning_rate": 4.78990368159424e-06,
"loss": 0.0809,
"step": 708
},
{
"epoch": 0.02836,
"grad_norm": 1.283028483390808,
"learning_rate": 4.76014094029921e-06,
"loss": 0.1119,
"step": 709
},
{
"epoch": 0.0284,
"grad_norm": 8.952004432678223,
"learning_rate": 4.7304420450332244e-06,
"loss": 0.144,
"step": 710
},
{
"epoch": 0.02844,
"grad_norm": 3.2001149654388428,
"learning_rate": 4.700807357667953e-06,
"loss": 0.0274,
"step": 711
},
{
"epoch": 0.02848,
"grad_norm": 1.8325859308242798,
"learning_rate": 4.671237239292699e-06,
"loss": 0.132,
"step": 712
},
{
"epoch": 0.02852,
"grad_norm": 1.060498833656311,
"learning_rate": 4.641732050210032e-06,
"loss": 0.1,
"step": 713
},
{
"epoch": 0.02856,
"grad_norm": 0.09238772839307785,
"learning_rate": 4.612292149931369e-06,
"loss": 0.0003,
"step": 714
},
{
"epoch": 0.0286,
"grad_norm": 11.677541732788086,
"learning_rate": 4.582917897172603e-06,
"loss": 0.3154,
"step": 715
},
{
"epoch": 0.02864,
"grad_norm": 1.742714524269104,
"learning_rate": 4.5536096498497295e-06,
"loss": 0.1209,
"step": 716
},
{
"epoch": 0.02868,
"grad_norm": 0.9469143748283386,
"learning_rate": 4.524367765074499e-06,
"loss": 0.063,
"step": 717
},
{
"epoch": 0.02872,
"grad_norm": 5.26619291305542,
"learning_rate": 4.495192599150045e-06,
"loss": 0.1045,
"step": 718
},
{
"epoch": 0.02876,
"grad_norm": 7.335450172424316,
"learning_rate": 4.46608450756656e-06,
"loss": 0.1781,
"step": 719
},
{
"epoch": 0.0288,
"grad_norm": 11.088786125183105,
"learning_rate": 4.437043844996952e-06,
"loss": 0.2475,
"step": 720
},
{
"epoch": 0.02884,
"grad_norm": 1.0706833600997925,
"learning_rate": 4.408070965292534e-06,
"loss": 0.0953,
"step": 721
},
{
"epoch": 0.02888,
"grad_norm": 0.7746567726135254,
"learning_rate": 4.379166221478697e-06,
"loss": 0.0098,
"step": 722
},
{
"epoch": 0.02892,
"grad_norm": 8.533845901489258,
"learning_rate": 4.350329965750622e-06,
"loss": 0.2352,
"step": 723
},
{
"epoch": 0.02896,
"grad_norm": 4.435492038726807,
"learning_rate": 4.321562549468991e-06,
"loss": 0.1459,
"step": 724
},
{
"epoch": 0.029,
"grad_norm": 4.611571788787842,
"learning_rate": 4.292864323155684e-06,
"loss": 0.1128,
"step": 725
},
{
"epoch": 0.02904,
"grad_norm": 0.7998501658439636,
"learning_rate": 4.264235636489542e-06,
"loss": 0.1025,
"step": 726
},
{
"epoch": 0.02908,
"grad_norm": 1.5966432094573975,
"learning_rate": 4.235676838302069e-06,
"loss": 0.0706,
"step": 727
},
{
"epoch": 0.02912,
"grad_norm": 0.37493979930877686,
"learning_rate": 4.207188276573214e-06,
"loss": 0.0653,
"step": 728
},
{
"epoch": 0.02916,
"grad_norm": 0.6811991333961487,
"learning_rate": 4.178770298427107e-06,
"loss": 0.065,
"step": 729
},
{
"epoch": 0.0292,
"grad_norm": 8.22618293762207,
"learning_rate": 4.150423250127846e-06,
"loss": 0.1721,
"step": 730
},
{
"epoch": 0.02924,
"grad_norm": 2.755448579788208,
"learning_rate": 4.12214747707527e-06,
"loss": 0.0349,
"step": 731
},
{
"epoch": 0.02928,
"grad_norm": 7.533814907073975,
"learning_rate": 4.093943323800746e-06,
"loss": 0.2147,
"step": 732
},
{
"epoch": 0.02932,
"grad_norm": 4.506439685821533,
"learning_rate": 4.065811133962987e-06,
"loss": 0.1272,
"step": 733
},
{
"epoch": 0.02936,
"grad_norm": 9.940001487731934,
"learning_rate": 4.037751250343841e-06,
"loss": 0.2193,
"step": 734
},
{
"epoch": 0.0294,
"grad_norm": 0.7324904203414917,
"learning_rate": 4.009764014844143e-06,
"loss": 0.0632,
"step": 735
},
{
"epoch": 0.02944,
"grad_norm": 3.1775712966918945,
"learning_rate": 3.981849768479516e-06,
"loss": 0.0777,
"step": 736
},
{
"epoch": 0.02948,
"grad_norm": 0.762550950050354,
"learning_rate": 3.954008851376252e-06,
"loss": 0.0869,
"step": 737
},
{
"epoch": 0.02952,
"grad_norm": 9.46602725982666,
"learning_rate": 3.9262416027671354e-06,
"loss": 0.2908,
"step": 738
},
{
"epoch": 0.02956,
"grad_norm": 0.6857216358184814,
"learning_rate": 3.898548360987325e-06,
"loss": 0.0058,
"step": 739
},
{
"epoch": 0.0296,
"grad_norm": 5.8143696784973145,
"learning_rate": 3.8709294634702374e-06,
"loss": 0.1393,
"step": 740
},
{
"epoch": 0.02964,
"grad_norm": 0.8191357254981995,
"learning_rate": 3.8433852467434175e-06,
"loss": 0.0633,
"step": 741
},
{
"epoch": 0.02968,
"grad_norm": 4.689929962158203,
"learning_rate": 3.81591604642446e-06,
"loss": 0.1243,
"step": 742
},
{
"epoch": 0.02972,
"grad_norm": 11.390048027038574,
"learning_rate": 3.7885221972168974e-06,
"loss": 0.243,
"step": 743
},
{
"epoch": 0.02976,
"grad_norm": 4.229440212249756,
"learning_rate": 3.7612040329061405e-06,
"loss": 0.1009,
"step": 744
},
{
"epoch": 0.0298,
"grad_norm": 0.7295369505882263,
