amburger66's picture
LoRA fine-tune on RobotSmith task07 tool bowl_lifter_modified2
6b28267 verified
Raw
History Blame Contribute Delete
175 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.03902439024390244,
"eval_steps": 50,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 3.902439024390244e-05,
"grad_norm": 17.488399505615234,
"learning_rate": 0.0,
"loss": 1.226,
"step": 1
},
{
"epoch": 7.804878048780487e-05,
"grad_norm": 6.759295463562012,
"learning_rate": 2.0000000000000002e-07,
"loss": 0.7835,
"step": 2
},
{
"epoch": 0.00011707317073170732,
"grad_norm": 6.967640399932861,
"learning_rate": 4.0000000000000003e-07,
"loss": 0.2753,
"step": 3
},
{
"epoch": 0.00015609756097560975,
"grad_norm": 14.539920806884766,
"learning_rate": 6.000000000000001e-07,
"loss": 1.4435,
"step": 4
},
{
"epoch": 0.0001951219512195122,
"grad_norm": 15.56311321258545,
"learning_rate": 8.000000000000001e-07,
"loss": 1.7648,
"step": 5
},
{
"epoch": 0.00023414634146341464,
"grad_norm": 3.013420820236206,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.1059,
"step": 6
},
{
"epoch": 0.00027317073170731707,
"grad_norm": 5.426943302154541,
"learning_rate": 1.2000000000000002e-06,
"loss": 0.3127,
"step": 7
},
{
"epoch": 0.0003121951219512195,
"grad_norm": 13.369946479797363,
"learning_rate": 1.4000000000000001e-06,
"loss": 2.0195,
"step": 8
},
{
"epoch": 0.0003512195121951219,
"grad_norm": 18.31296730041504,
"learning_rate": 1.6000000000000001e-06,
"loss": 1.4981,
"step": 9
},
{
"epoch": 0.0003902439024390244,
"grad_norm": 16.417490005493164,
"learning_rate": 1.8000000000000001e-06,
"loss": 1.171,
"step": 10
},
{
"epoch": 0.00042926829268292684,
"grad_norm": 6.436070442199707,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.336,
"step": 11
},
{
"epoch": 0.00046829268292682927,
"grad_norm": 14.296679496765137,
"learning_rate": 2.2e-06,
"loss": 1.6998,
"step": 12
},
{
"epoch": 0.0005073170731707317,
"grad_norm": 2.883645534515381,
"learning_rate": 2.4000000000000003e-06,
"loss": 0.2077,
"step": 13
},
{
"epoch": 0.0005463414634146341,
"grad_norm": 11.857547760009766,
"learning_rate": 2.6e-06,
"loss": 1.1927,
"step": 14
},
{
"epoch": 0.0005853658536585366,
"grad_norm": 8.715965270996094,
"learning_rate": 2.8000000000000003e-06,
"loss": 0.8397,
"step": 15
},
{
"epoch": 0.000624390243902439,
"grad_norm": 14.193154335021973,
"learning_rate": 3e-06,
"loss": 1.3847,
"step": 16
},
{
"epoch": 0.0006634146341463414,
"grad_norm": 20.103727340698242,
"learning_rate": 3.2000000000000003e-06,
"loss": 1.1781,
"step": 17
},
{
"epoch": 0.0007024390243902439,
"grad_norm": 19.597455978393555,
"learning_rate": 3.4000000000000005e-06,
"loss": 2.2673,
"step": 18
},
{
"epoch": 0.0007414634146341464,
"grad_norm": 7.589250564575195,
"learning_rate": 3.6000000000000003e-06,
"loss": 0.7986,
"step": 19
},
{
"epoch": 0.0007804878048780488,
"grad_norm": 8.664912223815918,
"learning_rate": 3.8000000000000005e-06,
"loss": 0.8342,
"step": 20
},
{
"epoch": 0.0008195121951219513,
"grad_norm": 20.323787689208984,
"learning_rate": 4.000000000000001e-06,
"loss": 1.7093,
"step": 21
},
{
"epoch": 0.0008585365853658537,
"grad_norm": 12.499859809875488,
"learning_rate": 4.2000000000000004e-06,
"loss": 1.07,
"step": 22
},
{
"epoch": 0.0008975609756097561,
"grad_norm": 10.358566284179688,
"learning_rate": 4.4e-06,
"loss": 0.5502,
"step": 23
},
{
"epoch": 0.0009365853658536585,
"grad_norm": 13.93114948272705,
"learning_rate": 4.600000000000001e-06,
"loss": 1.2819,
"step": 24
},
{
"epoch": 0.000975609756097561,
"grad_norm": 8.86787223815918,
"learning_rate": 4.800000000000001e-06,
"loss": 1.1975,
"step": 25
},
{
"epoch": 0.0010146341463414634,
"grad_norm": 7.953192710876465,
"learning_rate": 5e-06,
"loss": 0.329,
"step": 26
},
{
"epoch": 0.0010536585365853658,
"grad_norm": 4.725555419921875,
"learning_rate": 5.2e-06,
"loss": 0.1339,
"step": 27
},
{
"epoch": 0.0010926829268292683,
"grad_norm": 6.697360038757324,
"learning_rate": 5.400000000000001e-06,
"loss": 0.4998,
"step": 28
},
{
"epoch": 0.0011317073170731707,
"grad_norm": 12.212544441223145,
"learning_rate": 5.600000000000001e-06,
"loss": 0.4897,
"step": 29
},
{
"epoch": 0.0011707317073170731,
"grad_norm": 2.965939521789551,
"learning_rate": 5.8e-06,
"loss": 0.1951,
"step": 30
},
{
"epoch": 0.0012097560975609756,
"grad_norm": 9.593793869018555,
"learning_rate": 6e-06,
"loss": 0.6437,
"step": 31
},
{
"epoch": 0.001248780487804878,
"grad_norm": 9.148179054260254,
"learning_rate": 6.200000000000001e-06,
"loss": 0.9349,
"step": 32
},
{
"epoch": 0.0012878048780487804,
"grad_norm": 7.672140121459961,
"learning_rate": 6.4000000000000006e-06,
"loss": 0.6605,
"step": 33
},
{
"epoch": 0.0013268292682926828,
"grad_norm": 12.947052955627441,
"learning_rate": 6.600000000000001e-06,
"loss": 0.8352,
"step": 34
},
{
"epoch": 0.0013658536585365853,
"grad_norm": 10.349727630615234,
"learning_rate": 6.800000000000001e-06,
"loss": 0.8713,
"step": 35
},
{
"epoch": 0.0014048780487804877,
"grad_norm": 9.037214279174805,
"learning_rate": 7e-06,
"loss": 1.0487,
"step": 36
},
{
"epoch": 0.0014439024390243901,
"grad_norm": 9.46027660369873,
"learning_rate": 7.2000000000000005e-06,
"loss": 0.8676,
"step": 37
},
{
"epoch": 0.0014829268292682928,
"grad_norm": 8.622385025024414,
"learning_rate": 7.4e-06,
"loss": 0.4687,
"step": 38
},
{
"epoch": 0.0015219512195121952,
"grad_norm": 13.641250610351562,
"learning_rate": 7.600000000000001e-06,
"loss": 1.1929,
"step": 39
},
{
"epoch": 0.0015609756097560976,
"grad_norm": 12.384421348571777,
"learning_rate": 7.800000000000002e-06,
"loss": 0.7009,
"step": 40
},
{
"epoch": 0.0016,
"grad_norm": 3.4686012268066406,
"learning_rate": 8.000000000000001e-06,
"loss": 0.2169,
"step": 41
},
{
"epoch": 0.0016390243902439025,
"grad_norm": 8.398360252380371,
"learning_rate": 8.2e-06,
"loss": 0.7008,
"step": 42
},
{
"epoch": 0.001678048780487805,
"grad_norm": 4.720148086547852,
"learning_rate": 8.400000000000001e-06,
"loss": 0.3086,
"step": 43
},
{
"epoch": 0.0017170731707317074,
"grad_norm": 3.698089599609375,
"learning_rate": 8.6e-06,
"loss": 0.151,
"step": 44
},
{
"epoch": 0.0017560975609756098,
"grad_norm": 10.639657974243164,
"learning_rate": 8.8e-06,
"loss": 0.6401,
"step": 45
},
{
"epoch": 0.0017951219512195122,
"grad_norm": 1.8676412105560303,
"learning_rate": 9e-06,
"loss": 0.1222,
"step": 46
},
{
"epoch": 0.0018341463414634147,
"grad_norm": 5.136429786682129,
"learning_rate": 9.200000000000002e-06,
"loss": 0.52,
"step": 47
},
{
"epoch": 0.001873170731707317,
"grad_norm": 8.609760284423828,
"learning_rate": 9.4e-06,
"loss": 0.4534,
"step": 48
},
{
"epoch": 0.0019121951219512195,
"grad_norm": 12.459165573120117,
"learning_rate": 9.600000000000001e-06,
"loss": 0.6504,
"step": 49
},
{
"epoch": 0.001951219512195122,
"grad_norm": 7.804759979248047,
"learning_rate": 9.800000000000001e-06,
"loss": 0.6401,
"step": 50
},
{
"epoch": 0.0019902439024390246,
"grad_norm": 7.838076591491699,
"learning_rate": 1e-05,
"loss": 0.5352,
"step": 51
},
{
"epoch": 0.002029268292682927,
"grad_norm": 3.41353702545166,
"learning_rate": 1.02e-05,
"loss": 0.3086,
"step": 52
},
{
"epoch": 0.0020682926829268295,
"grad_norm": 2.9028987884521484,
"learning_rate": 1.04e-05,
"loss": 0.2201,
"step": 53
},
{
"epoch": 0.0021073170731707317,
"grad_norm": 5.654747486114502,
"learning_rate": 1.0600000000000002e-05,
"loss": 0.4777,
"step": 54
},
{
"epoch": 0.0021463414634146343,
"grad_norm": 7.974294662475586,
"learning_rate": 1.0800000000000002e-05,
"loss": 0.7391,
"step": 55
},
{
"epoch": 0.0021853658536585365,
"grad_norm": 3.1305031776428223,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.246,
"step": 56
},
{
"epoch": 0.002224390243902439,
"grad_norm": 5.731029033660889,
"learning_rate": 1.1200000000000001e-05,
"loss": 0.4327,
"step": 57
},
{
"epoch": 0.0022634146341463414,
"grad_norm": 1.9597899913787842,
"learning_rate": 1.14e-05,
"loss": 0.0765,
"step": 58
},
{
"epoch": 0.002302439024390244,
"grad_norm": 5.895092010498047,
"learning_rate": 1.16e-05,
"loss": 0.4113,
"step": 59
},
{
"epoch": 0.0023414634146341463,
"grad_norm": 3.9769833087921143,
"learning_rate": 1.18e-05,
"loss": 0.2684,
"step": 60
},
{
"epoch": 0.002380487804878049,
"grad_norm": 3.3503031730651855,
"learning_rate": 1.2e-05,
"loss": 0.1582,
"step": 61
},
{
"epoch": 0.002419512195121951,
"grad_norm": 4.7539381980896,
"learning_rate": 1.22e-05,
"loss": 0.3716,
"step": 62
},
{
"epoch": 0.0024585365853658538,
"grad_norm": 5.641966819763184,
"learning_rate": 1.2400000000000002e-05,
"loss": 0.44,
"step": 63
},
{
"epoch": 0.002497560975609756,
"grad_norm": 3.8150923252105713,
"learning_rate": 1.2600000000000001e-05,
"loss": 0.2021,
"step": 64
},
{
"epoch": 0.0025365853658536586,
"grad_norm": 1.833798885345459,
"learning_rate": 1.2800000000000001e-05,
"loss": 0.1273,
"step": 65
},
{
"epoch": 0.002575609756097561,
"grad_norm": 3.668488025665283,
"learning_rate": 1.3000000000000001e-05,
"loss": 0.2446,
"step": 66
},
{
"epoch": 0.0026146341463414635,
"grad_norm": 5.197065830230713,
"learning_rate": 1.3200000000000002e-05,
"loss": 0.2406,
"step": 67
},
{
"epoch": 0.0026536585365853657,
"grad_norm": 7.472379207611084,
"learning_rate": 1.3400000000000002e-05,
"loss": 0.4046,
"step": 68
},
{
"epoch": 0.0026926829268292683,
"grad_norm": 5.102828502655029,
"learning_rate": 1.3600000000000002e-05,
"loss": 0.1701,
"step": 69
},
{
"epoch": 0.0027317073170731706,
"grad_norm": 7.045306205749512,
"learning_rate": 1.38e-05,
"loss": 0.8455,
"step": 70
},
{
"epoch": 0.002770731707317073,
"grad_norm": 1.8381656408309937,
"learning_rate": 1.4e-05,
"loss": 0.1474,
"step": 71
},
{
"epoch": 0.0028097560975609754,
"grad_norm": 4.51398229598999,
"learning_rate": 1.4200000000000001e-05,
"loss": 0.3397,
"step": 72
},
{
"epoch": 0.002848780487804878,
"grad_norm": 4.925454616546631,
"learning_rate": 1.4400000000000001e-05,
"loss": 0.3386,
"step": 73
},
{
"epoch": 0.0028878048780487803,
"grad_norm": 7.905612468719482,
"learning_rate": 1.46e-05,
"loss": 0.2774,
"step": 74
},
{
"epoch": 0.002926829268292683,
"grad_norm": 6.0955023765563965,
"learning_rate": 1.48e-05,
"loss": 0.3257,
"step": 75
},
{
"epoch": 0.0029658536585365856,
"grad_norm": 4.479846954345703,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.2852,
"step": 76
},
{
"epoch": 0.003004878048780488,
"grad_norm": 1.678367257118225,
"learning_rate": 1.5200000000000002e-05,
"loss": 0.1093,
"step": 77
},
{
"epoch": 0.0030439024390243904,
"grad_norm": 4.245965957641602,
"learning_rate": 1.54e-05,
"loss": 0.1454,
"step": 78
},
{
"epoch": 0.0030829268292682926,
"grad_norm": 5.639206409454346,
"learning_rate": 1.5600000000000003e-05,
"loss": 0.3216,
"step": 79
},
{
"epoch": 0.0031219512195121953,
"grad_norm": 1.9889508485794067,
"learning_rate": 1.58e-05,
"loss": 0.1581,
"step": 80
},
{
"epoch": 0.0031609756097560975,
"grad_norm": 12.279495239257812,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.5877,
"step": 81
},
{
"epoch": 0.0032,
"grad_norm": 2.3625783920288086,
"learning_rate": 1.62e-05,
"loss": 0.1789,
"step": 82
},
{
"epoch": 0.0032390243902439024,
"grad_norm": 4.430482864379883,
"learning_rate": 1.64e-05,
"loss": 0.2241,
"step": 83
},
{
"epoch": 0.003278048780487805,
"grad_norm": 5.0593647956848145,
"learning_rate": 1.66e-05,
"loss": 0.1073,
"step": 84
},
{
"epoch": 0.0033170731707317072,
"grad_norm": 5.038410663604736,
"learning_rate": 1.6800000000000002e-05,
"loss": 0.2468,
"step": 85
},
{
"epoch": 0.00335609756097561,
"grad_norm": 6.7091240882873535,
"learning_rate": 1.7e-05,
"loss": 0.3885,
"step": 86
},
{
"epoch": 0.003395121951219512,
"grad_norm": 4.185050964355469,
"learning_rate": 1.72e-05,
"loss": 0.0824,
"step": 87
},
{
"epoch": 0.0034341463414634147,
"grad_norm": 6.81768798828125,
"learning_rate": 1.7400000000000003e-05,
"loss": 0.2585,
"step": 88
},
{
"epoch": 0.003473170731707317,
"grad_norm": 11.167231559753418,
"learning_rate": 1.76e-05,
"loss": 0.4026,
"step": 89
},
{
"epoch": 0.0035121951219512196,
"grad_norm": 5.705420970916748,
"learning_rate": 1.7800000000000002e-05,
"loss": 0.2463,
"step": 90
},
{
"epoch": 0.003551219512195122,
"grad_norm": 6.013692855834961,
"learning_rate": 1.8e-05,
"loss": 0.1471,
"step": 91
},
{
"epoch": 0.0035902439024390245,
"grad_norm": 6.8909149169921875,
"learning_rate": 1.8200000000000002e-05,
"loss": 0.2943,
"step": 92
},
{
"epoch": 0.0036292682926829267,
"grad_norm": 16.121662139892578,
"learning_rate": 1.8400000000000003e-05,
"loss": 0.9764,
"step": 93
},
{
"epoch": 0.0036682926829268293,
"grad_norm": 8.695253372192383,
"learning_rate": 1.86e-05,
"loss": 0.5182,
"step": 94
},
{
"epoch": 0.0037073170731707315,
"grad_norm": 13.130412101745605,
"learning_rate": 1.88e-05,
"loss": 0.8817,
"step": 95
},
{
"epoch": 0.003746341463414634,
"grad_norm": 12.404586791992188,
"learning_rate": 1.9e-05,
"loss": 0.4502,
"step": 96
},
{
"epoch": 0.0037853658536585364,
"grad_norm": 10.057884216308594,
"learning_rate": 1.9200000000000003e-05,
"loss": 0.292,
"step": 97
},
{
"epoch": 0.003824390243902439,
"grad_norm": 5.923167705535889,
"learning_rate": 1.94e-05,
"loss": 0.3237,
"step": 98
},
{
"epoch": 0.0038634146341463413,
"grad_norm": 4.915837287902832,
"learning_rate": 1.9600000000000002e-05,
"loss": 0.1174,
"step": 99
},
{
"epoch": 0.003902439024390244,
"grad_norm": 15.259949684143066,
"learning_rate": 1.98e-05,
"loss": 0.8936,
"step": 100
},
{
"epoch": 0.0039414634146341465,
"grad_norm": 1.4314539432525635,
"learning_rate": 2e-05,
"loss": 0.0978,
"step": 101
},
{
"epoch": 0.003980487804878049,
"grad_norm": 6.848033905029297,
"learning_rate": 1.9999939076577906e-05,
"loss": 0.4753,
"step": 102
},
{
"epoch": 0.004019512195121951,
"grad_norm": 8.512016296386719,
"learning_rate": 1.9999756307053947e-05,
"loss": 0.4204,
"step": 103
},
{
"epoch": 0.004058536585365854,
"grad_norm": 4.4925031661987305,
"learning_rate": 1.9999451693655125e-05,
"loss": 0.2196,
"step": 104
},
{
"epoch": 0.004097560975609756,
"grad_norm": 5.722268104553223,
"learning_rate": 1.9999025240093045e-05,
"loss": 0.3162,
"step": 105
},
{
"epoch": 0.004136585365853659,
"grad_norm": 1.9750213623046875,
"learning_rate": 1.9998476951563914e-05,
"loss": 0.1865,
"step": 106
},
{
"epoch": 0.004175609756097561,
"grad_norm": 2.2069339752197266,
"learning_rate": 1.9997806834748455e-05,
"loss": 0.1265,
"step": 107
},
{
"epoch": 0.004214634146341463,
"grad_norm": 1.1415362358093262,
"learning_rate": 1.9997014897811834e-05,
"loss": 0.1048,
"step": 108
},
{
"epoch": 0.004253658536585366,
"grad_norm": 2.7669618129730225,
"learning_rate": 1.9996101150403543e-05,
"loss": 0.1776,
"step": 109
},
{
"epoch": 0.004292682926829269,
"grad_norm": 4.738790988922119,
"learning_rate": 1.9995065603657317e-05,
"loss": 0.3257,
"step": 110
},
{
"epoch": 0.00433170731707317,
"grad_norm": 7.05373477935791,
"learning_rate": 1.999390827019096e-05,
"loss": 0.3921,
"step": 111
},
{
"epoch": 0.004370731707317073,
"grad_norm": 10.975568771362305,
"learning_rate": 1.999262916410621e-05,
"loss": 1.091,
"step": 112
},
{
"epoch": 0.004409756097560976,
"grad_norm": 1.9560153484344482,
"learning_rate": 1.9991228300988586e-05,
"loss": 0.1508,
"step": 113
},
{
"epoch": 0.004448780487804878,
"grad_norm": 9.060037612915039,
"learning_rate": 1.998970569790715e-05,
"loss": 0.2468,
"step": 114
},
{
"epoch": 0.00448780487804878,
"grad_norm": 3.928701639175415,
"learning_rate": 1.9988061373414342e-05,
"loss": 0.2342,
"step": 115
},
{
"epoch": 0.004526829268292683,
"grad_norm": 4.233069896697998,
"learning_rate": 1.9986295347545738e-05,
"loss": 0.2498,
"step": 116
},
{
"epoch": 0.004565853658536585,
"grad_norm": 3.7723169326782227,
"learning_rate": 1.9984407641819812e-05,
"loss": 0.0642,
"step": 117
},
{
"epoch": 0.004604878048780488,
"grad_norm": 6.083825588226318,
"learning_rate": 1.9982398279237657e-05,
"loss": 0.1499,
"step": 118
},
{
"epoch": 0.00464390243902439,
"grad_norm": 6.470362186431885,
"learning_rate": 1.9980267284282718e-05,
"loss": 0.2702,
"step": 119
},
{
"epoch": 0.0046829268292682925,
"grad_norm": 2.6331472396850586,
"learning_rate": 1.9978014682920503e-05,
"loss": 0.0493,
"step": 120
},
{
"epoch": 0.004721951219512195,
"grad_norm": 1.2624144554138184,
"learning_rate": 1.9975640502598243e-05,
"loss": 0.1,
"step": 121
},
{
"epoch": 0.004760975609756098,
"grad_norm": 3.1314401626586914,
"learning_rate": 1.997314477224458e-05,
"loss": 0.1177,
"step": 122
},
{
"epoch": 0.0048,
"grad_norm": 3.279804229736328,
"learning_rate": 1.9970527522269204e-05,
"loss": 0.1379,
"step": 123
},
{
"epoch": 0.004839024390243902,
"grad_norm": 1.9532862901687622,
"learning_rate": 1.9967788784562474e-05,
"loss": 0.1259,
"step": 124
},
{
"epoch": 0.004878048780487805,
"grad_norm": 3.9451792240142822,
"learning_rate": 1.9964928592495046e-05,
"loss": 0.168,
"step": 125
},
{
"epoch": 0.0049170731707317075,
"grad_norm": 12.864158630371094,
"learning_rate": 1.9961946980917457e-05,
"loss": 0.5974,
"step": 126
},
{
"epoch": 0.00495609756097561,
"grad_norm": 2.327380657196045,
"learning_rate": 1.9958843986159705e-05,
"loss": 0.1284,
"step": 127
},
{
"epoch": 0.004995121951219512,
"grad_norm": 6.298210620880127,
