barnav's picture
Training in progress, step 200, checkpoint
a26bffa verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.5925925925925926,
"eval_steps": 100,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002962962962962963,
"grad_norm": 4.3932013511657715,
"learning_rate": 0.0,
"loss": 1.122,
"step": 1
},
{
"epoch": 0.005925925925925926,
"grad_norm": 4.835041522979736,
"learning_rate": 4.000000000000001e-06,
"loss": 1.262,
"step": 2
},
{
"epoch": 0.008888888888888889,
"grad_norm": 4.558024883270264,
"learning_rate": 8.000000000000001e-06,
"loss": 1.1905,
"step": 3
},
{
"epoch": 0.011851851851851851,
"grad_norm": 4.187239170074463,
"learning_rate": 1.2e-05,
"loss": 1.0289,
"step": 4
},
{
"epoch": 0.014814814814814815,
"grad_norm": 3.6275532245635986,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.9252,
"step": 5
},
{
"epoch": 0.017777777777777778,
"grad_norm": 4.801448345184326,
"learning_rate": 2e-05,
"loss": 1.134,
"step": 6
},
{
"epoch": 0.02074074074074074,
"grad_norm": 4.969509601593018,
"learning_rate": 1.993993993993994e-05,
"loss": 1.1101,
"step": 7
},
{
"epoch": 0.023703703703703703,
"grad_norm": 6.065815448760986,
"learning_rate": 1.987987987987988e-05,
"loss": 1.2835,
"step": 8
},
{
"epoch": 0.02666666666666667,
"grad_norm": 4.675291538238525,
"learning_rate": 1.981981981981982e-05,
"loss": 0.9603,
"step": 9
},
{
"epoch": 0.02962962962962963,
"grad_norm": 5.45444917678833,
"learning_rate": 1.9759759759759763e-05,
"loss": 0.9926,
"step": 10
},
{
"epoch": 0.03259259259259259,
"grad_norm": 4.381354808807373,
"learning_rate": 1.9699699699699702e-05,
"loss": 0.7942,
"step": 11
},
{
"epoch": 0.035555555555555556,
"grad_norm": 6.248989105224609,
"learning_rate": 1.963963963963964e-05,
"loss": 1.1086,
"step": 12
},
{
"epoch": 0.03851851851851852,
"grad_norm": 5.123929023742676,
"learning_rate": 1.957957957957958e-05,
"loss": 0.778,
"step": 13
},
{
"epoch": 0.04148148148148148,
"grad_norm": 6.84102725982666,
"learning_rate": 1.951951951951952e-05,
"loss": 1.0916,
"step": 14
},
{
"epoch": 0.044444444444444446,
"grad_norm": 6.978565692901611,
"learning_rate": 1.9459459459459463e-05,
"loss": 0.982,
"step": 15
},
{
"epoch": 0.047407407407407405,
"grad_norm": 8.47054386138916,
"learning_rate": 1.9399399399399402e-05,
"loss": 1.0711,
"step": 16
},
{
"epoch": 0.05037037037037037,
"grad_norm": 7.508791923522949,
"learning_rate": 1.9339339339339342e-05,
"loss": 0.9112,
"step": 17
},
{
"epoch": 0.05333333333333334,
"grad_norm": 9.517326354980469,
"learning_rate": 1.927927927927928e-05,
"loss": 1.0273,
"step": 18
},
{
"epoch": 0.056296296296296296,
"grad_norm": 10.264245986938477,
"learning_rate": 1.921921921921922e-05,
"loss": 1.087,
"step": 19
},
{
"epoch": 0.05925925925925926,
"grad_norm": 9.43092155456543,
"learning_rate": 1.915915915915916e-05,
"loss": 1.0215,
"step": 20
},
{
"epoch": 0.06222222222222222,
"grad_norm": 8.558907508850098,
"learning_rate": 1.90990990990991e-05,
"loss": 0.82,
"step": 21
},
{
"epoch": 0.06518518518518518,
"grad_norm": 14.64515495300293,
"learning_rate": 1.903903903903904e-05,
"loss": 1.4202,
"step": 22
},
{
"epoch": 0.06814814814814815,
"grad_norm": 10.740450859069824,
"learning_rate": 1.8978978978978982e-05,
"loss": 1.1039,
"step": 23
},
{
"epoch": 0.07111111111111111,
"grad_norm": 9.0601224899292,
"learning_rate": 1.891891891891892e-05,
"loss": 1.0625,
"step": 24
},
{
"epoch": 0.07407407407407407,
"grad_norm": 11.073591232299805,
