openmathreasoning_100k / trainer_state.json
sedrickkeh's picture
Upload trainer_state.json with huggingface_hub
4d9783b verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 470,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.010638297872340425,
"grad_norm": 5.62164644869217,
"learning_rate": 1.7021276595744682e-06,
"loss": 0.8559,
"step": 1
},
{
"epoch": 0.02127659574468085,
"grad_norm": 5.541790955372414,
"learning_rate": 3.4042553191489363e-06,
"loss": 0.8537,
"step": 2
},
{
"epoch": 0.031914893617021274,
"grad_norm": 5.407901882681235,
"learning_rate": 5.106382978723404e-06,
"loss": 0.8414,
"step": 3
},
{
"epoch": 0.0425531914893617,
"grad_norm": 4.03578755597478,
"learning_rate": 6.808510638297873e-06,
"loss": 0.8026,
"step": 4
},
{
"epoch": 0.05319148936170213,
"grad_norm": 2.235595604545287,
"learning_rate": 8.510638297872341e-06,
"loss": 0.7586,
"step": 5
},
{
"epoch": 0.06382978723404255,
"grad_norm": 3.6713198410625205,
"learning_rate": 1.0212765957446808e-05,
"loss": 0.7542,
"step": 6
},
{
"epoch": 0.07446808510638298,
"grad_norm": 3.8053717898299015,
"learning_rate": 1.1914893617021277e-05,
"loss": 0.7525,
"step": 7
},
{
"epoch": 0.0851063829787234,
"grad_norm": 3.735929746217291,
"learning_rate": 1.3617021276595745e-05,
"loss": 0.7146,
"step": 8
},
{
"epoch": 0.09574468085106383,
"grad_norm": 3.4388722472257522,
"learning_rate": 1.5319148936170214e-05,
"loss": 0.7076,
"step": 9
},
{
"epoch": 0.10638297872340426,
"grad_norm": 2.0576402741081745,
"learning_rate": 1.7021276595744682e-05,
"loss": 0.6804,
"step": 10
},
{
"epoch": 0.11702127659574468,
"grad_norm": 2.194245295451263,
"learning_rate": 1.872340425531915e-05,
"loss": 0.6574,
"step": 11
},
{
"epoch": 0.1276595744680851,
"grad_norm": 1.970180537002589,
"learning_rate": 2.0425531914893616e-05,
"loss": 0.6351,
"step": 12
},
{
"epoch": 0.13829787234042554,
"grad_norm": 1.1822804981997306,
"learning_rate": 2.2127659574468088e-05,
"loss": 0.6205,
"step": 13
},
{
"epoch": 0.14893617021276595,
"grad_norm": 1.1602034908038812,
"learning_rate": 2.3829787234042553e-05,
"loss": 0.6102,
"step": 14
},
{
"epoch": 0.1595744680851064,
"grad_norm": 1.6168637308896043,
"learning_rate": 2.5531914893617025e-05,
"loss": 0.6058,
"step": 15
},
{
"epoch": 0.1702127659574468,
"grad_norm": 0.9268293511359277,
"learning_rate": 2.723404255319149e-05,
"loss": 0.5865,
"step": 16
},
{
"epoch": 0.18085106382978725,
"grad_norm": 1.3995612819702326,
"learning_rate": 2.8936170212765963e-05,
"loss": 0.5828,
"step": 17
},
{
"epoch": 0.19148936170212766,
"grad_norm": 0.9491984900355026,
"learning_rate": 3.063829787234043e-05,
"loss": 0.5814,
"step": 18
},
{
"epoch": 0.20212765957446807,
"grad_norm": 1.2461341580332967,
"learning_rate": 3.234042553191489e-05,
"loss": 0.5732,
"step": 19
},
{
"epoch": 0.2127659574468085,
"grad_norm": 0.9013655593666975,
"learning_rate": 3.4042553191489365e-05,
"loss": 0.572,
"step": 20
},
{
"epoch": 0.22340425531914893,
"grad_norm": 1.4385376985061347,
"learning_rate": 3.574468085106383e-05,
"loss": 0.5775,
"step": 21
},
{
"epoch": 0.23404255319148937,
"grad_norm": 1.1316548066081353,
"learning_rate": 3.74468085106383e-05,
"loss": 0.5589,
"step": 22
},
{
"epoch": 0.24468085106382978,
"grad_norm": 1.1348968485093565,
"learning_rate": 3.914893617021277e-05,
"loss": 0.5555,
"step": 23
},
{
"epoch": 0.2553191489361702,
"grad_norm": 0.9593256794712123,
"learning_rate": 4.085106382978723e-05,
"loss": 0.5577,
"step": 24
},
{
"epoch": 0.26595744680851063,
"grad_norm": 1.5810352206230227,
"learning_rate": 4.2553191489361704e-05,
"loss": 0.5506,
"step": 25
},
{
"epoch": 0.2765957446808511,
"grad_norm": 1.0552617330606424,
"learning_rate": 4.4255319148936176e-05,
"loss": 0.5373,
"step": 26
},
{
"epoch": 0.2872340425531915,
"grad_norm": 1.651827285414071,
"learning_rate": 4.595744680851065e-05,
"loss": 0.5465,
"step": 27
},
{
"epoch": 0.2978723404255319,
"grad_norm": 1.8485214296335475,
"learning_rate": 4.765957446808511e-05,
"loss": 0.5413,
"step": 28
},
{
"epoch": 0.30851063829787234,
"grad_norm": 0.8693252455438564,
"learning_rate": 4.936170212765958e-05,
"loss": 0.5416,
"step": 29
},
{
"epoch": 0.3191489361702128,
"grad_norm": 2.3515640310759522,
"learning_rate": 5.106382978723405e-05,
"loss": 0.547,
"step": 30
},
{
"epoch": 0.32978723404255317,
"grad_norm": 1.4756724644968964,
"learning_rate": 5.276595744680851e-05,
"loss": 0.5442,
"step": 31
},
{
"epoch": 0.3404255319148936,
"grad_norm": 1.9541127476117117,
"learning_rate": 5.446808510638298e-05,
"loss": 0.5376,
"step": 32
},
{
"epoch": 0.35106382978723405,
"grad_norm": 1.4385140916597212,
"learning_rate": 5.617021276595745e-05,
"loss": 0.5347,
"step": 33
},
{
"epoch": 0.3617021276595745,
"grad_norm": 1.6177174100025706,
"learning_rate": 5.7872340425531925e-05,
"loss": 0.5283,
"step": 34
},
{
"epoch": 0.3723404255319149,
"grad_norm": 1.5517064428797538,
"learning_rate": 5.9574468085106384e-05,
"loss": 0.5304,
"step": 35
},
{
"epoch": 0.3829787234042553,
"grad_norm": 1.3062103414138244,
"learning_rate": 6.127659574468086e-05,
"loss": 0.5277,
"step": 36
},
{
"epoch": 0.39361702127659576,
"grad_norm": 1.1717376687944345,
"learning_rate": 6.297872340425533e-05,
"loss": 0.5246,
"step": 37
},
{
"epoch": 0.40425531914893614,
"grad_norm": 1.5784372654316825,
"learning_rate": 6.468085106382979e-05,
"loss": 0.5211,
"step": 38
},
{
"epoch": 0.4148936170212766,
"grad_norm": 1.2570855650815358,
"learning_rate": 6.638297872340426e-05,
"loss": 0.531,
"step": 39
},
{
"epoch": 0.425531914893617,
"grad_norm": 1.1994392890244334,
"learning_rate": 6.808510638297873e-05,
"loss": 0.5098,
"step": 40
},
{
"epoch": 0.43617021276595747,
"grad_norm": 1.608445496858305,
"learning_rate": 6.97872340425532e-05,
"loss": 0.5176,
"step": 41
},
{
"epoch": 0.44680851063829785,
"grad_norm": 1.6126071175066283,
"learning_rate": 7.148936170212766e-05,
"loss": 0.5181,
"step": 42
},
{
"epoch": 0.4574468085106383,
"grad_norm": 1.3431412401241174,
"learning_rate": 7.319148936170213e-05,
"loss": 0.5135,
"step": 43
},
{
"epoch": 0.46808510638297873,
"grad_norm": 1.5807072225527692,
"learning_rate": 7.48936170212766e-05,
"loss": 0.5185,
"step": 44
},
{
"epoch": 0.4787234042553192,
"grad_norm": 2.0147344702502554,
"learning_rate": 7.659574468085108e-05,
"loss": 0.5281,
"step": 45
},
{
"epoch": 0.48936170212765956,
"grad_norm": 1.4169773369597476,
"learning_rate": 7.829787234042553e-05,
"loss": 0.5104,
"step": 46
},
{
"epoch": 0.5,
"grad_norm": 1.327014601864401,
"learning_rate": 8e-05,
"loss": 0.507,
"step": 47
},
{
"epoch": 0.5106382978723404,
"grad_norm": 1.463329858386686,
"learning_rate": 7.999889681839899e-05,
"loss": 0.513,
"step": 48
},
{
"epoch": 0.5212765957446809,
"grad_norm": 1.7226533644834205,
"learning_rate": 7.999558733444641e-05,
"loss": 0.5241,
"step": 49
},
{
"epoch": 0.5319148936170213,
"grad_norm": 1.4625043342314172,
"learning_rate": 7.999007173069037e-05,
"loss": 0.4992,
"step": 50
},
{
"epoch": 0.5425531914893617,
"grad_norm": 1.74638144425408,
"learning_rate": 7.998235031136648e-05,
"loss": 0.5129,
"step": 51
},
{
"epoch": 0.5531914893617021,
"grad_norm": 1.156051684687154,
"learning_rate": 7.997242350238117e-05,
"loss": 0.5017,
"step": 52
},
{
"epoch": 0.5638297872340425,
"grad_norm": 1.1484647512074981,
"learning_rate": 7.996029185128804e-05,
"loss": 0.5021,
"step": 53
},
{
"epoch": 0.574468085106383,
"grad_norm": 1.7889194838438403,
"learning_rate": 7.994595602725781e-05,
"loss": 0.5137,
"step": 54
},
{
"epoch": 0.5851063829787234,
"grad_norm": 1.2512024906203,
"learning_rate": 7.992941682104139e-05,
"loss": 0.4935,
"step": 55
},
{
"epoch": 0.5957446808510638,
"grad_norm": 1.9138223030633696,
"learning_rate": 7.991067514492614e-05,
"loss": 0.5084,
"step": 56
},
{
"epoch": 0.6063829787234043,
"grad_norm": 1.2776429355884351,
"learning_rate": 7.988973203268567e-05,
"loss": 0.5045,
"step": 57
},
{
"epoch": 0.6170212765957447,
"grad_norm": 2.141401973332067,
"learning_rate": 7.986658863952281e-05,
"loss": 0.4992,
"step": 58
