2035 lines
54 KiB
JSON
2035 lines
54 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.05904233335301411,
|
|
"eval_steps": 500,
|
|
"global_step": 1000,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.69201593399048,
|
|
"epoch": 0.00029521166676507054,
|
|
"grad_norm": 12.6875,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.8334,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 10275.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.691933250427246,
|
|
"epoch": 0.0005904233335301411,
|
|
"grad_norm": 13.1875,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7937,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 22020.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.691065883636474,
|
|
"epoch": 0.0008856350002952116,
|
|
"grad_norm": 9.5625,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.5174,
|
|
"mean_token_accuracy": 0.021099633118137718,
|
|
"num_tokens": 32612.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.68113021850586,
|
|
"epoch": 0.0011808466670602822,
|
|
"grad_norm": 6.21875,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.1848,
|
|
"mean_token_accuracy": 0.04274020157754421,
|
|
"num_tokens": 43270.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.621037483215332,
|
|
"epoch": 0.0014760583338253527,
|
|
"grad_norm": 3.84375,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 9.9202,
|
|
"mean_token_accuracy": 0.04267438128590584,
|
|
"num_tokens": 54178.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.595892238616944,
|
|
"epoch": 0.0017712700005904232,
|
|
"grad_norm": 3.234375,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 9.7559,
|
|
"mean_token_accuracy": 0.043279900401830676,
|
|
"num_tokens": 64390.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.602502059936523,
|
|
"epoch": 0.002066481667355494,
|
|
"grad_norm": 2.921875,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 9.6682,
|
|
"mean_token_accuracy": 0.043131088465452196,
|
|
"num_tokens": 73703.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.59941120147705,
|
|
"epoch": 0.0023616933341205643,
|
|
"grad_norm": 2.6875,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 9.6023,
|
|
"mean_token_accuracy": 0.04174515642225742,
|
|
"num_tokens": 83463.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.594866466522216,
|
|
"epoch": 0.002656905000885635,
|
|
"grad_norm": 2.515625,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.6097,
|
|
"mean_token_accuracy": 0.04398317113518715,
|
|
"num_tokens": 94129.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.567664051055909,
|
|
"epoch": 0.0029521166676507054,
|
|
"grad_norm": 2.421875,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.558,
|
|
"mean_token_accuracy": 0.043352384492754935,
|
|
"num_tokens": 105482.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.569747066497802,
|
|
"epoch": 0.003247328334415776,
|
|
"grad_norm": 2.453125,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.4562,
|
|
"mean_token_accuracy": 0.043072767183184625,
|
|
"num_tokens": 117342.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.565975093841553,
|
|
"epoch": 0.0035425400011808465,
|
|
"grad_norm": 2.25,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.44,
|
|
"mean_token_accuracy": 0.050405914708971974,
|
|
"num_tokens": 127691.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.458717250823975,
|
|
"epoch": 0.0038377516679459172,
|
|
"grad_norm": 2.46875,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.2492,
|
|
"mean_token_accuracy": 0.05396741069853306,
|
|
"num_tokens": 138186.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.420860767364502,
|
|
"epoch": 0.004132963334710988,
|
|
"grad_norm": 2.28125,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.2326,
|
|
"mean_token_accuracy": 0.06198026090860367,
|
|
"num_tokens": 149578.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.452775382995606,
|
|
"epoch": 0.004428175001476058,
|
|
"grad_norm": 2.21875,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.1186,
|
|
"mean_token_accuracy": 0.06664705500006676,
|
|
"num_tokens": 160139.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.454645347595214,
|
|
"epoch": 0.004723386668241129,
|
|
"grad_norm": 2.296875,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.037,
|
|
"mean_token_accuracy": 0.06905781887471676,
|
|
"num_tokens": 170768.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.332771587371827,
|
|
"epoch": 0.0050185983350062,
|
|
"grad_norm": 2.3125,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 8.8423,
|
|
"mean_token_accuracy": 0.07480080313980579,
|
|
"num_tokens": 180702.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.249947929382325,
|
|
"epoch": 0.00531381000177127,
|
|
"grad_norm": 2.328125,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 8.7738,
|
|
"mean_token_accuracy": 0.07588593885302544,
|
|
"num_tokens": 191188.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.206402206420899,
|
|
"epoch": 0.0056090216685363405,
|
|
"grad_norm": 3.609375,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 8.6866,
|
|
"mean_token_accuracy": 0.07205914407968521,
|
|
"num_tokens": 201094.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.155156707763672,
|
|
"epoch": 0.005904233335301411,
|
|
"grad_norm": 3.046875,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 8.5626,
|
|
"mean_token_accuracy": 0.076263627409935,
|
|
"num_tokens": 211718.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.151510620117188,
|
|
"epoch": 0.006199445002066482,
|
|
"grad_norm": 2.25,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 8.4394,
|
|
"mean_token_accuracy": 0.07763982266187668,
|
|
"num_tokens": 221446.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.075346565246582,
|
|
"epoch": 0.006494656668831552,
|
|
"grad_norm": 1.84375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 8.4619,
|
|
"mean_token_accuracy": 0.07559167668223381,
|
|
"num_tokens": 232848.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.015703201293945,
|
|
"epoch": 0.006789868335596623,
|
|
"grad_norm": 1.8046875,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.3798,
|
|
"mean_token_accuracy": 0.0727357666939497,
|
|
"num_tokens": 244008.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.909519577026368,
|
|
"epoch": 0.007085080002361693,
|
|
"grad_norm": 1.6171875,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.1771,