"learning_rate": 1.885885885885886e-05,
"loss": 0.9269,
"step": 25
},
{
"epoch": 0.07703703703703704,
"grad_norm": 10.829946517944336,
"learning_rate": 1.87987987987988e-05,
"loss": 0.9481,
"step": 26
},
{
"epoch": 0.08,
"grad_norm": 14.61500358581543,
"learning_rate": 1.873873873873874e-05,
"loss": 1.2452,
"step": 27
},
{
"epoch": 0.08296296296296296,
"grad_norm": 11.676919937133789,
"learning_rate": 1.8678678678678682e-05,
"loss": 1.0379,
"step": 28
},
{
"epoch": 0.08592592592592592,
"grad_norm": 10.817378997802734,
"learning_rate": 1.861861861861862e-05,
"loss": 0.8553,
"step": 29
},
{
"epoch": 0.08888888888888889,
"grad_norm": 11.56974983215332,
"learning_rate": 1.855855855855856e-05,
"loss": 0.8584,
"step": 30
},
{
"epoch": 0.09185185185185185,
"grad_norm": 10.188511848449707,
"learning_rate": 1.84984984984985e-05,
"loss": 0.7416,
"step": 31
},
{
"epoch": 0.09481481481481481,
"grad_norm": 10.488707542419434,
"learning_rate": 1.843843843843844e-05,
"loss": 0.5713,
"step": 32
},
{
"epoch": 0.09777777777777778,
"grad_norm": 11.057037353515625,
"learning_rate": 1.8378378378378383e-05,
"loss": 0.6974,
"step": 33
},
{
"epoch": 0.10074074074074074,
"grad_norm": 11.147880554199219,
"learning_rate": 1.831831831831832e-05,
"loss": 0.7529,
"step": 34
},
{
"epoch": 0.1037037037037037,
"grad_norm": 12.731097221374512,
"learning_rate": 1.8258258258258258e-05,
"loss": 0.9118,
"step": 35
},
{
"epoch": 0.10666666666666667,
"grad_norm": 13.77258014678955,
"learning_rate": 1.81981981981982e-05,
"loss": 0.8346,
"step": 36
},
{
"epoch": 0.10962962962962963,
"grad_norm": 13.153702735900879,
"learning_rate": 1.813813813813814e-05,
"loss": 0.9084,
"step": 37
},
{
"epoch": 0.11259259259259259,
"grad_norm": 12.729766845703125,
"learning_rate": 1.807807807807808e-05,
"loss": 0.7516,
"step": 38
},
{
"epoch": 0.11555555555555555,
"grad_norm": 12.333885192871094,
"learning_rate": 1.801801801801802e-05,
"loss": 0.7124,
"step": 39
},
{
"epoch": 0.11851851851851852,
"grad_norm": 10.395763397216797,
"learning_rate": 1.795795795795796e-05,
"loss": 0.6642,
"step": 40
},
{
"epoch": 0.12148148148148148,
"grad_norm": 11.341495513916016,
"learning_rate": 1.78978978978979e-05,
"loss": 0.7956,
"step": 41
},
{
"epoch": 0.12444444444444444,
"grad_norm": 12.643635749816895,
"learning_rate": 1.783783783783784e-05,
"loss": 1.012,
"step": 42
},
{
"epoch": 0.1274074074074074,
"grad_norm": 10.671773910522461,
"learning_rate": 1.7777777777777777e-05,
"loss": 0.6679,
"step": 43
},
{
"epoch": 0.13037037037037036,
"grad_norm": 10.571871757507324,
"learning_rate": 1.771771771771772e-05,
"loss": 0.6476,
"step": 44
},
{
"epoch": 0.13333333333333333,
"grad_norm": 9.936208724975586,
"learning_rate": 1.765765765765766e-05,
"loss": 0.6951,
"step": 45
},
{
"epoch": 0.1362962962962963,
"grad_norm": 13.039535522460938,
"learning_rate": 1.7597597597597598e-05,
"loss": 0.8956,
"step": 46
},
{
"epoch": 0.13925925925925925,
"grad_norm": 15.345221519470215,
"learning_rate": 1.7537537537537538e-05,
"loss": 1.0664,
"step": 47
},
{
"epoch": 0.14222222222222222,
"grad_norm": 12.190373420715332,
"learning_rate": 1.7477477477477477e-05,
"loss": 0.6127,
"step": 48
},
{
"epoch": 0.1451851851851852,
"grad_norm": 12.246444702148438,
"learning_rate": 1.741741741741742e-05,
"loss": 0.7511,
"step": 49
},
{
"epoch": 0.14814814814814814,
"grad_norm": 9.051176071166992,
"learning_rate": 1.735735735735736e-05,
"loss": 0.5273,
"step": 50
},
{
"epoch": 0.1511111111111111,
