danielm1405/lr-1e-05-epochs-1.0-cbqa-exqa-mcqa-paraphrase-sentiment-struct-summ-topic_cls-ddfb4b10
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity
Files
main
lr-1e-05-epochs-1.0-cbqa-ex…/trainer_state.json
ModelHub XC 0e47024d0a 初始化项目，由ModelHub XC社区提供模型 
Model: danielm1405/lr-1e-05-epochs-1.0-cbqa-exqa-mcqa-paraphrase-sentiment-struct-summ-topic_cls-ddfb4b10
Source: Original Platform
2026-07-26 09:03:10 +08:00

21389 lines
602 KiB
JSON
Raw Permalink Blame History

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 100,
"global_step": 19236,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"entropy": 2.5625,
"epoch": 0.0005198585984612185,
"grad_norm": 173.00236171303246,
"learning_rate": 1.5570934256055365e-07,
"loss": 2.7872,
"mean_token_accuracy": 0.540319162607193,
"num_tokens": 1302398.0,
"step": 10
},
{
"entropy": 2.55703125,
"epoch": 0.001039717196922437,
"grad_norm": 653.7176263429927,
"learning_rate": 3.28719723183391e-07,
"loss": 2.7474,
"mean_token_accuracy": 0.5395934134721756,
"num_tokens": 2607011.0,
"step": 20
},
{
"entropy": 2.6375,
"epoch": 0.0015595757953836558,
"grad_norm": 309.8554490522318,
"learning_rate": 5.017301038062284e-07,
"loss": 2.5912,
"mean_token_accuracy": 0.5403846651315689,
"num_tokens": 3908763.0,
"step": 30
},
{
"entropy": 2.6125,
"epoch": 0.002079434393844874,
"grad_norm": 35.18628584005397,
"learning_rate": 6.747404844290659e-07,
"loss": 2.3518,
"mean_token_accuracy": 0.5480821102857589,
"num_tokens": 5212330.0,
"step": 40
},
{
"entropy": 2.50625,
"epoch": 0.0025992929923060928,
"grad_norm": 226.13138768538084,
"learning_rate": 8.477508650519031e-07,
"loss": 2.0986,
"mean_token_accuracy": 0.5735700488090515,
"num_tokens": 6514897.0,
"step": 50
},
{
"entropy": 2.52421875,
"epoch": 0.0031191515907673115,
"grad_norm": 92.22199266365335,
"learning_rate": 1.0207612456747405e-06,
"loss": 2.0158,
"mean_token_accuracy": 0.5808026522397995,
"num_tokens": 7815337.0,
"step": 60
},
{
"entropy": 2.5046875,
"epoch": 0.00363901018922853,
"grad_norm": 56.90052530018388,
"learning_rate": 1.193771626297578e-06,
"loss": 1.9662,
"mean_token_accuracy": 0.5781357795000076,
"num_tokens": 9118609.0,
"step": 70
},
{
"entropy": 2.45859375,
"epoch": 0.004158868787689748,
"grad_norm": 110.37944061193666,
"learning_rate": 1.3667820069204154e-06,
"loss": 1.8477,
"mean_token_accuracy": 0.5997050851583481,
"num_tokens": 10421069.0,
"step": 80
},
{
"entropy": 2.45,
"epoch": 0.004678727386150967,
"grad_norm": 8.376974716797964,
"learning_rate": 1.5397923875432527e-06,
"loss": 1.7918,
"mean_token_accuracy": 0.6016785383224488,
"num_tokens": 11723320.0,
"step": 90
},
{
"entropy": 2.4,
"epoch": 0.0051985859846121855,
"grad_norm": 21.91872220290646,
"learning_rate": 1.71280276816609e-06,
"loss": 1.7713,
"mean_token_accuracy": 0.6012939006090164,
"num_tokens": 13026929.0,
"step": 100
},
{
"epoch": 0.0051985859846121855,
"eval_entropy": 2.423046875,
"eval_loss": 1.78825044631958,
"eval_mean_token_accuracy": 0.5983363449573517,
"eval_num_tokens": 13026929.0,
"eval_runtime": 27.3617,
"eval_samples_per_second": 92.611,
"eval_steps_per_second": 1.462,
"step": 100
},
{
"entropy": 2.44296875,
"epoch": 0.005718444583073404,
"grad_norm": 14.375722688364052,
"learning_rate": 1.8858131487889276e-06,
"loss": 1.7987,
"mean_token_accuracy": 0.6076627820730209,
"num_tokens": 14328199.0,
"step": 110
},
{
"entropy": 2.4578125,
"epoch": 0.006238303181534623,
"grad_norm": 35.69041079714868,
"learning_rate": 2.058823529411765e-06,
"loss": 1.781,
"mean_token_accuracy": 0.6253956496715546,
"num_tokens": 15627503.0,
"step": 120
},
{
"entropy": 2.5,
"epoch": 0.006758161779995841,
"grad_norm": 8.400019539117023,
"learning_rate": 2.2318339100346025e-06,
"loss": 1.744,
"mean_token_accuracy": 0.6284304618835449,
"num_tokens": 16929040.0,
"step": 130
},
{
"entropy": 2.4125,
"epoch": 0.00727802037845706,
"grad_norm": 102.93389432304647,
"learning_rate": 2.4048442906574396e-06,
"loss": 1.6901,
"mean_token_accuracy": 0.6349290311336517,
"num_tokens": 18231059.0,
"step": 140
},
{
"entropy": 2.42890625,
"epoch": 0.007797878976918278,
"grad_norm": 19.272194888320747,
"learning_rate": 2.577854671280277e-06,
"loss": 1.6287,
"mean_token_accuracy": 0.6461764484643936,
"num_tokens": 19531943.0,
"step": 150
},
{
"entropy": 2.43984375,
"epoch": 0.008317737575379496,
"grad_norm": 13.665872736925815,
"learning_rate": 2.7508650519031143e-06,
"loss": 1.598,
"mean_token_accuracy": 0.6472737610340118,
"num_tokens": 20833427.0,
"step": 160
},
{
"entropy": 2.38046875,
"epoch": 0.008837596173840716,
"grad_norm": 6.236866955752348,
"learning_rate": 2.9238754325259515e-06,
"loss": 1.5571,
"mean_token_accuracy": 0.6489362061023712,
"num_tokens": 22134792.0,
"step": 170
},
{
"entropy": 2.346875,
"epoch": 0.009357454772301934,
"grad_norm": 10.72783773703553,
"learning_rate": 3.0968858131487894e-06,
"loss": 1.58,
"mean_token_accuracy": 0.6393533855676651,
"num_tokens": 23435954.0,
"step": 180
},
{
"entropy": 2.3265625,
"epoch": 0.009877313370763153,
"grad_norm": 6.569709907458895,
"learning_rate": 3.2698961937716266e-06,
"loss": 1.5188,
"mean_token_accuracy": 0.6510779619216919,
"num_tokens": 24738507.0,
"step": 190
},
{
"entropy": 2.28046875,
"epoch": 0.010397171969224371,
"grad_norm": 9.334773327696508,
"learning_rate": 3.442906574394464e-06,
"loss": 1.5178,
"mean_token_accuracy": 0.6474015206098557,
"num_tokens": 26038289.0,
"step": 200
},
{
"epoch": 0.010397171969224371,
"eval_entropy": 2.2421875,
"eval_loss": 1.5215675830841064,
"eval_mean_token_accuracy": 0.645319864153862,
"eval_num_tokens": 26038289.0,
"eval_runtime": 27.0164,
"eval_samples_per_second": 93.795,
"eval_steps_per_second": 1.481,
"step": 200
},
{
"entropy": 2.25234375,
"epoch": 0.010917030567685589,
"grad_norm": 4.213769521737247,
"learning_rate": 3.6159169550173012e-06,
"loss": 1.5103,
"mean_token_accuracy": 0.6461672693490982,
"num_tokens": 27340266.0,
"step": 210
},
{
"entropy": 2.20703125,
"epoch": 0.011436889166146809,
"grad_norm": 6.152961450075137,
"learning_rate": 3.7889273356401384e-06,
"loss": 1.5447,
"mean_token_accuracy": 0.6397118270397186,
"num_tokens": 28642681.0,
"step": 220
},
{
"entropy": 2.203125,
"epoch": 0.011956747764608026,
"grad_norm": 11.216069496591544,
"learning_rate": 3.961937716262976e-06,
"loss": 1.4805,
"mean_token_accuracy": 0.6506381094455719,
"num_tokens": 29940158.0,
"step": 230
},
{
"entropy": 2.1640625,
"epoch": 0.012476606363069246,
"grad_norm": 23.232179980775868,
"learning_rate": 4.1349480968858135e-06,
"loss": 1.5193,
"mean_token_accuracy": 0.6417156457901001,
"num_tokens": 31242697.0,
"step": 240
},
{
"entropy": 2.16640625,
"epoch": 0.012996464961530464,
"grad_norm": 3.682986193504935,
"learning_rate": 4.307958477508651e-06,
"loss": 1.477,
"mean_token_accuracy": 0.6546868979930878,
"num_tokens": 32544908.0,
"step": 250
},
{
"entropy": 2.13515625,
"epoch": 0.013516323559991682,
"grad_norm": 4.094939061374869,
"learning_rate": 4.480968858131488e-06,
"loss": 1.4717,
"mean_token_accuracy": 0.6522895842790604,
"num_tokens": 33846529.0,
"step": 260
},
{
"entropy": 2.1484375,
"epoch": 0.014036182158452901,
"grad_norm": 4.494363903979221,
"learning_rate": 4.653979238754326e-06,
"loss": 1.4341,
"mean_token_accuracy": 0.6627357840538025,
"num_tokens": 35146020.0,
"step": 270
},
{
"entropy": 2.1234375,
"epoch": 0.01455604075691412,
"grad_norm": 3.533342196500187,
"learning_rate": 4.826989619377163e-06,
"loss": 1.5111,
"mean_token_accuracy": 0.6450612932443619,
"num_tokens": 36447257.0,
"step": 280
},
{
"entropy": 2.1046875,
"epoch": 0.015075899355375337,
"grad_norm": 4.348783488102715,
"learning_rate": 5e-06,
"loss": 1.5094,
"mean_token_accuracy": 0.6462055832147598,
"num_tokens": 37747089.0,
"step": 290
},
{
"entropy": 2.087890625,
"epoch": 0.015595757953836557,
"grad_norm": 2.8121419333266195,
"learning_rate": 5.173010380622838e-06,
"loss": 1.5192,
"mean_token_accuracy": 0.6424348950386047,
"num_tokens": 39051277.0,
"step": 300
},
{
"epoch": 0.015595757953836557,
"eval_entropy": 2.0732421875,
"eval_loss": 1.4926369190216064,
"eval_mean_token_accuracy": 0.6481952622532845,
"eval_num_tokens": 39051277.0,
"eval_runtime": 26.983,
"eval_samples_per_second": 93.911,
"eval_steps_per_second": 1.482,
"step": 300
},
{
"entropy": 2.06484375,
"epoch": 0.016115616552297776,
"grad_norm": 3.620740264379483,
"learning_rate": 5.346020761245675e-06,
"loss": 1.489,
"mean_token_accuracy": 0.6494882613420486,
"num_tokens": 40353493.0,
"step": 310
},
{
"entropy": 2.1078125,
"epoch": 0.016635475150758992,
"grad_norm": 3.4673697042127993,
"learning_rate": 5.519031141868513e-06,
"loss": 1.4958,
"mean_token_accuracy": 0.6502266526222229,
"num_tokens": 41655616.0,
"step": 320
},
{
"entropy": 2.1,
"epoch": 0.017155333749220212,
"grad_norm": 3.305182337991637,
"learning_rate": 5.69204152249135e-06,
"loss": 1.4863,
"mean_token_accuracy": 0.6506027430295944,
"num_tokens": 42958382.0,
"step": 330
},
{
"entropy": 2.1,
"epoch": 0.01767519234768143,
"grad_norm": 4.067633697286574,
"learning_rate": 5.865051903114187e-06,
"loss": 1.4702,
"mean_token_accuracy": 0.6524096846580505,
"num_tokens": 44262634.0,
"step": 340
},
{
"entropy": 2.12734375,
"epoch": 0.018195050946142648,
"grad_norm": 3.339679300532951,
"learning_rate": 6.038062283737025e-06,
"loss": 1.4721,
"mean_token_accuracy": 0.6526845216751098,
"num_tokens": 45564135.0,
"step": 350
},
{
"entropy": 2.1171875,
"epoch": 0.018714909544603867,
"grad_norm": 3.5112303034353807,
"learning_rate": 6.211072664359862e-06,
"loss": 1.4668,
"mean_token_accuracy": 0.6524525493383407,
"num_tokens": 46865688.0,
"step": 360
},
{
"entropy": 2.10546875,
"epoch": 0.019234768143065087,
"grad_norm": 2.9098234613114666,
"learning_rate": 6.3840830449827e-06,
"loss": 1.4852,
"mean_token_accuracy": 0.6528201282024384,
"num_tokens": 48168510.0,
"step": 370
},
{
"entropy": 2.11640625,
"epoch": 0.019754626741526306,
"grad_norm": 3.026576245453802,
"learning_rate": 6.5570934256055376e-06,
"loss": 1.4949,
"mean_token_accuracy": 0.6487828433513642,
"num_tokens": 49471798.0,
"step": 380
},
{
"entropy": 2.084375,
"epoch": 0.020274485339987523,
"grad_norm": 3.122464166043758,
"learning_rate": 6.730103806228374e-06,
"loss": 1.4454,
"mean_token_accuracy": 0.6613551646471023,
"num_tokens": 50773003.0,
"step": 390
},
{
"entropy": 2.090234375,
"epoch": 0.020794343938448742,
"grad_norm": 2.957254330987465,
"learning_rate": 6.903114186851212e-06,
"loss": 1.494,
"mean_token_accuracy": 0.6510500997304917,
"num_tokens": 52075603.0,
"step": 400
},
{
"epoch": 0.020794343938448742,
"eval_entropy": 2.0990234375,
"eval_loss": 1.4928697347640991,
"eval_mean_token_accuracy": 0.6480976983904838,
"eval_num_tokens": 52075603.0,
"eval_runtime": 27.0663,
"eval_samples_per_second": 93.622,
"eval_steps_per_second": 1.478,
"step": 400
},
{
"entropy": 2.044140625,
"epoch": 0.021314202536909962,
"grad_norm": 3.2313976222489833,
"learning_rate": 7.0761245674740485e-06,
"loss": 1.4534,
"mean_token_accuracy": 0.6563481628894806,
"num_tokens": 53378860.0,
"step": 410
},
{
"entropy": 2.104296875,
"epoch": 0.021834061135371178,
"grad_norm": 2.7907876838766725,
"learning_rate": 7.249134948096886e-06,
"loss": 1.4751,
"mean_token_accuracy": 0.6535353243350983,
"num_tokens": 54681181.0,
"step": 420
},
{
"entropy": 2.08046875,
"epoch": 0.022353919733832397,
"grad_norm": 3.1907761057018686,
"learning_rate": 7.4221453287197245e-06,
"loss": 1.4886,
"mean_token_accuracy": 0.6495951235294342,
"num_tokens": 55984560.0,
"step": 430
},
{
"entropy": 2.069140625,
"epoch": 0.022873778332293617,
"grad_norm": 3.6963945948731474,
"learning_rate": 7.595155709342561e-06,
"loss": 1.5012,
"mean_token_accuracy": 0.6491354495286942,
"num_tokens": 57287048.0,
"step": 440
},
{
"entropy": 2.08671875,
"epoch": 0.023393636930754833,
"grad_norm": 2.8708539076795163,
"learning_rate": 7.768166089965399e-06,
"loss": 1.5128,
"mean_token_accuracy": 0.6483903974294662,
"num_tokens": 58589477.0,
"step": 450
},
{
"entropy": 2.0765625,
"epoch": 0.023913495529216053,
"grad_norm": 3.005058942171316,
"learning_rate": 7.941176470588236e-06,
"loss": 1.4855,
"mean_token_accuracy": 0.6497094810009003,
"num_tokens": 59892225.0,
"step": 460
},
{
"entropy": 2.059375,
"epoch": 0.024433354127677272,
"grad_norm": 2.682637351906441,
"learning_rate": 8.114186851211074e-06,
"loss": 1.4966,
"mean_token_accuracy": 0.6504409044981003,
"num_tokens": 61194819.0,
"step": 470
},
{
"entropy": 2.058984375,
"epoch": 0.024953212726138492,
"grad_norm": 2.9240298923460895,
"learning_rate": 8.287197231833911e-06,
"loss": 1.5386,
"mean_token_accuracy": 0.6411254972219467,
"num_tokens": 62497232.0,
"step": 480
},
{
"entropy": 2.05390625,
"epoch": 0.025473071324599708,
"grad_norm": 2.9675010255382825,
"learning_rate": 8.460207612456747e-06,
"loss": 1.4771,
"mean_token_accuracy": 0.6519257396459579,
"num_tokens": 63798110.0,
"step": 490
},
{
"entropy": 2.0109375,
"epoch": 0.025992929923060928,
"grad_norm": 2.996099518295138,
"learning_rate": 8.633217993079585e-06,
"loss": 1.4362,
"mean_token_accuracy": 0.6603563964366913,
"num_tokens": 65101506.0,
"step": 500
},
{
"epoch": 0.025992929923060928,
"eval_entropy": 2.078515625,
"eval_loss": 1.490857720375061,
"eval_mean_token_accuracy": 0.6486845314502716,
"eval_num_tokens": 65101506.0,
"eval_runtime": 26.9649,
"eval_samples_per_second": 93.974,
"eval_steps_per_second": 1.483,
"step": 500
},
{
"entropy": 2.08203125,
"epoch": 0.026512788521522147,
"grad_norm": 2.953973470772088,
"learning_rate": 8.806228373702422e-06,
"loss": 1.4257,
"mean_token_accuracy": 0.6628826677799224,
"num_tokens": 66404425.0,
"step": 510
},
{
"entropy": 2.08984375,
"epoch": 0.027032647119983363,
"grad_norm": 2.733180753472642,
"learning_rate": 8.97923875432526e-06,
"loss": 1.5004,
"mean_token_accuracy": 0.6479833424091339,
"num_tokens": 67706316.0,
"step": 520
},
{
"entropy": 2.04453125,
"epoch": 0.027552505718444583,
"grad_norm": 2.907329412511417,
"learning_rate": 9.152249134948097e-06,
"loss": 1.4445,
"mean_token_accuracy": 0.6590858340263367,
"num_tokens": 69007161.0,
"step": 530
},
{
"entropy": 2.093359375,
"epoch": 0.028072364316905803,
"grad_norm": 3.6523070241011735,
"learning_rate": 9.325259515570935e-06,
"loss": 1.4712,
"mean_token_accuracy": 0.6559830635786057,
"num_tokens": 70309627.0,
"step": 540
},
{
"entropy": 2.0640625,
"epoch": 0.02859222291536702,
"grad_norm": 2.946331683196207,
"learning_rate": 9.498269896193773e-06,
"loss": 1.4714,
"mean_token_accuracy": 0.654650604724884,
"num_tokens": 71612466.0,
"step": 550
},
{
"entropy": 2.056640625,
"epoch": 0.02911208151382824,
"grad_norm": 3.2466724015437967,
"learning_rate": 9.67128027681661e-06,
"loss": 1.4822,
"mean_token_accuracy": 0.6531296879053116,
"num_tokens": 72912829.0,
"step": 560
},
{
"entropy": 2.061328125,
"epoch": 0.029631940112289458,
"grad_norm": 2.652578666886662,
"learning_rate": 9.844290657439448e-06,
"loss": 1.4826,
"mean_token_accuracy": 0.6525768607854843,
"num_tokens": 74214761.0,
"step": 570
},
{
"entropy": 2.05546875,
"epoch": 0.030151798710750674,
"grad_norm": 2.9016348623288635,
"learning_rate": 9.999464036874263e-06,
"loss": 1.4708,
"mean_token_accuracy": 0.6525696486234664,
"num_tokens": 75515289.0,
"step": 580
},
{
"entropy": 2.03671875,
"epoch": 0.030671657309211894,
"grad_norm": 2.9991876548536895,
"learning_rate": 9.994104405616894e-06,
"loss": 1.4584,
"mean_token_accuracy": 0.6555460274219513,
"num_tokens": 76815676.0,
"step": 590
},
{
"entropy": 2.0734375,
"epoch": 0.031191515907673113,
"grad_norm": 3.0405841067774557,
"learning_rate": 9.988744774359525e-06,
"loss": 1.4453,
"mean_token_accuracy": 0.66207515001297,
"num_tokens": 78119026.0,
"step": 600
},
{
"epoch": 0.031191515907673113,
"eval_entropy": 2.01875,
"eval_loss": 1.4894375801086426,
"eval_mean_token_accuracy": 0.6493151322007179,
"eval_num_tokens": 78119026.0,
"eval_runtime": 26.9704,
"eval_samples_per_second": 93.955,
"eval_steps_per_second": 1.483,
"step": 600
},
{
"entropy": 2.06640625,
"epoch": 0.03171137450613433,
"grad_norm": 2.9192745480529925,
"learning_rate": 9.983385143102156e-06,
"loss": 1.5119,
"mean_token_accuracy": 0.6473485946655273,
"num_tokens": 79420608.0,
"step": 610
},
{
"entropy": 2.0859375,
"epoch": 0.03223123310459555,
"grad_norm": 2.589615265030548,
"learning_rate": 9.978025511844785e-06,
"loss": 1.4652,
"mean_token_accuracy": 0.6533850491046905,
"num_tokens": 80722785.0,
"step": 620
},
{
"entropy": 2.09296875,
"epoch": 0.03275109170305677,
"grad_norm": 3.2433482337129527,
"learning_rate": 9.972665880587417e-06,
"loss": 1.4745,
"mean_token_accuracy": 0.650459548830986,
"num_tokens": 82025382.0,
"step": 630
},
{
"entropy": 2.094140625,
"epoch": 0.033270950301517985,
"grad_norm": 2.712870760818119,
"learning_rate": 9.967306249330047e-06,
"loss": 1.515,
"mean_token_accuracy": 0.6503280311822891,
"num_tokens": 83326986.0,
"step": 640
},
{
"entropy": 2.0515625,
"epoch": 0.03379080889997921,
"grad_norm": 2.6051431491720147,
"learning_rate": 9.961946618072677e-06,
"loss": 1.4609,
"mean_token_accuracy": 0.6562632411718369,
"num_tokens": 84628182.0,
"step": 650
},
{
"entropy": 2.046875,
"epoch": 0.034310667498440424,
"grad_norm": 2.854435184251635,
"learning_rate": 9.956586986815307e-06,
"loss": 1.5052,
"mean_token_accuracy": 0.6472997099161149,
"num_tokens": 85932409.0,
"step": 660
},
{
"entropy": 2.040625,
"epoch": 0.03483052609690164,
"grad_norm": 3.098754552074033,
"learning_rate": 9.951227355557939e-06,
"loss": 1.4143,
"mean_token_accuracy": 0.6645424753427506,
"num_tokens": 87234319.0,
"step": 670
},
{
"entropy": 2.02578125,
"epoch": 0.03535038469536286,
"grad_norm": 2.9414320331435073,
"learning_rate": 9.945867724300569e-06,
"loss": 1.5021,
"mean_token_accuracy": 0.64835484623909,
"num_tokens": 88535010.0,
"step": 680
},
{
"entropy": 2.0796875,
"epoch": 0.03587024329382408,
"grad_norm": 2.7934130993280655,
"learning_rate": 9.940508093043199e-06,
"loss": 1.4763,
"mean_token_accuracy": 0.6552160143852234,
"num_tokens": 89835746.0,
"step": 690
},
{
"entropy": 2.042578125,
"epoch": 0.036390101892285295,
"grad_norm": 3.1347359546773257,
"learning_rate": 9.93514846178583e-06,
"loss": 1.4826,
"mean_token_accuracy": 0.6510707646608352,
"num_tokens": 91137871.0,
"step": 700
},
{
"epoch": 0.036390101892285295,
"eval_entropy": 1.998046875,
"eval_loss": 1.4857159852981567,
"eval_mean_token_accuracy": 0.650704437494278,
"eval_num_tokens": 91137871.0,
"eval_runtime": 27.0065,
"eval_samples_per_second": 93.829,
"eval_steps_per_second": 1.481,
"step": 700
},
{
"entropy": 2.00859375,
"epoch": 0.03690996049074652,
"grad_norm": 3.1281660190423155,
"learning_rate": 9.929788830528461e-06,
"loss": 1.466,
"mean_token_accuracy": 0.6572742462158203,
"num_tokens": 92441464.0,
"step": 710
},
{
"entropy": 2.01328125,
"epoch": 0.037429819089207735,
"grad_norm": 2.6731307280447743,
"learning_rate": 9.924429199271091e-06,
"loss": 1.4701,
"mean_token_accuracy": 0.6554497301578521,
"num_tokens": 93741997.0,
"step": 720
},
{
"entropy": 2.009765625,
"epoch": 0.03794967768766895,
"grad_norm": 3.5412729207727267,
"learning_rate": 9.919069568013721e-06,
"loss": 1.4436,
"mean_token_accuracy": 0.6607381105422974,
"num_tokens": 95045372.0,
"step": 730
},
{
"entropy": 2.0296875,
"epoch": 0.038469536286130174,
"grad_norm": 4.46878809974768,
"learning_rate": 9.913709936756351e-06,
"loss": 1.434,
"mean_token_accuracy": 0.658992001414299,
"num_tokens": 96348340.0,
"step": 740
},
{
"entropy": 2.030859375,
"epoch": 0.03898939488459139,
"grad_norm": 3.123206546686491,
"learning_rate": 9.908350305498983e-06,
"loss": 1.4309,
"mean_token_accuracy": 0.6615214139223099,
"num_tokens": 97651304.0,
"step": 750
},
{
"entropy": 2.00390625,
"epoch": 0.03950925348305261,
"grad_norm": 2.7030837645374404,
"learning_rate": 9.902990674241613e-06,
"loss": 1.467,
"mean_token_accuracy": 0.6522508352994919,
"num_tokens": 98953989.0,
"step": 760
},
{
"entropy": 2.048828125,
"epoch": 0.04002911208151383,
"grad_norm": 2.9305573860007073,
"learning_rate": 9.897631042984243e-06,
"loss": 1.434,
"mean_token_accuracy": 0.6595852017402649,
"num_tokens": 100258068.0,
"step": 770
},
{
"entropy": 2.040625,
"epoch": 0.040548970679975045,
"grad_norm": 3.775025173199993,
"learning_rate": 9.892271411726874e-06,
"loss": 1.4698,
"mean_token_accuracy": 0.6581872284412384,
"num_tokens": 101559899.0,
"step": 780
},
{
"entropy": 2.01328125,
"epoch": 0.04106882927843627,
"grad_norm": 2.7597164411886275,
"learning_rate": 9.886911780469505e-06,
"loss": 1.4596,
"mean_token_accuracy": 0.6569606214761734,
"num_tokens": 102859594.0,
"step": 790
},
{
"entropy": 2.049609375,
"epoch": 0.041588687876897484,
"grad_norm": 2.8475350511506003,
"learning_rate": 9.881552149212136e-06,
"loss": 1.492,
"mean_token_accuracy": 0.6502189099788666,
"num_tokens": 104160615.0,
"step": 800
},
{
"epoch": 0.041588687876897484,
"eval_entropy": 2.01640625,
"eval_loss": 1.47556734085083,
"eval_mean_token_accuracy": 0.6519631192088127,
"eval_num_tokens": 104160615.0,
"eval_runtime": 27.0965,
"eval_samples_per_second": 93.518,
"eval_steps_per_second": 1.476,
"step": 800
},
{
"entropy": 2.02421875,
"epoch": 0.0421085464753587,
"grad_norm": 3.1643526915570437,
"learning_rate": 9.876192517954766e-06,
"loss": 1.4674,
"mean_token_accuracy": 0.6531112939119339,
"num_tokens": 105461759.0,
"step": 810
},
{
"entropy": 2.029296875,
"epoch": 0.042628405073819924,
"grad_norm": 3.0880496339447476,
"learning_rate": 9.870832886697396e-06,
"loss": 1.4181,
"mean_token_accuracy": 0.6645083844661712,
"num_tokens": 106765190.0,
"step": 820
},
{
"entropy": 1.994140625,
"epoch": 0.04314826367228114,
"grad_norm": 3.0587713049268914,
"learning_rate": 9.865473255440028e-06,
"loss": 1.4646,
"mean_token_accuracy": 0.6576838165521621,
"num_tokens": 108066682.0,
"step": 830
},
{
"entropy": 2.05,
"epoch": 0.043668122270742356,
"grad_norm": 3.005792270052235,
"learning_rate": 9.860113624182656e-06,
"loss": 1.4514,
"mean_token_accuracy": 0.6616773635149003,
"num_tokens": 109363733.0,
"step": 840
},
{
"entropy": 2.053125,
"epoch": 0.04418798086920358,
"grad_norm": 3.308815827315186,
"learning_rate": 9.854753992925288e-06,
"loss": 1.4562,
"mean_token_accuracy": 0.6562087327241898,
"num_tokens": 110666312.0,
"step": 850
},
{
"entropy": 2.059765625,