"learning_rate": 3.7339618863553983e-06,
"loss": 0.0927,
"step": 745
},
{
"epoch": 0.02984,
"grad_norm": 0.6325824856758118,
"learning_rate": 3.7067960895016277e-06,
"loss": 0.0559,
"step": 746
},
{
"epoch": 0.02988,
"grad_norm": 1.3186429738998413,
"learning_rate": 3.679706973351491e-06,
"loss": 0.0779,
"step": 747
},
{
"epoch": 0.02992,
"grad_norm": 5.386847972869873,
"learning_rate": 3.6526948679773256e-06,
"loss": 0.1398,
"step": 748
},
{
"epoch": 0.02996,
"grad_norm": 8.078812599182129,
"learning_rate": 3.625760102513103e-06,
"loss": 0.1664,
"step": 749
},
{
"epoch": 0.03,
"grad_norm": 4.324909210205078,
"learning_rate": 3.598903005150444e-06,
"loss": 0.2225,
"step": 750
},
{
"epoch": 0.03004,
"grad_norm": 0.5693206191062927,
"learning_rate": 3.5721239031346067e-06,
"loss": 0.0514,
"step": 751
},
{
"epoch": 0.03008,
"grad_norm": 1.579349160194397,
"learning_rate": 3.545423122760493e-06,
"loss": 0.1226,
"step": 752
},
{
"epoch": 0.03012,
"grad_norm": 1.4756170511245728,
"learning_rate": 3.5188009893686916e-06,
"loss": 0.1039,
"step": 753
},
{
"epoch": 0.03016,
"grad_norm": 1.485909104347229,
"learning_rate": 3.492257827341492e-06,
"loss": 0.0981,
"step": 754
},
{
"epoch": 0.0302,
"grad_norm": 0.7757060527801514,
"learning_rate": 3.4657939600989453e-06,
"loss": 0.088,
"step": 755
},
{
"epoch": 0.03024,
"grad_norm": 3.61979341506958,
"learning_rate": 3.4394097100949286e-06,
"loss": 0.1343,
"step": 756
},
{
"epoch": 0.03028,
"grad_norm": 4.138033390045166,
"learning_rate": 3.4131053988131947e-06,
"loss": 0.1116,
"step": 757
},
{
"epoch": 0.03032,
"grad_norm": 7.502678871154785,
"learning_rate": 3.3868813467634833e-06,
"loss": 0.1567,
"step": 758
},
{
"epoch": 0.03036,
"grad_norm": 1.0160504579544067,
"learning_rate": 3.360737873477584e-06,
"loss": 0.0824,
"step": 759
},
{
"epoch": 0.0304,
"grad_norm": 1.0232679843902588,
"learning_rate": 3.3346752975054763e-06,
"loss": 0.0852,
"step": 760
},
{
"epoch": 0.03044,
"grad_norm": 27.406278610229492,
"learning_rate": 3.308693936411421e-06,
"loss": 0.2068,
"step": 761
},
{
"epoch": 0.03048,
"grad_norm": 0.7658149600028992,
"learning_rate": 3.2827941067700996e-06,
"loss": 0.0823,
"step": 762
},
{
"epoch": 0.03052,
"grad_norm": 4.264142036437988,
"learning_rate": 3.2569761241627694e-06,
"loss": 0.1138,
"step": 763
},
{
"epoch": 0.03056,
"grad_norm": 0.46639883518218994,
"learning_rate": 3.2312403031733943e-06,
"loss": 0.045,
"step": 764
},
{
"epoch": 0.0306,
"grad_norm": 1.2367812395095825,
"learning_rate": 3.2055869573848374e-06,
"loss": 0.0785,
"step": 765
},
{
"epoch": 0.03064,
"grad_norm": 1.1481434106826782,
"learning_rate": 3.1800163993750166e-06,
"loss": 0.1012,
"step": 766
},
{
"epoch": 0.03068,
"grad_norm": 3.7116329669952393,
"learning_rate": 3.1545289407131128e-06,
"loss": 0.108,
"step": 767
},
{
"epoch": 0.03072,
"grad_norm": 1.4448208808898926,
"learning_rate": 3.1291248919557717e-06,
"loss": 0.0805,
"step": 768
},
{
"epoch": 0.03076,
"grad_norm": 1.178962230682373,
"learning_rate": 3.103804562643302e-06,
"loss": 0.1038,
"step": 769
},
{
"epoch": 0.0308,
"grad_norm": 3.160212993621826,
"learning_rate": 3.0785682612959334e-06,
"loss": 0.1154,
"step": 770
},
{
"epoch": 0.03084,
"grad_norm": 0.7983316779136658,
"learning_rate": 3.0534162954100264e-06,
"loss": 0.0742,
"step": 771
},
{
"epoch": 0.03088,
"grad_norm": 2.7964160442352295,
"learning_rate": 3.028348971454356e-06,
"loss": 0.0953,
"step": 772
},
{
"epoch": 0.03092,
"grad_norm": 1.5415724515914917,
"learning_rate": 3.003366594866345e-06,
"loss": 0.0728,
"step": 773
},
{
"epoch": 0.03096,
"grad_norm": 1.2726086378097534,
"learning_rate": 2.978469470048376e-06,
"loss": 0.0844,
"step": 774
},
{
"epoch": 0.031,
"grad_norm": 0.5856465101242065,
"learning_rate": 2.953657900364053e-06,
"loss": 0.0801,
"step": 775
},
{
"epoch": 0.03104,
"grad_norm": 1.6102033853530884,
"learning_rate": 2.9289321881345257e-06,
"loss": 0.0824,
"step": 776
},
{
"epoch": 0.03108,
"grad_norm": 0.9553947448730469,
"learning_rate": 2.9042926346347932e-06,
"loss": 0.0449,
"step": 777
},
{
"epoch": 0.03112,
"grad_norm": 7.113488674163818,
"learning_rate": 2.8797395400900362e-06,
"loss": 0.1516,
"step": 778
},
{
"epoch": 0.03116,
"grad_norm": 0.42258957028388977,
"learning_rate": 2.855273203671969e-06,
"loss": 0.0339,
"step": 779
},
{
"epoch": 0.0312,
"grad_norm": 1.1574292182922363,
"learning_rate": 2.830893923495173e-06,
"loss": 0.1051,
"step": 780
},
{
"epoch": 0.03124,
"grad_norm": 5.842410564422607,
"learning_rate": 2.8066019966134907e-06,
"loss": 0.1398,