"learning_rate": 1.99556196460308e-05,
"loss": 0.3122,
"step": 128
},
{
"epoch": 0.005034146341463415,
"grad_norm": 4.496811389923096,
"learning_rate": 1.9952273999818312e-05,
"loss": 0.1508,
"step": 129
},
{
"epoch": 0.005073170731707317,
"grad_norm": 3.48772931098938,
"learning_rate": 1.9948807088287884e-05,
"loss": 0.1321,
"step": 130
},
{
"epoch": 0.00511219512195122,
"grad_norm": 3.9864776134490967,
"learning_rate": 1.9945218953682736e-05,
"loss": 0.1646,
"step": 131
},
{
"epoch": 0.005151219512195122,
"grad_norm": 10.274496078491211,
"learning_rate": 1.9941509639723155e-05,
"loss": 0.2324,
"step": 132
},
{
"epoch": 0.005190243902439024,
"grad_norm": 5.216224670410156,
"learning_rate": 1.9937679191605964e-05,
"loss": 0.1829,
"step": 133
},
{
"epoch": 0.005229268292682927,
"grad_norm": 5.065738201141357,
"learning_rate": 1.9933727656003964e-05,
"loss": 0.1178,
"step": 134
},
{
"epoch": 0.00526829268292683,
"grad_norm": 1.7729227542877197,
"learning_rate": 1.992965508106537e-05,
"loss": 0.1556,
"step": 135
},
{
"epoch": 0.005307317073170731,
"grad_norm": 6.401525497436523,
"learning_rate": 1.9925461516413224e-05,
"loss": 0.4589,
"step": 136
},
{
"epoch": 0.005346341463414634,
"grad_norm": 1.491703748703003,
"learning_rate": 1.9921147013144782e-05,
"loss": 0.0592,
"step": 137
},
{
"epoch": 0.005385365853658537,
"grad_norm": 5.682543754577637,
"learning_rate": 1.9916711623830904e-05,
"loss": 0.243,
"step": 138
},
{
"epoch": 0.005424390243902439,
"grad_norm": 2.9709320068359375,
"learning_rate": 1.991215540251542e-05,
"loss": 0.1012,
"step": 139
},
{
"epoch": 0.005463414634146341,
"grad_norm": 6.430043697357178,
"learning_rate": 1.9907478404714438e-05,
"loss": 0.1594,
"step": 140
},
{
"epoch": 0.005502439024390244,
"grad_norm": 4.087896347045898,
"learning_rate": 1.9902680687415704e-05,
"loss": 0.1771,
"step": 141
},
{
"epoch": 0.005541463414634146,
"grad_norm": 2.6057660579681396,
"learning_rate": 1.989776230907789e-05,
"loss": 0.0375,
"step": 142
},
{
"epoch": 0.005580487804878049,
"grad_norm": 5.835061073303223,
"learning_rate": 1.9892723329629885e-05,
"loss": 0.291,
"step": 143
},
{
"epoch": 0.005619512195121951,
"grad_norm": 13.750198364257812,
"learning_rate": 1.988756381047006e-05,
"loss": 0.3837,
"step": 144
},
{
"epoch": 0.0056585365853658535,
"grad_norm": 2.4703943729400635,
"learning_rate": 1.988228381446553e-05,
"loss": 0.1973,
"step": 145
},
{
"epoch": 0.005697560975609756,
"grad_norm": 1.4389910697937012,
"learning_rate": 1.9876883405951378e-05,
"loss": 0.1136,
"step": 146
},
{
"epoch": 0.005736585365853659,
"grad_norm": 8.243542671203613,
"learning_rate": 1.987136265072988e-05,
"loss": 0.1831,
"step": 147
},
{
"epoch": 0.0057756097560975606,
"grad_norm": 1.2565394639968872,
"learning_rate": 1.9865721616069695e-05,
"loss": 0.107,
"step": 148
},
{
"epoch": 0.005814634146341463,
"grad_norm": 6.737248420715332,
"learning_rate": 1.985996037070505e-05,
"loss": 0.2623,
"step": 149
},
{
"epoch": 0.005853658536585366,
"grad_norm": 1.1391600370407104,
"learning_rate": 1.9854078984834904e-05,
"loss": 0.1409,
"step": 150
},
{
"epoch": 0.0058926829268292685,
"grad_norm": 2.5610439777374268,
"learning_rate": 1.9848077530122083e-05,
"loss": 0.1771,
"step": 151
},
{
"epoch": 0.005931707317073171,
"grad_norm": 5.796145439147949,
"learning_rate": 1.984195607969242e-05,
"loss": 0.2179,
"step": 152
},
{
"epoch": 0.005970731707317073,
"grad_norm": 2.511676549911499,
"learning_rate": 1.983571470813386e-05,
"loss": 0.1232,
"step": 153
},
{
"epoch": 0.006009756097560976,
"grad_norm": 6.647976875305176,
"learning_rate": 1.9829353491495545e-05,
"loss": 0.2521,
"step": 154
},
{
"epoch": 0.006048780487804878,
"grad_norm": 5.5812506675720215,
"learning_rate": 1.982287250728689e-05,
"loss": 0.1823,
"step": 155
},
{
"epoch": 0.006087804878048781,
"grad_norm": 11.388985633850098,
"learning_rate": 1.9816271834476642e-05,
"loss": 0.3122,
"step": 156
},
{
"epoch": 0.006126829268292683,
"grad_norm": 8.605890274047852,
"learning_rate": 1.9809551553491918e-05,
"loss": 0.694,
"step": 157
},
{
"epoch": 0.006165853658536585,
"grad_norm": 5.374488353729248,
"learning_rate": 1.9802711746217222e-05,
"loss": 0.0841,
"step": 158
},
{
"epoch": 0.006204878048780488,
"grad_norm": 0.8785826563835144,
"learning_rate": 1.979575249599344e-05,
"loss": 0.0445,
"step": 159
},
{
"epoch": 0.006243902439024391,
"grad_norm": 1.143341064453125,
"learning_rate": 1.9788673887616852e-05,
"loss": 0.0886,
"step": 160
},
{
"epoch": 0.006282926829268292,
"grad_norm": 1.6193596124649048,
"learning_rate": 1.9781476007338058e-05,
"loss": 0.0472,
"step": 161
},
{
"epoch": 0.006321951219512195,
"grad_norm": 4.978359699249268,
"learning_rate": 1.9774158942860962e-05,
"loss": 0.1315,
"step": 162
},
{
"epoch": 0.006360975609756098,
"grad_norm": 5.966531276702881,
"learning_rate": 1.9766722783341682e-05,
"loss": 0.1987,
"step": 163
},
{
"epoch": 0.0064,
"grad_norm": 1.901698350906372,
"learning_rate": 1.9759167619387474e-05,
"loss": 0.1075,
"step": 164
},
{
"epoch": 0.006439024390243902,
"grad_norm": 5.009815692901611,
"learning_rate": 1.9751493543055634e-05,
"loss": 0.0476,
"step": 165
},
{
"epoch": 0.006478048780487805,
"grad_norm": 8.462274551391602,
"learning_rate": 1.9743700647852356e-05,
"loss": 0.2288,
"step": 166
},
{
"epoch": 0.006517073170731707,
"grad_norm": 9.655773162841797,
"learning_rate": 1.9735789028731603e-05,
"loss": 0.2292,
"step": 167
},
{
"epoch": 0.00655609756097561,
"grad_norm": 1.6972038745880127,
"learning_rate": 1.972775878209397e-05,
"loss": 0.1048,
"step": 168
},
{
"epoch": 0.006595121951219512,
"grad_norm": 1.636698842048645,
"learning_rate": 1.9719610005785466e-05,
"loss": 0.1499,
"step": 169
},
{
"epoch": 0.0066341463414634145,
"grad_norm": 12.71432876586914,
"learning_rate": 1.971134279909636e-05,
"loss": 0.3148,
"step": 170
},
{
"epoch": 0.006673170731707317,
"grad_norm": 2.0866127014160156,
"learning_rate": 1.9702957262759964e-05,
"loss": 0.0721,
"step": 171
},
{
"epoch": 0.00671219512195122,
"grad_norm": 1.763672947883606,
"learning_rate": 1.9694453498951392e-05,
"loss": 0.1248,
"step": 172
},
{
"epoch": 0.0067512195121951215,
"grad_norm": 1.41291081905365,
"learning_rate": 1.9685831611286312e-05,
"loss": 0.1052,
"step": 173
},
{
"epoch": 0.006790243902439024,
"grad_norm": 1.3923869132995605,
"learning_rate": 1.9677091704819714e-05,
"loss": 0.1122,
"step": 174
},
{
"epoch": 0.006829268292682927,
"grad_norm": 7.990157604217529,
"learning_rate": 1.9668233886044597e-05,
"loss": 0.1776,
"step": 175
},
{
"epoch": 0.0068682926829268295,
"grad_norm": 0.5866390466690063,
"learning_rate": 1.9659258262890683e-05,
"loss": 0.0336,
"step": 176
},
{
"epoch": 0.006907317073170732,
"grad_norm": 1.2356959581375122,
"learning_rate": 1.9650164944723116e-05,
"loss": 0.1327,
"step": 177
},
{
"epoch": 0.006946341463414634,
"grad_norm": 0.13389591872692108,
"learning_rate": 1.96409540423411e-05,
"loss": 0.0019,
"step": 178
},
{
"epoch": 0.0069853658536585365,
"grad_norm": 1.350856900215149,
"learning_rate": 1.9631625667976584e-05,
"loss": 0.0961,
"step": 179
},
{
"epoch": 0.007024390243902439,
"grad_norm": 1.223233699798584,
"learning_rate": 1.9622179935292855e-05,
"loss": 0.1068,
"step": 180
},
{
"epoch": 0.007063414634146342,
"grad_norm": 7.109902381896973,
"learning_rate": 1.961261695938319e-05,
"loss": 0.2072,
"step": 181
},
{
"epoch": 0.007102439024390244,
"grad_norm": 3.1937339305877686,
"learning_rate": 1.9602936856769432e-05,
"loss": 0.0983,
"step": 182
},
{
"epoch": 0.007141463414634146,
"grad_norm": 0.9007237553596497,
"learning_rate": 1.9593139745400575e-05,
"loss": 0.0891,
"step": 183
},
{
"epoch": 0.007180487804878049,
"grad_norm": 7.184077739715576,
"learning_rate": 1.9583225744651334e-05,
"loss": 0.1117,
"step": 184
},
{
"epoch": 0.0072195121951219516,
"grad_norm": 3.034726619720459,
"learning_rate": 1.9573194975320672e-05,
"loss": 0.114,
"step": 185
},
{
"epoch": 0.007258536585365853,
"grad_norm": 2.3767752647399902,
"learning_rate": 1.9563047559630356e-05,
"loss": 0.0778,
"step": 186
},
{
"epoch": 0.007297560975609756,
"grad_norm": 0.5144960880279541,
"learning_rate": 1.9552783621223437e-05,
"loss": 0.034,
"step": 187
},
{
"epoch": 0.007336585365853659,
"grad_norm": 2.027453660964966,
"learning_rate": 1.954240328516277e-05,
"loss": 0.1038,
"step": 188
},
{
"epoch": 0.007375609756097561,
"grad_norm": 3.0588150024414062,
"learning_rate": 1.9531906677929472e-05,
"loss": 0.1259,
"step": 189
},
{
"epoch": 0.007414634146341463,
"grad_norm": 0.9642093181610107,
"learning_rate": 1.9521293927421388e-05,
"loss": 0.0687,
"step": 190
},
{
"epoch": 0.007453658536585366,
"grad_norm": 0.6058931946754456,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.0134,
"step": 191
},
{
"epoch": 0.007492682926829268,
"grad_norm": 5.5360541343688965,
"learning_rate": 1.9499720515246524e-05,
"loss": 0.1443,
"step": 192
},
{
"epoch": 0.007531707317073171,
"grad_norm": 24.87925148010254,
"learning_rate": 1.9488760116444966e-05,
"loss": 0.351,
"step": 193
},
{
"epoch": 0.007570731707317073,
"grad_norm": 2.0963354110717773,
"learning_rate": 1.947768410009586e-05,
"loss": 0.0934,
"step": 194
},
{
"epoch": 0.007609756097560975,
"grad_norm": 9.65958023071289,
"learning_rate": 1.9466492601156964e-05,
"loss": 0.4382,
"step": 195
},
{
"epoch": 0.007648780487804878,
"grad_norm": 1.131377100944519,
"learning_rate": 1.945518575599317e-05,
"loss": 0.0719,
"step": 196
},
{
"epoch": 0.007687804878048781,
"grad_norm": 1.1503591537475586,
"learning_rate": 1.944376370237481e-05,
"loss": 0.1202,
"step": 197
},
{
"epoch": 0.0077268292682926825,
"grad_norm": 7.936573505401611,
"learning_rate": 1.943222657947601e-05,
"loss": 0.1963,
"step": 198
},
{
"epoch": 0.007765853658536585,
"grad_norm": 13.797683715820312,
"learning_rate": 1.942057452787297e-05,
"loss": 0.1847,
"step": 199
},
{
"epoch": 0.007804878048780488,
"grad_norm": 0.8876925706863403,
"learning_rate": 1.9408807689542257e-05,
"loss": 0.0607,
"step": 200
},
{
"epoch": 0.00784390243902439,
"grad_norm": 0.717039942741394,
"learning_rate": 1.9396926207859085e-05,
"loss": 0.0984,
"step": 201
},
{
"epoch": 0.007882926829268293,
"grad_norm": 0.9834874272346497,
"learning_rate": 1.938493022759556e-05,
"loss": 0.0591,
"step": 202
},
{
"epoch": 0.007921951219512195,
"grad_norm": 8.773293495178223,
"learning_rate": 1.937281989491892e-05,
"loss": 0.2484,
"step": 203
},
{
"epoch": 0.007960975609756098,
"grad_norm": 6.789969444274902,
"learning_rate": 1.9360595357389735e-05,
"loss": 0.1926,
"step": 204
},
{
"epoch": 0.008,
"grad_norm": 1.8436493873596191,
"learning_rate": 1.9348256763960146e-05,
"loss": 0.0911,
"step": 205
},
{
"epoch": 0.008039024390243902,
"grad_norm": 0.7573915719985962,
"learning_rate": 1.9335804264972018e-05,
"loss": 0.0924,
"step": 206
},
{
"epoch": 0.008078048780487805,
"grad_norm": 8.017034530639648,
"learning_rate": 1.9323238012155125e-05,
"loss": 0.2224,
"step": 207
},
{
"epoch": 0.008117073170731707,
"grad_norm": 5.28120756149292,
"learning_rate": 1.9310558158625286e-05,
"loss": 0.1767,
"step": 208
},
{
"epoch": 0.008156097560975609,
"grad_norm": 0.8488652110099792,
"learning_rate": 1.9297764858882516e-05,
"loss": 0.1035,
"step": 209
},
{
"epoch": 0.008195121951219513,
"grad_norm": 3.5903353691101074,
"learning_rate": 1.9284858268809135e-05,
"loss": 0.1408,
"step": 210
},
{
"epoch": 0.008234146341463414,
"grad_norm": 4.732028961181641,
"learning_rate": 1.9271838545667876e-05,
"loss": 0.1164,
"step": 211
},
{
"epoch": 0.008273170731707318,
"grad_norm": 1.2205359935760498,
"learning_rate": 1.925870584809995e-05,
"loss": 0.0955,
"step": 212
},
{
"epoch": 0.00831219512195122,
"grad_norm": 7.764541149139404,
"learning_rate": 1.9245460336123136e-05,
"loss": 0.5587,
"step": 213
},
{
"epoch": 0.008351219512195121,
"grad_norm": 12.925251960754395,
"learning_rate": 1.923210217112981e-05,
"loss": 0.4424,
"step": 214
},
{
"epoch": 0.008390243902439025,
"grad_norm": 2.0783300399780273,
"learning_rate": 1.9218631515885007e-05,
"loss": 0.1959,
"step": 215
},
{
"epoch": 0.008429268292682927,
"grad_norm": 1.5903644561767578,
"learning_rate": 1.9205048534524405e-05,
"loss": 0.0991,
"step": 216
},
{
"epoch": 0.008468292682926828,
"grad_norm": 4.469902992248535,
"learning_rate": 1.9191353392552346e-05,
"loss": 0.1302,
"step": 217
},
{
"epoch": 0.008507317073170732,
"grad_norm": 1.0593162775039673,
"learning_rate": 1.9177546256839814e-05,
"loss": 0.043,
"step": 218
},
{
"epoch": 0.008546341463414634,
"grad_norm": 1.164799690246582,
"learning_rate": 1.9163627295622397e-05,
"loss": 0.109,
"step": 219
},
{
"epoch": 0.008585365853658537,
"grad_norm": 8.384074211120605,
"learning_rate": 1.914959667849825e-05,
"loss": 0.2114,
"step": 220
},
{
"epoch": 0.008624390243902439,
"grad_norm": 0.7901926636695862,
"learning_rate": 1.913545457642601e-05,
"loss": 0.0551,
"step": 221
},
{
"epoch": 0.00866341463414634,
"grad_norm": 1.8925807476043701,
"learning_rate": 1.9121201161722732e-05,
"loss": 0.0894,
"step": 222
},
{
"epoch": 0.008702439024390244,
"grad_norm": 10.837748527526855,
"learning_rate": 1.910683660806177e-05,
"loss": 0.4284,
"step": 223
},
{
"epoch": 0.008741463414634146,
"grad_norm": 1.533721923828125,
"learning_rate": 1.9092361090470688e-05,
"loss": 0.1186,
"step": 224
},
{
"epoch": 0.00878048780487805,
"grad_norm": 6.5838799476623535,
"learning_rate": 1.907777478532909e-05,
"loss": 0.2676,
"step": 225
},
{
"epoch": 0.008819512195121951,
"grad_norm": 11.218518257141113,
"learning_rate": 1.9063077870366504e-05,
"loss": 0.2635,
"step": 226
},
{
"epoch": 0.008858536585365853,
"grad_norm": 15.041123390197754,
"learning_rate": 1.9048270524660197e-05,
"loss": 0.2922,
"step": 227
},
{
"epoch": 0.008897560975609757,
"grad_norm": 1.7548902034759521,
"learning_rate": 1.903335292863301e-05,
"loss": 0.1029,
"step": 228
},
{
"epoch": 0.008936585365853658,
"grad_norm": 1.9405819177627563,
"learning_rate": 1.901832526405114e-05,
"loss": 0.0935,
"step": 229
},
{
"epoch": 0.00897560975609756,
"grad_norm": 2.6685397624969482,
"learning_rate": 1.9003187714021936e-05,
"loss": 0.0798,
"step": 230
},
{
"epoch": 0.009014634146341464,
"grad_norm": 0.7036566734313965,
"learning_rate": 1.8987940462991673e-05,
"loss": 0.0715,
"step": 231
},
{
"epoch": 0.009053658536585366,
"grad_norm": 1.298852801322937,
"learning_rate": 1.8972583696743284e-05,
"loss": 0.0772,
"step": 232
},
{
"epoch": 0.009092682926829269,
"grad_norm": 6.946676731109619,
"learning_rate": 1.895711760239413e-05,
"loss": 0.1485,
"step": 233
},
{
"epoch": 0.00913170731707317,
"grad_norm": 4.939789295196533,
"learning_rate": 1.8941542368393683e-05,
"loss": 0.1396,
"step": 234
},
{
"epoch": 0.009170731707317073,
"grad_norm": 1.0028315782546997,
"learning_rate": 1.892585818452126e-05,
"loss": 0.0545,
"step": 235
},
{
"epoch": 0.009209756097560976,
"grad_norm": 0.531678318977356,
"learning_rate": 1.891006524188368e-05,
"loss": 0.0474,
"step": 236
},
{
"epoch": 0.009248780487804878,
"grad_norm": 2.054960250854492,
"learning_rate": 1.889416373291298e-05,
"loss": 0.1086,
"step": 237
},
{
"epoch": 0.00928780487804878,
"grad_norm": 12.858938217163086,
"learning_rate": 1.8878153851364013e-05,
"loss": 0.7348,
"step": 238
},
{
"epoch": 0.009326829268292683,
"grad_norm": 0.15298770368099213,
"learning_rate": 1.8862035792312148e-05,
"loss": 0.0016,
"step": 239
},
{
"epoch": 0.009365853658536585,
"grad_norm": 8.230769157409668,
"learning_rate": 1.884580975215084e-05,
"loss": 0.1565,
"step": 240
},
{
"epoch": 0.009404878048780489,
"grad_norm": 9.449628829956055,
"learning_rate": 1.8829475928589272e-05,
"loss": 0.2137,
"step": 241
},
{
"epoch": 0.00944390243902439,
"grad_norm": 13.610590934753418,
"learning_rate": 1.8813034520649923e-05,
"loss": 0.2165,
"step": 242
},
{
"epoch": 0.009482926829268292,
"grad_norm": 15.66429328918457,
"learning_rate": 1.879648572866617e-05,
"loss": 0.4004,
"step": 243
},
{
"epoch": 0.009521951219512196,
"grad_norm": 8.834271430969238,
"learning_rate": 1.8779829754279806e-05,
"loss": 0.2625,
"step": 244
},
{
"epoch": 0.009560975609756097,
"grad_norm": 6.070154666900635,
"learning_rate": 1.8763066800438638e-05,
"loss": 0.162,
"step": 245
},
{
"epoch": 0.0096,
"grad_norm": 0.2640407681465149,
"learning_rate": 1.874619707139396e-05,
"loss": 0.0256,
"step": 246
},
{
"epoch": 0.009639024390243903,
"grad_norm": 0.8703957796096802,
"learning_rate": 1.8729220772698096e-05,
"loss": 0.0662,
"step": 247
},
{
"epoch": 0.009678048780487804,
"grad_norm": 1.0811554193496704,
"learning_rate": 1.8712138111201898e-05,
"loss": 0.0658,
"step": 248
},
{
"epoch": 0.009717073170731708,
"grad_norm": 8.928948402404785,
"learning_rate": 1.869494929505219e-05,
"loss": 0.2238,
"step": 249
},
{
"epoch": 0.00975609756097561,
"grad_norm": 1.734654426574707,
"learning_rate": 1.8677654533689287e-05,
"loss": 0.1008,
"step": 250
},
{
"epoch": 0.009795121951219512,
"grad_norm": 0.8071237802505493,
"learning_rate": 1.866025403784439e-05,
"loss": 0.0695,
"step": 251
},
{
"epoch": 0.009834146341463415,
"grad_norm": 6.491012096405029,
"learning_rate": 1.864274801953705e-05,
"loss": 0.1154,
"step": 252
},
{