},
{
"epoch": 0.6276595744680851,
"grad_norm": 1.454535688794727,
"learning_rate": 7.984124624200583e-05,
"loss": 0.5058,
"step": 59
},
{
"epoch": 0.6382978723404256,
"grad_norm": 1.7846076730422646,
"learning_rate": 7.981370623799803e-05,
"loss": 0.5013,
"step": 60
},
{
"epoch": 0.648936170212766,
"grad_norm": 1.4310227485726414,
"learning_rate": 7.978397014658075e-05,
"loss": 0.5079,
"step": 61
},
{
"epoch": 0.6595744680851063,
"grad_norm": 1.4094274678104701,
"learning_rate": 7.97520396079694e-05,
"loss": 0.4882,
"step": 62
},
{
"epoch": 0.6702127659574468,
"grad_norm": 1.6288107207366431,
"learning_rate": 7.971791638342313e-05,
"loss": 0.4977,
"step": 63
},
{
"epoch": 0.6808510638297872,
"grad_norm": 1.0461330955205717,
"learning_rate": 7.96816023551476e-05,
"loss": 0.4932,
"step": 64
},
{
"epoch": 0.6914893617021277,
"grad_norm": 1.8472118603297956,
"learning_rate": 7.964309952619121e-05,
"loss": 0.4956,
"step": 65
},
{
"epoch": 0.7021276595744681,
"grad_norm": 1.3785935023957434,
"learning_rate": 7.96024100203346e-05,
"loss": 0.5072,
"step": 66
},
{
"epoch": 0.7127659574468085,
"grad_norm": 1.321199071263235,
"learning_rate": 7.955953608197345e-05,
"loss": 0.4864,
"step": 67
},
{
"epoch": 0.723404255319149,
"grad_norm": 1.2126321080139488,
"learning_rate": 7.951448007599478e-05,
"loss": 0.5016,
"step": 68
},
{
"epoch": 0.7340425531914894,
"grad_norm": 1.1858808993782288,
"learning_rate": 7.946724448764644e-05,
"loss": 0.4943,
"step": 69
},
{
"epoch": 0.7446808510638298,
"grad_norm": 1.0062356288958618,
"learning_rate": 7.94178319224e-05,
"loss": 0.491,
"step": 70
},
{
"epoch": 0.7553191489361702,
"grad_norm": 1.9547579891672486,
"learning_rate": 7.936624510580712e-05,
"loss": 0.4879,
"step": 71
},
{
"epoch": 0.7659574468085106,
"grad_norm": 1.0537143803568765,
"learning_rate": 7.931248688334915e-05,
"loss": 0.4916,
"step": 72
},
{
"epoch": 0.776595744680851,
"grad_norm": 1.6913458685278957,
"learning_rate": 7.925656022028017e-05,
"loss": 0.4983,
"step": 73
},
{
"epoch": 0.7872340425531915,
"grad_norm": 1.016587494594598,
"learning_rate": 7.919846820146348e-05,
"loss": 0.4964,
"step": 74
},
{
"epoch": 0.7978723404255319,
"grad_norm": 1.6376648762774801,
"learning_rate": 7.913821403120139e-05,
"loss": 0.4916,
"step": 75
},
{
"epoch": 0.8085106382978723,
"grad_norm": 1.2262934120029891,
"learning_rate": 7.90758010330585e-05,
"loss": 0.4851,
"step": 76
},
{
"epoch": 0.8191489361702128,
"grad_norm": 1.381474776104582,
"learning_rate": 7.901123264967836e-05,
"loss": 0.4929,
"step": 77
},
{
"epoch": 0.8297872340425532,
"grad_norm": 1.3905933735533444,
"learning_rate": 7.894451244259363e-05,
"loss": 0.4924,
"step": 78
},
{
"epoch": 0.8404255319148937,
"grad_norm": 1.038270640845919,
"learning_rate": 7.887564409202953e-05,
"loss": 0.4893,
"step": 79
},
{
"epoch": 0.851063829787234,
"grad_norm": 1.4032705044700307,
"learning_rate": 7.880463139670091e-05,
"loss": 0.485,
"step": 80
},
{
"epoch": 0.8617021276595744,
"grad_norm": 1.1420897464578796,
"learning_rate": 7.873147827360273e-05,
"loss": 0.4867,
"step": 81
},
{
"epoch": 0.8723404255319149,
"grad_norm": 1.667140313423623,
"learning_rate": 7.865618875779398e-05,
"loss": 0.4775,
"step": 82
},
{
"epoch": 0.8829787234042553,
"grad_norm": 1.276151508655107,
"learning_rate": 7.857876700217508e-05,
"loss": 0.4786,
"step": 83
},
{
"epoch": 0.8936170212765957,
"grad_norm": 1.1090496719024276,
"learning_rate": 7.849921727725882e-05,
"loss": 0.4722,
"step": 84
},
{
"epoch": 0.9042553191489362,
"grad_norm": 0.991963411082399,
"learning_rate": 7.841754397093487e-05,
"loss": 0.478,
"step": 85
},
{
"epoch": 0.9148936170212766,
"grad_norm": 1.731621852488109,
"learning_rate": 7.833375158822766e-05,
"loss": 0.4773,
"step": 86
},
{
"epoch": 0.925531914893617,
"grad_norm": 1.3804671267005582,
"learning_rate": 7.824784475104795e-05,
"loss": 0.4787,
"step": 87
},
{
"epoch": 0.9361702127659575,
"grad_norm": 1.1135118775414654,
"learning_rate": 7.815982819793784e-05,
"loss": 0.4772,
"step": 88
},
{
"epoch": 0.9468085106382979,
"grad_norm": 1.9348194075183587,
"learning_rate": 7.806970678380943e-05,
"loss": 0.4768,
"step": 89
},
{
"epoch": 0.9574468085106383,
"grad_norm": 0.9278131511467589,
"learning_rate": 7.797748547967701e-05,
"loss": 0.4768,
"step": 90
},
{
"epoch": 0.9680851063829787,
"grad_norm": 1.5790755873743727,
"learning_rate": 7.788316937238287e-05,
"loss": 0.4777,
"step": 91
},
{
"epoch": 0.9787234042553191,
"grad_norm": 1.1589932356056467,
"learning_rate": 7.778676366431676e-05,
"loss": 0.4793,
"step": 92
},
{
"epoch": 0.9893617021276596,
"grad_norm": 0.9998066780828351,
"learning_rate": 7.768827367312882e-05,
"loss": 0.4736,
"step": 93
},
{
"epoch": 1.0,
"grad_norm": 0.9964032491428947,
"learning_rate": 7.758770483143634e-05,
"loss": 0.4841,
"step": 94
},
{
"epoch": 1.0106382978723405,
"grad_norm": 1.1948858983085593,
"learning_rate": 7.748506268652415e-05,
"loss": 0.461,
"step": 95
},
{
"epoch": 1.0212765957446808,
"grad_norm": 0.7073134184929624,
"learning_rate": 7.73803529000385e-05,
"loss": 0.4536,
"step": 96
},
{
"epoch": 1.0319148936170213,
"grad_norm": 1.1611831111477335,
"learning_rate": 7.727358124767491e-05,
"loss": 0.4612,
"step": 97
},
{
"epoch": 1.0425531914893618,
"grad_norm": 1.0669064153598073,
"learning_rate": 7.71647536188595e-05,
"loss": 0.4653,
"step": 98
},
{
"epoch": 1.053191489361702,
"grad_norm": 1.024613831141331,
"learning_rate": 7.705387601642416e-05,
"loss": 0.4584,
"step": 99
},
{
"epoch": 1.0638297872340425,
"grad_norm": 1.4884061518621656,
"learning_rate": 7.694095455627542e-05,
"loss": 0.4608,
"step": 100
},
{
"epoch": 1.074468085106383,
"grad_norm": 1.0414967862002833,
"learning_rate": 7.682599546705716e-05,
"loss": 0.4577,
"step": 101
},
{
"epoch": 1.0851063829787233,
"grad_norm": 1.3598647542639628,
"learning_rate": 7.670900508980697e-05,
"loss": 0.4587,
"step": 102
},
{
"epoch": 1.0957446808510638,
"grad_norm": 1.5748967124421995,
"learning_rate": 7.658998987760645e-05,
"loss": 0.4628,
"step": 103
},
{
"epoch": 1.1063829787234043,
"grad_norm": 1.142076727768023,
"learning_rate": 7.646895639522518e-05,
"loss": 0.4551,
"step": 104
},
{
"epoch": 1.1170212765957448,
"grad_norm": 1.3811816181823466,
"learning_rate": 7.634591131875875e-05,
"loss": 0.4594,
"step": 105
},
{
"epoch": 1.127659574468085,
"grad_norm": 0.9160727711980033,
"learning_rate": 7.622086143526036e-05,
"loss": 0.4518,
"step": 106
},
{
"epoch": 1.1382978723404256,
"grad_norm": 1.0917703855774061,
"learning_rate": 7.609381364236655e-05,
"loss": 0.4551,
"step": 107
},
{
"epoch": 1.148936170212766,
"grad_norm": 1.43226233951233,
"learning_rate": 7.59647749479167e-05,
"loss": 0.4603,
"step": 108
},
{
"epoch": 1.1595744680851063,
"grad_norm": 1.4134305371766467,
"learning_rate": 7.583375246956648e-05,
"loss": 0.4641,
"step": 109
},
{
"epoch": 1.1702127659574468,
"grad_norm": 0.7629510146350382,
"learning_rate": 7.570075343439526e-05,
"loss": 0.4559,
"step": 110
},
{
"epoch": 1.1808510638297873,
"grad_norm": 1.3551761060537464,
"learning_rate": 7.556578517850747e-05,
"loss": 0.4562,
"step": 111
},
{
"epoch": 1.1914893617021276,
"grad_norm": 1.1143038754484502,
"learning_rate": 7.542885514662794e-05,
"loss": 0.454,
"step": 112
},
{
"epoch": 1.202127659574468,
"grad_norm": 1.3365732739074656,
"learning_rate": 7.528997089169128e-05,
"loss": 0.4534,
"step": 113
},
{
"epoch": 1.2127659574468086,
"grad_norm": 0.9865960113031025,
"learning_rate": 7.51491400744252e-05,
"loss": 0.4522,
"step": 114
},
{
"epoch": 1.2234042553191489,
"grad_norm": 1.7137045691014816,
"learning_rate": 7.500637046292803e-05,
"loss": 0.4609,
"step": 115
},
{
"epoch": 1.2340425531914894,
"grad_norm": 1.0423122076919606,
"learning_rate": 7.48616699322402e-05,
"loss": 0.4582,
"step": 116
},
{
"epoch": 1.2446808510638299,
"grad_norm": 1.1897509411491205,
"learning_rate": 7.471504646390987e-05,
"loss": 0.4552,
"step": 117
},
{
"epoch": 1.2553191489361701,
"grad_norm": 1.707159683909229,
"learning_rate": 7.456650814555267e-05,
