6aa7ea38aa
Model: fpadovani/nld-latn-10mb-after-ppt-shuff-dyck-10mb-ckpt500_seed3407 Source: Original Platform
1046 lines
28 KiB
JSON
1046 lines
28 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.45433893684688775,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 4.856400108337402,
|
|
"epoch": 0.004543389368468878,
|
|
"grad_norm": 21.625,
|
|
"learning_rate": 2e-06,
|
|
"loss": 14.2106,
|
|
"mean_token_accuracy": 0.0006149356486275792,
|
|
"num_tokens": 10156.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 4.860177421569825,
|
|
"epoch": 0.009086778736937756,
|
|
"grad_norm": 21.375,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 14.2151,
|
|
"mean_token_accuracy": 0.0003687591990455985,
|
|
"num_tokens": 20669.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 4.876665878295898,
|
|
"epoch": 0.013630168105406633,
|
|
"grad_norm": 22.5,
|
|
"learning_rate": 7e-06,
|
|
"loss": 14.0374,
|
|
"mean_token_accuracy": 0.000468046945752576,
|
|
"num_tokens": 30952.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 4.957574272155762,
|
|
"epoch": 0.01817355747387551,
|
|
"grad_norm": 27.75,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 13.7887,
|
|
"mean_token_accuracy": 0.0006278944842051714,
|
|
"num_tokens": 41429.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 5.298445510864258,
|
|
"epoch": 0.02271694684234439,
|
|
"grad_norm": 35.0,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 13.1575,
|
|
"mean_token_accuracy": 0.0006482779106590896,
|
|
"num_tokens": 52166.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 6.667004203796386,
|
|
"epoch": 0.027260336210813266,
|
|
"grad_norm": 36.25,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 12.0099,
|
|
"mean_token_accuracy": 0.00022935778833925723,
|
|
"num_tokens": 61670.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.085978126525879,
|
|
"epoch": 0.03180372557928214,
|
|
"grad_norm": 4.0,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.8036,
|
|
"mean_token_accuracy": 0.006481384776998311,
|
|
"num_tokens": 71702.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.696837425231934,
|
|
"epoch": 0.03634711494775102,
|
|
"grad_norm": 3.5625,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.629,
|
|
"mean_token_accuracy": 0.014822260290384293,
|
|
"num_tokens": 82840.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.674850368499756,
|
|
"epoch": 0.0408905043162199,
|
|
"grad_norm": 3.09375,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.4551,
|
|
"mean_token_accuracy": 0.017661335971206428,
|
|
"num_tokens": 92290.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.59873275756836,
|
|
"epoch": 0.04543389368468878,
|
|
"grad_norm": 2.390625,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 10.2805,
|
|
"mean_token_accuracy": 0.017723128758370878,
|
|
"num_tokens": 101842.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.64423999786377,
|
|
"epoch": 0.049977283053157656,
|
|
"grad_norm": 2.140625,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 10.2044,
|
|
"mean_token_accuracy": 0.013444735389202832,
|
|
"num_tokens": 114030.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.725485229492188,
|
|
"epoch": 0.05452067242162653,
|
|
"grad_norm": 2.0,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 10.0243,
|
|
"mean_token_accuracy": 0.02563867699354887,
|
|
"num_tokens": 124349.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.725618076324462,
|
|
"epoch": 0.059064061790095414,
|
|
"grad_norm": 1.9453125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.8587,
|
|
"mean_token_accuracy": 0.03280576281249523,
|
|
"num_tokens": 133534.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.69280939102173,
|
|
"epoch": 0.06360745115856428,
|
|
"grad_norm": 2.15625,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.8074,
|
|
"mean_token_accuracy": 0.019261055253446104,
|
|
"num_tokens": 144383.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.707190322875977,
|
|
"epoch": 0.06815084052703317,
|
|
"grad_norm": 1.796875,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.7725,
|
|
"mean_token_accuracy": 0.028312317933887242,
|
|
"num_tokens": 155467.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.719540405273438,
|
|
"epoch": 0.07269422989550205,
|
|
"grad_norm": 1.9375,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.6826,
|
|
"mean_token_accuracy": 0.05074650049209595,
|
|
"num_tokens": 165903.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.707062530517579,
|
|
"epoch": 0.07723761926397092,
|
|
"grad_norm": 2.078125,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.5739,
|
|
"mean_token_accuracy": 0.04917050153017044,
|
|
"num_tokens": 175812.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.685445880889892,
|
|
"epoch": 0.0817810086324398,
|
|
"grad_norm": 1.859375,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.4801,
|
|
"mean_token_accuracy": 0.04583706185221672,
|
|
"num_tokens": 186357.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.684220504760741,
|
|
"epoch": 0.08632439800090867,
|
|
"grad_norm": 1.9296875,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.3766,