|
|
"mean_token_accuracy": 0.08037759214639664,
|
|
"num_tokens": 256211.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.780295848846436,
|
|
"epoch": 0.007380291669126764,
|
|
"grad_norm": 1.6171875,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.1207,
|
|
"mean_token_accuracy": 0.07622570730745792,
|
|
"num_tokens": 266688.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.564407634735108,
|
|
"epoch": 0.0076755033358918345,
|
|
"grad_norm": 1.4609375,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 7.9156,
|
|
"mean_token_accuracy": 0.07875195294618606,
|
|
"num_tokens": 276930.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.403815078735352,
|
|
"epoch": 0.007970715002656906,
|
|
"grad_norm": 1.5546875,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 7.7559,
|
|
"mean_token_accuracy": 0.08111081309616566,
|
|
"num_tokens": 287508.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 9.18376874923706,
|
|
"epoch": 0.008265926669421976,
|
|
"grad_norm": 1.4609375,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 7.6109,
|
|
"mean_token_accuracy": 0.08215704038739205,
|
|
"num_tokens": 298021.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 8.89451036453247,
|
|
"epoch": 0.008561138336187046,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 7.5483,
|
|
"mean_token_accuracy": 0.0828169234097004,
|
|
"num_tokens": 308301.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 8.718179512023926,
|
|
"epoch": 0.008856350002952117,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.4558,
|
|
"mean_token_accuracy": 0.07895909734070301,
|
|
"num_tokens": 319321.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 8.486823749542236,
|
|
"epoch": 0.009151561669717187,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.5026,
|
|
"mean_token_accuracy": 0.07978610247373581,
|
|
"num_tokens": 330926.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 8.353065967559814,
|
|
"epoch": 0.009446773336482257,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.3196,
|
|
"mean_token_accuracy": 0.08807190880179405,
|
|
"num_tokens": 341133.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 8.217400360107423,
|
|
"epoch": 0.009741985003247328,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.3374,
|
|
"mean_token_accuracy": 0.08286328092217446,
|
|
"num_tokens": 351897.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 8.153849840164185,
|
|
"epoch": 0.0100371966700124,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.3121,
|
|
"mean_token_accuracy": 0.08045563325285912,
|
|
"num_tokens": 362644.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.020972394943238,
|
|
"epoch": 0.01033240833677747,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.3254,
|
|
"mean_token_accuracy": 0.08039250746369361,
|
|
"num_tokens": 373902.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 7.979048109054565,
|
|
"epoch": 0.01062762000354254,
|
|
"grad_norm": 1.453125,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.2138,
|
|
"mean_token_accuracy": 0.09198684617877007,
|
|
"num_tokens": 384828.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 7.852319145202637,
|
|
"epoch": 0.01092283167030761,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.227,
|
|
"mean_token_accuracy": 0.08049636781215667,
|
|
"num_tokens": 395103.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 7.785118961334229,
|
|
"epoch": 0.011218043337072681,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.1629,
|
|
"mean_token_accuracy": 0.09020237326622009,
|
|
"num_tokens": 405283.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 7.803817939758301,
|
|
"epoch": 0.011513255003837751,
|
|
"grad_norm": 1.53125,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.1458,
|
|
"mean_token_accuracy": 0.08684369623661041,
|
|
"num_tokens": 416005.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 7.734343099594116,
|
|
"epoch": 0.011808466670602822,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.1475,
|
|
"mean_token_accuracy": 0.08818956762552262,
|
|
"num_tokens": 426547.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 7.702443599700928,
|
|
"epoch": 0.012103678337367892,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.0801,
|
|
"mean_token_accuracy": 0.09228792935609817,
|
|
"num_tokens": 436262.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 7.663109302520752,
|
|
"epoch": 0.012398890004132964,
|
|
"grad_norm": 1.578125,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.101,
|
|
"mean_token_accuracy": 0.08728625029325485,
|
|
"num_tokens": 446899.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 7.67796368598938,
|
|
"epoch": 0.012694101670898034,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.1407,
|
|
"mean_token_accuracy": 0.08986683264374733,
|
|
"num_tokens": 458224.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 7.675721073150635,
|
|
"epoch": 0.012989313337663105,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.0746,
|
|
"mean_token_accuracy": 0.08677861616015434,
|
|
"num_tokens": 469306.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.5801355838775635,
|
|
"epoch": 0.013284525004428175,
|
|
"grad_norm": 1.6953125,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.0774,
|
|
"mean_token_accuracy": 0.0875390887260437,
|
|
"num_tokens": 478685.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.611648845672607,
|
|
"epoch": 0.013579736671193245,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.0051,
|
|
"mean_token_accuracy": 0.09632268622517585,
|
|
"num_tokens": 489376.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.525375843048096,
|
|
"epoch": 0.013874948337958316,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.0117,
|
|
"mean_token_accuracy": 0.0921477772295475,
|
|
"num_tokens": 500260.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.426083660125732,
|
|
"epoch": 0.014170160004723386,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.0023,
|
|
"mean_token_accuracy": 0.09566814824938774,
|
|
"num_tokens": 511953.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.632671928405761,
|
|
"epoch": 0.014465371671488458,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.1873,
|
|
"mean_token_accuracy": 0.08889849931001663,
|
|
"num_tokens": 523055.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.509737873077393,
|
|