"epoch": 0.044707839467664795,
"grad_norm": 2.6047954855373994,
"learning_rate": 9.849394361667918e-06,
"loss": 1.4116,
"mean_token_accuracy": 0.6656414300203324,
"num_tokens": 111968752.0,
"step": 860
},
{
"entropy": 2.055859375,
"epoch": 0.04522769806612601,
"grad_norm": 2.7988446023477844,
"learning_rate": 9.844034730410548e-06,
"loss": 1.4906,
"mean_token_accuracy": 0.6486280649900437,
"num_tokens": 113270685.0,
"step": 870
},
{
"entropy": 2.1,
"epoch": 0.045747556664587234,
"grad_norm": 2.9936820179647863,
"learning_rate": 9.838675099153178e-06,
"loss": 1.4677,
"mean_token_accuracy": 0.6553049117326737,
"num_tokens": 114573889.0,
"step": 880
},
{
"entropy": 2.0625,
"epoch": 0.04626741526304845,
"grad_norm": 2.660892926972615,
"learning_rate": 9.83331546789581e-06,
"loss": 1.4419,
"mean_token_accuracy": 0.657513964176178,
"num_tokens": 115874810.0,
"step": 890
},
{
"entropy": 2.08359375,
"epoch": 0.046787273861509666,
"grad_norm": 2.6736521284345756,
"learning_rate": 9.82795583663844e-06,
"loss": 1.4846,
"mean_token_accuracy": 0.6504478216171264,
"num_tokens": 117178596.0,
"step": 900
},
{
"epoch": 0.046787273861509666,
"eval_entropy": 2.0328125,
"eval_loss": 1.4726715087890625,
"eval_mean_token_accuracy": 0.6528749153017998,
"eval_num_tokens": 117178596.0,
"eval_runtime": 27.0141,
"eval_samples_per_second": 93.803,
"eval_steps_per_second": 1.481,
"step": 900
},
{
"entropy": 2.03984375,
"epoch": 0.04730713245997089,
"grad_norm": 3.024566702107581,
"learning_rate": 9.82259620538107e-06,
"loss": 1.4614,
"mean_token_accuracy": 0.6563761234283447,
"num_tokens": 118478198.0,
"step": 910
},
{
"entropy": 2.0453125,
"epoch": 0.047826991058432106,
"grad_norm": 2.9173412533253553,
"learning_rate": 9.8172365741237e-06,
"loss": 1.4752,
"mean_token_accuracy": 0.6559763848781586,
"num_tokens": 119779511.0,
"step": 920
},
{
"entropy": 2.0640625,
"epoch": 0.04834684965689332,
"grad_norm": 2.716932057470356,
"learning_rate": 9.811876942866332e-06,
"loss": 1.4261,
"mean_token_accuracy": 0.6628080755472183,
"num_tokens": 121080412.0,
"step": 930
},
{
"entropy": 2.0265625,
"epoch": 0.048866708255354545,
"grad_norm": 2.7608920276304247,
"learning_rate": 9.806517311608962e-06,
"loss": 1.4482,
"mean_token_accuracy": 0.6566795647144318,
"num_tokens": 122381602.0,
"step": 940
},
{
"entropy": 2.05546875,
"epoch": 0.04938656685381576,
"grad_norm": 2.455582169305633,
"learning_rate": 9.801157680351592e-06,
"loss": 1.4046,
"mean_token_accuracy": 0.6685665279626847,
"num_tokens": 123683161.0,
"step": 950
},
{
"entropy": 1.99921875,
"epoch": 0.049906425452276984,
"grad_norm": 3.1389104392029616,
"learning_rate": 9.795798049094223e-06,
"loss": 1.4335,
"mean_token_accuracy": 0.6617822855710983,
"num_tokens": 124985833.0,
"step": 960
},
{
"entropy": 2.075,
"epoch": 0.0504262840507382,
"grad_norm": 2.757314412461097,
"learning_rate": 9.790438417836854e-06,
"loss": 1.4745,
"mean_token_accuracy": 0.6555801630020142,
"num_tokens": 126289017.0,
"step": 970
},
{
"entropy": 1.99921875,
"epoch": 0.050946142649199416,
"grad_norm": 2.845074834679372,
"learning_rate": 9.785078786579485e-06,
"loss": 1.4214,
"mean_token_accuracy": 0.6624303042888642,
"num_tokens": 127592117.0,
"step": 980
},
{
"entropy": 2.040625,
"epoch": 0.05146600124766064,
"grad_norm": 2.9003252490845415,
"learning_rate": 9.779719155322115e-06,
"loss": 1.4158,
"mean_token_accuracy": 0.6657815873622894,
"num_tokens": 128894144.0,
"step": 990
},
{
"entropy": 2.00234375,
"epoch": 0.051985859846121855,
"grad_norm": 2.356195354677558,
"learning_rate": 9.774359524064745e-06,
"loss": 1.4532,
"mean_token_accuracy": 0.6576653867959976,
"num_tokens": 130194963.0,
"step": 1000
},
{
"epoch": 0.051985859846121855,
"eval_entropy": 2.0388671875,
"eval_loss": 1.4649391174316406,
"eval_mean_token_accuracy": 0.6541767969727517,
"eval_num_tokens": 130194963.0,
"eval_runtime": 27.0417,
"eval_samples_per_second": 93.707,
"eval_steps_per_second": 1.479,
"step": 1000
},
{
"entropy": 2.062109375,
"epoch": 0.05250571844458307,
"grad_norm": 2.716269511162931,
"learning_rate": 9.768999892807377e-06,
"loss": 1.4396,
"mean_token_accuracy": 0.6611425369977951,
"num_tokens": 131496540.0,
"step": 1010
},
{
"entropy": 2.052734375,
"epoch": 0.053025577043044295,
"grad_norm": 2.598052948515733,
"learning_rate": 9.763640261550007e-06,
"loss": 1.4344,
"mean_token_accuracy": 0.6610976606607437,
"num_tokens": 132797502.0,
"step": 1020
},
{
"entropy": 2.048828125,
"epoch": 0.05354543564150551,
"grad_norm": 3.587080011212248,
"learning_rate": 9.758280630292637e-06,
"loss": 1.4678,
"mean_token_accuracy": 0.6552032768726349,
"num_tokens": 134100264.0,
"step": 1030
},
{
"entropy": 2.064453125,
"epoch": 0.05406529423996673,
"grad_norm": 2.981961179175914,
"learning_rate": 9.752920999035267e-06,
"loss": 1.4337,
"mean_token_accuracy": 0.6622271060943603,
"num_tokens": 135400359.0,
"step": 1040
},
{
"entropy": 1.99140625,
"epoch": 0.05458515283842795,
"grad_norm": 2.6027976873368988,
"learning_rate": 9.747561367777899e-06,
"loss": 1.4854,
"mean_token_accuracy": 0.6532850801944733,
"num_tokens": 136700864.0,
"step": 1050
},
{
"entropy": 1.973828125,
"epoch": 0.055105011436889166,
"grad_norm": 2.875961106963352,
"learning_rate": 9.742201736520527e-06,
"loss": 1.4088,
"mean_token_accuracy": 0.6645298361778259,
"num_tokens": 138001499.0,
"step": 1060
},
{
"entropy": 1.975390625,
"epoch": 0.05562487003535038,
"grad_norm": 2.800398345970527,
"learning_rate": 9.736842105263159e-06,
"loss": 1.4441,
"mean_token_accuracy": 0.6606097936630249,
"num_tokens": 139302971.0,
"step": 1070
},
{
"entropy": 1.985546875,
"epoch": 0.056144728633811605,
"grad_norm": 2.4865572825619413,
"learning_rate": 9.731482474005789e-06,
"loss": 1.3953,
"mean_token_accuracy": 0.6689865738153458,
"num_tokens": 140605374.0,
"step": 1080
},
{
"entropy": 2.001953125,
"epoch": 0.05666458723227282,
"grad_norm": 2.8653859880508623,
"learning_rate": 9.72612284274842e-06,
"loss": 1.4369,
"mean_token_accuracy": 0.6611711531877518,
"num_tokens": 141905165.0,
"step": 1090
},
{
"entropy": 2.037109375,
"epoch": 0.05718444583073404,
"grad_norm": 2.579525013369721,
"learning_rate": 9.72076321149105e-06,
"loss": 1.4235,
"mean_token_accuracy": 0.6633508086204529,
"num_tokens": 143205963.0,
"step": 1100
},
{
"epoch": 0.05718444583073404,
"eval_entropy": 2.0279296875,
"eval_loss": 1.458391785621643,
"eval_mean_token_accuracy": 0.6556931018829346,
"eval_num_tokens": 143205963.0,
"eval_runtime": 27.0163,
"eval_samples_per_second": 93.795,
"eval_steps_per_second": 1.481,
"step": 1100
},
{
"entropy": 2.01640625,
"epoch": 0.05770430442919526,
"grad_norm": 2.817300861064028,
"learning_rate": 9.715403580233681e-06,
"loss": 1.4735,
"mean_token_accuracy": 0.6563478857278824,
"num_tokens": 144507599.0,
"step": 1110
},
{
"entropy": 2.026171875,
"epoch": 0.05822416302765648,
"grad_norm": 2.733320765797777,
"learning_rate": 9.710043948976311e-06,
"loss": 1.4069,
"mean_token_accuracy": 0.6660541325807572,
"num_tokens": 145810141.0,
"step": 1120
},
{
"entropy": 2.04453125,
"epoch": 0.05874402162611769,
"grad_norm": 2.6476521431066415,
"learning_rate": 9.704684317718941e-06,
"loss": 1.4249,
"mean_token_accuracy": 0.6627670258283616,
"num_tokens": 147112414.0,
"step": 1130
},
{
"entropy": 2.065234375,
"epoch": 0.059263880224578916,
"grad_norm": 2.7740874118929693,
"learning_rate": 9.699324686461572e-06,
"loss": 1.4339,
"mean_token_accuracy": 0.6647266983985901,
"num_tokens": 148416498.0,
"step": 1140
},
{
"entropy": 2.05859375,
"epoch": 0.05978373882304013,
"grad_norm": 3.103989604370861,
"learning_rate": 9.693965055204203e-06,
"loss": 1.3971,
"mean_token_accuracy": 0.667517551779747,
"num_tokens": 149719140.0,
"step": 1150
},
{
"entropy": 2.07265625,
"epoch": 0.06030359742150135,
"grad_norm": 2.432109908088185,
"learning_rate": 9.688605423946833e-06,
"loss": 1.4384,
"mean_token_accuracy": 0.6609664559364319,
"num_tokens": 151019986.0,
"step": 1160
},
{
"entropy": 2.05546875,
"epoch": 0.06082345601996257,
"grad_norm": 2.5245605053903106,
"learning_rate": 9.683245792689464e-06,
"loss": 1.4765,
"mean_token_accuracy": 0.6526906937360764,
"num_tokens": 152322644.0,
"step": 1170
},
{
"entropy": 2.031640625,
"epoch": 0.06134331461842379,
"grad_norm": 2.7487928921874754,
"learning_rate": 9.677886161432094e-06,
"loss": 1.4488,
"mean_token_accuracy": 0.6603018581867218,
"num_tokens": 153626037.0,
"step": 1180
},
{
"entropy": 2.059765625,
"epoch": 0.06186317321688501,
"grad_norm": 2.9631985435015555,
"learning_rate": 9.672526530174726e-06,
"loss": 1.4751,
"mean_token_accuracy": 0.654591628909111,
"num_tokens": 154928300.0,
"step": 1190
},
{
"entropy": 2.05546875,
"epoch": 0.06238303181534623,
"grad_norm": 2.849220431526448,
"learning_rate": 9.667166898917356e-06,
"loss": 1.4476,
"mean_token_accuracy": 0.6598935633897781,
"num_tokens": 156231984.0,
"step": 1200
},
{
"epoch": 0.06238303181534623,
"eval_entropy": 2.001953125,
"eval_loss": 1.4500080347061157,
"eval_mean_token_accuracy": 0.6577001601457596,
"eval_num_tokens": 156231984.0,
"eval_runtime": 27.0663,
"eval_samples_per_second": 93.622,
"eval_steps_per_second": 1.478,
"step": 1200
},
{
"entropy": 2.047265625,
"epoch": 0.06290289041380745,
"grad_norm": 2.634615132737495,
"learning_rate": 9.661807267659986e-06,
"loss": 1.4365,
"mean_token_accuracy": 0.6642505705356598,
"num_tokens": 157534897.0,
"step": 1210
},
{
"entropy": 2.03671875,
"epoch": 0.06342274901226866,
"grad_norm": 2.7113712998791315,
"learning_rate": 9.656447636402616e-06,
"loss": 1.4474,
"mean_token_accuracy": 0.6567792683839798,
"num_tokens": 158836458.0,
"step": 1220
},
{
"entropy": 2.0171875,
"epoch": 0.06394260761072988,
"grad_norm": 2.878333567315318,
"learning_rate": 9.651088005145248e-06,
"loss": 1.4276,
"mean_token_accuracy": 0.6638014614582062,
"num_tokens": 160139143.0,
"step": 1230
},
{
"entropy": 2.061328125,
"epoch": 0.0644624662091911,
"grad_norm": 2.5974728287844555,
"learning_rate": 9.645728373887878e-06,
"loss": 1.4395,
"mean_token_accuracy": 0.6606919318437576,
"num_tokens": 161440221.0,
"step": 1240
},
{
"entropy": 2.0421875,
"epoch": 0.06498232480765231,
"grad_norm": 2.509222600707683,
"learning_rate": 9.640368742630508e-06,
"loss": 1.4125,
"mean_token_accuracy": 0.667571046948433,
"num_tokens": 162743424.0,
"step": 1250
},
{
"entropy": 2.06953125,
"epoch": 0.06550218340611354,
"grad_norm": 4.96634716219883,
"learning_rate": 9.635009111373138e-06,
"loss": 1.4061,
"mean_token_accuracy": 0.6698934137821198,
"num_tokens": 164045812.0,
"step": 1260
},
{
"entropy": 2.07890625,
"epoch": 0.06602204200457476,
"grad_norm": 2.6882780528795984,
"learning_rate": 9.62964948011577e-06,
"loss": 1.4323,
"mean_token_accuracy": 0.659720504283905,
"num_tokens": 165349641.0,
"step": 1270
},
{
"entropy": 2.05546875,
"epoch": 0.06654190060303597,
"grad_norm": 2.7689049397400973,
"learning_rate": 9.624289848858398e-06,
"loss": 1.394,
"mean_token_accuracy": 0.6697416126728057,
"num_tokens": 166650210.0,
"step": 1280
},
{
"entropy": 2.078125,
"epoch": 0.06706175920149719,
"grad_norm": 2.6772151859992346,
"learning_rate": 9.61893021760103e-06,
"loss": 1.4298,
"mean_token_accuracy": 0.6623325318098068,
"num_tokens": 167952353.0,
"step": 1290
},
{
"entropy": 2.019921875,
"epoch": 0.06758161779995842,
"grad_norm": 2.7985444284080394,
"learning_rate": 9.61357058634366e-06,
"loss": 1.4109,
"mean_token_accuracy": 0.6665020197629928,
"num_tokens": 169254892.0,
"step": 1300
},
{
"epoch": 0.06758161779995842,
"eval_entropy": 2.0751953125,
"eval_loss": 1.443481206893921,
"eval_mean_token_accuracy": 0.658372911810875,
"eval_num_tokens": 169254892.0,
"eval_runtime": 27.0712,
"eval_samples_per_second": 93.605,
"eval_steps_per_second": 1.478,
"step": 1300
},
{
"entropy": 2.058984375,
"epoch": 0.06810147639841962,
"grad_norm": 3.769690786712918,
"learning_rate": 9.60821095508629e-06,
"loss": 1.388,
"mean_token_accuracy": 0.6707593441009522,
"num_tokens": 170555970.0,
"step": 1310
},
{
"entropy": 2.0359375,
"epoch": 0.06862133499688085,
"grad_norm": 2.803096980297384,
"learning_rate": 9.60285132382892e-06,
"loss": 1.4074,
"mean_token_accuracy": 0.6653834074735642,
"num_tokens": 171855767.0,
"step": 1320
},
{
"entropy": 2.078125,
"epoch": 0.06914119359534207,
"grad_norm": 2.6438804468665404,
"learning_rate": 9.597491692571552e-06,
"loss": 1.4462,
"mean_token_accuracy": 0.6591643959283828,
"num_tokens": 173158594.0,
"step": 1330
},
{
"entropy": 2.03671875,
"epoch": 0.06966105219380328,
"grad_norm": 2.929393047646466,
"learning_rate": 9.592132061314182e-06,
"loss": 1.4309,
"mean_token_accuracy": 0.6621407389640808,
"num_tokens": 174459684.0,
"step": 1340
},
{
"entropy": 2.084375,
"epoch": 0.0701809107922645,
"grad_norm": 2.7909546896604875,
"learning_rate": 9.586772430056813e-06,
"loss": 1.4266,
"mean_token_accuracy": 0.6653582721948623,
"num_tokens": 175759779.0,
"step": 1350
},
{
"entropy": 2.052734375,
"epoch": 0.07070076939072573,
"grad_norm": 2.7175365018225737,
"learning_rate": 9.581412798799443e-06,
"loss": 1.3925,
"mean_token_accuracy": 0.6702443927526474,
"num_tokens": 177062874.0,
"step": 1360
},
{
"entropy": 2.039453125,
"epoch": 0.07122062798918694,
"grad_norm": 2.4908891014104237,
"learning_rate": 9.576053167542075e-06,
"loss": 1.3781,
"mean_token_accuracy": 0.6746690154075623,
"num_tokens": 178364954.0,
"step": 1370
},
{
"entropy": 2.09609375,
"epoch": 0.07174048658764816,
"grad_norm": 3.2112800815358726,
"learning_rate": 9.570693536284705e-06,
"loss": 1.4171,
"mean_token_accuracy": 0.6628296434879303,
"num_tokens": 179665812.0,
"step": 1380
},
{
"entropy": 2.014453125,
"epoch": 0.07226034518610938,
"grad_norm": 2.534586012522054,
"learning_rate": 9.565333905027335e-06,
"loss": 1.456,
"mean_token_accuracy": 0.6560030221939087,
"num_tokens": 180966747.0,
"step": 1390
},
{
"entropy": 1.981640625,
"epoch": 0.07278020378457059,
"grad_norm": 2.562779871243658,
"learning_rate": 9.559974273769965e-06,
"loss": 1.408,
"mean_token_accuracy": 0.6667282909154892,
"num_tokens": 182267345.0,
"step": 1400
},
{
"epoch": 0.07278020378457059,
"eval_entropy": 2.01953125,
"eval_loss": 1.4417048692703247,
"eval_mean_token_accuracy": 0.6589708909392357,
"eval_num_tokens": 182267345.0,
"eval_runtime": 26.9835,
"eval_samples_per_second": 93.909,
"eval_steps_per_second": 1.482,
"step": 1400
},
{
"entropy": 1.978125,
"epoch": 0.07330006238303181,
"grad_norm": 3.306965503665122,
"learning_rate": 9.554614642512597e-06,
"loss": 1.4175,
"mean_token_accuracy": 0.6639352053403854,
"num_tokens": 183569248.0,
"step": 1410
},
{
"entropy": 1.991015625,
"epoch": 0.07381992098149304,
"grad_norm": 2.941123092725448,
"learning_rate": 9.549255011255227e-06,
"loss": 1.398,
"mean_token_accuracy": 0.6701751917600631,
"num_tokens": 184871042.0,
"step": 1420
},
{
"entropy": 2.0328125,
"epoch": 0.07433977957995425,
"grad_norm": 2.6150032121028524,
"learning_rate": 9.543895379997857e-06,
"loss": 1.4336,
"mean_token_accuracy": 0.6624453902244568,
"num_tokens": 186174530.0,
"step": 1430
},
{
"entropy": 2.05859375,
"epoch": 0.07485963817841547,
"grad_norm": 2.7332956150312433,
"learning_rate": 9.538535748740487e-06,
"loss": 1.4201,
"mean_token_accuracy": 0.6651494354009628,
"num_tokens": 187476931.0,
"step": 1440
},
{
"entropy": 1.98984375,
"epoch": 0.07537949677687669,
"grad_norm": 2.4938129663780946,
"learning_rate": 9.533176117483119e-06,
"loss": 1.4176,
"mean_token_accuracy": 0.6665954530239105,
"num_tokens": 188776045.0,
"step": 1450
},
{
"entropy": 2.053125,
"epoch": 0.0758993553753379,
"grad_norm": 2.374623090360873,
"learning_rate": 9.527816486225749e-06,
"loss": 1.4135,
"mean_token_accuracy": 0.6634874373674393,
"num_tokens": 190078274.0,
"step": 1460
},
{
"entropy": 2.03671875,
"epoch": 0.07641921397379912,
"grad_norm": 2.6394887220491974,
"learning_rate": 9.52245685496838e-06,
"loss": 1.3691,
"mean_token_accuracy": 0.6745434761047363,
"num_tokens": 191379983.0,
"step": 1470
},
{
"entropy": 2.0328125,
"epoch": 0.07693907257226035,
"grad_norm": 2.4343024058513447,
"learning_rate": 9.51709722371101e-06,
"loss": 1.3865,
"mean_token_accuracy": 0.6698870331048965,
"num_tokens": 192681071.0,
"step": 1480
},
{
"entropy": 2.02734375,
"epoch": 0.07745893117072157,
"grad_norm": 2.5823376342186015,
"learning_rate": 9.511737592453641e-06,
"loss": 1.402,
"mean_token_accuracy": 0.6671822279691696,
"num_tokens": 193983728.0,
"step": 1490
},
{
"entropy": 2.038671875,
"epoch": 0.07797878976918278,
"grad_norm": 2.6024201505162354,
"learning_rate": 9.50637796119627e-06,
"loss": 1.3887,
"mean_token_accuracy": 0.668594816327095,
"num_tokens": 195287579.0,
"step": 1500
},
{
"epoch": 0.07797878976918278,
"eval_entropy": 2.0302734375,
"eval_loss": 1.435243010520935,
"eval_mean_token_accuracy": 0.6606044501066208,
"eval_num_tokens": 195287579.0,
"eval_runtime": 27.2068,
"eval_samples_per_second": 93.138,
"eval_steps_per_second": 1.47,
"step": 1500
},
{
"entropy": 2.02265625,
"epoch": 0.078498648367644,
"grad_norm": 2.5855803247661195,
"learning_rate": 9.501018329938901e-06,
"loss": 1.3895,
"mean_token_accuracy": 0.6713440209627152,
"num_tokens": 196591329.0,
"step": 1510
},
{
"entropy": 2.04609375,
"epoch": 0.07901850696610523,
"grad_norm": 2.7795656914567006,
"learning_rate": 9.495658698681531e-06,
"loss": 1.4315,
"mean_token_accuracy": 0.6607480376958847,
"num_tokens": 197890885.0,
"step": 1520
},
{
"entropy": 2.05625,
"epoch": 0.07953836556456644,
"grad_norm": 2.367645970708118,
"learning_rate": 9.490299067424162e-06,
"loss": 1.4146,
"mean_token_accuracy": 0.666170334815979,
"num_tokens": 199195384.0,
"step": 1530
},
{
"entropy": 2.062890625,
"epoch": 0.08005822416302766,
"grad_norm": 2.549154467458875,
"learning_rate": 9.484939436166792e-06,
"loss": 1.4092,
"mean_token_accuracy": 0.667629012465477,
"num_tokens": 200498501.0,
"step": 1540
},
{
"entropy": 2.03671875,
"epoch": 0.08057808276148888,
"grad_norm": 2.3488506515136254,
"learning_rate": 9.479579804909424e-06,
"loss": 1.3856,
"mean_token_accuracy": 0.6679800838232041,
"num_tokens": 201801209.0,
"step": 1550
},
{
"entropy": 2.07578125,
"epoch": 0.08109794135995009,
"grad_norm": 2.5642379296514517,
"learning_rate": 9.474220173652054e-06,
"loss": 1.4006,
"mean_token_accuracy": 0.6687598139047622,
"num_tokens": 203103345.0,
"step": 1560
},
{
"entropy": 2.05390625,
"epoch": 0.08161779995841131,
"grad_norm": 2.449029885700019,
"learning_rate": 9.468860542394684e-06,
"loss": 1.4193,
"mean_token_accuracy": 0.6652283906936646,
"num_tokens": 204405948.0,
"step": 1570
},
{
"entropy": 1.979296875,
"epoch": 0.08213765855687254,
"grad_norm": 3.0437637025766104,
"learning_rate": 9.463500911137314e-06,
"loss": 1.3955,
"mean_token_accuracy": 0.6658499509096145,
"num_tokens": 205707238.0,
"step": 1580
},
{
"entropy": 2.00703125,
"epoch": 0.08265751715533375,
"grad_norm": 3.1140732425343,
"learning_rate": 9.458141279879946e-06,
"loss": 1.3595,
"mean_token_accuracy": 0.6787232100963593,
"num_tokens": 207010502.0,
"step": 1590
},
{
"entropy": 2.0390625,
"epoch": 0.08317737575379497,
"grad_norm": 2.548702493340075,
"learning_rate": 9.452781648622576e-06,
"loss": 1.4185,
"mean_token_accuracy": 0.6632625550031662,
"num_tokens": 208311573.0,
"step": 1600
},
{
"epoch": 0.08317737575379497,
"eval_entropy": 2.004296875,
"eval_loss": 1.4317197799682617,
"eval_mean_token_accuracy": 0.6604633212089539,
"eval_num_tokens": 208311573.0,
"eval_runtime": 27.0037,
"eval_samples_per_second": 93.839,
"eval_steps_per_second": 1.481,
"step": 1600
},
{
"entropy": 1.997265625,
"epoch": 0.08369723435225619,
"grad_norm": 2.4567384212063628,
"learning_rate": 9.447422017365206e-06,
"loss": 1.3889,
"mean_token_accuracy": 0.6695194095373154,
"num_tokens": 209613844.0,
"step": 1610
},
{
"entropy": 1.978125,
"epoch": 0.0842170929507174,
"grad_norm": 2.7034706405083724,
"learning_rate": 9.442062386107836e-06,
"loss": 1.3773,
"mean_token_accuracy": 0.6750271588563919,
"num_tokens": 210916179.0,
"step": 1620
},
{
"entropy": 1.997265625,
"epoch": 0.08473695154917862,
"grad_norm": 2.5562181422275447,
"learning_rate": 9.436702754850468e-06,
"loss": 1.4301,
"mean_token_accuracy": 0.664066144824028,
"num_tokens": 212217362.0,
"step": 1630
},
{
"entropy": 2.02734375,
"epoch": 0.08525681014763985,
"grad_norm": 2.678502515222626,
"learning_rate": 9.431343123593098e-06,
"loss": 1.3841,
"mean_token_accuracy": 0.6694748312234878,
"num_tokens": 213518133.0,
"step": 1640
},
{
"entropy": 2.06328125,
"epoch": 0.08577666874610106,
"grad_norm": 2.607608097352403,
"learning_rate": 9.425983492335728e-06,
"loss": 1.4287,
"mean_token_accuracy": 0.6610083252191543,
"num_tokens": 214819908.0,
"step": 1650
},
{
"entropy": 2.022265625,
"epoch": 0.08629652734456228,
"grad_norm": 2.849344090876858,
"learning_rate": 9.420623861078358e-06,
"loss": 1.4219,
"mean_token_accuracy": 0.6640592664480209,
"num_tokens": 216121638.0,
"step": 1660
},
{
"entropy": 2.071484375,
"epoch": 0.0868163859430235,
"grad_norm": 2.922522673979407,
"learning_rate": 9.41526422982099e-06,
"loss": 1.3964,
"mean_token_accuracy": 0.6712499350309372,
"num_tokens": 217423164.0,
"step": 1670
},
{
"entropy": 2.045703125,
"epoch": 0.08733624454148471,
"grad_norm": 2.6177823848425126,
"learning_rate": 9.40990459856362e-06,
"loss": 1.3759,
"mean_token_accuracy": 0.6719487220048904,
"num_tokens": 218725967.0,
"step": 1680
},
{
"entropy": 2.03515625,
"epoch": 0.08785610313994593,
"grad_norm": 2.5785505086255376,
"learning_rate": 9.40454496730625e-06,
"loss": 1.3912,
"mean_token_accuracy": 0.6719380766153336,
"num_tokens": 220029274.0,
"step": 1690
},
{
"entropy": 2.066015625,
"epoch": 0.08837596173840716,
"grad_norm": 2.9368513925819935,
"learning_rate": 9.39918533604888e-06,
"loss": 1.3927,
"mean_token_accuracy": 0.6695972323417664,
"num_tokens": 221332034.0,
"step": 1700
},
{
"epoch": 0.08837596173840716,
"eval_entropy": 2.016796875,
"eval_loss": 1.420469880104065,
"eval_mean_token_accuracy": 0.6638253465294838,
"eval_num_tokens": 221332034.0,
"eval_runtime": 26.9803,
"eval_samples_per_second": 93.92,
"eval_steps_per_second": 1.483,
"step": 1700
},
{
"entropy": 2.06328125,
"epoch": 0.08889582033686837,
"grad_norm": 2.464298999864292,
"learning_rate": 9.393825704791512e-06,
"loss": 1.4245,
"mean_token_accuracy": 0.6632371693849564,