"grad_norm": 14.076638221740723,
"learning_rate": 1.72972972972973e-05,
"loss": 0.8716,
"step": 51
},
{
"epoch": 0.15407407407407409,
"grad_norm": 16.608592987060547,
"learning_rate": 1.7237237237237238e-05,
"loss": 0.7523,
"step": 52
},
{
"epoch": 0.15703703703703703,
"grad_norm": 11.700766563415527,
"learning_rate": 1.7177177177177177e-05,
"loss": 0.6512,
"step": 53
},
{
"epoch": 0.16,
"grad_norm": 18.934720993041992,
"learning_rate": 1.711711711711712e-05,
"loss": 0.8111,
"step": 54
},
{
"epoch": 0.16296296296296298,
"grad_norm": 12.533053398132324,
"learning_rate": 1.705705705705706e-05,
"loss": 0.5017,
"step": 55
},
{
"epoch": 0.16592592592592592,
"grad_norm": 16.80641746520996,
"learning_rate": 1.6996996996997e-05,
"loss": 0.7349,
"step": 56
},
{
"epoch": 0.1688888888888889,
"grad_norm": 17.351354598999023,
"learning_rate": 1.693693693693694e-05,
"loss": 0.5469,
"step": 57
},
{
"epoch": 0.17185185185185184,
"grad_norm": 12.17600154876709,
"learning_rate": 1.6876876876876878e-05,
"loss": 0.5454,
"step": 58
},
{
"epoch": 0.1748148148148148,
"grad_norm": 15.175498962402344,
"learning_rate": 1.6816816816816817e-05,
"loss": 0.7156,
"step": 59
},
{
"epoch": 0.17777777777777778,
"grad_norm": 11.376442909240723,
"learning_rate": 1.6756756756756757e-05,
"loss": 0.6287,
"step": 60
},
{
"epoch": 0.18074074074074073,
"grad_norm": 16.71291732788086,
"learning_rate": 1.6696696696696696e-05,
"loss": 0.5942,
"step": 61
},
{
"epoch": 0.1837037037037037,
"grad_norm": 10.104132652282715,
"learning_rate": 1.663663663663664e-05,
"loss": 0.4791,
"step": 62
},
{
"epoch": 0.18666666666666668,
"grad_norm": 11.949193000793457,
"learning_rate": 1.6576576576576578e-05,
"loss": 0.6793,
"step": 63
},
{
"epoch": 0.18962962962962962,
"grad_norm": 9.24198055267334,
"learning_rate": 1.6516516516516518e-05,
"loss": 0.5549,
"step": 64
},
{
"epoch": 0.1925925925925926,
"grad_norm": 9.324671745300293,
"learning_rate": 1.6456456456456457e-05,
"loss": 0.5719,
"step": 65
},
{
"epoch": 0.19555555555555557,
"grad_norm": 11.018183708190918,
"learning_rate": 1.6396396396396396e-05,
"loss": 0.5578,
"step": 66
},
{
"epoch": 0.1985185185185185,
"grad_norm": 9.80178165435791,
"learning_rate": 1.633633633633634e-05,
"loss": 0.648,
"step": 67
},
{
"epoch": 0.20148148148148148,
"grad_norm": 13.870157241821289,
"learning_rate": 1.627627627627628e-05,
"loss": 0.755,
"step": 68
},
{
"epoch": 0.20444444444444446,
"grad_norm": 14.881416320800781,
"learning_rate": 1.6216216216216218e-05,
"loss": 0.6568,
"step": 69
},
{
"epoch": 0.2074074074074074,
"grad_norm": 13.221749305725098,
"learning_rate": 1.6156156156156157e-05,
"loss": 0.4347,
"step": 70
},
{
"epoch": 0.21037037037037037,
"grad_norm": 15.9087553024292,
"learning_rate": 1.6096096096096097e-05,
"loss": 0.7353,
"step": 71
},
{
"epoch": 0.21333333333333335,
"grad_norm": 18.63846206665039,
"learning_rate": 1.6036036036036036e-05,
"loss": 0.6407,
"step": 72
},
{
"epoch": 0.2162962962962963,
"grad_norm": 16.938674926757812,
"learning_rate": 1.5975975975975976e-05,
"loss": 0.4618,
"step": 73
},
{
"epoch": 0.21925925925925926,
"grad_norm": 19.83721160888672,
"learning_rate": 1.591591591591592e-05,
"loss": 0.4019,
"step": 74
},
{
"epoch": 0.2222222222222222,
"grad_norm": 23.562578201293945,
"learning_rate": 1.5855855855855858e-05,
"loss": 0.5107,
"step": 75
},
{
"epoch": 0.22518518518518518,
"grad_norm": 22.505775451660156,
"learning_rate": 1.5795795795795797e-05,