"epoch": 0.014760583338253528,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 6.9947,
|
|
"mean_token_accuracy": 0.09188270270824432,
|
|
"num_tokens": 534337.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.46486268043518,
|
|
"epoch": 0.015055795005018599,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 0.000127,
|
|
"loss": 6.9868,
|
|
"mean_token_accuracy": 0.09533577933907508,
|
|
"num_tokens": 545189.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.538589096069336,
|
|
"epoch": 0.015351006671783669,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.0231,
|
|
"mean_token_accuracy": 0.09476621299982071,
|
|
"num_tokens": 555506.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.462624025344849,
|
|
"epoch": 0.01564621833854874,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.000132,
|
|
"loss": 6.9308,
|
|
"mean_token_accuracy": 0.0939927913248539,
|
|
"num_tokens": 565936.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.461940479278565,
|
|
"epoch": 0.01594143000531381,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.033,
|
|
"mean_token_accuracy": 0.09217713922262191,
|
|
"num_tokens": 575691.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.481535339355469,
|
|
"epoch": 0.01623664167207888,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 6.9921,
|
|
"mean_token_accuracy": 0.09237621873617172,
|
|
"num_tokens": 586170.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.461501359939575,
|
|
"epoch": 0.016531853338843952,
|
|
"grad_norm": 1.453125,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 6.9755,
|
|
"mean_token_accuracy": 0.09221490249037742,
|
|
"num_tokens": 596458.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.365391731262207,
|
|
"epoch": 0.016827065005609022,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 6.9201,
|
|
"mean_token_accuracy": 0.08767134547233582,
|
|
"num_tokens": 606214.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.444003868103027,
|
|
"epoch": 0.017122276672374093,
|
|
"grad_norm": 1.4296875,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 6.9609,
|
|
"mean_token_accuracy": 0.09265968054533005,
|
|
"num_tokens": 617074.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.425234937667847,
|
|
"epoch": 0.017417488339139163,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.000147,
|
|
"loss": 6.998,
|
|
"mean_token_accuracy": 0.08931455463171005,
|
|
"num_tokens": 628057.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.338685989379883,
|
|
"epoch": 0.017712700005904233,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 6.9406,
|
|
"mean_token_accuracy": 0.09267161414027214,
|
|
"num_tokens": 638150.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.352516222000122,
|
|
"epoch": 0.018007911672669304,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.000152,
|
|
"loss": 6.8101,
|
|
"mean_token_accuracy": 0.10107955634593964,
|
|
"num_tokens": 648377.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.305817174911499,
|
|
"epoch": 0.018303123339434374,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 6.8875,
|
|
"mean_token_accuracy": 0.09907981753349304,
|
|
"num_tokens": 658777.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.311208295822143,
|
|
"epoch": 0.018598335006199444,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.000157,
|
|
"loss": 6.9908,
|
|
"mean_token_accuracy": 0.09044492170214653,
|
|
"num_tokens": 669105.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.355552816390992,
|
|
"epoch": 0.018893546672964515,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 6.8608,
|
|
"mean_token_accuracy": 0.09483690708875656,
|
|
"num_tokens": 679294.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.222180414199829,
|
|
"epoch": 0.019188758339729585,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000162,
|
|
"loss": 6.829,
|
|
"mean_token_accuracy": 0.09829319342970848,
|
|
"num_tokens": 690495.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.220251893997192,
|
|
"epoch": 0.019483970006494655,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 6.8432,
|
|
"mean_token_accuracy": 0.09830230325460435,
|
|
"num_tokens": 701002.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.257207536697388,
|
|
"epoch": 0.019779181673259726,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 6.753,
|
|
"mean_token_accuracy": 0.09737947657704353,
|
|
"num_tokens": 711488.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.091268253326416,
|
|
"epoch": 0.0200743933400248,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 6.7863,
|
|
"mean_token_accuracy": 0.09428932666778564,
|
|
"num_tokens": 722637.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.228825855255127,
|
|
"epoch": 0.02036960500678987,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 6.8498,
|
|
"mean_token_accuracy": 0.09802942052483558,
|
|
"num_tokens": 733212.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.253208494186401,
|
|
"epoch": 0.02066481667355494,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 6.7639,
|
|
"mean_token_accuracy": 0.09950084388256072,
|
|
"num_tokens": 743602.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.207983446121216,
|
|
"epoch": 0.02096002834032001,
|
|
"grad_norm": 1.59375,
|
|
"learning_rate": 0.000177,
|
|
"loss": 6.8151,
|
|
"mean_token_accuracy": 0.09879285991191863,
|
|
"num_tokens": 754094.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.099168062210083,
|
|
"epoch": 0.02125524000708508,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 6.7724,
|
|
"mean_token_accuracy": 0.09406139552593232,
|
|
"num_tokens": 763823.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.163098907470703,
|
|
"epoch": 0.02155045167385015,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.000182,
|
|
"loss": 6.7096,
|
|
"mean_token_accuracy": 0.10476447790861129,
|
|
"num_tokens": 774128.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.131285381317139,
|
|
"epoch": 0.02184566334061522,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 6.7883,
|
|
"mean_token_accuracy": 0.09998859390616417,
|
|
"num_tokens": 784292.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.133191156387329,
|
|
"epoch": 0.02214087500738029,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000187,
|
|
"loss": 6.7794,
|
|
"mean_token_accuracy": 0.09937844276428223,
|
|
"num_tokens": 795437.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.124974298477173,