"num_tokens": 222632667.0,
"step": 1710
},
{
"entropy": 2.0390625,
"epoch": 0.08941567893532959,
"grad_norm": 2.7100483482009547,
"learning_rate": 9.38846607353414e-06,
"loss": 1.3366,
"mean_token_accuracy": 0.680601692199707,
"num_tokens": 223933320.0,
"step": 1720
},
{
"entropy": 2.073828125,
"epoch": 0.08993553753379081,
"grad_norm": 2.407558938300571,
"learning_rate": 9.383106442276773e-06,
"loss": 1.4176,
"mean_token_accuracy": 0.6645108669996261,
"num_tokens": 225233916.0,
"step": 1730
},
{
"entropy": 2.054296875,
"epoch": 0.09045539613225202,
"grad_norm": 2.6197908635542286,
"learning_rate": 9.377746811019403e-06,
"loss": 1.4105,
"mean_token_accuracy": 0.665682104229927,
"num_tokens": 226537516.0,
"step": 1740
},
{
"entropy": 2.028125,
"epoch": 0.09097525473071325,
"grad_norm": 2.5963703984942454,
"learning_rate": 9.372387179762033e-06,
"loss": 1.4031,
"mean_token_accuracy": 0.6695938050746918,
"num_tokens": 227838805.0,
"step": 1750
},
{
"entropy": 2.058984375,
"epoch": 0.09149511332917447,
"grad_norm": 2.7340601198404464,
"learning_rate": 9.367027548504663e-06,
"loss": 1.4108,
"mean_token_accuracy": 0.6670466840267182,
"num_tokens": 229141210.0,
"step": 1760
},
{
"entropy": 2.023046875,
"epoch": 0.09201497192763568,
"grad_norm": 2.4650465607155785,
"learning_rate": 9.361667917247295e-06,
"loss": 1.3645,
"mean_token_accuracy": 0.6754943817853928,
"num_tokens": 230441859.0,
"step": 1770
},
{
"entropy": 2.062890625,
"epoch": 0.0925348305260969,
"grad_norm": 2.5386388891164144,
"learning_rate": 9.356308285989925e-06,
"loss": 1.3955,
"mean_token_accuracy": 0.6698277622461319,
"num_tokens": 231743741.0,
"step": 1780
},
{
"entropy": 2.044140625,
"epoch": 0.09305468912455812,
"grad_norm": 2.698005424136433,
"learning_rate": 9.350948654732555e-06,
"loss": 1.3844,
"mean_token_accuracy": 0.6732267916202546,
"num_tokens": 233044446.0,
"step": 1790
},
{
"entropy": 2.038671875,
"epoch": 0.09357454772301933,
"grad_norm": 2.676908264455853,
"learning_rate": 9.345589023475185e-06,
"loss": 1.3943,
"mean_token_accuracy": 0.6691930711269378,
"num_tokens": 234348269.0,
"step": 1800
},
{
"epoch": 0.09357454772301933,
"eval_entropy": 2.007421875,
"eval_loss": 1.4178036451339722,
"eval_mean_token_accuracy": 0.6636074542999267,
"eval_num_tokens": 234348269.0,
"eval_runtime": 26.9853,
"eval_samples_per_second": 93.903,
"eval_steps_per_second": 1.482,
"step": 1800
},
{
"entropy": 2.02265625,
"epoch": 0.09409440632148056,
"grad_norm": 2.7390165861353486,
"learning_rate": 9.340229392217817e-06,
"loss": 1.395,
"mean_token_accuracy": 0.6693720579147339,
"num_tokens": 235650704.0,
"step": 1810
},
{
"entropy": 2.023828125,
"epoch": 0.09461426491994178,
"grad_norm": 2.507994043795811,
"learning_rate": 9.334869760960447e-06,
"loss": 1.376,
"mean_token_accuracy": 0.6738913178443908,
"num_tokens": 236951944.0,
"step": 1820
},
{
"entropy": 2.0328125,
"epoch": 0.09513412351840299,
"grad_norm": 3.240811141889355,
"learning_rate": 9.329510129703077e-06,
"loss": 1.3499,
"mean_token_accuracy": 0.6763323545455933,
"num_tokens": 238253416.0,
"step": 1830
},
{
"entropy": 1.99765625,
"epoch": 0.09565398211686421,
"grad_norm": 2.7346146361106283,
"learning_rate": 9.324150498445707e-06,
"loss": 1.4127,
"mean_token_accuracy": 0.6676194965839386,
"num_tokens": 239555031.0,
"step": 1840
},
{
"entropy": 2.006640625,
"epoch": 0.09617384071532543,
"grad_norm": 2.4928921712222123,
"learning_rate": 9.318790867188339e-06,
"loss": 1.3918,
"mean_token_accuracy": 0.670895129442215,
"num_tokens": 240856086.0,
"step": 1850
},
{
"entropy": 2.0640625,
"epoch": 0.09669369931378664,
"grad_norm": 2.5063360269997084,
"learning_rate": 9.31343123593097e-06,
"loss": 1.4056,
"mean_token_accuracy": 0.666303864121437,
"num_tokens": 242157854.0,
"step": 1860
},
{
"entropy": 2.01328125,
"epoch": 0.09721355791224787,
"grad_norm": 2.732554476391581,
"learning_rate": 9.3080716046736e-06,
"loss": 1.3824,
"mean_token_accuracy": 0.6771417498588562,
"num_tokens": 243460084.0,
"step": 1870
},
{
"entropy": 2.0515625,
"epoch": 0.09773341651070909,
"grad_norm": 2.5499294787066455,
"learning_rate": 9.30271197341623e-06,
"loss": 1.3656,
"mean_token_accuracy": 0.6766289532184601,
"num_tokens": 244762073.0,
"step": 1880
},
{
"entropy": 2.004296875,
"epoch": 0.0982532751091703,
"grad_norm": 2.538121085886189,
"learning_rate": 9.297352342158861e-06,
"loss": 1.3767,
"mean_token_accuracy": 0.671790137887001,
"num_tokens": 246064425.0,
"step": 1890
},
{
"entropy": 1.980078125,
"epoch": 0.09877313370763152,
"grad_norm": 2.76951925906396,
"learning_rate": 9.291992710901491e-06,
"loss": 1.4036,
"mean_token_accuracy": 0.6705715835094452,
"num_tokens": 247365824.0,
"step": 1900
},
{
"epoch": 0.09877313370763152,
"eval_entropy": 1.9462890625,
"eval_loss": 1.4127426147460938,
"eval_mean_token_accuracy": 0.6644406974315643,
"eval_num_tokens": 247365824.0,
"eval_runtime": 27.0054,
"eval_samples_per_second": 93.833,
"eval_steps_per_second": 1.481,
"step": 1900
},
{
"entropy": 2.026953125,
"epoch": 0.09929299230609274,
"grad_norm": 2.7238698426351893,
"learning_rate": 9.286633079644122e-06,
"loss": 1.3401,
"mean_token_accuracy": 0.6806715220212937,
"num_tokens": 248666870.0,
"step": 1910
},
{
"entropy": 1.994921875,
"epoch": 0.09981285090455397,
"grad_norm": 2.747682289147482,
"learning_rate": 9.281273448386752e-06,
"loss": 1.3739,
"mean_token_accuracy": 0.6740750968456268,
"num_tokens": 249969743.0,
"step": 1920
},
{
"entropy": 2.003125,
"epoch": 0.10033270950301518,
"grad_norm": 2.6595239579733314,
"learning_rate": 9.275913817129383e-06,
"loss": 1.3859,
"mean_token_accuracy": 0.6720771104097366,
"num_tokens": 251270177.0,
"step": 1930
},
{
"entropy": 2.019140625,
"epoch": 0.1008525681014764,
"grad_norm": 2.3939634162871406,
"learning_rate": 9.270554185872012e-06,
"loss": 1.4316,
"mean_token_accuracy": 0.6625065386295319,
"num_tokens": 252573929.0,
"step": 1940
},
{
"entropy": 1.98671875,
"epoch": 0.10137242669993762,
"grad_norm": 2.732857852778348,
"learning_rate": 9.265194554614644e-06,
"loss": 1.3877,
"mean_token_accuracy": 0.6720799267292022,
"num_tokens": 253875776.0,
"step": 1950
},
{
"entropy": 2.015234375,
"epoch": 0.10189228529839883,
"grad_norm": 2.533087746834989,
"learning_rate": 9.259834923357274e-06,
"loss": 1.4121,
"mean_token_accuracy": 0.6682723343372345,
"num_tokens": 255176042.0,
"step": 1960
},
{
"entropy": 2.024609375,
"epoch": 0.10241214389686006,
"grad_norm": 2.4218817580551697,
"learning_rate": 9.254475292099904e-06,
"loss": 1.4295,
"mean_token_accuracy": 0.6633944094181061,
"num_tokens": 256477359.0,
"step": 1970
},
{
"entropy": 1.975,
"epoch": 0.10293200249532128,
"grad_norm": 2.846881174548448,
"learning_rate": 9.249115660842534e-06,
"loss": 1.3588,
"mean_token_accuracy": 0.6769413322210311,
"num_tokens": 257779911.0,
"step": 1980
},
{
"entropy": 2.01953125,
"epoch": 0.10345186109378249,
"grad_norm": 2.9131234231296466,
"learning_rate": 9.243756029585166e-06,
"loss": 1.3517,
"mean_token_accuracy": 0.6770850718021393,
"num_tokens": 259081264.0,
"step": 1990
},
{
"entropy": 2.008203125,
"epoch": 0.10397171969224371,
"grad_norm": 2.6272952473750353,
"learning_rate": 9.238396398327796e-06,
"loss": 1.4158,
"mean_token_accuracy": 0.668806579709053,
"num_tokens": 260383494.0,
"step": 2000
},
{
"epoch": 0.10397171969224371,
"eval_entropy": 1.974609375,
"eval_loss": 1.4047836065292358,
"eval_mean_token_accuracy": 0.6666118755936623,
"eval_num_tokens": 260383494.0,
"eval_runtime": 27.1044,
"eval_samples_per_second": 93.49,
"eval_steps_per_second": 1.476,
"step": 2000
},
{
"entropy": 2.00078125,
"epoch": 0.10449157829070493,
"grad_norm": 3.070709524352511,
"learning_rate": 9.233036767070426e-06,
"loss": 1.3676,
"mean_token_accuracy": 0.6723472207784653,
"num_tokens": 261683016.0,
"step": 2010
},
{
"entropy": 2.0,
"epoch": 0.10501143688916614,
"grad_norm": 2.6265091862116585,
"learning_rate": 9.227677135813056e-06,
"loss": 1.382,
"mean_token_accuracy": 0.6716920107603073,
"num_tokens": 262986403.0,
"step": 2020
},
{
"entropy": 1.998828125,
"epoch": 0.10553129548762737,
"grad_norm": 2.6545799874321285,
"learning_rate": 9.222317504555688e-06,
"loss": 1.3805,
"mean_token_accuracy": 0.673284649848938,
"num_tokens": 264289615.0,
"step": 2030
},
{
"entropy": 2.03125,
"epoch": 0.10605115408608859,
"grad_norm": 2.9151022125053974,
"learning_rate": 9.216957873298318e-06,
"loss": 1.3453,
"mean_token_accuracy": 0.677672490477562,
"num_tokens": 265591910.0,
"step": 2040
},
{
"entropy": 2.007421875,
"epoch": 0.1065710126845498,
"grad_norm": 2.9782818218922915,
"learning_rate": 9.211598242040948e-06,
"loss": 1.3553,
"mean_token_accuracy": 0.6797802478075028,
"num_tokens": 266894394.0,
"step": 2050
},
{
"entropy": 2.031640625,
"epoch": 0.10709087128301102,
"grad_norm": 2.4002979406956886,
"learning_rate": 9.206238610783578e-06,
"loss": 1.3395,
"mean_token_accuracy": 0.6836020201444626,
"num_tokens": 268196218.0,
"step": 2060
},
{
"entropy": 1.967578125,
"epoch": 0.10761072988147224,
"grad_norm": 2.4076284230874703,
"learning_rate": 9.20087897952621e-06,
"loss": 1.3958,
"mean_token_accuracy": 0.6695514619350433,
"num_tokens": 269499140.0,
"step": 2070
},
{
"entropy": 2.10625,
"epoch": 0.10813058847993345,
"grad_norm": 2.8001726106408182,
"learning_rate": 9.19551934826884e-06,
"loss": 1.3643,
"mean_token_accuracy": 0.6767167150974274,
"num_tokens": 270799744.0,
"step": 2080
},
{
"entropy": 2.018359375,
"epoch": 0.10865044707839468,
"grad_norm": 8.157287764130137,
"learning_rate": 9.19015971701147e-06,
"loss": 1.3623,
"mean_token_accuracy": 0.6760893553495407,
"num_tokens": 272103344.0,
"step": 2090
},
{
"entropy": 2.031640625,
"epoch": 0.1091703056768559,
"grad_norm": 2.761874629938549,
"learning_rate": 9.1848000857541e-06,
"loss": 1.3851,
"mean_token_accuracy": 0.6700647860765457,
"num_tokens": 273402737.0,
"step": 2100
},
{
"epoch": 0.1091703056768559,
"eval_entropy": 2.040625,
"eval_loss": 1.4009929895401,
"eval_mean_token_accuracy": 0.667850948870182,
"eval_num_tokens": 273402737.0,
"eval_runtime": 27.0596,
"eval_samples_per_second": 93.645,
"eval_steps_per_second": 1.478,
"step": 2100
},
{
"entropy": 2.021875,
"epoch": 0.10969016427531711,
"grad_norm": 2.784601063977959,
"learning_rate": 9.17944045449673e-06,
"loss": 1.3379,
"mean_token_accuracy": 0.6825733125209809,
"num_tokens": 274702158.0,
"step": 2110
},
{
"entropy": 2.000390625,
"epoch": 0.11021002287377833,
"grad_norm": 2.591062518577273,
"learning_rate": 9.174080823239363e-06,
"loss": 1.3575,
"mean_token_accuracy": 0.6781095385551452,
"num_tokens": 276004579.0,
"step": 2120
},
{
"entropy": 2.0359375,
"epoch": 0.11072988147223956,
"grad_norm": 2.601686567633845,
"learning_rate": 9.168721191981991e-06,
"loss": 1.3777,
"mean_token_accuracy": 0.6740903496742249,
"num_tokens": 277305917.0,
"step": 2130
},
{
"entropy": 2.007421875,
"epoch": 0.11124974007070076,
"grad_norm": 2.7400501543741735,
"learning_rate": 9.163361560724623e-06,
"loss": 1.3723,
"mean_token_accuracy": 0.6745098769664765,
"num_tokens": 278609641.0,
"step": 2140
},
{
"entropy": 2.021875,
"epoch": 0.11176959866916199,
"grad_norm": 2.422440752528026,
"learning_rate": 9.158001929467253e-06,
"loss": 1.4219,
"mean_token_accuracy": 0.6665849059820175,
"num_tokens": 279910262.0,
"step": 2150
},
{
"entropy": 1.99296875,
"epoch": 0.11228945726762321,
"grad_norm": 2.5743034538552125,
"learning_rate": 9.152642298209885e-06,
"loss": 1.4015,
"mean_token_accuracy": 0.6684765040874481,
"num_tokens": 281212875.0,
"step": 2160
},
{
"entropy": 2.022265625,
"epoch": 0.11280931586608442,
"grad_norm": 2.7729910969564817,
"learning_rate": 9.147282666952513e-06,
"loss": 1.3554,
"mean_token_accuracy": 0.6798837929964066,
"num_tokens": 282515068.0,
"step": 2170
},
{
"entropy": 2.024609375,
"epoch": 0.11332917446454564,
"grad_norm": 2.6027295352329602,
"learning_rate": 9.141923035695145e-06,
"loss": 1.3623,
"mean_token_accuracy": 0.67585991024971,
"num_tokens": 283815784.0,
"step": 2180
},
{
"entropy": 1.9875,
"epoch": 0.11384903306300687,
"grad_norm": 2.4855342784314542,
"learning_rate": 9.136563404437775e-06,
"loss": 1.3857,
"mean_token_accuracy": 0.6701562643051148,
"num_tokens": 285116482.0,
"step": 2190
},
{
"entropy": 1.98515625,
"epoch": 0.11436889166146808,
"grad_norm": 2.5551981930725916,
"learning_rate": 9.131203773180405e-06,
"loss": 1.3495,
"mean_token_accuracy": 0.678818729519844,
"num_tokens": 286420060.0,
"step": 2200
},
{
"epoch": 0.11436889166146808,
"eval_entropy": 1.9970703125,
"eval_loss": 1.3960766792297363,
"eval_mean_token_accuracy": 0.6685970649123192,
"eval_num_tokens": 286420060.0,
"eval_runtime": 27.0608,
"eval_samples_per_second": 93.641,
"eval_steps_per_second": 1.478,
"step": 2200
},
{
"entropy": 2.001953125,
"epoch": 0.1148887502599293,
"grad_norm": 2.982405130046775,
"learning_rate": 9.125844141923035e-06,
"loss": 1.3781,
"mean_token_accuracy": 0.674466535449028,
"num_tokens": 287720878.0,
"step": 2210
},
{
"entropy": 2.017578125,
"epoch": 0.11540860885839052,
"grad_norm": 2.382276300062519,
"learning_rate": 9.120484510665667e-06,
"loss": 1.3691,
"mean_token_accuracy": 0.6744934886693954,
"num_tokens": 289023589.0,
"step": 2220
},
{
"entropy": 1.99765625,
"epoch": 0.11592846745685173,
"grad_norm": 2.5691691971572053,
"learning_rate": 9.115124879408297e-06,
"loss": 1.3518,
"mean_token_accuracy": 0.676643231511116,
"num_tokens": 290326305.0,
"step": 2230
},
{
"entropy": 2.0125,
"epoch": 0.11644832605531295,
"grad_norm": 2.4832074690838035,
"learning_rate": 9.109765248150927e-06,
"loss": 1.3769,
"mean_token_accuracy": 0.6709991842508316,
"num_tokens": 291629429.0,
"step": 2240
},
{
"entropy": 1.988671875,
"epoch": 0.11696818465377418,
"grad_norm": 2.652686434207374,
"learning_rate": 9.104405616893558e-06,
"loss": 1.3163,
"mean_token_accuracy": 0.6862490355968476,
"num_tokens": 292929974.0,
"step": 2250
},
{
"entropy": 1.975390625,
"epoch": 0.11748804325223539,
"grad_norm": 3.048981243113635,
"learning_rate": 9.09904598563619e-06,
"loss": 1.3594,
"mean_token_accuracy": 0.676978400349617,
"num_tokens": 294233843.0,
"step": 2260
},
{
"entropy": 2.021875,
"epoch": 0.11800790185069661,
"grad_norm": 2.608654437744063,
"learning_rate": 9.09368635437882e-06,
"loss": 1.3707,
"mean_token_accuracy": 0.6750653028488159,
"num_tokens": 295537371.0,
"step": 2270
},
{
"entropy": 1.98515625,
"epoch": 0.11852776044915783,
"grad_norm": 2.8141842646596515,
"learning_rate": 9.08832672312145e-06,
"loss": 1.3613,
"mean_token_accuracy": 0.6780448496341706,
"num_tokens": 296840438.0,
"step": 2280
},
{
"entropy": 2.022265625,
"epoch": 0.11904761904761904,
"grad_norm": 2.5998829340813554,
"learning_rate": 9.08296709186408e-06,
"loss": 1.356,
"mean_token_accuracy": 0.6770641416311264,
"num_tokens": 298140790.0,
"step": 2290
},
{
"entropy": 2.009375,
"epoch": 0.11956747764608026,
"grad_norm": 2.8125915143156304,
"learning_rate": 9.077607460606712e-06,
"loss": 1.33,
"mean_token_accuracy": 0.6845223695039749,
"num_tokens": 299441254.0,
"step": 2300
},
{
"epoch": 0.11956747764608026,
"eval_entropy": 1.9716796875,
"eval_loss": 1.3859691619873047,
"eval_mean_token_accuracy": 0.6709905341267586,
"eval_num_tokens": 299441254.0,
"eval_runtime": 27.0126,
"eval_samples_per_second": 93.808,
"eval_steps_per_second": 1.481,
"step": 2300
},
{
"entropy": 2.008203125,
"epoch": 0.12008733624454149,
"grad_norm": 2.624157211239115,
"learning_rate": 9.072247829349342e-06,
"loss": 1.3649,
"mean_token_accuracy": 0.6773047059774399,
"num_tokens": 300739436.0,
"step": 2310
},
{
"entropy": 2.0125,
"epoch": 0.1206071948430027,
"grad_norm": 2.8337814753648702,
"learning_rate": 9.066888198091972e-06,
"loss": 1.3914,
"mean_token_accuracy": 0.6691319733858109,
"num_tokens": 302043326.0,
"step": 2320
},
{
"entropy": 2.021875,
"epoch": 0.12112705344146392,
"grad_norm": 2.5653622912298766,
"learning_rate": 9.061528566834602e-06,
"loss": 1.3528,
"mean_token_accuracy": 0.6786749511957169,
"num_tokens": 303342498.0,
"step": 2330
},
{
"entropy": 2.005078125,
"epoch": 0.12164691203992514,
"grad_norm": 2.6674978787370636,
"learning_rate": 9.056168935577234e-06,
"loss": 1.3455,
"mean_token_accuracy": 0.6801716178655625,
"num_tokens": 304643812.0,
"step": 2340
},
{
"entropy": 2.054296875,
"epoch": 0.12216677063838637,
"grad_norm": 2.633751596710382,
"learning_rate": 9.050809304319864e-06,
"loss": 1.3695,
"mean_token_accuracy": 0.676679277420044,
"num_tokens": 305946029.0,
"step": 2350
},
{
"entropy": 2.041015625,
"epoch": 0.12268662923684757,
"grad_norm": 2.6676003058728086,
"learning_rate": 9.045449673062494e-06,
"loss": 1.3687,
"mean_token_accuracy": 0.6756788551807403,
"num_tokens": 307245241.0,
"step": 2360
},
{
"entropy": 2.040625,
"epoch": 0.1232064878353088,
"grad_norm": 2.5317864217738335,
"learning_rate": 9.040090041805124e-06,
"loss": 1.3732,
"mean_token_accuracy": 0.6745780855417252,
"num_tokens": 308546455.0,
"step": 2370
},
{
"entropy": 2.032421875,
"epoch": 0.12372634643377002,
"grad_norm": 2.642107073084182,
"learning_rate": 9.034730410547756e-06,
"loss": 1.3807,
"mean_token_accuracy": 0.6755623400211335,
"num_tokens": 309849524.0,
"step": 2380
},
{
"entropy": 2.030078125,
"epoch": 0.12424620503223123,
"grad_norm": 2.501819333032555,
"learning_rate": 9.029370779290384e-06,
"loss": 1.3359,
"mean_token_accuracy": 0.681933405995369,
"num_tokens": 311151059.0,
"step": 2390
},
{
"entropy": 2.045703125,
"epoch": 0.12476606363069245,
"grad_norm": 2.5157058130078664,
"learning_rate": 9.024011148033016e-06,
"loss": 1.3656,
"mean_token_accuracy": 0.6800758242607117,
"num_tokens": 312454559.0,
"step": 2400
},
{
"epoch": 0.12476606363069245,
"eval_entropy": 1.975390625,
"eval_loss": 1.3839225769042969,
"eval_mean_token_accuracy": 0.6713285893201828,
"eval_num_tokens": 312454559.0,
"eval_runtime": 27.0039,
"eval_samples_per_second": 93.838,
"eval_steps_per_second": 1.481,
"step": 2400
},
{
"entropy": 1.991015625,
"epoch": 0.12528592222915366,
"grad_norm": 2.360065329078026,
"learning_rate": 9.018651516775646e-06,
"loss": 1.3229,
"mean_token_accuracy": 0.6838825106620788,
"num_tokens": 313757539.0,
"step": 2410
},
{
"entropy": 1.990625,
"epoch": 0.1258057808276149,
"grad_norm": 2.501783883010503,
"learning_rate": 9.013291885518276e-06,
"loss": 1.362,
"mean_token_accuracy": 0.6750700205564499,
"num_tokens": 315058966.0,
"step": 2420
},
{
"entropy": 1.980859375,
"epoch": 0.1263256394260761,
"grad_norm": 2.47355760609497,
"learning_rate": 9.007932254260907e-06,
"loss": 1.3526,
"mean_token_accuracy": 0.6768155574798584,
"num_tokens": 316362342.0,
"step": 2430
},
{
"entropy": 2.007421875,
"epoch": 0.12684549802453732,
"grad_norm": 2.799988771098822,
"learning_rate": 9.002572623003538e-06,
"loss": 1.3233,
"mean_token_accuracy": 0.6848174929618835,
"num_tokens": 317667145.0,
"step": 2440
},
{
"entropy": 1.994921875,
"epoch": 0.12736535662299855,
"grad_norm": 2.527531475216805,
"learning_rate": 8.997212991746169e-06,
"loss": 1.3738,
"mean_token_accuracy": 0.6752849668264389,
"num_tokens": 318968175.0,
"step": 2450
},
{
"entropy": 1.978515625,
"epoch": 0.12788521522145976,
"grad_norm": 2.8430997230068287,
"learning_rate": 8.991853360488799e-06,
"loss": 1.3956,
"mean_token_accuracy": 0.6706503003835678,
"num_tokens": 320272349.0,
"step": 2460
},
{
"entropy": 1.988671875,
"epoch": 0.12840507381992097,
"grad_norm": 2.8777979592660485,
"learning_rate": 8.986493729231429e-06,
"loss": 1.3483,
"mean_token_accuracy": 0.6801801890134811,
"num_tokens": 321574263.0,
"step": 2470
},
{
"entropy": 2.004296875,
"epoch": 0.1289249324183822,
"grad_norm": 2.5268293675624665,
"learning_rate": 8.98113409797406e-06,
"loss": 1.3893,
"mean_token_accuracy": 0.6708779454231262,
"num_tokens": 322876530.0,
"step": 2480
},
{
"entropy": 1.96796875,
"epoch": 0.12944479101684342,
"grad_norm": 2.7115011170393135,
"learning_rate": 8.97577446671669e-06,
"loss": 1.3283,
"mean_token_accuracy": 0.6846076279878617,
"num_tokens": 324177574.0,
"step": 2490
},
{
"entropy": 1.962890625,
"epoch": 0.12996464961530463,
"grad_norm": 2.6053085660554456,
"learning_rate": 8.97041483545932e-06,
"loss": 1.3649,
"mean_token_accuracy": 0.675525963306427,
"num_tokens": 325477049.0,
"step": 2500
},
{
"epoch": 0.12996464961530463,
"eval_entropy": 1.972265625,
"eval_loss": 1.3792837858200073,
"eval_mean_token_accuracy": 0.6723639145493507,
"eval_num_tokens": 325477049.0,
"eval_runtime": 27.014,
"eval_samples_per_second": 93.803,
"eval_steps_per_second": 1.481,
"step": 2500
},
{
"entropy": 2.003515625,
"epoch": 0.13048450821376587,
"grad_norm": 2.4592758471166274,
"learning_rate": 8.965055204201951e-06,
"loss": 1.3708,
"mean_token_accuracy": 0.6743284314870834,
"num_tokens": 326776724.0,
"step": 2510
},
{
"entropy": 2.012109375,
"epoch": 0.13100436681222707,
"grad_norm": 2.574252296800668,
"learning_rate": 8.959695572944583e-06,
"loss": 1.3482,
"mean_token_accuracy": 0.6815009206533432,
"num_tokens": 328078930.0,
"step": 2520
},
{
"entropy": 1.94921875,
"epoch": 0.13152422541068828,
"grad_norm": 2.708487235558883,
"learning_rate": 8.954335941687213e-06,
"loss": 1.3569,
"mean_token_accuracy": 0.6750432759523392,
"num_tokens": 329380087.0,
"step": 2530
},
{
"entropy": 2.015234375,
"epoch": 0.13204408400914952,
"grad_norm": 2.4342382464506196,
"learning_rate": 8.948976310429843e-06,
"loss": 1.3757,
"mean_token_accuracy": 0.6739129811525345,
"num_tokens": 330682945.0,
"step": 2540
},
{
"entropy": 1.975390625,
"epoch": 0.13256394260761073,
"grad_norm": 2.5081017401653516,
"learning_rate": 8.943616679172473e-06,
"loss": 1.3718,
"mean_token_accuracy": 0.6725127309560776,
"num_tokens": 331982796.0,
"step": 2550
},
{
"entropy": 1.944921875,
"epoch": 0.13308380120607194,
"grad_norm": 2.793304458101309,
"learning_rate": 8.938257047915105e-06,
"loss": 1.3028,
"mean_token_accuracy": 0.6872901648283005,
"num_tokens": 333282751.0,
"step": 2560
},
{
"entropy": 1.984375,
"epoch": 0.13360365980453318,
"grad_norm": 2.5315368975828263,
"learning_rate": 8.932897416657735e-06,
"loss": 1.2981,
"mean_token_accuracy": 0.6897401094436646,
"num_tokens": 334585997.0,
"step": 2570
},
{
"entropy": 1.976171875,
"epoch": 0.13412351840299438,
"grad_norm": 2.5271811222319434,
"learning_rate": 8.927537785400365e-06,
"loss": 1.3197,
"mean_token_accuracy": 0.6846640080213546,