"epoch": 0.009873170731707317,
"grad_norm": 1.868613600730896,
"learning_rate": 1.8625136692072577e-05,
"loss": 0.0593,
"step": 253
},
{
"epoch": 0.00991219512195122,
"grad_norm": 8.784892082214355,
"learning_rate": 1.860742027003944e-05,
"loss": 0.1365,
"step": 254
},
{
"epoch": 0.009951219512195122,
"grad_norm": 0.5071855187416077,
"learning_rate": 1.8589598969306646e-05,
"loss": 0.0523,
"step": 255
},
{
"epoch": 0.009990243902439024,
"grad_norm": 1.2618370056152344,
"learning_rate": 1.8571673007021124e-05,
"loss": 0.0914,
"step": 256
},
{
"epoch": 0.010029268292682927,
"grad_norm": 19.47931671142578,
"learning_rate": 1.855364260160507e-05,
"loss": 0.653,
"step": 257
},
{
"epoch": 0.01006829268292683,
"grad_norm": 13.547938346862793,
"learning_rate": 1.8535507972753275e-05,
"loss": 0.3182,
"step": 258
},
{
"epoch": 0.010107317073170731,
"grad_norm": 12.62155818939209,
"learning_rate": 1.851726934143048e-05,
"loss": 0.182,
"step": 259
},
{
"epoch": 0.010146341463414634,
"grad_norm": 0.8845065832138062,
"learning_rate": 1.849892692986864e-05,
"loss": 0.0608,
"step": 260
},
{
"epoch": 0.010185365853658536,
"grad_norm": 0.7957473993301392,
"learning_rate": 1.848048096156426e-05,
"loss": 0.0636,
"step": 261
},
{
"epoch": 0.01022439024390244,
"grad_norm": 0.4376261234283447,
"learning_rate": 1.8461931661275642e-05,
"loss": 0.0062,
"step": 262
},
{
"epoch": 0.010263414634146342,
"grad_norm": 15.762907981872559,
"learning_rate": 1.8443279255020153e-05,
"loss": 0.2545,
"step": 263
},
{
"epoch": 0.010302439024390243,
"grad_norm": 13.374885559082031,
"learning_rate": 1.842452397007148e-05,
"loss": 0.1359,
"step": 264
},
{
"epoch": 0.010341463414634147,
"grad_norm": 11.982704162597656,
"learning_rate": 1.8405666034956842e-05,
"loss": 0.1324,
"step": 265
},
{
"epoch": 0.010380487804878049,
"grad_norm": 3.7512705326080322,
"learning_rate": 1.8386705679454243e-05,
"loss": 0.0956,
"step": 266
},
{
"epoch": 0.01041951219512195,
"grad_norm": 0.1455240398645401,
"learning_rate": 1.836764313458962e-05,
"loss": 0.0017,
"step": 267
},
{
"epoch": 0.010458536585365854,
"grad_norm": 0.38747304677963257,
"learning_rate": 1.8348478632634067e-05,
"loss": 0.0466,
"step": 268
},
{
"epoch": 0.010497560975609756,
"grad_norm": 7.600417137145996,
"learning_rate": 1.8329212407100996e-05,
"loss": 0.1524,
"step": 269
},
{
"epoch": 0.01053658536585366,
"grad_norm": 0.2582570016384125,
"learning_rate": 1.8309844692743283e-05,
"loss": 0.0032,
"step": 270
},
{
"epoch": 0.010575609756097561,
"grad_norm": 0.40249213576316833,
"learning_rate": 1.8290375725550417e-05,
"loss": 0.0475,
"step": 271
},
{
"epoch": 0.010614634146341463,
"grad_norm": 0.7164753079414368,
"learning_rate": 1.827080574274562e-05,
"loss": 0.0568,
"step": 272
},
{
"epoch": 0.010653658536585366,
"grad_norm": 1.23041832447052,
"learning_rate": 1.8251134982782952e-05,
"loss": 0.0575,
"step": 273
},
{
"epoch": 0.010692682926829268,
"grad_norm": 10.077104568481445,
"learning_rate": 1.8231363685344422e-05,
"loss": 0.1397,
"step": 274
},
{
"epoch": 0.010731707317073172,
"grad_norm": 1.252584457397461,
"learning_rate": 1.821149209133704e-05,
"loss": 0.0764,
"step": 275
},
{
"epoch": 0.010770731707317073,
"grad_norm": 1.1576642990112305,
"learning_rate": 1.819152044288992e-05,
"loss": 0.0829,
"step": 276
},
{
"epoch": 0.010809756097560975,
"grad_norm": 0.9366648197174072,
"learning_rate": 1.8171448983351284e-05,
"loss": 0.0673,
"step": 277
},
{
"epoch": 0.010848780487804879,
"grad_norm": 1.2474669218063354,
"learning_rate": 1.815127795728554e-05,
"loss": 0.0914,
"step": 278
},
{
"epoch": 0.01088780487804878,
"grad_norm": 1.8186805248260498,
"learning_rate": 1.8131007610470278e-05,
"loss": 0.0968,
"step": 279
},
{
"epoch": 0.010926829268292682,
"grad_norm": 0.17496399581432343,
"learning_rate": 1.8110638189893267e-05,
"loss": 0.0027,
"step": 280
},
{
"epoch": 0.010965853658536586,
"grad_norm": 5.077552318572998,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.0742,
"step": 281
},
{
"epoch": 0.011004878048780488,
"grad_norm": 1.5596524477005005,
"learning_rate": 1.806960312143802e-05,
"loss": 0.069,
"step": 282
},
{
"epoch": 0.011043902439024391,
"grad_norm": 2.0858726501464844,
"learning_rate": 1.804893797355914e-05,
"loss": 0.0978,
"step": 283
},
{
"epoch": 0.011082926829268293,
"grad_norm": 0.969219446182251,
"learning_rate": 1.8028174751911147e-05,
"loss": 0.092,
"step": 284
},
{
"epoch": 0.011121951219512195,
"grad_norm": 0.5313135981559753,
"learning_rate": 1.8007313709487334e-05,
"loss": 0.0699,
"step": 285
},
{
"epoch": 0.011160975609756098,
"grad_norm": 1.3951212167739868,
"learning_rate": 1.798635510047293e-05,
"loss": 0.0967,
"step": 286
},
{
"epoch": 0.0112,
"grad_norm": 3.1534245014190674,
"learning_rate": 1.7965299180241963e-05,
"loss": 0.0575,
"step": 287
},
{
"epoch": 0.011239024390243902,
"grad_norm": 6.24267578125,
"learning_rate": 1.7944146205354182e-05,
"loss": 0.1468,
"step": 288
},
{
"epoch": 0.011278048780487805,
"grad_norm": 12.395264625549316,
"learning_rate": 1.792289643355191e-05,
"loss": 0.5302,
"step": 289
},
{
"epoch": 0.011317073170731707,
"grad_norm": 1.8228356838226318,
"learning_rate": 1.7901550123756906e-05,
"loss": 0.0397,
"step": 290
},
{
"epoch": 0.01135609756097561,
"grad_norm": 1.1454330682754517,
"learning_rate": 1.788010753606722e-05,
"loss": 0.101,
"step": 291
},
{
"epoch": 0.011395121951219512,
"grad_norm": 0.22947749495506287,
"learning_rate": 1.785856893175402e-05,
"loss": 0.0227,
"step": 292
},
{
"epoch": 0.011434146341463414,
"grad_norm": 0.6581270098686218,
"learning_rate": 1.78369345732584e-05,
"loss": 0.0602,
"step": 293
},
{
"epoch": 0.011473170731707318,
"grad_norm": 0.2802331745624542,
"learning_rate": 1.781520472418819e-05,
"loss": 0.0461,
"step": 294
},
{
"epoch": 0.01151219512195122,
"grad_norm": 1.0553476810455322,
"learning_rate": 1.7793379649314743e-05,
"loss": 0.0615,
"step": 295
},
{
"epoch": 0.011551219512195121,
"grad_norm": 0.2524442672729492,
"learning_rate": 1.777145961456971e-05,
"loss": 0.0039,
"step": 296
},
{
"epoch": 0.011590243902439025,
"grad_norm": 2.464233875274658,
"learning_rate": 1.7749444887041797e-05,
"loss": 0.1106,
"step": 297
},
{
"epoch": 0.011629268292682926,
"grad_norm": 1.0873249769210815,
"learning_rate": 1.7727335734973512e-05,
"loss": 0.0585,
"step": 298
},
{
"epoch": 0.01166829268292683,
"grad_norm": 17.076723098754883,
"learning_rate": 1.7705132427757895e-05,
"loss": 0.3282,
"step": 299
},
{
"epoch": 0.011707317073170732,
"grad_norm": 1.1585568189620972,
"learning_rate": 1.7682835235935236e-05,
"loss": 0.0865,
"step": 300
},
{
"epoch": 0.011746341463414633,
"grad_norm": 1.6616325378417969,
"learning_rate": 1.766044443118978e-05,
"loss": 0.1263,
"step": 301
},
{
"epoch": 0.011785365853658537,
"grad_norm": 0.7904642820358276,
"learning_rate": 1.7637960286346423e-05,
"loss": 0.0841,
"step": 302
},
{
"epoch": 0.011824390243902439,
"grad_norm": 1.0754047632217407,
"learning_rate": 1.761538307536737e-05,
"loss": 0.0441,
"step": 303
},
{
"epoch": 0.011863414634146342,
"grad_norm": 0.8069223165512085,
"learning_rate": 1.759271307334881e-05,
"loss": 0.0574,
"step": 304
},
{
"epoch": 0.011902439024390244,
"grad_norm": 0.13271929323673248,
"learning_rate": 1.7569950556517566e-05,
"loss": 0.0201,
"step": 305
},
{
"epoch": 0.011941463414634146,
"grad_norm": 0.1488584578037262,
"learning_rate": 1.7547095802227723e-05,
"loss": 0.0026,
"step": 306
},
{
"epoch": 0.01198048780487805,
"grad_norm": 1.3673462867736816,
"learning_rate": 1.7524149088957244e-05,
"loss": 0.0726,
"step": 307
},
{
"epoch": 0.012019512195121951,
"grad_norm": 0.06994021683931351,
"learning_rate": 1.7501110696304598e-05,
"loss": 0.0008,
"step": 308
},
{
"epoch": 0.012058536585365853,
"grad_norm": 1.758649468421936,
"learning_rate": 1.747798090498532e-05,
"loss": 0.0141,
"step": 309
},
{
"epoch": 0.012097560975609756,
"grad_norm": 3.0488345623016357,
"learning_rate": 1.7454759996828622e-05,
"loss": 0.0994,
"step": 310
},
{
"epoch": 0.012136585365853658,
"grad_norm": 11.195862770080566,
"learning_rate": 1.7431448254773943e-05,
"loss": 0.16,
"step": 311
},
{
"epoch": 0.012175609756097562,
"grad_norm": 0.3210989534854889,
"learning_rate": 1.74080459628675e-05,
"loss": 0.0228,
"step": 312
},
{
"epoch": 0.012214634146341464,
"grad_norm": 1.3756961822509766,
"learning_rate": 1.7384553406258842e-05,
"loss": 0.0192,
"step": 313
},
{
"epoch": 0.012253658536585365,
"grad_norm": 1.4928122758865356,
"learning_rate": 1.7360970871197347e-05,
"loss": 0.0634,
"step": 314
},
{
"epoch": 0.012292682926829269,
"grad_norm": 0.5534517765045166,
"learning_rate": 1.7337298645028764e-05,
"loss": 0.0105,
"step": 315
},
{
"epoch": 0.01233170731707317,
"grad_norm": 1.438382625579834,
"learning_rate": 1.7313537016191706e-05,
"loss": 0.0863,
"step": 316
},
{
"epoch": 0.012370731707317072,
"grad_norm": 0.713753342628479,
"learning_rate": 1.7289686274214116e-05,
"loss": 0.0601,
"step": 317
},
{
"epoch": 0.012409756097560976,
"grad_norm": 0.856623113155365,
"learning_rate": 1.7265746709709762e-05,
"loss": 0.0895,
"step": 318
},
{
"epoch": 0.012448780487804878,
"grad_norm": 1.0639945268630981,
"learning_rate": 1.7241718614374678e-05,
"loss": 0.0926,
"step": 319
},
{
"epoch": 0.012487804878048781,
"grad_norm": 1.17190682888031,
"learning_rate": 1.7217602280983622e-05,
"loss": 0.0752,
"step": 320
},
{
"epoch": 0.012526829268292683,
"grad_norm": 19.120393753051758,
"learning_rate": 1.7193398003386514e-05,
"loss": 0.4092,
"step": 321
},
{
"epoch": 0.012565853658536585,
"grad_norm": 0.3367367684841156,
"learning_rate": 1.716910607650483e-05,
"loss": 0.0045,
"step": 322
},
{
"epoch": 0.012604878048780488,
"grad_norm": 1.0379791259765625,
"learning_rate": 1.7144726796328034e-05,
"loss": 0.0331,
"step": 323
},
{
"epoch": 0.01264390243902439,
"grad_norm": 0.7846254110336304,
"learning_rate": 1.712026045990997e-05,
"loss": 0.0564,
"step": 324
},
{
"epoch": 0.012682926829268294,
"grad_norm": 0.7981542348861694,
"learning_rate": 1.709570736536521e-05,
"loss": 0.0736,
"step": 325
},
{
"epoch": 0.012721951219512195,
"grad_norm": 0.4183460772037506,
"learning_rate": 1.7071067811865477e-05,
"loss": 0.0367,
"step": 326
},
{
"epoch": 0.012760975609756097,
"grad_norm": 0.8512915372848511,
"learning_rate": 1.7046342099635948e-05,
"loss": 0.0645,
"step": 327
},
{
"epoch": 0.0128,
"grad_norm": 1.0348637104034424,
"learning_rate": 1.7021530529951627e-05,
"loss": 0.0747,
"step": 328
},
{
"epoch": 0.012839024390243902,
"grad_norm": 0.2513549327850342,
"learning_rate": 1.6996633405133656e-05,
"loss": 0.0028,
"step": 329
},
{
"epoch": 0.012878048780487804,
"grad_norm": 0.9580299854278564,
"learning_rate": 1.697165102854565e-05,
"loss": 0.07,
"step": 330
},
{
"epoch": 0.012917073170731708,
"grad_norm": 2.5003881454467773,
"learning_rate": 1.6946583704589973e-05,
"loss": 0.0847,
"step": 331
},
{
"epoch": 0.01295609756097561,
"grad_norm": 1.1499512195587158,
"learning_rate": 1.692143173870407e-05,
"loss": 0.0632,
"step": 332
},
{
"epoch": 0.012995121951219513,
"grad_norm": 9.275592803955078,
"learning_rate": 1.68961954373567e-05,
"loss": 0.3353,
"step": 333
},
{
"epoch": 0.013034146341463415,
"grad_norm": 0.7970397472381592,
"learning_rate": 1.6870875108044233e-05,
"loss": 0.0685,
"step": 334
},
{
"epoch": 0.013073170731707317,
"grad_norm": 0.8982329964637756,
"learning_rate": 1.684547105928689e-05,
"loss": 0.0408,
"step": 335
},
{
"epoch": 0.01311219512195122,
"grad_norm": 0.5923863053321838,
"learning_rate": 1.6819983600624986e-05,
"loss": 0.072,
"step": 336
},
{
"epoch": 0.013151219512195122,
"grad_norm": 0.25536298751831055,
"learning_rate": 1.6794413042615168e-05,
"loss": 0.0221,
"step": 337
},
{
"epoch": 0.013190243902439024,
"grad_norm": 0.7973102927207947,
"learning_rate": 1.6768759696826608e-05,
"loss": 0.0572,
"step": 338
},
{
"epoch": 0.013229268292682927,
"grad_norm": 0.22800317406654358,
"learning_rate": 1.6743023875837233e-05,
"loss": 0.0024,
"step": 339
},
{
"epoch": 0.013268292682926829,
"grad_norm": 0.5794082880020142,
"learning_rate": 1.6717205893229904e-05,
"loss": 0.0508,
"step": 340
},
{
"epoch": 0.013307317073170732,
"grad_norm": 0.08247140794992447,
"learning_rate": 1.6691306063588583e-05,
"loss": 0.0014,
"step": 341
},
{
"epoch": 0.013346341463414634,
"grad_norm": 0.9693014621734619,
"learning_rate": 1.6665324702494524e-05,
"loss": 0.0605,
"step": 342
},
{
"epoch": 0.013385365853658536,
"grad_norm": 1.6244111061096191,
"learning_rate": 1.6639262126522417e-05,
"loss": 0.0614,
"step": 343
},
{
"epoch": 0.01342439024390244,
"grad_norm": 0.6118417382240295,
"learning_rate": 1.661311865323652e-05,
"loss": 0.0501,
"step": 344
},
{
"epoch": 0.013463414634146341,
"grad_norm": 22.77993392944336,
"learning_rate": 1.6586894601186804e-05,
"loss": 0.5808,
"step": 345
},
{
"epoch": 0.013502439024390243,
"grad_norm": 4.184711933135986,
"learning_rate": 1.6560590289905074e-05,
"loss": 0.1153,
"step": 346
},
{
"epoch": 0.013541463414634147,
"grad_norm": 0.958406925201416,
"learning_rate": 1.6534206039901057e-05,
"loss": 0.0546,
"step": 347
},
{
"epoch": 0.013580487804878048,
"grad_norm": 1.6430609226226807,
"learning_rate": 1.650774217265851e-05,
"loss": 0.0822,
"step": 348
},
{
"epoch": 0.013619512195121952,
"grad_norm": 0.8025966882705688,
"learning_rate": 1.6481199010631312e-05,
"loss": 0.0469,
"step": 349
},
{
"epoch": 0.013658536585365854,
"grad_norm": 0.5775986909866333,
"learning_rate": 1.645457687723951e-05,
"loss": 0.0551,
"step": 350
},
{
"epoch": 0.013697560975609755,
"grad_norm": 0.5034926533699036,
"learning_rate": 1.6427876096865394e-05,
"loss": 0.0509,
"step": 351
},
{
"epoch": 0.013736585365853659,
"grad_norm": 1.4661601781845093,
"learning_rate": 1.6401096994849558e-05,
"loss": 0.088,
"step": 352
},
{
"epoch": 0.01377560975609756,
"grad_norm": 1.0197827816009521,
"learning_rate": 1.63742398974869e-05,
"loss": 0.0651,
"step": 353
},
{
"epoch": 0.013814634146341464,
"grad_norm": 11.487275123596191,
"learning_rate": 1.6347305132022677e-05,
"loss": 0.1029,
"step": 354
},
{
"epoch": 0.013853658536585366,
"grad_norm": 1.9594494104385376,
"learning_rate": 1.632029302664851e-05,
"loss": 0.0645,
"step": 355
},
{
"epoch": 0.013892682926829268,
"grad_norm": 0.19847458600997925,
"learning_rate": 1.6293203910498375e-05,
"loss": 0.0324,
"step": 356
},
{
"epoch": 0.013931707317073171,
"grad_norm": 1.000249981880188,
"learning_rate": 1.6266038113644605e-05,
"loss": 0.0626,
"step": 357
},
{
"epoch": 0.013970731707317073,
"grad_norm": 0.7197471261024475,
"learning_rate": 1.6238795967093865e-05,
"loss": 0.0924,
"step": 358
},
{
"epoch": 0.014009756097560975,
"grad_norm": 0.2946488559246063,
"learning_rate": 1.6211477802783105e-05,
"loss": 0.0485,
"step": 359
},
{
"epoch": 0.014048780487804878,
"grad_norm": 7.890511989593506,
"learning_rate": 1.6184083953575543e-05,
"loss": 0.3441,
"step": 360
},
{
"epoch": 0.01408780487804878,
"grad_norm": 0.6663612127304077,
"learning_rate": 1.6156614753256583e-05,
"loss": 0.0735,
"step": 361
},
{
"epoch": 0.014126829268292684,
"grad_norm": 3.5433099269866943,
"learning_rate": 1.6129070536529767e-05,
"loss": 0.124,
"step": 362
},
{
"epoch": 0.014165853658536585,
"grad_norm": 0.9554763436317444,
"learning_rate": 1.610145163901268e-05,
"loss": 0.1157,
"step": 363
},
{
"epoch": 0.014204878048780487,
"grad_norm": 0.43955886363983154,
"learning_rate": 1.607375839723287e-05,
"loss": 0.0268,
"step": 364
},
{
"epoch": 0.01424390243902439,
"grad_norm": 3.2791476249694824,
"learning_rate": 1.6045991148623752e-05,
"loss": 0.0892,
"step": 365
},
{
"epoch": 0.014282926829268293,
"grad_norm": 1.370173692703247,
"learning_rate": 1.6018150231520486e-05,
"loss": 0.0699,
"step": 366
},
{
"epoch": 0.014321951219512194,
"grad_norm": 0.6852190494537354,
"learning_rate": 1.599023598515586e-05,
"loss": 0.0464,
"step": 367
},
{
"epoch": 0.014360975609756098,
"grad_norm": 26.119176864624023,
"learning_rate": 1.5962248749656158e-05,
"loss": 0.1162,
"step": 368
},
{
"epoch": 0.0144,
"grad_norm": 1.2278575897216797,
"learning_rate": 1.5934188866037017e-05,
"loss": 0.0833,
"step": 369
},
{
"epoch": 0.014439024390243903,
"grad_norm": 1.3735942840576172,
"learning_rate": 1.5906056676199256e-05,
"loss": 0.0718,
"step": 370
},
{
"epoch": 0.014478048780487805,
"grad_norm": 2.3541769981384277,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.0858,
"step": 371
},
{
"epoch": 0.014517073170731707,
"grad_norm": 1.5514541864395142,
"learning_rate": 1.584957674987216e-05,
"loss": 0.0703,
"step": 372
},
{
"epoch": 0.01455609756097561,
"grad_norm": 20.445016860961914,
"learning_rate": 1.5821229701572897e-05,
"loss": 0.3283,
"step": 373
},
{
"epoch": 0.014595121951219512,
"grad_norm": 0.8006591796875,
"learning_rate": 1.5792811723426787e-05,
"loss": 0.0525,
"step": 374
},
{
"epoch": 0.014634146341463415,
"grad_norm": 0.9361777305603027,
"learning_rate": 1.5764323161697933e-05,
"loss": 0.0505,
"step": 375
},
{
"epoch": 0.014673170731707317,
"grad_norm": 0.17962412536144257,
"learning_rate": 1.573576436351046e-05,
"loss": 0.0168,
"step": 376
},
{
"epoch": 0.014712195121951219,
"grad_norm": 2.0104076862335205,
