de25a3f955
Model: fpadovani/tur-latn-100mb-after-ppt-Dp-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": 8.055328845977783,
|
|
"epoch": 0.004972650422675286,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 2e-06,
|
|
"loss": 7.7699,
|
|
"mean_token_accuracy": 0.09289914444088936,
|
|
"num_tokens": 9089.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 7.9934410572052,
|
|
"epoch": 0.009945300845350571,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 7.9292,
|
|
"mean_token_accuracy": 0.0899883709847927,
|
|
"num_tokens": 21650.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 8.043873596191407,
|
|
"epoch": 0.014917951268025857,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 7e-06,
|
|
"loss": 7.8569,
|
|
"mean_token_accuracy": 0.0829312488436699,
|
|
"num_tokens": 31067.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 7.9948217391967775,
|
|
"epoch": 0.019890601690701143,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 7.8229,
|
|
"mean_token_accuracy": 0.08587435036897659,
|
|
"num_tokens": 40816.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 8.020876693725587,
|
|
"epoch": 0.02486325211337643,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 7.8415,
|
|
"mean_token_accuracy": 0.08777489773929119,
|
|
"num_tokens": 51065.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 8.016945314407348,
|
|
"epoch": 0.029835902536051714,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 7.8082,
|
|
"mean_token_accuracy": 0.09032100439071655,
|
|
"num_tokens": 60539.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 8.016125869750976,
|
|
"epoch": 0.034808552958727,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 7.8391,
|
|
"mean_token_accuracy": 0.08789670392870903,
|
|
"num_tokens": 71010.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 8.02649121284485,
|
|
"epoch": 0.039781203381402286,
|
|
"grad_norm": 0.98828125,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 7.9238,
|
|
"mean_token_accuracy": 0.08401809930801392,
|
|
"num_tokens": 81394.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 8.087946605682372,
|
|
"epoch": 0.04475385380407757,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 7.9606,
|
|
"mean_token_accuracy": 0.08528241962194442,
|
|
"num_tokens": 92491.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 8.078335189819336,
|
|
"epoch": 0.04972650422675286,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 7.788,
|
|
"mean_token_accuracy": 0.09470873698592186,
|
|
"num_tokens": 102113.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 8.062228298187256,
|
|
"epoch": 0.05469915464942814,
|
|
"grad_norm": 1.8125,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 7.9462,
|
|
"mean_token_accuracy": 0.08278393298387528,
|
|
"num_tokens": 112211.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 8.023606491088866,
|
|
"epoch": 0.05967180507210343,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 7.7747,
|
|
"mean_token_accuracy": 0.0901342697441578,
|
|
"num_tokens": 122166.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 8.003111219406128,
|
|
"epoch": 0.06464445549477872,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 7.7581,
|
|
"mean_token_accuracy": 0.09486681148409844,
|
|
"num_tokens": 131016.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 7.973823165893554,
|
|
"epoch": 0.069617105917454,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 7.8401,
|
|
"mean_token_accuracy": 0.09549511298537254,
|
|
"num_tokens": 141369.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 8.026473474502563,
|
|
"epoch": 0.07458975634012929,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 7.7438,
|
|
"mean_token_accuracy": 0.09063734859228134,
|
|
"num_tokens": 150939.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 8.038272190093995,
|
|
"epoch": 0.07956240676280457,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 7.7447,
|
|
"mean_token_accuracy": 0.09616999849677085,
|
|
"num_tokens": 159981.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 7.98382682800293,
|
|
"epoch": 0.08453505718547986,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 7.7952,
|
|
"mean_token_accuracy": 0.08998179510235786,
|
|
"num_tokens": 170139.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 8.070371198654176,
|
|
"epoch": 0.08950770760815514,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 7.7446,
|
|
"mean_token_accuracy": 0.09454798772931099,
|
|
"num_tokens": 180746.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 8.05376615524292,
|
|
"epoch": 0.09448035803083044,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 7.6953,
|
|
"mean_token_accuracy": 0.09375635012984276,
|
|
"num_tokens": 190250.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 7.960572814941406,
|
|
"epoch": 0.09945300845350571,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 7.851,
|
|
"mean_token_accuracy": 0.08996001482009888,
|
|
