1035 lines
27 KiB
JSON
1035 lines
27 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.0053732818931146765,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.742156791687012,
|
|
"epoch": 5.3732818931146765e-05,
|
|
"grad_norm": 4.5625,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.7975,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 12012.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.74215326309204,
|
|
"epoch": 0.00010746563786229353,
|
|
"grad_norm": 5.375,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7849,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 23548.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.742180919647216,
|
|
"epoch": 0.0001611984567934403,
|
|
"grad_norm": 4.96875,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.763,
|
|
"mean_token_accuracy": 0.00020668564829975366,
|
|
"num_tokens": 35237.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.742225646972656,
|
|
"epoch": 0.00021493127572458706,
|
|
"grad_norm": 5.03125,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.7267,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 47773.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.742192363739013,
|
|
"epoch": 0.00026866409465573383,
|
|
"grad_norm": 5.21875,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 10.6027,
|
|
"mean_token_accuracy": 0.0023876232909969984,
|
|
"num_tokens": 59398.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.741855812072753,
|
|
"epoch": 0.0003223969135868806,
|
|
"grad_norm": 4.0625,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.473,
|
|
"mean_token_accuracy": 0.020790282264351846,
|
|
"num_tokens": 70474.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.740978240966797,
|
|
"epoch": 0.00037612973251802736,
|
|
"grad_norm": 3.109375,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.3535,
|
|
"mean_token_accuracy": 0.03107255753129721,
|
|
"num_tokens": 81441.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.739280223846436,
|
|
"epoch": 0.0004298625514491741,
|
|
"grad_norm": 2.53125,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.23,
|
|
"mean_token_accuracy": 0.032152241468429564,
|
|
"num_tokens": 92819.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.736611366271973,
|
|
"epoch": 0.0004835953703803209,
|
|
"grad_norm": 2.484375,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.0679,
|
|
"mean_token_accuracy": 0.0433739360421896,
|
|
"num_tokens": 105565.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.732536792755127,
|
|
"epoch": 0.0005373281893114677,
|
|
"grad_norm": 2.1875,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.9911,
|
|
"mean_token_accuracy": 0.037867228500545025,
|
|
"num_tokens": 116424.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.72894058227539,
|
|
"epoch": 0.0005910610082426144,
|
|
"grad_norm": 1.984375,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.9525,
|
|
"mean_token_accuracy": 0.03848287593573332,
|
|
"num_tokens": 127807.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.727017021179199,
|
|
"epoch": 0.0006447938271737612,
|
|
"grad_norm": 2.109375,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.8619,
|
|
"mean_token_accuracy": 0.03968206029385328,
|
|
"num_tokens": 139177.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.726418113708496,
|
|
"epoch": 0.000698526646104908,
|
|
"grad_norm": 1.9765625,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.8295,
|
|
"mean_token_accuracy": 0.03538134638220072,
|
|
"num_tokens": 151296.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.725299739837647,
|
|
"epoch": 0.0007522594650360547,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.7022,
|
|
"mean_token_accuracy": 0.03797216359525919,
|
|
"num_tokens": 160488.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.723019790649413,
|
|
"epoch": 0.0008059922839672015,
|
|
"grad_norm": 1.984375,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.6965,
|
|
"mean_token_accuracy": 0.03595883902162313,
|
|
"num_tokens": 171627.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.718975067138672,
|
|
"epoch": 0.0008597251028983482,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.6005,
|
|
"mean_token_accuracy": 0.03552283551543951,
|
|
"num_tokens": 182978.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.713298988342284,
|
|
"epoch": 0.000913457921829495,
|
|
"grad_norm": 2.125,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.5237,
|
|
"mean_token_accuracy": 0.03787758443504572,
|
|
"num_tokens": 193362.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.705536365509033,
|
|
"epoch": 0.0009671907407606418,
|
|
"grad_norm": 1.9609375,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.4601,
|
|
"mean_token_accuracy": 0.037437847815454006,
|
|
"num_tokens": 204896.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.695236206054688,
|
|
"epoch": 0.0010209235596917885,
|
|
"grad_norm": 2.03125,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.3647,
|
|
"mean_token_accuracy": 0.03763183522969484,
|
|