"learning_rate": 1.570713567684432e-05,
"loss": 0.0421,
"step": 377
},
{
"epoch": 0.014751219512195123,
"grad_norm": 1.2183048725128174,
"learning_rate": 1.5678437450531014e-05,
"loss": 0.069,
"step": 378
},
{
"epoch": 0.014790243902439024,
"grad_norm": 0.82366544008255,
"learning_rate": 1.564967003424938e-05,
"loss": 0.0604,
"step": 379
},
{
"epoch": 0.014829268292682926,
"grad_norm": 2.1319451332092285,
"learning_rate": 1.5620833778521306e-05,
"loss": 0.0932,
"step": 380
},
{
"epoch": 0.01486829268292683,
"grad_norm": 0.9145373702049255,
"learning_rate": 1.5591929034707468e-05,
"loss": 0.0655,
"step": 381
},
{
"epoch": 0.014907317073170731,
"grad_norm": 0.7570858001708984,
"learning_rate": 1.556295615500305e-05,
"loss": 0.0696,
"step": 382
},
{
"epoch": 0.014946341463414635,
"grad_norm": 4.875082492828369,
"learning_rate": 1.553391549243344e-05,
"loss": 0.0807,
"step": 383
},
{
"epoch": 0.014985365853658537,
"grad_norm": 0.20908448100090027,
"learning_rate": 1.5504807400849957e-05,
"loss": 0.0288,
"step": 384
},
{
"epoch": 0.015024390243902438,
"grad_norm": 0.615444540977478,
"learning_rate": 1.5475632234925505e-05,
"loss": 0.0247,
"step": 385
},
{
"epoch": 0.015063414634146342,
"grad_norm": 0.7385594248771667,
"learning_rate": 1.5446390350150272e-05,
"loss": 0.0544,
"step": 386
},
{
"epoch": 0.015102439024390244,
"grad_norm": 0.9776281714439392,
"learning_rate": 1.54170821028274e-05,
"loss": 0.0093,
"step": 387
},
{
"epoch": 0.015141463414634146,
"grad_norm": 0.4464304447174072,
"learning_rate": 1.5387707850068633e-05,
"loss": 0.0251,
"step": 388
},
{
"epoch": 0.015180487804878049,
"grad_norm": 1.4949384927749634,
"learning_rate": 1.5358267949789968e-05,
"loss": 0.0766,
"step": 389
},
{
"epoch": 0.01521951219512195,
"grad_norm": 14.084965705871582,
"learning_rate": 1.53287627607073e-05,
"loss": 0.101,
"step": 390
},
{
"epoch": 0.015258536585365854,
"grad_norm": 0.4209037721157074,
"learning_rate": 1.529919264233205e-05,
"loss": 0.0588,
"step": 391
},
{
"epoch": 0.015297560975609756,
"grad_norm": 5.404710292816162,
"learning_rate": 1.5269557954966777e-05,
"loss": 0.1666,
"step": 392
},
{
"epoch": 0.015336585365853658,
"grad_norm": 23.280366897583008,
"learning_rate": 1.5239859059700794e-05,
"loss": 0.7821,
"step": 393
},
{
"epoch": 0.015375609756097561,
"grad_norm": 1.7488294839859009,
"learning_rate": 1.5210096318405768e-05,
"loss": 0.0667,
"step": 394
},
{
"epoch": 0.015414634146341463,
"grad_norm": 0.6551679372787476,
"learning_rate": 1.5180270093731305e-05,
"loss": 0.0225,
"step": 395
},
{
"epoch": 0.015453658536585365,
"grad_norm": 19.76372718811035,
"learning_rate": 1.5150380749100545e-05,
"loss": 0.3256,
"step": 396
},
{
"epoch": 0.015492682926829269,
"grad_norm": 1.006983995437622,
"learning_rate": 1.5120428648705716e-05,
"loss": 0.0585,
"step": 397
},
{
"epoch": 0.01553170731707317,
"grad_norm": 0.9461119174957275,
"learning_rate": 1.5090414157503715e-05,
"loss": 0.0772,
"step": 398
},
{
"epoch": 0.015570731707317074,
"grad_norm": 14.31990909576416,
"learning_rate": 1.5060337641211637e-05,
"loss": 0.1711,
"step": 399
},
{
"epoch": 0.015609756097560976,
"grad_norm": 5.607296943664551,
"learning_rate": 1.5030199466302354e-05,
"loss": 0.1085,
"step": 400
},
{
"epoch": 0.015648780487804877,
"grad_norm": 1.2967677116394043,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.0596,
"step": 401
},
{
"epoch": 0.01568780487804878,
"grad_norm": 0.7188644409179688,
"learning_rate": 1.4969739610275556e-05,
"loss": 0.0244,
"step": 402
},
{
"epoch": 0.015726829268292684,
"grad_norm": 1.3158472776412964,
"learning_rate": 1.493941866584231e-05,
"loss": 0.0718,
"step": 403
},
{
"epoch": 0.015765853658536586,
"grad_norm": 11.034093856811523,
"learning_rate": 1.490903753615141e-05,
"loss": 0.3683,
"step": 404
},
{
"epoch": 0.015804878048780488,
"grad_norm": 15.76632022857666,
"learning_rate": 1.4878596591387329e-05,
"loss": 0.1804,
"step": 405
},
{
"epoch": 0.01584390243902439,
"grad_norm": 5.973936557769775,
"learning_rate": 1.4848096202463373e-05,
"loss": 0.1055,
"step": 406
},
{
"epoch": 0.01588292682926829,
"grad_norm": 1.9282188415527344,
"learning_rate": 1.4817536741017153e-05,
"loss": 0.1188,
"step": 407
},
{
"epoch": 0.015921951219512197,
"grad_norm": 4.105005264282227,
"learning_rate": 1.478691857940607e-05,
"loss": 0.0611,
"step": 408
},
{
"epoch": 0.0159609756097561,
"grad_norm": 0.3579651117324829,
"learning_rate": 1.4756242090702756e-05,
"loss": 0.0052,
"step": 409
},
{
"epoch": 0.016,
"grad_norm": 0.3298042416572571,
"learning_rate": 1.4725507648690542e-05,
"loss": 0.0243,
"step": 410
},
{
"epoch": 0.016039024390243902,
"grad_norm": 13.815900802612305,
"learning_rate": 1.469471562785891e-05,
"loss": 0.2559,
"step": 411
},
{
"epoch": 0.016078048780487804,
"grad_norm": 2.3885419368743896,
"learning_rate": 1.4663866403398915e-05,
"loss": 0.0774,
"step": 412
},
{
"epoch": 0.016117073170731706,
"grad_norm": 0.913438618183136,
"learning_rate": 1.463296035119862e-05,
"loss": 0.0338,
"step": 413
},
{
"epoch": 0.01615609756097561,
"grad_norm": 2.0228524208068848,
"learning_rate": 1.4601997847838518e-05,
"loss": 0.0462,
"step": 414
},
{
"epoch": 0.016195121951219513,
"grad_norm": 4.598243236541748,
"learning_rate": 1.4570979270586944e-05,
"loss": 0.114,
"step": 415
},
{
"epoch": 0.016234146341463414,
"grad_norm": 14.568487167358398,
"learning_rate": 1.4539904997395468e-05,
"loss": 0.5411,
"step": 416
},
{
"epoch": 0.016273170731707316,
"grad_norm": 0.12189683318138123,
"learning_rate": 1.4508775406894308e-05,
"loss": 0.0009,
"step": 417
},
{
"epoch": 0.016312195121951218,
"grad_norm": 18.58679962158203,
"learning_rate": 1.4477590878387697e-05,
"loss": 0.2444,
"step": 418
},
{
"epoch": 0.016351219512195123,
"grad_norm": 0.3928692936897278,
"learning_rate": 1.4446351791849276e-05,
"loss": 0.0523,
"step": 419
},
{
"epoch": 0.016390243902439025,
"grad_norm": 0.3825647830963135,
"learning_rate": 1.4415058527917454e-05,
"loss": 0.0463,
"step": 420
},
{
"epoch": 0.016429268292682927,
"grad_norm": 0.24523496627807617,
"learning_rate": 1.4383711467890776e-05,
"loss": 0.0007,
"step": 421
},
{
"epoch": 0.01646829268292683,
"grad_norm": 0.6316486597061157,
"learning_rate": 1.4352310993723277e-05,
"loss": 0.0448,
"step": 422
},
{
"epoch": 0.01650731707317073,
"grad_norm": 0.5344034433364868,
"learning_rate": 1.4320857488019826e-05,
"loss": 0.061,
"step": 423
},
{
"epoch": 0.016546341463414636,
"grad_norm": 0.24341480433940887,
"learning_rate": 1.4289351334031461e-05,
"loss": 0.0241,
"step": 424
},
{
"epoch": 0.016585365853658537,
"grad_norm": 0.36743006110191345,
"learning_rate": 1.4257792915650728e-05,
"loss": 0.0463,
"step": 425
},
{
"epoch": 0.01662439024390244,
"grad_norm": 0.10371971875429153,
"learning_rate": 1.4226182617406996e-05,
"loss": 0.0007,
"step": 426
},
{
"epoch": 0.01666341463414634,
"grad_norm": 0.9512683153152466,
"learning_rate": 1.4194520824461773e-05,
"loss": 0.0201,
"step": 427
},
{
"epoch": 0.016702439024390243,
"grad_norm": 7.898250579833984,
"learning_rate": 1.4162807922604014e-05,
"loss": 0.1004,
"step": 428
},
{
"epoch": 0.016741463414634148,
"grad_norm": 0.547793447971344,
"learning_rate": 1.413104429824542e-05,
"loss": 0.0504,
"step": 429
},
{
"epoch": 0.01678048780487805,
"grad_norm": 0.6320812702178955,
"learning_rate": 1.4099230338415728e-05,
"loss": 0.0534,
"step": 430
},
{
"epoch": 0.01681951219512195,
"grad_norm": 0.8776715993881226,
"learning_rate": 1.4067366430758004e-05,
"loss": 0.0821,
"step": 431
},
{
"epoch": 0.016858536585365853,
"grad_norm": 0.28817278146743774,
"learning_rate": 1.4035452963523903e-05,
"loss": 0.0404,
"step": 432
},
{
"epoch": 0.016897560975609755,
"grad_norm": 0.7565488815307617,
"learning_rate": 1.4003490325568953e-05,
"loss": 0.0389,
"step": 433
},
{
"epoch": 0.016936585365853657,
"grad_norm": 0.20873385667800903,
"learning_rate": 1.3971478906347806e-05,
"loss": 0.0013,
"step": 434
},
{
"epoch": 0.016975609756097562,
"grad_norm": 0.4725866913795471,
"learning_rate": 1.3939419095909513e-05,
"loss": 0.0636,
"step": 435
},
{
"epoch": 0.017014634146341464,
"grad_norm": 0.4718579947948456,
"learning_rate": 1.3907311284892737e-05,
"loss": 0.0474,
"step": 436
},
{
"epoch": 0.017053658536585366,
"grad_norm": 0.4631275236606598,
"learning_rate": 1.3875155864521031e-05,
"loss": 0.0384,
"step": 437
},
{
"epoch": 0.017092682926829268,
"grad_norm": 1.1784242391586304,
"learning_rate": 1.3842953226598036e-05,
"loss": 0.016,
"step": 438
},
{
"epoch": 0.01713170731707317,
"grad_norm": 0.6740034818649292,
"learning_rate": 1.3810703763502744e-05,
"loss": 0.061,
"step": 439
},
{
"epoch": 0.017170731707317075,
"grad_norm": 0.3610188663005829,
"learning_rate": 1.3778407868184674e-05,
"loss": 0.0418,
"step": 440
},
{
"epoch": 0.017209756097560976,
"grad_norm": 1.5884325504302979,
"learning_rate": 1.3746065934159123e-05,
"loss": 0.0638,
"step": 441
},
{
"epoch": 0.017248780487804878,
"grad_norm": 2.0115463733673096,
"learning_rate": 1.371367835550235e-05,
"loss": 0.0644,
"step": 442
},
{
"epoch": 0.01728780487804878,
"grad_norm": 13.372712135314941,
"learning_rate": 1.3681245526846782e-05,
"loss": 0.3507,
"step": 443
},
{
"epoch": 0.01732682926829268,
"grad_norm": 1.6101962327957153,
"learning_rate": 1.3648767843376196e-05,
"loss": 0.0791,
"step": 444
},
{
"epoch": 0.017365853658536587,
"grad_norm": 3.6703128814697266,
"learning_rate": 1.3616245700820922e-05,
"loss": 0.0661,
"step": 445
},
{
"epoch": 0.01740487804878049,
"grad_norm": 0.19999340176582336,
"learning_rate": 1.3583679495453e-05,
"loss": 0.0023,
"step": 446
},
{
"epoch": 0.01744390243902439,
"grad_norm": 1.1666672229766846,
"learning_rate": 1.3551069624081372e-05,
"loss": 0.0623,
"step": 447
},
{
"epoch": 0.017482926829268292,
"grad_norm": 0.6350935101509094,
"learning_rate": 1.3518416484047018e-05,
"loss": 0.0459,
"step": 448
},
{
"epoch": 0.017521951219512194,
"grad_norm": 0.6683468222618103,
"learning_rate": 1.3485720473218153e-05,
"loss": 0.0598,
"step": 449
},
{
"epoch": 0.0175609756097561,
"grad_norm": 0.6237654685974121,
"learning_rate": 1.3452981989985347e-05,
"loss": 0.0735,
"step": 450
},
{
"epoch": 0.0176,
"grad_norm": 1.2072514295578003,
"learning_rate": 1.342020143325669e-05,
"loss": 0.0142,
"step": 451
},
{
"epoch": 0.017639024390243903,
"grad_norm": 0.8034166097640991,
"learning_rate": 1.3387379202452917e-05,
"loss": 0.0443,
"step": 452
},
{
"epoch": 0.017678048780487805,
"grad_norm": 0.36365169286727905,
"learning_rate": 1.3354515697502552e-05,
"loss": 0.0209,
"step": 453
},
{
"epoch": 0.017717073170731706,
"grad_norm": 1.6496081352233887,
"learning_rate": 1.3321611318837033e-05,
"loss": 0.0851,
"step": 454
},
{
"epoch": 0.017756097560975608,
"grad_norm": 0.6385945081710815,
"learning_rate": 1.3288666467385834e-05,
"loss": 0.0251,
"step": 455
},
{
"epoch": 0.017795121951219513,
"grad_norm": 0.20113211870193481,
"learning_rate": 1.3255681544571568e-05,
"loss": 0.0199,
"step": 456
},
{
"epoch": 0.017834146341463415,
"grad_norm": 1.1561201810836792,
"learning_rate": 1.3222656952305113e-05,
"loss": 0.0552,
"step": 457
},
{
"epoch": 0.017873170731707317,
"grad_norm": 2.386186122894287,
"learning_rate": 1.3189593092980701e-05,
"loss": 0.1009,
"step": 458
},
{
"epoch": 0.01791219512195122,
"grad_norm": 0.2110341489315033,
"learning_rate": 1.3156490369471026e-05,
"loss": 0.0287,
"step": 459
},
{
"epoch": 0.01795121951219512,
"grad_norm": 6.201526641845703,
"learning_rate": 1.3123349185122328e-05,
"loss": 0.1178,
"step": 460
},
{
"epoch": 0.017990243902439026,
"grad_norm": 0.5904516577720642,
"learning_rate": 1.3090169943749475e-05,
"loss": 0.055,
"step": 461
},
{
"epoch": 0.018029268292682928,
"grad_norm": 1.145871877670288,
"learning_rate": 1.3056953049631059e-05,
"loss": 0.0759,
"step": 462
},
{
"epoch": 0.01806829268292683,
"grad_norm": 0.4227178394794464,
"learning_rate": 1.3023698907504447e-05,
"loss": 0.0449,
"step": 463
},
{
"epoch": 0.01810731707317073,
"grad_norm": 0.22431926429271698,
"learning_rate": 1.2990407922560869e-05,
"loss": 0.0047,
"step": 464
},
{
"epoch": 0.018146341463414633,
"grad_norm": 0.4376838207244873,
"learning_rate": 1.2957080500440469e-05,
"loss": 0.003,
"step": 465
},
{
"epoch": 0.018185365853658538,
"grad_norm": 0.06425201892852783,
"learning_rate": 1.2923717047227368e-05,
"loss": 0.0007,
"step": 466
},
{
"epoch": 0.01822439024390244,
"grad_norm": 1.0954526662826538,
"learning_rate": 1.2890317969444716e-05,
"loss": 0.0649,
"step": 467
},
{
"epoch": 0.01826341463414634,
"grad_norm": 0.6397321820259094,
"learning_rate": 1.2856883674049736e-05,
"loss": 0.0711,
"step": 468
},
{
"epoch": 0.018302439024390244,
"grad_norm": 0.4302825927734375,
"learning_rate": 1.2823414568428767e-05,
"loss": 0.0333,
"step": 469
},
{
"epoch": 0.018341463414634145,
"grad_norm": 0.4080619812011719,
"learning_rate": 1.2789911060392295e-05,
"loss": 0.022,
"step": 470
},
{
"epoch": 0.018380487804878047,
"grad_norm": 0.3480626046657562,
"learning_rate": 1.2756373558169992e-05,
"loss": 0.0435,
"step": 471
},
{
"epoch": 0.018419512195121952,
"grad_norm": 18.184669494628906,
"learning_rate": 1.2722802470405744e-05,
"loss": 0.2973,
"step": 472
},
{
"epoch": 0.018458536585365854,
"grad_norm": 0.061509452760219574,
"learning_rate": 1.2689198206152657e-05,
"loss": 0.0008,
"step": 473
},
{
"epoch": 0.018497560975609756,
"grad_norm": 1.0081391334533691,
"learning_rate": 1.265556117486809e-05,
"loss": 0.0319,
"step": 474
},
{
"epoch": 0.018536585365853658,
"grad_norm": 0.29321637749671936,
"learning_rate": 1.2621891786408648e-05,
"loss": 0.03,
"step": 475
},
{
"epoch": 0.01857560975609756,
"grad_norm": 1.0678199529647827,
"learning_rate": 1.2588190451025209e-05,
"loss": 0.1247,
"step": 476
},
{
"epoch": 0.018614634146341465,
"grad_norm": 1.4400922060012817,
"learning_rate": 1.2554457579357906e-05,
"loss": 0.0468,
"step": 477
},
{
"epoch": 0.018653658536585366,
"grad_norm": 0.6483446359634399,
"learning_rate": 1.252069358243114e-05,
"loss": 0.0542,
"step": 478
},
{
"epoch": 0.018692682926829268,
"grad_norm": 1.5308302640914917,
"learning_rate": 1.2486898871648552e-05,
"loss": 0.1241,
"step": 479
},
{
"epoch": 0.01873170731707317,
"grad_norm": 0.6233421564102173,
"learning_rate": 1.2453073858788027e-05,
"loss": 0.0603,
"step": 480
},
{
"epoch": 0.018770731707317072,
"grad_norm": 1.5406410694122314,
"learning_rate": 1.2419218955996677e-05,
"loss": 0.101,
"step": 481
},
{
"epoch": 0.018809756097560977,
"grad_norm": 0.13375365734100342,
"learning_rate": 1.238533457578581e-05,
"loss": 0.0023,
"step": 482
},
{
"epoch": 0.01884878048780488,
"grad_norm": 1.7304381132125854,
"learning_rate": 1.23514211310259e-05,
"loss": 0.0997,
"step": 483
},
{
"epoch": 0.01888780487804878,
"grad_norm": 1.8434371948242188,
"learning_rate": 1.2317479034941572e-05,
"loss": 0.0891,
"step": 484
},
{
"epoch": 0.018926829268292682,
"grad_norm": 0.7466911673545837,
"learning_rate": 1.2283508701106559e-05,
"loss": 0.0575,
"step": 485
},
{
"epoch": 0.018965853658536584,
"grad_norm": 0.18972623348236084,
"learning_rate": 1.2249510543438652e-05,
"loss": 0.0154,
"step": 486
},
{
"epoch": 0.01900487804878049,
"grad_norm": 0.8122363686561584,
"learning_rate": 1.2215484976194675e-05,
"loss": 0.0531,
"step": 487
},
{
"epoch": 0.01904390243902439,
"grad_norm": 0.3305073380470276,
"learning_rate": 1.2181432413965428e-05,
"loss": 0.0327,
"step": 488
},
{
"epoch": 0.019082926829268293,
"grad_norm": 1.1868441104888916,
"learning_rate": 1.2147353271670634e-05,
"loss": 0.0661,
"step": 489
},
{
"epoch": 0.019121951219512195,
"grad_norm": 0.3918513357639313,
"learning_rate": 1.211324796455389e-05,
"loss": 0.0445,
"step": 490
},
{
"epoch": 0.019160975609756097,
"grad_norm": 1.5842230319976807,
"learning_rate": 1.2079116908177592e-05,
"loss": 0.0152,
"step": 491
},
{
"epoch": 0.0192,
"grad_norm": 0.6420873403549194,
"learning_rate": 1.2044960518417902e-05,
"loss": 0.0555,
"step": 492
},
{
"epoch": 0.019239024390243904,
"grad_norm": 0.7307310104370117,
"learning_rate": 1.2010779211459649e-05,
"loss": 0.0528,
"step": 493
},
{
"epoch": 0.019278048780487805,
"grad_norm": 0.5097520351409912,
"learning_rate": 1.1976573403791263e-05,
"loss": 0.0524,
"step": 494
},
{
"epoch": 0.019317073170731707,
"grad_norm": 0.9411758780479431,
"learning_rate": 1.194234351219972e-05,
"loss": 0.0599,
"step": 495
},
{
"epoch": 0.01935609756097561,
"grad_norm": 2.3555338382720947,
"learning_rate": 1.190808995376545e-05,
"loss": 0.1131,
"step": 496
},
{
"epoch": 0.01939512195121951,
"grad_norm": 0.47642064094543457,
"learning_rate": 1.187381314585725e-05,
"loss": 0.0334,
"step": 497
},
{
"epoch": 0.019434146341463416,
"grad_norm": 1.1976745128631592,
"learning_rate": 1.1839513506127202e-05,
"loss": 0.0624,
"step": 498
},
{
"epoch": 0.019473170731707318,
"grad_norm": 0.6874356865882874,
"learning_rate": 1.1805191452505602e-05,
"loss": 0.0523,
"step": 499
},
{
"epoch": 0.01951219512195122,
"grad_norm": 0.637179434299469,
"learning_rate": 1.1770847403195836e-05,
"loss": 0.0501,
"step": 500
},
{
"epoch": 0.01955121951219512,
"grad_norm": 0.9738136529922485,
"learning_rate": 1.1736481776669307e-05,
"loss": 0.0672,
