28185ab911
Model: fpadovani/tur-latn-100mb-after-ppt-shuff-dyck-10mb-ckpt500_seed3407 Source: Original Platform
1046 lines
28 KiB
JSON
1046 lines
28 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.4972650422675286,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 4.695967006683349,
|
|
"epoch": 0.004972650422675286,
|
|
"grad_norm": 14.4375,
|
|
"learning_rate": 2e-06,
|
|
"loss": 14.4267,
|
|
"mean_token_accuracy": 0.00011848341673612595,
|
|
"num_tokens": 9089.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 4.677842998504639,
|
|
"epoch": 0.009945300845350571,
|
|
"grad_norm": 15.3125,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 14.5334,
|
|
"mean_token_accuracy": 0.00015810276381671428,
|
|
"num_tokens": 21650.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 4.76645073890686,
|
|
"epoch": 0.014917951268025857,
|
|
"grad_norm": 19.25,
|
|
"learning_rate": 7e-06,
|
|
"loss": 14.113,
|
|
"mean_token_accuracy": 0.00020675990963354708,
|
|
"num_tokens": 31067.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 5.015227270126343,
|
|
"epoch": 0.019890601690701143,
|
|
"grad_norm": 34.25,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 13.4863,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 40816.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 7.510676717758178,
|
|
"epoch": 0.02486325211337643,
|
|
"grad_norm": 13.875,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 11.7346,
|
|
"mean_token_accuracy": 9.78473573923111e-05,
|
|
"num_tokens": 51065.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.519583129882813,
|
|
"epoch": 0.029835902536051714,
|
|
"grad_norm": 3.4375,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.7082,
|
|
"mean_token_accuracy": 0.005600619851611555,
|
|
"num_tokens": 60539.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.700261497497559,
|
|
"epoch": 0.034808552958727,
|
|
"grad_norm": 2.640625,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.5537,
|
|
"mean_token_accuracy": 0.028791588079184294,
|
|
"num_tokens": 71010.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.671732521057129,
|
|
"epoch": 0.039781203381402286,
|
|
"grad_norm": 2.265625,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.2564,
|
|
"mean_token_accuracy": 0.04515027813613415,
|
|
"num_tokens": 81394.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.483489990234375,
|
|
"epoch": 0.04475385380407757,
|
|
"grad_norm": 1.8046875,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.1167,
|
|
"mean_token_accuracy": 0.04342293459922075,
|
|
"num_tokens": 92491.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.399119567871093,
|
|
"epoch": 0.04972650422675286,
|
|
"grad_norm": 1.7109375,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.9358,
|
|
"mean_token_accuracy": 0.043518274836242196,
|
|
"num_tokens": 102113.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.48739938735962,
|
|
"epoch": 0.05469915464942814,
|
|
"grad_norm": 2.1875,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.8944,
|
|
"mean_token_accuracy": 0.04352726638317108,
|
|
"num_tokens": 112211.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.497431468963622,
|
|
"epoch": 0.05967180507210343,
|
|
"grad_norm": 1.5390625,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.7893,
|
|
"mean_token_accuracy": 0.040843716263771056,
|
|
"num_tokens": 122166.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.438471221923828,
|
|
"epoch": 0.06464445549477872,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.6965,
|
|
"mean_token_accuracy": 0.04762680903077125,
|
|
"num_tokens": 131016.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.383831214904784,
|
|
"epoch": 0.069617105917454,
|
|
"grad_norm": 1.515625,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.6514,
|
|
"mean_token_accuracy": 0.04989047236740589,
|
|
"num_tokens": 141369.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.395554351806641,
|
|
"epoch": 0.07458975634012929,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.6059,
|
|
"mean_token_accuracy": 0.045283081382513045,
|
|
"num_tokens": 150939.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.431264877319336,
|
|
"epoch": 0.07956240676280457,
|
|
"grad_norm": 1.5078125,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.5567,
|
|
"mean_token_accuracy": 0.05196296907961369,
|
|
"num_tokens": 159981.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.404547595977784,
|
|
"epoch": 0.08453505718547986,
|
|
"grad_norm": 1.4296875,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.4901,
|
|
"mean_token_accuracy": 0.05384046658873558,
|
|
"num_tokens": 170139.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.323073673248292,
|
|
"epoch": 0.08950770760815514,
|
|
"grad_norm": 1.578125,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.4153,
|
|
"mean_token_accuracy": 0.05128877349197865,
|
|
"num_tokens": 180746.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.234126758575439,
|
|
"epoch": 0.09448035803083044,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.2695,
|
|
"mean_token_accuracy": 0.057958731055259706,
|
|
"num_tokens": 190250.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.200442790985107,