"loss": 0.4668,
"step": 118
},
{
"epoch": 1.2659574468085106,
"grad_norm": 0.7299856606632058,
"learning_rate": 7.441606317040558e-05,
"loss": 0.4523,
"step": 119
},
{
"epoch": 1.2765957446808511,
"grad_norm": 1.7335201629368997,
"learning_rate": 7.426371983687503e-05,
"loss": 0.4645,
"step": 120
},
{
"epoch": 1.2872340425531914,
"grad_norm": 1.0376836462353736,
"learning_rate": 7.410948654807916e-05,
"loss": 0.4567,
"step": 121
},
{
"epoch": 1.297872340425532,
"grad_norm": 1.1899631811065923,
"learning_rate": 7.39533718113843e-05,
"loss": 0.4622,
"step": 122
},
{
"epoch": 1.3085106382978724,
"grad_norm": 1.2350801182171,
"learning_rate": 7.379538423793568e-05,
"loss": 0.4557,
"step": 123
},
{
"epoch": 1.3191489361702127,
"grad_norm": 0.8340292591613142,
"learning_rate": 7.363553254218253e-05,
"loss": 0.4492,
"step": 124
},
{
"epoch": 1.3297872340425532,
"grad_norm": 1.2214430144107857,
"learning_rate": 7.347382554139733e-05,
"loss": 0.4563,
"step": 125
},
{
"epoch": 1.3404255319148937,
"grad_norm": 0.7515274378448177,
"learning_rate": 7.331027215518949e-05,
"loss": 0.4574,
"step": 126
},
{
"epoch": 1.351063829787234,
"grad_norm": 1.0593819816413108,
"learning_rate": 7.31448814050133e-05,
"loss": 0.4586,
"step": 127
},
{
"epoch": 1.3617021276595744,
"grad_norm": 0.8257535268103262,
"learning_rate": 7.297766241367041e-05,
"loss": 0.4525,
"step": 128
},
{
"epoch": 1.372340425531915,
"grad_norm": 0.8489466970658132,
"learning_rate": 7.280862440480658e-05,
"loss": 0.4486,
"step": 129
},
{
"epoch": 1.3829787234042552,
"grad_norm": 1.0407070508163005,
"learning_rate": 7.263777670240282e-05,
"loss": 0.4518,
"step": 130
},
{
"epoch": 1.3936170212765957,
"grad_norm": 1.0870662063487706,
"learning_rate": 7.246512873026125e-05,
"loss": 0.4436,
"step": 131
},
{
"epoch": 1.4042553191489362,
"grad_norm": 1.040969884122318,
"learning_rate": 7.229069001148518e-05,
"loss": 0.4376,
"step": 132
},
{
"epoch": 1.4148936170212765,
"grad_norm": 0.8210287133544034,
"learning_rate": 7.211447016795388e-05,
"loss": 0.4459,
"step": 133
},
{
"epoch": 1.425531914893617,
"grad_norm": 0.6924416024292802,
"learning_rate": 7.193647891979177e-05,
"loss": 0.4492,
"step": 134
},
{
"epoch": 1.4361702127659575,
"grad_norm": 0.691504635298063,
"learning_rate": 7.17567260848324e-05,
"loss": 0.4382,
"step": 135
},
{
"epoch": 1.4468085106382977,
"grad_norm": 0.557066728314828,
"learning_rate": 7.157522157807675e-05,
"loss": 0.4475,
"step": 136
},
{
"epoch": 1.4574468085106382,
"grad_norm": 0.578188119392695,
"learning_rate": 7.139197541114645e-05,
"loss": 0.4416,
"step": 137
},
{
"epoch": 1.4680851063829787,
"grad_norm": 0.8694472070990833,
"learning_rate": 7.120699769173149e-05,
"loss": 0.4492,
"step": 138
},
{
"epoch": 1.4787234042553192,
"grad_norm": 1.0414676752540923,
"learning_rate": 7.10202986230327e-05,
"loss": 0.4371,
"step": 139
},
{
"epoch": 1.4893617021276595,
"grad_norm": 0.9048472155799596,
"learning_rate": 7.083188850319895e-05,
"loss": 0.4411,
"step": 140
},
{
"epoch": 1.5,
"grad_norm": 0.9161726375221393,
"learning_rate": 7.064177772475912e-05,
"loss": 0.4529,
"step": 141
},
{
"epoch": 1.5106382978723403,
"grad_norm": 1.1657596780724042,
"learning_rate": 7.044997677404888e-05,
"loss": 0.4433,
"step": 142
},
{
"epoch": 1.521276595744681,
"grad_norm": 0.8278836619484179,
"learning_rate": 7.025649623063223e-05,
"loss": 0.4432,
"step": 143
},
{
"epoch": 1.5319148936170213,
"grad_norm": 0.6982292221675673,
"learning_rate": 7.006134676671791e-05,
"loss": 0.4498,
"step": 144
},
{
"epoch": 1.5425531914893615,
"grad_norm": 1.1530549987417396,
"learning_rate": 6.986453914657083e-05,
"loss": 0.4484,
"step": 145
},
{
"epoch": 1.5531914893617023,
"grad_norm": 1.2290110194371844,
"learning_rate": 6.96660842259183e-05,
"loss": 0.4541,
"step": 146
},
{
"epoch": 1.5638297872340425,
"grad_norm": 0.6040374185478055,
"learning_rate": 6.946599295135116e-05,
"loss": 0.4356,
"step": 147
},
{
"epoch": 1.574468085106383,
"grad_norm": 0.5974414549578787,
"learning_rate": 6.926427635972003e-05,
"loss": 0.4374,
"step": 148
},
{
"epoch": 1.5851063829787235,
"grad_norm": 0.9321005443290319,
"learning_rate": 6.906094557752654e-05,
"loss": 0.4432,
"step": 149
},
{
"epoch": 1.5957446808510638,
"grad_norm": 1.0909193280528477,
"learning_rate": 6.885601182030958e-05,
"loss": 0.4474,
"step": 150
},
{
"epoch": 1.6063829787234043,
"grad_norm": 0.7679852263478857,
"learning_rate": 6.864948639202667e-05,
"loss": 0.4442,
"step": 151
},
{
"epoch": 1.6170212765957448,
"grad_norm": 0.7550553221749178,
"learning_rate": 6.844138068443043e-05,
"loss": 0.4452,
"step": 152
},
{
"epoch": 1.627659574468085,
"grad_norm": 1.1202478929189756,
"learning_rate": 6.823170617644029e-05,
"loss": 0.4391,
"step": 153
},
{
"epoch": 1.6382978723404256,
"grad_norm": 0.8957986675565583,
"learning_rate": 6.802047443350915e-05,
"loss": 0.44,
"step": 154
},
{
"epoch": 1.648936170212766,
"grad_norm": 0.7670887365074628,
"learning_rate": 6.78076971069857e-05,
"loss": 0.4427,
"step": 155
},
{
"epoch": 1.6595744680851063,
"grad_norm": 0.7780853531387013,
"learning_rate": 6.759338593347148e-05,
"loss": 0.4398,
"step": 156
},
{
"epoch": 1.6702127659574468,
"grad_norm": 0.7615691990716689,
"learning_rate": 6.737755273417367e-05,
"loss": 0.4456,
"step": 157
},
{
"epoch": 1.6808510638297873,
"grad_norm": 0.9908956798698196,
"learning_rate": 6.716020941425302e-05,
"loss": 0.4511,
"step": 158
},
{
"epoch": 1.6914893617021276,
"grad_norm": 1.1748417677944305,
"learning_rate": 6.694136796216706e-05,
"loss": 0.4442,
"step": 159
},
{
"epoch": 1.702127659574468,
"grad_norm": 0.7089766235204993,
"learning_rate": 6.672104044900901e-05,
"loss": 0.4437,
"step": 160
},
{
"epoch": 1.7127659574468086,
"grad_norm": 0.6183339909308402,
"learning_rate": 6.649923902784178e-05,
"loss": 0.4391,
"step": 161
},
{
"epoch": 1.7234042553191489,
"grad_norm": 0.6684140166542728,
"learning_rate": 6.627597593302772e-05,
"loss": 0.4543,
"step": 162
},
{
"epoch": 1.7340425531914894,
"grad_norm": 0.8241026926395437,
"learning_rate": 6.605126347955376e-05,
"loss": 0.446,
"step": 163
},
{
"epoch": 1.7446808510638299,
"grad_norm": 0.8675844769376447,
"learning_rate": 6.58251140623521e-05,
"loss": 0.4394,
"step": 164
},
{
"epoch": 1.7553191489361701,
"grad_norm": 0.7195091122160452,
"learning_rate": 6.559754015561655e-05,
"loss": 0.4361,
"step": 165
},
{
"epoch": 1.7659574468085106,
"grad_norm": 0.7586387307080512,
"learning_rate": 6.536855431211445e-05,
"loss": 0.4452,
"step": 166
},
{
"epoch": 1.7765957446808511,
"grad_norm": 0.8241109857261576,
"learning_rate": 6.513816916249427e-05,
"loss": 0.4462,
"step": 167
},
{
"epoch": 1.7872340425531914,
"grad_norm": 0.7828717831576161,
"learning_rate": 6.490639741458891e-05,
"loss": 0.4387,
"step": 168
},
{
"epoch": 1.797872340425532,
"grad_norm": 0.9841281333161127,
"learning_rate": 6.46732518527148e-05,
"loss": 0.4433,
"step": 169
},
{
"epoch": 1.8085106382978724,
"grad_norm": 0.8948930714592511,
"learning_rate": 6.443874533696662e-05,
"loss": 0.4329,
"step": 170
},
{
"epoch": 1.8191489361702127,
"grad_norm": 0.6167915147089398,
"learning_rate": 6.420289080250804e-05,
"loss": 0.4423,
"step": 171
},
{
"epoch": 1.8297872340425532,
"grad_norm": 0.7688689128970628,
"learning_rate": 6.396570125885823e-05,
"loss": 0.448,
"step": 172
},
{
"epoch": 1.8404255319148937,
"grad_norm": 0.7366333884582301,
"learning_rate": 6.372718978917421e-05,
"loss": 0.4421,
"step": 173
},
{
"epoch": 1.851063829787234,
"grad_norm": 0.6292983833226043,
"learning_rate": 6.348736954952923e-05,
"loss": 0.4345,
"step": 174
},
{
"epoch": 1.8617021276595744,
"grad_norm": 0.6094161942129656,
"learning_rate": 6.324625376818707e-05,
"loss": 0.4289,
"step": 175
},
{
"epoch": 1.872340425531915,
"grad_norm": 0.5223920021177019,
"learning_rate": 6.300385574487242e-05,
"loss": 0.4281,
"step": 176
},
{
"epoch": 1.8829787234042552,
"grad_norm": 0.43105768007225675,
"learning_rate": 6.276018885003727e-05,
"loss": 0.43,
"step": 177
},
{