"step": 501
},
{
"epoch": 0.019590243902439023,
"grad_norm": 0.07047262042760849,
"learning_rate": 1.1702094991660326e-05,
"loss": 0.0004,
"step": 502
},
{
"epoch": 0.01962926829268293,
"grad_norm": 0.5129433870315552,
"learning_rate": 1.1667687467161025e-05,
"loss": 0.0418,
"step": 503
},
{
"epoch": 0.01966829268292683,
"grad_norm": 0.2185850888490677,
"learning_rate": 1.1633259622416224e-05,
"loss": 0.0184,
"step": 504
},
{
"epoch": 0.019707317073170732,
"grad_norm": 0.05486743524670601,
"learning_rate": 1.159881187691835e-05,
"loss": 0.0012,
"step": 505
},
{
"epoch": 0.019746341463414634,
"grad_norm": 0.686374843120575,
"learning_rate": 1.156434465040231e-05,
"loss": 0.0539,
"step": 506
},
{
"epoch": 0.019785365853658535,
"grad_norm": 1.2067912817001343,
"learning_rate": 1.1529858362840383e-05,
"loss": 0.0344,
"step": 507
},
{
"epoch": 0.01982439024390244,
"grad_norm": 1.3713297843933105,
"learning_rate": 1.1495353434437098e-05,
"loss": 0.0721,
"step": 508
},
{
"epoch": 0.019863414634146342,
"grad_norm": 0.22708605229854584,
"learning_rate": 1.1460830285624119e-05,
"loss": 0.0165,
"step": 509
},
{
"epoch": 0.019902439024390244,
"grad_norm": 1.5374773740768433,
"learning_rate": 1.1426289337055119e-05,
"loss": 0.0899,
"step": 510
},
{
"epoch": 0.019941463414634146,
"grad_norm": 3.9283974170684814,
"learning_rate": 1.1391731009600655e-05,
"loss": 0.1541,
"step": 511
},
{
"epoch": 0.019980487804878048,
"grad_norm": 0.17813588678836823,
"learning_rate": 1.1357155724343046e-05,
"loss": 0.0404,
"step": 512
},
{
"epoch": 0.02001951219512195,
"grad_norm": 12.337963104248047,
"learning_rate": 1.1322563902571227e-05,
"loss": 0.0941,
"step": 513
},
{
"epoch": 0.020058536585365855,
"grad_norm": 0.43604737520217896,
"learning_rate": 1.128795596577563e-05,
"loss": 0.0535,
"step": 514
},
{
"epoch": 0.020097560975609757,
"grad_norm": 0.4784732162952423,
"learning_rate": 1.1253332335643043e-05,
"loss": 0.0454,
"step": 515
},
{
"epoch": 0.02013658536585366,
"grad_norm": 1.3681074380874634,
"learning_rate": 1.1218693434051475e-05,
"loss": 0.0122,
"step": 516
},
{
"epoch": 0.02017560975609756,
"grad_norm": 1.0038574934005737,
"learning_rate": 1.1184039683065014e-05,
"loss": 0.0706,
"step": 517
},
{
"epoch": 0.020214634146341462,
"grad_norm": 0.20179997384548187,
"learning_rate": 1.1149371504928667e-05,
"loss": 0.0175,
"step": 518
},
{
"epoch": 0.020253658536585367,
"grad_norm": 0.33986353874206543,
"learning_rate": 1.1114689322063255e-05,
"loss": 0.0323,
"step": 519
},
{
"epoch": 0.02029268292682927,
"grad_norm": 0.15242138504981995,
"learning_rate": 1.1079993557060228e-05,
"loss": 0.041,
"step": 520
},
{
"epoch": 0.02033170731707317,
"grad_norm": 1.1438884735107422,
"learning_rate": 1.1045284632676535e-05,
"loss": 0.0582,
"step": 521
},
{
"epoch": 0.020370731707317073,
"grad_norm": 0.05870413780212402,
"learning_rate": 1.1010562971829464e-05,
"loss": 0.001,
"step": 522
},
{
"epoch": 0.020409756097560974,
"grad_norm": 0.18979205191135406,
"learning_rate": 1.0975828997591496e-05,
"loss": 0.0237,
"step": 523
},
{
"epoch": 0.02044878048780488,
"grad_norm": 5.487180709838867,
"learning_rate": 1.0941083133185146e-05,
"loss": 0.0909,
"step": 524
},
{
"epoch": 0.02048780487804878,
"grad_norm": 0.6563659906387329,
"learning_rate": 1.0906325801977804e-05,
"loss": 0.0607,
"step": 525
},
{
"epoch": 0.020526829268292683,
"grad_norm": 0.7054755091667175,
"learning_rate": 1.0871557427476585e-05,
"loss": 0.0689,
"step": 526
},
{
"epoch": 0.020565853658536585,
"grad_norm": 0.20347629487514496,
"learning_rate": 1.083677843332316e-05,
"loss": 0.0299,
"step": 527
},
{
"epoch": 0.020604878048780487,
"grad_norm": 0.8572469353675842,
"learning_rate": 1.0801989243288588e-05,
"loss": 0.0511,
"step": 528
},
{
"epoch": 0.020643902439024392,
"grad_norm": 0.3179641366004944,
"learning_rate": 1.0767190281268187e-05,
"loss": 0.0275,
"step": 529
},
{
"epoch": 0.020682926829268294,
"grad_norm": 0.3560505509376526,
"learning_rate": 1.0732381971276318e-05,
"loss": 0.0027,
"step": 530
},
{
"epoch": 0.020721951219512195,
"grad_norm": 1.8273065090179443,
"learning_rate": 1.0697564737441254e-05,
"loss": 0.0772,
"step": 531
},
{
"epoch": 0.020760975609756097,
"grad_norm": 1.628716230392456,
"learning_rate": 1.0662739004000005e-05,
"loss": 0.1436,
"step": 532
},
{
"epoch": 0.0208,
"grad_norm": 0.2696378529071808,
"learning_rate": 1.0627905195293135e-05,
"loss": 0.0274,
"step": 533
},
{
"epoch": 0.0208390243902439,
"grad_norm": 0.1852053999900818,
"learning_rate": 1.0593063735759619e-05,
"loss": 0.0151,
"step": 534
},
{
"epoch": 0.020878048780487806,
"grad_norm": 1.2610079050064087,
"learning_rate": 1.055821504993164e-05,
"loss": 0.0607,
"step": 535
},
{
"epoch": 0.020917073170731708,
"grad_norm": 1.4779349565505981,
"learning_rate": 1.0523359562429441e-05,
"loss": 0.0509,
"step": 536
},
{
"epoch": 0.02095609756097561,
"grad_norm": 1.5421380996704102,
"learning_rate": 1.0488497697956134e-05,
"loss": 0.0586,
"step": 537
},
{
"epoch": 0.02099512195121951,
"grad_norm": 0.3786073625087738,
"learning_rate": 1.0453629881292537e-05,
"loss": 0.0408,
"step": 538
},
{
"epoch": 0.021034146341463413,
"grad_norm": 1.097656011581421,
"learning_rate": 1.0418756537291996e-05,
"loss": 0.1187,
"step": 539
},
{
"epoch": 0.02107317073170732,
"grad_norm": 0.1372087001800537,
"learning_rate": 1.03838780908752e-05,
"loss": 0.0009,
"step": 540
},
{
"epoch": 0.02111219512195122,
"grad_norm": 4.703077793121338,
"learning_rate": 1.0348994967025012e-05,
"loss": 0.0869,
"step": 541
},
{
"epoch": 0.021151219512195122,
"grad_norm": 9.926535606384277,
"learning_rate": 1.0314107590781284e-05,
"loss": 0.1128,
"step": 542
},
{
"epoch": 0.021190243902439024,
"grad_norm": 0.28871986269950867,
"learning_rate": 1.0279216387235691e-05,
"loss": 0.0304,
"step": 543
},
{
"epoch": 0.021229268292682926,
"grad_norm": 1.8993171453475952,
"learning_rate": 1.0244321781526533e-05,
"loss": 0.0704,
"step": 544
},
{
"epoch": 0.02126829268292683,
"grad_norm": 0.044724393635988235,
"learning_rate": 1.0209424198833571e-05,
"loss": 0.0014,
"step": 545
},
{
"epoch": 0.021307317073170733,
"grad_norm": 0.6550346612930298,
"learning_rate": 1.0174524064372837e-05,
"loss": 0.0524,
"step": 546
},
{
"epoch": 0.021346341463414634,
"grad_norm": 0.04702272266149521,
"learning_rate": 1.0139621803391454e-05,
"loss": 0.0014,
"step": 547
},
{
"epoch": 0.021385365853658536,
"grad_norm": 0.9153466820716858,
"learning_rate": 1.010471784116246e-05,
"loss": 0.0708,
"step": 548
},
{
"epoch": 0.021424390243902438,
"grad_norm": 0.7554747462272644,
"learning_rate": 1.0069812602979617e-05,
"loss": 0.0461,
"step": 549
},
{
"epoch": 0.021463414634146343,
"grad_norm": 23.34663200378418,
"learning_rate": 1.0034906514152239e-05,
"loss": 0.5614,
"step": 550
},
{
"epoch": 0.021502439024390245,
"grad_norm": 0.15118105709552765,
"learning_rate": 1e-05,
"loss": 0.014,
"step": 551
},
{
"epoch": 0.021541463414634147,
"grad_norm": 0.45870324969291687,
"learning_rate": 9.965093485847766e-06,
"loss": 0.0481,
"step": 552
},
{
"epoch": 0.02158048780487805,
"grad_norm": 6.843705654144287,
"learning_rate": 9.930187397020385e-06,
"loss": 0.0517,
"step": 553
},
{
"epoch": 0.02161951219512195,
"grad_norm": 0.5138716101646423,
"learning_rate": 9.895282158837545e-06,
"loss": 0.049,
"step": 554
},
{
"epoch": 0.021658536585365852,
"grad_norm": 0.27236583828926086,
"learning_rate": 9.860378196608549e-06,
"loss": 0.0417,
"step": 555
},
{
"epoch": 0.021697560975609757,
"grad_norm": 0.48508819937705994,
"learning_rate": 9.825475935627165e-06,
"loss": 0.0405,
"step": 556
},
{
"epoch": 0.02173658536585366,
"grad_norm": 0.4914189279079437,
"learning_rate": 9.790575801166432e-06,
"loss": 0.0466,
"step": 557
},
{
"epoch": 0.02177560975609756,
"grad_norm": 0.047320615500211716,
"learning_rate": 9.75567821847347e-06,
"loss": 0.0004,
"step": 558
},
{
"epoch": 0.021814634146341463,
"grad_norm": 0.16130925714969635,
"learning_rate": 9.720783612764314e-06,
"loss": 0.0147,
"step": 559
},
{
"epoch": 0.021853658536585364,
"grad_norm": 0.42710769176483154,
"learning_rate": 9.685892409218718e-06,
"loss": 0.0564,
"step": 560
},
{
"epoch": 0.02189268292682927,
"grad_norm": 0.4221307337284088,
"learning_rate": 9.651005032974994e-06,
"loss": 0.0399,
"step": 561
},
{
"epoch": 0.02193170731707317,
"grad_norm": 0.39911699295043945,
"learning_rate": 9.616121909124801e-06,
"loss": 0.0586,
"step": 562
},
{
"epoch": 0.021970731707317073,
"grad_norm": 0.04985790699720383,
"learning_rate": 9.581243462708007e-06,
"loss": 0.0006,
"step": 563
},
{
"epoch": 0.022009756097560975,
"grad_norm": 13.271845817565918,
"learning_rate": 9.546370118707463e-06,
"loss": 0.2957,
"step": 564
},
{
"epoch": 0.022048780487804877,
"grad_norm": 1.0403120517730713,
"learning_rate": 9.511502302043867e-06,
"loss": 0.077,
"step": 565
},
{
"epoch": 0.022087804878048782,
"grad_norm": 1.5088798999786377,
"learning_rate": 9.476640437570562e-06,
"loss": 0.0576,
"step": 566
},
{
"epoch": 0.022126829268292684,
"grad_norm": 0.41252589225769043,
"learning_rate": 9.441784950068362e-06,
"loss": 0.0505,
"step": 567
},
{
"epoch": 0.022165853658536586,
"grad_norm": 0.2599889039993286,
"learning_rate": 9.406936264240386e-06,
"loss": 0.0325,
"step": 568
},
{
"epoch": 0.022204878048780487,
"grad_norm": 0.7953629493713379,
"learning_rate": 9.372094804706867e-06,
"loss": 0.0406,
"step": 569
},
{
"epoch": 0.02224390243902439,
"grad_norm": 0.3637285530567169,
"learning_rate": 9.337260996000002e-06,
"loss": 0.0002,
"step": 570
},
{
"epoch": 0.02228292682926829,
"grad_norm": 0.8682350516319275,
"learning_rate": 9.302435262558748e-06,
"loss": 0.0458,
"step": 571
},
{
"epoch": 0.022321951219512196,
"grad_norm": 0.21154120564460754,
"learning_rate": 9.267618028723687e-06,
"loss": 0.0004,
"step": 572
},
{
"epoch": 0.022360975609756098,
"grad_norm": 1.1550549268722534,
"learning_rate": 9.232809718731815e-06,
"loss": 0.0698,
"step": 573
},
{
"epoch": 0.0224,
"grad_norm": 0.4845518171787262,
"learning_rate": 9.198010756711413e-06,
"loss": 0.0432,
"step": 574
},
{
"epoch": 0.0224390243902439,
"grad_norm": 0.31368234753608704,
"learning_rate": 9.163221566676847e-06,
"loss": 0.0404,
"step": 575
},
{
"epoch": 0.022478048780487803,
"grad_norm": 0.7459965944290161,
"learning_rate": 9.128442572523418e-06,
"loss": 0.0492,
"step": 576
},
{
"epoch": 0.02251707317073171,
"grad_norm": 0.3097609281539917,
"learning_rate": 9.093674198022201e-06,
"loss": 0.0382,
"step": 577
},
{
"epoch": 0.02255609756097561,
"grad_norm": 0.05860842764377594,
"learning_rate": 9.058916866814857e-06,
"loss": 0.0006,
"step": 578
},
{
"epoch": 0.022595121951219512,
"grad_norm": 0.32315728068351746,
"learning_rate": 9.024171002408507e-06,
"loss": 0.0308,
"step": 579
},
{
"epoch": 0.022634146341463414,
"grad_norm": 0.9359681010246277,
"learning_rate": 8.989437028170537e-06,
"loss": 0.0551,
"step": 580
},
{
"epoch": 0.022673170731707316,
"grad_norm": 0.35661354660987854,
"learning_rate": 8.954715367323468e-06,
"loss": 0.0469,
"step": 581
},
{
"epoch": 0.02271219512195122,
"grad_norm": 1.917565941810608,
"learning_rate": 8.920006442939772e-06,
"loss": 0.0739,
"step": 582
},
{
"epoch": 0.022751219512195123,
"grad_norm": 0.504372239112854,
"learning_rate": 8.885310677936746e-06,
"loss": 0.025,
"step": 583
},
{
"epoch": 0.022790243902439025,
"grad_norm": 0.29174262285232544,
"learning_rate": 8.850628495071336e-06,
"loss": 0.0048,
"step": 584
},
{
"epoch": 0.022829268292682926,
"grad_norm": 0.3026125133037567,
"learning_rate": 8.815960316934991e-06,
"loss": 0.0252,
"step": 585
},
{
"epoch": 0.022868292682926828,
"grad_norm": 0.6295485496520996,
"learning_rate": 8.781306565948528e-06,
"loss": 0.021,
"step": 586
},
{
"epoch": 0.022907317073170733,
"grad_norm": 0.07715345174074173,
"learning_rate": 8.746667664356957e-06,
"loss": 0.0015,
"step": 587
},
{
"epoch": 0.022946341463414635,
"grad_norm": 0.31300753355026245,
"learning_rate": 8.712044034224374e-06,
"loss": 0.0292,
"step": 588
},
{
"epoch": 0.022985365853658537,
"grad_norm": 0.2406529039144516,
"learning_rate": 8.677436097428775e-06,
"loss": 0.0165,
"step": 589
},
{
"epoch": 0.02302439024390244,
"grad_norm": 0.2978267967700958,
"learning_rate": 8.642844275656957e-06,
"loss": 0.0462,
"step": 590
},
{
"epoch": 0.02306341463414634,
"grad_norm": 1.345794916152954,
"learning_rate": 8.60826899039935e-06,
"loss": 0.0528,
"step": 591
},
{
"epoch": 0.023102439024390242,
"grad_norm": 0.14432461559772491,
"learning_rate": 8.573710662944884e-06,
"loss": 0.0398,
"step": 592
},
{
"epoch": 0.023141463414634147,
"grad_norm": 1.0637136697769165,
"learning_rate": 8.539169714375885e-06,
"loss": 0.0727,
"step": 593
},
{
"epoch": 0.02318048780487805,
"grad_norm": 1.6893724203109741,
"learning_rate": 8.504646565562907e-06,
"loss": 0.0291,
"step": 594
},
{
"epoch": 0.02321951219512195,
"grad_norm": 0.588715136051178,
"learning_rate": 8.47014163715962e-06,
"loss": 0.0441,
"step": 595
},
{
"epoch": 0.023258536585365853,
"grad_norm": 1.1892492771148682,
"learning_rate": 8.43565534959769e-06,
"loss": 0.0894,
"step": 596
},
{
"epoch": 0.023297560975609755,
"grad_norm": 0.1478816419839859,
"learning_rate": 8.401188123081653e-06,
"loss": 0.0192,
"step": 597
},
{
"epoch": 0.02333658536585366,
"grad_norm": 19.229629516601562,
"learning_rate": 8.366740377583781e-06,
"loss": 0.1443,
"step": 598
},
{
"epoch": 0.02337560975609756,
"grad_norm": 0.8096135258674622,
"learning_rate": 8.332312532838978e-06,
"loss": 0.058,
"step": 599
},
{
"epoch": 0.023414634146341463,
"grad_norm": 2.723881721496582,
"learning_rate": 8.297905008339677e-06,
"loss": 0.11,
"step": 600
},
{
"epoch": 0.023453658536585365,
"grad_norm": 6.2353515625,
"learning_rate": 8.263518223330698e-06,
"loss": 0.088,
"step": 601
},
{
"epoch": 0.023492682926829267,
"grad_norm": 0.15220575034618378,
"learning_rate": 8.22915259680417e-06,
"loss": 0.0302,
"step": 602
},
{
"epoch": 0.023531707317073172,
"grad_norm": 0.1424236297607422,
"learning_rate": 8.194808547494401e-06,
"loss": 0.0134,
"step": 603
},
{
"epoch": 0.023570731707317074,
"grad_norm": 1.4514447450637817,
"learning_rate": 8.1604864938728e-06,
"loss": 0.0909,
"step": 604
},
{
"epoch": 0.023609756097560976,
"grad_norm": 1.7440004348754883,
"learning_rate": 8.126186854142752e-06,
"loss": 0.061,
"step": 605
},
{
"epoch": 0.023648780487804878,
"grad_norm": 0.5406599640846252,
"learning_rate": 8.091910046234552e-06,
"loss": 0.0345,
"step": 606
},
{
"epoch": 0.02368780487804878,
"grad_norm": 0.13408121466636658,
"learning_rate": 8.057656487800283e-06,
"loss": 0.0134,
"step": 607
},
{
"epoch": 0.023726829268292685,
"grad_norm": 0.46480971574783325,
"learning_rate": 8.023426596208739e-06,
"loss": 0.0499,
"step": 608
},
{
"epoch": 0.023765853658536586,
"grad_norm": 0.4142056703567505,
"learning_rate": 7.989220788540356e-06,
"loss": 0.0364,
"step": 609
},
{
"epoch": 0.023804878048780488,
"grad_norm": 1.00434410572052,
"learning_rate": 7.955039481582098e-06,
"loss": 0.099,
"step": 610
},
{
"epoch": 0.02384390243902439,
"grad_norm": 0.22845646739006042,
"learning_rate": 7.92088309182241e-06,
"loss": 0.0164,
"step": 611
},
{
"epoch": 0.02388292682926829,
"grad_norm": 0.36398613452911377,
"learning_rate": 7.886752035446116e-06,
"loss": 0.02,
"step": 612
},
{
"epoch": 0.023921951219512193,
"grad_norm": 0.40852734446525574,
"learning_rate": 7.852646728329368e-06,
"loss": 0.0537,
"step": 613
},
{
"epoch": 0.0239609756097561,
"grad_norm": 1.0155836343765259,
"learning_rate": 7.818567586034578e-06,
"loss": 0.077,
"step": 614
},
{
"epoch": 0.024,
"grad_norm": 0.10167985409498215,
"learning_rate": 7.784515023805328e-06,
"loss": 0.0011,
"step": 615
},
{
"epoch": 0.024039024390243902,
"grad_norm": 0.9250698685646057,
"learning_rate": 7.750489456561351e-06,
"loss": 0.059,
"step": 616
},
{
"epoch": 0.024078048780487804,
"grad_norm": 1.5903898477554321,
"learning_rate": 7.716491298893443e-06,
"loss": 0.0599,
"step": 617
},
{
"epoch": 0.024117073170731706,
"grad_norm": 0.7739226818084717,
"learning_rate": 7.68252096505843e-06,
"loss": 0.044,
"step": 618
},
{
"epoch": 0.02415609756097561,
"grad_norm": 0.3847982585430145,
"learning_rate": 7.6485788689741e-06,
"loss": 0.0267,
"step": 619
},
{
"epoch": 0.024195121951219513,
"grad_norm": 0.024815106764435768,
"learning_rate": 7.6146654242141935e-06,
"loss": 0.0009,
"step": 620
},
{
"epoch": 0.024234146341463415,
"grad_norm": 0.6505560278892517,
"learning_rate": 7.580781044003324e-06,
"loss": 0.07,
"step": 621
},
{
"epoch": 0.024273170731707316,
"grad_norm": 0.049012988805770874,
"learning_rate": 7.546926141211975e-06,
"loss": 0.0003,
"step": 622
},
{
"epoch": 0.024312195121951218,
"grad_norm": 1.9184240102767944,
"learning_rate": 7.513101128351454e-06,
"loss": 0.0996,
"step": 623
},
{
"epoch": 0.024351219512195123,
"grad_norm": 0.11216408014297485,
"learning_rate": 7.4793064175688635e-06,
"loss": 0.0242,
"step": 624
},
{
"epoch": 0.024390243902439025,
"grad_norm": 0.4607756733894348,
"learning_rate": 7.445542420642097e-06,
"loss": 0.0481,
"step": 625
},
{
"epoch": 0.024429268292682927,
"grad_norm": 0.1728745996952057,
"learning_rate": 7.411809548974792e-06,