"num_tokens": 335885944.0,
"step": 2580
},
{
"entropy": 1.993359375,
"epoch": 0.1346433770014556,
"grad_norm": 2.855655468025504,
"learning_rate": 8.922178154142995e-06,
"loss": 1.3676,
"mean_token_accuracy": 0.6748760938644409,
"num_tokens": 337187228.0,
"step": 2590
},
{
"entropy": 1.983203125,
"epoch": 0.13516323559991683,
"grad_norm": 2.5039621531937115,
"learning_rate": 8.916818522885627e-06,
"loss": 1.3328,
"mean_token_accuracy": 0.6825832277536392,
"num_tokens": 338488164.0,
"step": 2600
},
{
"epoch": 0.13516323559991683,
"eval_entropy": 1.9767578125,
"eval_loss": 1.3755197525024414,
"eval_mean_token_accuracy": 0.6733179241418839,
"eval_num_tokens": 338488164.0,
"eval_runtime": 27.008,
"eval_samples_per_second": 93.824,
"eval_steps_per_second": 1.481,
"step": 2600
},
{
"entropy": 2.01640625,
"epoch": 0.13568309419837804,
"grad_norm": 2.5312937937310864,
"learning_rate": 8.911458891628256e-06,
"loss": 1.3267,
"mean_token_accuracy": 0.6848561882972717,
"num_tokens": 339790594.0,
"step": 2610
},
{
"entropy": 1.994921875,
"epoch": 0.13620295279683925,
"grad_norm": 2.3059126461180357,
"learning_rate": 8.906099260370887e-06,
"loss": 1.3702,
"mean_token_accuracy": 0.6771656483411789,
"num_tokens": 341094157.0,
"step": 2620
},
{
"entropy": 2.0,
"epoch": 0.1367228113953005,
"grad_norm": 2.6807827388338126,
"learning_rate": 8.900739629113517e-06,
"loss": 1.3225,
"mean_token_accuracy": 0.6845659852027893,
"num_tokens": 342398122.0,
"step": 2630
},
{
"entropy": 1.971484375,
"epoch": 0.1372426699937617,
"grad_norm": 2.51924850073416,
"learning_rate": 8.895379997856148e-06,
"loss": 1.3048,
"mean_token_accuracy": 0.689801812171936,
"num_tokens": 343700717.0,
"step": 2640
},
{
"entropy": 1.96328125,
"epoch": 0.1377625285922229,
"grad_norm": 2.3982882079984735,
"learning_rate": 8.890020366598778e-06,
"loss": 1.3311,
"mean_token_accuracy": 0.6822763621807099,
"num_tokens": 344999022.0,
"step": 2650
},
{
"entropy": 1.9734375,
"epoch": 0.13828238719068414,
"grad_norm": 2.291264463741086,
"learning_rate": 8.88466073534141e-06,
"loss": 1.3007,
"mean_token_accuracy": 0.6913455426692963,
"num_tokens": 346301080.0,
"step": 2660
},
{
"entropy": 1.98125,
"epoch": 0.13880224578914535,
"grad_norm": 2.4229439218028896,
"learning_rate": 8.87930110408404e-06,
"loss": 1.3457,
"mean_token_accuracy": 0.6779372036457062,
"num_tokens": 347605133.0,
"step": 2670
},
{
"entropy": 1.988671875,
"epoch": 0.13932210438760656,
"grad_norm": 2.7586423758544627,
"learning_rate": 8.87394147282667e-06,
"loss": 1.3605,
"mean_token_accuracy": 0.6751784056425094,
"num_tokens": 348907828.0,
"step": 2680
},
{
"entropy": 2.021484375,
"epoch": 0.1398419629860678,
"grad_norm": 2.681990574272719,
"learning_rate": 8.8685818415693e-06,
"loss": 1.3682,
"mean_token_accuracy": 0.6760987550020218,
"num_tokens": 350208334.0,
"step": 2690
},
{
"entropy": 2.00234375,
"epoch": 0.140361821584529,
"grad_norm": 2.928317940950406,
"learning_rate": 8.863222210311932e-06,
"loss": 1.337,
"mean_token_accuracy": 0.6807793468236923,
"num_tokens": 351509096.0,
"step": 2700
},
{
"epoch": 0.140361821584529,
"eval_entropy": 1.9724609375,
"eval_loss": 1.3685216903686523,
"eval_mean_token_accuracy": 0.6740984484553337,
"eval_num_tokens": 351509096.0,
"eval_runtime": 27.0434,
"eval_samples_per_second": 93.701,
"eval_steps_per_second": 1.479,
"step": 2700
},
{
"entropy": 2.016015625,
"epoch": 0.14088168018299022,
"grad_norm": 2.4812603690102284,
"learning_rate": 8.857862579054562e-06,
"loss": 1.3212,
"mean_token_accuracy": 0.6844858020544052,
"num_tokens": 352809085.0,
"step": 2710
},
{
"entropy": 1.993359375,
"epoch": 0.14140153878145145,
"grad_norm": 2.8405704970246606,
"learning_rate": 8.852502947797192e-06,
"loss": 1.3565,
"mean_token_accuracy": 0.6776537209749222,
"num_tokens": 354112358.0,
"step": 2720
},
{
"entropy": 1.969921875,
"epoch": 0.14192139737991266,
"grad_norm": 2.4264903116994625,
"learning_rate": 8.847143316539822e-06,
"loss": 1.3127,
"mean_token_accuracy": 0.6866943329572678,
"num_tokens": 355414619.0,
"step": 2730
},
{
"entropy": 2.0296875,
"epoch": 0.14244125597837387,
"grad_norm": 2.6577250750002928,
"learning_rate": 8.841783685282454e-06,
"loss": 1.3039,
"mean_token_accuracy": 0.6873502671718598,
"num_tokens": 356714862.0,
"step": 2740
},
{
"entropy": 1.971875,
"epoch": 0.1429611145768351,
"grad_norm": 2.6797247114469673,
"learning_rate": 8.836424054025084e-06,
"loss": 1.3788,
"mean_token_accuracy": 0.6743247807025909,
"num_tokens": 358018141.0,
"step": 2750
},
{
"entropy": 1.992578125,
"epoch": 0.14348097317529632,
"grad_norm": 2.537164654047222,
"learning_rate": 8.831064422767714e-06,
"loss": 1.3247,
"mean_token_accuracy": 0.6858545809984207,
"num_tokens": 359320684.0,
"step": 2760
},
{
"entropy": 2.00546875,
"epoch": 0.14400083177375753,
"grad_norm": 2.444024008418957,
"learning_rate": 8.825704791510344e-06,
"loss": 1.333,
"mean_token_accuracy": 0.6815615147352219,
"num_tokens": 360623362.0,
"step": 2770
},
{
"entropy": 1.951953125,
"epoch": 0.14452069037221876,
"grad_norm": 2.9034091639344246,
"learning_rate": 8.820345160252976e-06,
"loss": 1.2783,
"mean_token_accuracy": 0.695491886138916,
"num_tokens": 361923777.0,
"step": 2780
},
{
"entropy": 1.980078125,
"epoch": 0.14504054897067997,
"grad_norm": 2.7259757079111915,
"learning_rate": 8.814985528995606e-06,
"loss": 1.3381,
"mean_token_accuracy": 0.680729228258133,
"num_tokens": 363227113.0,
"step": 2790
},
{
"entropy": 2.020703125,
"epoch": 0.14556040756914118,
"grad_norm": 2.648938911928046,
"learning_rate": 8.809625897738236e-06,
"loss": 1.3458,
"mean_token_accuracy": 0.6786447942256928,
"num_tokens": 364530086.0,
"step": 2800
},
{
"epoch": 0.14556040756914118,
"eval_entropy": 2.01640625,
"eval_loss": 1.3597148656845093,
"eval_mean_token_accuracy": 0.6760232791304588,
"eval_num_tokens": 364530086.0,
"eval_runtime": 26.983,
"eval_samples_per_second": 93.911,
"eval_steps_per_second": 1.482,
"step": 2800
},
{
"entropy": 2.005859375,
"epoch": 0.14608026616760242,
"grad_norm": 2.868676709985562,
"learning_rate": 8.804266266480866e-06,
"loss": 1.3356,
"mean_token_accuracy": 0.6796860009431839,
"num_tokens": 365832735.0,
"step": 2810
},
{
"entropy": 1.98984375,
"epoch": 0.14660012476606363,
"grad_norm": 2.5283796981020346,
"learning_rate": 8.798906635223498e-06,
"loss": 1.3473,
"mean_token_accuracy": 0.681627282500267,
"num_tokens": 367136645.0,
"step": 2820
},
{
"entropy": 2.012109375,
"epoch": 0.14711998336452484,
"grad_norm": 2.392401812557231,
"learning_rate": 8.793547003966127e-06,
"loss": 1.3135,
"mean_token_accuracy": 0.68596131503582,
"num_tokens": 368439060.0,
"step": 2830
},
{
"entropy": 2.028125,
"epoch": 0.14763984196298607,
"grad_norm": 2.8483478752110374,
"learning_rate": 8.788187372708759e-06,
"loss": 1.2941,
"mean_token_accuracy": 0.6914515793323517,
"num_tokens": 369739968.0,
"step": 2840
},
{
"entropy": 1.987890625,
"epoch": 0.14815970056144728,
"grad_norm": 2.370904946985651,
"learning_rate": 8.782827741451389e-06,
"loss": 1.3239,
"mean_token_accuracy": 0.6835605353116989,
"num_tokens": 371040465.0,
"step": 2850
},
{
"entropy": 2.009375,
"epoch": 0.1486795591599085,
"grad_norm": 2.494085338086606,
"learning_rate": 8.777468110194019e-06,
"loss": 1.3613,
"mean_token_accuracy": 0.6772917687892914,
"num_tokens": 372336652.0,
"step": 2860
},
{
"entropy": 2.02421875,
"epoch": 0.14919941775836973,
"grad_norm": 2.662846593941635,
"learning_rate": 8.772108478936649e-06,
"loss": 1.3648,
"mean_token_accuracy": 0.6763909608125687,
"num_tokens": 373639925.0,
"step": 2870
},
{
"entropy": 2.00078125,
"epoch": 0.14971927635683094,
"grad_norm": 2.5244790450499446,
"learning_rate": 8.76674884767928e-06,
"loss": 1.3239,
"mean_token_accuracy": 0.6852716267108917,
"num_tokens": 374940750.0,
"step": 2880
},
{
"entropy": 1.99609375,
"epoch": 0.15023913495529215,
"grad_norm": 2.6111772819587395,
"learning_rate": 8.761389216421911e-06,
"loss": 1.3203,
"mean_token_accuracy": 0.6862067461013794,
"num_tokens": 376243176.0,
"step": 2890
},
{
"entropy": 2.01796875,
"epoch": 0.15075899355375338,
"grad_norm": 2.427344833467654,
"learning_rate": 8.756029585164541e-06,
"loss": 1.3772,
"mean_token_accuracy": 0.6720719128847122,
"num_tokens": 377544961.0,
"step": 2900
},
{
"epoch": 0.15075899355375338,
"eval_entropy": 1.9890625,
"eval_loss": 1.3540679216384888,
"eval_mean_token_accuracy": 0.6764335855841637,
"eval_num_tokens": 377544961.0,
"eval_runtime": 27.037,
"eval_samples_per_second": 93.723,
"eval_steps_per_second": 1.479,
"step": 2900
},
{
"entropy": 2.008203125,
"epoch": 0.1512788521522146,
"grad_norm": 2.623787264729052,
"learning_rate": 8.750669953907171e-06,
"loss": 1.293,
"mean_token_accuracy": 0.6916057914495468,
"num_tokens": 378847897.0,
"step": 2910
},
{
"entropy": 2.004296875,
"epoch": 0.1517987107506758,
"grad_norm": 2.8201179956244142,
"learning_rate": 8.745310322649803e-06,
"loss": 1.3522,
"mean_token_accuracy": 0.6789193272590637,
"num_tokens": 380147711.0,
"step": 2920
},
{
"entropy": 2.012890625,
"epoch": 0.15231856934913704,
"grad_norm": 2.504506206973736,
"learning_rate": 8.739950691392433e-06,
"loss": 1.3012,
"mean_token_accuracy": 0.6893252015113831,
"num_tokens": 381447257.0,
"step": 2930
},
{
"entropy": 1.9984375,
"epoch": 0.15283842794759825,
"grad_norm": 3.0347064329799753,
"learning_rate": 8.734591060135063e-06,
"loss": 1.3349,
"mean_token_accuracy": 0.6818037271499634,
"num_tokens": 382750228.0,
"step": 2940
},
{
"entropy": 2.005859375,
"epoch": 0.15335828654605946,
"grad_norm": 3.033374672693164,
"learning_rate": 8.729231428877693e-06,
"loss": 1.349,
"mean_token_accuracy": 0.677243971824646,
"num_tokens": 384051557.0,
"step": 2950
},
{
"entropy": 1.98671875,
"epoch": 0.1538781451445207,
"grad_norm": 2.7068525147559397,
"learning_rate": 8.723871797620325e-06,
"loss": 1.3232,
"mean_token_accuracy": 0.6841682553291321,
"num_tokens": 385354158.0,
"step": 2960
},
{
"entropy": 1.991796875,
"epoch": 0.1543980037429819,
"grad_norm": 2.477533880018443,
"learning_rate": 8.718512166362955e-06,
"loss": 1.3898,
"mean_token_accuracy": 0.6690465480089187,
"num_tokens": 386656193.0,
"step": 2970
},
{
"entropy": 2.001953125,
"epoch": 0.15491786234144314,
"grad_norm": 2.5264189363847165,
"learning_rate": 8.713152535105585e-06,
"loss": 1.3267,
"mean_token_accuracy": 0.6818209290504456,
"num_tokens": 387954360.0,
"step": 2980
},
{
"entropy": 1.986328125,
"epoch": 0.15543772093990435,
"grad_norm": 2.8619970024445074,
"learning_rate": 8.707792903848215e-06,
"loss": 1.318,
"mean_token_accuracy": 0.6846384286880494,
"num_tokens": 389256533.0,
"step": 2990
},
{
"entropy": 2.015234375,
"epoch": 0.15595757953836556,
"grad_norm": 2.402134420877998,
"learning_rate": 8.702433272590847e-06,
"loss": 1.3495,
"mean_token_accuracy": 0.6791374206542968,
"num_tokens": 390555737.0,
"step": 3000
},
{
"epoch": 0.15595757953836556,
"eval_entropy": 2.0166015625,
"eval_loss": 1.3420400619506836,
"eval_mean_token_accuracy": 0.6798159137368203,
"eval_num_tokens": 390555737.0,
"eval_runtime": 27.0128,
"eval_samples_per_second": 93.807,
"eval_steps_per_second": 1.481,
"step": 3000
},
{
"entropy": 2.012890625,
"epoch": 0.1564774381368268,
"grad_norm": 2.712488471425981,
"learning_rate": 8.697073641333477e-06,
"loss": 1.357,
"mean_token_accuracy": 0.6774081468582154,
"num_tokens": 391854248.0,
"step": 3010
},
{
"entropy": 2.030078125,
"epoch": 0.156997296735288,
"grad_norm": 3.029286445533591,
"learning_rate": 8.691714010076108e-06,
"loss": 1.2992,
"mean_token_accuracy": 0.690914198756218,
"num_tokens": 393156664.0,
"step": 3020
},
{
"entropy": 2.00234375,
"epoch": 0.15751715533374921,
"grad_norm": 2.549421364195977,
"learning_rate": 8.686354378818738e-06,
"loss": 1.3324,
"mean_token_accuracy": 0.6826860338449479,
"num_tokens": 394457749.0,
"step": 3030
},
{
"entropy": 1.95703125,
"epoch": 0.15803701393221045,
"grad_norm": 2.6946139036663572,
"learning_rate": 8.68099474756137e-06,
"loss": 1.2933,
"mean_token_accuracy": 0.6927355051040649,
"num_tokens": 395761962.0,
"step": 3040
},
{
"entropy": 2.014453125,
"epoch": 0.15855687253067166,
"grad_norm": 2.659494572770039,
"learning_rate": 8.675635116303998e-06,
"loss": 1.3212,
"mean_token_accuracy": 0.6813899397850036,
"num_tokens": 397066622.0,
"step": 3050
},
{
"entropy": 1.990234375,
"epoch": 0.15907673112913287,
"grad_norm": 3.0257146340036782,
"learning_rate": 8.67027548504663e-06,
"loss": 1.3041,
"mean_token_accuracy": 0.6869647830724717,
"num_tokens": 398367746.0,
"step": 3060
},
{
"entropy": 1.99375,
"epoch": 0.1595965897275941,
"grad_norm": 2.7044476167817435,
"learning_rate": 8.66491585378926e-06,
"loss": 1.2603,
"mean_token_accuracy": 0.6964340448379517,
"num_tokens": 399667719.0,
"step": 3070
},
{
"entropy": 1.947265625,
"epoch": 0.16011644832605532,
"grad_norm": 2.7140401230398763,
"learning_rate": 8.65955622253189e-06,
"loss": 1.3319,
"mean_token_accuracy": 0.6832780838012695,
"num_tokens": 400971140.0,
"step": 3080
},
{
"entropy": 2.014453125,
"epoch": 0.16063630692451653,
"grad_norm": 2.918491437563472,
"learning_rate": 8.65419659127452e-06,
"loss": 1.3342,
"mean_token_accuracy": 0.6809035211801528,
"num_tokens": 402274282.0,
"step": 3090
},
{
"entropy": 1.971875,
"epoch": 0.16115616552297776,
"grad_norm": 2.7732057095829523,
"learning_rate": 8.648836960017152e-06,
"loss": 1.2304,
"mean_token_accuracy": 0.7058033257722854,
"num_tokens": 403574862.0,
"step": 3100
},
{
"epoch": 0.16115616552297776,
"eval_entropy": 1.9900390625,
"eval_loss": 1.3368927240371704,
"eval_mean_token_accuracy": 0.6812232688069344,
"eval_num_tokens": 403574862.0,
"eval_runtime": 26.9888,
"eval_samples_per_second": 93.891,
"eval_steps_per_second": 1.482,
"step": 3100
},
{
"entropy": 2.011328125,
"epoch": 0.16167602412143897,
"grad_norm": 2.516195651589745,
"learning_rate": 8.643477328759782e-06,
"loss": 1.3139,
"mean_token_accuracy": 0.6866344630718231,
"num_tokens": 404877102.0,
"step": 3110
},
{
"entropy": 2.023046875,
"epoch": 0.16219588271990018,
"grad_norm": 2.8050918029508702,
"learning_rate": 8.638117697502412e-06,
"loss": 1.3172,
"mean_token_accuracy": 0.6869854718446732,
"num_tokens": 406174990.0,
"step": 3120
},
{
"entropy": 2.0234375,
"epoch": 0.16271574131836142,
"grad_norm": 2.287828406152705,
"learning_rate": 8.632758066245042e-06,
"loss": 1.3403,
"mean_token_accuracy": 0.680439755320549,
"num_tokens": 407477403.0,
"step": 3130
},
{
"entropy": 1.99921875,
"epoch": 0.16323559991682263,
"grad_norm": 2.6916159432386753,
"learning_rate": 8.627398434987674e-06,
"loss": 1.2908,
"mean_token_accuracy": 0.6886514484882355,
"num_tokens": 408776929.0,
"step": 3140
},
{
"entropy": 2.0109375,
"epoch": 0.16375545851528384,
"grad_norm": 2.532492301779656,
"learning_rate": 8.622038803730304e-06,
"loss": 1.2909,
"mean_token_accuracy": 0.6927838683128357,
"num_tokens": 410077344.0,
"step": 3150
},
{
"entropy": 1.96953125,
"epoch": 0.16427531711374507,
"grad_norm": 2.504364126745504,
"learning_rate": 8.616679172472934e-06,
"loss": 1.237,
"mean_token_accuracy": 0.7020116597414017,
"num_tokens": 411378804.0,
"step": 3160
},
{
"entropy": 2.01640625,
"epoch": 0.16479517571220628,
"grad_norm": 2.539504973347424,
"learning_rate": 8.611319541215564e-06,
"loss": 1.2895,
"mean_token_accuracy": 0.6912752240896225,
"num_tokens": 412679952.0,
"step": 3170
},
{
"entropy": 1.9890625,
"epoch": 0.1653150343106675,
"grad_norm": 2.78779166383061,
"learning_rate": 8.605959909958196e-06,
"loss": 1.367,
"mean_token_accuracy": 0.6729890555143356,
"num_tokens": 413980620.0,
"step": 3180
},
{
"entropy": 2.005078125,
"epoch": 0.16583489290912873,
"grad_norm": 2.857762333085046,
"learning_rate": 8.600600278700826e-06,
"loss": 1.3104,
"mean_token_accuracy": 0.6861138582229614,
"num_tokens": 415281204.0,
"step": 3190
},
{
"entropy": 2.0140625,
"epoch": 0.16635475150758994,
"grad_norm": 2.8126719227011794,
"learning_rate": 8.595240647443457e-06,
"loss": 1.3586,
"mean_token_accuracy": 0.6755324244499207,
"num_tokens": 416584005.0,
"step": 3200
},
{
"epoch": 0.16635475150758994,
"eval_entropy": 1.9892578125,
"eval_loss": 1.332979679107666,
"eval_mean_token_accuracy": 0.6817536249756813,
"eval_num_tokens": 416584005.0,
"eval_runtime": 27.0403,
"eval_samples_per_second": 93.712,
"eval_steps_per_second": 1.479,
"step": 3200
},
{
"entropy": 1.958203125,
"epoch": 0.16687461010605115,
"grad_norm": 2.527329884267192,
"learning_rate": 8.589881016186087e-06,
"loss": 1.2701,
"mean_token_accuracy": 0.6953856438398361,
"num_tokens": 417886755.0,
"step": 3210
},
{
"entropy": 1.998828125,
"epoch": 0.16739446870451238,
"grad_norm": 2.8303040621343674,
"learning_rate": 8.584521384928718e-06,
"loss": 1.2925,
"mean_token_accuracy": 0.692959401011467,
"num_tokens": 419191708.0,
"step": 3220
},
{
"entropy": 2.033984375,
"epoch": 0.1679143273029736,
"grad_norm": 2.9512079802506013,
"learning_rate": 8.579161753671349e-06,
"loss": 1.3584,
"mean_token_accuracy": 0.679707270860672,
"num_tokens": 420492166.0,
"step": 3230
},
{
"entropy": 2.020703125,
"epoch": 0.1684341859014348,
"grad_norm": 3.0911987676160964,
"learning_rate": 8.573802122413979e-06,
"loss": 1.297,
"mean_token_accuracy": 0.693993216753006,
"num_tokens": 421794872.0,
"step": 3240
},
{
"entropy": 2.05546875,
"epoch": 0.16895404449989604,
"grad_norm": 2.5517025238720414,
"learning_rate": 8.568442491156609e-06,
"loss": 1.2924,
"mean_token_accuracy": 0.6882816791534424,
"num_tokens": 423096086.0,
"step": 3250
},
{
"entropy": 2.031640625,
"epoch": 0.16947390309835725,
"grad_norm": 2.6138027777136044,
"learning_rate": 8.56308285989924e-06,
"loss": 1.3599,
"mean_token_accuracy": 0.6792479693889618,
"num_tokens": 424398022.0,
"step": 3260
},
{
"entropy": 2.03671875,
"epoch": 0.16999376169681846,
"grad_norm": 2.9756071137844016,
"learning_rate": 8.557723228641869e-06,
"loss": 1.3238,
"mean_token_accuracy": 0.6868531584739686,
"num_tokens": 425700797.0,
"step": 3270
},
{
"entropy": 2.044140625,
"epoch": 0.1705136202952797,
"grad_norm": 2.4922172520563044,
"learning_rate": 8.552363597384501e-06,
"loss": 1.3235,
"mean_token_accuracy": 0.6865602463483811,
"num_tokens": 427003805.0,
"step": 3280
},
{
"entropy": 2.008203125,
"epoch": 0.1710334788937409,
"grad_norm": 2.6717049657698615,
"learning_rate": 8.547003966127131e-06,
"loss": 1.2725,
"mean_token_accuracy": 0.6955862492322922,
"num_tokens": 428303464.0,
"step": 3290
},
{
"entropy": 2.041015625,
"epoch": 0.1715533374922021,
"grad_norm": 2.5719492034233284,
"learning_rate": 8.541644334869761e-06,
"loss": 1.3053,
"mean_token_accuracy": 0.6873748898506165,
"num_tokens": 429605993.0,
"step": 3300
},
{
"epoch": 0.1715533374922021,
"eval_entropy": 1.955859375,
"eval_loss": 1.323106288909912,
"eval_mean_token_accuracy": 0.6837490022182464,
"eval_num_tokens": 429605993.0,
"eval_runtime": 27.0297,
"eval_samples_per_second": 93.749,
"eval_steps_per_second": 1.48,
"step": 3300
},
{
"entropy": 1.978125,
"epoch": 0.17207319609066335,
"grad_norm": 2.523205090147758,
"learning_rate": 8.536284703612391e-06,
"loss": 1.2787,
"mean_token_accuracy": 0.6939565598964691,
"num_tokens": 430907381.0,
"step": 3310
},
{
"entropy": 2.005859375,
"epoch": 0.17259305468912456,
"grad_norm": 2.60512260688254,
"learning_rate": 8.530925072355023e-06,
"loss": 1.2896,
"mean_token_accuracy": 0.692984688282013,
"num_tokens": 432209977.0,
"step": 3320
},
{
"entropy": 1.98515625,
"epoch": 0.17311291328758577,
"grad_norm": 2.8771897714768757,
"learning_rate": 8.525565441097653e-06,
"loss": 1.3234,
"mean_token_accuracy": 0.6854825973510742,
"num_tokens": 433512715.0,
"step": 3330
},
{
"entropy": 2.016015625,
"epoch": 0.173632771886047,
"grad_norm": 2.4729343844799794,
"learning_rate": 8.520205809840283e-06,
"loss": 1.2783,
"mean_token_accuracy": 0.6929464250802994,
"num_tokens": 434816458.0,
"step": 3340
},
{
"entropy": 2.0078125,
"epoch": 0.17415263048450821,
"grad_norm": 2.7381887133802243,
"learning_rate": 8.514846178582913e-06,
"loss": 1.3193,
"mean_token_accuracy": 0.683475187420845,
"num_tokens": 436119244.0,
"step": 3350
},
{
"entropy": 1.98984375,
"epoch": 0.17467248908296942,
"grad_norm": 2.6715332120065103,
"learning_rate": 8.509486547325545e-06,
"loss": 1.2694,
"mean_token_accuracy": 0.6946573346853256,
"num_tokens": 437421645.0,
"step": 3360
},
{
"entropy": 1.990625,
"epoch": 0.17519234768143066,
"grad_norm": 2.6689245007680493,
"learning_rate": 8.504126916068175e-06,
"loss": 1.2839,
"mean_token_accuracy": 0.693576556444168,
"num_tokens": 438724839.0,
"step": 3370
},
{
"entropy": 1.992578125,
"epoch": 0.17571220627989187,
"grad_norm": 2.8395756091416904,
"learning_rate": 8.498767284810806e-06,
"loss": 1.2748,
"mean_token_accuracy": 0.6937979161739349,
"num_tokens": 440025116.0,
"step": 3380
},
{
"entropy": 2.016796875,
"epoch": 0.17623206487835308,
"grad_norm": 2.692005356181763,
"learning_rate": 8.493407653553436e-06,
"loss": 1.3114,
"mean_token_accuracy": 0.6859481185674667,
"num_tokens": 441328183.0,
"step": 3390
},
{
"entropy": 1.994921875,
"epoch": 0.17675192347681432,
"grad_norm": 2.868946104175745,
"learning_rate": 8.488048022296067e-06,
"loss": 1.2796,
"mean_token_accuracy": 0.6929744601249694,
"num_tokens": 442628565.0,
"step": 3400
},
{
"epoch": 0.17675192347681432,
"eval_entropy": 1.9943359375,
"eval_loss": 1.3183141946792603,
"eval_mean_token_accuracy": 0.6847504064440727,
"eval_num_tokens": 442628565.0,
"eval_runtime": 27.0644,
"eval_samples_per_second": 93.628,
"eval_steps_per_second": 1.478,
"step": 3400
},
{
"entropy": 2.009375,
"epoch": 0.17727178207527552,
"grad_norm": 2.4337424964078007,
"learning_rate": 8.482688391038698e-06,
"loss": 1.257,
"mean_token_accuracy": 0.6964388340711594,
"num_tokens": 443932477.0,
"step": 3410
},
{
"entropy": 2.00703125,
"epoch": 0.17779164067373673,
"grad_norm": 2.4398435548540713,
"learning_rate": 8.477328759781328e-06,
"loss": 1.3215,
"mean_token_accuracy": 0.6860674232244491,
"num_tokens": 445235027.0,
"step": 3420
},
{
"entropy": 2.01484375,
"epoch": 0.17831149927219797,
"grad_norm": 2.617082222937643,
"learning_rate": 8.471969128523958e-06,
"loss": 1.3375,
"mean_token_accuracy": 0.6841282606124878,
"num_tokens": 446538050.0,
"step": 3430
},
{
"entropy": 2.009375,
"epoch": 0.17883135787065918,
"grad_norm": 2.7321256317847773,