|
|
"epoch": 0.022436086674145362,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 6.687,
|
|
"mean_token_accuracy": 0.10222809389233589,
|
|
"num_tokens": 806543.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.1218994140625,
|
|
"epoch": 0.022731298340910432,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.000192,
|
|
"loss": 6.843,
|
|
"mean_token_accuracy": 0.09876690059900284,
|
|
"num_tokens": 817179.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.118319416046143,
|
|
"epoch": 0.023026510007675503,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.7608,
|
|
"mean_token_accuracy": 0.10439011678099633,
|
|
"num_tokens": 826749.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.08788104057312,
|
|
"epoch": 0.023321721674440573,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 6.798,
|
|
"mean_token_accuracy": 0.09762646481394768,
|
|
"num_tokens": 838807.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.131559801101685,
|
|
"epoch": 0.023616933341205643,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.7766,
|
|
"mean_token_accuracy": 0.10133634880185127,
|
|
"num_tokens": 849651.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.159732913970947,
|
|
"epoch": 0.023912145007970714,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.8434,
|
|
"mean_token_accuracy": 0.09272509887814522,
|
|
"num_tokens": 859753.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.0073809146881105,
|
|
"epoch": 0.024207356674735784,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.7134,
|
|
"mean_token_accuracy": 0.10226179882884026,
|
|
"num_tokens": 869904.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 6.957417392730713,
|
|
"epoch": 0.024502568341500858,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.73,
|
|
"mean_token_accuracy": 0.10192711651325226,
|
|
"num_tokens": 880948.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.170308446884155,
|
|
"epoch": 0.024797780008265928,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.7966,
|
|
"mean_token_accuracy": 0.09502052217721939,
|
|
"num_tokens": 892331.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 6.980013370513916,
|
|
"epoch": 0.025092991675031,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.628,
|
|
"mean_token_accuracy": 0.1050586886703968,
|
|
"num_tokens": 901691.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.007831001281739,
|
|
"epoch": 0.02538820334179607,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.6769,
|
|
"mean_token_accuracy": 0.10150302425026894,
|
|
"num_tokens": 912495.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.010691833496094,
|
|
"epoch": 0.02568341500856114,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.636,
|
|
"mean_token_accuracy": 0.1017840936779976,
|
|
"num_tokens": 923420.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 6.8785758972167965,
|
|
"epoch": 0.02597862667532621,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.6396,
|
|
"mean_token_accuracy": 0.09978873506188393,
|
|
"num_tokens": 933598.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 6.98797607421875,
|
|
"epoch": 0.02627383834209128,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.7659,
|
|
"mean_token_accuracy": 0.1045355699956417,
|
|
"num_tokens": 943229.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 6.983278131484985,
|
|
"epoch": 0.02656905000885635,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.6247,
|
|
"mean_token_accuracy": 0.10298437029123306,
|
|
"num_tokens": 954425.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.073006629943848,
|
|
"epoch": 0.02686426167562142,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.745,
|
|
"mean_token_accuracy": 0.10184216126799583,
|
|
"num_tokens": 965516.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 6.900727653503418,
|
|
"epoch": 0.02715947334238649,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.6285,
|
|
"mean_token_accuracy": 0.10302256718277931,
|
|
"num_tokens": 976469.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 6.955856561660767,
|
|
"epoch": 0.02745468500915156,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.7139,
|
|
"mean_token_accuracy": 0.09732916429638863,
|
|
"num_tokens": 986946.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 6.982401418685913,
|
|
"epoch": 0.02774989667591663,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.612,
|
|
"mean_token_accuracy": 0.10779485702514649,
|
|
"num_tokens": 997101.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 6.9505868434906,
|
|
"epoch": 0.0280451083426817,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.7346,
|
|
"mean_token_accuracy": 0.10266673490405083,
|
|
"num_tokens": 1007866.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.019359874725342,
|
|
"epoch": 0.028340320009446772,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.7355,
|
|
"mean_token_accuracy": 0.09643693268299103,
|
|
"num_tokens": 1019377.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 6.936902332305908,
|
|
"epoch": 0.028635531676211842,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.7267,
|
|
"mean_token_accuracy": 0.09898768290877343,
|
|
"num_tokens": 1031278.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 6.772035408020019,
|
|
"epoch": 0.028930743342976916,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.5192,
|
|
"mean_token_accuracy": 0.1116688534617424,
|
|
"num_tokens": 1041663.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 6.908066129684448,
|
|
"epoch": 0.029225955009741986,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.5969,
|
|
"mean_token_accuracy": 0.10450899600982666,
|
|
"num_tokens": 1052286.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 6.807862234115601,
|
|
"epoch": 0.029521166676507057,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.6629,
|
|
"mean_token_accuracy": 0.10684632584452629,
|
|
"num_tokens": 1061945.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"entropy": 6.962998628616333,
|
|
"epoch": 0.029816378343272127,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.000252,
|
|
"loss": 6.6436,
|
|
"mean_token_accuracy": 0.10230686739087105,
|
|
"num_tokens": 1074057.0,
|
|
"step": 505
|
|
},
|
|
{
|
|
"entropy": 6.809896945953369,
|
|
"epoch": 0.030111590010037197,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.0002545,
|
