mink-ep2-ckpt / trainer_state.json
Fizzarolli's picture
Upload folder using huggingface_hub
c474ed8 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9953521630318198,
"eval_steps": 500,
"global_step": 174,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01144082946013586,
"grad_norm": 4.3125,
"learning_rate": 2.0000000000000002e-07,
"loss": 1.0962,
"step": 1
},
{
"epoch": 0.02288165892027172,
"grad_norm": 40.25,
"learning_rate": 4.0000000000000003e-07,
"loss": 1.0789,
"step": 2
},
{
"epoch": 0.03432248838040758,
"grad_norm": 14.8125,
"learning_rate": 6.000000000000001e-07,
"loss": 1.0865,
"step": 3
},
{
"epoch": 0.04576331784054344,
"grad_norm": 4.1875,
"learning_rate": 8.000000000000001e-07,
"loss": 1.0257,
"step": 4
},
{
"epoch": 0.0572041473006793,
"grad_norm": 5.09375,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.0766,
"step": 5
},
{
"epoch": 0.06864497676081516,
"grad_norm": 27.375,
"learning_rate": 1.2000000000000002e-06,
"loss": 1.0548,
"step": 6
},
{
"epoch": 0.08008580622095102,
"grad_norm": 2.734375,
"learning_rate": 1.4000000000000001e-06,
"loss": 1.0669,
"step": 7
},
{
"epoch": 0.09152663568108688,
"grad_norm": 6.5,
"learning_rate": 1.6000000000000001e-06,
"loss": 1.0566,
"step": 8
},
{
"epoch": 0.10296746514122274,
"grad_norm": 17.875,
"learning_rate": 1.8000000000000001e-06,
"loss": 1.2024,
"step": 9
},
{
"epoch": 0.1144082946013586,
"grad_norm": 3.5,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.0421,
"step": 10
},
{
"epoch": 0.12584912406149446,
"grad_norm": 3.5,
"learning_rate": 2.2e-06,
"loss": 1.1209,
"step": 11
},
{
"epoch": 0.13728995352163031,
"grad_norm": 3.734375,
"learning_rate": 2.4000000000000003e-06,
"loss": 1.1189,
"step": 12
},
{
"epoch": 0.14873078298176617,
"grad_norm": 2.609375,
"learning_rate": 2.6e-06,
"loss": 1.1231,
"step": 13
},
{
"epoch": 0.16017161244190203,
"grad_norm": 6.125,
"learning_rate": 2.8000000000000003e-06,
"loss": 1.1184,
"step": 14
},
{
"epoch": 0.1716124419020379,
"grad_norm": 5.0,
"learning_rate": 3e-06,
"loss": 1.0939,
"step": 15
},
{
"epoch": 0.18305327136217375,
"grad_norm": 55.5,
"learning_rate": 3.2000000000000003e-06,
"loss": 1.121,
"step": 16
},
{
"epoch": 0.1944941008223096,
"grad_norm": 2.96875,
"learning_rate": 3.4000000000000005e-06,
"loss": 1.0549,
"step": 17
},
{
"epoch": 0.20593493028244547,
"grad_norm": 2.65625,
"learning_rate": 3.6000000000000003e-06,
"loss": 1.0634,
"step": 18
},
{
"epoch": 0.21737575974258133,
"grad_norm": 2.640625,
"learning_rate": 3.8000000000000005e-06,
"loss": 1.1134,
"step": 19
},
{
"epoch": 0.2288165892027172,
"grad_norm": 3.328125,
"learning_rate": 4.000000000000001e-06,
"loss": 1.1133,
"step": 20
},
{
"epoch": 0.24025741866285305,