"learning_rate": 8.46660949726659e-06,
"loss": 1.321,
"mean_token_accuracy": 0.6862499833106994,
"num_tokens": 447839302.0,
"step": 3440
},
{
"entropy": 2.0171875,
"epoch": 0.1793512164691204,
"grad_norm": 2.6974604509712177,
"learning_rate": 8.46124986600922e-06,
"loss": 1.3052,
"mean_token_accuracy": 0.6891430199146271,
"num_tokens": 449142346.0,
"step": 3450
},
{
"entropy": 2.00390625,
"epoch": 0.17987107506758163,
"grad_norm": 2.7264701334593413,
"learning_rate": 8.45589023475185e-06,
"loss": 1.356,
"mean_token_accuracy": 0.6807908654212952,
"num_tokens": 450443178.0,
"step": 3460
},
{
"entropy": 1.988671875,
"epoch": 0.18039093366604284,
"grad_norm": 2.7002689956386643,
"learning_rate": 8.45053060349448e-06,
"loss": 1.2852,
"mean_token_accuracy": 0.692778468132019,
"num_tokens": 451744762.0,
"step": 3470
},
{
"entropy": 1.978515625,
"epoch": 0.18091079226450404,
"grad_norm": 2.741609991884545,
"learning_rate": 8.445170972237112e-06,
"loss": 1.3458,
"mean_token_accuracy": 0.677189102768898,
"num_tokens": 453045638.0,
"step": 3480
},
{
"entropy": 2.03046875,
"epoch": 0.18143065086296528,
"grad_norm": 2.3442319335334076,
"learning_rate": 8.43981134097974e-06,
"loss": 1.3438,
"mean_token_accuracy": 0.680471408367157,
"num_tokens": 454347947.0,
"step": 3490
},
{
"entropy": 2.009765625,
"epoch": 0.1819505094614265,
"grad_norm": 2.7523290252894324,
"learning_rate": 8.434451709722372e-06,
"loss": 1.3109,
"mean_token_accuracy": 0.6865796416997909,
"num_tokens": 455649492.0,
"step": 3500
},
{
"epoch": 0.1819505094614265,
"eval_entropy": 2.0052734375,
"eval_loss": 1.3143655061721802,
"eval_mean_token_accuracy": 0.6855434864759445,
"eval_num_tokens": 455649492.0,
"eval_runtime": 27.0247,
"eval_samples_per_second": 93.766,
"eval_steps_per_second": 1.48,
"step": 3500
},
{
"entropy": 1.99609375,
"epoch": 0.1824703680598877,
"grad_norm": 2.5665071503164936,
"learning_rate": 8.429092078465002e-06,
"loss": 1.2854,
"mean_token_accuracy": 0.6944643348455429,
"num_tokens": 456951845.0,
"step": 3510
},
{
"entropy": 1.991015625,
"epoch": 0.18299022665834894,
"grad_norm": 2.438561834050929,
"learning_rate": 8.423732447207632e-06,
"loss": 1.2758,
"mean_token_accuracy": 0.69437175989151,
"num_tokens": 458253374.0,
"step": 3520
},
{
"entropy": 1.96640625,
"epoch": 0.18351008525681015,
"grad_norm": 3.0673431878177144,
"learning_rate": 8.418372815950262e-06,
"loss": 1.2808,
"mean_token_accuracy": 0.69155213534832,
"num_tokens": 459555430.0,
"step": 3530
},
{
"entropy": 2.073046875,
"epoch": 0.18402994385527136,
"grad_norm": 2.7386691525736397,
"learning_rate": 8.413013184692894e-06,
"loss": 1.308,
"mean_token_accuracy": 0.6892219990491867,
"num_tokens": 460857872.0,
"step": 3540
},
{
"entropy": 1.930859375,
"epoch": 0.1845498024537326,
"grad_norm": 2.4647760007854846,
"learning_rate": 8.407653553435524e-06,
"loss": 1.282,
"mean_token_accuracy": 0.6934405654668808,
"num_tokens": 462157419.0,
"step": 3550
},
{
"entropy": 2.026953125,
"epoch": 0.1850696610521938,
"grad_norm": 2.5405250826454147,
"learning_rate": 8.402293922178155e-06,
"loss": 1.322,
"mean_token_accuracy": 0.6844764918088913,
"num_tokens": 463459803.0,
"step": 3560
},
{
"entropy": 1.939453125,
"epoch": 0.185589519650655,
"grad_norm": 2.636741248874388,
"learning_rate": 8.396934290920785e-06,
"loss": 1.2639,
"mean_token_accuracy": 0.6986570179462432,
"num_tokens": 464759174.0,
"step": 3570
},
{
"entropy": 2.0109375,
"epoch": 0.18610937824911625,
"grad_norm": 2.5076749130934965,
"learning_rate": 8.391574659663416e-06,
"loss": 1.2842,
"mean_token_accuracy": 0.6903380155563354,
"num_tokens": 466061583.0,
"step": 3580
},
{
"entropy": 1.962109375,
"epoch": 0.18662923684757746,
"grad_norm": 2.3636116706484103,
"learning_rate": 8.386215028406047e-06,
"loss": 1.3061,
"mean_token_accuracy": 0.688317421078682,
"num_tokens": 467363808.0,
"step": 3590
},
{
"entropy": 2.028125,
"epoch": 0.18714909544603867,
"grad_norm": 2.874949091760469,
"learning_rate": 8.380855397148677e-06,
"loss": 1.3182,
"mean_token_accuracy": 0.6887200832366943,
"num_tokens": 468666231.0,
"step": 3600
},
{
"epoch": 0.18714909544603867,
"eval_entropy": 1.9822265625,
"eval_loss": 1.308958649635315,
"eval_mean_token_accuracy": 0.6863628134131432,
"eval_num_tokens": 468666231.0,
"eval_runtime": 27.072,
"eval_samples_per_second": 93.602,
"eval_steps_per_second": 1.478,
"step": 3600
},
{
"entropy": 2.01171875,
"epoch": 0.1876689540444999,
"grad_norm": 2.791887464596478,
"learning_rate": 8.375495765891307e-06,
"loss": 1.204,
"mean_token_accuracy": 0.7074621051549912,
"num_tokens": 469970504.0,
"step": 3610
},
{
"entropy": 1.987890625,
"epoch": 0.1881888126429611,
"grad_norm": 2.722553955621076,
"learning_rate": 8.370136134633939e-06,
"loss": 1.2791,
"mean_token_accuracy": 0.6947989016771317,
"num_tokens": 471273322.0,
"step": 3620
},
{
"entropy": 1.9734375,
"epoch": 0.18870867124142232,
"grad_norm": 2.4391336595832156,
"learning_rate": 8.364776503376569e-06,
"loss": 1.2643,
"mean_token_accuracy": 0.6956757009029388,
"num_tokens": 472574776.0,
"step": 3630
},
{
"entropy": 2.012890625,
"epoch": 0.18922852983988356,
"grad_norm": 2.6069622941941293,
"learning_rate": 8.359416872119199e-06,
"loss": 1.3037,
"mean_token_accuracy": 0.6898593544960022,
"num_tokens": 473875431.0,
"step": 3640
},
{
"entropy": 1.976953125,
"epoch": 0.18974838843834477,
"grad_norm": 2.567210877725929,
"learning_rate": 8.354057240861829e-06,
"loss": 1.3121,
"mean_token_accuracy": 0.6894363254308701,
"num_tokens": 475177422.0,
"step": 3650
},
{
"entropy": 2.0046875,
"epoch": 0.19026824703680598,
"grad_norm": 2.709222740127343,
"learning_rate": 8.34869760960446e-06,
"loss": 1.2627,
"mean_token_accuracy": 0.6976218849420548,
"num_tokens": 476481085.0,
"step": 3660
},
{
"entropy": 1.982421875,
"epoch": 0.1907881056352672,
"grad_norm": 2.5324418986681105,
"learning_rate": 8.343337978347091e-06,
"loss": 1.2398,
"mean_token_accuracy": 0.7029396176338196,
"num_tokens": 477782950.0,
"step": 3670
},
{
"entropy": 2.007421875,
"epoch": 0.19130796423372842,
"grad_norm": 2.6446650653129673,
"learning_rate": 8.337978347089721e-06,
"loss": 1.2633,
"mean_token_accuracy": 0.6963121145963669,
"num_tokens": 479085834.0,
"step": 3680
},
{
"entropy": 2.001171875,
"epoch": 0.19182782283218963,
"grad_norm": 2.3918911540727334,
"learning_rate": 8.332618715832351e-06,
"loss": 1.2306,
"mean_token_accuracy": 0.7040454655885696,
"num_tokens": 480389075.0,
"step": 3690
},
{
"entropy": 1.971875,
"epoch": 0.19234768143065087,
"grad_norm": 2.8090615003673522,
"learning_rate": 8.327259084574983e-06,
"loss": 1.2888,
"mean_token_accuracy": 0.6914811164140702,
"num_tokens": 481689466.0,
"step": 3700
},
{
"epoch": 0.19234768143065087,
"eval_entropy": 1.98125,
"eval_loss": 1.305329442024231,
"eval_mean_token_accuracy": 0.6875247940421104,
"eval_num_tokens": 481689466.0,
"eval_runtime": 27.0036,
"eval_samples_per_second": 93.839,
"eval_steps_per_second": 1.481,
"step": 3700
},
{
"entropy": 1.986328125,
"epoch": 0.19286754002911208,
"grad_norm": 2.4978895167913238,
"learning_rate": 8.321899453317611e-06,
"loss": 1.2909,
"mean_token_accuracy": 0.6916386902332305,
"num_tokens": 482990667.0,
"step": 3710
},
{
"entropy": 1.95625,
"epoch": 0.1933873986275733,
"grad_norm": 2.627870609217679,
"learning_rate": 8.316539822060243e-06,
"loss": 1.2313,
"mean_token_accuracy": 0.7025312125682831,
"num_tokens": 484292456.0,
"step": 3720
},
{
"entropy": 1.9734375,
"epoch": 0.19390725722603452,
"grad_norm": 2.7206781186754743,
"learning_rate": 8.311180190802873e-06,
"loss": 1.2815,
"mean_token_accuracy": 0.6926350146532059,
"num_tokens": 485594945.0,
"step": 3730
},
{
"entropy": 1.97265625,
"epoch": 0.19442711582449573,
"grad_norm": 2.4546147295958027,
"learning_rate": 8.305820559545505e-06,
"loss": 1.3165,
"mean_token_accuracy": 0.6854215890169144,
"num_tokens": 486896310.0,
"step": 3740
},
{
"entropy": 1.987109375,
"epoch": 0.19494697442295694,
"grad_norm": 2.8530802124586963,
"learning_rate": 8.300460928288134e-06,
"loss": 1.2283,
"mean_token_accuracy": 0.708053606748581,
"num_tokens": 488198153.0,
"step": 3750
},
{
"entropy": 2.02265625,
"epoch": 0.19546683302141818,
"grad_norm": 2.4606670253073033,
"learning_rate": 8.295101297030765e-06,
"loss": 1.2523,
"mean_token_accuracy": 0.7028962075710297,
"num_tokens": 489501010.0,
"step": 3760
},
{
"entropy": 1.9484375,
"epoch": 0.1959866916198794,
"grad_norm": 2.497375010260242,
"learning_rate": 8.289741665773396e-06,
"loss": 1.262,
"mean_token_accuracy": 0.6947780072689056,
"num_tokens": 490802380.0,
"step": 3770
},
{
"entropy": 1.9703125,
"epoch": 0.1965065502183406,
"grad_norm": 2.6341600196231676,
"learning_rate": 8.284382034516026e-06,
"loss": 1.3042,
"mean_token_accuracy": 0.6902482360601425,
"num_tokens": 492102855.0,
"step": 3780
},
{
"entropy": 1.984765625,
"epoch": 0.19702640881680183,
"grad_norm": 2.739342860507875,
"learning_rate": 8.279022403258656e-06,
"loss": 1.2691,
"mean_token_accuracy": 0.6948835045099259,
"num_tokens": 493404317.0,
"step": 3790
},
{
"entropy": 1.98515625,
"epoch": 0.19754626741526304,
"grad_norm": 2.791173949344963,
"learning_rate": 8.273662772001288e-06,
"loss": 1.3142,
"mean_token_accuracy": 0.6858647584915161,
"num_tokens": 494705781.0,
"step": 3800
},
{
"epoch": 0.19754626741526304,
"eval_entropy": 1.980078125,
"eval_loss": 1.3015981912612915,
"eval_mean_token_accuracy": 0.6882883310317993,
"eval_num_tokens": 494705781.0,
"eval_runtime": 27.0681,
"eval_samples_per_second": 93.616,
"eval_steps_per_second": 1.478,
"step": 3800
},
{
"entropy": 2.00859375,
"epoch": 0.19806612601372428,
"grad_norm": 2.534698212392668,
"learning_rate": 8.268303140743918e-06,
"loss": 1.2944,
"mean_token_accuracy": 0.6921063899993897,
"num_tokens": 496009169.0,
"step": 3810
},
{
"entropy": 1.973828125,
"epoch": 0.1985859846121855,
"grad_norm": 2.470064591884078,
"learning_rate": 8.262943509486548e-06,
"loss": 1.3613,
"mean_token_accuracy": 0.6791955471038819,
"num_tokens": 497311535.0,
"step": 3820
},
{
"entropy": 1.99921875,
"epoch": 0.1991058432106467,
"grad_norm": 2.7951788308052845,
"learning_rate": 8.257583878229178e-06,
"loss": 1.2318,
"mean_token_accuracy": 0.7071437776088715,
"num_tokens": 498614592.0,
"step": 3830
},
{
"entropy": 2.020703125,
"epoch": 0.19962570180910794,
"grad_norm": 2.492487923436212,
"learning_rate": 8.25222424697181e-06,
"loss": 1.313,
"mean_token_accuracy": 0.6868307888507843,
"num_tokens": 499916664.0,
"step": 3840
},
{
"entropy": 2.009375,
"epoch": 0.20014556040756915,
"grad_norm": 2.6167655707165145,
"learning_rate": 8.24686461571444e-06,
"loss": 1.2949,
"mean_token_accuracy": 0.6911433279514313,
"num_tokens": 501219159.0,
"step": 3850
},
{
"entropy": 1.98515625,
"epoch": 0.20066541900603035,
"grad_norm": 2.331722970763142,
"learning_rate": 8.24150498445707e-06,
"loss": 1.28,
"mean_token_accuracy": 0.6937197148799896,
"num_tokens": 502522148.0,
"step": 3860
},
{
"entropy": 2.0046875,
"epoch": 0.2011852776044916,
"grad_norm": 2.8402910369022396,
"learning_rate": 8.2361453531997e-06,
"loss": 1.2966,
"mean_token_accuracy": 0.6902586638927459,
"num_tokens": 503823802.0,
"step": 3870
},
{
"entropy": 1.971484375,
"epoch": 0.2017051362029528,
"grad_norm": 2.60745005509188,
"learning_rate": 8.230785721942332e-06,
"loss": 1.3045,
"mean_token_accuracy": 0.6883846968412399,
"num_tokens": 505124519.0,
"step": 3880
},
{
"entropy": 1.9828125,
"epoch": 0.202224994801414,
"grad_norm": 2.8247292680192007,
"learning_rate": 8.225426090684962e-06,
"loss": 1.2928,
"mean_token_accuracy": 0.690139764547348,
"num_tokens": 506425627.0,
"step": 3890
},
{
"entropy": 1.97890625,
"epoch": 0.20274485339987525,
"grad_norm": 2.7618637289665964,
"learning_rate": 8.220066459427592e-06,
"loss": 1.2556,
"mean_token_accuracy": 0.6991171002388,
"num_tokens": 507727622.0,
"step": 3900
},
{
"epoch": 0.20274485339987525,
"eval_entropy": 1.9361328125,
"eval_loss": 1.2959157228469849,
"eval_mean_token_accuracy": 0.6889487415552139,
"eval_num_tokens": 507727622.0,
"eval_runtime": 27.0219,
"eval_samples_per_second": 93.776,
"eval_steps_per_second": 1.48,
"step": 3900
},
{
"entropy": 1.957421875,
"epoch": 0.20326471199833646,
"grad_norm": 2.697529046630667,
"learning_rate": 8.214706828170222e-06,
"loss": 1.2216,
"mean_token_accuracy": 0.708197021484375,
"num_tokens": 509029655.0,
"step": 3910
},
{
"entropy": 1.923828125,
"epoch": 0.20378457059679767,
"grad_norm": 2.4987978086734435,
"learning_rate": 8.209347196912854e-06,
"loss": 1.273,
"mean_token_accuracy": 0.6991938650608063,
"num_tokens": 510331697.0,
"step": 3920
},
{
"entropy": 1.996875,
"epoch": 0.2043044291952589,
"grad_norm": 2.541166684942553,
"learning_rate": 8.203987565655484e-06,
"loss": 1.3083,
"mean_token_accuracy": 0.6929676026105881,
"num_tokens": 511634148.0,
"step": 3930
},
{
"entropy": 1.980078125,
"epoch": 0.2048242877937201,
"grad_norm": 3.3719193163838392,
"learning_rate": 8.198627934398114e-06,
"loss": 1.2598,
"mean_token_accuracy": 0.6964832454919815,
"num_tokens": 512936685.0,
"step": 3940
},
{
"entropy": 1.995703125,
"epoch": 0.20534414639218132,
"grad_norm": 2.402589221143996,
"learning_rate": 8.193268303140745e-06,
"loss": 1.2753,
"mean_token_accuracy": 0.6921638280153275,
"num_tokens": 514239946.0,
"step": 3950
},
{
"entropy": 1.969140625,
"epoch": 0.20586400499064256,
"grad_norm": 2.705219812521976,
"learning_rate": 8.187908671883376e-06,
"loss": 1.2852,
"mean_token_accuracy": 0.6914204746484757,
"num_tokens": 515541879.0,
"step": 3960
},
{
"entropy": 1.943359375,
"epoch": 0.20638386358910377,
"grad_norm": 2.521041085825172,
"learning_rate": 8.182549040626005e-06,
"loss": 1.2858,
"mean_token_accuracy": 0.6972074031829834,
"num_tokens": 516845304.0,
"step": 3970
},
{
"entropy": 1.9578125,
"epoch": 0.20690372218756498,
"grad_norm": 2.8151569839490684,
"learning_rate": 8.177189409368637e-06,
"loss": 1.2536,
"mean_token_accuracy": 0.6993647307157517,
"num_tokens": 518147170.0,
"step": 3980
},
{
"entropy": 1.952734375,
"epoch": 0.2074235807860262,
"grad_norm": 2.772263110047239,
"learning_rate": 8.171829778111267e-06,
"loss": 1.239,
"mean_token_accuracy": 0.7034797668457031,
"num_tokens": 519449157.0,
"step": 3990
},
{
"entropy": 1.930859375,
"epoch": 0.20794343938448742,
"grad_norm": 2.309370772097019,
"learning_rate": 8.166470146853897e-06,
"loss": 1.3225,
"mean_token_accuracy": 0.6844330430030823,
"num_tokens": 520750553.0,
"step": 4000
},
{
"epoch": 0.20794343938448742,
"eval_entropy": 1.9474609375,
"eval_loss": 1.293915867805481,
"eval_mean_token_accuracy": 0.6892638981342316,
"eval_num_tokens": 520750553.0,
"eval_runtime": 26.9809,
"eval_samples_per_second": 93.918,
"eval_steps_per_second": 1.483,
"step": 4000
},
{
"entropy": 1.979296875,
"epoch": 0.20846329798294863,
"grad_norm": 2.7243920225408917,
"learning_rate": 8.161110515596527e-06,
"loss": 1.2604,
"mean_token_accuracy": 0.6993803977966309,
"num_tokens": 522051854.0,
"step": 4010
},
{
"entropy": 1.98828125,
"epoch": 0.20898315658140987,
"grad_norm": 2.4033750829753884,
"learning_rate": 8.155750884339159e-06,
"loss": 1.2924,
"mean_token_accuracy": 0.6909846574068069,
"num_tokens": 523353191.0,
"step": 4020
},
{
"entropy": 1.966015625,
"epoch": 0.20950301517987108,
"grad_norm": 2.4125808653062255,
"learning_rate": 8.150391253081789e-06,
"loss": 1.2456,
"mean_token_accuracy": 0.7014933854341507,
"num_tokens": 524656877.0,
"step": 4030
},
{
"entropy": 1.941796875,
"epoch": 0.2100228737783323,
"grad_norm": 2.761482824549148,
"learning_rate": 8.145031621824419e-06,
"loss": 1.284,
"mean_token_accuracy": 0.6947420746088028,
"num_tokens": 525959737.0,
"step": 4040
},
{
"entropy": 1.958203125,
"epoch": 0.21054273237679352,
"grad_norm": 3.087415419775397,
"learning_rate": 8.13967199056705e-06,
"loss": 1.2438,
"mean_token_accuracy": 0.7026516288518906,
"num_tokens": 527261861.0,
"step": 4050
},
{
"entropy": 1.983203125,
"epoch": 0.21106259097525473,
"grad_norm": 3.3365103963970064,
"learning_rate": 8.134312359309681e-06,
"loss": 1.2278,
"mean_token_accuracy": 0.7061191201210022,
"num_tokens": 528564862.0,
"step": 4060
},
{
"entropy": 1.96953125,
"epoch": 0.21158244957371594,
"grad_norm": 2.8661334089703967,
"learning_rate": 8.128952728052311e-06,
"loss": 1.2967,
"mean_token_accuracy": 0.6898200452327728,
"num_tokens": 529865367.0,
"step": 4070
},
{
"entropy": 1.98125,
"epoch": 0.21210230817217718,
"grad_norm": 2.8587404182379688,
"learning_rate": 8.123593096794941e-06,
"loss": 1.3221,
"mean_token_accuracy": 0.6857372492551803,
"num_tokens": 531169755.0,
"step": 4080
},
{
"entropy": 1.970703125,
"epoch": 0.2126221667706384,
"grad_norm": 2.7418586676102157,
"learning_rate": 8.118233465537571e-06,
"loss": 1.2394,
"mean_token_accuracy": 0.7002495944499969,
"num_tokens": 532472304.0,
"step": 4090
},
{
"entropy": 1.96796875,
"epoch": 0.2131420253690996,
"grad_norm": 2.2730523707204378,
"learning_rate": 8.112873834280203e-06,
"loss": 1.2728,
"mean_token_accuracy": 0.6959463089704514,
"num_tokens": 533775711.0,
"step": 4100
},
{
"epoch": 0.2131420253690996,
"eval_entropy": 1.9625,
"eval_loss": 1.2918890714645386,
"eval_mean_token_accuracy": 0.6903177663683892,
"eval_num_tokens": 533775711.0,
"eval_runtime": 27.0558,
"eval_samples_per_second": 93.658,
"eval_steps_per_second": 1.478,
"step": 4100
},
{
"entropy": 1.9890625,
"epoch": 0.21366188396756083,
"grad_norm": 2.505371363894896,
"learning_rate": 8.107514203022833e-06,
"loss": 1.2692,
"mean_token_accuracy": 0.694954690337181,
"num_tokens": 535076281.0,
"step": 4110
},
{
"entropy": 1.97890625,
"epoch": 0.21418174256602204,
"grad_norm": 2.632991263138149,
"learning_rate": 8.102154571765463e-06,
"loss": 1.2384,
"mean_token_accuracy": 0.7034430623054504,
"num_tokens": 536377500.0,
"step": 4120
},
{
"entropy": 1.97421875,
"epoch": 0.21470160116448325,
"grad_norm": 2.7053712123242892,
"learning_rate": 8.096794940508094e-06,
"loss": 1.2692,
"mean_token_accuracy": 0.6977686077356339,
"num_tokens": 537680545.0,
"step": 4130
},
{
"entropy": 1.945703125,
"epoch": 0.2152214597629445,
"grad_norm": 2.799276783016702,
"learning_rate": 8.091435309250725e-06,
"loss": 1.2984,
"mean_token_accuracy": 0.6907071709632874,
"num_tokens": 538982214.0,
"step": 4140
},
{
"entropy": 1.976953125,
"epoch": 0.2157413183614057,
"grad_norm": 2.432775626114827,
"learning_rate": 8.086075677993355e-06,
"loss": 1.2963,
"mean_token_accuracy": 0.6919291108846665,
"num_tokens": 540284485.0,
"step": 4150
},
{
"entropy": 1.973046875,
"epoch": 0.2162611769598669,
"grad_norm": 2.4249770641048363,
"learning_rate": 8.080716046735986e-06,
"loss": 1.2991,
"mean_token_accuracy": 0.687791109085083,
"num_tokens": 541585054.0,
"step": 4160
},
{
"entropy": 1.98046875,
"epoch": 0.21678103555832814,
"grad_norm": 2.991384722555285,
"learning_rate": 8.075356415478616e-06,
"loss": 1.2873,
"mean_token_accuracy": 0.6940852135419846,
"num_tokens": 542884636.0,
"step": 4170
},
{
"entropy": 1.969140625,
"epoch": 0.21730089415678935,
"grad_norm": 3.1558672164409134,
"learning_rate": 8.069996784221248e-06,
"loss": 1.2511,
"mean_token_accuracy": 0.6998495638370514,
"num_tokens": 544184740.0,
"step": 4180
},
{
"entropy": 1.94375,
"epoch": 0.21782075275525056,
"grad_norm": 2.8617532745872136,
"learning_rate": 8.064637152963876e-06,
"loss": 1.2617,
"mean_token_accuracy": 0.6979638487100601,
"num_tokens": 545487560.0,
"step": 4190
},
{
"entropy": 1.983984375,
"epoch": 0.2183406113537118,
"grad_norm": 2.410635918096622,
"learning_rate": 8.059277521706508e-06,
"loss": 1.286,
"mean_token_accuracy": 0.6914912313222885,
"num_tokens": 546784877.0,
"step": 4200
},
{
"epoch": 0.2183406113537118,
"eval_entropy": 1.9771484375,
"eval_loss": 1.2872940301895142,
"eval_mean_token_accuracy": 0.6916115820407868,
"eval_num_tokens": 546784877.0,
"eval_runtime": 26.9692,
"eval_samples_per_second": 93.959,
"eval_steps_per_second": 1.483,
"step": 4200
},
{
"entropy": 1.966015625,
"epoch": 0.218860469952173,
"grad_norm": 2.563933512523068,
"learning_rate": 8.053917890449138e-06,
"loss": 1.3183,
"mean_token_accuracy": 0.6864300340414047,
"num_tokens": 548085990.0,
"step": 4210
},
{
"entropy": 1.957421875,
"epoch": 0.21938032855063422,
"grad_norm": 2.672627927496643,
"learning_rate": 8.048558259191768e-06,
"loss": 1.2172,
"mean_token_accuracy": 0.7049766331911087,
"num_tokens": 549388453.0,
"step": 4220
},
{
"entropy": 1.96015625,
"epoch": 0.21990018714909546,
"grad_norm": 2.7305334701622486,
"learning_rate": 8.043198627934398e-06,
"loss": 1.2864,
"mean_token_accuracy": 0.6947871923446656,
"num_tokens": 550688889.0,
"step": 4230
},
{
"entropy": 2.003515625,
"epoch": 0.22042004574755666,
"grad_norm": 2.725722402094167,
"learning_rate": 8.03783899667703e-06,
"loss": 1.236,
"mean_token_accuracy": 0.7025877058506012,
"num_tokens": 551991734.0,
"step": 4240
},
{
"entropy": 1.9578125,
"epoch": 0.22093990434601787,
"grad_norm": 2.522171785903581,
"learning_rate": 8.03247936541966e-06,
"loss": 1.2953,
"mean_token_accuracy": 0.6870299607515336,
"num_tokens": 553292475.0,
"step": 4250
},
{
"entropy": 1.977734375,
"epoch": 0.2214597629444791,
"grad_norm": 2.817799279564371,
"learning_rate": 8.02711973416229e-06,
"loss": 1.264,
"mean_token_accuracy": 0.6977929383516311,
"num_tokens": 554595117.0,
"step": 4260
},
{
"entropy": 1.97890625,
"epoch": 0.22197962154294032,
"grad_norm": 2.4648277748686844,
"learning_rate": 8.02176010290492e-06,
"loss": 1.2974,
"mean_token_accuracy": 0.6897315353155136,
"num_tokens": 555897412.0,
"step": 4270
},
{
"entropy": 1.97578125,
"epoch": 0.22249948014140153,
"grad_norm": 2.7628151573610946,
"learning_rate": 8.01640047164755e-06,
"loss": 1.2755,
"mean_token_accuracy": 0.6931395530700684,
"num_tokens": 557199556.0,
"step": 4280
},
{
"entropy": 1.93671875,
"epoch": 0.22301933873986277,
"grad_norm": 3.0867003123761823,
"learning_rate": 8.011040840390182e-06,
"loss": 1.2714,
"mean_token_accuracy": 0.6966659635305404,
"num_tokens": 558500766.0,
"step": 4290
},
{
"entropy": 1.972265625,
"epoch": 0.22353919733832397,
"grad_norm": 2.778052082079629,
"learning_rate": 8.005681209132812e-06,
"loss": 1.273,
"mean_token_accuracy": 0.695733654499054,
"num_tokens": 559803780.0,
"step": 4300
},
{
"epoch": 0.22353919733832397,
"eval_entropy": 1.938671875,
"eval_loss": 1.2812213897705078,