"epoch": 1.8936170212765957,
"grad_norm": 0.4154261253991378,
"learning_rate": 6.251526652412335e-05,
"loss": 0.4272,
"step": 178
},
{
"epoch": 1.9042553191489362,
"grad_norm": 0.5129738950204691,
"learning_rate": 6.226910227682087e-05,
"loss": 0.4404,
"step": 179
},
{
"epoch": 1.9148936170212765,
"grad_norm": 0.4069384888530334,
"learning_rate": 6.202170968632324e-05,
"loss": 0.4379,
"step": 180
},
{
"epoch": 1.925531914893617,
"grad_norm": 0.4346407097225191,
"learning_rate": 6.177310239857815e-05,
"loss": 0.4393,
"step": 181
},
{
"epoch": 1.9361702127659575,
"grad_norm": 0.7030752370713033,
"learning_rate": 6.152329412653491e-05,
"loss": 0.4359,
"step": 182
},
{
"epoch": 1.9468085106382977,
"grad_norm": 0.7931293701516278,
"learning_rate": 6.127229864938798e-05,
"loss": 0.4324,
"step": 183
},
{
"epoch": 1.9574468085106385,
"grad_norm": 0.8569387442215538,
"learning_rate": 6.1020129811816985e-05,
"loss": 0.439,
"step": 184
},
{
"epoch": 1.9680851063829787,
"grad_norm": 0.9366916666402957,
"learning_rate": 6.076680152322302e-05,
"loss": 0.4365,
"step": 185
},
{
"epoch": 1.978723404255319,
"grad_norm": 0.9953725617355069,
"learning_rate": 6.051232775696143e-05,
"loss": 0.436,
"step": 186
},
{
"epoch": 1.9893617021276597,
"grad_norm": 1.0794213458303916,
"learning_rate": 6.025672254957106e-05,
"loss": 0.4423,
"step": 187
},
{
"epoch": 2.0,
"grad_norm": 0.724853495699258,
"learning_rate": 6.000000000000001e-05,
"loss": 0.4317,
"step": 188
},
{
"epoch": 2.0106382978723403,
"grad_norm": 0.5489002192429712,
"learning_rate": 5.9742174268827936e-05,
"loss": 0.4137,
"step": 189
},
{
"epoch": 2.021276595744681,
"grad_norm": 0.5979267122970597,
"learning_rate": 5.948325957748498e-05,
"loss": 0.4107,
"step": 190
},
{
"epoch": 2.0319148936170213,
"grad_norm": 0.6769157942235892,
"learning_rate": 5.9223270207467355e-05,
"loss": 0.415,
"step": 191
},
{
"epoch": 2.0425531914893615,
"grad_norm": 0.694157788349472,
"learning_rate": 5.896222049954951e-05,
"loss": 0.4138,
"step": 192
},
{
"epoch": 2.0531914893617023,
"grad_norm": 0.7900658328855281,
"learning_rate": 5.870012485299318e-05,
"loss": 0.415,
"step": 193
},
{
"epoch": 2.0638297872340425,
"grad_norm": 0.6937916417075501,
"learning_rate": 5.843699772475312e-05,
"loss": 0.4132,
"step": 194
},
{
"epoch": 2.074468085106383,
"grad_norm": 0.6698631990082241,
"learning_rate": 5.8172853628679676e-05,
"loss": 0.4121,
"step": 195
},
{
"epoch": 2.0851063829787235,
"grad_norm": 0.6656108376135776,
"learning_rate": 5.790770713471816e-05,
"loss": 0.4127,
"step": 196
},
{
"epoch": 2.095744680851064,
"grad_norm": 0.5836102935781896,
"learning_rate": 5.764157286810527e-05,
"loss": 0.4068,
"step": 197
},
{
"epoch": 2.106382978723404,
"grad_norm": 0.6583841935134351,
"learning_rate": 5.7374465508562324e-05,
"loss": 0.4063,
"step": 198
},
{
"epoch": 2.117021276595745,
"grad_norm": 0.5654983891994843,
"learning_rate": 5.710639978948555e-05,
"loss": 0.412,
"step": 199
},
{
"epoch": 2.127659574468085,
"grad_norm": 0.423035149723838,
"learning_rate": 5.6837390497133406e-05,
"loss": 0.4038,
"step": 200
},
{
"epoch": 2.1382978723404253,
"grad_norm": 0.6023613024743625,
"learning_rate": 5.6567452469810984e-05,
"loss": 0.4072,
"step": 201
},
{
"epoch": 2.148936170212766,
"grad_norm": 0.5872005390949395,
"learning_rate": 5.629660059705153e-05,
"loss": 0.4064,
"step": 202
},
{
"epoch": 2.1595744680851063,
"grad_norm": 0.38488472443140903,
"learning_rate": 5.602484981879519e-05,
"loss": 0.4088,
"step": 203
},
{
"epoch": 2.1702127659574466,
"grad_norm": 0.34867056077161274,
"learning_rate": 5.5752215124564895e-05,
"loss": 0.4165,
"step": 204
},
{
"epoch": 2.1808510638297873,
"grad_norm": 0.5189695063587118,
"learning_rate": 5.547871155263955e-05,
"loss": 0.403,
"step": 205
},
{
"epoch": 2.1914893617021276,
"grad_norm": 0.5349240957528729,
"learning_rate": 5.5204354189224596e-05,
"loss": 0.4076,
"step": 206
},
{
"epoch": 2.202127659574468,
"grad_norm": 0.4097332185439771,
"learning_rate": 5.492915816761979e-05,
"loss": 0.4055,
"step": 207
},
{
"epoch": 2.2127659574468086,
"grad_norm": 0.29969289622943385,
"learning_rate": 5.465313866738454e-05,
"loss": 0.4102,
"step": 208
},
{
"epoch": 2.223404255319149,
"grad_norm": 0.3210843691588522,
"learning_rate": 5.4376310913500514e-05,
"loss": 0.4082,
"step": 209
},
{
"epoch": 2.2340425531914896,
"grad_norm": 0.3342881154938907,
"learning_rate": 5.409869017553199e-05,
"loss": 0.4043,
"step": 210
},
{
"epoch": 2.24468085106383,
"grad_norm": 0.3152754379746835,
"learning_rate": 5.382029176678345e-05,
"loss": 0.4091,
"step": 211
},
{
"epoch": 2.25531914893617,
"grad_norm": 0.27611840786578706,
"learning_rate": 5.354113104345503e-05,
"loss": 0.4039,
"step": 212
},
{
"epoch": 2.2659574468085104,
"grad_norm": 0.34803624809617745,
"learning_rate": 5.326122340379539e-05,
"loss": 0.4063,
"step": 213
},
{
"epoch": 2.276595744680851,
"grad_norm": 0.36164081157802974,
"learning_rate": 5.2980584287252456e-05,
"loss": 0.4054,
"step": 214
},
{
"epoch": 2.2872340425531914,
"grad_norm": 0.28080091164784143,
"learning_rate": 5.269922917362171e-05,
"loss": 0.4137,
"step": 215
},
{
"epoch": 2.297872340425532,
"grad_norm": 0.3123559500548175,
"learning_rate": 5.241717358219239e-05,
"loss": 0.4136,
"step": 216
},
{
"epoch": 2.3085106382978724,
"grad_norm": 0.33068468215663827,
"learning_rate": 5.213443307089144e-05,
"loss": 0.4067,
"step": 217
},
{
"epoch": 2.3191489361702127,
"grad_norm": 0.2984606793541823,
"learning_rate": 5.1851023235425366e-05,
"loss": 0.3982,
"step": 218
},
{
"epoch": 2.329787234042553,
"grad_norm": 0.3545208178999491,
"learning_rate": 5.156695970841997e-05,
"loss": 0.41,
"step": 219
},
{
"epoch": 2.3404255319148937,
"grad_norm": 0.3590473616023012,
"learning_rate": 5.128225815855805e-05,
"loss": 0.4119,
"step": 220
},
{
"epoch": 2.351063829787234,
"grad_norm": 0.3815124462451615,
"learning_rate": 5.099693428971522e-05,
"loss": 0.4117,
"step": 221
},
{
"epoch": 2.3617021276595747,
"grad_norm": 0.3777633888640618,
"learning_rate": 5.0711003840093583e-05,
"loss": 0.4064,
"step": 222
},
{
"epoch": 2.372340425531915,
"grad_norm": 0.42841655218775976,
"learning_rate": 5.042448258135371e-05,
"loss": 0.4059,
"step": 223
},
{
"epoch": 2.382978723404255,
"grad_norm": 0.3957041429267785,
"learning_rate": 5.013738631774463e-05,
"loss": 0.4082,
"step": 224
},
{
"epoch": 2.393617021276596,
"grad_norm": 0.23548416366582237,
"learning_rate": 4.984973088523216e-05,
"loss": 0.4069,
"step": 225
},
{
"epoch": 2.404255319148936,
"grad_norm": 0.3532835233128077,
"learning_rate": 4.9561532150625305e-05,
"loss": 0.4123,
"step": 226
},
{
"epoch": 2.4148936170212765,
"grad_norm": 0.41182185945267635,
"learning_rate": 4.927280601070114e-05,
"loss": 0.4043,
"step": 227
},
{
"epoch": 2.425531914893617,
"grad_norm": 0.3444115639955698,
"learning_rate": 4.898356839132793e-05,
"loss": 0.4085,
"step": 228
},
{
"epoch": 2.4361702127659575,
"grad_norm": 0.3641718763669911,
"learning_rate": 4.869383524658668e-05,
"loss": 0.4044,
"step": 229
},
{
"epoch": 2.4468085106382977,
"grad_norm": 0.31325645363197496,
"learning_rate": 4.840362255789112e-05,
"loss": 0.4088,
"step": 230
},
{
"epoch": 2.4574468085106385,
"grad_norm": 0.22787584781999606,
"learning_rate": 4.811294633310617e-05,
"loss": 0.4072,
"step": 231
},
{
"epoch": 2.4680851063829787,
"grad_norm": 0.34212691771738396,
"learning_rate": 4.782182260566498e-05,
"loss": 0.4029,
"step": 232
},
{
"epoch": 2.478723404255319,
"grad_norm": 0.4204486154126815,
"learning_rate": 4.7530267433684546e-05,
"loss": 0.4082,
"step": 233
},
{
"epoch": 2.4893617021276597,
"grad_norm": 0.4183586609670886,
"learning_rate": 4.723829689907993e-05,
"loss": 0.4069,
"step": 234
},
{
"epoch": 2.5,
"grad_norm": 0.361420219603079,
"learning_rate": 4.694592710667723e-05,
"loss": 0.4051,
"step": 235
},
{
"epoch": 2.5106382978723403,
"grad_norm": 0.3230255284520481,
"learning_rate": 4.665317418332521e-05,
"loss": 0.4032,
"step": 236
},
{
"epoch": 2.521276595744681,
"grad_norm": 0.370284885999409,