"grad_norm": 2.96875,
"learning_rate": 4.2000000000000004e-06,
"loss": 1.1252,
"step": 21
},
{
"epoch": 0.2516982481229889,
"grad_norm": 4.21875,
"learning_rate": 4.4e-06,
"loss": 1.187,
"step": 22
},
{
"epoch": 0.26313907758312477,
"grad_norm": 2.46875,
"learning_rate": 4.600000000000001e-06,
"loss": 1.103,
"step": 23
},
{
"epoch": 0.27457990704326063,
"grad_norm": 3.625,
"learning_rate": 4.800000000000001e-06,
"loss": 1.1059,
"step": 24
},
{
"epoch": 0.2860207365033965,
"grad_norm": 1.7265625,
"learning_rate": 5e-06,
"loss": 1.0725,
"step": 25
},
{
"epoch": 0.29746156596353235,
"grad_norm": 1.6875,
"learning_rate": 5.2e-06,
"loss": 1.0673,
"step": 26
},
{
"epoch": 0.3089023954236682,
"grad_norm": 1.828125,
"learning_rate": 5.400000000000001e-06,
"loss": 1.0517,
"step": 27
},
{
"epoch": 0.32034322488380407,
"grad_norm": 1.9375,
"learning_rate": 5.600000000000001e-06,
"loss": 1.1265,
"step": 28
},
{
"epoch": 0.3317840543439399,
"grad_norm": 1.6875,
"learning_rate": 5.8e-06,
"loss": 1.0353,
"step": 29
},
{
"epoch": 0.3432248838040758,
"grad_norm": 1.65625,
"learning_rate": 6e-06,
"loss": 1.1089,
"step": 30
},
{
"epoch": 0.35466571326421165,
"grad_norm": 2.28125,
"learning_rate": 6.200000000000001e-06,
"loss": 1.0071,
"step": 31
},
{
"epoch": 0.3661065427243475,
"grad_norm": 2.34375,
"learning_rate": 6.4000000000000006e-06,
"loss": 1.129,
"step": 32
},
{
"epoch": 0.37754737218448337,
"grad_norm": 2.09375,
"learning_rate": 6.600000000000001e-06,
"loss": 1.0258,
"step": 33
},
{
"epoch": 0.3889882016446192,
"grad_norm": 2.890625,
"learning_rate": 6.800000000000001e-06,
"loss": 1.0924,
"step": 34
},
{
"epoch": 0.4004290311047551,
"grad_norm": 4.4375,
"learning_rate": 7e-06,
"loss": 1.1275,
"step": 35
},
{
"epoch": 0.41186986056489094,
"grad_norm": 1.6328125,
"learning_rate": 7.2000000000000005e-06,
"loss": 1.0683,
"step": 36
},
{
"epoch": 0.4233106900250268,
"grad_norm": 2.34375,
"learning_rate": 7.4e-06,
"loss": 1.1141,
"step": 37
},
{
"epoch": 0.43475151948516266,
"grad_norm": 5.34375,
"learning_rate": 7.600000000000001e-06,
"loss": 1.0,
"step": 38
},
{
"epoch": 0.4461923489452985,
"grad_norm": 4.09375,
"learning_rate": 7.800000000000002e-06,
"loss": 1.1454,
"step": 39
},
{
"epoch": 0.4576331784054344,
"grad_norm": 3.3125,
"learning_rate": 8.000000000000001e-06,
"loss": 1.0984,
"step": 40
},
{
"epoch": 0.46907400786557024,
"grad_norm": 1.578125,
"learning_rate": 8.2e-06,
"loss": 1.0194,
"step": 41
},
{
"epoch": 0.4805148373257061,
"grad_norm": 7.03125,
"learning_rate": 8.400000000000001e-06,
"loss": 1.122,
"step": 42
},
{
"epoch": 0.49195566678584196,
"grad_norm": 9.0625,
"learning_rate": 8.6e-06,
"loss": 1.112,
"step": 43
},
{
"epoch": 0.5033964962459778,
"grad_norm": 7.75,