"num_tokens": 215828.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.683992767333985,
|
|
"epoch": 0.0010746563786229353,
|
|
"grad_norm": 2.203125,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.2553,
|
|
"mean_token_accuracy": 0.04709151349961758,
|
|
"num_tokens": 226311.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.66820011138916,
|
|
"epoch": 0.001128389197554082,
|
|
"grad_norm": 1.96875,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.1779,
|
|
"mean_token_accuracy": 0.04788613952696323,
|
|
"num_tokens": 237817.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.641796970367432,
|
|
"epoch": 0.0011821220164852288,
|
|
"grad_norm": 1.9375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.0389,
|
|
"mean_token_accuracy": 0.05182615555822849,
|
|
"num_tokens": 248778.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.600088024139405,
|
|
"epoch": 0.0012358548354163756,
|
|
"grad_norm": 1.890625,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.9776,
|
|
"mean_token_accuracy": 0.048598403483629225,
|
|
"num_tokens": 260638.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.544377613067628,
|
|
"epoch": 0.0012895876543475224,
|
|
"grad_norm": 1.671875,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.8365,
|
|
"mean_token_accuracy": 0.049195655062794684,
|
|
"num_tokens": 272223.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.488489818572997,
|
|
"epoch": 0.0013433204732786691,
|
|
"grad_norm": 1.8359375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.6543,
|
|
"mean_token_accuracy": 0.06183730475604534,
|
|
"num_tokens": 282883.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.427958393096924,
|
|
"epoch": 0.001397053292209816,
|
|
"grad_norm": 1.71875,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.5628,
|
|
"mean_token_accuracy": 0.05788363590836525,
|
|
"num_tokens": 293318.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.361266994476319,
|
|
"epoch": 0.0014507861111409627,
|
|
"grad_norm": 1.7109375,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.3999,
|
|
"mean_token_accuracy": 0.06325617916882038,
|
|
"num_tokens": 303563.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.266337871551514,
|
|
"epoch": 0.0015045189300721094,
|
|
"grad_norm": 1.6796875,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.2536,
|
|
"mean_token_accuracy": 0.07024482376873493,
|
|
"num_tokens": 313968.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.139619827270508,
|
|
"epoch": 0.0015582517490032562,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.2142,
|
|
"mean_token_accuracy": 0.06537894494831561,
|
|
"num_tokens": 325261.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 10.008739376068116,
|
|
"epoch": 0.001611984567934403,
|
|
"grad_norm": 1.546875,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.9979,
|
|
"mean_token_accuracy": 0.07281787917017937,
|
|
"num_tokens": 335803.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.818781661987305,
|
|
"epoch": 0.0016657173868655497,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.8841,
|
|
"mean_token_accuracy": 0.0677342727780342,
|
|
"num_tokens": 346582.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.64435920715332,
|
|
"epoch": 0.0017194502057966965,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.8699,
|
|
"mean_token_accuracy": 0.07075726985931396,
|
|
"num_tokens": 357779.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.435050868988037,
|
|
"epoch": 0.0017731830247278433,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.6867,
|
|
"mean_token_accuracy": 0.07693751528859138,
|
|
"num_tokens": 369191.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.177427005767822,
|
|
"epoch": 0.00182691584365899,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.6553,
|
|
"mean_token_accuracy": 0.07154584787786007,
|
|
"num_tokens": 381374.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.965550804138184,
|
|
"epoch": 0.0018806486625901368,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.5673,
|
|
"mean_token_accuracy": 0.07786881327629089,
|
|
"num_tokens": 392152.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.720534420013427,
|
|
"epoch": 0.0019343814815212836,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.5062,
|
|
"mean_token_accuracy": 0.07432210743427277,
|
|
"num_tokens": 404121.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.521537494659423,
|
|
"epoch": 0.0019881143004524303,
|
|
"grad_norm": 0.86328125,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.4599,
|
|
"mean_token_accuracy": 0.0757918406277895,
|
|
"num_tokens": 416144.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.401978397369385,
|
|
"epoch": 0.002041847119383577,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.3406,
|
|
"mean_token_accuracy": 0.08547781854867935,
|
|
"num_tokens": 427141.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.218191432952882,
|
|
"epoch": 0.002095579938314724,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.2438,
|
|
"mean_token_accuracy": 0.07949326783418656,
|
|