"loss": 0.4955,
"step": 76
},
{
"epoch": 0.22814814814814816,
"grad_norm": 23.18890380859375,
"learning_rate": 1.5735735735735737e-05,
"loss": 0.5082,
"step": 77
},
{
"epoch": 0.2311111111111111,
"grad_norm": 29.059083938598633,
"learning_rate": 1.5675675675675676e-05,
"loss": 0.5518,
"step": 78
},
{
"epoch": 0.23407407407407407,
"grad_norm": 24.563520431518555,
"learning_rate": 1.5615615615615616e-05,
"loss": 0.3839,
"step": 79
},
{
"epoch": 0.23703703703703705,
"grad_norm": 18.441802978515625,
"learning_rate": 1.555555555555556e-05,
"loss": 0.4532,
"step": 80
},
{
"epoch": 0.24,
"grad_norm": 19.27843475341797,
"learning_rate": 1.5495495495495498e-05,
"loss": 0.431,
"step": 81
},
{
"epoch": 0.24296296296296296,
"grad_norm": 25.197582244873047,
"learning_rate": 1.5435435435435437e-05,
"loss": 0.3855,
"step": 82
},
{
"epoch": 0.24592592592592594,
"grad_norm": 14.846511840820312,
"learning_rate": 1.5375375375375377e-05,
"loss": 0.3664,
"step": 83
},
{
"epoch": 0.24888888888888888,
"grad_norm": 19.469579696655273,
"learning_rate": 1.5315315315315316e-05,
"loss": 0.4523,
"step": 84
},
{
"epoch": 0.2518518518518518,
"grad_norm": 15.252752304077148,
"learning_rate": 1.5255255255255257e-05,
"loss": 0.551,
"step": 85
},
{
"epoch": 0.2548148148148148,
"grad_norm": 11.770284652709961,
"learning_rate": 1.5195195195195196e-05,
"loss": 0.3996,
"step": 86
},
{
"epoch": 0.2577777777777778,
"grad_norm": 12.881220817565918,
"learning_rate": 1.5135135135135138e-05,
"loss": 0.5693,
"step": 87
},
{
"epoch": 0.2607407407407407,
"grad_norm": 11.658583641052246,
"learning_rate": 1.5075075075075077e-05,
"loss": 0.4513,
"step": 88
},
{
"epoch": 0.2637037037037037,
"grad_norm": 10.149765968322754,
"learning_rate": 1.5015015015015015e-05,
"loss": 0.3656,
"step": 89
},
{
"epoch": 0.26666666666666666,
"grad_norm": 7.5767822265625,
"learning_rate": 1.4954954954954957e-05,
"loss": 0.4198,
"step": 90
},
{
"epoch": 0.2696296296296296,
"grad_norm": 18.215837478637695,
"learning_rate": 1.4894894894894895e-05,
"loss": 0.4751,
"step": 91
},
{
"epoch": 0.2725925925925926,
"grad_norm": 15.47228717803955,
"learning_rate": 1.4834834834834836e-05,
"loss": 0.3909,
"step": 92
},
{
"epoch": 0.27555555555555555,
"grad_norm": 8.646990776062012,
"learning_rate": 1.4774774774774776e-05,
"loss": 0.2954,
"step": 93
},
{
"epoch": 0.2785185185185185,
"grad_norm": 10.278739929199219,
"learning_rate": 1.4714714714714715e-05,
"loss": 0.5366,
"step": 94
},
{
"epoch": 0.2814814814814815,
"grad_norm": 14.884415626525879,
"learning_rate": 1.4654654654654656e-05,
"loss": 0.4875,
"step": 95
},
{
"epoch": 0.28444444444444444,
"grad_norm": 7.219095230102539,
"learning_rate": 1.4594594594594596e-05,
"loss": 0.4121,
"step": 96
},
{
"epoch": 0.2874074074074074,
"grad_norm": 6.076047897338867,
"learning_rate": 1.4534534534534537e-05,
"loss": 0.4601,
"step": 97
},
{
"epoch": 0.2903703703703704,
"grad_norm": 5.5280680656433105,
"learning_rate": 1.4474474474474476e-05,
"loss": 0.3693,
"step": 98
},
{
"epoch": 0.29333333333333333,
"grad_norm": 5.4252777099609375,
"learning_rate": 1.4414414414414416e-05,
"loss": 0.5719,
"step": 99
},
{
"epoch": 0.2962962962962963,
"grad_norm": 3.8267464637756348,
"learning_rate": 1.4354354354354357e-05,
"loss": 0.3806,
"step": 100
},
{
"epoch": 0.2962962962962963,
"eval_loss": 0.3932308554649353,
"eval_runtime": 12.1133,
"eval_samples_per_second": 49.532,
"eval_steps_per_second": 24.766,
"step": 100
},
{