"learning_rate": 8.8e-06,
"loss": 1.1003,
"step": 44
},
{
"epoch": 0.5148373257061137,
"grad_norm": 27.375,
"learning_rate": 9e-06,
"loss": 1.1063,
"step": 45
},
{
"epoch": 0.5262781551662495,
"grad_norm": 1.8984375,
"learning_rate": 9.200000000000002e-06,
"loss": 1.0891,
"step": 46
},
{
"epoch": 0.5377189846263855,
"grad_norm": 2.21875,
"learning_rate": 9.4e-06,
"loss": 1.0974,
"step": 47
},
{
"epoch": 0.5491598140865213,
"grad_norm": 1.1875,
"learning_rate": 9.600000000000001e-06,
"loss": 1.0568,
"step": 48
},
{
"epoch": 0.5606006435466572,
"grad_norm": 1.78125,
"learning_rate": 9.800000000000001e-06,
"loss": 1.0272,
"step": 49
},
{
"epoch": 0.572041473006793,
"grad_norm": 4.6875,
"learning_rate": 1e-05,
"loss": 1.0282,
"step": 50
},
{
"epoch": 0.5834823024669289,
"grad_norm": 1.7890625,
"learning_rate": 9.998395376482152e-06,
"loss": 1.0128,
"step": 51
},
{
"epoch": 0.5949231319270647,
"grad_norm": 2.5625,
"learning_rate": 9.993582535855265e-06,
"loss": 1.0956,
"step": 52
},
{
"epoch": 0.6063639613872006,
"grad_norm": 3.59375,
"learning_rate": 9.985564567238237e-06,
"loss": 1.0804,
"step": 53
},
{
"epoch": 0.6178047908473364,
"grad_norm": 10.4375,
"learning_rate": 9.974346616959476e-06,
"loss": 1.1,
"step": 54
},
{
"epoch": 0.6292456203074723,
"grad_norm": 4.75,
"learning_rate": 9.959935885253715e-06,
"loss": 1.0653,
"step": 55
},
{
"epoch": 0.6406864497676081,
"grad_norm": 1.875,
"learning_rate": 9.942341621640558e-06,
"loss": 1.0784,
"step": 56
},
{
"epoch": 0.652127279227744,
"grad_norm": 3.265625,
"learning_rate": 9.921575118987672e-06,
"loss": 1.0597,
"step": 57
},
{
"epoch": 0.6635681086878799,
"grad_norm": 2.078125,
"learning_rate": 9.897649706262474e-06,
"loss": 1.0695,
"step": 58
},
{
"epoch": 0.6750089381480158,
"grad_norm": 6.65625,
"learning_rate": 9.870580739976936e-06,
"loss": 1.0896,
"step": 59
},
{
"epoch": 0.6864497676081516,
"grad_norm": 2.453125,
"learning_rate": 9.840385594331022e-06,
"loss": 1.0653,
"step": 60
},
{
"epoch": 0.6978905970682875,
"grad_norm": 2.859375,
"learning_rate": 9.807083650061063e-06,
"loss": 1.0338,
"step": 61
},
{
"epoch": 0.7093314265284233,
"grad_norm": 2.234375,
"learning_rate": 9.770696282000245e-06,
"loss": 1.0404,
"step": 62
},
{
"epoch": 0.7207722559885592,
"grad_norm": 1.53125,
"learning_rate": 9.731246845359187e-06,
"loss": 1.0658,
"step": 63
},
{
"epoch": 0.732213085448695,
"grad_norm": 3.15625,
"learning_rate": 9.688760660735403e-06,
"loss": 1.0667,
"step": 64
},
{
"epoch": 0.7436539149088309,
"grad_norm": 2.515625,
"learning_rate": 9.643264997861312e-06,
"loss": 0.9907,
"step": 65
},
{
"epoch": 0.7550947443689667,
"grad_norm": 2.3125,
"learning_rate": 9.594789058101154e-06,
"loss": 1.0233,
"step": 66