"learning_rate": 4.6360054277005826e-05,
"loss": 0.406,
"step": 237
},
{
"epoch": 2.5319148936170213,
"grad_norm": 0.30097725112497414,
"learning_rate": 4.606658355594344e-05,
"loss": 0.3993,
"step": 238
},
{
"epoch": 2.5425531914893615,
"grad_norm": 0.32435130844112303,
"learning_rate": 4.577277820771307e-05,
"loss": 0.4102,
"step": 239
},
{
"epoch": 2.5531914893617023,
"grad_norm": 0.3080820117631687,
"learning_rate": 4.5478654438347414e-05,
"loss": 0.4026,
"step": 240
},
{
"epoch": 2.5638297872340425,
"grad_norm": 0.39816170983067595,
"learning_rate": 4.518422847144304e-05,
"loss": 0.406,
"step": 241
},
{
"epoch": 2.574468085106383,
"grad_norm": 0.3800310483317905,
"learning_rate": 4.488951654726539e-05,
"loss": 0.4073,
"step": 242
},
{
"epoch": 2.5851063829787235,
"grad_norm": 0.2671985596247953,
"learning_rate": 4.4594534921853096e-05,
"loss": 0.4038,
"step": 243
},
{
"epoch": 2.595744680851064,
"grad_norm": 0.3417581375350411,
"learning_rate": 4.429929986612125e-05,
"loss": 0.4047,
"step": 244
},
{
"epoch": 2.6063829787234045,
"grad_norm": 0.3200819491909366,
"learning_rate": 4.400382766496394e-05,
"loss": 0.4051,
"step": 245
},
{
"epoch": 2.617021276595745,
"grad_norm": 0.2966057678537604,
"learning_rate": 4.3708134616355934e-05,
"loss": 0.4106,
"step": 246
},
{
"epoch": 2.627659574468085,
"grad_norm": 0.31666189726009564,
"learning_rate": 4.341223703045379e-05,
"loss": 0.4044,
"step": 247
},
{
"epoch": 2.6382978723404253,
"grad_norm": 0.2379399207090575,
"learning_rate": 4.311615122869613e-05,
"loss": 0.4049,
"step": 248
},
{
"epoch": 2.648936170212766,
"grad_norm": 0.33070194627221655,
"learning_rate": 4.281989354290341e-05,
"loss": 0.4104,
"step": 249
},
{
"epoch": 2.6595744680851063,
"grad_norm": 0.29152996582110796,
"learning_rate": 4.2523480314376996e-05,
"loss": 0.4083,
"step": 250
},
{
"epoch": 2.670212765957447,
"grad_norm": 0.2768236406010446,
"learning_rate": 4.222692789299794e-05,
"loss": 0.4005,
"step": 251
},
{
"epoch": 2.6808510638297873,
"grad_norm": 0.32822073352952436,
"learning_rate": 4.193025263632495e-05,
"loss": 0.4098,
"step": 252
},
{
"epoch": 2.6914893617021276,
"grad_norm": 0.20759537070759246,
"learning_rate": 4.163347090869227e-05,
"loss": 0.4015,
"step": 253
},
{
"epoch": 2.702127659574468,
"grad_norm": 0.2450986247008019,
"learning_rate": 4.133659908030699e-05,
"loss": 0.4011,
"step": 254
},
{
"epoch": 2.7127659574468086,
"grad_norm": 0.3094748307247204,
"learning_rate": 4.103965352634604e-05,
"loss": 0.4128,
"step": 255
},
{
"epoch": 2.723404255319149,
"grad_norm": 0.2811649837424367,
"learning_rate": 4.0742650626053004e-05,
"loss": 0.409,
"step": 256
},
{
"epoch": 2.7340425531914896,
"grad_norm": 0.2666126452683894,
"learning_rate": 4.044560676183462e-05,
"loss": 0.4052,
"step": 257
},
{
"epoch": 2.74468085106383,
"grad_norm": 0.3043401200996446,
"learning_rate": 4.014853831835721e-05,
"loss": 0.4075,
"step": 258
},
{
"epoch": 2.75531914893617,
"grad_norm": 0.26121736766165093,
"learning_rate": 3.985146168164281e-05,
"loss": 0.4114,
"step": 259
},
{
"epoch": 2.7659574468085104,
"grad_norm": 0.25762450388003205,
"learning_rate": 3.9554393238165386e-05,
"loss": 0.4066,
"step": 260
},
{
"epoch": 2.776595744680851,
"grad_norm": 0.2541891436719679,
"learning_rate": 3.9257349373947016e-05,
"loss": 0.4059,
"step": 261
},
{
"epoch": 2.7872340425531914,
"grad_norm": 0.2722839322667676,
"learning_rate": 3.896034647365398e-05,
"loss": 0.4058,
"step": 262
},
{
"epoch": 2.797872340425532,
"grad_norm": 0.28452361022067413,
"learning_rate": 3.866340091969303e-05,
"loss": 0.4077,
"step": 263
},
{
"epoch": 2.8085106382978724,
"grad_norm": 0.2530140333177483,
"learning_rate": 3.836652909130774e-05,
"loss": 0.4093,
"step": 264
},
{
"epoch": 2.8191489361702127,
"grad_norm": 0.45406695750367465,
"learning_rate": 3.806974736367507e-05,
"loss": 0.4117,
"step": 265
},
{
"epoch": 2.829787234042553,
"grad_norm": 0.2988144349921366,
"learning_rate": 3.7773072107002084e-05,
"loss": 0.4023,
"step": 266
},
{
"epoch": 2.8404255319148937,
"grad_norm": 0.22739583662187962,
"learning_rate": 3.747651968562302e-05,
"loss": 0.4054,
"step": 267
},
{
"epoch": 2.851063829787234,
"grad_norm": 0.2636532613327371,
"learning_rate": 3.718010645709661e-05,
"loss": 0.4106,
"step": 268
},
{
"epoch": 2.8617021276595747,
"grad_norm": 0.24726890660848583,
"learning_rate": 3.688384877130388e-05,
"loss": 0.4059,
"step": 269
},
{
"epoch": 2.872340425531915,
"grad_norm": 0.2944014064456227,
"learning_rate": 3.658776296954622e-05,
"loss": 0.4051,
"step": 270
},
{
"epoch": 2.882978723404255,
"grad_norm": 0.22859295611193128,
"learning_rate": 3.629186538364408e-05,
"loss": 0.4108,
"step": 271
},
{
"epoch": 2.8936170212765955,
"grad_norm": 0.2939557054489036,
"learning_rate": 3.5996172335036065e-05,
"loss": 0.4004,
"step": 272
},
{
"epoch": 2.904255319148936,
"grad_norm": 0.26529631331687625,
"learning_rate": 3.570070013387876e-05,
"loss": 0.4076,
"step": 273
},
{
"epoch": 2.9148936170212765,
"grad_norm": 0.25501324633080297,
"learning_rate": 3.540546507814692e-05,
"loss": 0.3984,
"step": 274
},
{
"epoch": 2.925531914893617,
"grad_norm": 0.21459730505126537,
"learning_rate": 3.5110483452734633e-05,
"loss": 0.4025,
"step": 275
},
{
"epoch": 2.9361702127659575,
"grad_norm": 0.2046281102038189,
"learning_rate": 3.4815771528556976e-05,
"loss": 0.4066,
"step": 276
},
{
"epoch": 2.9468085106382977,
"grad_norm": 0.25983079535813713,
"learning_rate": 3.452134556165259e-05,
"loss": 0.4062,
"step": 277
},
{
"epoch": 2.9574468085106385,
"grad_norm": 0.18624346254163252,
"learning_rate": 3.4227221792286945e-05,
"loss": 0.4101,
"step": 278
},
{
"epoch": 2.9680851063829787,
"grad_norm": 0.25858622274413134,
"learning_rate": 3.393341644405657e-05,
"loss": 0.3966,
"step": 279
},
{
"epoch": 2.978723404255319,
"grad_norm": 0.21292012777230374,
"learning_rate": 3.363994572299418e-05,
"loss": 0.4061,
"step": 280
},
{
"epoch": 2.9893617021276597,
"grad_norm": 0.2509387412075419,
"learning_rate": 3.3346825816674804e-05,
"loss": 0.4108,
"step": 281
},
{
"epoch": 3.0,
"grad_norm": 0.23091413343534087,
"learning_rate": 3.305407289332279e-05,
"loss": 0.3937,
"step": 282
},
{
"epoch": 3.0106382978723403,
"grad_norm": 0.20905420882080966,
"learning_rate": 3.276170310092008e-05,
"loss": 0.3778,
"step": 283
},
{
"epoch": 3.021276595744681,
"grad_norm": 0.23991076346040097,
"learning_rate": 3.246973256631546e-05,
"loss": 0.3798,
"step": 284
},
{
"epoch": 3.0319148936170213,
"grad_norm": 0.24449335128970107,
"learning_rate": 3.217817739433502e-05,
"loss": 0.3802,
"step": 285
},
{
"epoch": 3.0425531914893615,
"grad_norm": 0.245315113180501,
"learning_rate": 3.1887053666893834e-05,
"loss": 0.3875,
"step": 286
},
{
"epoch": 3.0531914893617023,
"grad_norm": 0.24057913704130401,
"learning_rate": 3.159637744210888e-05,
"loss": 0.3738,
"step": 287
},
{
"epoch": 3.0638297872340425,
"grad_norm": 0.2549241752014073,
"learning_rate": 3.130616475341332e-05,
"loss": 0.3814,
"step": 288
},
{
"epoch": 3.074468085106383,
"grad_norm": 0.21160294231186252,
"learning_rate": 3.101643160867208e-05,
"loss": 0.3815,
"step": 289
},
{
"epoch": 3.0851063829787235,
"grad_norm": 0.23805410062655133,
"learning_rate": 3.072719398929887e-05,
"loss": 0.3801,
"step": 290
},
{
"epoch": 3.095744680851064,
"grad_norm": 0.20195712835328383,
"learning_rate": 3.0438467849374702e-05,
"loss": 0.3797,
"step": 291
},
{
"epoch": 3.106382978723404,
"grad_norm": 0.23312007432367898,
"learning_rate": 3.0150269114767862e-05,
"loss": 0.3764,
"step": 292
},
{
"epoch": 3.117021276595745,
"grad_norm": 0.18274301420799274,
"learning_rate": 2.9862613682255383e-05,
"loss": 0.3754,
"step": 293
},
{
"epoch": 3.127659574468085,
"grad_norm": 0.2478292014777071,
"learning_rate": 2.957551741864631e-05,
"loss": 0.3767,
"step": 294
},
{
"epoch": 3.1382978723404253,
"grad_norm": 0.1927262119265681,
"learning_rate": 2.928899615990643e-05,
"loss": 0.375,
"step": 295
},
{
"epoch": 3.148936170212766,
"grad_norm": 0.2868229283783766,