|
|
"epoch": 0.09945300845350571,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.3164,
|
|
"mean_token_accuracy": 0.061629551276564595,
|
|
"num_tokens": 201185.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.167423057556153,
|
|
"epoch": 0.10442565887618101,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.2292,
|
|
"mean_token_accuracy": 0.057263224571943286,
|
|
"num_tokens": 211476.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.147284317016602,
|
|
"epoch": 0.10939830929885629,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.1279,
|
|
"mean_token_accuracy": 0.054998861998319624,
|
|
"num_tokens": 220965.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.161615943908691,
|
|
"epoch": 0.11437095972153158,
|
|
"grad_norm": 1.4453125,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 9.0647,
|
|
"mean_token_accuracy": 0.04764294326305389,
|
|
"num_tokens": 231539.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.009593105316162,
|
|
"epoch": 0.11934361014420686,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.9565,
|
|
"mean_token_accuracy": 0.0596030842512846,
|
|
"num_tokens": 241206.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.899549961090088,
|
|
"epoch": 0.12431626056688215,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.8673,
|
|
"mean_token_accuracy": 0.056177464127540586,
|
|
"num_tokens": 250629.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.890707683563232,
|
|
"epoch": 0.12928891098955744,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.8412,
|
|
"mean_token_accuracy": 0.05115443356335163,
|
|
"num_tokens": 260705.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.684529113769532,
|
|
"epoch": 0.13426156141223272,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.7676,
|
|
"mean_token_accuracy": 0.05398021526634693,
|
|
"num_tokens": 270499.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 9.696455478668213,
|
|
"epoch": 0.139234211834908,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.7417,
|
|
"mean_token_accuracy": 0.05484265871345997,
|
|
"num_tokens": 279998.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 9.45594367980957,
|
|
"epoch": 0.14420686225758328,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.6184,
|
|
"mean_token_accuracy": 0.058821605145931245,
|
|
"num_tokens": 290322.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 9.307820510864257,
|
|
"epoch": 0.14917951268025859,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.5725,
|
|
"mean_token_accuracy": 0.06243101172149181,
|
|
"num_tokens": 299893.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.206488990783692,
|
|
"epoch": 0.15415216310293386,
|
|
"grad_norm": 0.9140625,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.5498,
|
|
"mean_token_accuracy": 0.06206914372742176,
|
|
"num_tokens": 309191.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.06636447906494,
|
|
"epoch": 0.15912481352560914,
|
|
"grad_norm": 0.88671875,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 8.4815,
|
|
"mean_token_accuracy": 0.06140335313975811,
|
|
"num_tokens": 318494.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.044106483459473,
|
|
"epoch": 0.16409746394828442,
|
|
"grad_norm": 0.90234375,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 8.4775,
|
|
"mean_token_accuracy": 0.0596194114536047,
|
|
"num_tokens": 328299.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 8.939890956878662,
|
|
"epoch": 0.16907011437095973,
|
|
"grad_norm": 0.81640625,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 8.5442,
|
|
"mean_token_accuracy": 0.057165585830807684,
|
|
"num_tokens": 338878.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.874685859680175,
|
|
"epoch": 0.174042764793635,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 8.468,
|
|
"mean_token_accuracy": 0.05600897297263145,
|
|
"num_tokens": 348508.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.797854232788087,
|
|
"epoch": 0.1790154152163103,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 8.4283,
|
|
"mean_token_accuracy": 0.06139532402157784,
|
|
"num_tokens": 358233.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.801108264923096,
|
|
"epoch": 0.1839880656389856,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 8.3868,
|
|
"mean_token_accuracy": 0.06326259039342404,
|
|
"num_tokens": 367297.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.760680103302002,
|
|
"epoch": 0.18896071606166087,
|
|
"grad_norm": 0.87109375,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 8.3884,
|
|
"mean_token_accuracy": 0.058847269788384435,
|
|
"num_tokens": 376880.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.77523078918457,
|
|
"epoch": 0.19393336648433615,
|
|
"grad_norm": 0.81640625,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 8.3879,
|
|
"mean_token_accuracy": 0.06135675199329853,
|
|
"num_tokens": 386806.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.665618419647217,
|
|
"epoch": 0.19890601690701143,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 8.349,
|
|
"mean_token_accuracy": 0.06269419603049756,
|
|
"num_tokens": 396161.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.676361560821533,