"num_tokens": 201185.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 8.003534412384033,
|
|
"epoch": 0.10442565887618101,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 7.8286,
|
|
"mean_token_accuracy": 0.08292106837034226,
|
|
"num_tokens": 211476.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 8.028951740264892,
|
|
"epoch": 0.10939830929885629,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 7.7808,
|
|
"mean_token_accuracy": 0.08876777328550815,
|
|
"num_tokens": 220965.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 8.002874565124511,
|
|
"epoch": 0.11437095972153158,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 7.8037,
|
|
"mean_token_accuracy": 0.08823544830083847,
|
|
"num_tokens": 231539.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 8.01924533843994,
|
|
"epoch": 0.11934361014420686,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 7.7812,
|
|
"mean_token_accuracy": 0.08827254623174667,
|
|
"num_tokens": 241206.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 7.982138299942017,
|
|
"epoch": 0.12431626056688215,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 7.7187,
|
|
"mean_token_accuracy": 0.09137163013219833,
|
|
"num_tokens": 250629.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 8.054266786575317,
|
|
"epoch": 0.12928891098955744,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 7.8412,
|
|
"mean_token_accuracy": 0.09213275685906411,
|
|
"num_tokens": 260705.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 8.1447190284729,
|
|
"epoch": 0.13426156141223272,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 7.7988,
|
|
"mean_token_accuracy": 0.09144297614693642,
|
|
"num_tokens": 270499.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 8.07759428024292,
|
|
"epoch": 0.139234211834908,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 7.8214,
|
|
"mean_token_accuracy": 0.08136414065957069,
|
|
"num_tokens": 279998.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 7.998452997207641,
|
|
"epoch": 0.14420686225758328,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 7.7959,
|
|
"mean_token_accuracy": 0.09346884936094284,
|
|
"num_tokens": 290322.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 8.026957654953003,
|
|
"epoch": 0.14917951268025859,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.741,
|
|
"mean_token_accuracy": 0.09033143669366836,
|
|
"num_tokens": 299893.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 7.940546894073487,
|
|
"epoch": 0.15415216310293386,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.7927,
|
|
"mean_token_accuracy": 0.09222338274121285,
|
|
"num_tokens": 309191.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 8.115492153167725,
|
|
"epoch": 0.15912481352560914,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.7385,
|
|
"mean_token_accuracy": 0.09139059036970139,
|
|
"num_tokens": 318494.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 7.960078859329224,
|
|
"epoch": 0.16409746394828442,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.7788,
|
|
"mean_token_accuracy": 0.0923788644373417,
|
|
"num_tokens": 328299.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 8.144480419158935,
|
|
"epoch": 0.16907011437095973,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.8496,
|
|
"mean_token_accuracy": 0.08552220612764358,
|
|
"num_tokens": 338878.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.092591524124146,
|
|
"epoch": 0.174042764793635,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.7971,
|
|
"mean_token_accuracy": 0.08752194195985794,
|
|
"num_tokens": 348508.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 7.965788459777832,
|
|
"epoch": 0.1790154152163103,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.7449,
|
|
"mean_token_accuracy": 0.08802809938788414,
|
|
"num_tokens": 358233.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.010887241363525,
|
|
"epoch": 0.1839880656389856,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.7062,
|
|
"mean_token_accuracy": 0.09302832186222076,
|
|
"num_tokens": 367297.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.047781038284302,
|
|
"epoch": 0.18896071606166087,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.7319,
|
|
"mean_token_accuracy": 0.08684638142585754,
|
|
"num_tokens": 376880.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.015840673446656,
|
|
"epoch": 0.19393336648433615,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.7585,
|
|
"mean_token_accuracy": 0.09129639342427254,
|
|
"num_tokens": 386806.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 7.969292211532593,
|
|
"epoch": 0.19890601690701143,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.7064,
|
|
"mean_token_accuracy": 0.0939807079732418,
|
|
"num_tokens": 396161.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.018733930587768,
|
|
"epoch": 0.20387866732968674,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.7352,