"num_tokens": 439113.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.094591236114502,
|
|
"epoch": 0.0021493127572458706,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.3,
|
|
"mean_token_accuracy": 0.08100042939186096,
|
|
"num_tokens": 450595.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.087145900726318,
|
|
"epoch": 0.0022030455761770174,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.3575,
|
|
"mean_token_accuracy": 0.08150793164968491,
|
|
"num_tokens": 462811.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 7.998901224136352,
|
|
"epoch": 0.002256778395108164,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.3188,
|
|
"mean_token_accuracy": 0.08503394201397896,
|
|
"num_tokens": 473629.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.012935066223145,
|
|
"epoch": 0.002310511214039311,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.2985,
|
|
"mean_token_accuracy": 0.08433945029973984,
|
|
"num_tokens": 484731.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 7.911737585067749,
|
|
"epoch": 0.0023642440329704577,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.2521,
|
|
"mean_token_accuracy": 0.08716428875923157,
|
|
"num_tokens": 496309.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.943577146530151,
|
|
"epoch": 0.0024179768519016044,
|
|
"grad_norm": 0.93359375,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.2935,
|
|
"mean_token_accuracy": 0.08446456864476204,
|
|
"num_tokens": 508715.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.8531383037567135,
|
|
"epoch": 0.002471709670832751,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.1316,
|
|
"mean_token_accuracy": 0.0886182926595211,
|
|
"num_tokens": 519791.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.81738133430481,
|
|
"epoch": 0.002525442489763898,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.2519,
|
|
"mean_token_accuracy": 0.08779674172401428,
|
|
"num_tokens": 531897.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.859956073760986,
|
|
"epoch": 0.0025791753086950447,
|
|
"grad_norm": 0.91015625,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.2388,
|
|
"mean_token_accuracy": 0.0916181556880474,
|
|
"num_tokens": 543243.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.7477397441864015,
|
|
"epoch": 0.0026329081276261915,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.0962,
|
|
"mean_token_accuracy": 0.09303439110517502,
|
|
"num_tokens": 554139.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.774733972549439,
|
|
"epoch": 0.0026866409465573383,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.1076,
|
|
"mean_token_accuracy": 0.09551994875073433,
|
|
"num_tokens": 565378.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.715104246139527,
|
|
"epoch": 0.002740373765488485,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.0605,
|
|
"mean_token_accuracy": 0.09414524212479591,
|
|
"num_tokens": 576628.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.768790674209595,
|
|
"epoch": 0.002794106584419632,
|
|
"grad_norm": 0.93359375,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.1575,
|
|
"mean_token_accuracy": 0.09156634658575058,
|
|
"num_tokens": 588760.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.687872219085693,
|
|
"epoch": 0.0028478394033507786,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.0791,
|
|
"mean_token_accuracy": 0.10132888033986091,
|
|
"num_tokens": 600024.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.683035945892334,
|
|
"epoch": 0.0029015722222819253,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.0399,
|
|
"mean_token_accuracy": 0.09685612693428994,
|
|
"num_tokens": 613150.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.63238935470581,
|
|
"epoch": 0.002955305041213072,
|
|
"grad_norm": 0.9296875,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.0179,
|
|
"mean_token_accuracy": 0.09851899445056915,
|
|
"num_tokens": 624762.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.715113353729248,
|
|
"epoch": 0.003009037860144219,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.0639,
|
|
"mean_token_accuracy": 0.09616550207138061,
|
|
"num_tokens": 635761.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.620885419845581,
|
|
"epoch": 0.0030627706790753656,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.0092,
|
|
"mean_token_accuracy": 0.09458725675940513,
|
|
"num_tokens": 647882.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.5718300342559814,
|
|
"epoch": 0.0031165034980065124,
|
|
"grad_norm": 0.92578125,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.0483,
|
|
"mean_token_accuracy": 0.09655040949583053,
|
|
"num_tokens": 659749.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.516621065139771,
|
|
"epoch": 0.003170236316937659,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000147,
|
|
"loss": 6.9872,
|
|
"mean_token_accuracy": 0.0997149519622326,
|
|
"num_tokens": 672002.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.547021913528442,
|
|