"learning_rate": 2.90030657102848e-05,
"loss": 0.3808,
"step": 296
},
{
"epoch": 3.1595744680851063,
"grad_norm": 0.18547034376158922,
"learning_rate": 2.8717741841441964e-05,
"loss": 0.3767,
"step": 297
},
{
"epoch": 3.1702127659574466,
"grad_norm": 0.278737988337013,
"learning_rate": 2.8433040291580053e-05,
"loss": 0.3809,
"step": 298
},
{
"epoch": 3.1808510638297873,
"grad_norm": 0.1707915913816605,
"learning_rate": 2.8148976764574648e-05,
"loss": 0.377,
"step": 299
},
{
"epoch": 3.1914893617021276,
"grad_norm": 0.2487011269688432,
"learning_rate": 2.7865566929108573e-05,
"loss": 0.3838,
"step": 300
},
{
"epoch": 3.202127659574468,
"grad_norm": 0.17967682539404906,
"learning_rate": 2.758282641780762e-05,
"loss": 0.382,
"step": 301
},
{
"epoch": 3.2127659574468086,
"grad_norm": 0.21709328897403127,
"learning_rate": 2.7300770826378302e-05,
"loss": 0.3816,
"step": 302
},
{
"epoch": 3.223404255319149,
"grad_norm": 0.178520628301021,
"learning_rate": 2.7019415712747558e-05,
"loss": 0.3834,
"step": 303
},
{
"epoch": 3.2340425531914896,
"grad_norm": 0.21223809263671378,
"learning_rate": 2.6738776596204624e-05,
"loss": 0.3832,
"step": 304
},
{
"epoch": 3.24468085106383,
"grad_norm": 0.15649445787021687,
"learning_rate": 2.6458868956544984e-05,
"loss": 0.3726,
"step": 305
},
{
"epoch": 3.25531914893617,
"grad_norm": 0.1764738407571,
"learning_rate": 2.6179708233216557e-05,
"loss": 0.3831,
"step": 306
},
{
"epoch": 3.2659574468085104,
"grad_norm": 0.15542729747310566,
"learning_rate": 2.590130982446802e-05,
"loss": 0.3736,
"step": 307
},
{
"epoch": 3.276595744680851,
"grad_norm": 0.1747611677030782,
"learning_rate": 2.5623689086499496e-05,
"loss": 0.3793,
"step": 308
},
{
"epoch": 3.2872340425531914,
"grad_norm": 0.13212244563255485,
"learning_rate": 2.5346861332615476e-05,
"loss": 0.3767,
"step": 309
},
{
"epoch": 3.297872340425532,
"grad_norm": 0.17023514875738952,
"learning_rate": 2.5070841832380212e-05,
"loss": 0.3879,
"step": 310
},
{
"epoch": 3.3085106382978724,
"grad_norm": 0.15961770967934297,
"learning_rate": 2.4795645810775414e-05,
"loss": 0.3769,
"step": 311
},
{
"epoch": 3.3191489361702127,
"grad_norm": 0.18863144944962443,
"learning_rate": 2.4521288447360457e-05,
"loss": 0.3814,
"step": 312
},
{
"epoch": 3.329787234042553,
"grad_norm": 0.17296200115217117,
"learning_rate": 2.424778487543512e-05,
"loss": 0.38,
"step": 313
},
{
"epoch": 3.3404255319148937,
"grad_norm": 0.18291540188393604,
"learning_rate": 2.3975150181204817e-05,
"loss": 0.3736,
"step": 314
},
{
"epoch": 3.351063829787234,
"grad_norm": 0.1539347041742154,
"learning_rate": 2.370339940294848e-05,
"loss": 0.374,
"step": 315
},
{
"epoch": 3.3617021276595747,
"grad_norm": 0.17067354232177032,
"learning_rate": 2.3432547530189033e-05,
"loss": 0.375,
"step": 316
},
{
"epoch": 3.372340425531915,
"grad_norm": 0.16125243954444898,
"learning_rate": 2.316260950286661e-05,
"loss": 0.386,
"step": 317
},
{
"epoch": 3.382978723404255,
"grad_norm": 0.17706581816471947,
"learning_rate": 2.2893600210514464e-05,
"loss": 0.3859,
"step": 318
},
{
"epoch": 3.393617021276596,
"grad_norm": 0.19387187739363282,
"learning_rate": 2.2625534491437672e-05,
"loss": 0.3818,
"step": 319
},
{
"epoch": 3.404255319148936,
"grad_norm": 0.16113410459885513,
"learning_rate": 2.2358427131894732e-05,
"loss": 0.3762,
"step": 320
},
{
"epoch": 3.4148936170212765,
"grad_norm": 0.20116908168809555,
"learning_rate": 2.2092292865281845e-05,
"loss": 0.3803,
"step": 321
},
{
"epoch": 3.425531914893617,
"grad_norm": 0.16099148978300246,
"learning_rate": 2.1827146371320334e-05,
"loss": 0.3822,
"step": 322
},
{
"epoch": 3.4361702127659575,
"grad_norm": 0.18990059248934593,
"learning_rate": 2.156300227524688e-05,
"loss": 0.3825,
"step": 323
},
{
"epoch": 3.4468085106382977,
"grad_norm": 0.16540658424680887,
"learning_rate": 2.1299875147006838e-05,
"loss": 0.3781,
"step": 324
},
{
"epoch": 3.4574468085106385,
"grad_norm": 0.16752244552028583,
"learning_rate": 2.10377795004505e-05,
"loss": 0.3809,
"step": 325
},
{
"epoch": 3.4680851063829787,
"grad_norm": 0.1810597095974272,
"learning_rate": 2.0776729792532652e-05,
"loss": 0.3836,
"step": 326
},
{
"epoch": 3.478723404255319,
"grad_norm": 0.1417871274980451,
"learning_rate": 2.0516740422515022e-05,
"loss": 0.3878,
"step": 327
},
{
"epoch": 3.4893617021276597,
"grad_norm": 0.15754394939129449,
"learning_rate": 2.0257825731172077e-05,
"loss": 0.3833,
"step": 328
},
{
"epoch": 3.5,
"grad_norm": 0.12745856435079908,
"learning_rate": 2.0000000000000012e-05,
"loss": 0.3736,
"step": 329
},
{
"epoch": 3.5106382978723403,
"grad_norm": 0.13505830495843163,
"learning_rate": 1.9743277450428962e-05,
"loss": 0.3807,
"step": 330
},
{
"epoch": 3.521276595744681,
"grad_norm": 0.12857257976193204,
"learning_rate": 1.9487672243038594e-05,
"loss": 0.381,
"step": 331
},
{
"epoch": 3.5319148936170213,
"grad_norm": 0.13541243433649724,
"learning_rate": 1.9233198476777003e-05,
"loss": 0.385,
"step": 332
},
{
"epoch": 3.5425531914893615,
"grad_norm": 0.13836684138720814,
"learning_rate": 1.897987018818302e-05,
"loss": 0.3808,
"step": 333
},
{
"epoch": 3.5531914893617023,
"grad_norm": 0.12028403797052283,
"learning_rate": 1.8727701350612026e-05,
"loss": 0.3787,
"step": 334
},
{
"epoch": 3.5638297872340425,
"grad_norm": 0.13853549831902906,
"learning_rate": 1.8476705873465097e-05,
"loss": 0.3845,
"step": 335
},
{
"epoch": 3.574468085106383,
"grad_norm": 0.12529122214780986,
"learning_rate": 1.8226897601421858e-05,
"loss": 0.3773,
"step": 336
},
{
"epoch": 3.5851063829787235,
"grad_norm": 0.13718627686941767,
"learning_rate": 1.7978290313676774e-05,
"loss": 0.3771,
"step": 337
},
{
"epoch": 3.595744680851064,
"grad_norm": 0.13452451362504975,
"learning_rate": 1.7730897723179144e-05,
"loss": 0.3799,
"step": 338
},
{
"epoch": 3.6063829787234045,
"grad_norm": 0.13390152462264926,
"learning_rate": 1.748473347587666e-05,
"loss": 0.3784,
"step": 339
},
{
"epoch": 3.617021276595745,
"grad_norm": 0.1381149397634833,
"learning_rate": 1.7239811149962756e-05,
"loss": 0.3759,
"step": 340
},
{
"epoch": 3.627659574468085,
"grad_norm": 0.12941518047233325,
"learning_rate": 1.6996144255127586e-05,
"loss": 0.377,
"step": 341
},
{
"epoch": 3.6382978723404253,
"grad_norm": 0.14340217919626994,
"learning_rate": 1.675374623181294e-05,
"loss": 0.3799,
"step": 342
},
{
"epoch": 3.648936170212766,
"grad_norm": 0.11836717659820724,
"learning_rate": 1.6512630450470784e-05,
"loss": 0.3768,
"step": 343
},
{
"epoch": 3.6595744680851063,
"grad_norm": 0.15778520770347512,
"learning_rate": 1.6272810210825794e-05,
"loss": 0.3832,
"step": 344
},
{
"epoch": 3.670212765957447,
"grad_norm": 0.1204529845934934,
"learning_rate": 1.6034298741141768e-05,
"loss": 0.382,
"step": 345
},
{
"epoch": 3.6808510638297873,
"grad_norm": 0.15750579133093734,
"learning_rate": 1.579710919749196e-05,
"loss": 0.3798,
"step": 346
},
{
"epoch": 3.6914893617021276,
"grad_norm": 0.1386144145229538,
"learning_rate": 1.5561254663033393e-05,
"loss": 0.3764,
"step": 347
},
{
"epoch": 3.702127659574468,
"grad_norm": 0.1239231423498722,
"learning_rate": 1.53267481472852e-05,
"loss": 0.3827,
"step": 348
},
{
"epoch": 3.7127659574468086,
"grad_norm": 0.14157764395949052,
"learning_rate": 1.5093602585411078e-05,
"loss": 0.3773,
"step": 349
},
{
"epoch": 3.723404255319149,
"grad_norm": 0.140686345314232,
"learning_rate": 1.4861830837505733e-05,
"loss": 0.3784,
"step": 350
},
{
"epoch": 3.7340425531914896,
"grad_norm": 0.12320059023506465,
"learning_rate": 1.4631445687885553e-05,
"loss": 0.3844,
"step": 351
},
{
"epoch": 3.74468085106383,
"grad_norm": 0.14431820751164123,
"learning_rate": 1.4402459844383451e-05,
"loss": 0.3803,
"step": 352
},
{
"epoch": 3.75531914893617,
"grad_norm": 0.12143326106626687,
"learning_rate": 1.4174885937647905e-05,
"loss": 0.383,
"step": 353
},
{
"epoch": 3.7659574468085104,
"grad_norm": 0.11735894401221722,
"learning_rate": 1.3948736520446246e-05,
"loss": 0.3751,
"step": 354
},
{
"epoch": 3.776595744680851,