"step": 781
},
{
"epoch": 0.03128,
"grad_norm": 0.2941426932811737,
"learning_rate": 2.7823977190163788e-06,
"loss": 0.0033,
"step": 782
},
{
"epoch": 0.03132,
"grad_norm": 7.133516311645508,
"learning_rate": 2.7582813856253276e-06,
"loss": 0.1564,
"step": 783
},
{
"epoch": 0.03136,
"grad_norm": 4.029329776763916,
"learning_rate": 2.7342532902902418e-06,
"loss": 0.1201,
"step": 784
},
{
"epoch": 0.0314,
"grad_norm": 0.6362094283103943,
"learning_rate": 2.7103137257858867e-06,
"loss": 0.0784,
"step": 785
},
{
"epoch": 0.03144,
"grad_norm": 2.0200414657592773,
"learning_rate": 2.6864629838082957e-06,
"loss": 0.1227,
"step": 786
},
{
"epoch": 0.03148,
"grad_norm": 9.439911842346191,
"learning_rate": 2.6627013549712355e-06,
"loss": 0.1405,
"step": 787
},
{
"epoch": 0.03152,
"grad_norm": 6.89825439453125,
"learning_rate": 2.639029128802657e-06,
"loss": 0.1839,
"step": 788
},
{
"epoch": 0.03156,
"grad_norm": 2.551635503768921,
"learning_rate": 2.615446593741161e-06,
"loss": 0.1176,
"step": 789
},
{
"epoch": 0.0316,
"grad_norm": 4.661273002624512,
"learning_rate": 2.5919540371325005e-06,
"loss": 0.1188,
"step": 790
},
{
"epoch": 0.03164,
"grad_norm": 1.7053903341293335,
"learning_rate": 2.5685517452260566e-06,
"loss": 0.1292,
"step": 791
},
{
"epoch": 0.03168,
"grad_norm": 1.1583722829818726,
"learning_rate": 2.5452400031713786e-06,
"loss": 0.0697,
"step": 792
},
{
"epoch": 0.03172,
"grad_norm": 15.050021171569824,
"learning_rate": 2.522019095014683e-06,
"loss": 0.3301,
"step": 793
},
{
"epoch": 0.03176,
"grad_norm": 0.8170905113220215,
"learning_rate": 2.4988893036954045e-06,
"loss": 0.07,
"step": 794
},
{
"epoch": 0.0318,
"grad_norm": 0.44838589429855347,
"learning_rate": 2.4758509110427576e-06,
"loss": 0.0478,
"step": 795
},
{
"epoch": 0.03184,
"grad_norm": 0.7005017995834351,
"learning_rate": 2.45290419777228e-06,
"loss": 0.0649,
"step": 796
},
{
"epoch": 0.03188,
"grad_norm": 6.050832271575928,
"learning_rate": 2.4300494434824373e-06,
"loss": 0.1514,
"step": 797
},
{
"epoch": 0.03192,
"grad_norm": 5.461383819580078,
"learning_rate": 2.407286926651192e-06,
"loss": 0.1264,
"step": 798
},
{
"epoch": 0.03196,
"grad_norm": 1.6498404741287231,
"learning_rate": 2.3846169246326345e-06,
"loss": 0.0822,
"step": 799
},
{
"epoch": 0.032,
"grad_norm": 4.357858180999756,
"learning_rate": 2.362039713653581e-06,
"loss": 0.1021,
"step": 800
},
{
"epoch": 0.03204,
"grad_norm": 15.552661895751953,
"learning_rate": 2.339555568810221e-06,
"loss": 0.6585,
"step": 801
},
{
"epoch": 0.03208,
"grad_norm": 1.0037691593170166,
"learning_rate": 2.317164764064769e-06,
"loss": 0.1092,
"step": 802
},
{
"epoch": 0.03212,
"grad_norm": 1.530813217163086,
"learning_rate": 2.2948675722421086e-06,
"loss": 0.0157,
"step": 803
},
{
"epoch": 0.03216,
"grad_norm": 1.1743625402450562,
"learning_rate": 2.27266426502649e-06,
"loss": 0.0967,
"step": 804
},
{
"epoch": 0.0322,
"grad_norm": 4.888083457946777,
"learning_rate": 2.2505551129582047e-06,
"loss": 0.1417,
"step": 805
},
{
"epoch": 0.03224,
"grad_norm": 16.029897689819336,
"learning_rate": 2.2285403854302912e-06,
"loss": 0.4153,
"step": 806
},
{
"epoch": 0.03228,
"grad_norm": 1.5436164140701294,
"learning_rate": 2.206620350685257e-06,
"loss": 0.0725,
"step": 807
},
{
"epoch": 0.03232,
"grad_norm": 6.853541374206543,
"learning_rate": 2.1847952758118118e-06,
"loss": 0.1056,
"step": 808
},
{
"epoch": 0.03236,
"grad_norm": 1.7744520902633667,
"learning_rate": 2.163065426741603e-06,
"loss": 0.1047,
"step": 809
},
{
"epoch": 0.0324,
"grad_norm": 0.15737532079219818,
"learning_rate": 2.1414310682459805e-06,
"loss": 0.0021,
"step": 810
},
{
"epoch": 0.03244,
"grad_norm": 0.9227983355522156,
"learning_rate": 2.119892463932781e-06,
"loss": 0.0769,
"step": 811
},
{
"epoch": 0.03248,
"grad_norm": 0.08678217977285385,
"learning_rate": 2.098449876243096e-06,
"loss": 0.0015,
"step": 812
},
{
"epoch": 0.03252,
"grad_norm": 1.087418556213379,
"learning_rate": 2.0771035664480944e-06,
"loss": 0.0105,
"step": 813
},
{
"epoch": 0.03256,
"grad_norm": 4.054409503936768,
"learning_rate": 2.0558537946458177e-06,
"loss": 0.0838,
"step": 814
},
{
"epoch": 0.0326,
"grad_norm": 0.9547904133796692,
"learning_rate": 2.0347008197580376e-06,
"loss": 0.1007,
"step": 815
},
{
"epoch": 0.03264,
"grad_norm": 1.0841913223266602,
"learning_rate": 2.013644899527074e-06,
"loss": 0.0748,
"step": 816
},
{
"epoch": 0.03268,
"grad_norm": 4.705662727355957,
"learning_rate": 1.9926862905126663e-06,
"loss": 0.1135,
"step": 817
},
{
"epoch": 0.03272,
"grad_norm": 0.5187758207321167,
