495 lines
14 KiB
JSON
495 lines
14 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 1.0,
|
|
"eval_steps": 500,
|
|
"global_step": 2331,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 1.529844079017639,
|
|
"epoch": 0.021458076283461188,
|
|
"grad_norm": 2.6875,
|
|
"learning_rate": 6.282051282051282e-06,
|
|
"loss": 1.6806,
|
|
"mean_token_accuracy": 0.6537430156767369,
|
|
"num_tokens": 13027304.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 1.488607188165188,
|
|
"epoch": 0.042916152566922376,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 1.2692307692307693e-05,
|
|
"loss": 1.5052,
|
|
"mean_token_accuracy": 0.673151359707117,
|
|
"num_tokens": 26060152.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 1.3883284440636634,
|
|
"epoch": 0.06437422885038356,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 1.9102564102564103e-05,
|
|
"loss": 1.3595,
|
|
"mean_token_accuracy": 0.6914066426455975,
|
|
"num_tokens": 39093818.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 1.309543669819832,
|
|
"epoch": 0.08583230513384475,
|
|
"grad_norm": 0.91015625,
|
|
"learning_rate": 2.551282051282051e-05,
|
|
"loss": 1.2897,
|
|
"mean_token_accuracy": 0.7024075204133987,
|
|
"num_tokens": 52123695.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 1.2504261362552642,
|
|
"epoch": 0.10729038141730594,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 2.9996212711496212e-05,
|
|
"loss": 1.2349,
|
|
"mean_token_accuracy": 0.7132975848019123,
|
|
"num_tokens": 65156301.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 1.2115551654994487,
|
|
"epoch": 0.12874845770076712,
|
|
"grad_norm": 0.87890625,
|
|
"learning_rate": 2.992893632784019e-05,
|
|
"loss": 1.2009,
|
|
"mean_token_accuracy": 0.7190598705410958,
|
|
"num_tokens": 78189155.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 1.1798285143077374,
|
|
"epoch": 0.1502065339842283,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 2.9777932337106776e-05,
|
|
"loss": 1.1708,
|
|
"mean_token_accuracy": 0.7244446368515491,
|
|
"num_tokens": 91225387.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 1.1685248683393001,
|
|
"epoch": 0.1716646102676895,
|
|
"grad_norm": 0.82421875,
|
|
"learning_rate": 2.95440476317018e-05,
|
|
"loss": 1.1594,
|
|
"mean_token_accuracy": 0.7263273361325264,
|
|
"num_tokens": 104264395.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 1.1535844399034978,
|
|
"epoch": 0.1931226865511507,
|
|
"grad_norm": 0.83984375,
|
|
"learning_rate": 2.9228593933122738e-05,
|
|
"loss": 1.1456,
|
|
"mean_token_accuracy": 0.729079841375351,
|
|
"num_tokens": 117304006.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 1.14171948492527,
|
|
"epoch": 0.2145807628346119,
|
|
"grad_norm": 0.8046875,
|
|
"learning_rate": 2.8833340435285732e-05,
|
|
"loss": 1.1335,
|
|
"mean_token_accuracy": 0.7310337677598,
|
|
"num_tokens": 130330369.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"entropy": 1.1242267568409443,
|
|
"epoch": 0.23603883911807305,
|
|
"grad_norm": 0.81640625,
|
|
"learning_rate": 2.8360503882159413e-05,
|
|
"loss": 1.1166,
|
|
"mean_token_accuracy": 0.7342477452754974,
|
|
"num_tokens": 143357759.0,
|
|
"step": 550
|
|
},
|
|
{
|
|
"entropy": 1.1167650932073594,
|
|
"epoch": 0.25749691540153424,
|
|
"grad_norm": 0.87109375,