"epoch": 0.2992592592592593,
"grad_norm": 8.116653442382812,
"learning_rate": 1.4294294294294296e-05,
"loss": 0.3736,
"step": 101
},
{
"epoch": 0.3022222222222222,
"grad_norm": 5.39288330078125,
"learning_rate": 1.4234234234234234e-05,
"loss": 0.4315,
"step": 102
},
{
"epoch": 0.30518518518518517,
"grad_norm": 5.903439044952393,
"learning_rate": 1.4174174174174175e-05,
"loss": 0.5395,
"step": 103
},
{
"epoch": 0.30814814814814817,
"grad_norm": 3.8141722679138184,
"learning_rate": 1.4114114114114114e-05,
"loss": 0.3634,
"step": 104
},
{
"epoch": 0.3111111111111111,
"grad_norm": 4.634343147277832,
"learning_rate": 1.4054054054054055e-05,
"loss": 0.4299,
"step": 105
},
{
"epoch": 0.31407407407407406,
"grad_norm": 4.35814094543457,
"learning_rate": 1.3993993993993995e-05,
"loss": 0.335,
"step": 106
},
{
"epoch": 0.31703703703703706,
"grad_norm": 6.833477973937988,
"learning_rate": 1.3933933933933934e-05,
"loss": 0.4607,
"step": 107
},
{
"epoch": 0.32,
"grad_norm": 3.177086591720581,
"learning_rate": 1.3873873873873875e-05,
"loss": 0.3536,
"step": 108
},
{
"epoch": 0.32296296296296295,
"grad_norm": 7.129577159881592,
"learning_rate": 1.3813813813813815e-05,
"loss": 0.3965,
"step": 109
},
{
"epoch": 0.32592592592592595,
"grad_norm": 4.180286407470703,
"learning_rate": 1.3753753753753756e-05,
"loss": 0.381,
"step": 110
},
{
"epoch": 0.3288888888888889,
"grad_norm": 3.6946301460266113,
"learning_rate": 1.3693693693693695e-05,
"loss": 0.4034,
"step": 111
},
{
"epoch": 0.33185185185185184,
"grad_norm": 3.360057830810547,
"learning_rate": 1.3633633633633635e-05,
"loss": 0.3664,
"step": 112
},
{
"epoch": 0.3348148148148148,
"grad_norm": 3.8995361328125,
"learning_rate": 1.3573573573573576e-05,
"loss": 0.3204,
"step": 113
},
{
"epoch": 0.3377777777777778,
"grad_norm": 4.03761100769043,
"learning_rate": 1.3513513513513515e-05,
"loss": 0.4419,
"step": 114
},
{
"epoch": 0.34074074074074073,
"grad_norm": 4.310070514678955,
"learning_rate": 1.3453453453453456e-05,
"loss": 0.3727,
"step": 115
},
{
"epoch": 0.3437037037037037,
"grad_norm": 4.510432243347168,
"learning_rate": 1.3393393393393394e-05,
"loss": 0.5438,
"step": 116
},
{
"epoch": 0.3466666666666667,
"grad_norm": 3.73860239982605,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.3544,
"step": 117
},
{
"epoch": 0.3496296296296296,
"grad_norm": 4.66259765625,
"learning_rate": 1.3273273273273274e-05,
"loss": 0.4867,
"step": 118
},
{
"epoch": 0.35259259259259257,
"grad_norm": 5.015610694885254,
"learning_rate": 1.3213213213213214e-05,
"loss": 0.5052,
"step": 119
},
{
"epoch": 0.35555555555555557,
"grad_norm": 2.8242692947387695,
"learning_rate": 1.3153153153153155e-05,
"loss": 0.3435,
"step": 120
},
{
"epoch": 0.3585185185185185,
"grad_norm": 3.423738956451416,
"learning_rate": 1.3093093093093094e-05,
"loss": 0.4255,
"step": 121
},
{
"epoch": 0.36148148148148146,
"grad_norm": 4.251372337341309,
"learning_rate": 1.3033033033033034e-05,
"loss": 0.3419,
"step": 122
},
{
"epoch": 0.36444444444444446,
"grad_norm": 3.6724612712860107,
"learning_rate": 1.2972972972972975e-05,
"loss": 0.4328,
"step": 123
},
{
"epoch": 0.3674074074074074,
"grad_norm": 4.145914077758789,
"learning_rate": 1.2912912912912914e-05,
"loss": 0.4995,
"step": 124
},
{
"epoch": 0.37037037037037035,
"grad_norm": 4.318111419677734,
"learning_rate": 1.2852852852852854e-05,
"loss": 0.3894,
"step": 125
},
{
"epoch": 0.37333333333333335,
"grad_norm": 4.244164943695068,