"grad_norm": 0.13185033729281154,
"learning_rate": 1.372402406697229e-05,
"loss": 0.3737,
"step": 355
},
{
"epoch": 3.7872340425531914,
"grad_norm": 0.1154428148825697,
"learning_rate": 1.3500760972158223e-05,
"loss": 0.378,
"step": 356
},
{
"epoch": 3.797872340425532,
"grad_norm": 0.12842073282310573,
"learning_rate": 1.3278959550991011e-05,
"loss": 0.3775,
"step": 357
},
{
"epoch": 3.8085106382978724,
"grad_norm": 0.12670592325854088,
"learning_rate": 1.3058632037832957e-05,
"loss": 0.3828,
"step": 358
},
{
"epoch": 3.8191489361702127,
"grad_norm": 0.11334725671446295,
"learning_rate": 1.2839790585747008e-05,
"loss": 0.3784,
"step": 359
},
{
"epoch": 3.829787234042553,
"grad_norm": 0.11759887567137367,
"learning_rate": 1.2622447265826345e-05,
"loss": 0.3823,
"step": 360
},
{
"epoch": 3.8404255319148937,
"grad_norm": 0.11200560690429617,
"learning_rate": 1.2406614066528543e-05,
"loss": 0.3839,
"step": 361
},
{
"epoch": 3.851063829787234,
"grad_norm": 0.10929402999457798,
"learning_rate": 1.219230289301431e-05,
"loss": 0.38,
"step": 362
},
{
"epoch": 3.8617021276595747,
"grad_norm": 0.12132649925473797,
"learning_rate": 1.1979525566490845e-05,
"loss": 0.3752,
"step": 363
},
{
"epoch": 3.872340425531915,
"grad_norm": 0.11204838908548817,
"learning_rate": 1.176829382355973e-05,
"loss": 0.3752,
"step": 364
},
{
"epoch": 3.882978723404255,
"grad_norm": 0.11317597743783404,
"learning_rate": 1.1558619315569572e-05,
"loss": 0.3815,
"step": 365
},
{
"epoch": 3.8936170212765955,
"grad_norm": 0.13136090336497436,
"learning_rate": 1.1350513607973351e-05,
"loss": 0.3839,
"step": 366
},
{
"epoch": 3.904255319148936,
"grad_norm": 0.11332184292598117,
"learning_rate": 1.1143988179690441e-05,
"loss": 0.3784,
"step": 367
},
{
"epoch": 3.9148936170212765,
"grad_norm": 0.12049799122378656,
"learning_rate": 1.093905442247348e-05,
"loss": 0.3737,
"step": 368
},
{
"epoch": 3.925531914893617,
"grad_norm": 0.12112505146396743,
"learning_rate": 1.073572364027999e-05,
"loss": 0.3784,
"step": 369
},
{
"epoch": 3.9361702127659575,
"grad_norm": 0.11152808514251403,
"learning_rate": 1.0534007048648846e-05,
"loss": 0.3836,
"step": 370
},
{
"epoch": 3.9468085106382977,
"grad_norm": 0.11026956462916394,
"learning_rate": 1.0333915774081698e-05,
"loss": 0.382,
"step": 371
},
{
"epoch": 3.9574468085106385,
"grad_norm": 0.10840535830420614,
"learning_rate": 1.0135460853429166e-05,
"loss": 0.3734,
"step": 372
},
{
"epoch": 3.9680851063829787,
"grad_norm": 0.1236335149703349,
"learning_rate": 9.938653233282105e-06,
"loss": 0.3763,
"step": 373
},
{
"epoch": 3.978723404255319,
"grad_norm": 0.12252596041380151,
"learning_rate": 9.74350376936779e-06,
"loss": 0.3716,
"step": 374
},
{
"epoch": 3.9893617021276597,
"grad_norm": 0.10540461633402395,
"learning_rate": 9.550023225951124e-06,
"loss": 0.3725,
"step": 375
},
{
"epoch": 4.0,
"grad_norm": 0.12520513195330457,
"learning_rate": 9.358222275240884e-06,
"loss": 0.3755,
"step": 376
},
{
"epoch": 4.01063829787234,
"grad_norm": 0.14847057489702098,
"learning_rate": 9.168111496801071e-06,
"loss": 0.3652,
"step": 377
},
{
"epoch": 4.0212765957446805,
"grad_norm": 0.11332943867336272,
"learning_rate": 8.979701376967313e-06,
"loss": 0.364,
"step": 378
},
{
"epoch": 4.031914893617022,
"grad_norm": 0.1232583568124227,
"learning_rate": 8.793002308268521e-06,
"loss": 0.3577,
"step": 379
},
{
"epoch": 4.042553191489362,
"grad_norm": 0.12780375441362188,
"learning_rate": 8.608024588853561e-06,
"loss": 0.3635,
"step": 380
},
{
"epoch": 4.053191489361702,
"grad_norm": 0.12259673347799972,
"learning_rate": 8.424778421923258e-06,
"loss": 0.3622,
"step": 381
},
{
"epoch": 4.0638297872340425,
"grad_norm": 0.11562199484437485,
"learning_rate": 8.24327391516761e-06,
"loss": 0.3619,
"step": 382
},
{
"epoch": 4.074468085106383,
"grad_norm": 0.11649371701938849,
"learning_rate": 8.06352108020823e-06,
"loss": 0.367,
"step": 383
},
{
"epoch": 4.085106382978723,
"grad_norm": 0.10570361375232744,
"learning_rate": 7.885529832046134e-06,
"loss": 0.3582,
"step": 384
},
{
"epoch": 4.095744680851064,
"grad_norm": 0.10461915141244181,
"learning_rate": 7.709309988514824e-06,
"loss": 0.3608,
"step": 385
},
{
"epoch": 4.1063829787234045,
"grad_norm": 0.11511232586029703,
"learning_rate": 7.534871269738753e-06,
"loss": 0.3647,
"step": 386
},
{
"epoch": 4.117021276595745,
"grad_norm": 0.10867329200636058,
"learning_rate": 7.362223297597184e-06,
"loss": 0.3629,
"step": 387
},
{
"epoch": 4.127659574468085,
"grad_norm": 0.10982240516813774,
"learning_rate": 7.191375595193433e-06,
"loss": 0.3691,
"step": 388
},
{
"epoch": 4.138297872340425,
"grad_norm": 0.11776263009120118,
"learning_rate": 7.022337586329597e-06,
"loss": 0.3617,
"step": 389
},
{
"epoch": 4.148936170212766,
"grad_norm": 0.10569398652654016,
"learning_rate": 6.855118594986718e-06,
"loss": 0.3567,
"step": 390
},
{
"epoch": 4.159574468085107,
"grad_norm": 0.10505309486877212,
"learning_rate": 6.6897278448105405e-06,
"loss": 0.3639,
"step": 391
},
{
"epoch": 4.170212765957447,
"grad_norm": 0.1079113458903354,
"learning_rate": 6.526174458602681e-06,
"loss": 0.368,
"step": 392
},
{
"epoch": 4.180851063829787,
"grad_norm": 0.11197666687109296,
"learning_rate": 6.364467457817482e-06,
"loss": 0.3587,
"step": 393
},
{
"epoch": 4.191489361702128,
"grad_norm": 0.11278688179148387,
"learning_rate": 6.20461576206433e-06,
"loss": 0.3611,
"step": 394
},
{
"epoch": 4.202127659574468,
"grad_norm": 0.11202771282958791,
"learning_rate": 6.046628188615718e-06,
"loss": 0.3632,
"step": 395
},
{
"epoch": 4.212765957446808,
"grad_norm": 0.098500856006095,
"learning_rate": 5.890513451920843e-06,
"loss": 0.3538,
"step": 396
},
{
"epoch": 4.223404255319149,
"grad_norm": 0.09784156477819876,
"learning_rate": 5.736280163124974e-06,
"loss": 0.365,
"step": 397
},
{
"epoch": 4.23404255319149,
"grad_norm": 0.10953538933768776,
"learning_rate": 5.583936829594434e-06,
"loss": 0.3603,
"step": 398
},
{
"epoch": 4.24468085106383,
"grad_norm": 0.10639360562326262,
"learning_rate": 5.4334918544473436e-06,
"loss": 0.3659,
"step": 399
},
{
"epoch": 4.25531914893617,
"grad_norm": 0.10211474883073758,
"learning_rate": 5.284953536090131e-06,
"loss": 0.3641,
"step": 400
},
{
"epoch": 4.26595744680851,
"grad_norm": 0.10574255358672346,
"learning_rate": 5.1383300677598024e-06,
"loss": 0.3643,
"step": 401
},
{
"epoch": 4.276595744680851,
"grad_norm": 0.11115010018265371,
"learning_rate": 4.993629537071978e-06,
"loss": 0.3701,
"step": 402
},
{
"epoch": 4.287234042553192,
"grad_norm": 0.09127212979081487,
"learning_rate": 4.850859925574809e-06,
"loss": 0.3662,
"step": 403
},
{
"epoch": 4.297872340425532,
"grad_norm": 0.08625180913725822,
"learning_rate": 4.710029108308733e-06,
"loss": 0.3548,
"step": 404
},
{
"epoch": 4.308510638297872,
"grad_norm": 0.10510607272882863,
"learning_rate": 4.571144853372063e-06,
"loss": 0.363,
"step": 405
},
{
"epoch": 4.319148936170213,
"grad_norm": 0.09294640053501395,
"learning_rate": 4.434214821492542e-06,
"loss": 0.3659,
"step": 406
},
{
"epoch": 4.329787234042553,
"grad_norm": 0.10322126590894673,
"learning_rate": 4.299246565604756e-06,
"loss": 0.3636,
"step": 407
},
{
"epoch": 4.340425531914893,
"grad_norm": 0.08513844389462451,
"learning_rate": 4.166247530433531e-06,
"loss": 0.361,
"step": 408
},
{
"epoch": 4.351063829787234,
"grad_norm": 0.09456400637267966,
"learning_rate": 4.035225052083309e-06,
"loss": 0.3631,
"step": 409
},
{
"epoch": 4.361702127659575,
"grad_norm": 0.09912377148014549,
"learning_rate": 3.906186357633455e-06,
"loss": 0.3627,
"step": 410
},
{
"epoch": 4.372340425531915,
"grad_norm": 0.1007261033413922,
"learning_rate": 3.779138564739646e-06,
"loss": 0.3625,
"step": 411
},
{
"epoch": 4.382978723404255,
"grad_norm": 0.08934724234816895,
"learning_rate": 3.6540886812412547e-06,
"loss": 0.3648,
"step": 412
},
{
"epoch": 4.3936170212765955,
"grad_norm": 0.0925468177451363,
"learning_rate": 3.5310436047748263e-06,
"loss": 0.3581,
"step": 413
},
{
"epoch": 4.404255319148936,