|
|
"learning_rate": 2.7812736135354192e-05,
|
|
"loss": 1.1092,
|
|
"mean_token_accuracy": 0.7356279300153256,
|
|
"num_tokens": 156390933.0,
|
|
"step": 600
|
|
},
|
|
{
|
|
"entropy": 1.1086778999865055,
|
|
"epoch": 0.27895499168499543,
|
|
"grad_norm": 0.8046875,
|
|
"learning_rate": 2.7193109301393063e-05,
|
|
"loss": 1.1008,
|
|
"mean_token_accuracy": 0.7369775851070881,
|
|
"num_tokens": 169421819.0,
|
|
"step": 650
|
|
},
|
|
{
|
|
"entropy": 1.0990354216098785,
|
|
"epoch": 0.3004130679684566,
|
|
"grad_norm": 0.78515625,
|
|
"learning_rate": 2.6505098502076353e-05,
|
|
"loss": 1.0916,
|
|
"mean_token_accuracy": 0.7388592520356179,
|
|
"num_tokens": 182458556.0,
|
|
"step": 700
|
|
},
|
|
{
|
|
"entropy": 1.095612591356039,
|
|
"epoch": 0.3218711442519178,
|
|
"grad_norm": 0.80859375,
|
|
"learning_rate": 2.575256238457122e-05,
|
|
"loss": 1.088,
|
|
"mean_token_accuracy": 0.7395522198081017,
|
|
"num_tokens": 195498872.0,
|
|
"step": 750
|
|
},
|
|
{
|
|
"entropy": 1.094855775386095,
|
|
"epoch": 0.343329220535379,
|
|
"grad_norm": 0.8125,
|
|
"learning_rate": 2.4939721480533394e-05,
|
|
"loss": 1.0875,
|
|
"mean_token_accuracy": 0.7393953828513622,
|
|
"num_tokens": 208529244.0,
|
|
"step": 800
|
|
},
|
|
{
|
|
"entropy": 1.086953050494194,
|
|
"epoch": 0.3647872968188402,
|
|
"grad_norm": 0.75390625,
|
|
"learning_rate": 2.4071134535632042e-05,
|
|
"loss": 1.0799,
|
|
"mean_token_accuracy": 0.7406895773112774,
|
|
"num_tokens": 221559622.0,
|
|
"step": 850
|
|
},
|
|
{
|
|
"entropy": 1.0810946352779864,
|
|
"epoch": 0.3862453731023014,
|
|
"grad_norm": 0.75390625,
|
|
"learning_rate": 2.3151672942231506e-05,
|
|
"loss": 1.0738,
|
|
"mean_token_accuracy": 0.7419975420832634,
|
|
"num_tokens": 234596401.0,
|
|
"step": 900
|
|
},
|
|
{
|
|
"entropy": 1.077113990187645,
|
|
"epoch": 0.4077034493857626,
|
|
"grad_norm": 0.8515625,
|
|
"learning_rate": 2.218649341862198e-05,
|
|
"loss": 1.0697,
|
|
"mean_token_accuracy": 0.7425307250022888,
|
|
"num_tokens": 247619938.0,
|
|
"step": 950
|
|
},
|
|
{
|
|
"entropy": 1.0798053552210332,
|
|
"epoch": 0.4291615256692238,
|
|
"grad_norm": 0.78125,
|
|
"learning_rate": 2.1181009088025206e-05,
|
|
"loss": 1.0725,
|
|
"mean_token_accuracy": 0.7423897469043732,
|
|
"num_tokens": 260654172.0,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"entropy": 1.0781676921248435,
|
|
"epoch": 0.45061960195268497,
|
|
"grad_norm": 0.76171875,
|
|
"learning_rate": 2.014085911957592e-05,
|
|
"loss": 1.0711,
|
|
"mean_token_accuracy": 0.7425947175920009,
|
|
"num_tokens": 273684574.0,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"entropy": 1.0795374807715417,
|
|
"epoch": 0.4720776782361461,
|
|
"grad_norm": 0.8203125,
|
|
"learning_rate": 1.907187710154509e-05,
|
|
"loss": 1.0717,
|
|
"mean_token_accuracy": 0.7422310309112072,
|
|
"num_tokens": 286722109.0,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"entropy": 1.0711059863865375,
|
|
"epoch": 0.4935357545196073,
|
|
"grad_norm": 0.74609375,
|
|
"learning_rate": 1.7980058324180654e-05,
|
|
"loss": 1.0637,
|
|
"mean_token_accuracy": 0.7439769515395165,
|
|
"num_tokens": 299759049.0,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"entropy": 1.0707093475759029,
|
|
"epoch": 0.5149938308030685,
|
|
"grad_norm": 0.75,
|
|