},
{
"epoch": 0.7665355738291026,
"grad_norm": 1.6328125,
"learning_rate": 9.543363955708124e-06,
"loss": 1.0892,
"step": 67
},
{
"epoch": 0.7779764032892384,
"grad_norm": 1.859375,
"learning_rate": 9.48902269785371e-06,
"loss": 1.0364,
"step": 68
},
{
"epoch": 0.7894172327493744,
"grad_norm": 1.6953125,
"learning_rate": 9.431800163442043e-06,
"loss": 0.9652,
"step": 69
},
{
"epoch": 0.8008580622095102,
"grad_norm": 36.0,
"learning_rate": 9.371733080722911e-06,
"loss": 1.0376,
"step": 70
},
{
"epoch": 0.8122988916696461,
"grad_norm": 3.890625,
"learning_rate": 9.308860003717748e-06,
"loss": 1.037,
"step": 71
},
{
"epoch": 0.8237397211297819,
"grad_norm": 2.640625,
"learning_rate": 9.243221287473755e-06,
"loss": 1.0471,
"step": 72
},
{
"epoch": 0.8351805505899178,
"grad_norm": 5.84375,
"learning_rate": 9.174859062162037e-06,
"loss": 1.0588,
"step": 73
},
{
"epoch": 0.8466213800500536,
"grad_norm": 1.3359375,
"learning_rate": 9.103817206036383e-06,
"loss": 1.0636,
"step": 74
},
{
"epoch": 0.8580622095101895,
"grad_norm": 39.25,
"learning_rate": 9.030141317270026e-06,
"loss": 1.0492,
"step": 75
},
{
"epoch": 0.8695030389703253,
"grad_norm": 35.0,
"learning_rate": 8.953878684688492e-06,
"loss": 1.1284,
"step": 76
},
{
"epoch": 0.8809438684304612,
"grad_norm": 101.0,
"learning_rate": 8.875078257417294e-06,
"loss": 1.0219,
"step": 77
},
{
"epoch": 0.892384697890597,
"grad_norm": 4.4375,
"learning_rate": 8.793790613463956e-06,
"loss": 1.1058,
"step": 78
},
{
"epoch": 0.903825527350733,
"grad_norm": 8.0625,
"learning_rate": 8.710067927254555e-06,
"loss": 1.0518,
"step": 79
},
{
"epoch": 0.9152663568108688,
"grad_norm": 8.875,
"learning_rate": 8.6239639361456e-06,
"loss": 1.0308,
"step": 80
},
{
"epoch": 0.9267071862710047,
"grad_norm": 1.6484375,
"learning_rate": 8.535533905932739e-06,
"loss": 1.0269,
"step": 81
},
{
"epoch": 0.9381480157311405,
"grad_norm": 2.21875,
"learning_rate": 8.444834595378434e-06,
"loss": 1.0251,
"step": 82
},
{
"epoch": 0.9495888451912764,
"grad_norm": 2.765625,
"learning_rate": 8.351924219781393e-06,
"loss": 1.0167,
"step": 83
},
{
"epoch": 0.9610296746514122,
"grad_norm": 1.3984375,
"learning_rate": 8.256862413611113e-06,
"loss": 1.0286,
"step": 84
},
{
"epoch": 0.9724705041115481,
"grad_norm": 1.6796875,
"learning_rate": 8.15971019223152e-06,
"loss": 1.0564,
"step": 85
},
{
"epoch": 0.9839113335716839,
"grad_norm": 5.3125,
"learning_rate": 8.060529912738316e-06,
"loss": 1.027,
"step": 86
},
{
"epoch": 0.9953521630318198,
"grad_norm": 1.4296875,
"learning_rate": 7.959385233935087e-06,
"loss": 1.0048,
"step": 87
},
{
"epoch": 1.0114408294601358,
"grad_norm": 2.109375,
"learning_rate": 7.856341075473963e-06,
"loss": 1.4279,
"step": 88
},