"grad_norm": 0.08919561633872664,
"learning_rate": 3.4100101223935743e-06,
"loss": 0.3571,
"step": 414
},
{
"epoch": 4.414893617021277,
"grad_norm": 0.09499084946810057,
"learning_rate": 3.290994910193037e-06,
"loss": 0.3579,
"step": 415
},
{
"epoch": 4.425531914893617,
"grad_norm": 0.09014647001183278,
"learning_rate": 3.174004532942845e-06,
"loss": 0.3689,
"step": 416
},
{
"epoch": 4.4361702127659575,
"grad_norm": 0.08581817647958025,
"learning_rate": 3.059045443724582e-06,
"loss": 0.3623,
"step": 417
},
{
"epoch": 4.446808510638298,
"grad_norm": 0.09478322839033952,
"learning_rate": 2.946123983575846e-06,
"loss": 0.3621,
"step": 418
},
{
"epoch": 4.457446808510638,
"grad_norm": 0.09127292341411823,
"learning_rate": 2.8352463811404952e-06,
"loss": 0.352,
"step": 419
},
{
"epoch": 4.468085106382979,
"grad_norm": 0.09314064563107814,
"learning_rate": 2.726418752325084e-06,
"loss": 0.3604,
"step": 420
},
{
"epoch": 4.4787234042553195,
"grad_norm": 0.08258953047242434,
"learning_rate": 2.619647099961502e-06,
"loss": 0.3613,
"step": 421
},
{
"epoch": 4.48936170212766,
"grad_norm": 0.08671649480194962,
"learning_rate": 2.514937313475865e-06,
"loss": 0.3619,
"step": 422
},
{
"epoch": 4.5,
"grad_norm": 0.08348749730392281,
"learning_rate": 2.4122951685636674e-06,
"loss": 0.362,
"step": 423
},
{
"epoch": 4.51063829787234,
"grad_norm": 0.08847421831587975,
"learning_rate": 2.3117263268712e-06,
"loss": 0.3643,
"step": 424
},
{
"epoch": 4.5212765957446805,
"grad_norm": 0.08541200353819858,
"learning_rate": 2.2132363356832532e-06,
"loss": 0.3617,
"step": 425
},
{
"epoch": 4.531914893617021,
"grad_norm": 0.09072202451972457,
"learning_rate": 2.11683062761713e-06,
"loss": 0.3637,
"step": 426
},
{
"epoch": 4.542553191489362,
"grad_norm": 0.08771913713301659,
"learning_rate": 2.0225145203230044e-06,
"loss": 0.3635,
"step": 427
},
{
"epoch": 4.553191489361702,
"grad_norm": 0.08227630624973846,
"learning_rate": 1.930293216190586e-06,
"loss": 0.3677,
"step": 428
},
{
"epoch": 4.5638297872340425,
"grad_norm": 0.08142816323777968,
"learning_rate": 1.8401718020621694e-06,
"loss": 0.3653,
"step": 429
},
{
"epoch": 4.574468085106383,
"grad_norm": 0.08626743211827775,
"learning_rate": 1.7521552489520566e-06,
"loss": 0.3685,
"step": 430
},
{
"epoch": 4.585106382978723,
"grad_norm": 0.08468387346146096,
"learning_rate": 1.666248411772342e-06,
"loss": 0.3651,
"step": 431
},
{
"epoch": 4.595744680851064,
"grad_norm": 0.08278070326775537,
"learning_rate": 1.5824560290651404e-06,
"loss": 0.3638,
"step": 432
},
{
"epoch": 4.6063829787234045,
"grad_norm": 0.08542627322845557,
"learning_rate": 1.5007827227411942e-06,
"loss": 0.3608,
"step": 433
},
{
"epoch": 4.617021276595745,
"grad_norm": 0.0850283799369069,
"learning_rate": 1.4212329978249417e-06,
"loss": 0.3694,
"step": 434
},
{
"epoch": 4.627659574468085,
"grad_norm": 0.08375767787904646,
"learning_rate": 1.3438112422060256e-06,
"loss": 0.3572,
"step": 435
},
{
"epoch": 4.638297872340425,
"grad_norm": 0.08181511297244645,
"learning_rate": 1.2685217263972693e-06,
"loss": 0.3625,
"step": 436
},
{
"epoch": 4.648936170212766,
"grad_norm": 0.08280046685526832,
"learning_rate": 1.1953686032990964e-06,
"loss": 0.3604,
"step": 437
},
{
"epoch": 4.659574468085106,
"grad_norm": 0.07768111306576052,
"learning_rate": 1.124355907970487e-06,
"loss": 0.3642,
"step": 438
},
{
"epoch": 4.670212765957447,
"grad_norm": 0.08142474629210497,
"learning_rate": 1.0554875574063784e-06,
"loss": 0.3645,
"step": 439
},
{
"epoch": 4.680851063829787,
"grad_norm": 0.0795668987754362,
"learning_rate": 9.887673503216422e-07,
"loss": 0.3663,
"step": 440
},
{
"epoch": 4.691489361702128,
"grad_norm": 0.07508730919255964,
"learning_rate": 9.241989669415097e-07,
"loss": 0.3661,
"step": 441
},
{
"epoch": 4.702127659574468,
"grad_norm": 0.08240799525472663,
"learning_rate": 8.61785968798623e-07,
"loss": 0.3608,
"step": 442
},
{
"epoch": 4.712765957446808,
"grad_norm": 0.07849402490031887,
"learning_rate": 8.015317985365301e-07,
"loss": 0.3616,
"step": 443
},
{
"epoch": 4.723404255319149,
"grad_norm": 0.07995230536661296,
"learning_rate": 7.434397797198367e-07,
"loss": 0.3599,
"step": 444
},
{
"epoch": 4.73404255319149,
"grad_norm": 0.08161632016821302,
"learning_rate": 6.875131166508553e-07,
"loss": 0.36,
"step": 445
},
{
"epoch": 4.74468085106383,
"grad_norm": 0.07946253233083479,
"learning_rate": 6.337548941928839e-07,
"loss": 0.3626,
"step": 446
},
{
"epoch": 4.75531914893617,
"grad_norm": 0.07786315052046529,
"learning_rate": 5.821680776000049e-07,
"loss": 0.3619,
"step": 447
},
{
"epoch": 4.76595744680851,
"grad_norm": 0.07728112424595031,
"learning_rate": 5.32755512353571e-07,
"loss": 0.3607,
"step": 448
},
{
"epoch": 4.776595744680851,
"grad_norm": 0.07848088350738662,
"learning_rate": 4.8551992400522e-07,
"loss": 0.361,
"step": 449
},
{
"epoch": 4.787234042553192,
"grad_norm": 0.07839104640568627,
"learning_rate": 4.4046391802655463e-07,
"loss": 0.3597,
"step": 450
},
{
"epoch": 4.797872340425532,
"grad_norm": 0.0757611098723252,
"learning_rate": 3.975899796654137e-07,
"loss": 0.3671,
"step": 451
},
{
"epoch": 4.808510638297872,
"grad_norm": 0.07660526674986573,
"learning_rate": 3.569004738087989e-07,
"loss": 0.3589,
"step": 452
},
{
"epoch": 4.819148936170213,
"grad_norm": 0.07303253050833616,
"learning_rate": 3.183976448524106e-07,
"loss": 0.363,
"step": 453
},
{
"epoch": 4.829787234042553,
"grad_norm": 0.0754318214355068,
"learning_rate": 2.8208361657688474e-07,
"loss": 0.3643,
"step": 454
},
{
"epoch": 4.840425531914894,
"grad_norm": 0.07876372271022951,
"learning_rate": 2.479603920306106e-07,
"loss": 0.3631,
"step": 455
},
{
"epoch": 4.851063829787234,
"grad_norm": 0.07633688369964613,
"learning_rate": 2.1602985341925953e-07,
"loss": 0.3667,
"step": 456
},
{
"epoch": 4.861702127659575,
"grad_norm": 0.07789291049580865,
"learning_rate": 1.8629376200197004e-07,
"loss": 0.3626,
"step": 457
},
{
"epoch": 4.872340425531915,
"grad_norm": 0.07875357482220077,
"learning_rate": 1.5875375799419e-07,
"loss": 0.3634,
"step": 458
},
{
"epoch": 4.882978723404255,
"grad_norm": 0.0786747469968429,
"learning_rate": 1.3341136047719805e-07,
"loss": 0.3643,
"step": 459
},
{
"epoch": 4.8936170212765955,
"grad_norm": 0.07411973494338685,
"learning_rate": 1.1026796731433475e-07,
"loss": 0.3619,
"step": 460
},
{
"epoch": 4.904255319148936,
"grad_norm": 0.07325052452508983,
"learning_rate": 8.932485507387345e-08,
"loss": 0.3538,
"step": 461
},
{
"epoch": 4.914893617021277,
"grad_norm": 0.07867411963780692,
"learning_rate": 7.058317895861866e-08,
"loss": 0.3627,
"step": 462
},
{
"epoch": 4.925531914893617,
"grad_norm": 0.07601171107784903,
"learning_rate": 5.40439727421882e-08,
"loss": 0.3645,
"step": 463
},
{
"epoch": 4.9361702127659575,
"grad_norm": 0.07741148103950671,
"learning_rate": 3.970814871197437e-08,
"loss": 0.3611,
"step": 464
},
{
"epoch": 4.946808510638298,
"grad_norm": 0.07590962518579049,
"learning_rate": 2.75764976188464e-08,
"loss": 0.3594,
"step": 465
},
{
"epoch": 4.957446808510638,
"grad_norm": 0.07705705973848143,
"learning_rate": 1.764968863351424e-08,
"loss": 0.3619,
"step": 466
},
{
"epoch": 4.968085106382979,
"grad_norm": 0.0738876803957471,
"learning_rate": 9.928269309638083e-09,
"loss": 0.3604,
"step": 467
},
{
"epoch": 4.9787234042553195,
"grad_norm": 0.07315503801724814,
"learning_rate": 4.412665553594764e-09,
"loss": 0.3638,
"step": 468
},
{
"epoch": 4.98936170212766,
"grad_norm": 0.07954351258156074,
"learning_rate": 1.103181601020964e-09,
"loss": 0.3628,
"step": 469
},
{
"epoch": 5.0,
"grad_norm": 0.07533000239349877,
"learning_rate": 0.0,
"loss": 0.3558,
"step": 470
},
{
"epoch": 5.0,
"step": 470,
"total_flos": 7884786861342720.0,
"train_loss": 0.4284587557011462,
"train_runtime": 6272.7597,
"train_samples_per_second": 38.218,
"train_steps_per_second": 0.075
}
],
"logging_steps": 1,
"max_steps": 470,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 7884786861342720.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}