"learning_rate": 1.687152615565724e-05,
|
|
"loss": 1.0631,
|
|
"mean_token_accuracy": 0.7438222867250442,
|
|
"num_tokens": 312785989.0,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"entropy": 1.0649618780612946,
|
|
"epoch": 0.5364519070865297,
|
|
"grad_norm": 0.7421875,
|
|
"learning_rate": 1.5752497699712207e-05,
|
|
"loss": 1.0577,
|
|
"mean_token_accuracy": 0.7450838792324066,
|
|
"num_tokens": 325818163.0,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"entropy": 1.0698006546497345,
|
|
"epoch": 0.5579099833699909,
|
|
"grad_norm": 0.76171875,
|
|
"learning_rate": 1.4629248927574217e-05,
|
|
"loss": 1.062,
|
|
"mean_token_accuracy": 0.7441214697062969,
|
|
"num_tokens": 338855461.0,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"entropy": 1.0641462428867816,
|
|
"epoch": 0.5793680596534521,
|
|
"grad_norm": 0.7109375,
|
|
"learning_rate": 1.3508079479738632e-05,
|
|
"loss": 1.057,
|
|
"mean_token_accuracy": 0.7451571583747864,
|
|
"num_tokens": 351888193.0,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"entropy": 1.0668337062001227,
|
|
"epoch": 0.6008261359369133,
|
|
"grad_norm": 0.7109375,
|
|
"learning_rate": 1.2395277334996045e-05,
|
|
"loss": 1.0593,
|
|
"mean_token_accuracy": 0.7446440425515175,
|
|
"num_tokens": 364920249.0,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"entropy": 1.0648700632154942,
|
|
"epoch": 0.6222842122203744,
|
|
"grad_norm": 0.73046875,
|
|
"learning_rate": 1.1297083544864386e-05,
|
|
"loss": 1.0574,
|
|
"mean_token_accuracy": 0.7448987066745758,
|
|
"num_tokens": 377954848.0,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"entropy": 1.0625084039568902,
|
|
"epoch": 0.6437422885038356,
|
|
"grad_norm": 0.7421875,
|
|
"learning_rate": 1.0219657231208455e-05,
|
|
"loss": 1.0555,
|
|
"mean_token_accuracy": 0.7454155634343624,
|
|
"num_tokens": 390982547.0,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"entropy": 1.0593435187637805,
|
|
"epoch": 0.6652003647872968,
|
|
"grad_norm": 0.71484375,
|
|
"learning_rate": 9.1690410433546e-06,
|
|
"loss": 1.0516,
|
|
"mean_token_accuracy": 0.7464735677838326,
|
|
"num_tokens": 404012732.0,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"entropy": 1.0650947599112988,
|
|
"epoch": 0.686658441070758,
|
|
"grad_norm": 0.68359375,
|
|
"learning_rate": 8.151127268431263e-06,
|
|
"loss": 1.058,
|
|
"mean_token_accuracy": 0.744667494893074,
|
|
"num_tokens": 417050367.0,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"entropy": 1.061756199300289,
|
|
"epoch": 0.7081165173542192,
|
|
"grad_norm": 0.72265625,
|
|
"learning_rate": 7.171624785002444e-06,
|
|
"loss": 1.0541,
|
|
"mean_token_accuracy": 0.7454431113600731,
|
|
"num_tokens": 430085128.0,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"entropy": 1.066297420412302,
|
|
"epoch": 0.7295745936376804,
|
|
"grad_norm": 0.7109375,
|
|
"learning_rate": 6.236027045331792e-06,
|
|
"loss": 1.0591,
|
|
"mean_token_accuracy": 0.7445771189033985,
|
|
"num_tokens": 443116587.0,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"entropy": 1.0651886078715325,
|
|
"epoch": 0.7510326699211416,
|
|
"grad_norm": 0.70703125,
|
|
"learning_rate": 5.349581265845804e-06,
|
|
"loss": 1.0572,
|
|
"mean_token_accuracy": 0.7450574149191379,
|
|
"num_tokens": 456150190.0,
|
|
"step": 1750
|
|
},
|
|
{
|