|
|
"mean_token_accuracy": 0.04345137663185596,
|
|
"num_tokens": 196124.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.67547779083252,
|
|
"epoch": 0.09086778736937756,
|
|
"grad_norm": 1.9140625,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.2652,
|
|
"mean_token_accuracy": 0.04514526352286339,
|
|
"num_tokens": 205655.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.653054904937743,
|
|
"epoch": 0.09541117673784644,
|
|
"grad_norm": 1.8125,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.1821,
|
|
"mean_token_accuracy": 0.050565244257450105,
|
|
"num_tokens": 215870.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.627846431732177,
|
|
"epoch": 0.09995456610631531,
|
|
"grad_norm": 1.828125,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.0638,
|
|
"mean_token_accuracy": 0.05385896116495133,
|
|
"num_tokens": 225027.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.588682270050048,
|
|
"epoch": 0.10449795547478419,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.9803,
|
|
"mean_token_accuracy": 0.05429651960730553,
|
|
"num_tokens": 236352.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.560179710388184,
|
|
"epoch": 0.10904134484325306,
|
|
"grad_norm": 1.75,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.8435,
|
|
"mean_token_accuracy": 0.0539657898247242,
|
|
"num_tokens": 246713.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.498886871337891,
|
|
"epoch": 0.11358473421172194,
|
|
"grad_norm": 1.7109375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.6823,
|
|
"mean_token_accuracy": 0.05030508264899254,
|
|
"num_tokens": 256950.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.440382480621338,
|
|
"epoch": 0.11812812358019083,
|
|
"grad_norm": 1.6484375,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.6331,
|
|
"mean_token_accuracy": 0.05017514526844025,
|
|
"num_tokens": 267345.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.351335620880127,
|
|
"epoch": 0.1226715129486597,
|
|
"grad_norm": 2.0,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.4422,
|
|
"mean_token_accuracy": 0.054022491723299024,
|
|
"num_tokens": 276782.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.266972255706786,
|
|
"epoch": 0.12721490231712856,
|
|
"grad_norm": 1.59375,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.3194,
|
|
"mean_token_accuracy": 0.05632990375161171,
|
|
"num_tokens": 287024.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.132027339935302,
|
|
"epoch": 0.13175829168559747,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.2954,
|
|
"mean_token_accuracy": 0.05156242474913597,
|
|
"num_tokens": 296597.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 10.018408966064452,
|
|
"epoch": 0.13630168105406634,
|
|
"grad_norm": 1.453125,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.1037,
|
|
"mean_token_accuracy": 0.05491906628012657,
|
|
"num_tokens": 307709.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.880233860015869,
|
|
"epoch": 0.14084507042253522,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.0694,
|
|
"mean_token_accuracy": 0.05425432473421097,
|
|
"num_tokens": 317976.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.71201810836792,
|
|
"epoch": 0.1453884597910041,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.8984,
|
|
"mean_token_accuracy": 0.054642430320382115,
|
|
"num_tokens": 327319.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.520874309539796,
|
|
"epoch": 0.14993184915947297,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.8208,
|
|
"mean_token_accuracy": 0.05670260861515999,
|
|
"num_tokens": 337337.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.264504623413085,
|
|
"epoch": 0.15447523852794184,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.7862,
|
|
"mean_token_accuracy": 0.0514026153832674,
|
|
"num_tokens": 347779.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 9.049597454071044,
|
|
"epoch": 0.15901862789641072,
|
|
"grad_norm": 0.90234375,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.7529,
|
|
"mean_token_accuracy": 0.06217045783996582,
|
|
"num_tokens": 358088.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.819415187835693,
|
|
"epoch": 0.1635620172648796,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.6422,
|
|
"mean_token_accuracy": 0.059666012227535245,
|
|
"num_tokens": 368314.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.560069179534912,
|
|
"epoch": 0.16810540663334847,
|
|
"grad_norm": 0.828125,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.5159,
|
|
"mean_token_accuracy": 0.07182336524128914,
|
|
"num_tokens": 378406.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.396781253814698,
|
|
"epoch": 0.17264879600181735,
|
|
"grad_norm": 0.78125,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.5248,
|
|
"mean_token_accuracy": 0.07036102451384067,
|
|
"num_tokens": 388912.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.250075912475586,
|
|
"epoch": 0.17719218537028622,
|
|
"grad_norm": 0.828125,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.5042,
|
|
"mean_token_accuracy": 0.07041607536375523,