"eval_mean_token_accuracy": 0.6922902554273606,
"eval_num_tokens": 559803780.0,
"eval_runtime": 27.0434,
"eval_samples_per_second": 93.701,
"eval_steps_per_second": 1.479,
"step": 4300
},
{
"entropy": 1.95390625,
"epoch": 0.22405905593678518,
"grad_norm": 2.567467303473037,
"learning_rate": 8.000321577875443e-06,
"loss": 1.2922,
"mean_token_accuracy": 0.6903115957975388,
"num_tokens": 561104473.0,
"step": 4310
},
{
"entropy": 1.911328125,
"epoch": 0.22457891453524642,
"grad_norm": 2.817023476007252,
"learning_rate": 7.994961946618073e-06,
"loss": 1.2239,
"mean_token_accuracy": 0.7066559463739395,
"num_tokens": 562407193.0,
"step": 4320
},
{
"entropy": 1.958984375,
"epoch": 0.22509877313370763,
"grad_norm": 2.5880717553221007,
"learning_rate": 7.989602315360704e-06,
"loss": 1.2349,
"mean_token_accuracy": 0.70477976500988,
"num_tokens": 563707267.0,
"step": 4330
},
{
"entropy": 1.9484375,
"epoch": 0.22561863173216884,
"grad_norm": 2.8274250167618398,
"learning_rate": 7.984242684103335e-06,
"loss": 1.2449,
"mean_token_accuracy": 0.703673678636551,
"num_tokens": 565008099.0,
"step": 4340
},
{
"entropy": 1.95078125,
"epoch": 0.22613849033063008,
"grad_norm": 2.431494361366356,
"learning_rate": 7.978883052845965e-06,
"loss": 1.293,
"mean_token_accuracy": 0.6909309953451157,
"num_tokens": 566309878.0,
"step": 4350
},
{
"entropy": 1.97265625,
"epoch": 0.22665834892909129,
"grad_norm": 2.9158414524326615,
"learning_rate": 7.973523421588595e-06,
"loss": 1.2485,
"mean_token_accuracy": 0.6994311362504959,
"num_tokens": 567611888.0,
"step": 4360
},
{
"entropy": 1.951953125,
"epoch": 0.2271782075275525,
"grad_norm": 2.818797463966022,
"learning_rate": 7.968163790331227e-06,
"loss": 1.2632,
"mean_token_accuracy": 0.6981148570775986,
"num_tokens": 568915332.0,
"step": 4370
},
{
"entropy": 1.959375,
"epoch": 0.22769806612601373,
"grad_norm": 2.664253370826872,
"learning_rate": 7.962804159073855e-06,
"loss": 1.2747,
"mean_token_accuracy": 0.6949633628129959,
"num_tokens": 570216135.0,
"step": 4380
},
{
"entropy": 1.959375,
"epoch": 0.22821792472447494,
"grad_norm": 2.6872459393832733,
"learning_rate": 7.957444527816487e-06,
"loss": 1.2417,
"mean_token_accuracy": 0.7018421232700348,
"num_tokens": 571515176.0,
"step": 4390
},
{
"entropy": 1.920703125,
"epoch": 0.22873778332293615,
"grad_norm": 3.1655730275105087,
"learning_rate": 7.952084896559117e-06,
"loss": 1.1863,
"mean_token_accuracy": 0.7133954465389252,
"num_tokens": 572817502.0,
"step": 4400
},
{
"epoch": 0.22873778332293615,
"eval_entropy": 1.9025390625,
"eval_loss": 1.2765703201293945,
"eval_mean_token_accuracy": 0.6941077694296837,
"eval_num_tokens": 572817502.0,
"eval_runtime": 26.992,
"eval_samples_per_second": 93.88,
"eval_steps_per_second": 1.482,
"step": 4400
},
{
"entropy": 1.9671875,
"epoch": 0.2292576419213974,
"grad_norm": 2.5225591095126503,
"learning_rate": 7.946725265301747e-06,
"loss": 1.2535,
"mean_token_accuracy": 0.7007995426654816,
"num_tokens": 574119446.0,
"step": 4410
},
{
"entropy": 1.964453125,
"epoch": 0.2297775005198586,
"grad_norm": 2.6196134337531594,
"learning_rate": 7.941365634044377e-06,
"loss": 1.2899,
"mean_token_accuracy": 0.6930925190448761,
"num_tokens": 575421333.0,
"step": 4420
},
{
"entropy": 1.96640625,
"epoch": 0.2302973591183198,
"grad_norm": 2.660438735542575,
"learning_rate": 7.936006002787009e-06,
"loss": 1.238,
"mean_token_accuracy": 0.7024450659751892,
"num_tokens": 576724339.0,
"step": 4430
},
{
"entropy": 1.986328125,
"epoch": 0.23081721771678104,
"grad_norm": 2.8884538412502025,
"learning_rate": 7.93064637152964e-06,
"loss": 1.2791,
"mean_token_accuracy": 0.6945829808712005,
"num_tokens": 578026159.0,
"step": 4440
},
{
"entropy": 1.930859375,
"epoch": 0.23133707631524225,
"grad_norm": 2.491346355104527,
"learning_rate": 7.92528674027227e-06,
"loss": 1.2697,
"mean_token_accuracy": 0.6986850589513779,
"num_tokens": 579326836.0,
"step": 4450
},
{
"entropy": 1.996484375,
"epoch": 0.23185693491370346,
"grad_norm": 2.7792443799151916,
"learning_rate": 7.9199271090149e-06,
"loss": 1.2505,
"mean_token_accuracy": 0.7027821063995361,
"num_tokens": 580628261.0,
"step": 4460
},
{
"entropy": 1.9625,
"epoch": 0.2323767935121647,
"grad_norm": 2.7762796020397147,
"learning_rate": 7.914567477757531e-06,
"loss": 1.2498,
"mean_token_accuracy": 0.7014061719179153,
"num_tokens": 581931706.0,
"step": 4470
},
{
"entropy": 1.975,
"epoch": 0.2328966521106259,
"grad_norm": 2.6204349058972505,
"learning_rate": 7.909207846500161e-06,
"loss": 1.2775,
"mean_token_accuracy": 0.6945300877094269,
"num_tokens": 583234826.0,
"step": 4480
},
{
"entropy": 1.981640625,
"epoch": 0.23341651070908712,
"grad_norm": 2.7785022938050505,
"learning_rate": 7.903848215242792e-06,
"loss": 1.2204,
"mean_token_accuracy": 0.7053841650485992,
"num_tokens": 584537431.0,
"step": 4490
},
{
"entropy": 1.939453125,
"epoch": 0.23393636930754835,
"grad_norm": 2.7796243528329576,
"learning_rate": 7.898488583985422e-06,
"loss": 1.2486,
"mean_token_accuracy": 0.6998362272977829,
"num_tokens": 585838607.0,
"step": 4500
},
{
"epoch": 0.23393636930754835,
"eval_entropy": 1.947265625,
"eval_loss": 1.2723462581634521,
"eval_mean_token_accuracy": 0.6942752122879028,
"eval_num_tokens": 585838607.0,
"eval_runtime": 27.005,
"eval_samples_per_second": 93.834,
"eval_steps_per_second": 1.481,
"step": 4500
},
{
"entropy": 1.953515625,
"epoch": 0.23445622790600956,
"grad_norm": 2.9422313134137856,
"learning_rate": 7.893128952728053e-06,
"loss": 1.2936,
"mean_token_accuracy": 0.6911924839019775,
"num_tokens": 587140996.0,
"step": 4510
},
{
"entropy": 1.94609375,
"epoch": 0.23497608650447077,
"grad_norm": 2.7066539753403656,
"learning_rate": 7.887769321470684e-06,
"loss": 1.2479,
"mean_token_accuracy": 0.6973439574241638,
"num_tokens": 588442997.0,
"step": 4520
},
{
"entropy": 1.985546875,
"epoch": 0.235495945102932,
"grad_norm": 2.397851381725743,
"learning_rate": 7.882409690213314e-06,
"loss": 1.2953,
"mean_token_accuracy": 0.6903478235006333,
"num_tokens": 589743748.0,
"step": 4530
},
{
"entropy": 1.9875,
"epoch": 0.23601580370139322,
"grad_norm": 2.229749220703251,
"learning_rate": 7.877050058955944e-06,
"loss": 1.2229,
"mean_token_accuracy": 0.7060524135828018,
"num_tokens": 591047698.0,
"step": 4540
},
{
"entropy": 1.941015625,
"epoch": 0.23653566229985443,
"grad_norm": 3.0382881187950495,
"learning_rate": 7.871690427698576e-06,
"loss": 1.2092,
"mean_token_accuracy": 0.7101279973983765,
"num_tokens": 592350904.0,
"step": 4550
},
{
"entropy": 1.963671875,
"epoch": 0.23705552089831566,
"grad_norm": 2.8037789296932103,
"learning_rate": 7.866330796441206e-06,
"loss": 1.2233,
"mean_token_accuracy": 0.7067608326673508,
"num_tokens": 593651874.0,
"step": 4560
},
{
"entropy": 1.98828125,
"epoch": 0.23757537949677687,
"grad_norm": 3.0252127647852984,
"learning_rate": 7.860971165183836e-06,
"loss": 1.2796,
"mean_token_accuracy": 0.6942305028438568,
"num_tokens": 594954491.0,
"step": 4570
},
{
"entropy": 1.961328125,
"epoch": 0.23809523809523808,
"grad_norm": 3.0716444124866036,
"learning_rate": 7.855611533926466e-06,
"loss": 1.3134,
"mean_token_accuracy": 0.6880674183368682,
"num_tokens": 596258329.0,
"step": 4580
},
{
"entropy": 2.012890625,
"epoch": 0.23861509669369932,
"grad_norm": 2.7237861128600156,
"learning_rate": 7.850251902669098e-06,
"loss": 1.2167,
"mean_token_accuracy": 0.7102824777364731,
"num_tokens": 597560934.0,
"step": 4590
},
{
"entropy": 1.9734375,
"epoch": 0.23913495529216053,
"grad_norm": 2.4614198874993627,
"learning_rate": 7.844892271411726e-06,
"loss": 1.2389,
"mean_token_accuracy": 0.7028800249099731,
"num_tokens": 598861678.0,
"step": 4600
},
{
"epoch": 0.23913495529216053,
"eval_entropy": 1.9740234375,
"eval_loss": 1.267970085144043,
"eval_mean_token_accuracy": 0.6954253003001213,
"eval_num_tokens": 598861678.0,
"eval_runtime": 26.9936,
"eval_samples_per_second": 93.874,
"eval_steps_per_second": 1.482,
"step": 4600
},
{
"entropy": 2.02265625,
"epoch": 0.23965481389062174,
"grad_norm": 2.754118862167845,
"learning_rate": 7.839532640154358e-06,
"loss": 1.272,
"mean_token_accuracy": 0.6969685256481171,
"num_tokens": 600164130.0,
"step": 4610
},
{
"entropy": 1.971875,
"epoch": 0.24017467248908297,
"grad_norm": 2.8445732961587757,
"learning_rate": 7.834173008896988e-06,
"loss": 1.283,
"mean_token_accuracy": 0.6953945219516754,
"num_tokens": 601467242.0,
"step": 4620
},
{
"entropy": 1.957421875,
"epoch": 0.24069453108754418,
"grad_norm": 2.889952630402841,
"learning_rate": 7.828813377639618e-06,
"loss": 1.2016,
"mean_token_accuracy": 0.7125820994377137,
"num_tokens": 602769094.0,
"step": 4630
},
{
"entropy": 1.973046875,
"epoch": 0.2412143896860054,
"grad_norm": 2.8984992124069984,
"learning_rate": 7.823453746382248e-06,
"loss": 1.2051,
"mean_token_accuracy": 0.7100563675165177,
"num_tokens": 604069158.0,
"step": 4640
},
{
"entropy": 1.95859375,
"epoch": 0.24173424828446663,
"grad_norm": 2.779076274595125,
"learning_rate": 7.81809411512488e-06,
"loss": 1.2804,
"mean_token_accuracy": 0.6951503753662109,
"num_tokens": 605369563.0,
"step": 4650
},
{
"entropy": 1.98828125,
"epoch": 0.24225410688292784,
"grad_norm": 2.6466330299849274,
"learning_rate": 7.81273448386751e-06,
"loss": 1.2673,
"mean_token_accuracy": 0.6982070684432984,
"num_tokens": 606673467.0,
"step": 4660
},
{
"entropy": 1.9296875,
"epoch": 0.24277396548138908,
"grad_norm": 3.273098206066536,
"learning_rate": 7.80737485261014e-06,
"loss": 1.2581,
"mean_token_accuracy": 0.698051118850708,
"num_tokens": 607975841.0,
"step": 4670
},
{
"entropy": 2.016796875,
"epoch": 0.24329382407985028,
"grad_norm": 3.0443250662788888,
"learning_rate": 7.80201522135277e-06,
"loss": 1.263,
"mean_token_accuracy": 0.6979148924350739,
"num_tokens": 609275582.0,
"step": 4680
},
{
"entropy": 1.95859375,
"epoch": 0.2438136826783115,
"grad_norm": 2.600061039698163,
"learning_rate": 7.796655590095402e-06,
"loss": 1.2929,
"mean_token_accuracy": 0.6898707449436188,
"num_tokens": 610577817.0,
"step": 4690
},
{
"entropy": 1.961328125,
"epoch": 0.24433354127677273,
"grad_norm": 2.938702598725605,
"learning_rate": 7.791295958838033e-06,
"loss": 1.2697,
"mean_token_accuracy": 0.69442478120327,
"num_tokens": 611878682.0,
"step": 4700
},
{
"epoch": 0.24433354127677273,
"eval_entropy": 1.941015625,
"eval_loss": 1.2665934562683105,
"eval_mean_token_accuracy": 0.695571257174015,
"eval_num_tokens": 611878682.0,
"eval_runtime": 27.0148,
"eval_samples_per_second": 93.801,
"eval_steps_per_second": 1.481,
"step": 4700
},
{
"entropy": 1.93359375,
"epoch": 0.24485339987523394,
"grad_norm": 2.7431271514342352,
"learning_rate": 7.785936327580663e-06,
"loss": 1.1924,
"mean_token_accuracy": 0.7101903825998306,
"num_tokens": 613181600.0,
"step": 4710
},
{
"entropy": 1.95,
"epoch": 0.24537325847369515,
"grad_norm": 2.9153966228033066,
"learning_rate": 7.780576696323293e-06,
"loss": 1.2867,
"mean_token_accuracy": 0.6935893803834915,
"num_tokens": 614484864.0,
"step": 4720
},
{
"entropy": 1.921875,
"epoch": 0.2458931170721564,
"grad_norm": 2.911701980008095,
"learning_rate": 7.775217065065925e-06,
"loss": 1.2297,
"mean_token_accuracy": 0.7043861538171768,
"num_tokens": 615787631.0,
"step": 4730
},
{
"entropy": 1.953515625,
"epoch": 0.2464129756706176,
"grad_norm": 2.7963284598386338,
"learning_rate": 7.769857433808555e-06,
"loss": 1.2518,
"mean_token_accuracy": 0.7025377005338669,
"num_tokens": 617089202.0,
"step": 4740
},
{
"entropy": 1.96875,
"epoch": 0.2469328342690788,
"grad_norm": 2.7694482417090684,
"learning_rate": 7.764497802551185e-06,
"loss": 1.2206,
"mean_token_accuracy": 0.7073077768087387,
"num_tokens": 618391693.0,
"step": 4750
},
{
"entropy": 1.965625,
"epoch": 0.24745269286754004,
"grad_norm": 2.4505868233426242,
"learning_rate": 7.759138171293815e-06,
"loss": 1.2544,
"mean_token_accuracy": 0.6981542706489563,
"num_tokens": 619692216.0,
"step": 4760
},
{
"entropy": 1.94140625,
"epoch": 0.24797255146600125,
"grad_norm": 2.5687393911466145,
"learning_rate": 7.753778540036447e-06,
"loss": 1.238,
"mean_token_accuracy": 0.7030215322971344,
"num_tokens": 620995529.0,
"step": 4770
},
{
"entropy": 1.954296875,
"epoch": 0.24849241006446246,
"grad_norm": 2.6776346755134397,
"learning_rate": 7.748418908779077e-06,
"loss": 1.2011,
"mean_token_accuracy": 0.7112019300460816,
"num_tokens": 622299689.0,
"step": 4780
},
{
"entropy": 1.979296875,
"epoch": 0.2490122686629237,
"grad_norm": 2.3820367592869265,
"learning_rate": 7.743059277521707e-06,
"loss": 1.2577,
"mean_token_accuracy": 0.697287705540657,
"num_tokens": 623600265.0,
"step": 4790
},
{
"entropy": 1.955859375,
"epoch": 0.2495321272613849,
"grad_norm": 2.729346977356678,
"learning_rate": 7.737699646264337e-06,
"loss": 1.2853,
"mean_token_accuracy": 0.6956493765115738,
"num_tokens": 624901504.0,
"step": 4800
},
{
"epoch": 0.2495321272613849,
"eval_entropy": 1.92890625,
"eval_loss": 1.263953447341919,
"eval_mean_token_accuracy": 0.6961007863283157,
"eval_num_tokens": 624901504.0,
"eval_runtime": 27.0147,
"eval_samples_per_second": 93.801,
"eval_steps_per_second": 1.481,
"step": 4800
},
{
"entropy": 1.971484375,
"epoch": 0.2500519858598461,
"grad_norm": 2.441479557851593,
"learning_rate": 7.732340015006969e-06,
"loss": 1.2356,
"mean_token_accuracy": 0.7034515768289566,
"num_tokens": 626201333.0,
"step": 4810
},
{
"entropy": 1.917578125,
"epoch": 0.2505718444583073,
"grad_norm": 2.6088645411292566,
"learning_rate": 7.726980383749597e-06,
"loss": 1.1708,
"mean_token_accuracy": 0.7162606179714203,
"num_tokens": 627504821.0,
"step": 4820
},
{
"entropy": 1.9234375,
"epoch": 0.25109170305676853,
"grad_norm": 2.7467832552504117,
"learning_rate": 7.72162075249223e-06,
"loss": 1.2108,
"mean_token_accuracy": 0.7109028309583664,
"num_tokens": 628806938.0,
"step": 4830
},
{
"entropy": 1.934375,
"epoch": 0.2516115616552298,
"grad_norm": 2.4764321790173227,
"learning_rate": 7.71626112123486e-06,
"loss": 1.1775,
"mean_token_accuracy": 0.7154791980981827,
"num_tokens": 630108540.0,
"step": 4840
},
{
"entropy": 1.953515625,
"epoch": 0.252131420253691,
"grad_norm": 2.6648686297452104,
"learning_rate": 7.71090148997749e-06,
"loss": 1.2916,
"mean_token_accuracy": 0.6932720243930817,
"num_tokens": 631410744.0,
"step": 4850
},
{
"entropy": 1.9546875,
"epoch": 0.2526512788521522,
"grad_norm": 3.1307085065084155,
"learning_rate": 7.70554185872012e-06,
"loss": 1.2779,
"mean_token_accuracy": 0.693497759103775,
"num_tokens": 632711937.0,
"step": 4860
},
{
"entropy": 1.953515625,
"epoch": 0.2531711374506134,
"grad_norm": 2.703892075800755,
"learning_rate": 7.700182227462751e-06,
"loss": 1.2922,
"mean_token_accuracy": 0.6932292968034744,
"num_tokens": 634013852.0,
"step": 4870
},
{
"entropy": 1.937890625,
"epoch": 0.25369099604907464,
"grad_norm": 2.554634700626667,
"learning_rate": 7.694822596205382e-06,
"loss": 1.1974,
"mean_token_accuracy": 0.7117091089487075,
"num_tokens": 635315692.0,
"step": 4880
},
{
"entropy": 1.944921875,
"epoch": 0.25421085464753584,
"grad_norm": 3.1334748392378384,
"learning_rate": 7.689462964948012e-06,
"loss": 1.2728,
"mean_token_accuracy": 0.6964755386114121,
"num_tokens": 636616125.0,
"step": 4890
},
{
"entropy": 1.955078125,
"epoch": 0.2547307132459971,
"grad_norm": 2.3176982738828853,
"learning_rate": 7.684103333690642e-06,
"loss": 1.2586,
"mean_token_accuracy": 0.6991387635469437,
"num_tokens": 637917436.0,
"step": 4900
},
{
"epoch": 0.2547307132459971,
"eval_entropy": 1.94453125,
"eval_loss": 1.2611573934555054,
"eval_mean_token_accuracy": 0.696676367521286,
"eval_num_tokens": 637917436.0,
"eval_runtime": 27.0436,
"eval_samples_per_second": 93.701,
"eval_steps_per_second": 1.479,
"step": 4900
},
{
"entropy": 1.9546875,
"epoch": 0.2552505718444583,
"grad_norm": 2.659233051515967,
"learning_rate": 7.678743702433274e-06,
"loss": 1.217,
"mean_token_accuracy": 0.7063292741775513,
"num_tokens": 639217421.0,
"step": 4910
},
{
"entropy": 1.938671875,
"epoch": 0.2557704304429195,
"grad_norm": 3.0453077797197583,
"learning_rate": 7.673384071175904e-06,
"loss": 1.2405,
"mean_token_accuracy": 0.7005271553993225,
"num_tokens": 640521049.0,
"step": 4920
},
{
"entropy": 1.971484375,
"epoch": 0.25629028904138074,
"grad_norm": 2.6361358273140736,
"learning_rate": 7.668024439918534e-06,
"loss": 1.2518,
"mean_token_accuracy": 0.7005126535892486,
"num_tokens": 641822027.0,
"step": 4930
},
{
"entropy": 1.94140625,
"epoch": 0.25681014763984195,
"grad_norm": 2.615960216357449,
"learning_rate": 7.662664808661164e-06,
"loss": 1.1607,
"mean_token_accuracy": 0.7182939171791076,
"num_tokens": 643126386.0,
"step": 4940
},
{
"entropy": 1.91484375,
"epoch": 0.25733000623830316,
"grad_norm": 3.0572120674280723,
"learning_rate": 7.657305177403796e-06,
"loss": 1.2201,
"mean_token_accuracy": 0.7099931478500366,
"num_tokens": 644428334.0,
"step": 4950
},
{
"entropy": 1.92109375,
"epoch": 0.2578498648367644,
"grad_norm": 2.940475985908865,
"learning_rate": 7.651945546146426e-06,
"loss": 1.228,
"mean_token_accuracy": 0.7041484266519547,
"num_tokens": 645728962.0,
"step": 4960
},
{
"entropy": 1.91640625,
"epoch": 0.25836972343522563,
"grad_norm": 2.692853034600281,
"learning_rate": 7.646585914889056e-06,
"loss": 1.2815,
"mean_token_accuracy": 0.6926003754138946,
"num_tokens": 647030970.0,
"step": 4970
},
{
"entropy": 1.92578125,
"epoch": 0.25888958203368684,
"grad_norm": 2.741276751751905,
"learning_rate": 7.641226283631686e-06,
"loss": 1.2139,
"mean_token_accuracy": 0.7093022853136063,
"num_tokens": 648331923.0,
"step": 4980
},
{
"entropy": 1.957421875,
"epoch": 0.25940944063214805,
"grad_norm": 2.6883898510435618,
"learning_rate": 7.635866652374318e-06,
"loss": 1.272,
"mean_token_accuracy": 0.6946966707706451,
"num_tokens": 649634172.0,
"step": 4990
},
{
"entropy": 1.9421875,
"epoch": 0.25992929923060926,
"grad_norm": 2.7248434434916144,
"learning_rate": 7.630507021116948e-06,
"loss": 1.2086,
"mean_token_accuracy": 0.7105143129825592,
"num_tokens": 650933969.0,
"step": 5000
},
{
"epoch": 0.25992929923060926,
"eval_entropy": 1.9123046875,
"eval_loss": 1.2598856687545776,
"eval_mean_token_accuracy": 0.6974375113844872,
"eval_num_tokens": 650933969.0,
"eval_runtime": 26.9823,
"eval_samples_per_second": 93.913,
"eval_steps_per_second": 1.482,
"step": 5000
},
{
"entropy": 1.9484375,
"epoch": 0.26044915782907047,
"grad_norm": 2.416606017613257,
"learning_rate": 7.625147389859578e-06,
"loss": 1.2183,
"mean_token_accuracy": 0.7071460753679275,
"num_tokens": 652236167.0,
"step": 5010
},
{
"entropy": 1.948046875,
"epoch": 0.26096901642753173,
"grad_norm": 2.5151498558038767,
"learning_rate": 7.619787758602208e-06,
"loss": 1.2299,
"mean_token_accuracy": 0.7075282990932464,
"num_tokens": 653536853.0,
"step": 5020
},
{
"entropy": 1.985546875,
"epoch": 0.26148887502599294,
"grad_norm": 2.8357485126352424,
"learning_rate": 7.614428127344839e-06,
"loss": 1.2431,
"mean_token_accuracy": 0.7025340616703033,
"num_tokens": 654840099.0,
"step": 5030
},
{
"entropy": 1.946875,
"epoch": 0.26200873362445415,
"grad_norm": 2.747044244398687,
"learning_rate": 7.6090684960874695e-06,
"loss": 1.1984,
"mean_token_accuracy": 0.7127739608287811,
"num_tokens": 656141458.0,
"step": 5040
},
{
"entropy": 1.95390625,
"epoch": 0.26252859222291536,
"grad_norm": 3.1906310283387547,
"learning_rate": 7.6037088648301004e-06,
"loss": 1.2314,
"mean_token_accuracy": 0.707214081287384,
"num_tokens": 657444399.0,
"step": 5050
},
{
"entropy": 2.006640625,
"epoch": 0.26304845082137657,
"grad_norm": 2.8080095740282394,
"learning_rate": 7.5983492335727306e-06,
"loss": 1.2382,
"mean_token_accuracy": 0.7069886833429336,
"num_tokens": 658745627.0,
"step": 5060
},
{
"entropy": 1.924609375,
"epoch": 0.26356830941983783,
"grad_norm": 2.5262985349840883,
"learning_rate": 7.5929896023153615e-06,
"loss": 1.2587,
"mean_token_accuracy": 0.6995914459228516,
"num_tokens": 660047288.0,
"step": 5070
},
{
"entropy": 1.92265625,
"epoch": 0.26408816801829904,
"grad_norm": 2.8010083945012085,
"learning_rate": 7.587629971057992e-06,
"loss": 1.1933,
"mean_token_accuracy": 0.7122175902128219,
"num_tokens": 661348718.0,
"step": 5080
},
{
"entropy": 1.913671875,
"epoch": 0.26460802661676025,
"grad_norm": 2.689126726942353,
"learning_rate": 7.582270339800623e-06,
"loss": 1.208,
"mean_token_accuracy": 0.7095348536968231,
"num_tokens": 662651355.0,
"step": 5090
},
{
"entropy": 1.98046875,
"epoch": 0.26512788521522146,
"grad_norm": 2.578696315845084,
"learning_rate": 7.576910708543253e-06,
"loss": 1.2306,
"mean_token_accuracy": 0.7059185087680817,
"num_tokens": 663953044.0,
"step": 5100
},
{
"epoch": 0.26512788521522146,
"eval_entropy": 1.9509765625,
"eval_loss": 1.2543494701385498,
"eval_mean_token_accuracy": 0.6989290028810501,
"eval_num_tokens": 663953044.0,
"eval_runtime": 27.0024,
"eval_samples_per_second": 93.843,
"eval_steps_per_second": 1.481,
"step": 5100
},
{
"entropy": 1.977734375,
"epoch": 0.26564774381368267,
"grad_norm": 2.876734245620972,
"learning_rate": 7.571551077285884e-06,
"loss": 1.2509,
"mean_token_accuracy": 0.7007964789867401,
"num_tokens": 665253351.0,
"step": 5110
},
{
"entropy": 1.963671875,
"epoch": 0.2661676024121439,
"grad_norm": 2.821649371135593,
"learning_rate": 7.566191446028514e-06,
"loss": 1.2077,
"mean_token_accuracy": 0.7099482923746109,
"num_tokens": 666555147.0,
"step": 5120
},
{
"entropy": 1.965234375,
"epoch": 0.26668746101060514,
"grad_norm": 2.6804982182820143,
"learning_rate": 7.560831814771145e-06,
"loss": 1.21,
"mean_token_accuracy": 0.7121542394161224,
"num_tokens": 667857330.0,
"step": 5130
},
{
"entropy": 1.96171875,
"epoch": 0.26720731960906635,
"grad_norm": 3.2332078225653667,
"learning_rate": 7.555472183513775e-06,
"loss": 1.1995,
"mean_token_accuracy": 0.7113998055458068,
"num_tokens": 669160376.0,
"step": 5140
},
{
"entropy": 1.969140625,
"epoch": 0.26772717820752756,
"grad_norm": 2.5990006942741224,
"learning_rate": 7.550112552256406e-06,
"loss": 1.27,
"mean_token_accuracy": 0.6997297942638397,
"num_tokens": 670461127.0,
"step": 5150
},
{
"entropy": 1.9734375,
"epoch": 0.26824703680598877,
"grad_norm": 3.0065160478644652,
"learning_rate": 7.544752920999035e-06,
"loss": 1.2043,
"mean_token_accuracy": 0.7129088431596756,
"num_tokens": 671761598.0,
"step": 5160
},
{
"entropy": 1.957421875,
"epoch": 0.26876689540445,