"learning_rate": 1.2792792792792795e-05,
"loss": 0.4767,
"step": 126
},
{
"epoch": 0.3762962962962963,
"grad_norm": 3.5205907821655273,
"learning_rate": 1.2732732732732732e-05,
"loss": 0.321,
"step": 127
},
{
"epoch": 0.37925925925925924,
"grad_norm": 3.4363930225372314,
"learning_rate": 1.2672672672672675e-05,
"loss": 0.2923,
"step": 128
},
{
"epoch": 0.38222222222222224,
"grad_norm": 3.424726963043213,
"learning_rate": 1.2612612612612613e-05,
"loss": 0.3507,
"step": 129
},
{
"epoch": 0.3851851851851852,
"grad_norm": 5.809702396392822,
"learning_rate": 1.2552552552552552e-05,
"loss": 0.3021,
"step": 130
},
{
"epoch": 0.38814814814814813,
"grad_norm": 4.665796756744385,
"learning_rate": 1.2492492492492493e-05,
"loss": 0.4211,
"step": 131
},
{
"epoch": 0.39111111111111113,
"grad_norm": 3.970022201538086,
"learning_rate": 1.2432432432432433e-05,
"loss": 0.353,
"step": 132
},
{
"epoch": 0.3940740740740741,
"grad_norm": 3.3230977058410645,
"learning_rate": 1.2372372372372374e-05,
"loss": 0.4631,
"step": 133
},
{
"epoch": 0.397037037037037,
"grad_norm": 2.7031373977661133,
"learning_rate": 1.2312312312312313e-05,
"loss": 0.426,
"step": 134
},
{
"epoch": 0.4,
"grad_norm": 4.042789459228516,
"learning_rate": 1.2252252252252253e-05,
"loss": 0.3452,
"step": 135
},
{
"epoch": 0.40296296296296297,
"grad_norm": 4.957043170928955,
"learning_rate": 1.2192192192192194e-05,
"loss": 0.3273,
"step": 136
},
{
"epoch": 0.4059259259259259,
"grad_norm": 4.024094581604004,
"learning_rate": 1.2132132132132133e-05,
"loss": 0.3293,
"step": 137
},
{
"epoch": 0.4088888888888889,
"grad_norm": 3.3985772132873535,
"learning_rate": 1.2072072072072074e-05,
"loss": 0.3796,
"step": 138
},
{
"epoch": 0.41185185185185186,
"grad_norm": 2.8760104179382324,
"learning_rate": 1.2012012012012014e-05,
"loss": 0.3718,
"step": 139
},
{
"epoch": 0.4148148148148148,
"grad_norm": 4.576436519622803,
"learning_rate": 1.1951951951951951e-05,
"loss": 0.5758,
"step": 140
},
{
"epoch": 0.4177777777777778,
"grad_norm": 3.618910551071167,
"learning_rate": 1.1891891891891894e-05,
"loss": 0.4362,
"step": 141
},
{
"epoch": 0.42074074074074075,
"grad_norm": 7.472576141357422,
"learning_rate": 1.1831831831831832e-05,
"loss": 0.2875,
"step": 142
},
{
"epoch": 0.4237037037037037,
"grad_norm": 3.3167757987976074,
"learning_rate": 1.1771771771771771e-05,
"loss": 0.4415,
"step": 143
},
{
"epoch": 0.4266666666666667,
"grad_norm": 3.0888941287994385,
"learning_rate": 1.1711711711711713e-05,
"loss": 0.3728,
"step": 144
},
{
"epoch": 0.42962962962962964,
"grad_norm": 4.755927562713623,
"learning_rate": 1.1651651651651652e-05,
"loss": 0.3541,
"step": 145
},
{
"epoch": 0.4325925925925926,
"grad_norm": 2.7480132579803467,
"learning_rate": 1.1591591591591593e-05,
"loss": 0.2377,
"step": 146
},
{
"epoch": 0.43555555555555553,
"grad_norm": 4.149758338928223,
"learning_rate": 1.1531531531531532e-05,
"loss": 0.2628,
"step": 147
},
{
"epoch": 0.43851851851851853,
"grad_norm": 2.735320806503296,
"learning_rate": 1.1471471471471472e-05,
"loss": 0.36,
"step": 148
},
{
"epoch": 0.4414814814814815,
"grad_norm": 3.5400054454803467,
"learning_rate": 1.1411411411411413e-05,
"loss": 0.4208,
"step": 149
},
{
"epoch": 0.4444444444444444,
"grad_norm": 4.006497383117676,
"learning_rate": 1.1351351351351352e-05,
"loss": 0.4096,
"step": 150
},
{
"epoch": 0.4474074074074074,
"grad_norm": 5.038761138916016,
"learning_rate": 1.1291291291291293e-05,
"loss": 0.4994,
"step": 151
},
{