|
"loss": 6.6465,
|
|
"mean_token_accuracy": 0.10390325114130974,
|
|
"num_tokens": 1084830.0,
|
|
"step": 510
|
|
},
|
|
{
|
|
"entropy": 6.909462213516235,
|
|
"epoch": 0.030406801676802268,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.000257,
|
|
"loss": 6.5865,
|
|
"mean_token_accuracy": 0.11080023497343064,
|
|
"num_tokens": 1095540.0,
|
|
"step": 515
|
|
},
|
|
{
|
|
"entropy": 6.853788089752197,
|
|
"epoch": 0.030702013343567338,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.0002595,
|
|
"loss": 6.5906,
|
|
"mean_token_accuracy": 0.10921324118971824,
|
|
"num_tokens": 1105395.0,
|
|
"step": 520
|
|
},
|
|
{
|
|
"entropy": 6.8561993598937985,
|
|
"epoch": 0.03099722501033241,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.000262,
|
|
"loss": 6.5617,
|
|
"mean_token_accuracy": 0.0991988554596901,
|
|
"num_tokens": 1117007.0,
|
|
"step": 525
|
|
},
|
|
{
|
|
"entropy": 6.876309776306153,
|
|
"epoch": 0.03129243667709748,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.00026450000000000003,
|
|
"loss": 6.5751,
|
|
"mean_token_accuracy": 0.11366137936711311,
|
|
"num_tokens": 1128531.0,
|
|
"step": 530
|
|
},
|
|
{
|
|
"entropy": 6.810136461257935,
|
|
"epoch": 0.03158764834386255,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00026700000000000004,
|
|
"loss": 6.5986,
|
|
"mean_token_accuracy": 0.10317134782671929,
|
|
"num_tokens": 1139194.0,
|
|
"step": 535
|
|
},
|
|
{
|
|
"entropy": 6.84046630859375,
|
|
"epoch": 0.03188286001062762,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.00026950000000000005,
|
|
"loss": 6.64,
|
|
"mean_token_accuracy": 0.09497019648551941,
|
|
"num_tokens": 1150371.0,
|
|
"step": 540
|
|
},
|
|
{
|
|
"entropy": 6.778982639312744,
|
|
"epoch": 0.03217807167739269,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00027200000000000005,
|
|
"loss": 6.5427,
|
|
"mean_token_accuracy": 0.10634316056966782,
|
|
"num_tokens": 1160840.0,
|
|
"step": 545
|
|
},
|
|
{
|
|
"entropy": 6.843386459350586,
|
|
"epoch": 0.03247328334415776,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0002745,
|
|
"loss": 6.595,
|
|
"mean_token_accuracy": 0.10080070197582244,
|
|
"num_tokens": 1171438.0,
|
|
"step": 550
|
|
},
|
|
{
|
|
"entropy": 6.788133335113526,
|
|
"epoch": 0.032768495010922834,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.000277,
|
|
"loss": 6.5448,
|
|
"mean_token_accuracy": 0.11157862991094589,
|
|
"num_tokens": 1180889.0,
|
|
"step": 555
|
|
},
|
|
{
|
|
"entropy": 6.7335409164428714,
|
|
"epoch": 0.033063706677687904,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.0002795,
|
|
"loss": 6.5536,
|
|
"mean_token_accuracy": 0.10646847784519195,
|
|
"num_tokens": 1190755.0,
|
|
"step": 560
|
|
},
|
|
{
|
|
"entropy": 6.849792003631592,
|
|
"epoch": 0.033358918344452974,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.00028199999999999997,
|
|
"loss": 6.6392,
|
|
"mean_token_accuracy": 0.10497644990682602,
|
|
"num_tokens": 1202002.0,
|
|
"step": 565
|
|
},
|
|
{
|
|
"entropy": 6.836796712875366,
|
|
"epoch": 0.033654130011218045,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0002845,
|
|
"loss": 6.6549,
|
|
"mean_token_accuracy": 0.10088638663291931,
|
|
"num_tokens": 1213611.0,
|
|
"step": 570
|
|
},
|
|
{
|
|
"entropy": 6.87101526260376,
|
|
"epoch": 0.033949341677983115,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.000287,
|
|
"loss": 6.665,
|
|
"mean_token_accuracy": 0.10538341179490089,
|
|
"num_tokens": 1224450.0,
|
|
"step": 575
|
|
},
|
|
{
|
|
"entropy": 6.789998292922974,
|
|
"epoch": 0.034244553344748185,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0002895,
|
|
"loss": 6.6039,
|
|
"mean_token_accuracy": 0.10749325305223464,
|
|
"num_tokens": 1235229.0,
|
|
"step": 580
|
|
},
|
|
{
|
|
"entropy": 6.817308807373047,
|
|
"epoch": 0.034539765011513256,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.000292,
|
|
"loss": 6.6003,
|
|
"mean_token_accuracy": 0.10448379814624786,
|
|
"num_tokens": 1245015.0,
|
|
"step": 585
|
|
},
|
|
{
|
|
"entropy": 6.658805990219117,
|
|
"epoch": 0.034834976678278326,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.0002945,
|
|
"loss": 6.4771,
|
|
"mean_token_accuracy": 0.11165011152625084,
|
|
"num_tokens": 1256307.0,
|
|
"step": 590
|
|
},
|
|
{
|
|
"entropy": 6.781815767288208,
|
|
"epoch": 0.035130188345043396,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000297,
|
|
"loss": 6.5426,
|
|
"mean_token_accuracy": 0.1121568076312542,
|
|
"num_tokens": 1267042.0,
|
|
"step": 595
|
|
},
|
|
{
|
|
"entropy": 6.763632011413574,
|
|
"epoch": 0.03542540001180847,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0002995,
|
|
"loss": 6.5296,
|
|
"mean_token_accuracy": 0.11234459504485131,
|
|
"num_tokens": 1277571.0,
|
|
"step": 600
|
|
},
|
|
{
|
|
"entropy": 6.766745567321777,
|
|
"epoch": 0.03572061167857354,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000302,
|
|
"loss": 6.6043,
|
|
"mean_token_accuracy": 0.10907782688736915,
|
|
"num_tokens": 1288203.0,
|
|
"step": 605
|
|
},
|
|
{
|
|
"entropy": 6.755765724182129,
|
|
"epoch": 0.03601582334533861,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.0003045,
|
|
"loss": 6.5908,
|
|
"mean_token_accuracy": 0.11033066883683204,
|
|
"num_tokens": 1299510.0,
|
|
"step": 610
|
|
},
|
|
{
|
|
"entropy": 6.666236639022827,
|
|
"epoch": 0.03631103501210368,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.000307,
|
|
"loss": 6.4499,
|
|
"mean_token_accuracy": 0.11379646882414818,
|
|
"num_tokens": 1309573.0,
|
|
"step": 615
|
|
},
|
|
{
|
|
"entropy": 6.696754121780396,
|
|
"epoch": 0.03660624667886875,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.0003095,
|
|
"loss": 6.5194,
|
|
"mean_token_accuracy": 0.11005613952875137,
|
|
"num_tokens": 1321104.0,
|
|
"step": 620
|
|
},
|
|
{
|
|
"entropy": 6.693728637695313,
|
|
"epoch": 0.03690145834563382,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.000312,
|
|
"loss": 6.536,
|
|
"mean_token_accuracy": 0.10800015479326248,
|
|
"num_tokens": 1332671.0,
|
|
"step": 625
|
|
},
|
|
{
|
|
"entropy": 6.623374509811401,
|
|
"epoch": 0.03719667001239889,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0003145,
|
|
"loss": 6.4132,
|
|
"mean_token_accuracy": 0.11599879413843155,
|
|
"num_tokens": 1344416.0,
|
|
"step": 630
|
|
},
|
|
{
|
|
"entropy": 6.784224557876587,
|
|