|
|
"num_tokens": 399290.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.157392597198486,
|
|
"epoch": 0.18173557473875512,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.4089,
|
|
"mean_token_accuracy": 0.07651857286691666,
|
|
"num_tokens": 409757.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.099771499633789,
|
|
"epoch": 0.186278964107224,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.5407,
|
|
"mean_token_accuracy": 0.07969405055046082,
|
|
"num_tokens": 420182.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.01094584465027,
|
|
"epoch": 0.19082235347569287,
|
|
"grad_norm": 0.90625,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.4483,
|
|
"mean_token_accuracy": 0.07581862695515156,
|
|
"num_tokens": 429897.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 7.986596918106079,
|
|
"epoch": 0.19536574284416175,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.3481,
|
|
"mean_token_accuracy": 0.07829344719648361,
|
|
"num_tokens": 441131.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 7.942036485671997,
|
|
"epoch": 0.19990913221263062,
|
|
"grad_norm": 0.85546875,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.3161,
|
|
"mean_token_accuracy": 0.0775233581662178,
|
|
"num_tokens": 451894.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.821108770370484,
|
|
"epoch": 0.2044525215810995,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.2593,
|
|
"mean_token_accuracy": 0.07877892255783081,
|
|
"num_tokens": 462613.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.904732894897461,
|
|
"epoch": 0.20899591094956838,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.3536,
|
|
"mean_token_accuracy": 0.07972405850887299,
|
|
"num_tokens": 471606.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.824862718582153,
|
|
"epoch": 0.21353930031803725,
|
|
"grad_norm": 0.81640625,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.2458,
|
|
"mean_token_accuracy": 0.08337627723813057,
|
|
"num_tokens": 481559.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.791320323944092,
|
|
"epoch": 0.21808268968650613,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.2611,
|
|
"mean_token_accuracy": 0.08033445328474045,
|
|
"num_tokens": 492766.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.805171489715576,
|
|
"epoch": 0.222626079054975,
|
|
"grad_norm": 0.80859375,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.3194,
|
|
"mean_token_accuracy": 0.08032750971615314,
|
|
"num_tokens": 503357.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.816235494613648,
|
|
"epoch": 0.22716946842344388,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.1556,
|
|
"mean_token_accuracy": 0.08248411491513252,
|
|
"num_tokens": 513098.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.719041204452514,
|
|
"epoch": 0.23171285779191278,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.2404,
|
|
"mean_token_accuracy": 0.08086299896240234,
|
|
"num_tokens": 522807.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.781749677658081,
|
|
"epoch": 0.23625624716038165,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.259,
|
|
"mean_token_accuracy": 0.08047396913170815,
|
|
"num_tokens": 532670.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.758613204956054,
|
|
"epoch": 0.24079963652885053,
|
|
"grad_norm": 0.71875,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.2064,
|
|
"mean_token_accuracy": 0.07659858912229538,
|
|
"num_tokens": 543647.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.7667632579803465,
|
|
"epoch": 0.2453430258973194,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.2166,
|
|
"mean_token_accuracy": 0.08752154409885407,
|
|
"num_tokens": 553047.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.720886993408203,
|
|
"epoch": 0.24988641526578828,
|
|
"grad_norm": 0.79296875,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.2571,
|
|
"mean_token_accuracy": 0.0864359587430954,
|
|
"num_tokens": 563413.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.675430965423584,
|
|
"epoch": 0.25442980463425713,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.2182,
|
|
"mean_token_accuracy": 0.08417232409119606,
|
|
"num_tokens": 573790.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.737106895446777,
|
|
"epoch": 0.25897319400272606,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.1452,
|
|
"mean_token_accuracy": 0.08157877773046493,
|
|
"num_tokens": 582958.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.618109798431396,
|
|
"epoch": 0.26351658337119493,
|
|
"grad_norm": 0.99609375,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.0887,
|
|
"mean_token_accuracy": 0.09229333996772766,
|
|
"num_tokens": 592886.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.615639877319336,
|
|
"epoch": 0.2680599727396638,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.1444,
|
|
"mean_token_accuracy": 0.08253698348999024,
|
|
"num_tokens": 603692.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.630686283111572,
|
|
"epoch": 0.2726033621081327,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.1779,
|
|