"grad_norm": 2.3455366809362173,
"learning_rate": 7.539393289741667e-06,
"loss": 1.2094,
"mean_token_accuracy": 0.7085828363895417,
"num_tokens": 673065355.0,
"step": 5170
},
{
"entropy": 1.96015625,
"epoch": 0.2692867540029112,
"grad_norm": 2.8812315837238316,
"learning_rate": 7.534033658484296e-06,
"loss": 1.2166,
"mean_token_accuracy": 0.7104844450950623,
"num_tokens": 674365317.0,
"step": 5180
},
{
"entropy": 1.978515625,
"epoch": 0.26980661260137245,
"grad_norm": 2.8474115733640506,
"learning_rate": 7.528674027226927e-06,
"loss": 1.2409,
"mean_token_accuracy": 0.7038198083639144,
"num_tokens": 675668003.0,
"step": 5190
},
{
"entropy": 1.955078125,
"epoch": 0.27032647119983366,
"grad_norm": 2.439489940556994,
"learning_rate": 7.523314395969557e-06,
"loss": 1.2346,
"mean_token_accuracy": 0.7021261721849441,
"num_tokens": 676971440.0,
"step": 5200
},
{
"epoch": 0.27032647119983366,
"eval_entropy": 1.9626953125,
"eval_loss": 1.2539491653442383,
"eval_mean_token_accuracy": 0.6983672767877579,
"eval_num_tokens": 676971440.0,
"eval_runtime": 27.0084,
"eval_samples_per_second": 93.823,
"eval_steps_per_second": 1.481,
"step": 5200
},
{
"entropy": 1.99609375,
"epoch": 0.27084632979829487,
"grad_norm": 2.687407631245056,
"learning_rate": 7.517954764712188e-06,
"loss": 1.2515,
"mean_token_accuracy": 0.7003001749515534,
"num_tokens": 678272114.0,
"step": 5210
},
{
"entropy": 1.9921875,
"epoch": 0.2713661883967561,
"grad_norm": 2.631541255470272,
"learning_rate": 7.5125951334548185e-06,
"loss": 1.2608,
"mean_token_accuracy": 0.6989294230937958,
"num_tokens": 679573426.0,
"step": 5220
},
{
"entropy": 1.963671875,
"epoch": 0.2718860469952173,
"grad_norm": 2.512259997804729,
"learning_rate": 7.5072355021974494e-06,
"loss": 1.2627,
"mean_token_accuracy": 0.6988277584314346,
"num_tokens": 680876184.0,
"step": 5230
},
{
"entropy": 1.95234375,
"epoch": 0.2724059055936785,
"grad_norm": 2.5250982549228915,
"learning_rate": 7.5018758709400796e-06,
"loss": 1.2317,
"mean_token_accuracy": 0.7018355280160904,
"num_tokens": 682177066.0,
"step": 5240
},
{
"entropy": 1.991015625,
"epoch": 0.27292576419213976,
"grad_norm": 2.9056371374889185,
"learning_rate": 7.4965162396827105e-06,
"loss": 1.2464,
"mean_token_accuracy": 0.7020098388195037,
"num_tokens": 683475436.0,
"step": 5250
},
{
"entropy": 1.9578125,
"epoch": 0.273445622790601,
"grad_norm": 2.695060392521751,
"learning_rate": 7.491156608425341e-06,
"loss": 1.227,
"mean_token_accuracy": 0.7065005242824555,
"num_tokens": 684776901.0,
"step": 5260
},
{
"entropy": 1.97734375,
"epoch": 0.2739654813890622,
"grad_norm": 2.8328596886005517,
"learning_rate": 7.485796977167972e-06,
"loss": 1.2335,
"mean_token_accuracy": 0.7066089063882828,
"num_tokens": 686074434.0,
"step": 5270
},
{
"entropy": 2.012890625,
"epoch": 0.2744853399875234,
"grad_norm": 2.990207155628477,
"learning_rate": 7.480437345910602e-06,
"loss": 1.2901,
"mean_token_accuracy": 0.6942520767450333,
"num_tokens": 687378742.0,
"step": 5280
},
{
"entropy": 1.959765625,
"epoch": 0.2750051985859846,
"grad_norm": 2.701827746100283,
"learning_rate": 7.475077714653233e-06,
"loss": 1.1579,
"mean_token_accuracy": 0.7188471108675003,
"num_tokens": 688680947.0,
"step": 5290
},
{
"entropy": 1.95859375,
"epoch": 0.2755250571844458,
"grad_norm": 2.892736229955728,
"learning_rate": 7.469718083395863e-06,
"loss": 1.2476,
"mean_token_accuracy": 0.7039466559886932,
"num_tokens": 689981525.0,
"step": 5300
},
{
"epoch": 0.2755250571844458,
"eval_entropy": 1.9962890625,
"eval_loss": 1.2514336109161377,
"eval_mean_token_accuracy": 0.6991482481360436,
"eval_num_tokens": 689981525.0,
"eval_runtime": 26.9854,
"eval_samples_per_second": 93.902,
"eval_steps_per_second": 1.482,
"step": 5300
},
{
"entropy": 1.95,
"epoch": 0.2760449157829071,
"grad_norm": 2.8279829585711274,
"learning_rate": 7.464358452138494e-06,
"loss": 1.1922,
"mean_token_accuracy": 0.7123861104249954,
"num_tokens": 691284424.0,
"step": 5310
},
{
"entropy": 1.972265625,
"epoch": 0.2765647743813683,
"grad_norm": 2.619503957702591,
"learning_rate": 7.458998820881124e-06,
"loss": 1.1994,
"mean_token_accuracy": 0.7118939399719239,
"num_tokens": 692587099.0,
"step": 5320
},
{
"entropy": 1.982421875,
"epoch": 0.2770846329798295,
"grad_norm": 2.667220353000756,
"learning_rate": 7.453639189623755e-06,
"loss": 1.2235,
"mean_token_accuracy": 0.7055204808712006,
"num_tokens": 693890122.0,
"step": 5330
},
{
"entropy": 1.951953125,
"epoch": 0.2776044915782907,
"grad_norm": 2.7460628318176616,
"learning_rate": 7.448279558366385e-06,
"loss": 1.2486,
"mean_token_accuracy": 0.7022708028554916,
"num_tokens": 695190033.0,
"step": 5340
},
{
"entropy": 1.983203125,
"epoch": 0.2781243501767519,
"grad_norm": 2.4794967917592396,
"learning_rate": 7.442919927109016e-06,
"loss": 1.2117,
"mean_token_accuracy": 0.7078495651483536,
"num_tokens": 696492208.0,
"step": 5350
},
{
"entropy": 1.95546875,
"epoch": 0.2786442087752131,
"grad_norm": 2.943898706007821,
"learning_rate": 7.437560295851646e-06,
"loss": 1.2235,
"mean_token_accuracy": 0.7081861704587936,
"num_tokens": 697793707.0,
"step": 5360
},
{
"entropy": 1.94453125,
"epoch": 0.2791640673736744,
"grad_norm": 2.6026968777551738,
"learning_rate": 7.432200664594277e-06,
"loss": 1.1997,
"mean_token_accuracy": 0.7086179912090301,
"num_tokens": 699095505.0,
"step": 5370
},
{
"entropy": 1.94140625,
"epoch": 0.2796839259721356,
"grad_norm": 3.0881571351717727,
"learning_rate": 7.426841033336906e-06,
"loss": 1.1961,
"mean_token_accuracy": 0.712966400384903,
"num_tokens": 700397791.0,
"step": 5380
},
{
"entropy": 1.9734375,
"epoch": 0.2802037845705968,
"grad_norm": 2.976183344082867,
"learning_rate": 7.421481402079538e-06,
"loss": 1.2296,
"mean_token_accuracy": 0.7033983498811722,
"num_tokens": 701699500.0,
"step": 5390
},
{
"entropy": 1.9984375,
"epoch": 0.280723643169058,
"grad_norm": 2.56089398801131,
"learning_rate": 7.4161217708221675e-06,
"loss": 1.2536,
"mean_token_accuracy": 0.7002466976642608,
"num_tokens": 703001384.0,
"step": 5400
},
{
"epoch": 0.280723643169058,
"eval_entropy": 1.979296875,
"eval_loss": 1.2450758218765259,
"eval_mean_token_accuracy": 0.7007353320717812,
"eval_num_tokens": 703001384.0,
"eval_runtime": 26.992,
"eval_samples_per_second": 93.88,
"eval_steps_per_second": 1.482,
"step": 5400
},
{
"entropy": 1.98515625,
"epoch": 0.2812435017675192,
"grad_norm": 2.5272425463662698,
"learning_rate": 7.4107621395647984e-06,
"loss": 1.2359,
"mean_token_accuracy": 0.704455754160881,
"num_tokens": 704302786.0,
"step": 5410
},
{
"entropy": 1.974609375,
"epoch": 0.28176336036598043,
"grad_norm": 2.718031542927409,
"learning_rate": 7.4054025083074286e-06,
"loss": 1.251,
"mean_token_accuracy": 0.7005485534667969,
"num_tokens": 705606198.0,
"step": 5420
},
{
"entropy": 1.955859375,
"epoch": 0.2822832189644417,
"grad_norm": 2.868456651558638,
"learning_rate": 7.4000428770500595e-06,
"loss": 1.2203,
"mean_token_accuracy": 0.7065345704555511,
"num_tokens": 706907340.0,
"step": 5430
},
{
"entropy": 1.97109375,
"epoch": 0.2828030775629029,
"grad_norm": 3.5612536867129716,
"learning_rate": 7.39468324579269e-06,
"loss": 1.2465,
"mean_token_accuracy": 0.697993329167366,
"num_tokens": 708209631.0,
"step": 5440
},
{
"entropy": 1.962890625,
"epoch": 0.2833229361613641,
"grad_norm": 2.9724414889429496,
"learning_rate": 7.389323614535321e-06,
"loss": 1.2394,
"mean_token_accuracy": 0.7033718883991241,
"num_tokens": 709510884.0,
"step": 5450
},
{
"entropy": 1.987890625,
"epoch": 0.2838427947598253,
"grad_norm": 2.6042989737540196,
"learning_rate": 7.383963983277951e-06,
"loss": 1.2174,
"mean_token_accuracy": 0.7064710199832916,
"num_tokens": 710812280.0,
"step": 5460
},
{
"entropy": 1.95859375,
"epoch": 0.28436265335828653,
"grad_norm": 2.834259800528809,
"learning_rate": 7.378604352020582e-06,
"loss": 1.2289,
"mean_token_accuracy": 0.7042302817106247,
"num_tokens": 712113146.0,
"step": 5470
},
{
"entropy": 1.941015625,
"epoch": 0.28488251195674774,
"grad_norm": 2.500747200653327,
"learning_rate": 7.373244720763212e-06,
"loss": 1.2343,
"mean_token_accuracy": 0.7041598945856095,
"num_tokens": 713415335.0,
"step": 5480
},
{
"entropy": 2.00703125,
"epoch": 0.285402370555209,
"grad_norm": 2.3217741425486875,
"learning_rate": 7.367885089505843e-06,
"loss": 1.2292,
"mean_token_accuracy": 0.7065331846475601,
"num_tokens": 714716281.0,
"step": 5490
},
{
"entropy": 1.94765625,
"epoch": 0.2859222291536702,
"grad_norm": 2.955004563822571,
"learning_rate": 7.362525458248473e-06,
"loss": 1.2435,
"mean_token_accuracy": 0.704520583152771,
"num_tokens": 716019366.0,
"step": 5500
},
{
"epoch": 0.2859222291536702,
"eval_entropy": 1.90703125,
"eval_loss": 1.2407454252243042,
"eval_mean_token_accuracy": 0.7014473259449006,
"eval_num_tokens": 716019366.0,
"eval_runtime": 27.0343,
"eval_samples_per_second": 93.733,
"eval_steps_per_second": 1.48,
"step": 5500
},
{
"entropy": 1.979296875,
"epoch": 0.2864420877521314,
"grad_norm": 2.7898187028644292,
"learning_rate": 7.357165826991104e-06,
"loss": 1.2585,
"mean_token_accuracy": 0.6986296743154525,
"num_tokens": 717322715.0,
"step": 5510
},
{
"entropy": 1.946484375,
"epoch": 0.28696194635059263,
"grad_norm": 2.6696185517109745,
"learning_rate": 7.351806195733734e-06,
"loss": 1.2704,
"mean_token_accuracy": 0.6985262989997864,
"num_tokens": 718622646.0,
"step": 5520
},
{
"entropy": 1.9421875,
"epoch": 0.28748180494905384,
"grad_norm": 2.280946157307394,
"learning_rate": 7.346446564476365e-06,
"loss": 1.2616,
"mean_token_accuracy": 0.6985371917486191,
"num_tokens": 719923705.0,
"step": 5530
},
{
"entropy": 1.9328125,
"epoch": 0.28800166354751505,
"grad_norm": 2.628551349750335,
"learning_rate": 7.341086933218995e-06,
"loss": 1.2352,
"mean_token_accuracy": 0.7038856953382492,
"num_tokens": 721222960.0,
"step": 5540
},
{
"entropy": 1.979296875,
"epoch": 0.2885215221459763,
"grad_norm": 2.616332823086885,
"learning_rate": 7.335727301961626e-06,
"loss": 1.234,
"mean_token_accuracy": 0.7032134979963303,
"num_tokens": 722523146.0,
"step": 5550
},
{
"entropy": 1.925,
"epoch": 0.2890413807444375,
"grad_norm": 2.6537815931775457,
"learning_rate": 7.330367670704256e-06,
"loss": 1.2425,
"mean_token_accuracy": 0.6997834593057632,
"num_tokens": 723824022.0,
"step": 5560
},
{
"entropy": 1.94609375,
"epoch": 0.28956123934289874,
"grad_norm": 2.858015675005538,
"learning_rate": 7.325008039446887e-06,
"loss": 1.1743,
"mean_token_accuracy": 0.7179679811000824,
"num_tokens": 725127294.0,
"step": 5570
},
{
"entropy": 1.948046875,
"epoch": 0.29008109794135994,
"grad_norm": 2.662750986703342,
"learning_rate": 7.319648408189517e-06,
"loss": 1.2061,
"mean_token_accuracy": 0.7117459863424301,
"num_tokens": 726429434.0,
"step": 5580
},
{
"entropy": 1.99375,
"epoch": 0.29060095653982115,
"grad_norm": 2.4546920844375997,
"learning_rate": 7.314288776932148e-06,
"loss": 1.2585,
"mean_token_accuracy": 0.701487022638321,
"num_tokens": 727732659.0,
"step": 5590
},
{
"entropy": 1.91953125,
"epoch": 0.29112081513828236,
"grad_norm": 2.703076283971951,
"learning_rate": 7.3089291456747775e-06,
"loss": 1.186,
"mean_token_accuracy": 0.7143799424171448,
"num_tokens": 729033912.0,
"step": 5600
},
{
"epoch": 0.29112081513828236,
"eval_entropy": 1.9388671875,
"eval_loss": 1.2393319606781006,
"eval_mean_token_accuracy": 0.7023263826966286,
"eval_num_tokens": 729033912.0,
"eval_runtime": 27.013,
"eval_samples_per_second": 93.807,
"eval_steps_per_second": 1.481,
"step": 5600
},
{
"entropy": 1.966796875,
"epoch": 0.2916406737367436,
"grad_norm": 2.8920521361195397,
"learning_rate": 7.303569514417409e-06,
"loss": 1.2709,
"mean_token_accuracy": 0.6949620723724366,
"num_tokens": 730331039.0,
"step": 5610
},
{
"entropy": 1.9421875,
"epoch": 0.29216053233520484,
"grad_norm": 2.9340418796089844,
"learning_rate": 7.298209883160039e-06,
"loss": 1.2404,
"mean_token_accuracy": 0.7014801740646363,
"num_tokens": 731633271.0,
"step": 5620
},
{
"entropy": 1.9796875,
"epoch": 0.29268039093366605,
"grad_norm": 2.510714761382756,
"learning_rate": 7.2928502519026705e-06,
"loss": 1.1528,
"mean_token_accuracy": 0.7216533690690994,
"num_tokens": 732935824.0,
"step": 5630
},
{
"entropy": 1.940234375,
"epoch": 0.29320024953212726,
"grad_norm": 2.6868705037856597,
"learning_rate": 7.2874906206453e-06,
"loss": 1.1683,
"mean_token_accuracy": 0.7170694202184678,
"num_tokens": 734235568.0,
"step": 5640
},
{
"entropy": 1.94296875,
"epoch": 0.29372010813058846,
"grad_norm": 2.8718912586834544,
"learning_rate": 7.282130989387931e-06,
"loss": 1.243,
"mean_token_accuracy": 0.701151841878891,
"num_tokens": 735539025.0,
"step": 5650
},
{
"entropy": 1.973828125,
"epoch": 0.2942399667290497,
"grad_norm": 2.3714935588266695,
"learning_rate": 7.276771358130561e-06,
"loss": 1.1907,
"mean_token_accuracy": 0.7156265556812287,
"num_tokens": 736841815.0,
"step": 5660
},
{
"entropy": 1.9296875,
"epoch": 0.29475982532751094,
"grad_norm": 2.806036207171397,
"learning_rate": 7.271411726873192e-06,
"loss": 1.2018,
"mean_token_accuracy": 0.7100878655910492,
"num_tokens": 738143672.0,
"step": 5670
},
{
"entropy": 1.92578125,
"epoch": 0.29527968392597215,
"grad_norm": 2.690060247896926,
"learning_rate": 7.266052095615822e-06,
"loss": 1.2065,
"mean_token_accuracy": 0.7097260385751725,
"num_tokens": 739444791.0,
"step": 5680
},
{
"entropy": 1.92109375,
"epoch": 0.29579954252443336,
"grad_norm": 2.8585923334619205,
"learning_rate": 7.260692464358453e-06,
"loss": 1.185,
"mean_token_accuracy": 0.7135227739810943,
"num_tokens": 740748103.0,
"step": 5690
},
{
"entropy": 1.958984375,
"epoch": 0.29631940112289457,
"grad_norm": 2.7970665728540993,
"learning_rate": 7.255332833101083e-06,
"loss": 1.1889,
"mean_token_accuracy": 0.7155257254838944,
"num_tokens": 742047768.0,
"step": 5700
},
{
"epoch": 0.29631940112289457,
"eval_entropy": 1.9203125,
"eval_loss": 1.2334680557250977,
"eval_mean_token_accuracy": 0.7034152701497078,
"eval_num_tokens": 742047768.0,
"eval_runtime": 27.0486,
"eval_samples_per_second": 93.683,
"eval_steps_per_second": 1.479,
"step": 5700
},
{
"entropy": 1.938671875,
"epoch": 0.2968392597213558,
"grad_norm": 2.8142297194301125,
"learning_rate": 7.249973201843714e-06,
"loss": 1.2063,
"mean_token_accuracy": 0.7137587189674377,
"num_tokens": 743349586.0,
"step": 5710
},
{
"entropy": 1.944140625,
"epoch": 0.297359118319817,
"grad_norm": 2.8493328771563426,
"learning_rate": 7.244613570586344e-06,
"loss": 1.2048,
"mean_token_accuracy": 0.7134816825389863,
"num_tokens": 744647253.0,
"step": 5720
},
{
"entropy": 1.921875,
"epoch": 0.29787897691827825,
"grad_norm": 3.008008991130602,
"learning_rate": 7.239253939328975e-06,
"loss": 1.2255,
"mean_token_accuracy": 0.7058566868305206,
"num_tokens": 745948165.0,
"step": 5730
},
{
"entropy": 1.91796875,
"epoch": 0.29839883551673946,
"grad_norm": 2.8979156391887697,
"learning_rate": 7.233894308071605e-06,
"loss": 1.1776,
"mean_token_accuracy": 0.7174970537424088,
"num_tokens": 747251696.0,
"step": 5740
},
{
"entropy": 1.932421875,
"epoch": 0.29891869411520067,
"grad_norm": 2.8239340383729625,
"learning_rate": 7.228534676814236e-06,
"loss": 1.2298,
"mean_token_accuracy": 0.7044425338506699,
"num_tokens": 748553625.0,
"step": 5750
},
{
"entropy": 1.916015625,
"epoch": 0.2994385527136619,
"grad_norm": 2.6370636785054815,
"learning_rate": 7.223175045556866e-06,
"loss": 1.2075,
"mean_token_accuracy": 0.7122007071971893,
"num_tokens": 749856650.0,
"step": 5760
},
{
"entropy": 1.9625,
"epoch": 0.2999584113121231,
"grad_norm": 2.7210008299307584,
"learning_rate": 7.217815414299497e-06,
"loss": 1.1966,
"mean_token_accuracy": 0.7159106642007828,
"num_tokens": 751159771.0,
"step": 5770
},
{
"entropy": 1.958203125,
"epoch": 0.3004782699105843,
"grad_norm": 2.2879938362694707,
"learning_rate": 7.212455783042127e-06,
"loss": 1.1966,
"mean_token_accuracy": 0.716105169057846,
"num_tokens": 752461520.0,
"step": 5780
},
{
"entropy": 1.9453125,
"epoch": 0.30099812850904556,
"grad_norm": 2.85304969639525,
"learning_rate": 7.207096151784758e-06,
"loss": 1.1694,
"mean_token_accuracy": 0.7198493242263794,
"num_tokens": 753762459.0,
"step": 5790
},
{
"entropy": 1.94765625,
"epoch": 0.30151798710750677,
"grad_norm": 2.591943536802794,
"learning_rate": 7.2017365205273885e-06,
"loss": 1.2079,
"mean_token_accuracy": 0.709328094124794,
"num_tokens": 755062065.0,
"step": 5800
},
{
"epoch": 0.30151798710750677,
"eval_entropy": 1.9150390625,
"eval_loss": 1.2302395105361938,
"eval_mean_token_accuracy": 0.7037526503205299,
"eval_num_tokens": 755062065.0,
"eval_runtime": 26.9781,
"eval_samples_per_second": 93.928,
"eval_steps_per_second": 1.483,
"step": 5800
},
{
"entropy": 1.913671875,
"epoch": 0.302037845705968,
"grad_norm": 2.712499601729641,
"learning_rate": 7.1963768892700194e-06,
"loss": 1.1681,
"mean_token_accuracy": 0.7172327995300293,
"num_tokens": 756365499.0,
"step": 5810
},
{
"entropy": 1.91875,
"epoch": 0.3025577043044292,
"grad_norm": 2.442301799557076,
"learning_rate": 7.1910172580126496e-06,
"loss": 1.1419,
"mean_token_accuracy": 0.7224777072668076,
"num_tokens": 757666496.0,
"step": 5820
},
{
"entropy": 1.94609375,
"epoch": 0.3030775629028904,
"grad_norm": 3.142181441761507,
"learning_rate": 7.1856576267552805e-06,
"loss": 1.2279,
"mean_token_accuracy": 0.7056004136800766,
"num_tokens": 758969431.0,
"step": 5830
},
{
"entropy": 1.944140625,
"epoch": 0.3035974215013516,
"grad_norm": 3.14692615239008,
"learning_rate": 7.18029799549791e-06,
"loss": 1.1822,
"mean_token_accuracy": 0.7149683803319931,
"num_tokens": 760271012.0,
"step": 5840
},
{
"entropy": 1.946484375,
"epoch": 0.30411728009981287,
"grad_norm": 2.978009623324451,
"learning_rate": 7.174938364240542e-06,
"loss": 1.2225,
"mean_token_accuracy": 0.7082667648792267,
"num_tokens": 761574176.0,
"step": 5850
},
{
"entropy": 1.921875,
"epoch": 0.3046371386982741,
"grad_norm": 3.257659737027142,
"learning_rate": 7.169578732983171e-06,
"loss": 1.1507,
"mean_token_accuracy": 0.7247169196605683,
"num_tokens": 762875523.0,
"step": 5860
},
{
"entropy": 1.9734375,
"epoch": 0.3051569972967353,
"grad_norm": 2.7525531045796874,
"learning_rate": 7.164219101725802e-06,
"loss": 1.1805,
"mean_token_accuracy": 0.7174414724111557,
"num_tokens": 764176952.0,
"step": 5870
},
{
"entropy": 1.94609375,
"epoch": 0.3056768558951965,
"grad_norm": 2.7107855479228973,
"learning_rate": 7.158859470468432e-06,
"loss": 1.2246,
"mean_token_accuracy": 0.7051614582538605,
"num_tokens": 765480060.0,
"step": 5880
},
{
"entropy": 1.959765625,
"epoch": 0.3061967144936577,
"grad_norm": 2.670492972252768,
"learning_rate": 7.153499839211063e-06,
"loss": 1.2309,
"mean_token_accuracy": 0.7050163090229035,
"num_tokens": 766782181.0,
"step": 5890
},
{
"entropy": 1.94921875,
"epoch": 0.3067165730921189,
"grad_norm": 2.744962393459733,
"learning_rate": 7.148140207953693e-06,
"loss": 1.2095,
"mean_token_accuracy": 0.7110066026449203,
"num_tokens": 768083448.0,
"step": 5900
},
{
"epoch": 0.3067165730921189,
"eval_entropy": 1.9322265625,
"eval_loss": 1.2244899272918701,
"eval_mean_token_accuracy": 0.7049604192376137,
"eval_num_tokens": 768083448.0,
"eval_runtime": 26.9488,
"eval_samples_per_second": 94.03,
"eval_steps_per_second": 1.484,
"step": 5900
},
{
"entropy": 1.9484375,
"epoch": 0.3072364316905802,
"grad_norm": 2.3462907707509966,
"learning_rate": 7.142780576696324e-06,
"loss": 1.2192,
"mean_token_accuracy": 0.7091288506984711,
"num_tokens": 769387166.0,
"step": 5910
},
{
"entropy": 1.94453125,
"epoch": 0.3077562902890414,
"grad_norm": 2.742541311601078,
"learning_rate": 7.137420945438954e-06,
"loss": 1.1785,
"mean_token_accuracy": 0.716044420003891,
"num_tokens": 770688983.0,
"step": 5920
},
{
"entropy": 1.953515625,
"epoch": 0.3082761488875026,
"grad_norm": 2.925236111803703,
"learning_rate": 7.132061314181585e-06,
"loss": 1.218,
"mean_token_accuracy": 0.7094991773366928,
"num_tokens": 771992704.0,
"step": 5930
},
{
"entropy": 1.98046875,
"epoch": 0.3087960074859638,
"grad_norm": 2.560140368591148,
"learning_rate": 7.126701682924215e-06,
"loss": 1.2093,
"mean_token_accuracy": 0.7107134282588958,
"num_tokens": 773294777.0,
"step": 5940
},
{
"entropy": 1.9546875,
"epoch": 0.309315866084425,
"grad_norm": 2.7099295862166692,
"learning_rate": 7.121342051666846e-06,
"loss": 1.1573,
"mean_token_accuracy": 0.7208248853683472,
"num_tokens": 774596868.0,
"step": 5950
},
{
"entropy": 1.908203125,
"epoch": 0.3098357246828863,
"grad_norm": 2.7380772315897013,
"learning_rate": 7.115982420409476e-06,
"loss": 1.1468,
"mean_token_accuracy": 0.7225511878728866,
"num_tokens": 775896662.0,
"step": 5960
},
{
"entropy": 1.94453125,
"epoch": 0.3103555832813475,
"grad_norm": 2.812014274505157,
"learning_rate": 7.110622789152107e-06,
"loss": 1.2159,
"mean_token_accuracy": 0.708139282464981,
"num_tokens": 777201030.0,
"step": 5970
},
{
"entropy": 1.9125,
"epoch": 0.3108754418798087,
"grad_norm": 3.0114917756534654,
"learning_rate": 7.1052631578947375e-06,
"loss": 1.1506,
"mean_token_accuracy": 0.7221070230007172,
"num_tokens": 778501578.0,
"step": 5980
},
{
"entropy": 1.97265625,
"epoch": 0.3113953004782699,
"grad_norm": 2.6938514563537668,
"learning_rate": 7.0999035266373684e-06,
"loss": 1.2198,
"mean_token_accuracy": 0.708455890417099,
"num_tokens": 779804048.0,
"step": 5990
},
{
"entropy": 1.926171875,
"epoch": 0.3119151590767311,
"grad_norm": 2.4481502720826334,
"learning_rate": 7.0945438953799986e-06,
"loss": 1.1594,
"mean_token_accuracy": 0.7198601931333541,
"num_tokens": 781106750.0,
"step": 6000
},
{
"epoch": 0.3119151590767311,
"eval_entropy": 1.9181640625,
"eval_loss": 1.2212631702423096,
"eval_mean_token_accuracy": 0.7061094477772712,
"eval_num_tokens": 781106750.0,
"eval_runtime": 27.0318,
"eval_samples_per_second": 93.741,
"eval_steps_per_second": 1.48,
"step": 6000
},
{
"entropy": 1.941796875,
"epoch": 0.31243501767519233,
"grad_norm": 2.7416825946542507,
"learning_rate": 7.0891842641226295e-06,
"loss": 1.1671,
"mean_token_accuracy": 0.7184388726949692,
"num_tokens": 782409993.0,
"step": 6010
},
{
"entropy": 1.944140625,
"epoch": 0.3129548762736536,
"grad_norm": 2.575273598980468,
"learning_rate": 7.08382463286526e-06,
"loss": 1.2067,
"mean_token_accuracy": 0.7113319844007492,
"num_tokens": 783710696.0,
"step": 6020
},
{