"epoch": 0.03749188167916396,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000317,
|
|
"loss": 6.5499,
|
|
"mean_token_accuracy": 0.10508671477437019,
|
|
"num_tokens": 1355740.0,
|
|
"step": 635
|
|
},
|
|
{
|
|
"entropy": 6.663637495040893,
|
|
"epoch": 0.03778709334592903,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.0003195,
|
|
"loss": 6.5103,
|
|
"mean_token_accuracy": 0.10943645983934402,
|
|
"num_tokens": 1366465.0,
|
|
"step": 640
|
|
},
|
|
{
|
|
"entropy": 6.570268201828003,
|
|
"epoch": 0.0380823050126941,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000322,
|
|
"loss": 6.4257,
|
|
"mean_token_accuracy": 0.10816865861415863,
|
|
"num_tokens": 1376476.0,
|
|
"step": 645
|
|
},
|
|
{
|
|
"entropy": 6.72826361656189,
|
|
"epoch": 0.03837751667945917,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 0.00032450000000000003,
|
|
"loss": 6.4271,
|
|
"mean_token_accuracy": 0.11982382684946061,
|
|
"num_tokens": 1386172.0,
|
|
"step": 650
|
|
},
|
|
{
|
|
"entropy": 6.599955940246582,
|
|
"epoch": 0.03867272834622424,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.00032700000000000003,
|
|
"loss": 6.4205,
|
|
"mean_token_accuracy": 0.10781983509659768,
|
|
"num_tokens": 1396655.0,
|
|
"step": 655
|
|
},
|
|
{
|
|
"entropy": 6.604927682876587,
|
|
"epoch": 0.03896794001298931,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00032950000000000004,
|
|
"loss": 6.4394,
|
|
"mean_token_accuracy": 0.10952353551983833,
|
|
"num_tokens": 1407816.0,
|
|
"step": 660
|
|
},
|
|
{
|
|
"entropy": 6.50361123085022,
|
|
"epoch": 0.03926315167975438,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.00033200000000000005,
|
|
"loss": 6.4102,
|
|
"mean_token_accuracy": 0.11540385410189628,
|
|
"num_tokens": 1418265.0,
|
|
"step": 665
|
|
},
|
|
{
|
|
"entropy": 6.59829306602478,
|
|
"epoch": 0.03955836334651945,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00033450000000000005,
|
|
"loss": 6.404,
|
|
"mean_token_accuracy": 0.11367250382900237,
|
|
"num_tokens": 1429776.0,
|
|
"step": 670
|
|
},
|
|
{
|
|
"entropy": 6.546575927734375,
|
|
"epoch": 0.03985357501328453,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.000337,
|
|
"loss": 6.4344,
|
|
"mean_token_accuracy": 0.11163216009736061,
|
|
"num_tokens": 1440295.0,
|
|
"step": 675
|
|
},
|
|
{
|
|
"entropy": 6.640892171859742,
|
|
"epoch": 0.0401487866800496,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0003395,
|
|
"loss": 6.4677,
|
|
"mean_token_accuracy": 0.11355267614126205,
|
|
"num_tokens": 1450505.0,
|
|
"step": 680
|
|
},
|
|
{
|
|
"entropy": 6.712115478515625,
|
|
"epoch": 0.04044399834681467,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000342,
|
|
"loss": 6.5529,
|
|
"mean_token_accuracy": 0.11008020266890525,
|
|
"num_tokens": 1461768.0,
|
|
"step": 685
|
|
},
|
|
{
|
|
"entropy": 6.6160581588745115,
|
|
"epoch": 0.04073921001357974,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.00034449999999999997,
|
|
"loss": 6.4961,
|
|
"mean_token_accuracy": 0.11170130670070648,
|
|
"num_tokens": 1473965.0,
|
|
"step": 690
|
|
},
|
|
{
|
|
"entropy": 6.568212413787842,
|
|
"epoch": 0.04103442168034481,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.000347,
|
|
"loss": 6.4378,
|
|
"mean_token_accuracy": 0.11968905776739121,
|
|
"num_tokens": 1484750.0,
|
|
"step": 695
|
|
},
|
|
{
|
|
"entropy": 6.620304441452026,
|
|
"epoch": 0.04132963334710988,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0003495,
|
|
"loss": 6.4357,
|
|
"mean_token_accuracy": 0.10914467945694924,
|
|
"num_tokens": 1496474.0,
|
|
"step": 700
|
|
},
|
|
{
|
|
"entropy": 6.5506373882293705,
|
|
"epoch": 0.04162484501387495,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000352,
|
|
"loss": 6.5706,
|
|
"mean_token_accuracy": 0.1017179697751999,
|
|
"num_tokens": 1507880.0,
|
|
"step": 705
|
|
},
|
|
{
|
|
"entropy": 6.623064422607422,
|
|
"epoch": 0.04192005668064002,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 0.0003545,
|
|
"loss": 6.4875,
|
|
"mean_token_accuracy": 0.10796806961297989,
|
|
"num_tokens": 1519115.0,
|
|
"step": 710
|
|
},
|
|
{
|
|
"entropy": 6.600681447982788,
|
|
"epoch": 0.04221526834740509,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.000357,
|
|
"loss": 6.4319,
|
|
"mean_token_accuracy": 0.11712071001529693,
|
|
"num_tokens": 1529031.0,
|
|
"step": 715
|
|
},
|
|
{
|
|
"entropy": 6.581659746170044,
|
|
"epoch": 0.04251048001417016,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0003595,
|
|
"loss": 6.3756,
|
|
"mean_token_accuracy": 0.11652529910206795,
|
|
"num_tokens": 1539954.0,
|
|
"step": 720
|
|
},
|
|
{
|
|
"entropy": 6.566905403137207,
|
|
"epoch": 0.04280569168093523,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000362,
|
|
"loss": 6.5397,
|
|
"mean_token_accuracy": 0.10744471997022628,
|
|
"num_tokens": 1550713.0,
|
|
"step": 725
|
|
},
|
|
{
|
|
"entropy": 6.5449799537658695,
|
|
"epoch": 0.0431009033477003,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.0003645,
|
|
"loss": 6.4371,
|
|
"mean_token_accuracy": 0.10990587547421456,
|
|
"num_tokens": 1562569.0,
|
|
"step": 730
|
|
},
|
|
{
|
|
"entropy": 6.6069684505462645,
|
|
"epoch": 0.04339611501446537,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.000367,
|
|
"loss": 6.4711,
|
|
"mean_token_accuracy": 0.11592478230595589,
|
|
"num_tokens": 1572999.0,
|
|
"step": 735
|
|
},
|
|
{
|
|
"entropy": 6.60914249420166,
|
|
"epoch": 0.04369132668123044,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.0003695,
|
|
"loss": 6.4636,
|
|
"mean_token_accuracy": 0.10721064656972885,
|
|
"num_tokens": 1583560.0,
|
|
"step": 740
|
|
},
|
|
{
|
|
"entropy": 6.513941383361816,
|
|
"epoch": 0.04398653834799551,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000372,
|
|
"loss": 6.3622,
|
|
"mean_token_accuracy": 0.11819557920098304,
|
|
"num_tokens": 1593931.0,
|
|
"step": 745
|
|
},
|
|
{
|
|
"entropy": 6.53983850479126,
|
|
"epoch": 0.04428175001476058,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0003745,
|
|
"loss": 6.4451,
|
|
"mean_token_accuracy": 0.10905564352869987,
|
|
"num_tokens": 1605182.0,
|
|
"step": 750
|
|
},
|
|
{
|
|
"entropy": 6.547965335845947,
|
|
"epoch": 0.044576961681525654,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000377,
|
|
"loss": 6.3938,
|
|
"mean_token_accuracy": 0.11519684940576554,
|
|
"num_tokens": 1615912.0,
|
|
"step": 755
|
|
},
|
|
{
|