"mean_token_accuracy": 0.08081393018364906,
|
|
"num_tokens": 613534.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.684085559844971,
|
|
"epoch": 0.27714675147660156,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.1265,
|
|
"mean_token_accuracy": 0.0869300678372383,
|
|
"num_tokens": 623862.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.708980464935303,
|
|
"epoch": 0.28169014084507044,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.2041,
|
|
"mean_token_accuracy": 0.0886037103831768,
|
|
"num_tokens": 633759.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.514765357971191,
|
|
"epoch": 0.2862335302135393,
|
|
"grad_norm": 0.859375,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.0843,
|
|
"mean_token_accuracy": 0.0886026717722416,
|
|
"num_tokens": 643553.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.49160099029541,
|
|
"epoch": 0.2907769195820082,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.0034,
|
|
"mean_token_accuracy": 0.08772724494338036,
|
|
"num_tokens": 653714.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.63023567199707,
|
|
"epoch": 0.29532030895047706,
|
|
"grad_norm": 0.92578125,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.033,
|
|
"mean_token_accuracy": 0.09123715907335281,
|
|
"num_tokens": 662817.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.539647436141967,
|
|
"epoch": 0.29986369831894594,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.0547,
|
|
"mean_token_accuracy": 0.0879720188677311,
|
|
"num_tokens": 672381.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.440394639968872,
|
|
"epoch": 0.3044070876874148,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 6.9355,
|
|
"mean_token_accuracy": 0.08899887949228287,
|
|
"num_tokens": 682525.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.532028865814209,
|
|
"epoch": 0.3089504770558837,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.0283,
|
|
"mean_token_accuracy": 0.08837513625621796,
|
|
"num_tokens": 691826.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.473609542846679,
|
|
"epoch": 0.31349386642435256,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.007,
|
|
"mean_token_accuracy": 0.0950367122888565,
|
|
"num_tokens": 701319.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.459249973297119,
|
|
"epoch": 0.31803725579282144,
|
|
"grad_norm": 1.453125,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.0947,
|
|
"mean_token_accuracy": 0.08535790592432022,
|
|
"num_tokens": 710888.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.56165566444397,
|
|
"epoch": 0.3225806451612903,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.0342,
|
|
"mean_token_accuracy": 0.09012823775410653,
|
|
"num_tokens": 722516.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.527815628051758,
|
|
"epoch": 0.3271240345297592,
|
|
"grad_norm": 1.65625,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.0169,
|
|
"mean_token_accuracy": 0.08682614788413048,
|
|
"num_tokens": 733058.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.45044322013855,
|
|
"epoch": 0.33166742389822806,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000182,
|
|
"loss": 7.0592,
|
|
"mean_token_accuracy": 0.09242936447262765,
|
|
"num_tokens": 744832.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.463720369338989,
|
|
"epoch": 0.33621081326669694,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 7.0049,
|
|
"mean_token_accuracy": 0.09356781840324402,
|
|
"num_tokens": 755387.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.43138747215271,
|
|
"epoch": 0.3407542026351658,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.000187,
|
|
"loss": 6.9355,
|
|
"mean_token_accuracy": 0.09147335067391396,
|
|
"num_tokens": 764455.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.4005152702331545,
|
|
"epoch": 0.3452975920036347,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 6.98,
|
|
"mean_token_accuracy": 0.09236079826951027,
|
|
"num_tokens": 774513.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.440145206451416,
|
|
"epoch": 0.34984098137210357,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.0819,
|
|
"mean_token_accuracy": 0.08978344351053238,
|
|
"num_tokens": 785540.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.4839293479919435,
|
|
"epoch": 0.35438437074057244,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.947,
|
|
"mean_token_accuracy": 0.09177564308047295,
|
|
"num_tokens": 795233.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.431085395812988,
|
|
"epoch": 0.35892776010904137,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 6.9969,
|
|
"mean_token_accuracy": 0.09167078733444214,
|
|
"num_tokens": 804855.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.446485424041748,
|
|
"epoch": 0.36347114947751025,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.9593,
|
|
"mean_token_accuracy": 0.0934673197567463,
|
|
"num_tokens": 814818.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.352738714218139,
|
|
"epoch": 0.3680145388459791,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.9751,
|
|
"mean_token_accuracy": 0.09469850584864617,