"epoch": 0.45037037037037037,
"grad_norm": 4.316198348999023,
"learning_rate": 1.1231231231231233e-05,
"loss": 0.3145,
"step": 152
},
{
"epoch": 0.4533333333333333,
"grad_norm": 4.305347919464111,
"learning_rate": 1.117117117117117e-05,
"loss": 0.3376,
"step": 153
},
{
"epoch": 0.4562962962962963,
"grad_norm": 5.134220123291016,
"learning_rate": 1.1111111111111113e-05,
"loss": 0.4161,
"step": 154
},
{
"epoch": 0.45925925925925926,
"grad_norm": 3.8545286655426025,
"learning_rate": 1.1051051051051051e-05,
"loss": 0.3128,
"step": 155
},
{
"epoch": 0.4622222222222222,
"grad_norm": 3.019829511642456,
"learning_rate": 1.0990990990990992e-05,
"loss": 0.151,
"step": 156
},
{
"epoch": 0.4651851851851852,
"grad_norm": 4.200697422027588,
"learning_rate": 1.0930930930930932e-05,
"loss": 0.4173,
"step": 157
},
{
"epoch": 0.46814814814814815,
"grad_norm": 3.808044195175171,
"learning_rate": 1.0870870870870871e-05,
"loss": 0.3183,
"step": 158
},
{
"epoch": 0.4711111111111111,
"grad_norm": 3.6768927574157715,
"learning_rate": 1.0810810810810812e-05,
"loss": 0.4224,
"step": 159
},
{
"epoch": 0.4740740740740741,
"grad_norm": 5.188108444213867,
"learning_rate": 1.0750750750750751e-05,
"loss": 0.324,
"step": 160
},
{
"epoch": 0.47703703703703704,
"grad_norm": 4.247463703155518,
"learning_rate": 1.0690690690690693e-05,
"loss": 0.3786,
"step": 161
},
{
"epoch": 0.48,
"grad_norm": 5.044620037078857,
"learning_rate": 1.0630630630630632e-05,
"loss": 0.3001,
"step": 162
},
{
"epoch": 0.482962962962963,
"grad_norm": 4.548891544342041,
"learning_rate": 1.0570570570570571e-05,
"loss": 0.3384,
"step": 163
},
{
"epoch": 0.48592592592592593,
"grad_norm": 4.440279960632324,
"learning_rate": 1.0510510510510512e-05,
"loss": 0.2924,
"step": 164
},
{
"epoch": 0.4888888888888889,
"grad_norm": 3.8252761363983154,
"learning_rate": 1.0450450450450452e-05,
"loss": 0.3676,
"step": 165
},
{
"epoch": 0.4918518518518519,
"grad_norm": 5.11232852935791,
"learning_rate": 1.039039039039039e-05,
"loss": 0.2977,
"step": 166
},
{
"epoch": 0.4948148148148148,
"grad_norm": 3.689526081085205,
"learning_rate": 1.033033033033033e-05,
"loss": 0.3096,
"step": 167
},
{
"epoch": 0.49777777777777776,
"grad_norm": 5.140994071960449,
"learning_rate": 1.027027027027027e-05,
"loss": 0.5585,
"step": 168
},
{
"epoch": 0.5007407407407407,
"grad_norm": 5.232369899749756,
"learning_rate": 1.0210210210210211e-05,
"loss": 0.3111,
"step": 169
},
{
"epoch": 0.5037037037037037,
"grad_norm": 4.676846504211426,
"learning_rate": 1.015015015015015e-05,
"loss": 0.4102,
"step": 170
},
{
"epoch": 0.5066666666666667,
"grad_norm": 3.288276195526123,
"learning_rate": 1.009009009009009e-05,
"loss": 0.4111,
"step": 171
},
{
"epoch": 0.5096296296296297,
"grad_norm": 4.381916522979736,
"learning_rate": 1.0030030030030031e-05,
"loss": 0.4808,
"step": 172
},
{
"epoch": 0.5125925925925926,
"grad_norm": 4.530202388763428,
"learning_rate": 9.96996996996997e-06,
"loss": 0.4134,
"step": 173
},
{
"epoch": 0.5155555555555555,
"grad_norm": 3.8073766231536865,
"learning_rate": 9.90990990990991e-06,
"loss": 0.4222,
"step": 174
},
{
"epoch": 0.5185185185185185,
"grad_norm": 4.142769813537598,
"learning_rate": 9.849849849849851e-06,
"loss": 0.3753,
"step": 175
},
{
"epoch": 0.5214814814814814,
"grad_norm": 6.5730485916137695,
"learning_rate": 9.78978978978979e-06,
"loss": 0.4778,
"step": 176
},
{
"epoch": 0.5244444444444445,
"grad_norm": 3.555816173553467,
"learning_rate": 9.729729729729732e-06,