|
|
"mean_token_accuracy": 0.09067206978797912,
|
|
"num_tokens": 405643.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.042714643478394,
|
|
"epoch": 0.20885131775236201,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.7571,
|
|
"mean_token_accuracy": 0.09590560421347619,
|
|
"num_tokens": 414367.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.020010566711425,
|
|
"epoch": 0.2138239681750373,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.7188,
|
|
"mean_token_accuracy": 0.09117217287421227,
|
|
"num_tokens": 424430.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.063075876235962,
|
|
"epoch": 0.21879661859771257,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.7448,
|
|
"mean_token_accuracy": 0.09226949140429497,
|
|
"num_tokens": 435176.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.007707548141479,
|
|
"epoch": 0.22376926902038788,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.8191,
|
|
"mean_token_accuracy": 0.09036103039979934,
|
|
"num_tokens": 445611.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.024854516983032,
|
|
"epoch": 0.22874191944306316,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.732,
|
|
"mean_token_accuracy": 0.09320453926920891,
|
|
"num_tokens": 455143.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.067090892791748,
|
|
"epoch": 0.23371456986573844,
|
|
"grad_norm": 1.6953125,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.7643,
|
|
"mean_token_accuracy": 0.09478386417031288,
|
|
"num_tokens": 465507.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.055903482437134,
|
|
"epoch": 0.23868722028841372,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.8112,
|
|
"mean_token_accuracy": 0.08999223560094834,
|
|
"num_tokens": 476152.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.923675012588501,
|
|
"epoch": 0.24365987071108902,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.6274,
|
|
"mean_token_accuracy": 0.09889271780848503,
|
|
"num_tokens": 486341.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.941055393218994,
|
|
"epoch": 0.2486325211337643,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.5837,
|
|
"mean_token_accuracy": 0.09929879531264305,
|
|
"num_tokens": 495915.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.947218942642212,
|
|
"epoch": 0.2536051715564396,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.7126,
|
|
"mean_token_accuracy": 0.09258004277944565,
|
|
"num_tokens": 505666.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.92978081703186,
|
|
"epoch": 0.2585778219791149,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.5981,
|
|
"mean_token_accuracy": 0.09562918171286583,
|
|
"num_tokens": 515411.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.961107540130615,
|
|
"epoch": 0.26355047240179014,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.7116,
|
|
"mean_token_accuracy": 0.09054910317063332,
|
|
"num_tokens": 525018.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.943254613876343,
|
|
"epoch": 0.26852312282446544,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.6326,
|
|
"mean_token_accuracy": 0.09534763172268867,
|
|
"num_tokens": 535056.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.991715574264527,
|
|
"epoch": 0.27349577324714075,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.6291,
|
|
"mean_token_accuracy": 0.09336584955453872,
|
|
"num_tokens": 545115.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 8.080969667434692,
|
|
"epoch": 0.278468423669816,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.6495,
|
|
"mean_token_accuracy": 0.09909325316548348,
|
|
"num_tokens": 554429.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.849652528762817,
|
|
"epoch": 0.2834410740924913,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.5983,
|
|
"mean_token_accuracy": 0.10442101955413818,
|
|
"num_tokens": 563357.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.9693444728851315,
|
|
"epoch": 0.28841372451516656,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.6591,
|
|
"mean_token_accuracy": 0.09824569448828697,
|
|
"num_tokens": 574480.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.956569957733154,
|
|
"epoch": 0.29338637493784187,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.6345,
|
|
"mean_token_accuracy": 0.0936326451599598,
|
|
"num_tokens": 583722.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.912858390808106,
|
|
"epoch": 0.29835902536051717,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.5976,
|
|
"mean_token_accuracy": 0.10199215114116669,
|
|
"num_tokens": 593644.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.891701698303223,
|
|
"epoch": 0.3033316757831924,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.6595,
|
|
"mean_token_accuracy": 0.09747006073594093,
|
|
"num_tokens": 603547.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.942541217803955,