"learning_rate": 1.9718252480888567e-06,
"loss": 0.0501,
"step": 818
},
{
"epoch": 0.03276,
"grad_norm": 7.7622294425964355,
"learning_rate": 1.95106202644086e-06,
"loss": 0.3285,
"step": 819
},
{
"epoch": 0.0328,
"grad_norm": 0.38402506709098816,
"learning_rate": 1.930396878561983e-06,
"loss": 0.0677,
"step": 820
},
{
"epoch": 0.03284,
"grad_norm": 4.81564998626709,
"learning_rate": 1.9098300562505266e-06,
"loss": 0.1483,
"step": 821
},
{
"epoch": 0.03288,
"grad_norm": 7.7534565925598145,
"learning_rate": 1.8893618101067357e-06,
"loss": 0.2352,
"step": 822
},
{
"epoch": 0.03292,
"grad_norm": 0.7130910158157349,
"learning_rate": 1.8689923895297247e-06,
"loss": 0.0993,
"step": 823
},
{
"epoch": 0.03296,
"grad_norm": 2.1166458129882812,
"learning_rate": 1.848722042714457e-06,
"loss": 0.0703,
"step": 824
},
{
"epoch": 0.033,
"grad_norm": 0.7550241351127625,
"learning_rate": 1.8285510166487154e-06,
"loss": 0.0663,
"step": 825
},
{
"epoch": 0.03304,
"grad_norm": 19.060657501220703,
"learning_rate": 1.808479557110081e-06,
"loss": 0.1482,
"step": 826
},
{
"epoch": 0.03308,
"grad_norm": 1.1805157661437988,
"learning_rate": 1.7885079086629598e-06,
"loss": 0.1234,
"step": 827
},
{
"epoch": 0.03312,
"grad_norm": 7.956192493438721,
"learning_rate": 1.7686363146555807e-06,
"loss": 0.1578,
"step": 828
},
{
"epoch": 0.03316,
"grad_norm": 1.6199045181274414,
"learning_rate": 1.7488650172170496e-06,
"loss": 0.0723,
"step": 829
},
{
"epoch": 0.0332,
"grad_norm": 2.616152763366699,
"learning_rate": 1.7291942572543806e-06,
"loss": 0.1063,
"step": 830
},
{
"epoch": 0.03324,
"grad_norm": 1.3882237672805786,
"learning_rate": 1.709624274449584e-06,
"loss": 0.0148,
"step": 831
},
{
"epoch": 0.03328,
"grad_norm": 1.7266179323196411,
"learning_rate": 1.6901553072567189e-06,
"loss": 0.0899,
"step": 832
},
{
"epoch": 0.03332,
"grad_norm": 1.3904311656951904,
"learning_rate": 1.6707875928990059e-06,
"loss": 0.1024,
"step": 833
},
{
"epoch": 0.03336,
"grad_norm": 5.440541744232178,
"learning_rate": 1.651521367365936e-06,
"loss": 0.1339,
"step": 834
},
{
"epoch": 0.0334,
"grad_norm": 8.045305252075195,
"learning_rate": 1.6323568654103838e-06,
"loss": 0.0584,
"step": 835
},
{
"epoch": 0.03344,
"grad_norm": 0.8960226774215698,
"learning_rate": 1.6132943205457607e-06,
"loss": 0.0888,
"step": 836
},
{
"epoch": 0.03348,
"grad_norm": 0.42670223116874695,
"learning_rate": 1.5943339650431578e-06,
"loss": 0.0326,
"step": 837
},
{
"epoch": 0.03352,
"grad_norm": 5.909890651702881,
"learning_rate": 1.5754760299285255e-06,
"loss": 0.1434,
"step": 838
},
{
"epoch": 0.03356,
"grad_norm": 0.8122259378433228,
"learning_rate": 1.5567207449798517e-06,
"loss": 0.0754,
"step": 839
},
{
"epoch": 0.0336,
"grad_norm": 15.18900203704834,
"learning_rate": 1.538068338724361e-06,
"loss": 0.4247,
"step": 840
},
{
"epoch": 0.03364,
"grad_norm": 7.332209587097168,
"learning_rate": 1.5195190384357405e-06,
"loss": 0.1517,
"step": 841
},
{
"epoch": 0.03368,
"grad_norm": 5.032659530639648,
"learning_rate": 1.5010730701313626e-06,
"loss": 0.101,
"step": 842
},
{
"epoch": 0.03372,
"grad_norm": 0.8301838636398315,
"learning_rate": 1.4827306585695234e-06,
"loss": 0.0741,
"step": 843
},
{
"epoch": 0.03376,
"grad_norm": 1.060888648033142,
"learning_rate": 1.4644920272467245e-06,
"loss": 0.0613,
"step": 844
},
{
"epoch": 0.0338,
"grad_norm": 1.6646232604980469,
"learning_rate": 1.446357398394934e-06,
"loss": 0.0493,
"step": 845
},
{
"epoch": 0.03384,
"grad_norm": 0.7017843723297119,
"learning_rate": 1.4283269929788779e-06,
"loss": 0.0781,
"step": 846
},
{
"epoch": 0.03388,
"grad_norm": 0.9491502046585083,
"learning_rate": 1.4104010306933558e-06,
"loss": 0.1041,
"step": 847
},
{
"epoch": 0.03392,
"grad_norm": 0.5201511383056641,
"learning_rate": 1.3925797299605649e-06,
"loss": 0.0486,
"step": 848
},
{
"epoch": 0.03396,
"grad_norm": 3.370574951171875,
"learning_rate": 1.3748633079274254e-06,
"loss": 0.1074,
"step": 849
},
{
"epoch": 0.034,
"grad_norm": 0.7470880150794983,
"learning_rate": 1.3572519804629537e-06,
"loss": 0.0629,
"step": 850
},
{
"epoch": 0.03404,
"grad_norm": 1.4284977912902832,
"learning_rate": 1.339745962155613e-06,
"loss": 0.0748,
"step": 851
},
{
"epoch": 0.03408,
"grad_norm": 0.7702114582061768,
"learning_rate": 1.322345466310717e-06,
"loss": 0.0601,
"step": 852
},
{
"epoch": 0.03412,
"grad_norm": 9.254502296447754,
"learning_rate": 1.30505070494781e-06,
"loss": 0.2391,
"step": 853
},
{
"epoch": 0.03416,
"grad_norm": 1.3638499975204468,
"learning_rate": 1.2878618887981064e-06,
"loss": 0.0946,
"step": 854
},
{