|
"entropy": 6.417063283920288,
|
|
"epoch": 0.044872173348290724,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.0003795,
|
|
"loss": 6.3974,
|
|
"mean_token_accuracy": 0.11966018378734589,
|
|
"num_tokens": 1626322.0,
|
|
"step": 760
|
|
},
|
|
{
|
|
"entropy": 6.590868949890137,
|
|
"epoch": 0.045167385015055794,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.000382,
|
|
"loss": 6.3744,
|
|
"mean_token_accuracy": 0.11621762439608574,
|
|
"num_tokens": 1636859.0,
|
|
"step": 765
|
|
},
|
|
{
|
|
"entropy": 6.502106046676635,
|
|
"epoch": 0.045462596681820865,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0003845,
|
|
"loss": 6.4183,
|
|
"mean_token_accuracy": 0.10947346538305283,
|
|
"num_tokens": 1648511.0,
|
|
"step": 770
|
|
},
|
|
{
|
|
"entropy": 6.496643590927124,
|
|
"epoch": 0.045757808348585935,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.00038700000000000003,
|
|
"loss": 6.3729,
|
|
"mean_token_accuracy": 0.11914441511034965,
|
|
"num_tokens": 1659350.0,
|
|
"step": 775
|
|
},
|
|
{
|
|
"entropy": 6.433891248703003,
|
|
"epoch": 0.046053020015351005,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.00038950000000000003,
|
|
"loss": 6.4166,
|
|
"mean_token_accuracy": 0.11184289306402206,
|
|
"num_tokens": 1670077.0,
|
|
"step": 780
|
|
},
|
|
{
|
|
"entropy": 6.547736930847168,
|
|
"epoch": 0.046348231682116076,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.00039200000000000004,
|
|
"loss": 6.4018,
|
|
"mean_token_accuracy": 0.11911343410611153,
|
|
"num_tokens": 1681877.0,
|
|
"step": 785
|
|
},
|
|
{
|
|
"entropy": 6.486341857910157,
|
|
"epoch": 0.046643443348881146,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.00039450000000000005,
|
|
"loss": 6.3895,
|
|
"mean_token_accuracy": 0.11960405185818672,
|
|
"num_tokens": 1692448.0,
|
|
"step": 790
|
|
},
|
|
{
|
|
"entropy": 6.552379465103149,
|
|
"epoch": 0.046938655015646216,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.00039700000000000005,
|
|
"loss": 6.4569,
|
|
"mean_token_accuracy": 0.11376626342535019,
|
|
"num_tokens": 1702437.0,
|
|
"step": 795
|
|
},
|
|
{
|
|
"entropy": 6.533510732650757,
|
|
"epoch": 0.04723386668241129,
|
|
"grad_norm": 0.9609375,
|
|
"learning_rate": 0.0003995,
|
|
"loss": 6.3491,
|
|
"mean_token_accuracy": 0.11783147305250168,
|
|
"num_tokens": 1712587.0,
|
|
"step": 800
|
|
},
|
|
{
|
|
"entropy": 6.516396141052246,
|
|
"epoch": 0.04752907834917636,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000402,
|
|
"loss": 6.4434,
|
|
"mean_token_accuracy": 0.11643455773591996,
|
|
"num_tokens": 1722782.0,
|
|
"step": 805
|
|
},
|
|
{
|
|
"entropy": 6.544632863998413,
|
|
"epoch": 0.04782429001594143,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0004045,
|
|
"loss": 6.3938,
|
|
"mean_token_accuracy": 0.1165135882794857,
|
|
"num_tokens": 1733993.0,
|
|
"step": 810
|
|
},
|
|
{
|
|
"entropy": 6.4083781242370605,
|
|
"epoch": 0.0481195016827065,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.00040699999999999997,
|
|
"loss": 6.4062,
|
|
"mean_token_accuracy": 0.11843438744544983,
|
|
"num_tokens": 1744095.0,
|
|
"step": 815
|
|
},
|
|
{
|
|
"entropy": 6.446332406997681,
|
|
"epoch": 0.04841471334947157,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.0004095,
|
|
"loss": 6.3581,
|
|
"mean_token_accuracy": 0.11571272611618041,
|
|
"num_tokens": 1754356.0,
|
|
"step": 820
|
|
},
|
|
{
|
|
"entropy": 6.452772045135498,
|
|
"epoch": 0.048709925016236645,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.000412,
|
|
"loss": 6.4377,
|
|
"mean_token_accuracy": 0.11594505831599236,
|
|
"num_tokens": 1765628.0,
|
|
"step": 825
|
|
},
|
|
{
|
|
"entropy": 6.456363677978516,
|
|
"epoch": 0.049005136683001715,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.0004145,
|
|
"loss": 6.3107,
|
|
"mean_token_accuracy": 0.11763904914259911,
|
|
"num_tokens": 1775697.0,
|
|
"step": 830
|
|
},
|
|
{
|
|
"entropy": 6.446932458877564,
|
|
"epoch": 0.049300348349766786,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.000417,
|
|
"loss": 6.412,
|
|
"mean_token_accuracy": 0.1122404970228672,
|
|
"num_tokens": 1786066.0,
|
|
"step": 835
|
|
},
|
|
{
|
|
"entropy": 6.421263313293457,
|
|
"epoch": 0.049595560016531856,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0004195,
|
|
"loss": 6.2323,
|
|
"mean_token_accuracy": 0.12168351784348488,
|
|
"num_tokens": 1796735.0,
|
|
"step": 840
|
|
},
|
|
{
|
|
"entropy": 6.4752278327941895,
|
|
"epoch": 0.049890771683296926,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000422,
|
|
"loss": 6.3938,
|
|
"mean_token_accuracy": 0.11445808708667755,
|
|
"num_tokens": 1807164.0,
|
|
"step": 845
|
|
},
|
|
{
|
|
"entropy": 6.44355959892273,
|
|
"epoch": 0.050185983350062,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0004245,
|
|
"loss": 6.4254,
|
|
"mean_token_accuracy": 0.11221889480948448,
|
|
"num_tokens": 1818788.0,
|
|
"step": 850
|
|
},
|
|
{
|
|
"entropy": 6.524218511581421,
|
|
"epoch": 0.05048119501682707,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000427,
|
|
"loss": 6.4196,
|
|
"mean_token_accuracy": 0.11620973348617554,
|
|
"num_tokens": 1830488.0,
|
|
"step": 855
|
|
},
|
|
{
|
|
"entropy": 6.335734176635742,
|
|
"epoch": 0.05077640668359214,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0004295,
|
|
"loss": 6.3469,
|
|
"mean_token_accuracy": 0.11381080970168114,
|
|
"num_tokens": 1841085.0,
|
|
"step": 860
|
|
},
|
|
{
|
|
"entropy": 6.448235368728637,
|
|
"epoch": 0.05107161835035721,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.000432,
|
|
"loss": 6.324,
|
|
"mean_token_accuracy": 0.11611942201852798,
|
|
"num_tokens": 1852831.0,
|
|
"step": 865
|
|
},
|
|
{
|
|
"entropy": 6.353146409988403,
|
|
"epoch": 0.05136683001712228,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0004345,
|
|
"loss": 6.2856,
|
|
"mean_token_accuracy": 0.12149567529559135,
|
|
"num_tokens": 1863868.0,
|
|
"step": 870
|
|
},
|
|
{
|
|
"entropy": 6.34972448348999,
|
|
"epoch": 0.05166204168388735,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000437,
|
|
"loss": 6.2641,
|
|
"mean_token_accuracy": 0.11966291964054107,
|
|
"num_tokens": 1874227.0,
|
|
"step": 875
|
|
},
|
|
{
|
|
"entropy": 6.424135446548462,
|
|
"epoch": 0.05195725335065242,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.0004395,
|
|
"loss": 6.2741,
|
|
"mean_token_accuracy": 0.11863231882452965,