"epoch": 0.003223969135868806,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 6.9284,
|
|
"mean_token_accuracy": 0.10656724348664284,
|
|
"num_tokens": 682312.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.456842422485352,
|
|
"epoch": 0.0032777019547999527,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000152,
|
|
"loss": 6.965,
|
|
"mean_token_accuracy": 0.09908917099237442,
|
|
"num_tokens": 692188.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.561665821075439,
|
|
"epoch": 0.0033314347737310995,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 6.9484,
|
|
"mean_token_accuracy": 0.10072302371263504,
|
|
"num_tokens": 703789.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.566508626937866,
|
|
"epoch": 0.0033851675926622462,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.0206,
|
|
"mean_token_accuracy": 0.10419768020510674,
|
|
"num_tokens": 714752.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.47095513343811,
|
|
"epoch": 0.003438900411593393,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 6.9571,
|
|
"mean_token_accuracy": 0.10431606322526932,
|
|
"num_tokens": 726716.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.433886337280273,
|
|
"epoch": 0.0034926332305245398,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.000162,
|
|
"loss": 6.9813,
|
|
"mean_token_accuracy": 0.1021119087934494,
|
|
"num_tokens": 737267.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.5219165802001955,
|
|
"epoch": 0.0035463660494556865,
|
|
"grad_norm": 0.99609375,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.0025,
|
|
"mean_token_accuracy": 0.10279990583658219,
|
|
"num_tokens": 750180.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.45853362083435,
|
|
"epoch": 0.0036000988683868333,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 6.8296,
|
|
"mean_token_accuracy": 0.10508297532796859,
|
|
"num_tokens": 761337.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.402671909332275,
|
|
"epoch": 0.00365383168731798,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 6.8015,
|
|
"mean_token_accuracy": 0.1093854360282421,
|
|
"num_tokens": 771379.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.37822003364563,
|
|
"epoch": 0.003707564506249127,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 6.8888,
|
|
"mean_token_accuracy": 0.10919447690248489,
|
|
"num_tokens": 782508.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.4207008361816404,
|
|
"epoch": 0.0037612973251802736,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 6.9096,
|
|
"mean_token_accuracy": 0.10041842758655548,
|
|
"num_tokens": 795541.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.383993768692017,
|
|
"epoch": 0.0038150301441114204,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.000177,
|
|
"loss": 6.8261,
|
|
"mean_token_accuracy": 0.10757922530174255,
|
|
"num_tokens": 805665.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.3851542472839355,
|
|
"epoch": 0.003868762963042567,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 6.9727,
|
|
"mean_token_accuracy": 0.1046277977526188,
|
|
"num_tokens": 817393.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.4332190990448,
|
|
"epoch": 0.003922495781973714,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.000182,
|
|
"loss": 6.9061,
|
|
"mean_token_accuracy": 0.1035635381937027,
|
|
"num_tokens": 828436.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.383777904510498,
|
|
"epoch": 0.003976228600904861,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 6.7789,
|
|
"mean_token_accuracy": 0.10867122486233712,
|
|
"num_tokens": 839223.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.291258668899536,
|
|
"epoch": 0.004029961419836007,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.000187,
|
|
"loss": 6.7478,
|
|
"mean_token_accuracy": 0.11749055162072182,
|
|
"num_tokens": 848953.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.338685035705566,
|
|
"epoch": 0.004083694238767154,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 6.8829,
|
|
"mean_token_accuracy": 0.10381969884037971,
|
|
"num_tokens": 860349.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.314834403991699,
|
|
"epoch": 0.004137427057698301,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.000192,
|
|
"loss": 6.8047,
|
|
"mean_token_accuracy": 0.10881992653012276,
|
|
"num_tokens": 870993.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.348164367675781,
|
|
"epoch": 0.004191159876629448,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.8523,
|
|
"mean_token_accuracy": 0.10951408222317696,
|
|
"num_tokens": 880950.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.376342344284057,
|
|
"epoch": 0.0042448926955605945,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 6.8578,
|
|
"mean_token_accuracy": 0.11057165712118149,
|
|
"num_tokens": 892583.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.21990008354187,
|
|
"epoch": 0.004298625514491741,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.6711,
|
|
"mean_token_accuracy": 0.1182975672185421,
|
|