"epoch": 0.0342,
"grad_norm": 7.902167320251465,
"learning_rate": 1.2707792273019049e-06,
"loss": 0.2335,
"step": 855
},
{
"epoch": 0.03424,
"grad_norm": 0.5616679191589355,
"learning_rate": 1.2538029286060428e-06,
"loss": 0.0722,
"step": 856
},
{
"epoch": 0.03428,
"grad_norm": 0.8284444212913513,
"learning_rate": 1.2369331995613664e-06,
"loss": 0.0853,
"step": 857
},
{
"epoch": 0.03432,
"grad_norm": 0.4530278742313385,
"learning_rate": 1.2201702457201948e-06,
"loss": 0.0529,
"step": 858
},
{
"epoch": 0.03436,
"grad_norm": 0.7437230944633484,
"learning_rate": 1.2035142713338366e-06,
"loss": 0.0644,
"step": 859
},
{
"epoch": 0.0344,
"grad_norm": 2.391450881958008,
"learning_rate": 1.1869654793500784e-06,
"loss": 0.0902,
"step": 860
},
{
"epoch": 0.03444,
"grad_norm": 3.210803747177124,
"learning_rate": 1.1705240714107301e-06,
"loss": 0.0445,
"step": 861
},
{
"epoch": 0.03448,
"grad_norm": 20.66632652282715,
"learning_rate": 1.1541902478491607e-06,
"loss": 0.3445,
"step": 862
},
{
"epoch": 0.03452,
"grad_norm": 19.06394386291504,
"learning_rate": 1.1379642076878528e-06,
"loss": 0.3957,
"step": 863
},
{
"epoch": 0.03456,
"grad_norm": 2.497035264968872,
"learning_rate": 1.1218461486359878e-06,
"loss": 0.133,
"step": 864
},
{
"epoch": 0.0346,
"grad_norm": 1.5445897579193115,
"learning_rate": 1.1058362670870248e-06,
"loss": 0.09,
"step": 865
},
{
"epoch": 0.03464,
"grad_norm": 9.944012641906738,
"learning_rate": 1.0899347581163222e-06,
"loss": 0.1585,
"step": 866
},
{
"epoch": 0.03468,
"grad_norm": 1.9828206300735474,
"learning_rate": 1.0741418154787443e-06,
"loss": 0.093,
"step": 867
},
{
"epoch": 0.03472,
"grad_norm": 0.5986301302909851,
"learning_rate": 1.058457631606319e-06,
"loss": 0.0741,
"step": 868
},
{
"epoch": 0.03476,
"grad_norm": 1.3600716590881348,
"learning_rate": 1.042882397605871e-06,
"loss": 0.0607,
"step": 869
},
{
"epoch": 0.0348,
"grad_norm": 0.6756635904312134,
"learning_rate": 1.0274163032567165e-06,
"loss": 0.0521,
"step": 870
},
{
"epoch": 0.03484,
"grad_norm": 0.1648545265197754,
"learning_rate": 1.012059537008332e-06,
"loss": 0.0008,
"step": 871
},
{
"epoch": 0.03488,
"grad_norm": 1.113710641860962,
"learning_rate": 9.968122859780648e-07,
"loss": 0.0816,
"step": 872
},
{
"epoch": 0.03492,
"grad_norm": 8.441840171813965,
"learning_rate": 9.816747359488632e-07,
"loss": 0.1579,
"step": 873
},
{
"epoch": 0.03496,
"grad_norm": 0.8907444477081299,
"learning_rate": 9.666470713669918e-07,
"loss": 0.0706,
"step": 874
},
{
"epoch": 0.035,
"grad_norm": 1.0123701095581055,
"learning_rate": 9.517294753398066e-07,
"loss": 0.0159,
"step": 875
},
{
"epoch": 0.03504,
"grad_norm": 1.3048282861709595,
"learning_rate": 9.369221296335007e-07,
"loss": 0.0659,
"step": 876
},
{
"epoch": 0.03508,
"grad_norm": 0.9157322645187378,
"learning_rate": 9.222252146709143e-07,
"loss": 0.1019,
"step": 877
},
{
"epoch": 0.03512,
"grad_norm": 5.971161365509033,
"learning_rate": 9.076389095293148e-07,
"loss": 0.2539,
"step": 878
},
{
"epoch": 0.03516,
"grad_norm": 7.711389064788818,
"learning_rate": 8.931633919382299e-07,
"loss": 0.1518,
"step": 879
},
{
"epoch": 0.0352,
"grad_norm": 2.8643109798431396,
"learning_rate": 8.787988382772705e-07,
"loss": 0.0743,
"step": 880
},
{
"epoch": 0.03524,
"grad_norm": 0.962252676486969,
"learning_rate": 8.645454235739903e-07,
"loss": 0.0891,
"step": 881
},
{
"epoch": 0.03528,
"grad_norm": 16.95307731628418,
"learning_rate": 8.504033215017527e-07,
"loss": 0.2249,
"step": 882
},
{
"epoch": 0.03532,
"grad_norm": 1.3776546716690063,
"learning_rate": 8.363727043776037e-07,
"loss": 0.0751,
"step": 883
},
{
"epoch": 0.03536,
"grad_norm": 9.943365097045898,
"learning_rate": 8.224537431601886e-07,
"loss": 0.1206,
"step": 884
},
{
"epoch": 0.0354,
"grad_norm": 11.410050392150879,
"learning_rate": 8.086466074476562e-07,
"loss": 0.19,
"step": 885
},
{
"epoch": 0.03544,
"grad_norm": 1.0532925128936768,
"learning_rate": 7.949514654755963e-07,
"loss": 0.0638,
"step": 886
},
{
"epoch": 0.03548,
"grad_norm": 1.5347234010696411,
"learning_rate": 7.81368484114996e-07,
"loss": 0.1067,
"step": 887
},
{
"epoch": 0.03552,
"grad_norm": 0.5733872056007385,
"learning_rate": 7.678978288701911e-07,
"loss": 0.0662,
"step": 888
},
{
"epoch": 0.03556,
"grad_norm": 4.4785661697387695,
"learning_rate": 7.545396638768698e-07,
"loss": 0.141,
"step": 889
},
{
"epoch": 0.0356,
"grad_norm": 1.1501100063323975,
"learning_rate": 7.412941519000527e-07,
"loss": 0.0631,
"step": 890
},
{
"epoch": 0.03564,
"grad_norm": 0.8512274622917175,
"learning_rate": 7.281614543321269e-07,