|
"entropy": 1.0632556484639644,
|
|
"epoch": 0.7724907462046028,
|
|
"grad_norm": 0.7109375,
|
|
"learning_rate": 4.517258998588792e-06,
|
|
"loss": 1.0556,
|
|
"mean_token_accuracy": 0.7453512494266034,
|
|
"num_tokens": 469187645.0,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"entropy": 1.056969922631979,
|
|
"epoch": 0.793948822488064,
|
|
"grad_norm": 0.69140625,
|
|
"learning_rate": 3.743728248716951e-06,
|
|
"loss": 1.0491,
|
|
"mean_token_accuracy": 0.7466551867127419,
|
|
"num_tokens": 482216628.0,
|
|
"step": 1850
|
|
},
|
|
{
|
|
"entropy": 1.0633544081449509,
|
|
"epoch": 0.8154068987715252,
|
|
"grad_norm": 0.72265625,
|
|
"learning_rate": 3.033327294407988e-06,
|
|
"loss": 1.0554,
|
|
"mean_token_accuracy": 0.7453896969556808,
|
|
"num_tokens": 495240948.0,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"entropy": 1.0602363917231559,
|
|
"epoch": 0.8368649750549864,
|
|
"grad_norm": 0.69921875,
|
|
"learning_rate": 2.390040356015151e-06,
|
|
"loss": 1.0531,
|
|
"mean_token_accuracy": 0.7463011787831784,
|
|
"num_tokens": 508270012.0,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"entropy": 1.0584601506590843,
|
|
"epoch": 0.8583230513384476,
|
|
"grad_norm": 0.7265625,
|
|
"learning_rate": 1.817475250922957e-06,
|
|
"loss": 1.0512,
|
|
"mean_token_accuracy": 0.7463495220243931,
|
|
"num_tokens": 521303590.0,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"entropy": 1.0634522691369057,
|
|
"epoch": 0.8797811276219087,
|
|
"grad_norm": 0.70703125,
|
|
"learning_rate": 1.318843159425332e-06,
|
|
"loss": 1.0566,
|
|
"mean_token_accuracy": 0.7450326754152775,
|
|
"num_tokens": 534336739.0,
|
|
"step": 2050
|
|
},
|
|
{
|
|
"entropy": 1.0665240594744683,
|
|
"epoch": 0.9012392039053699,
|
|
"grad_norm": 0.73046875,
|
|
"learning_rate": 8.969406151075494e-07,
|
|
"loss": 1.0598,
|
|
"mean_token_accuracy": 0.7446090675890445,
|
|
"num_tokens": 547357092.0,
|
|
"step": 2100
|
|
},
|
|
{
|
|
"entropy": 1.0626434257626534,
|
|
"epoch": 0.9226972801888311,
|
|
"grad_norm": 0.7265625,
|
|
"learning_rate": 5.541338207371605e-07,
|
|
"loss": 1.055,
|
|
"mean_token_accuracy": 0.7456535014510155,
|
|
"num_tokens": 560392664.0,
|
|
"step": 2150
|
|
},
|
|
{
|
|
"entropy": 1.0659497360885144,
|
|
"epoch": 0.9441553564722922,
|
|
"grad_norm": 0.6875,
|
|
"learning_rate": 2.923453776269558e-07,
|
|
"loss": 1.0578,
|
|
"mean_token_accuracy": 0.7447389586269856,
|
|
"num_tokens": 573423859.0,
|
|
"step": 2200
|
|
},
|
|
{
|
|
"entropy": 1.0699594897031783,
|
|
"epoch": 0.9656134327557534,
|
|
"grad_norm": 0.74609375,
|
|
"learning_rate": 1.130435028970872e-07,
|
|
"loss": 1.0617,
|
|
"mean_token_accuracy": 0.7441424219310284,
|
|
"num_tokens": 586459673.0,
|
|
"step": 2250
|
|
},
|
|
{
|
|
"entropy": 1.0601028017699718,
|
|
"epoch": 0.9870715090392146,
|
|
"grad_norm": 0.7421875,
|
|
"learning_rate": 1.7233795110367935e-08,
|
|
"loss": 1.0515,
|
|
"mean_token_accuracy": 0.7460673457384109,
|
|
"num_tokens": 599492881.0,
|
|
"step": 2300
|
|
}
|
|
],
|
|
"logging_steps": 50,
|
|
"max_steps": 2331,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 1,
|
|
"save_steps": 50000,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 5.135994653597762e+18,
|
|
"train_batch_size": 2,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|