|
|
"epoch": 0.20387866732968674,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000102,
|
|
"loss": 8.3986,
|
|
"mean_token_accuracy": 0.06113377064466476,
|
|
"num_tokens": 405643.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.612883186340332,
|
|
"epoch": 0.20885131775236201,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 8.3532,
|
|
"mean_token_accuracy": 0.06992297060787678,
|
|
"num_tokens": 414367.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.678384780883789,
|
|
"epoch": 0.2138239681750373,
|
|
"grad_norm": 0.90234375,
|
|
"learning_rate": 0.000107,
|
|
"loss": 8.4084,
|
|
"mean_token_accuracy": 0.06210144981741905,
|
|
"num_tokens": 424430.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.648359966278075,
|
|
"epoch": 0.21879661859771257,
|
|
"grad_norm": 0.98828125,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 8.3776,
|
|
"mean_token_accuracy": 0.06364397667348384,
|
|
"num_tokens": 435176.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.576762390136718,
|
|
"epoch": 0.22376926902038788,
|
|
"grad_norm": 0.86328125,
|
|
"learning_rate": 0.000112,
|
|
"loss": 8.3963,
|
|
"mean_token_accuracy": 0.06432987973093987,
|
|
"num_tokens": 445611.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.57848825454712,
|
|
"epoch": 0.22874191944306316,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 8.3247,
|
|
"mean_token_accuracy": 0.0661028627306223,
|
|
"num_tokens": 455143.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.5846848487854,
|
|
"epoch": 0.23371456986573844,
|
|
"grad_norm": 1.7421875,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 8.4127,
|
|
"mean_token_accuracy": 0.06825519613921642,
|
|
"num_tokens": 465507.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.642884063720704,
|
|
"epoch": 0.23868722028841372,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 8.413,
|
|
"mean_token_accuracy": 0.06387550197541714,
|
|
"num_tokens": 476152.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.646145343780518,
|
|
"epoch": 0.24365987071108902,
|
|
"grad_norm": 0.92578125,
|
|
"learning_rate": 0.000122,
|
|
"loss": 8.2597,
|
|
"mean_token_accuracy": 0.06847746260464191,
|
|
"num_tokens": 486341.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.544342041015625,
|
|
"epoch": 0.2486325211337643,
|
|
"grad_norm": 0.84375,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 8.218,
|
|
"mean_token_accuracy": 0.06961140334606171,
|
|
"num_tokens": 495915.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.424121761322022,
|
|
"epoch": 0.2536051715564396,
|
|
"grad_norm": 0.84375,
|
|
"learning_rate": 0.000127,
|
|
"loss": 8.2949,
|
|
"mean_token_accuracy": 0.07351046912372113,
|
|
"num_tokens": 505666.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 8.567151832580567,
|
|
"epoch": 0.2585778219791149,
|
|
"grad_norm": 0.9609375,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 8.2218,
|
|
"mean_token_accuracy": 0.06917047686874866,
|
|
"num_tokens": 515411.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 8.503114604949952,
|
|
"epoch": 0.26355047240179014,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000132,
|
|
"loss": 8.2975,
|
|
"mean_token_accuracy": 0.06331334635615349,
|
|
"num_tokens": 525018.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 8.563315486907959,
|
|
"epoch": 0.26852312282446544,
|
|
"grad_norm": 1.5390625,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 8.255,
|
|
"mean_token_accuracy": 0.06826438158750534,
|
|
"num_tokens": 535056.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 8.556770515441894,
|
|
"epoch": 0.27349577324714075,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 8.2081,
|
|
"mean_token_accuracy": 0.06950257569551468,
|
|
"num_tokens": 545115.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 8.51298542022705,
|
|
"epoch": 0.278468423669816,
|
|
"grad_norm": 2.4375,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 8.2583,
|
|
"mean_token_accuracy": 0.06896053329110145,
|
|
"num_tokens": 554429.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 8.453236293792724,
|
|
"epoch": 0.2834410740924913,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 8.2007,
|
|
"mean_token_accuracy": 0.07584292404353618,
|
|
"num_tokens": 563357.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 8.559890079498292,
|
|
"epoch": 0.28841372451516656,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 8.3728,
|
|
"mean_token_accuracy": 0.06644579172134399,
|
|
"num_tokens": 574480.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 8.503904342651367,
|
|
"epoch": 0.29338637493784187,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 8.2295,
|
|
"mean_token_accuracy": 0.06927415728569031,
|
|
"num_tokens": 583722.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 8.454716300964355,
|
|
"epoch": 0.29835902536051717,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 8.1738,
|
|
"mean_token_accuracy": 0.07493777833878994,
|
|
"num_tokens": 593644.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 8.49923391342163,
|
|
"epoch": 0.3033316757831924,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000152,
|
|
"loss": 8.2147,
|
|