"loss": 0.094,
"step": 891
},
{
"epoch": 0.03568,
"grad_norm": 1.5823286771774292,
"learning_rate": 7.151417311908648e-07,
"loss": 0.0918,
"step": 892
},
{
"epoch": 0.03572,
"grad_norm": 1.838431477546692,
"learning_rate": 7.022351411174866e-07,
"loss": 0.0831,
"step": 893
},
{
"epoch": 0.03576,
"grad_norm": 0.5003334283828735,
"learning_rate": 6.894418413747183e-07,
"loss": 0.0683,
"step": 894
},
{
"epoch": 0.0358,
"grad_norm": 1.2628241777420044,
"learning_rate": 6.767619878448783e-07,
"loss": 0.0447,
"step": 895
},
{
"epoch": 0.03584,
"grad_norm": 0.48141220211982727,
"learning_rate": 6.641957350279838e-07,
"loss": 0.0395,
"step": 896
},
{
"epoch": 0.03588,
"grad_norm": 1.504714012145996,
"learning_rate": 6.517432360398556e-07,
"loss": 0.0944,
"step": 897
},
{
"epoch": 0.03592,
"grad_norm": 1.8488521575927734,
"learning_rate": 6.394046426102673e-07,
"loss": 0.0779,
"step": 898
},
{
"epoch": 0.03596,
"grad_norm": 8.644623756408691,
"learning_rate": 6.271801050810856e-07,
"loss": 0.1898,
"step": 899
},
{
"epoch": 0.036,
"grad_norm": 1.2239041328430176,
"learning_rate": 6.150697724044407e-07,
"loss": 0.0992,
"step": 900
},
{
"epoch": 0.03604,
"grad_norm": 1.6652538776397705,
"learning_rate": 6.030737921409169e-07,
"loss": 0.0747,
"step": 901
},
{
"epoch": 0.03608,
"grad_norm": 1.8840757608413696,
"learning_rate": 5.911923104577455e-07,
"loss": 0.108,
"step": 902
},
{
"epoch": 0.03612,
"grad_norm": 1.5210579633712769,
"learning_rate": 5.794254721270331e-07,
"loss": 0.1155,
"step": 903
},
{
"epoch": 0.03616,
"grad_norm": 0.69416344165802,
"learning_rate": 5.677734205239904e-07,
"loss": 0.0577,
"step": 904
},
{
"epoch": 0.0362,
"grad_norm": 1.242936372756958,
"learning_rate": 5.562362976251901e-07,
"loss": 0.0783,
"step": 905
},
{
"epoch": 0.03624,
"grad_norm": 4.353142738342285,
"learning_rate": 5.448142440068316e-07,
"loss": 0.1161,
"step": 906
},
{
"epoch": 0.03628,
"grad_norm": 1.2351945638656616,
"learning_rate": 5.335073988430373e-07,
"loss": 0.0691,
"step": 907
},
{
"epoch": 0.03632,
"grad_norm": 2.496166706085205,
"learning_rate": 5.223158999041444e-07,
"loss": 0.0967,
"step": 908
},
{
"epoch": 0.03636,
"grad_norm": 0.8021566867828369,
"learning_rate": 5.112398835550348e-07,
"loss": 0.085,
"step": 909
},
{
"epoch": 0.0364,
"grad_norm": 1.1814556121826172,
"learning_rate": 5.002794847534765e-07,
"loss": 0.077,
"step": 910
},
{
"epoch": 0.03644,
"grad_norm": 1.4265468120574951,
"learning_rate": 4.894348370484648e-07,
"loss": 0.0969,
"step": 911
},
{
"epoch": 0.03648,
"grad_norm": 1.2824418544769287,
"learning_rate": 4.787060725786141e-07,
"loss": 0.1004,
"step": 912
},
{
"epoch": 0.03652,
"grad_norm": 1.06431245803833,
"learning_rate": 4.6809332207053083e-07,
"loss": 0.0673,
"step": 913
},
{
"epoch": 0.03656,
"grad_norm": 2.5703158378601074,
"learning_rate": 4.575967148372318e-07,
"loss": 0.0883,
"step": 914
},
{
"epoch": 0.0366,
"grad_norm": 1.1215368509292603,
"learning_rate": 4.4721637877656377e-07,
"loss": 0.0141,
"step": 915
},
{
"epoch": 0.03664,
"grad_norm": 1.490781307220459,
"learning_rate": 4.3695244036964567e-07,
"loss": 0.0916,
"step": 916
},
{
"epoch": 0.03668,
"grad_norm": 1.1962960958480835,
"learning_rate": 4.268050246793276e-07,
"loss": 0.1029,
"step": 917
},
{
"epoch": 0.03672,
"grad_norm": 0.25713813304901123,
"learning_rate": 4.167742553486676e-07,
"loss": 0.001,
"step": 918
},
{
"epoch": 0.03676,
"grad_norm": 1.092668890953064,
"learning_rate": 4.068602545994249e-07,
"loss": 0.1002,
"step": 919
},
{
"epoch": 0.0368,
"grad_norm": 0.9069550037384033,
"learning_rate": 3.9706314323056936e-07,
"loss": 0.0609,
"step": 920
},
{
"epoch": 0.03684,
"grad_norm": 2.2207298278808594,
"learning_rate": 3.8738304061681107e-07,
"loss": 0.1173,
"step": 921
},
{
"epoch": 0.03688,
"grad_norm": 0.8923022150993347,
"learning_rate": 3.7782006470714614e-07,
"loss": 0.0852,
"step": 922
},
{
"epoch": 0.03692,
"grad_norm": 0.5463240146636963,
"learning_rate": 3.68374332023419e-07,
"loss": 0.0621,
"step": 923
},
{
"epoch": 0.03696,
"grad_norm": 16.21967124938965,
"learning_rate": 3.590459576589e-07,
"loss": 0.1394,
"step": 924
},
{
"epoch": 0.037,
"grad_norm": 0.7936307787895203,
"learning_rate": 3.498350552768859e-07,
"loss": 0.0726,
"step": 925
},
{
"epoch": 0.03704,
"grad_norm": 0.4272739589214325,
"learning_rate": 3.4074173710931804e-07,
"loss": 0.0325,
"step": 926
},
{
"epoch": 0.03708,
"grad_norm": 4.404773235321045,
"learning_rate": 3.3176611395540625e-07,
"loss": 0.0672,
"step": 927
},
{
"epoch": 0.03712,
"grad_norm": 0.842936098575592,