{
"epoch": 1.0228816589202716,
"grad_norm": 1.375,
"learning_rate": 7.751463576186957e-06,
"loss": 1.0041,
"step": 89
},
{
"epoch": 1.0343224883804076,
"grad_norm": 2.578125,
"learning_rate": 7.644820051634813e-06,
"loss": 1.0455,
"step": 90
},
{
"epoch": 1.0457633178405434,
"grad_norm": 2.90625,
"learning_rate": 7.536478950900537e-06,
"loss": 1.0663,
"step": 91
},
{
"epoch": 1.0572041473006792,
"grad_norm": 5.6875,
"learning_rate": 7.4265098126554065e-06,
"loss": 1.0023,
"step": 92
},
{
"epoch": 1.0686449767608153,
"grad_norm": 1.2265625,
"learning_rate": 7.314983220525604e-06,
"loss": 1.0352,
"step": 93
},
{
"epoch": 1.080085806220951,
"grad_norm": 3.28125,
"learning_rate": 7.201970757788172e-06,
"loss": 1.1086,
"step": 94
},
{
"epoch": 1.0915266356810869,
"grad_norm": 2.8125,
"learning_rate": 7.087544961425317e-06,
"loss": 1.0692,
"step": 95
},
{
"epoch": 1.1029674651412227,
"grad_norm": 4.40625,
"learning_rate": 6.971779275566593e-06,
"loss": 1.0666,
"step": 96
},
{
"epoch": 1.1144082946013585,
"grad_norm": 4.21875,
"learning_rate": 6.85474800434884e-06,
"loss": 1.0401,
"step": 97
},
{
"epoch": 1.1258491240614945,
"grad_norm": 1.671875,
"learning_rate": 6.736526264224101e-06,
"loss": 1.062,
"step": 98
},
{
"epoch": 1.1372899535216303,
"grad_norm": 2.90625,
"learning_rate": 6.617189935746191e-06,
"loss": 1.1024,
"step": 99
},
{
"epoch": 1.1487307829817661,
"grad_norm": 4.46875,
"learning_rate": 6.496815614866792e-06,
"loss": 1.1216,
"step": 100
},
{
"epoch": 1.1601716124419021,
"grad_norm": 1.9453125,
"learning_rate": 6.375480563772391e-06,
"loss": 1.0347,
"step": 101
},
{
"epoch": 1.171612441902038,
"grad_norm": 2.015625,
"learning_rate": 6.2532626612936035e-06,
"loss": 1.1006,
"step": 102
},
{
"epoch": 1.1830532713621738,
"grad_norm": 1.8671875,
"learning_rate": 6.130240352918675e-06,
"loss": 1.0257,
"step": 103
},
{
"epoch": 1.1944941008223096,
"grad_norm": 3.53125,
"learning_rate": 6.006492600443301e-06,
"loss": 1.091,
"step": 104
},
{
"epoch": 1.2059349302824454,
"grad_norm": 2.890625,
"learning_rate": 5.882098831289044e-06,
"loss": 1.0359,
"step": 105
},
{
"epoch": 1.2173757597425814,
"grad_norm": 3.421875,
"learning_rate": 5.757138887522884e-06,
"loss": 1.1039,
"step": 106
},
{
"epoch": 1.2288165892027172,
"grad_norm": 12.4375,
"learning_rate": 5.631692974610647e-06,
"loss": 1.1496,
"step": 107
},
{
"epoch": 1.240257418662853,
"grad_norm": 16.25,
"learning_rate": 5.505841609937162e-06,
"loss": 1.2067,
"step": 108
},
{
"epoch": 1.251698248122989,
"grad_norm": 8.1875,
"learning_rate": 5.379665571126232e-06,
"loss": 1.2643,
"step": 109
},
{
"epoch": 1.2631390775831248,
"grad_norm": 6.03125,
"learning_rate": 5.253245844193564e-06,
"loss": 1.0771,