"mean_token_accuracy": 0.07160871513187886,
|
|
"num_tokens": 603547.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 8.44572868347168,
|
|
"epoch": 0.30830432620586773,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 8.2459,
|
|
"mean_token_accuracy": 0.06570485346019268,
|
|
"num_tokens": 613880.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 8.436261844635009,
|
|
"epoch": 0.31327697662854304,
|
|
"grad_norm": 1.609375,
|
|
"learning_rate": 0.000157,
|
|
"loss": 8.2238,
|
|
"mean_token_accuracy": 0.07062076106667518,
|
|
"num_tokens": 624213.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 8.490541934967041,
|
|
"epoch": 0.3182496270512183,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 8.2437,
|
|
"mean_token_accuracy": 0.07502996549010277,
|
|
"num_tokens": 634350.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 8.546083354949952,
|
|
"epoch": 0.3232222774738936,
|
|
"grad_norm": 1.4296875,
|
|
"learning_rate": 0.000162,
|
|
"loss": 8.2438,
|
|
"mean_token_accuracy": 0.07064831554889679,
|
|
"num_tokens": 644040.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 8.492809295654297,
|
|
"epoch": 0.32819492789656884,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 8.2033,
|
|
"mean_token_accuracy": 0.07537714764475822,
|
|
"num_tokens": 652915.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 8.426844501495362,
|
|
"epoch": 0.33316757831924415,
|
|
"grad_norm": 0.859375,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 8.1858,
|
|
"mean_token_accuracy": 0.06924654506146907,
|
|
"num_tokens": 662665.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 8.427862071990967,
|
|
"epoch": 0.33814022874191946,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 8.1955,
|
|
"mean_token_accuracy": 0.07384301200509072,
|
|
"num_tokens": 671667.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 8.440595722198486,
|
|
"epoch": 0.3431128791645947,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 8.2177,
|
|
"mean_token_accuracy": 0.06726174391806125,
|
|
"num_tokens": 680432.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 8.477052116394043,
|
|
"epoch": 0.34808552958727,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 8.2011,
|
|
"mean_token_accuracy": 0.0729417447000742,
|
|
"num_tokens": 690184.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 8.372831344604492,
|
|
"epoch": 0.3530581800099453,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.000177,
|
|
"loss": 8.092,
|
|
"mean_token_accuracy": 0.07851836830377579,
|
|
"num_tokens": 699147.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 8.329792785644532,
|
|
"epoch": 0.3580308304326206,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 8.1201,
|
|
"mean_token_accuracy": 0.0765114389359951,
|
|
"num_tokens": 708696.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 8.382805728912354,
|
|
"epoch": 0.3630034808552959,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000182,
|
|
"loss": 8.0897,
|
|
"mean_token_accuracy": 0.07439367957413197,
|
|
"num_tokens": 718438.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 8.393447971343994,
|
|
"epoch": 0.3679761312779712,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 8.2278,
|
|
"mean_token_accuracy": 0.0699504878371954,
|
|
"num_tokens": 728157.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 8.350840854644776,
|
|
"epoch": 0.37294878170064644,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.000187,
|
|
"loss": 8.1211,
|
|
"mean_token_accuracy": 0.07720480635762214,
|
|
"num_tokens": 737007.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 8.357489013671875,
|
|
"epoch": 0.37792143212332174,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 8.0726,
|
|
"mean_token_accuracy": 0.07758348956704139,
|
|
"num_tokens": 746873.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 8.33244686126709,
|
|
"epoch": 0.382894082545997,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.9852,
|
|
"mean_token_accuracy": 0.08200759738683701,
|
|
"num_tokens": 755486.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 8.285024738311767,
|
|
"epoch": 0.3878667329686723,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 8.0944,
|
|
"mean_token_accuracy": 0.07533938214182853,
|
|
"num_tokens": 765481.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 8.345268440246581,
|
|
"epoch": 0.3928393833913476,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 8.0464,
|
|
"mean_token_accuracy": 0.07712380215525627,
|
|
"num_tokens": 774335.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 8.38193359375,
|
|
"epoch": 0.39781203381402286,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 8.1296,
|
|
"mean_token_accuracy": 0.07453210800886154,
|
|
"num_tokens": 784307.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 8.411716270446778,
|
|
"epoch": 0.40278468423669817,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.000202,
|
|
"loss": 8.152,
|
|
"mean_token_accuracy": 0.07242761701345443,
|
|
"num_tokens": 795058.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 8.320662117004394,
|
|
"epoch": 0.40775733465937347,
|
|
"grad_norm": 1.6953125,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 8.0567,