|
|
"epoch": 0.30830432620586773,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.6749,
|
|
"mean_token_accuracy": 0.09077658727765084,
|
|
"num_tokens": 613880.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 8.01622257232666,
|
|
"epoch": 0.31327697662854304,
|
|
"grad_norm": 1.5078125,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.6791,
|
|
"mean_token_accuracy": 0.09485002011060714,
|
|
"num_tokens": 624213.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.940730810165405,
|
|
"epoch": 0.3182496270512183,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.7188,
|
|
"mean_token_accuracy": 0.09535183757543564,
|
|
"num_tokens": 634350.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 8.01810598373413,
|
|
"epoch": 0.3232222774738936,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.6764,
|
|
"mean_token_accuracy": 0.09407038241624832,
|
|
"num_tokens": 644040.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.90651569366455,
|
|
"epoch": 0.32819492789656884,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.6464,
|
|
"mean_token_accuracy": 0.09886646866798401,
|
|
"num_tokens": 652915.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.997863626480102,
|
|
"epoch": 0.33316757831924415,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.663,
|
|
"mean_token_accuracy": 0.09337838441133499,
|
|
"num_tokens": 662665.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.880599927902222,
|
|
"epoch": 0.33814022874191946,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.6934,
|
|
"mean_token_accuracy": 0.09364098682999611,
|
|
"num_tokens": 671667.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 8.07152271270752,
|
|
"epoch": 0.3431128791645947,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.6974,
|
|
"mean_token_accuracy": 0.08758332729339599,
|
|
"num_tokens": 680432.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.930040788650513,
|
|
"epoch": 0.34808552958727,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.7059,
|
|
"mean_token_accuracy": 0.09117975533008575,
|
|
"num_tokens": 690184.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.958481740951538,
|
|
"epoch": 0.3530581800099453,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.5809,
|
|
"mean_token_accuracy": 0.09757120013237,
|
|
"num_tokens": 699147.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.778140354156494,
|
|
"epoch": 0.3580308304326206,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.6241,
|
|
"mean_token_accuracy": 0.09795154929161072,
|
|
"num_tokens": 708696.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.94059271812439,
|
|
"epoch": 0.3630034808552959,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000182,
|
|
"loss": 7.5868,
|
|
"mean_token_accuracy": 0.09651872366666794,
|
|
"num_tokens": 718438.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.958788824081421,
|
|
"epoch": 0.3679761312779712,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 7.7347,
|
|
"mean_token_accuracy": 0.08726413100957871,
|
|
"num_tokens": 728157.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.920962810516357,
|
|
"epoch": 0.37294878170064644,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.000187,
|
|
"loss": 7.6345,
|
|
"mean_token_accuracy": 0.09856286793947219,
|
|
"num_tokens": 737007.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.974191665649414,
|
|
"epoch": 0.37792143212332174,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 7.5487,
|
|
"mean_token_accuracy": 0.1027902714908123,
|
|
"num_tokens": 746873.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.777252101898194,
|
|
"epoch": 0.382894082545997,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.4764,
|
|
"mean_token_accuracy": 0.10671861693263054,
|
|
"num_tokens": 755486.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.895697546005249,
|
|
"epoch": 0.3878667329686723,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 7.6079,
|
|
"mean_token_accuracy": 0.09227629154920577,
|
|
"num_tokens": 765481.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.837667560577392,
|
|
"epoch": 0.3928393833913476,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 7.5488,
|
|
"mean_token_accuracy": 0.09985539466142654,
|
|
"num_tokens": 774335.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 8.047065830230713,
|
|
"epoch": 0.39781203381402286,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 7.6338,
|
|
"mean_token_accuracy": 0.09441028982400894,
|
|
"num_tokens": 784307.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.9048755168914795,
|
|
"epoch": 0.40278468423669817,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.000202,
|
|
"loss": 7.6761,
|
|
"mean_token_accuracy": 0.09370537623763084,
|
|
"num_tokens": 795058.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 8.047796821594238,
|
|
"epoch": 0.40775733465937347,
|
|
"grad_norm": 1.8515625,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 7.6291,