|
|
"num_tokens": 1883712.0,
|
|
"step": 880
|
|
},
|
|
{
|
|
"entropy": 6.426289415359497,
|
|
"epoch": 0.05225246501741749,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000442,
|
|
"loss": 6.3813,
|
|
"mean_token_accuracy": 0.12182779237627983,
|
|
"num_tokens": 1894909.0,
|
|
"step": 885
|
|
},
|
|
{
|
|
"entropy": 6.322753953933716,
|
|
"epoch": 0.05254767668418256,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.0004445,
|
|
"loss": 6.2976,
|
|
"mean_token_accuracy": 0.12649846225976943,
|
|
"num_tokens": 1904973.0,
|
|
"step": 890
|
|
},
|
|
{
|
|
"entropy": 6.352806806564331,
|
|
"epoch": 0.05284288835094763,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000447,
|
|
"loss": 6.2755,
|
|
"mean_token_accuracy": 0.1184873789548874,
|
|
"num_tokens": 1915131.0,
|
|
"step": 895
|
|
},
|
|
{
|
|
"entropy": 6.3792942523956295,
|
|
"epoch": 0.0531381000177127,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.00044950000000000003,
|
|
"loss": 6.3776,
|
|
"mean_token_accuracy": 0.11576106548309326,
|
|
"num_tokens": 1925893.0,
|
|
"step": 900
|
|
},
|
|
{
|
|
"entropy": 6.471661901473999,
|
|
"epoch": 0.05343331168447777,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.00045200000000000004,
|
|
"loss": 6.3869,
|
|
"mean_token_accuracy": 0.11486705616116524,
|
|
"num_tokens": 1936830.0,
|
|
"step": 905
|
|
},
|
|
{
|
|
"entropy": 6.366755723953247,
|
|
"epoch": 0.05372852335124284,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.00045450000000000004,
|
|
"loss": 6.32,
|
|
"mean_token_accuracy": 0.12052246332168579,
|
|
"num_tokens": 1948291.0,
|
|
"step": 910
|
|
},
|
|
{
|
|
"entropy": 6.413037490844727,
|
|
"epoch": 0.05402373501800791,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.00045700000000000005,
|
|
"loss": 6.3619,
|
|
"mean_token_accuracy": 0.11181799247860909,
|
|
"num_tokens": 1959168.0,
|
|
"step": 915
|
|
},
|
|
{
|
|
"entropy": 6.320397138595581,
|
|
"epoch": 0.05431894668477298,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00045950000000000006,
|
|
"loss": 6.3454,
|
|
"mean_token_accuracy": 0.11968773975968361,
|
|
"num_tokens": 1970046.0,
|
|
"step": 920
|
|
},
|
|
{
|
|
"entropy": 6.349421739578247,
|
|
"epoch": 0.05461415835153805,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000462,
|
|
"loss": 6.333,
|
|
"mean_token_accuracy": 0.11495715156197547,
|
|
"num_tokens": 1980430.0,
|
|
"step": 925
|
|
},
|
|
{
|
|
"entropy": 6.475516033172608,
|
|
"epoch": 0.05490937001830312,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0004645,
|
|
"loss": 6.3077,
|
|
"mean_token_accuracy": 0.12274432182312012,
|
|
"num_tokens": 1991035.0,
|
|
"step": 930
|
|
},
|
|
{
|
|
"entropy": 6.42578911781311,
|
|
"epoch": 0.05520458168506819,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000467,
|
|
"loss": 6.3654,
|
|
"mean_token_accuracy": 0.11354202851653099,
|
|
"num_tokens": 2002122.0,
|
|
"step": 935
|
|
},
|
|
{
|
|
"entropy": 6.304956388473511,
|
|
"epoch": 0.05549979335183326,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0004695,
|
|
"loss": 6.2948,
|
|
"mean_token_accuracy": 0.11706151142716407,
|
|
"num_tokens": 2012676.0,
|
|
"step": 940
|
|
},
|
|
{
|
|
"entropy": 6.362958908081055,
|
|
"epoch": 0.05579500501859833,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.000472,
|
|
"loss": 6.3144,
|
|
"mean_token_accuracy": 0.11494619697332382,
|
|
"num_tokens": 2023124.0,
|
|
"step": 945
|
|
},
|
|
{
|
|
"entropy": 6.352200746536255,
|
|
"epoch": 0.0560902166853634,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.0004745,
|
|
"loss": 6.3036,
|
|
"mean_token_accuracy": 0.1179499588906765,
|
|
"num_tokens": 2034030.0,
|
|
"step": 950
|
|
},
|
|
{
|
|
"entropy": 6.315423631668091,
|
|
"epoch": 0.056385428352128474,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000477,
|
|
"loss": 6.3389,
|
|
"mean_token_accuracy": 0.11249075978994369,
|
|
"num_tokens": 2044499.0,
|
|
"step": 955
|
|
},
|
|
{
|
|
"entropy": 6.378019380569458,
|
|
"epoch": 0.056680640018893544,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0004795,
|
|
"loss": 6.3761,
|
|
"mean_token_accuracy": 0.1221327118575573,
|
|
"num_tokens": 2055242.0,
|
|
"step": 960
|
|
},
|
|
{
|
|
"entropy": 6.296391916275025,
|
|
"epoch": 0.056975851685658614,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.000482,
|
|
"loss": 6.3379,
|
|
"mean_token_accuracy": 0.11753152683377266,
|
|
"num_tokens": 2065987.0,
|
|
"step": 965
|
|
},
|
|
{
|
|
"entropy": 6.374587059020996,
|
|
"epoch": 0.057271063352423685,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.0004845,
|
|
"loss": 6.2512,
|
|
"mean_token_accuracy": 0.12176798135042191,
|
|
"num_tokens": 2077015.0,
|
|
"step": 970
|
|
},
|
|
{
|
|
"entropy": 6.433490657806397,
|
|
"epoch": 0.05756627501918876,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000487,
|
|
"loss": 6.3711,
|
|
"mean_token_accuracy": 0.12369481921195984,
|
|
"num_tokens": 2087411.0,
|
|
"step": 975
|
|
},
|
|
{
|
|
"entropy": 6.3158018589019775,
|
|
"epoch": 0.05786148668595383,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.0004895,
|
|
"loss": 6.2923,
|
|
"mean_token_accuracy": 0.12385761588811875,
|
|
"num_tokens": 2097612.0,
|
|
"step": 980
|
|
},
|
|
{
|
|
"entropy": 6.208762693405151,
|
|
"epoch": 0.0581566983527189,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000492,
|
|
"loss": 6.2291,
|
|
"mean_token_accuracy": 0.1223476842045784,
|
|
"num_tokens": 2108118.0,
|
|
"step": 985
|
|
},
|
|
{
|
|
"entropy": 6.340756034851074,
|
|
"epoch": 0.05845191001948397,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0004945,
|
|
"loss": 6.3148,
|
|
"mean_token_accuracy": 0.11962561458349227,
|
|
"num_tokens": 2118262.0,
|
|
"step": 990
|
|
},
|
|
{
|
|
"entropy": 6.331036233901978,
|
|
"epoch": 0.05874712168624904,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.000497,
|
|
"loss": 6.3164,
|
|
"mean_token_accuracy": 0.11812316924333573,
|
|
"num_tokens": 2128819.0,
|
|
"step": 995
|
|
},
|
|
{
|
|
"entropy": 6.350315570831299,
|
|
"epoch": 0.05904233335301411,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0004995,
|
|
"loss": 6.2294,
|
|
"mean_token_accuracy": 0.12225014418363571,
|
|
"num_tokens": 2139852.0,
|
|
"step": 1000
|
|
}
|
|
],
|
|
"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": 3132581179392000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|