"loss": 0.014,
"step": 626
},
{
"epoch": 0.02446829268292683,
"grad_norm": 2.9097442626953125,
"learning_rate": 7.378108213591355e-06,
"loss": 0.0707,
"step": 627
},
{
"epoch": 0.02450731707317073,
"grad_norm": 0.612389326095581,
"learning_rate": 7.344438825131912e-06,
"loss": 0.0575,
"step": 628
},
{
"epoch": 0.024546341463414636,
"grad_norm": 0.3870827853679657,
"learning_rate": 7.310801793847344e-06,
"loss": 0.0139,
"step": 629
},
{
"epoch": 0.024585365853658538,
"grad_norm": 0.5372679829597473,
"learning_rate": 7.277197529594257e-06,
"loss": 0.0375,
"step": 630
},
{
"epoch": 0.02462439024390244,
"grad_norm": 0.16969643533229828,
"learning_rate": 7.243626441830009e-06,
"loss": 0.0178,
"step": 631
},
{
"epoch": 0.02466341463414634,
"grad_norm": 0.5604158043861389,
"learning_rate": 7.210088939607709e-06,
"loss": 0.0551,
"step": 632
},
{
"epoch": 0.024702439024390243,
"grad_norm": 0.2329418808221817,
"learning_rate": 7.176585431571235e-06,
"loss": 0.0303,
"step": 633
},
{
"epoch": 0.024741463414634145,
"grad_norm": 5.765641689300537,
"learning_rate": 7.143116325950266e-06,
"loss": 0.1118,
"step": 634
},
{
"epoch": 0.02478048780487805,
"grad_norm": 1.0145319700241089,
"learning_rate": 7.109682030555283e-06,
"loss": 0.0648,
"step": 635
},
{
"epoch": 0.024819512195121952,
"grad_norm": 0.19974470138549805,
"learning_rate": 7.076282952772634e-06,
"loss": 0.0244,
"step": 636
},
{
"epoch": 0.024858536585365854,
"grad_norm": 0.031927406787872314,
"learning_rate": 7.042919499559538e-06,
"loss": 0.0005,
"step": 637
},
{
"epoch": 0.024897560975609755,
"grad_norm": 0.2905659079551697,
"learning_rate": 7.009592077439135e-06,
"loss": 0.0358,
"step": 638
},
{
"epoch": 0.024936585365853657,
"grad_norm": 0.5015307664871216,
"learning_rate": 6.976301092495556e-06,
"loss": 0.0564,
"step": 639
},
{
"epoch": 0.024975609756097562,
"grad_norm": 0.2731625437736511,
"learning_rate": 6.943046950368944e-06,
"loss": 0.0405,
"step": 640
},
{
"epoch": 0.025014634146341464,
"grad_norm": 0.1554478108882904,
"learning_rate": 6.909830056250527e-06,
"loss": 0.0185,
"step": 641
},
{
"epoch": 0.025053658536585366,
"grad_norm": 0.8015327453613281,
"learning_rate": 6.876650814877675e-06,
"loss": 0.0059,
"step": 642
},
{
"epoch": 0.025092682926829268,
"grad_norm": 1.0645581483840942,
"learning_rate": 6.843509630528977e-06,
"loss": 0.0744,
"step": 643
},
{
"epoch": 0.02513170731707317,
"grad_norm": 0.35431545972824097,
"learning_rate": 6.8104069070193e-06,
"loss": 0.0385,
"step": 644
},
{
"epoch": 0.025170731707317075,
"grad_norm": 3.227726459503174,
"learning_rate": 6.777343047694891e-06,
"loss": 0.0649,
"step": 645
},
{
"epoch": 0.025209756097560976,
"grad_norm": 0.44207656383514404,
"learning_rate": 6.744318455428436e-06,
"loss": 0.0345,
"step": 646
},
{
"epoch": 0.02524878048780488,
"grad_norm": 0.09259597212076187,
"learning_rate": 6.711333532614168e-06,
"loss": 0.0129,
"step": 647
},
{
"epoch": 0.02528780487804878,
"grad_norm": 0.05886169150471687,
"learning_rate": 6.67838868116297e-06,
"loss": 0.0006,
"step": 648
},
{
"epoch": 0.025326829268292682,
"grad_norm": 0.222504124045372,
"learning_rate": 6.645484302497452e-06,
"loss": 0.0168,
"step": 649
},
{
"epoch": 0.025365853658536587,
"grad_norm": 0.2820782959461212,
"learning_rate": 6.612620797547087e-06,
"loss": 0.0189,
"step": 650
},
{
"epoch": 0.02540487804878049,
"grad_norm": 0.6021156907081604,
"learning_rate": 6.579798566743314e-06,
"loss": 0.0204,
"step": 651
},
{
"epoch": 0.02544390243902439,
"grad_norm": 0.2560287117958069,
"learning_rate": 6.547018010014654e-06,
"loss": 0.044,
"step": 652
},
{
"epoch": 0.025482926829268292,
"grad_norm": 0.2795514166355133,
"learning_rate": 6.5142795267818505e-06,
"loss": 0.0466,
"step": 653
},
{
"epoch": 0.025521951219512194,
"grad_norm": 0.6416069269180298,
"learning_rate": 6.481583515952983e-06,
"loss": 0.0384,
"step": 654
},
{
"epoch": 0.025560975609756096,
"grad_norm": 1.3182569742202759,
"learning_rate": 6.448930375918632e-06,
"loss": 0.034,
"step": 655
},
{
"epoch": 0.0256,
"grad_norm": 0.13717637956142426,
"learning_rate": 6.4163205045469975e-06,
"loss": 0.0285,
"step": 656
},
{
"epoch": 0.025639024390243903,
"grad_norm": 0.2880706191062927,
"learning_rate": 6.383754299179079e-06,
"loss": 0.0096,
"step": 657
},
{
"epoch": 0.025678048780487805,
"grad_norm": 0.6714202761650085,
"learning_rate": 6.351232156623803e-06,
"loss": 0.0492,
"step": 658
},
{
"epoch": 0.025717073170731707,
"grad_norm": 0.19599796831607819,
"learning_rate": 6.318754473153221e-06,
"loss": 0.0084,
"step": 659
},
{
"epoch": 0.02575609756097561,
"grad_norm": 0.8548691272735596,
"learning_rate": 6.286321644497655e-06,
"loss": 0.0671,
"step": 660
},
{
"epoch": 0.025795121951219514,
"grad_norm": 0.53550124168396,
"learning_rate": 6.25393406584088e-06,
"loss": 0.0332,
"step": 661
},
{
"epoch": 0.025834146341463415,
"grad_norm": 0.44911351799964905,
"learning_rate": 6.22159213181533e-06,
"loss": 0.0288,
"step": 662
},
{
"epoch": 0.025873170731707317,
"grad_norm": 1.8297635316848755,
"learning_rate": 6.18929623649726e-06,
"loss": 0.0605,
"step": 663
},
{
"epoch": 0.02591219512195122,
"grad_norm": 0.38161206245422363,
"learning_rate": 6.157046773401964e-06,
"loss": 0.0437,
"step": 664
},
{
"epoch": 0.02595121951219512,
"grad_norm": 1.1536452770233154,
"learning_rate": 6.124844135478971e-06,
"loss": 0.0662,
"step": 665
},
{
"epoch": 0.025990243902439026,
"grad_norm": 0.2946390211582184,
"learning_rate": 6.092688715107265e-06,
"loss": 0.0205,
"step": 666
},
{
"epoch": 0.026029268292682928,
"grad_norm": 0.33534473180770874,
"learning_rate": 6.06058090409049e-06,
"loss": 0.0413,
"step": 667
},
{
"epoch": 0.02606829268292683,
"grad_norm": 1.0039939880371094,
"learning_rate": 6.028521093652195e-06,
"loss": 0.0261,
"step": 668
},
{
"epoch": 0.02610731707317073,
"grad_norm": 0.9835073947906494,
"learning_rate": 5.996509674431053e-06,
"loss": 0.0556,
"step": 669
},
{
"epoch": 0.026146341463414633,
"grad_norm": 0.8067487478256226,
"learning_rate": 5.9645470364761e-06,
"loss": 0.0483,
"step": 670
},
{
"epoch": 0.026185365853658535,
"grad_norm": 0.1388138234615326,
"learning_rate": 5.932633569242e-06,
"loss": 0.0282,
"step": 671
},
{
"epoch": 0.02622439024390244,
"grad_norm": 0.29007747769355774,
"learning_rate": 5.900769661584273e-06,
"loss": 0.0439,
"step": 672
},
{
"epoch": 0.026263414634146342,
"grad_norm": 0.4608079493045807,
"learning_rate": 5.868955701754584e-06,
"loss": 0.043,
"step": 673
},
{
"epoch": 0.026302439024390244,
"grad_norm": 1.7731486558914185,
"learning_rate": 5.83719207739599e-06,
"loss": 0.1131,
"step": 674
},
{
"epoch": 0.026341463414634145,
"grad_norm": 0.09589545428752899,
"learning_rate": 5.8054791755382286e-06,
"loss": 0.0174,
"step": 675
},
{
"epoch": 0.026380487804878047,
"grad_norm": 0.21543467044830322,
"learning_rate": 5.773817382593008e-06,
"loss": 0.0265,
"step": 676
},
{
"epoch": 0.026419512195121952,
"grad_norm": 0.25877299904823303,
"learning_rate": 5.742207084349274e-06,
"loss": 0.0057,
"step": 677
},
{
"epoch": 0.026458536585365854,
"grad_norm": 0.22479064762592316,
"learning_rate": 5.710648665968543e-06,
"loss": 0.0184,
"step": 678
},
{
"epoch": 0.026497560975609756,
"grad_norm": 0.7022895216941833,
"learning_rate": 5.679142511980176e-06,
"loss": 0.0411,
"step": 679
},
{
"epoch": 0.026536585365853658,
"grad_norm": 0.3547297716140747,
"learning_rate": 5.647689006276727e-06,
"loss": 0.0472,
"step": 680
},
{
"epoch": 0.02657560975609756,
"grad_norm": 0.36916905641555786,
"learning_rate": 5.616288532109225e-06,
"loss": 0.0186,
"step": 681
},
{
"epoch": 0.026614634146341465,
"grad_norm": 0.06366704404354095,
"learning_rate": 5.584941472082549e-06,
"loss": 0.012,
"step": 682
},
{
"epoch": 0.026653658536585367,
"grad_norm": 0.7081243395805359,
"learning_rate": 5.553648208150728e-06,
"loss": 0.0453,
"step": 683
},
{
"epoch": 0.02669268292682927,
"grad_norm": 1.618084192276001,
"learning_rate": 5.522409121612304e-06,
"loss": 0.0345,
"step": 684
},
{
"epoch": 0.02673170731707317,
"grad_norm": 0.9531549215316772,
"learning_rate": 5.491224593105695e-06,
"loss": 0.0446,
"step": 685
},
{
"epoch": 0.026770731707317072,
"grad_norm": 0.24412359297275543,
"learning_rate": 5.460095002604533e-06,
"loss": 0.0272,
"step": 686
},
{
"epoch": 0.026809756097560977,
"grad_norm": 0.976974368095398,
"learning_rate": 5.429020729413062e-06,
"loss": 0.0591,
"step": 687
},
{
"epoch": 0.02684878048780488,
"grad_norm": 0.4555303752422333,
"learning_rate": 5.398002152161484e-06,
"loss": 0.0375,
"step": 688
},
{
"epoch": 0.02688780487804878,
"grad_norm": 0.6136086583137512,
"learning_rate": 5.367039648801386e-06,
"loss": 0.0585,
"step": 689
},
{
"epoch": 0.026926829268292683,
"grad_norm": 0.05675547197461128,
"learning_rate": 5.336133596601089e-06,
"loss": 0.0008,
"step": 690
},
{
"epoch": 0.026965853658536584,
"grad_norm": 0.20970018208026886,
"learning_rate": 5.305284372141095e-06,
"loss": 0.0172,
"step": 691
},
{
"epoch": 0.027004878048780486,
"grad_norm": 0.7311350703239441,
"learning_rate": 5.274492351309462e-06,
"loss": 0.0467,
"step": 692
},
{
"epoch": 0.02704390243902439,
"grad_norm": 1.0485392808914185,
"learning_rate": 5.243757909297247e-06,
"loss": 0.0437,
"step": 693
},
{
"epoch": 0.027082926829268293,
"grad_norm": 0.9493411779403687,
"learning_rate": 5.213081420593933e-06,
"loss": 0.0564,
"step": 694
},
{
"epoch": 0.027121951219512195,
"grad_norm": 2.038564443588257,
"learning_rate": 5.1824632589828465e-06,
"loss": 0.0239,
"step": 695
},
{
"epoch": 0.027160975609756097,
"grad_norm": 1.2097431421279907,
"learning_rate": 5.151903797536631e-06,
"loss": 0.0842,
"step": 696
},
{
"epoch": 0.0272,
"grad_norm": 1.0269906520843506,
"learning_rate": 5.121403408612672e-06,
"loss": 0.0256,
"step": 697
},
{
"epoch": 0.027239024390243904,
"grad_norm": 0.3099774718284607,
"learning_rate": 5.090962463848592e-06,
"loss": 0.0348,
"step": 698
},
{
"epoch": 0.027278048780487806,
"grad_norm": 0.3374386131763458,
"learning_rate": 5.060581334157693e-06,
"loss": 0.0437,
"step": 699
},
{
"epoch": 0.027317073170731707,
"grad_norm": 34.19160461425781,
"learning_rate": 5.030260389724447e-06,
"loss": 0.7619,
"step": 700
},
{
"epoch": 0.02735609756097561,
"grad_norm": 0.5533769130706787,
"learning_rate": 5.000000000000003e-06,
"loss": 0.0515,
"step": 701
},
{
"epoch": 0.02739512195121951,
"grad_norm": 0.42525023221969604,
"learning_rate": 4.96980053369765e-06,
"loss": 0.0545,
"step": 702
},
{
"epoch": 0.027434146341463416,
"grad_norm": 14.860175132751465,
"learning_rate": 4.939662358788364e-06,
"loss": 0.1756,
"step": 703
},
{
"epoch": 0.027473170731707318,
"grad_norm": 0.18521559238433838,
"learning_rate": 4.909585842496287e-06,
"loss": 0.0399,
"step": 704
},
{
"epoch": 0.02751219512195122,
"grad_norm": 0.8066133856773376,
"learning_rate": 4.879571351294287e-06,
"loss": 0.0487,
"step": 705
},
{
"epoch": 0.02755121951219512,
"grad_norm": 0.26865291595458984,
"learning_rate": 4.849619250899458e-06,
"loss": 0.0148,
"step": 706
},
{
"epoch": 0.027590243902439023,
"grad_norm": 0.5729135870933533,
"learning_rate": 4.8197299062687e-06,
"loss": 0.0207,
"step": 707
},
{
"epoch": 0.02762926829268293,
"grad_norm": 4.557161331176758,
"learning_rate": 4.78990368159424e-06,
"loss": 0.0453,
"step": 708
},
{
"epoch": 0.02766829268292683,
"grad_norm": 0.10228562355041504,
"learning_rate": 4.76014094029921e-06,
"loss": 0.012,
"step": 709
},
{
"epoch": 0.027707317073170732,
"grad_norm": 0.3522282540798187,
"learning_rate": 4.7304420450332244e-06,
"loss": 0.0347,
"step": 710
},
{
"epoch": 0.027746341463414634,
"grad_norm": 0.06579418480396271,
"learning_rate": 4.700807357667953e-06,
"loss": 0.0007,
"step": 711
},
{
"epoch": 0.027785365853658536,
"grad_norm": 0.4105568528175354,
"learning_rate": 4.671237239292699e-06,
"loss": 0.036,
"step": 712
},
{
"epoch": 0.027824390243902437,
"grad_norm": 0.4862738847732544,
"learning_rate": 4.641732050210032e-06,
"loss": 0.0647,
"step": 713
},
{
"epoch": 0.027863414634146343,
"grad_norm": 0.27456459403038025,
"learning_rate": 4.612292149931369e-06,
"loss": 0.0412,
"step": 714
},
{
"epoch": 0.027902439024390244,
"grad_norm": 0.7907021045684814,
"learning_rate": 4.582917897172603e-06,
"loss": 0.0571,
"step": 715
},
{
"epoch": 0.027941463414634146,
"grad_norm": 0.2606887221336365,
"learning_rate": 4.5536096498497295e-06,
"loss": 0.0274,
"step": 716
},
{
"epoch": 0.027980487804878048,
"grad_norm": 1.0683624744415283,
"learning_rate": 4.524367765074499e-06,
"loss": 0.0576,
"step": 717
},
{
"epoch": 0.02801951219512195,
"grad_norm": 0.1247168481349945,
"learning_rate": 4.495192599150045e-06,
"loss": 0.0172,
"step": 718
},
{
"epoch": 0.028058536585365855,
"grad_norm": 2.4952852725982666,
"learning_rate": 4.46608450756656e-06,
"loss": 0.0783,
"step": 719
},
{
"epoch": 0.028097560975609757,
"grad_norm": 1.9848724603652954,
"learning_rate": 4.437043844996952e-06,
"loss": 0.0582,
"step": 720
},
{
"epoch": 0.02813658536585366,
"grad_norm": 0.2863747477531433,
"learning_rate": 4.408070965292534e-06,
"loss": 0.0485,
"step": 721
},
{
"epoch": 0.02817560975609756,
"grad_norm": 0.043544329702854156,
"learning_rate": 4.379166221478697e-06,
"loss": 0.0009,
"step": 722
},
{
"epoch": 0.028214634146341462,
"grad_norm": 0.24715405702590942,
"learning_rate": 4.350329965750622e-06,
"loss": 0.0391,
"step": 723
},
{
"epoch": 0.028253658536585367,
"grad_norm": 0.6251457333564758,
"learning_rate": 4.321562549468991e-06,
"loss": 0.0577,
"step": 724
},
{
"epoch": 0.02829268292682927,
"grad_norm": 0.10604850202798843,
"learning_rate": 4.292864323155684e-06,
"loss": 0.013,
"step": 725
},
{
"epoch": 0.02833170731707317,
"grad_norm": 0.7458891868591309,
"learning_rate": 4.264235636489542e-06,
"loss": 0.0341,
"step": 726
},
{
"epoch": 0.028370731707317073,
"grad_norm": 0.6399158239364624,
"learning_rate": 4.235676838302069e-06,
"loss": 0.065,
"step": 727
},
{
"epoch": 0.028409756097560974,
"grad_norm": 0.35858163237571716,
"learning_rate": 4.207188276573214e-06,
"loss": 0.0043,
"step": 728
},
{
"epoch": 0.02844878048780488,
"grad_norm": 0.2830521762371063,
"learning_rate": 4.178770298427107e-06,
"loss": 0.0159,
"step": 729
},
{
"epoch": 0.02848780487804878,
"grad_norm": 0.1780395358800888,
"learning_rate": 4.150423250127846e-06,
"loss": 0.0331,
"step": 730
},
{
"epoch": 0.028526829268292683,
"grad_norm": 48.76409149169922,
"learning_rate": 4.12214747707527e-06,
"loss": 0.4586,
"step": 731
},
{
"epoch": 0.028565853658536585,
"grad_norm": 0.8201588988304138,
"learning_rate": 4.093943323800746e-06,
"loss": 0.076,
"step": 732
},
{
"epoch": 0.028604878048780487,
"grad_norm": 0.38445425033569336,
"learning_rate": 4.065811133962987e-06,
"loss": 0.0419,
"step": 733
},
{
"epoch": 0.02864390243902439,
"grad_norm": 1.271559476852417,
"learning_rate": 4.037751250343841e-06,
"loss": 0.0789,
"step": 734
},
{
"epoch": 0.028682926829268294,
"grad_norm": 0.4847497344017029,
"learning_rate": 4.009764014844143e-06,
"loss": 0.0487,
"step": 735
},
{
"epoch": 0.028721951219512196,
"grad_norm": 43.94010543823242,
"learning_rate": 3.981849768479516e-06,
"loss": 0.1357,
"step": 736
},
{
"epoch": 0.028760975609756097,
"grad_norm": 0.5430985689163208,
"learning_rate": 3.954008851376252e-06,
"loss": 0.0463,
"step": 737
},
{
"epoch": 0.0288,
"grad_norm": 0.06355824321508408,
"learning_rate": 3.9262416027671354e-06,
"loss": 0.0014,
"step": 738
},
{
"epoch": 0.0288390243902439,
"grad_norm": 1.003113031387329,
"learning_rate": 3.898548360987325e-06,
"loss": 0.0626,
"step": 739
},
{
"epoch": 0.028878048780487806,
"grad_norm": 1.6102895736694336,
"learning_rate": 3.8709294634702374e-06,
"loss": 0.022,
"step": 740
},
{
"epoch": 0.028917073170731708,
"grad_norm": 2.137524366378784,
"learning_rate": 3.8433852467434175e-06,
"loss": 0.0323,
"step": 741
},
{
"epoch": 0.02895609756097561,
"grad_norm": 0.2510572373867035,
"learning_rate": 3.81591604642446e-06,
"loss": 0.0013,
"step": 742
},
{
"epoch": 0.02899512195121951,
"grad_norm": 0.43281295895576477,
"learning_rate": 3.7885221972168974e-06,
"loss": 0.0453,
"step": 743
},
{
"epoch": 0.029034146341463413,
"grad_norm": 1.7247326374053955,
"learning_rate": 3.7612040329061405e-06,
"loss": 0.0478,
"step": 744
},
{
"epoch": 0.02907317073170732,
"grad_norm": 6.772576808929443,
"learning_rate": 3.7339618863553983e-06,
"loss": 0.0602,
"step": 745
},
{
"epoch": 0.02911219512195122,
"grad_norm": 0.5640243291854858,
"learning_rate": 3.7067960895016277e-06,
"loss": 0.0576,
"step": 746
},
{
"epoch": 0.029151219512195122,
"grad_norm": 0.28753384947776794,
"learning_rate": 3.679706973351491e-06,
"loss": 0.0407,
"step": 747
},
{
"epoch": 0.029190243902439024,
"grad_norm": 0.15965646505355835,
"learning_rate": 3.6526948679773256e-06,
"loss": 0.0312,
"step": 748
},
{
"epoch": 0.029229268292682926,
"grad_norm": 0.09254062920808792,
"learning_rate": 3.625760102513103e-06,
"loss": 0.0016,
"step": 749
},
{
"epoch": 0.02926829268292683,
"grad_norm": 0.2519022524356842,
"learning_rate": 3.598903005150444e-06,
"loss": 0.0332,
"step": 750
},
{
"epoch": 0.029307317073170733,
"grad_norm": 0.48083826899528503,
"learning_rate": 3.5721239031346067e-06,