|
|
"mean_token_accuracy": 0.09702575877308846,
|
|
"num_tokens": 805687.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.804214286804199,
|
|
"epoch": 0.4127299850820487,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.000207,
|
|
"loss": 7.6174,
|
|
"mean_token_accuracy": 0.09554455727338791,
|
|
"num_tokens": 815138.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.951880931854248,
|
|
"epoch": 0.41770263550472403,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 7.5742,
|
|
"mean_token_accuracy": 0.09677162915468215,
|
|
"num_tokens": 825009.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.795875835418701,
|
|
"epoch": 0.4226752859273993,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.000212,
|
|
"loss": 7.5637,
|
|
"mean_token_accuracy": 0.09957250207662582,
|
|
"num_tokens": 834708.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.925895738601684,
|
|
"epoch": 0.4276479363500746,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 7.6117,
|
|
"mean_token_accuracy": 0.0967609629034996,
|
|
"num_tokens": 843888.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.853721952438354,
|
|
"epoch": 0.4326205867727499,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 7.6049,
|
|
"mean_token_accuracy": 0.09610669240355492,
|
|
"num_tokens": 853057.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.923298740386963,
|
|
"epoch": 0.43759323719542514,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 7.6546,
|
|
"mean_token_accuracy": 0.09225371703505517,
|
|
"num_tokens": 862701.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.815232610702514,
|
|
"epoch": 0.44256588761810045,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.000222,
|
|
"loss": 7.549,
|
|
"mean_token_accuracy": 0.10032494440674782,
|
|
"num_tokens": 872129.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.797481489181519,
|
|
"epoch": 0.44753853804077576,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 7.5867,
|
|
"mean_token_accuracy": 0.0956971988081932,
|
|
"num_tokens": 883166.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.8453751564025875,
|
|
"epoch": 0.452511188463451,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 7.4193,
|
|
"mean_token_accuracy": 0.10789886862039566,
|
|
"num_tokens": 894159.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.877769851684571,
|
|
"epoch": 0.4574838388861263,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 7.5812,
|
|
"mean_token_accuracy": 0.09706785902380943,
|
|
"num_tokens": 904588.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.8562257289886475,
|
|
"epoch": 0.46245648930880157,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 7.5406,
|
|
"mean_token_accuracy": 0.09594281017780304,
|
|
"num_tokens": 915554.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.823633670806885,
|
|
"epoch": 0.4674291397314769,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 7.593,
|
|
"mean_token_accuracy": 0.09437503591179848,
|
|
"num_tokens": 925546.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.860237646102905,
|
|
"epoch": 0.4724017901541522,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.000237,
|
|
"loss": 7.5458,
|
|
"mean_token_accuracy": 0.09933575987815857,
|
|
"num_tokens": 936615.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.841644430160523,
|
|
"epoch": 0.47737444057682743,
|
|
"grad_norm": 1.5234375,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 7.5854,
|
|
"mean_token_accuracy": 0.09283004775643348,
|
|
"num_tokens": 947374.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.803491544723511,
|
|
"epoch": 0.48234709099950274,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.000242,
|
|
"loss": 7.5083,
|
|
"mean_token_accuracy": 0.09690938144922256,
|
|
"num_tokens": 957772.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.840833377838135,
|
|
"epoch": 0.48731974142217804,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 7.6166,
|
|
"mean_token_accuracy": 0.09776200726628304,
|
|
"num_tokens": 967724.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.732542610168457,
|
|
"epoch": 0.4922923918448533,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.000247,
|
|
"loss": 7.5253,
|
|
"mean_token_accuracy": 0.0953903116285801,
|
|
"num_tokens": 976497.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.777756214141846,
|
|
"epoch": 0.4972650422675286,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 7.5113,
|
|
"mean_token_accuracy": 0.09828274846076965,
|
|
"num_tokens": 986207.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.4972650422675286,
|
|
"eval_entropy": 7.791762825530795,
|
|
"eval_loss": 7.572336673736572,
|
|
"eval_mean_token_accuracy": 0.09290712282559427,
|
|
"eval_num_tokens": 986207.0,
|
|
"eval_runtime": 10.6134,
|
|
"eval_samples_per_second": 326.475,
|
|
"eval_steps_per_second": 40.892,
|
|
"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
|
|
}
|