"learning_rate": 3.2290829518028867e-07,
"loss": 0.084,
"step": 928
},
{
"epoch": 0.03716,
"grad_norm": 0.5097290277481079,
"learning_rate": 3.1416838871368925e-07,
"loss": 0.061,
"step": 929
},
{
"epoch": 0.0372,
"grad_norm": 2.257333517074585,
"learning_rate": 3.0554650104861137e-07,
"loss": 0.0942,
"step": 930
},
{
"epoch": 0.03724,
"grad_norm": 5.119732856750488,
"learning_rate": 2.970427372400353e-07,
"loss": 0.0565,
"step": 931
},
{
"epoch": 0.03728,
"grad_norm": 12.634729385375977,
"learning_rate": 2.8865720090364037e-07,
"loss": 0.2283,
"step": 932
},
{
"epoch": 0.03732,
"grad_norm": 1.8135788440704346,
"learning_rate": 2.8038999421453827e-07,
"loss": 0.1045,
"step": 933
},
{
"epoch": 0.03736,
"grad_norm": 9.891157150268555,
"learning_rate": 2.7224121790603517e-07,
"loss": 0.0543,
"step": 934
},
{
"epoch": 0.0374,
"grad_norm": 2.102281332015991,
"learning_rate": 2.6421097126839714e-07,
"loss": 0.0141,
"step": 935
},
{
"epoch": 0.03744,
"grad_norm": 11.036881446838379,
"learning_rate": 2.5629935214764866e-07,
"loss": 0.2305,
"step": 936
},
{
"epoch": 0.03748,
"grad_norm": 3.1347999572753906,
"learning_rate": 2.4850645694436736e-07,
"loss": 0.1497,
"step": 937
},
{
"epoch": 0.03752,
"grad_norm": 0.5600032210350037,
"learning_rate": 2.4083238061252565e-07,
"loss": 0.0328,
"step": 938
},
{
"epoch": 0.03756,
"grad_norm": 0.13085992634296417,
"learning_rate": 2.332772166583208e-07,
"loss": 0.0011,
"step": 939
},
{
"epoch": 0.0376,
"grad_norm": 0.8646296858787537,
"learning_rate": 2.2584105713904126e-07,
"loss": 0.0599,
"step": 940
},
{
"epoch": 0.03764,
"grad_norm": 0.9335747361183167,
"learning_rate": 2.1852399266194312e-07,
"loss": 0.0623,
"step": 941
},
{
"epoch": 0.03768,
"grad_norm": 1.0552181005477905,
"learning_rate": 2.1132611238315004e-07,
"loss": 0.0799,
"step": 942
},
{
"epoch": 0.03772,
"grad_norm": 14.19265365600586,
"learning_rate": 2.0424750400655947e-07,
"loss": 0.3822,
"step": 943
},
{
"epoch": 0.03776,
"grad_norm": 10.433021545410156,
"learning_rate": 1.9728825378278248e-07,
"loss": 0.1455,
"step": 944
},
{
"epoch": 0.0378,
"grad_norm": 1.599608063697815,
"learning_rate": 1.9044844650808468e-07,
"loss": 0.1132,
"step": 945
},
{
"epoch": 0.03784,
"grad_norm": 1.117100477218628,
"learning_rate": 1.8372816552336025e-07,
"loss": 0.0874,
"step": 946
},
{
"epoch": 0.03788,
"grad_norm": 0.5097063183784485,
"learning_rate": 1.7712749271311392e-07,
"loss": 0.0097,
"step": 947
},
{
"epoch": 0.03792,
"grad_norm": 0.9047200083732605,
"learning_rate": 1.706465085044584e-07,
"loss": 0.1129,
"step": 948
},
{
"epoch": 0.03796,
"grad_norm": 1.0145995616912842,
"learning_rate": 1.6428529186614195e-07,
"loss": 0.0704,
"step": 949
},
{
"epoch": 0.038,
"grad_norm": 7.475941181182861,
"learning_rate": 1.580439203075812e-07,
"loss": 0.0802,
"step": 950
},
{
"epoch": 0.03804,
"grad_norm": 7.637581825256348,
"learning_rate": 1.519224698779198e-07,
"loss": 0.1485,
"step": 951
},
{
"epoch": 0.03808,
"grad_norm": 0.5519956350326538,
"learning_rate": 1.4592101516509916e-07,
"loss": 0.05,
"step": 952
},
{
"epoch": 0.03812,
"grad_norm": 11.45514965057373,
"learning_rate": 1.400396292949513e-07,
"loss": 0.1382,
"step": 953
},
{
"epoch": 0.03816,
"grad_norm": 5.387925148010254,
"learning_rate": 1.3427838393030634e-07,
"loss": 0.1025,
"step": 954
},
{
"epoch": 0.0382,
"grad_norm": 3.8086771965026855,
"learning_rate": 1.2863734927012094e-07,
"loss": 0.0838,
"step": 955
},
{
"epoch": 0.03824,
"grad_norm": 6.967965602874756,
"learning_rate": 1.231165940486234e-07,
"loss": 0.0585,
"step": 956
},
{
"epoch": 0.03828,
"grad_norm": 6.010990142822266,
"learning_rate": 1.1771618553447217e-07,
"loss": 0.1598,
"step": 957
},
{
"epoch": 0.03832,
"grad_norm": 0.637279748916626,
"learning_rate": 1.1243618952994195e-07,
"loss": 0.0642,
"step": 958
},
{
"epoch": 0.03836,
"grad_norm": 0.9717149138450623,
"learning_rate": 1.0727667037011668e-07,
"loss": 0.0826,
"step": 959
},
{
"epoch": 0.0384,
"grad_norm": 7.8545403480529785,
"learning_rate": 1.0223769092211012e-07,
"loss": 0.17,
"step": 960
},
{
"epoch": 0.03844,
"grad_norm": 0.8411500453948975,
"learning_rate": 9.731931258429638e-08,
"loss": 0.0874,
"step": 961
},
{
"epoch": 0.03848,
"grad_norm": 2.80416202545166,
"learning_rate": 9.252159528556404e-08,
"loss": 0.1147,
"step": 962
},
{
"epoch": 0.03852,
"grad_norm": 0.8255104422569275,
"learning_rate": 8.784459748458318e-08,
"loss": 0.0718,
"step": 963
},
{
"epoch": 0.03856,
"grad_norm": 0.5604941844940186,
"learning_rate": 8.328837616909612e-08,
"loss": 0.0061,
"step": 964
},
{