"step": 110
},
{
"epoch": 1.2745799070432606,
"grad_norm": 3.03125,
"learning_rate": 5.12666357156594e-06,
"loss": 1.0801,
"step": 111
},
{
"epoch": 1.2860207365033964,
"grad_norm": 4.0625,
"learning_rate": 5e-06,
"loss": 1.125,
"step": 112
},
{
"epoch": 1.2974615659635322,
"grad_norm": 2.875,
"learning_rate": 4.873336428434062e-06,
"loss": 1.0785,
"step": 113
},
{
"epoch": 1.3089023954236683,
"grad_norm": 4.65625,
"learning_rate": 4.746754155806437e-06,
"loss": 1.1186,
"step": 114
},
{
"epoch": 1.320343224883804,
"grad_norm": 1.671875,
"learning_rate": 4.62033442887377e-06,
"loss": 1.0442,
"step": 115
},
{
"epoch": 1.3317840543439399,
"grad_norm": 13.125,
"learning_rate": 4.49415839006284e-06,
"loss": 0.992,
"step": 116
},
{
"epoch": 1.343224883804076,
"grad_norm": 2.515625,
"learning_rate": 4.368307025389355e-06,
"loss": 1.0311,
"step": 117
},
{
"epoch": 1.3546657132642117,
"grad_norm": 2.234375,
"learning_rate": 4.2428611124771184e-06,
"loss": 1.1222,
"step": 118
},
{
"epoch": 1.3661065427243475,
"grad_norm": 3.859375,
"learning_rate": 4.11790116871096e-06,
"loss": 1.0476,
"step": 119
},
{
"epoch": 1.3775473721844833,
"grad_norm": 2.875,
"learning_rate": 3.993507399556699e-06,
"loss": 1.1549,
"step": 120
},
{
"epoch": 1.3889882016446191,
"grad_norm": 3.203125,
"learning_rate": 3.869759647081326e-06,
"loss": 1.0162,
"step": 121
},
{
"epoch": 1.4004290311047551,
"grad_norm": 2.515625,
"learning_rate": 3.7467373387063973e-06,
"loss": 1.0692,
"step": 122
},
{
"epoch": 1.411869860564891,
"grad_norm": 9.5,
"learning_rate": 3.62451943622761e-06,
"loss": 1.1584,
"step": 123
},
{
"epoch": 1.4233106900250267,
"grad_norm": 2.0625,
"learning_rate": 3.5031843851332105e-06,
"loss": 1.1119,
"step": 124
},
{
"epoch": 1.4347515194851628,
"grad_norm": 3.453125,
"learning_rate": 3.3828100642538097e-06,
"loss": 1.0913,
"step": 125
},
{
"epoch": 1.4461923489452986,
"grad_norm": 2.046875,
"learning_rate": 3.2634737357758994e-06,
"loss": 1.0886,
"step": 126
},
{
"epoch": 1.4576331784054344,
"grad_norm": 2.453125,
"learning_rate": 3.145251995651162e-06,
"loss": 1.064,
"step": 127
},
{
"epoch": 1.4690740078655702,
"grad_norm": 1.1953125,
"learning_rate": 3.0282207244334084e-06,
"loss": 1.029,
"step": 128
},
{
"epoch": 1.480514837325706,
"grad_norm": 3.21875,
"learning_rate": 2.912455038574686e-06,
"loss": 1.0867,
"step": 129
},
{
"epoch": 1.491955666785842,
"grad_norm": 2.828125,
"learning_rate": 2.7980292422118282e-06,
"loss": 1.1117,
"step": 130
},
{
"epoch": 1.5033964962459778,
"grad_norm": 6.21875,
"learning_rate": 2.6850167794743966e-06,
"loss": 1.0614,
"step": 131
},
{
"epoch": 1.5148373257061136,
"grad_norm": 1.0078125,
"learning_rate": 2.573490187344596e-06,
"loss": 1.016,