"loss": 0.032,
"step": 751
},
{
"epoch": 0.029346341463414635,
"grad_norm": 2.171731948852539,
"learning_rate": 3.545423122760493e-06,
"loss": 0.0513,
"step": 752
},
{
"epoch": 0.029385365853658536,
"grad_norm": 0.5155346989631653,
"learning_rate": 3.5188009893686916e-06,
"loss": 0.0404,
"step": 753
},
{
"epoch": 0.029424390243902438,
"grad_norm": 1.0810842514038086,
"learning_rate": 3.492257827341492e-06,
"loss": 0.0487,
"step": 754
},
{
"epoch": 0.02946341463414634,
"grad_norm": 32.556636810302734,
"learning_rate": 3.4657939600989453e-06,
"loss": 0.412,
"step": 755
},
{
"epoch": 0.029502439024390245,
"grad_norm": 0.6079092025756836,
"learning_rate": 3.4394097100949286e-06,
"loss": 0.0583,
"step": 756
},
{
"epoch": 0.029541463414634147,
"grad_norm": 0.7667189836502075,
"learning_rate": 3.4131053988131947e-06,
"loss": 0.0404,
"step": 757
},
{
"epoch": 0.02958048780487805,
"grad_norm": 1.863149881362915,
"learning_rate": 3.3868813467634833e-06,
"loss": 0.0663,
"step": 758
},
{
"epoch": 0.02961951219512195,
"grad_norm": 0.27520114183425903,
"learning_rate": 3.360737873477584e-06,
"loss": 0.0425,
"step": 759
},
{
"epoch": 0.029658536585365852,
"grad_norm": 0.3419632315635681,
"learning_rate": 3.3346752975054763e-06,
"loss": 0.0392,
"step": 760
},
{
"epoch": 0.029697560975609757,
"grad_norm": 0.4352755844593048,
"learning_rate": 3.308693936411421e-06,
"loss": 0.0412,
"step": 761
},
{
"epoch": 0.02973658536585366,
"grad_norm": 19.8631591796875,
"learning_rate": 3.2827941067700996e-06,
"loss": 0.462,
"step": 762
},
{
"epoch": 0.02977560975609756,
"grad_norm": 15.159614562988281,
"learning_rate": 3.2569761241627694e-06,
"loss": 0.0978,
"step": 763
},
{
"epoch": 0.029814634146341463,
"grad_norm": 1.176772117614746,
"learning_rate": 3.2312403031733943e-06,
"loss": 0.0566,
"step": 764
},
{
"epoch": 0.029853658536585365,
"grad_norm": 14.415740013122559,
"learning_rate": 3.2055869573848374e-06,
"loss": 0.9163,
"step": 765
},
{
"epoch": 0.02989268292682927,
"grad_norm": 0.8066580891609192,
"learning_rate": 3.1800163993750166e-06,
"loss": 0.0217,
"step": 766
},
{
"epoch": 0.02993170731707317,
"grad_norm": 6.320103168487549,
"learning_rate": 3.1545289407131128e-06,
"loss": 0.0327,
"step": 767
},
{
"epoch": 0.029970731707317073,
"grad_norm": 0.06332789361476898,
"learning_rate": 3.1291248919557717e-06,
"loss": 0.0011,
"step": 768
},
{
"epoch": 0.030009756097560975,
"grad_norm": 0.6345298886299133,
"learning_rate": 3.103804562643302e-06,
"loss": 0.0551,
"step": 769
},
{
"epoch": 0.030048780487804877,
"grad_norm": 0.7310605645179749,
"learning_rate": 3.0785682612959334e-06,
"loss": 0.0413,
"step": 770
},
{
"epoch": 0.03008780487804878,
"grad_norm": 22.530820846557617,
"learning_rate": 3.0534162954100264e-06,
"loss": 0.1556,
"step": 771
},
{
"epoch": 0.030126829268292684,
"grad_norm": 1.038345456123352,
"learning_rate": 3.028348971454356e-06,
"loss": 0.0549,
"step": 772
},
{
"epoch": 0.030165853658536586,
"grad_norm": 0.22657279670238495,
"learning_rate": 3.003366594866345e-06,
"loss": 0.0427,
"step": 773
},
{
"epoch": 0.030204878048780488,
"grad_norm": 1.2644487619400024,
"learning_rate": 2.978469470048376e-06,
"loss": 0.0509,
"step": 774
},
{
"epoch": 0.03024390243902439,
"grad_norm": 2.099160671234131,
"learning_rate": 2.953657900364053e-06,
"loss": 0.0295,
"step": 775
},
{
"epoch": 0.03028292682926829,
"grad_norm": 0.0801204964518547,
"learning_rate": 2.9289321881345257e-06,
"loss": 0.0162,
"step": 776
},
{
"epoch": 0.030321951219512196,
"grad_norm": 0.16855764389038086,
"learning_rate": 2.9042926346347932e-06,
"loss": 0.0262,
"step": 777
},
{
"epoch": 0.030360975609756098,
"grad_norm": 0.06617280840873718,
"learning_rate": 2.8797395400900362e-06,
"loss": 0.016,
"step": 778
},
{
"epoch": 0.0304,
"grad_norm": 0.23229794204235077,
"learning_rate": 2.855273203671969e-06,
"loss": 0.0066,
"step": 779
},
{
"epoch": 0.0304390243902439,
"grad_norm": 0.4215773344039917,
"learning_rate": 2.830893923495173e-06,
"loss": 0.0566,
"step": 780
},
{
"epoch": 0.030478048780487803,
"grad_norm": 0.498444527387619,
"learning_rate": 2.8066019966134907e-06,
"loss": 0.0427,
"step": 781
},
{
"epoch": 0.03051707317073171,
"grad_norm": 0.7319293022155762,
"learning_rate": 2.7823977190163788e-06,
"loss": 0.0458,
"step": 782
},
{
"epoch": 0.03055609756097561,
"grad_norm": 0.6238008141517639,
"learning_rate": 2.7582813856253276e-06,
"loss": 0.0194,
"step": 783
},
{
"epoch": 0.030595121951219512,
"grad_norm": 0.23200275003910065,
"learning_rate": 2.7342532902902418e-06,
"loss": 0.0448,
"step": 784
},
{
"epoch": 0.030634146341463414,
"grad_norm": 0.7218787670135498,
"learning_rate": 2.7103137257858867e-06,
"loss": 0.0074,
"step": 785
},
{
"epoch": 0.030673170731707316,
"grad_norm": 8.98879623413086,
"learning_rate": 2.6864629838082957e-06,
"loss": 0.042,
"step": 786
},
{
"epoch": 0.03071219512195122,
"grad_norm": 0.029952067881822586,
"learning_rate": 2.6627013549712355e-06,
"loss": 0.0009,
"step": 787
},
{
"epoch": 0.030751219512195123,
"grad_norm": 0.4528684616088867,
"learning_rate": 2.639029128802657e-06,
"loss": 0.0058,
"step": 788
},
{
"epoch": 0.030790243902439025,
"grad_norm": 0.2058892399072647,
"learning_rate": 2.615446593741161e-06,
"loss": 0.0344,
"step": 789
},
{
"epoch": 0.030829268292682926,
"grad_norm": 0.05679577961564064,
"learning_rate": 2.5919540371325005e-06,
"loss": 0.0009,
"step": 790
},
{
"epoch": 0.030868292682926828,
"grad_norm": 0.49694156646728516,
"learning_rate": 2.5685517452260566e-06,
"loss": 0.0461,
"step": 791
},
{
"epoch": 0.03090731707317073,
"grad_norm": 0.16773158311843872,
"learning_rate": 2.5452400031713786e-06,
"loss": 0.0431,
"step": 792
},
{
"epoch": 0.030946341463414635,
"grad_norm": 8.313273429870605,
"learning_rate": 2.522019095014683e-06,
"loss": 0.0941,
"step": 793
},
{
"epoch": 0.030985365853658537,
"grad_norm": 0.1104949563741684,
"learning_rate": 2.4988893036954045e-06,
"loss": 0.0176,
"step": 794
},
{
"epoch": 0.03102439024390244,
"grad_norm": 0.2046859860420227,
"learning_rate": 2.4758509110427576e-06,
"loss": 0.0146,
"step": 795
},
{
"epoch": 0.03106341463414634,
"grad_norm": 0.37380310893058777,
"learning_rate": 2.45290419777228e-06,
"loss": 0.0411,
"step": 796
},
{
"epoch": 0.031102439024390242,
"grad_norm": 0.09917093813419342,
"learning_rate": 2.4300494434824373e-06,
"loss": 0.0238,
"step": 797
},
{
"epoch": 0.031141463414634148,
"grad_norm": 0.7636759877204895,
"learning_rate": 2.407286926651192e-06,
"loss": 0.0391,
"step": 798
},
{
"epoch": 0.03118048780487805,
"grad_norm": 0.6203073859214783,
"learning_rate": 2.3846169246326345e-06,
"loss": 0.0443,
"step": 799
},
{
"epoch": 0.03121951219512195,
"grad_norm": 0.292289674282074,
"learning_rate": 2.362039713653581e-06,
"loss": 0.042,
"step": 800
},
{
"epoch": 0.031258536585365856,
"grad_norm": 0.15751637518405914,
"learning_rate": 2.339555568810221e-06,
"loss": 0.0296,
"step": 801
},
{
"epoch": 0.031297560975609755,
"grad_norm": 0.4948784112930298,
"learning_rate": 2.317164764064769e-06,
"loss": 0.044,
"step": 802
},
{
"epoch": 0.03133658536585366,
"grad_norm": 0.5928308963775635,
"learning_rate": 2.2948675722421086e-06,
"loss": 0.049,
"step": 803
},
{
"epoch": 0.03137560975609756,
"grad_norm": 0.563982367515564,
"learning_rate": 2.27266426502649e-06,
"loss": 0.0567,
"step": 804
},
{
"epoch": 0.031414634146341464,
"grad_norm": 0.12261471152305603,
"learning_rate": 2.2505551129582047e-06,
"loss": 0.0146,
"step": 805
},
{
"epoch": 0.03145365853658537,
"grad_norm": 0.4826788008213043,
"learning_rate": 2.2285403854302912e-06,
"loss": 0.0335,
"step": 806
},
{
"epoch": 0.03149268292682927,
"grad_norm": 0.0629960224032402,
"learning_rate": 2.206620350685257e-06,
"loss": 0.0008,
"step": 807
},
{
"epoch": 0.03153170731707317,
"grad_norm": 0.3132227659225464,
"learning_rate": 2.1847952758118118e-06,
"loss": 0.0212,
"step": 808
},
{
"epoch": 0.03157073170731707,
"grad_norm": 0.7135205268859863,
"learning_rate": 2.163065426741603e-06,
"loss": 0.0114,
"step": 809
},
{
"epoch": 0.031609756097560976,
"grad_norm": 0.03485079109668732,
"learning_rate": 2.1414310682459805e-06,
"loss": 0.0006,
"step": 810
},
{
"epoch": 0.03164878048780488,
"grad_norm": 1.211981177330017,
"learning_rate": 2.119892463932781e-06,
"loss": 0.0314,
"step": 811
},
{
"epoch": 0.03168780487804878,
"grad_norm": 0.528056263923645,
"learning_rate": 2.098449876243096e-06,
"loss": 0.0303,
"step": 812
},
{
"epoch": 0.031726829268292685,
"grad_norm": 0.34719961881637573,
"learning_rate": 2.0771035664480944e-06,
"loss": 0.0374,
"step": 813
},
{
"epoch": 0.03176585365853658,
"grad_norm": 0.8061074614524841,
"learning_rate": 2.0558537946458177e-06,
"loss": 0.0558,
"step": 814
},
{
"epoch": 0.03180487804878049,
"grad_norm": 0.7810201048851013,
"learning_rate": 2.0347008197580376e-06,
"loss": 0.0513,
"step": 815
},
{
"epoch": 0.031843902439024394,
"grad_norm": 0.6814368367195129,
"learning_rate": 2.013644899527074e-06,
"loss": 0.0443,
"step": 816
},
{
"epoch": 0.03188292682926829,
"grad_norm": 0.4110775291919708,
"learning_rate": 1.9926862905126663e-06,
"loss": 0.0403,
"step": 817
},
{
"epoch": 0.0319219512195122,
"grad_norm": 17.279312133789062,
"learning_rate": 1.9718252480888567e-06,
"loss": 0.0964,
"step": 818
},
{
"epoch": 0.031960975609756095,
"grad_norm": 0.8680134415626526,
"learning_rate": 1.95106202644086e-06,
"loss": 0.0574,
"step": 819
},
{
"epoch": 0.032,
"grad_norm": 0.8446072936058044,
"learning_rate": 1.930396878561983e-06,
"loss": 0.0689,
"step": 820
},
{
"epoch": 0.0320390243902439,
"grad_norm": 0.665858268737793,
"learning_rate": 1.9098300562505266e-06,
"loss": 0.0592,
"step": 821
},
{
"epoch": 0.032078048780487804,
"grad_norm": 0.3089471757411957,
"learning_rate": 1.8893618101067357e-06,
"loss": 0.0437,
"step": 822
},
{
"epoch": 0.03211707317073171,
"grad_norm": 0.08234549313783646,
"learning_rate": 1.8689923895297247e-06,
"loss": 0.0125,
"step": 823
},
{
"epoch": 0.03215609756097561,
"grad_norm": 0.7744914293289185,
"learning_rate": 1.848722042714457e-06,
"loss": 0.0573,
"step": 824
},
{
"epoch": 0.03219512195121951,
"grad_norm": 0.0758761540055275,
"learning_rate": 1.8285510166487154e-06,
"loss": 0.0119,
"step": 825
},
{
"epoch": 0.03223414634146341,
"grad_norm": 0.2736808657646179,
"learning_rate": 1.808479557110081e-06,
"loss": 0.0352,
"step": 826
},
{
"epoch": 0.03227317073170732,
"grad_norm": 0.06399993598461151,
"learning_rate": 1.7885079086629598e-06,
"loss": 0.0004,
"step": 827
},
{
"epoch": 0.03231219512195122,
"grad_norm": 0.17318545281887054,
"learning_rate": 1.7686363146555807e-06,
"loss": 0.0237,
"step": 828
},
{
"epoch": 0.03235121951219512,
"grad_norm": 0.10776067525148392,
"learning_rate": 1.7488650172170496e-06,
"loss": 0.0122,
"step": 829
},
{
"epoch": 0.032390243902439025,
"grad_norm": 0.11919575929641724,
"learning_rate": 1.7291942572543806e-06,
"loss": 0.0146,
"step": 830
},
{
"epoch": 0.032429268292682924,
"grad_norm": 6.274477005004883,
"learning_rate": 1.709624274449584e-06,
"loss": 0.0915,
"step": 831
},
{
"epoch": 0.03246829268292683,
"grad_norm": 0.12984342873096466,
"learning_rate": 1.6901553072567189e-06,
"loss": 0.0171,
"step": 832
},
{
"epoch": 0.032507317073170734,
"grad_norm": 0.34555119276046753,
"learning_rate": 1.6707875928990059e-06,
"loss": 0.0177,
"step": 833
},
{
"epoch": 0.03254634146341463,
"grad_norm": 0.3377743661403656,
"learning_rate": 1.651521367365936e-06,
"loss": 0.0362,
"step": 834
},
{
"epoch": 0.03258536585365854,
"grad_norm": 0.3195168077945709,
"learning_rate": 1.6323568654103838e-06,
"loss": 0.0379,
"step": 835
},
{
"epoch": 0.032624390243902436,
"grad_norm": 0.5070753693580627,
"learning_rate": 1.6132943205457607e-06,
"loss": 0.0351,
"step": 836
},
{
"epoch": 0.03266341463414634,
"grad_norm": 0.07364442944526672,
"learning_rate": 1.5943339650431578e-06,
"loss": 0.0172,
"step": 837
},
{
"epoch": 0.03270243902439025,
"grad_norm": 29.415647506713867,
"learning_rate": 1.5754760299285255e-06,
"loss": 0.4915,
"step": 838
},
{
"epoch": 0.032741463414634145,
"grad_norm": 0.2161831557750702,
"learning_rate": 1.5567207449798517e-06,
"loss": 0.0264,
"step": 839
},
{
"epoch": 0.03278048780487805,
"grad_norm": 2.044379949569702,
"learning_rate": 1.538068338724361e-06,
"loss": 0.0312,
"step": 840
},
{
"epoch": 0.03281951219512195,
"grad_norm": 0.29258430004119873,
"learning_rate": 1.5195190384357405e-06,
"loss": 0.0282,
"step": 841
},
{
"epoch": 0.032858536585365854,
"grad_norm": 1.3051424026489258,
"learning_rate": 1.5010730701313626e-06,
"loss": 0.0683,
"step": 842
},
{
"epoch": 0.03289756097560976,
"grad_norm": 0.8376571536064148,
"learning_rate": 1.4827306585695234e-06,
"loss": 0.0427,
"step": 843
},
{
"epoch": 0.03293658536585366,
"grad_norm": 0.4196270704269409,
"learning_rate": 1.4644920272467245e-06,
"loss": 0.0445,
"step": 844
},
{
"epoch": 0.03297560975609756,
"grad_norm": 0.37262892723083496,
"learning_rate": 1.446357398394934e-06,
"loss": 0.0538,
"step": 845
},
{
"epoch": 0.03301463414634146,
"grad_norm": 0.5560579299926758,
"learning_rate": 1.4283269929788779e-06,
"loss": 0.0044,
"step": 846
},
{
"epoch": 0.033053658536585366,
"grad_norm": 0.5738880634307861,
"learning_rate": 1.4104010306933558e-06,
"loss": 0.032,
"step": 847
},
{
"epoch": 0.03309268292682927,
"grad_norm": 1.707336664199829,
"learning_rate": 1.3925797299605649e-06,
"loss": 0.1044,
"step": 848
},
{
"epoch": 0.03313170731707317,
"grad_norm": 0.5831617116928101,
"learning_rate": 1.3748633079274254e-06,
"loss": 0.041,
"step": 849
},
{
"epoch": 0.033170731707317075,
"grad_norm": 0.5975121855735779,
"learning_rate": 1.3572519804629537e-06,
"loss": 0.0579,
"step": 850
},
{
"epoch": 0.03320975609756097,
"grad_norm": 0.5416494011878967,
"learning_rate": 1.339745962155613e-06,
"loss": 0.0465,
"step": 851
},
{
"epoch": 0.03324878048780488,
"grad_norm": 0.17728735506534576,
"learning_rate": 1.322345466310717e-06,
"loss": 0.0265,
"step": 852
},
{
"epoch": 0.033287804878048784,
"grad_norm": 1.6175427436828613,
"learning_rate": 1.30505070494781e-06,
"loss": 0.061,
"step": 853
},
{
"epoch": 0.03332682926829268,
"grad_norm": 0.4316226840019226,
"learning_rate": 1.2878618887981064e-06,
"loss": 0.0388,
"step": 854
},
{
"epoch": 0.03336585365853659,
"grad_norm": 4.9184699058532715,
"learning_rate": 1.2707792273019049e-06,
"loss": 0.0584,
"step": 855
},
{
"epoch": 0.033404878048780486,
"grad_norm": 0.1798275262117386,
"learning_rate": 1.2538029286060428e-06,
"loss": 0.0318,
"step": 856
},
{
"epoch": 0.03344390243902439,
"grad_norm": 0.606025218963623,
"learning_rate": 1.2369331995613664e-06,
"loss": 0.0511,
"step": 857
},
{
"epoch": 0.033482926829268296,
"grad_norm": 0.616762101650238,
"learning_rate": 1.2201702457201948e-06,
"loss": 0.0473,
"step": 858
},
{
"epoch": 0.033521951219512194,
"grad_norm": 0.9533067941665649,
"learning_rate": 1.2035142713338366e-06,
"loss": 0.0792,
"step": 859
},
{
"epoch": 0.0335609756097561,
"grad_norm": 0.5307132601737976,
"learning_rate": 1.1869654793500784e-06,
"loss": 0.0328,
"step": 860
},
{
"epoch": 0.0336,
"grad_norm": 0.6104608178138733,
"learning_rate": 1.1705240714107301e-06,
"loss": 0.0195,
"step": 861
},
{
"epoch": 0.0336390243902439,
"grad_norm": 0.11029266566038132,
"learning_rate": 1.1541902478491607e-06,
"loss": 0.0126,
"step": 862
},
{
"epoch": 0.0336780487804878,
"grad_norm": 0.34786757826805115,
"learning_rate": 1.1379642076878528e-06,
"loss": 0.0379,
"step": 863
},
{
"epoch": 0.03371707317073171,
"grad_norm": 0.649806022644043,
"learning_rate": 1.1218461486359878e-06,
"loss": 0.0455,
"step": 864
},
{
"epoch": 0.03375609756097561,
"grad_norm": 0.1312018781900406,
"learning_rate": 1.1058362670870248e-06,
"loss": 0.0139,
"step": 865
},
{
"epoch": 0.03379512195121951,
"grad_norm": 0.06650601327419281,
"learning_rate": 1.0899347581163222e-06,
"loss": 0.0146,
"step": 866
},
{
"epoch": 0.033834146341463416,
"grad_norm": 0.09275788813829422,
"learning_rate": 1.0741418154787443e-06,
"loss": 0.0109,
"step": 867
},
{
"epoch": 0.033873170731707314,
"grad_norm": 0.6223591566085815,
"learning_rate": 1.058457631606319e-06,
"loss": 0.0576,
"step": 868
},
{
"epoch": 0.03391219512195122,
"grad_norm": 16.083154678344727,
"learning_rate": 1.042882397605871e-06,
"loss": 0.4109,
"step": 869
},
{
"epoch": 0.033951219512195124,
"grad_norm": 0.22150637209415436,
"learning_rate": 1.0274163032567165e-06,
"loss": 0.0271,
"step": 870
},
{
"epoch": 0.03399024390243902,
"grad_norm": 20.84095573425293,
"learning_rate": 1.012059537008332e-06,
"loss": 0.1342,
"step": 871
},
{
"epoch": 0.03402926829268293,
"grad_norm": 0.11181342601776123,
"learning_rate": 9.968122859780648e-07,
"loss": 0.0315,
"step": 872
},
{
"epoch": 0.034068292682926826,
"grad_norm": 0.1724962592124939,
"learning_rate": 9.816747359488632e-07,
"loss": 0.0287,
"step": 873
},
{
"epoch": 0.03410731707317073,
"grad_norm": 0.09344766288995743,
"learning_rate": 9.666470713669918e-07,
"loss": 0.0169,
"step": 874
},
{
"epoch": 0.03414634146341464,
"grad_norm": 0.7909224033355713,
"learning_rate": 9.517294753398066e-07,
"loss": 0.0068,
"step": 875
},
{
"epoch": 0.034185365853658535,
"grad_norm": 1.2862979173660278,