"epoch": 0.0386,
"grad_norm": 1.7548247575759888,
"learning_rate": 7.885298685522235e-08,
"loss": 0.0752,
"step": 965
},
{
"epoch": 0.03864,
"grad_norm": 1.4195278882980347,
"learning_rate": 7.453848358678018e-08,
"loss": 0.0828,
"step": 966
},
{
"epoch": 0.03868,
"grad_norm": 8.356043815612793,
"learning_rate": 7.034491893463059e-08,
"loss": 0.2693,
"step": 967
},
{
"epoch": 0.03872,
"grad_norm": 1.3332304954528809,
"learning_rate": 6.627234399603554e-08,
"loss": 0.0984,
"step": 968
},
{
"epoch": 0.03876,
"grad_norm": 0.6685734987258911,
"learning_rate": 6.232080839403631e-08,
"loss": 0.0328,
"step": 969
},
{
"epoch": 0.0388,
"grad_norm": 2.5308942794799805,
"learning_rate": 5.849036027684607e-08,
"loss": 0.0815,
"step": 970
},
{
"epoch": 0.03884,
"grad_norm": 3.929035186767578,
"learning_rate": 5.4781046317267103e-08,
"loss": 0.0841,
"step": 971
},
{
"epoch": 0.03888,
"grad_norm": 0.42925241589546204,
"learning_rate": 5.119291171211793e-08,
"loss": 0.05,
"step": 972
},
{
"epoch": 0.03892,
"grad_norm": 2.9308767318725586,
"learning_rate": 4.772600018168816e-08,
"loss": 0.1107,
"step": 973
},
{
"epoch": 0.03896,
"grad_norm": 1.3381779193878174,
"learning_rate": 4.438035396920004e-08,
"loss": 0.0121,
"step": 974
},
{
"epoch": 0.039,
"grad_norm": 0.9549940228462219,
"learning_rate": 4.115601384029666e-08,
"loss": 0.1026,
"step": 975
},
{
"epoch": 0.03904,
"grad_norm": 3.592839479446411,
"learning_rate": 3.805301908254455e-08,
"loss": 0.1116,
"step": 976
},
{
"epoch": 0.03908,
"grad_norm": 12.554839134216309,
"learning_rate": 3.50714075049563e-08,
"loss": 0.2162,
"step": 977
},
{
"epoch": 0.03912,
"grad_norm": 0.7680001854896545,
"learning_rate": 3.22112154375287e-08,
"loss": 0.0836,
"step": 978
},
{
"epoch": 0.03916,
"grad_norm": 0.2087625116109848,
"learning_rate": 2.947247773079753e-08,
"loss": 0.0006,
"step": 979
},
{
"epoch": 0.0392,
"grad_norm": 0.7369223237037659,
"learning_rate": 2.6855227755419046e-08,
"loss": 0.0913,
"step": 980
},
{
"epoch": 0.03924,
"grad_norm": 0.8005729913711548,
"learning_rate": 2.4359497401758026e-08,
"loss": 0.1034,
"step": 981
},
{
"epoch": 0.03928,
"grad_norm": 0.6573168635368347,
"learning_rate": 2.1985317079500358e-08,
"loss": 0.0723,
"step": 982
},
{
"epoch": 0.03932,
"grad_norm": 0.8846660256385803,
"learning_rate": 1.973271571728441e-08,
"loss": 0.0848,
"step": 983
},
{
"epoch": 0.03936,
"grad_norm": 0.8547536730766296,
"learning_rate": 1.7601720762346895e-08,
"loss": 0.0764,
"step": 984
},
{
"epoch": 0.0394,
"grad_norm": 5.158422470092773,
"learning_rate": 1.5592358180189782e-08,
"loss": 0.0911,
"step": 985
},
{
"epoch": 0.03944,
"grad_norm": 3.070943832397461,
"learning_rate": 1.370465245426167e-08,
"loss": 0.099,
"step": 986
},
{
"epoch": 0.03948,
"grad_norm": 0.8975648880004883,
"learning_rate": 1.1938626585660252e-08,
"loss": 0.0676,
"step": 987
},
{
"epoch": 0.03952,
"grad_norm": 0.8555331826210022,
"learning_rate": 1.0294302092853647e-08,
"loss": 0.0793,
"step": 988
},
{
"epoch": 0.03956,
"grad_norm": 4.783902645111084,
"learning_rate": 8.771699011416169e-09,
"loss": 0.1303,
"step": 989
},
{
"epoch": 0.0396,
"grad_norm": 0.45209425687789917,
"learning_rate": 7.370835893788508e-09,
"loss": 0.0379,
"step": 990
},
{
"epoch": 0.03964,
"grad_norm": 3.3422908782958984,
"learning_rate": 6.091729809042379e-09,
"loss": 0.0372,
"step": 991
},
{
"epoch": 0.03968,
"grad_norm": 1.3098170757293701,
"learning_rate": 4.9343963426840006e-09,
"loss": 0.103,
"step": 992
},
{
"epoch": 0.03972,
"grad_norm": 1.089207410812378,
"learning_rate": 3.898849596456477e-09,
"loss": 0.0101,
"step": 993
},
{
"epoch": 0.03976,
"grad_norm": 0.5473456382751465,
"learning_rate": 2.9851021881688314e-09,
"loss": 0.0568,
"step": 994
},
{
"epoch": 0.0398,
"grad_norm": 4.233986854553223,
"learning_rate": 2.193165251545004e-09,
"loss": 0.1283,
"step": 995
},
{
"epoch": 0.03984,
"grad_norm": 2.9097344875335693,
"learning_rate": 1.5230484360873043e-09,
"loss": 0.0755,
"step": 996
},
{
"epoch": 0.03988,
"grad_norm": 3.9271740913391113,
"learning_rate": 9.74759906957612e-10,
"loss": 0.0473,
"step": 997
},
{
"epoch": 0.03992,
"grad_norm": 5.4770121574401855,
"learning_rate": 5.483063448785686e-10,
"loss": 0.1066,
"step": 998
},
{
"epoch": 0.03996,
"grad_norm": 10.889819145202637,
"learning_rate": 2.436929460525317e-10,
"loss": 0.133,
"step": 999
},
{
"epoch": 0.04,
"grad_norm": 2.4631078243255615,
"learning_rate": 6.092342209607083e-11,
"loss": 0.0875,
"step": 1000
}
],
"logging_steps": 1,
"max_steps": 1000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}