"loss": 0.244,
"step": 177
},
{
"epoch": 0.5274074074074074,
"grad_norm": 4.206595420837402,
"learning_rate": 9.669669669669671e-06,
"loss": 0.384,
"step": 178
},
{
"epoch": 0.5303703703703704,
"grad_norm": 5.356174945831299,
"learning_rate": 9.60960960960961e-06,
"loss": 0.4418,
"step": 179
},
{
"epoch": 0.5333333333333333,
"grad_norm": 2.7669811248779297,
"learning_rate": 9.54954954954955e-06,
"loss": 0.2553,
"step": 180
},
{
"epoch": 0.5362962962962963,
"grad_norm": 2.8105010986328125,
"learning_rate": 9.489489489489491e-06,
"loss": 0.3328,
"step": 181
},
{
"epoch": 0.5392592592592592,
"grad_norm": 4.279818534851074,
"learning_rate": 9.42942942942943e-06,
"loss": 0.3403,
"step": 182
},
{
"epoch": 0.5422222222222223,
"grad_norm": 2.6417016983032227,
"learning_rate": 9.36936936936937e-06,
"loss": 0.2617,
"step": 183
},
{
"epoch": 0.5451851851851852,
"grad_norm": 6.554925441741943,
"learning_rate": 9.30930930930931e-06,
"loss": 0.4989,
"step": 184
},
{
"epoch": 0.5481481481481482,
"grad_norm": 4.862616062164307,
"learning_rate": 9.24924924924925e-06,
"loss": 0.2661,
"step": 185
},
{
"epoch": 0.5511111111111111,
"grad_norm": 5.0014543533325195,
"learning_rate": 9.189189189189191e-06,
"loss": 0.4386,
"step": 186
},
{
"epoch": 0.554074074074074,
"grad_norm": 7.454826354980469,
"learning_rate": 9.129129129129129e-06,
"loss": 0.4617,
"step": 187
},
{
"epoch": 0.557037037037037,
"grad_norm": 3.3302102088928223,
"learning_rate": 9.06906906906907e-06,
"loss": 0.3056,
"step": 188
},
{
"epoch": 0.56,
"grad_norm": 5.1727681159973145,
"learning_rate": 9.00900900900901e-06,
"loss": 0.3877,
"step": 189
},
{
"epoch": 0.562962962962963,
"grad_norm": 5.255517482757568,
"learning_rate": 8.94894894894895e-06,
"loss": 0.41,
"step": 190
},
{
"epoch": 0.5659259259259259,
"grad_norm": 3.0652015209198,
"learning_rate": 8.888888888888888e-06,
"loss": 0.3192,
"step": 191
},
{
"epoch": 0.5688888888888889,
"grad_norm": 4.500412940979004,
"learning_rate": 8.82882882882883e-06,
"loss": 0.3249,
"step": 192
},
{
"epoch": 0.5718518518518518,
"grad_norm": 4.083544731140137,
"learning_rate": 8.768768768768769e-06,
"loss": 0.4206,
"step": 193
},
{
"epoch": 0.5748148148148148,
"grad_norm": 3.433180093765259,
"learning_rate": 8.70870870870871e-06,
"loss": 0.3948,
"step": 194
},
{
"epoch": 0.5777777777777777,
"grad_norm": 5.444077968597412,
"learning_rate": 8.64864864864865e-06,
"loss": 0.3302,
"step": 195
},
{
"epoch": 0.5807407407407408,
"grad_norm": 3.774658441543579,
"learning_rate": 8.588588588588589e-06,
"loss": 0.2912,
"step": 196
},
{
"epoch": 0.5837037037037037,
"grad_norm": 2.684931755065918,
"learning_rate": 8.52852852852853e-06,
"loss": 0.3227,
"step": 197
},
{
"epoch": 0.5866666666666667,
"grad_norm": 4.436760425567627,
"learning_rate": 8.46846846846847e-06,
"loss": 0.3615,
"step": 198
},
{
"epoch": 0.5896296296296296,
"grad_norm": 3.151691198348999,
"learning_rate": 8.408408408408409e-06,
"loss": 0.2977,
"step": 199
},
{
"epoch": 0.5925925925925926,
"grad_norm": 4.275235176086426,
"learning_rate": 8.348348348348348e-06,
"loss": 0.2899,
"step": 200
},
{
"epoch": 0.5925925925925926,
"eval_loss": 0.34511882066726685,
"eval_runtime": 11.6694,
"eval_samples_per_second": 51.416,
"eval_steps_per_second": 25.708,
"step": 200
}
],
"logging_steps": 1,
"max_steps": 338,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 7235103620736000.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}