|
|
"num_tokens": 824622.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.417457246780396,
|
|
"epoch": 0.372557928214448,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.8664,
|
|
"mean_token_accuracy": 0.09812097549438477,
|
|
"num_tokens": 834813.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.318056726455689,
|
|
"epoch": 0.3771013175829169,
|
|
"grad_norm": 0.84765625,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.9516,
|
|
"mean_token_accuracy": 0.09538300186395646,
|
|
"num_tokens": 844887.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.351348257064819,
|
|
"epoch": 0.38164470695138575,
|
|
"grad_norm": 0.85546875,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.9529,
|
|
"mean_token_accuracy": 0.0924244724214077,
|
|
"num_tokens": 855430.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.276928663253784,
|
|
"epoch": 0.3861880963198546,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.794,
|
|
"mean_token_accuracy": 0.09646925926208497,
|
|
"num_tokens": 866052.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.296294832229615,
|
|
"epoch": 0.3907314856883235,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.9297,
|
|
"mean_token_accuracy": 0.09176526665687561,
|
|
"num_tokens": 877961.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.363173007965088,
|
|
"epoch": 0.3952748750567924,
|
|
"grad_norm": 0.87890625,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.8984,
|
|
"mean_token_accuracy": 0.1007479526102543,
|
|
"num_tokens": 888386.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.295883321762085,
|
|
"epoch": 0.39981826442526125,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.9439,
|
|
"mean_token_accuracy": 0.09971338957548141,
|
|
"num_tokens": 898272.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.353458261489868,
|
|
"epoch": 0.4043616537937301,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.9055,
|
|
"mean_token_accuracy": 0.10149330794811248,
|
|
"num_tokens": 908317.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.281500577926636,
|
|
"epoch": 0.408905043162199,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.867,
|
|
"mean_token_accuracy": 0.09293231442570686,
|
|
"num_tokens": 917970.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.282706642150879,
|
|
"epoch": 0.4134484325306679,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.9115,
|
|
"mean_token_accuracy": 0.0950246959924698,
|
|
"num_tokens": 928132.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.30927529335022,
|
|
"epoch": 0.41799182189913675,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.8424,
|
|
"mean_token_accuracy": 0.09944264888763428,
|
|
"num_tokens": 938092.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.37966022491455,
|
|
"epoch": 0.4225352112676056,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.9599,
|
|
"mean_token_accuracy": 0.09508244097232818,
|
|
"num_tokens": 948650.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.15450644493103,
|
|
"epoch": 0.4270786006360745,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.868,
|
|
"mean_token_accuracy": 0.10101077854633331,
|
|
"num_tokens": 960195.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.254466104507446,
|
|
"epoch": 0.4316219900045434,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.8956,
|
|
"mean_token_accuracy": 0.10054716616868972,
|
|
"num_tokens": 970608.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.369555997848511,
|
|
"epoch": 0.43616537937301225,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.8689,
|
|
"mean_token_accuracy": 0.09615805298089981,
|
|
"num_tokens": 981053.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.163026285171509,
|
|
"epoch": 0.4407087687414811,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.7827,
|
|
"mean_token_accuracy": 0.10568929314613343,
|
|
"num_tokens": 991560.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.18547511100769,
|
|
"epoch": 0.44525215810995,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.8472,
|
|
"mean_token_accuracy": 0.09352557733654976,
|
|
"num_tokens": 1002008.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.173881673812867,
|
|
"epoch": 0.4497955474784189,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.8306,
|
|
"mean_token_accuracy": 0.10517893880605697,
|
|
"num_tokens": 1012349.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.266806602478027,
|
|
"epoch": 0.45433893684688775,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.8159,
|
|
"mean_token_accuracy": 0.10154995843768119,
|
|
"num_tokens": 1022886.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.45433893684688775,
|
|
"eval_entropy": 7.165732025545696,
|
|
"eval_loss": 6.923176288604736,
|
|
"eval_mean_token_accuracy": 0.10160443299385004,
|
|
"eval_num_tokens": 1022886.0,
|
|
"eval_runtime": 2.056,
|
|
"eval_samples_per_second": 1670.696,
|
|
"eval_steps_per_second": 209.141,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 11000,
|
|
"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": 221422467317760.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|