"step": 132
},
{
"epoch": 1.5262781551662497,
"grad_norm": 5.65625,
"learning_rate": 2.4635210490994648e-06,
"loss": 1.0849,
"step": 133
},
{
"epoch": 1.5377189846263855,
"grad_norm": 2.5625,
"learning_rate": 2.3551799483651894e-06,
"loss": 1.0158,
"step": 134
},
{
"epoch": 1.5491598140865213,
"grad_norm": 0.9453125,
"learning_rate": 2.2485364238130435e-06,
"loss": 1.0059,
"step": 135
},
{
"epoch": 1.5606006435466573,
"grad_norm": 1.71875,
"learning_rate": 2.1436589245260375e-06,
"loss": 1.0081,
"step": 136
},
{
"epoch": 1.5720414730067929,
"grad_norm": 2.09375,
"learning_rate": 2.040614766064913e-06,
"loss": 1.0354,
"step": 137
},
{
"epoch": 1.583482302466929,
"grad_norm": 2.875,
"learning_rate": 1.9394700872616856e-06,
"loss": 1.0662,
"step": 138
},
{
"epoch": 1.5949231319270647,
"grad_norm": 2.390625,
"learning_rate": 1.8402898077684806e-06,
"loss": 1.1284,
"step": 139
},
{
"epoch": 1.6063639613872005,
"grad_norm": 1.9609375,
"learning_rate": 1.74313758638889e-06,
"loss": 1.0253,
"step": 140
},
{
"epoch": 1.6178047908473365,
"grad_norm": 2.328125,
"learning_rate": 1.648075780218607e-06,
"loss": 1.1114,
"step": 141
},
{
"epoch": 1.6292456203074723,
"grad_norm": 1.3359375,
"learning_rate": 1.555165404621567e-06,
"loss": 1.0888,
"step": 142
},
{
"epoch": 1.6406864497676081,
"grad_norm": 2.4375,
"learning_rate": 1.4644660940672628e-06,
"loss": 1.0505,
"step": 143
},
{
"epoch": 1.6521272792277442,
"grad_norm": 6.0,
"learning_rate": 1.3760360638544012e-06,
"loss": 1.2607,
"step": 144
},
{
"epoch": 1.6635681086878797,
"grad_norm": 4.1875,
"learning_rate": 1.2899320727454472e-06,
"loss": 1.1066,
"step": 145
},
{
"epoch": 1.6750089381480158,
"grad_norm": 2.71875,
"learning_rate": 1.2062093865360458e-06,
"loss": 1.0927,
"step": 146
},
{
"epoch": 1.6864497676081516,
"grad_norm": 5.1875,
"learning_rate": 1.1249217425827063e-06,
"loss": 1.0638,
"step": 147
},
{
"epoch": 1.6978905970682874,
"grad_norm": 2.125,
"learning_rate": 1.046121315311508e-06,
"loss": 1.0502,
"step": 148
},
{
"epoch": 1.7093314265284234,
"grad_norm": 4.34375,
"learning_rate": 9.69858682729976e-07,
"loss": 1.0722,
"step": 149
},
{
"epoch": 1.7207722559885592,
"grad_norm": 3.84375,
"learning_rate": 8.961827939636198e-07,
"loss": 1.1203,
"step": 150
},
{
"epoch": 1.732213085448695,
"grad_norm": 4.09375,
"learning_rate": 8.251409378379638e-07,
"loss": 1.1094,
"step": 151
},
{
"epoch": 1.743653914908831,
"grad_norm": 2.859375,
"learning_rate": 7.567787125262449e-07,
"loss": 1.0351,
"step": 152
},
{
"epoch": 1.7550947443689666,
"grad_norm": 16.125,
"learning_rate": 6.911399962822518e-07,
"loss": 1.2069,
"step": 153
},
{
"epoch": 1.7665355738291026,
"grad_norm": 1.296875,
"learning_rate": 6.282669192770896e-07,