"num_tokens": 902191.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.244998121261597,
|
|
"epoch": 0.004352358333422888,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.7288,
|
|
"mean_token_accuracy": 0.11365634202957153,
|
|
"num_tokens": 914274.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.147398805618286,
|
|
"epoch": 0.004406091152354035,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.6684,
|
|
"mean_token_accuracy": 0.10877410843968391,
|
|
"num_tokens": 925253.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.30414662361145,
|
|
"epoch": 0.0044598239712851815,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.8047,
|
|
"mean_token_accuracy": 0.11197239682078361,
|
|
"num_tokens": 935520.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.248277902603149,
|
|
"epoch": 0.004513556790216328,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.7966,
|
|
"mean_token_accuracy": 0.11800143346190453,
|
|
"num_tokens": 947817.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.1241779804229735,
|
|
"epoch": 0.004567289609147475,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.7073,
|
|
"mean_token_accuracy": 0.1153214804828167,
|
|
"num_tokens": 959474.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.291252708435058,
|
|
"epoch": 0.004621022428078622,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.7,
|
|
"mean_token_accuracy": 0.11340035423636437,
|
|
"num_tokens": 971271.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.1269285678863525,
|
|
"epoch": 0.004674755247009769,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.7119,
|
|
"mean_token_accuracy": 0.11621161103248596,
|
|
"num_tokens": 981995.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.202797365188599,
|
|
"epoch": 0.004728488065940915,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.7741,
|
|
"mean_token_accuracy": 0.11647263541817665,
|
|
"num_tokens": 994346.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.202062654495239,
|
|
"epoch": 0.004782220884872062,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.7803,
|
|
"mean_token_accuracy": 0.11048730537295341,
|
|
"num_tokens": 1005552.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.2215375900268555,
|
|
"epoch": 0.004835953703803209,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.7147,
|
|
"mean_token_accuracy": 0.10997768640518188,
|
|
"num_tokens": 1017202.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.02616319656372,
|
|
"epoch": 0.004889686522734356,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.6729,
|
|
"mean_token_accuracy": 0.11669943779706955,
|
|
"num_tokens": 1027450.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.129533100128174,
|
|
"epoch": 0.004943419341665502,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.6599,
|
|
"mean_token_accuracy": 0.11500771567225457,
|
|
"num_tokens": 1038990.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.127964210510254,
|
|
"epoch": 0.004997152160596649,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.6597,
|
|
"mean_token_accuracy": 0.11586256548762322,
|
|
"num_tokens": 1050643.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.037399816513061,
|
|
"epoch": 0.005050884979527796,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.5917,
|
|
"mean_token_accuracy": 0.12084505632519722,
|
|
"num_tokens": 1061182.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.155901622772217,
|
|
"epoch": 0.005104617798458943,
|
|
"grad_norm": 0.875,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.7526,
|
|
"mean_token_accuracy": 0.11448139399290085,
|
|
"num_tokens": 1073074.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.101831865310669,
|
|
"epoch": 0.0051583506173900895,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.7746,
|
|
"mean_token_accuracy": 0.1180665597319603,
|
|
"num_tokens": 1085061.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.091040754318238,
|
|
"epoch": 0.005212083436321236,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.6384,
|
|
"mean_token_accuracy": 0.11752296015620231,
|
|
"num_tokens": 1096148.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.014005517959594,
|
|
"epoch": 0.005265816255252383,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.5661,
|
|
"mean_token_accuracy": 0.11984377726912498,
|
|
"num_tokens": 1106159.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.034162092208862,
|
|
"epoch": 0.00531954907418353,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.7162,
|
|
"mean_token_accuracy": 0.11838811710476875,
|
|
"num_tokens": 1117159.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.145173740386963,
|
|
"epoch": 0.0053732818931146765,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.6168,
|
|
"mean_token_accuracy": 0.12091756090521813,
|
|
"num_tokens": 1129633.0,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 4000,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 1,
|
|
"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": 262024031600640.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|