"learning_rate": 9.369221296335007e-07,
"loss": 0.0825,
"step": 876
},
{
"epoch": 0.03422439024390244,
"grad_norm": 1.6369551420211792,
"learning_rate": 9.222252146709143e-07,
"loss": 0.058,
"step": 877
},
{
"epoch": 0.03426341463414634,
"grad_norm": 0.33004558086395264,
"learning_rate": 9.076389095293148e-07,
"loss": 0.0518,
"step": 878
},
{
"epoch": 0.034302439024390244,
"grad_norm": 0.3090706467628479,
"learning_rate": 8.931633919382299e-07,
"loss": 0.0499,
"step": 879
},
{
"epoch": 0.03434146341463415,
"grad_norm": 0.4096774756908417,
"learning_rate": 8.787988382772705e-07,
"loss": 0.0375,
"step": 880
},
{
"epoch": 0.03438048780487805,
"grad_norm": 20.11399269104004,
"learning_rate": 8.645454235739903e-07,
"loss": 0.142,
"step": 881
},
{
"epoch": 0.03441951219512195,
"grad_norm": 0.16637049615383148,
"learning_rate": 8.504033215017527e-07,
"loss": 0.0145,
"step": 882
},
{
"epoch": 0.03445853658536585,
"grad_norm": 1.362591028213501,
"learning_rate": 8.363727043776037e-07,
"loss": 0.0775,
"step": 883
},
{
"epoch": 0.034497560975609756,
"grad_norm": 1.0016076564788818,
"learning_rate": 8.224537431601886e-07,
"loss": 0.045,
"step": 884
},
{
"epoch": 0.03453658536585366,
"grad_norm": 22.13150978088379,
"learning_rate": 8.086466074476562e-07,
"loss": 0.2224,
"step": 885
},
{
"epoch": 0.03457560975609756,
"grad_norm": 0.5086643695831299,
"learning_rate": 7.949514654755963e-07,
"loss": 0.0404,
"step": 886
},
{
"epoch": 0.034614634146341465,
"grad_norm": 0.1973675787448883,
"learning_rate": 7.81368484114996e-07,
"loss": 0.0268,
"step": 887
},
{
"epoch": 0.03465365853658536,
"grad_norm": 0.30816933512687683,
"learning_rate": 7.678978288701911e-07,
"loss": 0.0312,
"step": 888
},
{
"epoch": 0.03469268292682927,
"grad_norm": 24.030494689941406,
"learning_rate": 7.545396638768698e-07,
"loss": 0.0924,
"step": 889
},
{
"epoch": 0.034731707317073174,
"grad_norm": 0.7220743894577026,
"learning_rate": 7.412941519000527e-07,
"loss": 0.0211,
"step": 890
},
{
"epoch": 0.03477073170731707,
"grad_norm": 0.36591020226478577,
"learning_rate": 7.281614543321269e-07,
"loss": 0.0477,
"step": 891
},
{
"epoch": 0.03480975609756098,
"grad_norm": 0.42255067825317383,
"learning_rate": 7.151417311908648e-07,
"loss": 0.0492,
"step": 892
},
{
"epoch": 0.034848780487804876,
"grad_norm": 0.09390468895435333,
"learning_rate": 7.022351411174866e-07,
"loss": 0.0217,
"step": 893
},
{
"epoch": 0.03488780487804878,
"grad_norm": 0.0377718061208725,
"learning_rate": 6.894418413747183e-07,
"loss": 0.0009,
"step": 894
},
{
"epoch": 0.034926829268292686,
"grad_norm": 0.4957011640071869,
"learning_rate": 6.767619878448783e-07,
"loss": 0.0551,
"step": 895
},
{
"epoch": 0.034965853658536585,
"grad_norm": 0.45442405343055725,
"learning_rate": 6.641957350279838e-07,
"loss": 0.0423,
"step": 896
},
{
"epoch": 0.03500487804878049,
"grad_norm": 0.6611925363540649,
"learning_rate": 6.517432360398556e-07,
"loss": 0.0444,
"step": 897
},
{
"epoch": 0.03504390243902439,
"grad_norm": 0.17193956673145294,
"learning_rate": 6.394046426102673e-07,
"loss": 0.0026,
"step": 898
},
{
"epoch": 0.03508292682926829,
"grad_norm": 0.10275353491306305,
"learning_rate": 6.271801050810856e-07,
"loss": 0.0127,
"step": 899
},
{
"epoch": 0.0351219512195122,
"grad_norm": 0.38216477632522583,
"learning_rate": 6.150697724044407e-07,
"loss": 0.0342,
"step": 900
},
{
"epoch": 0.0351609756097561,
"grad_norm": 0.9646055102348328,
"learning_rate": 6.030737921409169e-07,
"loss": 0.0672,
"step": 901
},
{
"epoch": 0.0352,
"grad_norm": 0.07141842693090439,
"learning_rate": 5.911923104577455e-07,
"loss": 0.0176,
"step": 902
},
{
"epoch": 0.0352390243902439,
"grad_norm": 0.3345755636692047,
"learning_rate": 5.794254721270331e-07,
"loss": 0.0329,
"step": 903
},
{
"epoch": 0.035278048780487806,
"grad_norm": 0.48479580879211426,
"learning_rate": 5.677734205239904e-07,
"loss": 0.0535,
"step": 904
},
{
"epoch": 0.035317073170731704,
"grad_norm": 1.3086276054382324,
"learning_rate": 5.562362976251901e-07,
"loss": 0.03,
"step": 905
},
{
"epoch": 0.03535609756097561,
"grad_norm": 0.6228365302085876,
"learning_rate": 5.448142440068316e-07,
"loss": 0.0053,
"step": 906
},
{
"epoch": 0.035395121951219514,
"grad_norm": 7.785345554351807,
"learning_rate": 5.335073988430373e-07,
"loss": 0.0669,
"step": 907
},
{
"epoch": 0.03543414634146341,
"grad_norm": 29.245267868041992,
"learning_rate": 5.223158999041444e-07,
"loss": 0.1556,
"step": 908
},
{
"epoch": 0.03547317073170732,
"grad_norm": 0.6782280206680298,
"learning_rate": 5.112398835550348e-07,
"loss": 0.0184,
"step": 909
},
{
"epoch": 0.035512195121951216,
"grad_norm": 0.19054441154003143,
"learning_rate": 5.002794847534765e-07,
"loss": 0.0159,
"step": 910
},
{
"epoch": 0.03555121951219512,
"grad_norm": 10.456110954284668,
"learning_rate": 4.894348370484648e-07,
"loss": 0.0797,
"step": 911
},
{
"epoch": 0.03559024390243903,
"grad_norm": 0.2572881579399109,
"learning_rate": 4.787060725786141e-07,
"loss": 0.0451,
"step": 912
},
{
"epoch": 0.035629268292682925,
"grad_norm": 0.19994540512561798,
"learning_rate": 4.6809332207053083e-07,
"loss": 0.0371,
"step": 913
},
{
"epoch": 0.03566829268292683,
"grad_norm": 0.43133410811424255,
"learning_rate": 4.575967148372318e-07,
"loss": 0.0336,
"step": 914
},
{
"epoch": 0.03570731707317073,
"grad_norm": 0.08243723958730698,
"learning_rate": 4.4721637877656377e-07,
"loss": 0.0011,
"step": 915
},
{
"epoch": 0.035746341463414634,
"grad_norm": 0.2883797883987427,
"learning_rate": 4.3695244036964567e-07,
"loss": 0.0242,
"step": 916
},
{
"epoch": 0.03578536585365854,
"grad_norm": 0.8832783699035645,
"learning_rate": 4.268050246793276e-07,
"loss": 0.052,
"step": 917
},
{
"epoch": 0.03582439024390244,
"grad_norm": 0.6831177473068237,
"learning_rate": 4.167742553486676e-07,
"loss": 0.044,
"step": 918
},
{
"epoch": 0.03586341463414634,
"grad_norm": 10.518945693969727,
"learning_rate": 4.068602545994249e-07,
"loss": 0.75,
"step": 919
},
{
"epoch": 0.03590243902439024,
"grad_norm": 0.2913833558559418,
"learning_rate": 3.9706314323056936e-07,
"loss": 0.0039,
"step": 920
},
{
"epoch": 0.035941463414634146,
"grad_norm": 0.8756797909736633,
"learning_rate": 3.8738304061681107e-07,
"loss": 0.0977,
"step": 921
},
{
"epoch": 0.03598048780487805,
"grad_norm": 0.5055662989616394,
"learning_rate": 3.7782006470714614e-07,
"loss": 0.0325,
"step": 922
},
{
"epoch": 0.03601951219512195,
"grad_norm": 0.1131315529346466,
"learning_rate": 3.68374332023419e-07,
"loss": 0.0119,
"step": 923
},
{
"epoch": 0.036058536585365855,
"grad_norm": 0.29950669407844543,
"learning_rate": 3.590459576589e-07,
"loss": 0.0382,
"step": 924
},
{
"epoch": 0.03609756097560975,
"grad_norm": 0.48019495606422424,
"learning_rate": 3.498350552768859e-07,
"loss": 0.0343,
"step": 925
},
{
"epoch": 0.03613658536585366,
"grad_norm": 0.6012889742851257,
"learning_rate": 3.4074173710931804e-07,
"loss": 0.0342,
"step": 926
},
{
"epoch": 0.036175609756097564,
"grad_norm": 0.5712623000144958,
"learning_rate": 3.3176611395540625e-07,
"loss": 0.0357,
"step": 927
},
{
"epoch": 0.03621463414634146,
"grad_norm": 2.0280637741088867,
"learning_rate": 3.2290829518028867e-07,
"loss": 0.0462,
"step": 928
},
{
"epoch": 0.03625365853658537,
"grad_norm": 0.1502760946750641,
"learning_rate": 3.1416838871368925e-07,
"loss": 0.0012,
"step": 929
},
{
"epoch": 0.036292682926829266,
"grad_norm": 0.3682931363582611,
"learning_rate": 3.0554650104861137e-07,
"loss": 0.0415,
"step": 930
},
{
"epoch": 0.03633170731707317,
"grad_norm": 4.736273765563965,
"learning_rate": 2.970427372400353e-07,
"loss": 0.0597,
"step": 931
},
{
"epoch": 0.036370731707317076,
"grad_norm": 0.020756475627422333,
"learning_rate": 2.8865720090364037e-07,
"loss": 0.0005,
"step": 932
},
{
"epoch": 0.036409756097560975,
"grad_norm": 0.31484755873680115,
"learning_rate": 2.8038999421453827e-07,
"loss": 0.04,
"step": 933
},
{
"epoch": 0.03644878048780488,
"grad_norm": 1.5251492261886597,
"learning_rate": 2.7224121790603517e-07,
"loss": 0.052,
"step": 934
},
{
"epoch": 0.03648780487804878,
"grad_norm": 1.0087127685546875,
"learning_rate": 2.6421097126839714e-07,
"loss": 0.0089,
"step": 935
},
{
"epoch": 0.03652682926829268,
"grad_norm": 0.13508859276771545,
"learning_rate": 2.5629935214764866e-07,
"loss": 0.03,
"step": 936
},
{
"epoch": 0.03656585365853659,
"grad_norm": 0.07740826904773712,
"learning_rate": 2.4850645694436736e-07,
"loss": 0.0132,
"step": 937
},
{
"epoch": 0.03660487804878049,
"grad_norm": 0.4256102740764618,
"learning_rate": 2.4083238061252565e-07,
"loss": 0.0039,
"step": 938
},
{
"epoch": 0.03664390243902439,
"grad_norm": 0.86114901304245,
"learning_rate": 2.332772166583208e-07,
"loss": 0.0461,
"step": 939
},
{
"epoch": 0.03668292682926829,
"grad_norm": 3.0873754024505615,
"learning_rate": 2.2584105713904126e-07,
"loss": 0.0505,
"step": 940
},
{
"epoch": 0.036721951219512196,
"grad_norm": 0.3900286555290222,
"learning_rate": 2.1852399266194312e-07,
"loss": 0.0397,
"step": 941
},
{
"epoch": 0.036760975609756094,
"grad_norm": 23.157546997070312,
"learning_rate": 2.1132611238315004e-07,
"loss": 0.2699,
"step": 942
},
{
"epoch": 0.0368,
"grad_norm": 1.5414388179779053,
"learning_rate": 2.0424750400655947e-07,
"loss": 0.0478,
"step": 943
},
{
"epoch": 0.036839024390243905,
"grad_norm": 0.04199296236038208,
"learning_rate": 1.9728825378278248e-07,
"loss": 0.0008,
"step": 944
},
{
"epoch": 0.0368780487804878,
"grad_norm": 0.20770829916000366,
"learning_rate": 1.9044844650808468e-07,
"loss": 0.0241,
"step": 945
},
{
"epoch": 0.03691707317073171,
"grad_norm": 0.5355384349822998,
"learning_rate": 1.8372816552336025e-07,
"loss": 0.016,
"step": 946
},
{
"epoch": 0.036956097560975606,
"grad_norm": 0.4140547215938568,
"learning_rate": 1.7712749271311392e-07,
"loss": 0.0527,
"step": 947
},
{
"epoch": 0.03699512195121951,
"grad_norm": 0.5766732692718506,
"learning_rate": 1.706465085044584e-07,
"loss": 0.0207,
"step": 948
},
{
"epoch": 0.03703414634146342,
"grad_norm": 0.2273063063621521,
"learning_rate": 1.6428529186614195e-07,
"loss": 0.0438,
"step": 949
},
{
"epoch": 0.037073170731707315,
"grad_norm": 0.17417757213115692,
"learning_rate": 1.580439203075812e-07,
"loss": 0.0113,
"step": 950
},
{
"epoch": 0.03711219512195122,
"grad_norm": 0.3251909911632538,
"learning_rate": 1.519224698779198e-07,
"loss": 0.0327,
"step": 951
},
{
"epoch": 0.03715121951219512,
"grad_norm": 1.1465661525726318,
"learning_rate": 1.4592101516509916e-07,
"loss": 0.0341,
"step": 952
},
{
"epoch": 0.037190243902439024,
"grad_norm": 9.988484382629395,
"learning_rate": 1.400396292949513e-07,
"loss": 0.7207,
"step": 953
},
{
"epoch": 0.03722926829268293,
"grad_norm": 0.3501976430416107,
"learning_rate": 1.3427838393030634e-07,
"loss": 0.0384,
"step": 954
},
{
"epoch": 0.03726829268292683,
"grad_norm": 0.3496639132499695,
"learning_rate": 1.2863734927012094e-07,
"loss": 0.029,
"step": 955
},
{
"epoch": 0.03730731707317073,
"grad_norm": 0.40560710430145264,
"learning_rate": 1.231165940486234e-07,
"loss": 0.0392,
"step": 956
},
{
"epoch": 0.03734634146341463,
"grad_norm": 0.4083835482597351,
"learning_rate": 1.1771618553447217e-07,
"loss": 0.0511,
"step": 957
},
{
"epoch": 0.037385365853658536,
"grad_norm": 0.30038732290267944,
"learning_rate": 1.1243618952994195e-07,
"loss": 0.0332,
"step": 958
},
{
"epoch": 0.03742439024390244,
"grad_norm": 0.4680466949939728,
"learning_rate": 1.0727667037011668e-07,
"loss": 0.0375,
"step": 959
},
{
"epoch": 0.03746341463414634,
"grad_norm": 4.1751203536987305,
"learning_rate": 1.0223769092211012e-07,
"loss": 0.0625,
"step": 960
},
{
"epoch": 0.037502439024390245,
"grad_norm": 0.3359062373638153,
"learning_rate": 9.731931258429638e-08,
"loss": 0.0339,
"step": 961
},
{
"epoch": 0.037541463414634144,
"grad_norm": 0.24274514615535736,
"learning_rate": 9.252159528556404e-08,
"loss": 0.0201,
"step": 962
},
{
"epoch": 0.03758048780487805,
"grad_norm": 0.1267894208431244,
"learning_rate": 8.784459748458318e-08,
"loss": 0.0012,
"step": 963
},
{
"epoch": 0.037619512195121954,
"grad_norm": 0.03637217357754707,
"learning_rate": 8.328837616909612e-08,
"loss": 0.0006,
"step": 964
},
{
"epoch": 0.03765853658536585,
"grad_norm": 5.283059597015381,
"learning_rate": 7.885298685522235e-08,
"loss": 0.0772,
"step": 965
},
{
"epoch": 0.03769756097560976,
"grad_norm": 0.20719438791275024,
"learning_rate": 7.453848358678018e-08,
"loss": 0.0364,
"step": 966
},
{
"epoch": 0.037736585365853656,
"grad_norm": 0.4098169207572937,
"learning_rate": 7.034491893463059e-08,
"loss": 0.0423,
"step": 967
},
{
"epoch": 0.03777560975609756,
"grad_norm": 0.4853972792625427,
"learning_rate": 6.627234399603554e-08,
"loss": 0.0411,
"step": 968
},
{
"epoch": 0.037814634146341466,
"grad_norm": 0.75942462682724,
"learning_rate": 6.232080839403631e-08,
"loss": 0.0452,
"step": 969
},
{
"epoch": 0.037853658536585365,
"grad_norm": 0.039565883576869965,
"learning_rate": 5.849036027684607e-08,
"loss": 0.0009,
"step": 970
},
{
"epoch": 0.03789268292682927,
"grad_norm": 0.23365069925785065,
"learning_rate": 5.4781046317267103e-08,
"loss": 0.0253,
"step": 971
},
{
"epoch": 0.03793170731707317,
"grad_norm": 0.16438551247119904,
"learning_rate": 5.119291171211793e-08,
"loss": 0.0385,
"step": 972
},
{
"epoch": 0.037970731707317074,
"grad_norm": 0.6953434944152832,
"learning_rate": 4.772600018168816e-08,
"loss": 0.0368,
"step": 973
},
{
"epoch": 0.03800975609756098,
"grad_norm": 0.37828877568244934,
"learning_rate": 4.438035396920004e-08,
"loss": 0.051,
"step": 974
},
{
"epoch": 0.03804878048780488,
"grad_norm": 19.395727157592773,
"learning_rate": 4.115601384029666e-08,
"loss": 0.0747,
"step": 975
},
{
"epoch": 0.03808780487804878,
"grad_norm": 0.08213439583778381,
"learning_rate": 3.805301908254455e-08,
"loss": 0.0155,
"step": 976
},
{
"epoch": 0.03812682926829268,
"grad_norm": 19.913890838623047,
"learning_rate": 3.50714075049563e-08,
"loss": 0.338,
"step": 977
},
{
"epoch": 0.038165853658536586,
"grad_norm": 0.12111137807369232,
"learning_rate": 3.22112154375287e-08,
"loss": 0.0115,
"step": 978
},
{
"epoch": 0.03820487804878049,
"grad_norm": 1.0126404762268066,
"learning_rate": 2.947247773079753e-08,
"loss": 0.0616,
"step": 979
},
{
"epoch": 0.03824390243902439,
"grad_norm": 0.5435498356819153,
"learning_rate": 2.6855227755419046e-08,
"loss": 0.0404,
"step": 980
},
{
"epoch": 0.038282926829268295,
"grad_norm": 0.7325711846351624,
"learning_rate": 2.4359497401758026e-08,
"loss": 0.0458,
"step": 981
},
{
"epoch": 0.03832195121951219,
"grad_norm": 2.582336187362671,
"learning_rate": 2.1985317079500358e-08,
"loss": 0.0515,
"step": 982
},
{
"epoch": 0.0383609756097561,
"grad_norm": 0.045450713485479355,
"learning_rate": 1.973271571728441e-08,
"loss": 0.0005,
"step": 983
},
{
"epoch": 0.0384,
"grad_norm": 0.1935480535030365,
"learning_rate": 1.7601720762346895e-08,
"loss": 0.0184,
"step": 984
},
{
"epoch": 0.0384390243902439,
"grad_norm": 0.7397153377532959,
"learning_rate": 1.5592358180189782e-08,
"loss": 0.0432,
"step": 985
},
{
"epoch": 0.03847804878048781,
"grad_norm": 0.3976181745529175,
"learning_rate": 1.370465245426167e-08,
"loss": 0.0323,
"step": 986
},
{
"epoch": 0.038517073170731705,
"grad_norm": 0.12866108119487762,
"learning_rate": 1.1938626585660252e-08,
"loss": 0.0122,
"step": 987
},
{
"epoch": 0.03855609756097561,
"grad_norm": 0.47660142183303833,
"learning_rate": 1.0294302092853647e-08,
"loss": 0.0424,
"step": 988
},
{
"epoch": 0.03859512195121951,
"grad_norm": 0.2763524353504181,
"learning_rate": 8.771699011416169e-09,
"loss": 0.0163,
"step": 989
},
{
"epoch": 0.038634146341463414,
"grad_norm": 0.20618335902690887,
"learning_rate": 7.370835893788508e-09,
"loss": 0.0444,
"step": 990
},
{
"epoch": 0.03867317073170732,
"grad_norm": 1.1208971738815308,
"learning_rate": 6.091729809042379e-09,
"loss": 0.0256,
"step": 991
},
{
"epoch": 0.03871219512195122,
"grad_norm": 0.04721367359161377,
"learning_rate": 4.9343963426840006e-09,
"loss": 0.0161,
"step": 992
},
{
"epoch": 0.03875121951219512,
"grad_norm": 0.10911668092012405,
"learning_rate": 3.898849596456477e-09,
"loss": 0.0009,
"step": 993
},
{
"epoch": 0.03879024390243902,
"grad_norm": 0.1727072298526764,
"learning_rate": 2.9851021881688314e-09,
"loss": 0.0327,
"step": 994
},
{
"epoch": 0.03882926829268293,
"grad_norm": 0.45987993478775024,
"learning_rate": 2.193165251545004e-09,
"loss": 0.0577,
"step": 995
},
{
"epoch": 0.03886829268292683,
"grad_norm": 0.7591833472251892,
"learning_rate": 1.5230484360873043e-09,
"loss": 0.057,
"step": 996
},
{
"epoch": 0.03890731707317073,
"grad_norm": 0.7695837020874023,
"learning_rate": 9.74759906957612e-10,
"loss": 0.0649,
"step": 997
},
{
"epoch": 0.038946341463414635,
"grad_norm": 1.0995266437530518,
"learning_rate": 5.483063448785686e-10,
"loss": 0.0745,
"step": 998
},
{
"epoch": 0.038985365853658534,
"grad_norm": 0.21566379070281982,
"learning_rate": 2.436929460525317e-10,
"loss": 0.0013,
"step": 999
},
{
"epoch": 0.03902439024390244,
"grad_norm": 0.3683568835258484,
"learning_rate": 6.092342209607083e-11,
"loss": 0.0338,
"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
}