"loss": 1.0696,
"step": 154
},
{
"epoch": 1.7779764032892384,
"grad_norm": 3.140625,
"learning_rate": 5.681998365579594e-07,
"loss": 1.117,
"step": 155
},
{
"epoch": 1.7894172327493743,
"grad_norm": 1.609375,
"learning_rate": 5.109773021462921e-07,
"loss": 1.0426,
"step": 156
},
{
"epoch": 1.8008580622095103,
"grad_norm": 1.171875,
"learning_rate": 4.5663604429187547e-07,
"loss": 1.0841,
"step": 157
},
{
"epoch": 1.812298891669646,
"grad_norm": 1.8984375,
"learning_rate": 4.05210941898847e-07,
"loss": 0.9866,
"step": 158
},
{
"epoch": 1.8237397211297819,
"grad_norm": 3.25,
"learning_rate": 3.567350021386895e-07,
"loss": 0.9696,
"step": 159
},
{
"epoch": 1.835180550589918,
"grad_norm": 1.8984375,
"learning_rate": 3.112393392645985e-07,
"loss": 1.0818,
"step": 160
},
{
"epoch": 1.8466213800500535,
"grad_norm": 2.71875,
"learning_rate": 2.6875315464081566e-07,
"loss": 1.102,
"step": 161
},
{
"epoch": 1.8580622095101895,
"grad_norm": 5.625,
"learning_rate": 2.2930371799975593e-07,
"loss": 1.0949,
"step": 162
},
{
"epoch": 1.8695030389703253,
"grad_norm": 2.671875,
"learning_rate": 1.9291634993893803e-07,
"loss": 1.0649,
"step": 163
},
{
"epoch": 1.8809438684304611,
"grad_norm": 6.375,
"learning_rate": 1.5961440566897913e-07,
"loss": 1.1652,
"step": 164
},
{
"epoch": 1.8923846978905972,
"grad_norm": 1.515625,
"learning_rate": 1.2941926002306536e-07,
"loss": 1.088,
"step": 165
},
{
"epoch": 1.903825527350733,
"grad_norm": 2.875,
"learning_rate": 1.0235029373752758e-07,
"loss": 1.091,
"step": 166
},
{
"epoch": 1.9152663568108688,
"grad_norm": 4.21875,
"learning_rate": 7.842488101232893e-08,
"loss": 1.0127,
"step": 167
},
{
"epoch": 1.9267071862710048,
"grad_norm": 1.5,
"learning_rate": 5.7658378359443104e-08,
"loss": 1.0575,
"step": 168
},
{
"epoch": 1.9381480157311404,
"grad_norm": 2.265625,
"learning_rate": 4.006411474628491e-08,
"loss": 1.0529,
"step": 169
},
{
"epoch": 1.9495888451912764,
"grad_norm": 3.4375,
"learning_rate": 2.5653383040524228e-08,
"loss": 1.1027,
"step": 170
},
{
"epoch": 1.9610296746514122,
"grad_norm": 4.96875,
"learning_rate": 1.4435432761762958e-08,
"loss": 1.1598,
"step": 171
},
{
"epoch": 1.972470504111548,
"grad_norm": 2.5,
"learning_rate": 6.417464144736208e-09,
"loss": 1.001,
"step": 172
},
{
"epoch": 1.983911333571684,
"grad_norm": 5.875,
"learning_rate": 1.6046235178474034e-09,
"loss": 1.161,
"step": 173
},
{
"epoch": 1.9953521630318198,
"grad_norm": 1.4296875,
"learning_rate": 0.0,
"loss": 1.0751,
"step": 174
}
],
"logging_steps": 1,
"max_steps": 174,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 44,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.2837861025752023e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}