|
|
"mean_token_accuracy": 0.07517294511198998,
|
|
"num_tokens": 805687.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 8.350330257415772,
|
|
"epoch": 0.4127299850820487,
|
|
"grad_norm": 0.99609375,
|
|
"learning_rate": 0.000207,
|
|
"loss": 8.0842,
|
|
"mean_token_accuracy": 0.07857826426625251,
|
|
"num_tokens": 815138.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 8.33483533859253,
|
|
"epoch": 0.41770263550472403,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 8.0588,
|
|
"mean_token_accuracy": 0.07447804734110833,
|
|
"num_tokens": 825009.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 8.324224948883057,
|
|
"epoch": 0.4226752859273993,
|
|
"grad_norm": 1.96875,
|
|
"learning_rate": 0.000212,
|
|
"loss": 8.0464,
|
|
"mean_token_accuracy": 0.07719542868435383,
|
|
"num_tokens": 834708.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 8.2837308883667,
|
|
"epoch": 0.4276479363500746,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 8.0994,
|
|
"mean_token_accuracy": 0.07782356329262256,
|
|
"num_tokens": 843888.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 8.280590438842774,
|
|
"epoch": 0.4326205867727499,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 8.0652,
|
|
"mean_token_accuracy": 0.07787982597947121,
|
|
"num_tokens": 853057.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 8.313239860534669,
|
|
"epoch": 0.43759323719542514,
|
|
"grad_norm": 1.546875,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 8.1067,
|
|
"mean_token_accuracy": 0.0732115276157856,
|
|
"num_tokens": 862701.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 8.202727890014648,
|
|
"epoch": 0.44256588761810045,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000222,
|
|
"loss": 8.0037,
|
|
"mean_token_accuracy": 0.07995785847306251,
|
|
"num_tokens": 872129.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 8.249781703948974,
|
|
"epoch": 0.44753853804077576,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 8.0233,
|
|
"mean_token_accuracy": 0.0809432752430439,
|
|
"num_tokens": 883166.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 8.261981916427612,
|
|
"epoch": 0.452511188463451,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 7.8287,
|
|
"mean_token_accuracy": 0.09492504075169564,
|
|
"num_tokens": 894159.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 8.228504180908203,
|
|
"epoch": 0.4574838388861263,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 8.0265,
|
|
"mean_token_accuracy": 0.07806817516684532,
|
|
"num_tokens": 904588.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 8.212932109832764,
|
|
"epoch": 0.46245648930880157,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 8.0094,
|
|
"mean_token_accuracy": 0.07935516387224198,
|
|
"num_tokens": 915554.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 8.191864347457885,
|
|
"epoch": 0.4674291397314769,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 8.03,
|
|
"mean_token_accuracy": 0.07978576868772506,
|
|
"num_tokens": 925546.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 8.266866683959961,
|
|
"epoch": 0.4724017901541522,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.000237,
|
|
"loss": 7.9894,
|
|
"mean_token_accuracy": 0.082433320581913,
|
|
"num_tokens": 936615.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 8.207193756103516,
|
|
"epoch": 0.47737444057682743,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 8.023,
|
|
"mean_token_accuracy": 0.08058698140084744,
|
|
"num_tokens": 947374.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 8.204709148406982,
|
|
"epoch": 0.48234709099950274,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000242,
|
|
"loss": 7.955,
|
|
"mean_token_accuracy": 0.08018621355295182,
|
|
"num_tokens": 957772.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 8.176499080657958,
|
|
"epoch": 0.48731974142217804,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 8.0309,
|
|
"mean_token_accuracy": 0.08311204165220261,
|
|
"num_tokens": 967724.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 8.139270210266114,
|
|
"epoch": 0.4922923918448533,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000247,
|
|
"loss": 7.9264,
|
|
"mean_token_accuracy": 0.08439484089612961,
|
|
"num_tokens": 976497.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 8.097719097137452,
|
|
"epoch": 0.4972650422675286,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 7.9124,
|
|
"mean_token_accuracy": 0.0810380220413208,
|
|
"num_tokens": 986207.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.4972650422675286,
|
|
"eval_entropy": 8.243832484917707,
|
|
"eval_loss": 8.042750358581543,
|
|
"eval_mean_token_accuracy": 0.07886241382623117,
|
|
"eval_num_tokens": 986207.0,
|
|
"eval_runtime": 2.7268,
|
|
"eval_samples_per_second": 1270.725,
|
|
"eval_steps_per_second": 159.162,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 10050,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 10,
|
|
"save_steps": 500,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": false
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 1606088632320000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|