"entropy": 1.94140625,
"epoch": 0.3134747348721148,
"grad_norm": 2.8730659744328255,
"learning_rate": 7.078465001607891e-06,
"loss": 1.2538,
"mean_token_accuracy": 0.7004183262586594,
"num_tokens": 785013611.0,
"step": 6030
},
{
"entropy": 1.951171875,
"epoch": 0.313994593470576,
"grad_norm": 2.6072224724601787,
"learning_rate": 7.073105370350521e-06,
"loss": 1.1666,
"mean_token_accuracy": 0.7210355818271637,
"num_tokens": 786317112.0,
"step": 6040
},
{
"entropy": 1.926171875,
"epoch": 0.3145144520690372,
"grad_norm": 3.338051099285416,
"learning_rate": 7.067745739093152e-06,
"loss": 1.1716,
"mean_token_accuracy": 0.7183846563100815,
"num_tokens": 787620132.0,
"step": 6050
},
{
"entropy": 1.958984375,
"epoch": 0.31503431066749843,
"grad_norm": 2.8202697692662357,
"learning_rate": 7.062386107835781e-06,
"loss": 1.203,
"mean_token_accuracy": 0.7138849049806595,
"num_tokens": 788922585.0,
"step": 6060
},
{
"entropy": 1.929296875,
"epoch": 0.31555416926595964,
"grad_norm": 2.6930688551859068,
"learning_rate": 7.057026476578413e-06,
"loss": 1.1717,
"mean_token_accuracy": 0.7199897319078445,
"num_tokens": 790224217.0,
"step": 6070
},
{
"entropy": 1.9546875,
"epoch": 0.3160740278644209,
"grad_norm": 2.913171880142254,
"learning_rate": 7.051666845321042e-06,
"loss": 1.1897,
"mean_token_accuracy": 0.7115429103374481,
"num_tokens": 791523276.0,
"step": 6080
},
{
"entropy": 1.929296875,
"epoch": 0.3165938864628821,
"grad_norm": 3.0161386941930077,
"learning_rate": 7.046307214063673e-06,
"loss": 1.1935,
"mean_token_accuracy": 0.7147217184305191,
"num_tokens": 792827049.0,
"step": 6090
},
{
"entropy": 1.962890625,
"epoch": 0.3171137450613433,
"grad_norm": 2.7334606951054403,
"learning_rate": 7.040947582806303e-06,
"loss": 1.1751,
"mean_token_accuracy": 0.718577778339386,
"num_tokens": 794129469.0,
"step": 6100
},
{
"epoch": 0.3171137450613433,
"eval_entropy": 1.9142578125,
"eval_loss": 1.2191351652145386,
"eval_mean_token_accuracy": 0.706866504251957,
"eval_num_tokens": 794129469.0,
"eval_runtime": 26.9836,
"eval_samples_per_second": 93.909,
"eval_steps_per_second": 1.482,
"step": 6100
},
{
"entropy": 1.937109375,
"epoch": 0.31763360365980453,
"grad_norm": 2.395471892301908,
"learning_rate": 7.035587951548934e-06,
"loss": 1.1632,
"mean_token_accuracy": 0.7210221737623215,
"num_tokens": 795429912.0,
"step": 6110
},
{
"entropy": 1.942578125,
"epoch": 0.31815346225826574,
"grad_norm": 2.8298670374339605,
"learning_rate": 7.030228320291564e-06,
"loss": 1.1571,
"mean_token_accuracy": 0.7205399483442306,
"num_tokens": 796732576.0,
"step": 6120
},
{
"entropy": 1.96484375,
"epoch": 0.31867332085672695,
"grad_norm": 2.5771561032440866,
"learning_rate": 7.024868689034195e-06,
"loss": 1.1921,
"mean_token_accuracy": 0.7149586290121078,
"num_tokens": 798034208.0,
"step": 6130
},
{
"entropy": 1.958203125,
"epoch": 0.3191931794551882,
"grad_norm": 2.6129221797648774,
"learning_rate": 7.019509057776825e-06,
"loss": 1.2371,
"mean_token_accuracy": 0.7032031327486038,
"num_tokens": 799336493.0,
"step": 6140
},
{
"entropy": 1.963671875,
"epoch": 0.3197130380536494,
"grad_norm": 2.861705815619681,
"learning_rate": 7.014149426519456e-06,
"loss": 1.1802,
"mean_token_accuracy": 0.7148268252611161,
"num_tokens": 800638614.0,
"step": 6150
},
{
"entropy": 1.937109375,
"epoch": 0.32023289665211063,
"grad_norm": 2.818153741149106,
"learning_rate": 7.0087897952620865e-06,
"loss": 1.2165,
"mean_token_accuracy": 0.7065550684928894,
"num_tokens": 801940912.0,
"step": 6160
},
{
"entropy": 1.962890625,
"epoch": 0.32075275525057184,
"grad_norm": 2.76254147923918,
"learning_rate": 7.0034301640047174e-06,
"loss": 1.1982,
"mean_token_accuracy": 0.7154376953840256,
"num_tokens": 803242885.0,
"step": 6170
},
{
"entropy": 1.941015625,
"epoch": 0.32127261384903305,
"grad_norm": 2.752784583594428,
"learning_rate": 6.9980705327473476e-06,
"loss": 1.1973,
"mean_token_accuracy": 0.7128362476825714,
"num_tokens": 804545449.0,
"step": 6180
},
{
"entropy": 1.95703125,
"epoch": 0.32179247244749426,
"grad_norm": 2.675559801146392,
"learning_rate": 6.9927109014899785e-06,
"loss": 1.1833,
"mean_token_accuracy": 0.7163683325052261,
"num_tokens": 805848065.0,
"step": 6190
},
{
"entropy": 1.95625,
"epoch": 0.3223123310459555,
"grad_norm": 2.641351668213455,
"learning_rate": 6.987351270232609e-06,
"loss": 1.1954,
"mean_token_accuracy": 0.7132179349660873,
"num_tokens": 807149892.0,
"step": 6200
},
{
"epoch": 0.3223123310459555,
"eval_entropy": 1.9537109375,
"eval_loss": 1.2142773866653442,
"eval_mean_token_accuracy": 0.7078522220253944,
"eval_num_tokens": 807149892.0,
"eval_runtime": 27.0008,
"eval_samples_per_second": 93.849,
"eval_steps_per_second": 1.481,
"step": 6200
},
{
"entropy": 1.9234375,
"epoch": 0.32283218964441673,
"grad_norm": 2.745046830354981,
"learning_rate": 6.98199163897524e-06,
"loss": 1.1875,
"mean_token_accuracy": 0.7160315811634064,
"num_tokens": 808452586.0,
"step": 6210
},
{
"entropy": 1.99453125,
"epoch": 0.32335204824287794,
"grad_norm": 2.6854458552429223,
"learning_rate": 6.97663200771787e-06,
"loss": 1.1802,
"mean_token_accuracy": 0.7137223988771438,
"num_tokens": 809754125.0,
"step": 6220
},
{
"entropy": 1.944921875,
"epoch": 0.32387190684133915,
"grad_norm": 2.8926314791837413,
"learning_rate": 6.971272376460501e-06,
"loss": 1.2117,
"mean_token_accuracy": 0.7113638281822204,
"num_tokens": 811055888.0,
"step": 6230
},
{
"entropy": 1.91953125,
"epoch": 0.32439176543980036,
"grad_norm": 2.759113972371384,
"learning_rate": 6.965912745203131e-06,
"loss": 1.1671,
"mean_token_accuracy": 0.7212203621864319,
"num_tokens": 812359439.0,
"step": 6240
},
{
"entropy": 1.959765625,
"epoch": 0.32491162403826157,
"grad_norm": 2.6625994150248102,
"learning_rate": 6.960553113945762e-06,
"loss": 1.1511,
"mean_token_accuracy": 0.722861933708191,
"num_tokens": 813662523.0,
"step": 6250
},
{
"entropy": 1.926171875,
"epoch": 0.32543148263672284,
"grad_norm": 2.62312166136983,
"learning_rate": 6.955193482688392e-06,
"loss": 1.1785,
"mean_token_accuracy": 0.7169151067733764,
"num_tokens": 814965009.0,
"step": 6260
},
{
"entropy": 1.95625,
"epoch": 0.32595134123518404,
"grad_norm": 2.557341753831714,
"learning_rate": 6.949833851431023e-06,
"loss": 1.238,
"mean_token_accuracy": 0.7049805134534836,
"num_tokens": 816266629.0,
"step": 6270
},
{
"entropy": 1.94765625,
"epoch": 0.32647119983364525,
"grad_norm": 2.7271373884064736,
"learning_rate": 6.944474220173652e-06,
"loss": 1.1864,
"mean_token_accuracy": 0.7149619698524475,
"num_tokens": 817568560.0,
"step": 6280
},
{
"entropy": 1.967578125,
"epoch": 0.32699105843210646,
"grad_norm": 2.54727797578136,
"learning_rate": 6.939114588916284e-06,
"loss": 1.249,
"mean_token_accuracy": 0.700292906165123,
"num_tokens": 818871044.0,
"step": 6290
},
{
"entropy": 1.9484375,
"epoch": 0.32751091703056767,
"grad_norm": 3.6215754995428635,
"learning_rate": 6.933754957658913e-06,
"loss": 1.2059,
"mean_token_accuracy": 0.7123150020837784,
"num_tokens": 820171524.0,
"step": 6300
},
{
"epoch": 0.32751091703056767,
"eval_entropy": 1.9076171875,
"eval_loss": 1.2089036703109741,
"eval_mean_token_accuracy": 0.708921417593956,
"eval_num_tokens": 820171524.0,
"eval_runtime": 27.0533,
"eval_samples_per_second": 93.667,
"eval_steps_per_second": 1.479,
"step": 6300
},
{
"entropy": 1.991015625,
"epoch": 0.3280307756290289,
"grad_norm": 2.69254967992242,
"learning_rate": 6.928395326401545e-06,
"loss": 1.2314,
"mean_token_accuracy": 0.707514488697052,
"num_tokens": 821471846.0,
"step": 6310
},
{
"entropy": 1.931640625,
"epoch": 0.32855063422749015,
"grad_norm": 2.6926904657111894,
"learning_rate": 6.923035695144174e-06,
"loss": 1.1762,
"mean_token_accuracy": 0.7191505700349807,
"num_tokens": 822775210.0,
"step": 6320
},
{
"entropy": 1.984375,
"epoch": 0.32907049282595136,
"grad_norm": 2.5051966025034074,
"learning_rate": 6.917676063886805e-06,
"loss": 1.1784,
"mean_token_accuracy": 0.7165213465690613,
"num_tokens": 824079836.0,
"step": 6330
},
{
"entropy": 1.902734375,
"epoch": 0.32959035142441256,
"grad_norm": 2.773864859843364,
"learning_rate": 6.9123164326294355e-06,
"loss": 1.1934,
"mean_token_accuracy": 0.7146458566188812,
"num_tokens": 825381869.0,
"step": 6340
},
{
"entropy": 1.9734375,
"epoch": 0.3301102100228738,
"grad_norm": 2.640319875748386,
"learning_rate": 6.906956801372066e-06,
"loss": 1.258,
"mean_token_accuracy": 0.6998931586742401,
"num_tokens": 826683473.0,
"step": 6350
},
{
"entropy": 1.946875,
"epoch": 0.330630068621335,
"grad_norm": 2.748296945185621,
"learning_rate": 6.9015971701146965e-06,
"loss": 1.1625,
"mean_token_accuracy": 0.7236717879772187,
"num_tokens": 827984153.0,
"step": 6360
},
{
"entropy": 1.976953125,
"epoch": 0.3311499272197962,
"grad_norm": 2.5870279649747707,
"learning_rate": 6.8962375388573275e-06,
"loss": 1.1674,
"mean_token_accuracy": 0.7201027631759643,
"num_tokens": 829287475.0,
"step": 6370
},
{
"entropy": 1.971875,
"epoch": 0.33166978581825746,
"grad_norm": 2.7295967283481057,
"learning_rate": 6.890877907599958e-06,
"loss": 1.1602,
"mean_token_accuracy": 0.72014821767807,
"num_tokens": 830589166.0,
"step": 6380
},
{
"entropy": 1.937109375,
"epoch": 0.33218964441671867,
"grad_norm": 2.9087373432015524,
"learning_rate": 6.885518276342589e-06,
"loss": 1.1516,
"mean_token_accuracy": 0.7199266254901886,
"num_tokens": 831890870.0,
"step": 6390
},
{
"entropy": 1.96328125,
"epoch": 0.3327095030151799,
"grad_norm": 2.6241484751811703,
"learning_rate": 6.880158645085219e-06,
"loss": 1.1757,
"mean_token_accuracy": 0.718605899810791,
"num_tokens": 833191900.0,
"step": 6400
},
{
"epoch": 0.3327095030151799,
"eval_entropy": 1.934375,
"eval_loss": 1.2063891887664795,
"eval_mean_token_accuracy": 0.7098991841077804,
"eval_num_tokens": 833191900.0,
"eval_runtime": 26.9814,
"eval_samples_per_second": 93.917,
"eval_steps_per_second": 1.483,
"step": 6400
},
{
"entropy": 1.95625,
"epoch": 0.3332293616136411,
"grad_norm": 2.4964313577584942,
"learning_rate": 6.87479901382785e-06,
"loss": 1.2236,
"mean_token_accuracy": 0.7071185767650604,
"num_tokens": 834493818.0,
"step": 6410
},
{
"entropy": 1.9328125,
"epoch": 0.3337492202121023,
"grad_norm": 2.774012265283333,
"learning_rate": 6.86943938257048e-06,
"loss": 1.1597,
"mean_token_accuracy": 0.7188451290130615,
"num_tokens": 835795200.0,
"step": 6420
},
{
"entropy": 1.89296875,
"epoch": 0.3342690788105635,
"grad_norm": 2.7643889391319183,
"learning_rate": 6.864079751313111e-06,
"loss": 1.1456,
"mean_token_accuracy": 0.7241940170526504,
"num_tokens": 837098349.0,
"step": 6430
},
{
"entropy": 1.941015625,
"epoch": 0.33478893740902477,
"grad_norm": 2.437923941288673,
"learning_rate": 6.858720120055741e-06,
"loss": 1.2215,
"mean_token_accuracy": 0.7108740687370301,
"num_tokens": 838397852.0,
"step": 6440
},
{
"entropy": 1.955078125,
"epoch": 0.335308796007486,
"grad_norm": 3.3186446294014247,
"learning_rate": 6.853360488798371e-06,
"loss": 1.1832,
"mean_token_accuracy": 0.7171518921852111,
"num_tokens": 839698940.0,
"step": 6450
},
{
"entropy": 1.914453125,
"epoch": 0.3358286546059472,
"grad_norm": 2.984795625387781,
"learning_rate": 6.848000857541002e-06,
"loss": 1.1745,
"mean_token_accuracy": 0.7176398009061813,
"num_tokens": 841002062.0,
"step": 6460
},
{
"entropy": 1.89453125,
"epoch": 0.3363485132044084,
"grad_norm": 2.7768653462055504,
"learning_rate": 6.842641226283631e-06,
"loss": 1.1386,
"mean_token_accuracy": 0.7280478745698928,
"num_tokens": 842302988.0,
"step": 6470
},
{
"entropy": 1.939453125,
"epoch": 0.3368683718028696,
"grad_norm": 3.0792346171316796,
"learning_rate": 6.837281595026263e-06,
"loss": 1.1958,
"mean_token_accuracy": 0.7157926440238953,
"num_tokens": 843604468.0,
"step": 6480
},
{
"entropy": 1.9703125,
"epoch": 0.3373882304013308,
"grad_norm": 2.7554741695183758,
"learning_rate": 6.831921963768892e-06,
"loss": 1.2262,
"mean_token_accuracy": 0.706649187207222,
"num_tokens": 844906706.0,
"step": 6490
},
{
"entropy": 1.951171875,
"epoch": 0.3379080889997921,
"grad_norm": 2.511251857958229,
"learning_rate": 6.826562332511524e-06,
"loss": 1.1539,
"mean_token_accuracy": 0.7230695873498917,
"num_tokens": 846209073.0,
"step": 6500
},
{
"epoch": 0.3379080889997921,
"eval_entropy": 1.9439453125,
"eval_loss": 1.2014882564544678,
"eval_mean_token_accuracy": 0.7102363556623459,
"eval_num_tokens": 846209073.0,
"eval_runtime": 26.9689,
"eval_samples_per_second": 93.96,
"eval_steps_per_second": 1.483,
"step": 6500
},
{
"entropy": 1.9140625,
"epoch": 0.3384279475982533,
"grad_norm": 2.669414119234351,
"learning_rate": 6.8212027012541535e-06,
"loss": 1.1632,
"mean_token_accuracy": 0.7189427733421325,
"num_tokens": 847509293.0,
"step": 6510
},
{
"entropy": 1.9546875,
"epoch": 0.3389478061967145,
"grad_norm": 2.825197461696076,
"learning_rate": 6.8158430699967844e-06,
"loss": 1.1814,
"mean_token_accuracy": 0.7159137547016143,
"num_tokens": 848809529.0,
"step": 6520
},
{
"entropy": 1.93671875,
"epoch": 0.3394676647951757,
"grad_norm": 2.91585176965479,
"learning_rate": 6.8104834387394146e-06,
"loss": 1.2252,
"mean_token_accuracy": 0.7063211143016815,
"num_tokens": 850113283.0,
"step": 6530
},
{
"entropy": 1.95625,
"epoch": 0.3399875233936369,
"grad_norm": 2.846973761028311,
"learning_rate": 6.8051238074820455e-06,
"loss": 1.1641,
"mean_token_accuracy": 0.7196978271007538,
"num_tokens": 851413746.0,
"step": 6540
},
{
"entropy": 1.930859375,
"epoch": 0.3405073819920981,
"grad_norm": 2.8221678288507897,
"learning_rate": 6.799764176224676e-06,
"loss": 1.1968,
"mean_token_accuracy": 0.7108127981424331,
"num_tokens": 852714521.0,
"step": 6550
},
{
"entropy": 1.93203125,
"epoch": 0.3410272405905594,
"grad_norm": 2.8682442992662187,
"learning_rate": 6.794404544967307e-06,
"loss": 1.145,
"mean_token_accuracy": 0.7231799691915513,
"num_tokens": 854016048.0,
"step": 6560
},
{
"entropy": 1.9625,
"epoch": 0.3415470991890206,
"grad_norm": 2.4238462001417926,
"learning_rate": 6.789044913709937e-06,
"loss": 1.1983,
"mean_token_accuracy": 0.7144931375980377,
"num_tokens": 855316696.0,
"step": 6570
},
{
"entropy": 1.88984375,
"epoch": 0.3420669577874818,
"grad_norm": 3.0683828424527966,
"learning_rate": 6.783685282452568e-06,
"loss": 1.189,
"mean_token_accuracy": 0.7125275641679764,
"num_tokens": 856620812.0,
"step": 6580
},
{
"entropy": 1.95390625,
"epoch": 0.342586816385943,
"grad_norm": 2.5900780721299643,
"learning_rate": 6.778325651195198e-06,
"loss": 1.1807,
"mean_token_accuracy": 0.7170392394065856,
"num_tokens": 857919522.0,
"step": 6590
},
{
"entropy": 1.9625,
"epoch": 0.3431066749844042,
"grad_norm": 2.8177910375379276,
"learning_rate": 6.772966019937829e-06,
"loss": 1.1113,
"mean_token_accuracy": 0.7344673156738282,
"num_tokens": 859220784.0,
"step": 6600
},
{
"epoch": 0.3431066749844042,
"eval_entropy": 1.9400390625,
"eval_loss": 1.199386477470398,
"eval_mean_token_accuracy": 0.7108535036444664,
"eval_num_tokens": 859220784.0,
"eval_runtime": 27.001,
"eval_samples_per_second": 93.848,
"eval_steps_per_second": 1.481,
"step": 6600
},
{
"entropy": 1.93984375,
"epoch": 0.34362653358286543,
"grad_norm": 2.818528268391227,
"learning_rate": 6.767606388680459e-06,
"loss": 1.1735,
"mean_token_accuracy": 0.717198520898819,
"num_tokens": 860523206.0,
"step": 6610
},
{
"entropy": 1.9578125,
"epoch": 0.3441463921813267,
"grad_norm": 2.697909090214905,
"learning_rate": 6.76224675742309e-06,
"loss": 1.1586,
"mean_token_accuracy": 0.7228897750377655,
"num_tokens": 861822348.0,
"step": 6620
},
{
"entropy": 1.944140625,
"epoch": 0.3446662507797879,
"grad_norm": 2.7717538084609865,
"learning_rate": 6.75688712616572e-06,
"loss": 1.1914,
"mean_token_accuracy": 0.7141066640615463,
"num_tokens": 863122647.0,
"step": 6630
},
{
"entropy": 1.95546875,
"epoch": 0.3451861093782491,
"grad_norm": 2.622097443854084,
"learning_rate": 6.751527494908351e-06,
"loss": 1.1408,
"mean_token_accuracy": 0.7276633977890015,
"num_tokens": 864421463.0,
"step": 6640
},
{
"entropy": 1.97421875,
"epoch": 0.3457059679767103,
"grad_norm": 2.7649037967117125,
"learning_rate": 6.746167863650981e-06,
"loss": 1.1709,
"mean_token_accuracy": 0.7196216821670532,
"num_tokens": 865723803.0,
"step": 6650
},
{
"entropy": 1.945703125,
"epoch": 0.34622582657517154,
"grad_norm": 2.2573271800184127,
"learning_rate": 6.740808232393612e-06,
"loss": 1.157,
"mean_token_accuracy": 0.7230482935905457,
"num_tokens": 867025587.0,
"step": 6660
},
{
"entropy": 1.96953125,
"epoch": 0.34674568517363275,
"grad_norm": 2.479024683233382,
"learning_rate": 6.735448601136242e-06,
"loss": 1.22,
"mean_token_accuracy": 0.7064239323139191,
"num_tokens": 868328217.0,
"step": 6670
},
{
"entropy": 1.92421875,
"epoch": 0.347265543772094,
"grad_norm": 2.9169891102808956,
"learning_rate": 6.730088969878873e-06,
"loss": 1.1652,
"mean_token_accuracy": 0.7186296939849853,
"num_tokens": 869629921.0,
"step": 6680
},
{
"entropy": 1.953515625,
"epoch": 0.3477854023705552,
"grad_norm": 2.8341222563948367,
"learning_rate": 6.724729338621503e-06,
"loss": 1.213,
"mean_token_accuracy": 0.710431706905365,
"num_tokens": 870931329.0,
"step": 6690
},
{
"entropy": 1.9796875,
"epoch": 0.34830526096901643,
"grad_norm": 2.9209615074802078,
"learning_rate": 6.719369707364134e-06,
"loss": 1.1544,
"mean_token_accuracy": 0.7231431782245636,
"num_tokens": 872232882.0,
"step": 6700
},
{
"epoch": 0.34830526096901643,
"eval_entropy": 1.9072265625,
"eval_loss": 1.1972352266311646,
"eval_mean_token_accuracy": 0.7116771385073661,
"eval_num_tokens": 872232882.0,
"eval_runtime": 26.9703,
"eval_samples_per_second": 93.955,
"eval_steps_per_second": 1.483,
"step": 6700
},
{
"entropy": 1.951171875,
"epoch": 0.34882511956747764,
"grad_norm": 2.787520774351182,
"learning_rate": 6.7140100761067636e-06,
"loss": 1.1626,
"mean_token_accuracy": 0.7212519347667694,
"num_tokens": 873534943.0,
"step": 6710
},
{
"entropy": 1.926953125,
"epoch": 0.34934497816593885,
"grad_norm": 2.517338958986832,
"learning_rate": 6.708650444849395e-06,
"loss": 1.1297,
"mean_token_accuracy": 0.7273350149393082,
"num_tokens": 874835577.0,
"step": 6720
},
{
"entropy": 1.94765625,
"epoch": 0.34986483676440006,
"grad_norm": 2.428098778552561,
"learning_rate": 6.703290813592025e-06,
"loss": 1.102,
"mean_token_accuracy": 0.7335271000862121,
"num_tokens": 876138531.0,
"step": 6730
},
{
"entropy": 1.940234375,
"epoch": 0.3503846953628613,
"grad_norm": 2.788258655699573,
"learning_rate": 6.697931182334656e-06,
"loss": 1.1525,
"mean_token_accuracy": 0.7215675085783004,
"num_tokens": 877440550.0,
"step": 6740
},
{
"entropy": 1.962109375,
"epoch": 0.35090455396132253,
"grad_norm": 2.985966265488703,
"learning_rate": 6.692571551077286e-06,
"loss": 1.1885,
"mean_token_accuracy": 0.718495887517929,
"num_tokens": 878743636.0,
"step": 6750
},
{
"entropy": 1.932421875,
"epoch": 0.35142441255978374,
"grad_norm": 2.7821670662298694,
"learning_rate": 6.687211919819917e-06,
"loss": 1.1435,
"mean_token_accuracy": 0.7251176506280899,
"num_tokens": 880044750.0,
"step": 6760
},
{
"entropy": 1.94921875,
"epoch": 0.35194427115824495,
"grad_norm": 2.396550474038095,
"learning_rate": 6.681852288562547e-06,
"loss": 1.1924,
"mean_token_accuracy": 0.7125939130783081,
"num_tokens": 881348648.0,
"step": 6770
},
{
"entropy": 1.9375,
"epoch": 0.35246412975670616,
"grad_norm": 2.9681778018037797,
"learning_rate": 6.676492657305178e-06,
"loss": 1.1332,
"mean_token_accuracy": 0.7237930655479431,
"num_tokens": 882647148.0,
"step": 6780
},
{
"entropy": 1.940625,
"epoch": 0.3529839883551674,
"grad_norm": 2.685655997560711,
"learning_rate": 6.671133026047808e-06,
"loss": 1.1581,
"mean_token_accuracy": 0.7218601882457734,
"num_tokens": 883948440.0,
"step": 6790
},
{
"entropy": 1.94375,
"epoch": 0.35350384695362863,
"grad_norm": 2.4575104201462317,
"learning_rate": 6.665773394790439e-06,
"loss": 1.1628,
"mean_token_accuracy": 0.7186234414577484,
"num_tokens": 885247191.0,
"step": 6800
},
{
"epoch": 0.35350384695362863,
"eval_entropy": 1.9322265625,
"eval_loss": 1.1942270994186401,
"eval_mean_token_accuracy": 0.7126296281814575,
"eval_num_tokens": 885247191.0,
"eval_runtime": 27.0047,
"eval_samples_per_second": 93.836,
"eval_steps_per_second": 1.481,
"step": 6800
},
{
"entropy": 1.9625,
"epoch": 0.35402370555208984,
"grad_norm": 2.760318457437893,
"learning_rate": 6.660413763533069e-06,
"loss": 1.2264,
"mean_token_accuracy": 0.709084814786911,
"num_tokens": 886550137.0,
"step": 6810
},
{
"entropy": 1.937109375,
"epoch": 0.35454356415055105,
"grad_norm": 2.646887318169609,
"learning_rate": 6.6550541322757e-06,
"loss": 1.1787,
"mean_token_accuracy": 0.7182801574468612,
"num_tokens": 887850575.0,
"step": 6820
},
{
"entropy": 1.9515625,
"epoch": 0.35506342274901226,
"grad_norm": 2.8086068476221597,
"learning_rate": 6.64969450101833e-06,
"loss": 1.1852,
"mean_token_accuracy": 0.7169524222612381,
"num_tokens": 889152807.0,
"step": 6830
},
{
"entropy": 1.976171875,
"epoch": 0.35558328134747347,
"grad_norm": 2.887517901815175,
"learning_rate": 6.644334869760961e-06,
"loss": 1.2385,
"mean_token_accuracy": 0.7056964993476867,
"num_tokens": 890455531.0,
"step": 6840
},
{
"entropy": 1.95625,
"epoch": 0.35610313994593473,
"grad_norm": 2.5849251926887216,
"learning_rate": 6.638975238503591e-06,
"loss": 1.1437,
"mean_token_accuracy": 0.7245052099227905,
"num_tokens": 891757744.0,
"step": 6850
},
{
"entropy": 1.93984375,
"epoch": 0.35662299854439594,
"grad_norm": 2.4522942341464646,
"learning_rate": 6.633615607246222e-06,
"loss": 1.1862,
"mean_token_accuracy": 0.7176487267017364,
"num_tokens": 893061618.0,
"step": 6860
},
{
"entropy": 1.9484375,
"epoch": 0.35714285714285715,
"grad_norm": 2.6037344214149805,
"learning_rate": 6.628255975988852e-06,
"loss": 1.1558,
"mean_token_accuracy": 0.7214541405439376,
"num_tokens": 894363853.0,
"step": 6870
},
{
"entropy": 1.941796875,
"epoch": 0.35766271574131836,
"grad_norm": 2.605556265623727,
"learning_rate": 6.622896344731483e-06,
"loss": 1.1082,
"mean_token_accuracy": 0.7319776952266693,
"num_tokens": 895664421.0,
"step": 6880
},
{
"entropy": 1.93125,
"epoch": 0.35818257433977957,
"grad_norm": 2.8198364797875275,
"learning_rate": 6.617536713474113e-06,
"loss": 1.1414,
"mean_token_accuracy": 0.7239604979753494,
"num_tokens": 896966952.0,
"step": 6890
},
{
"entropy": 1.935546875,
"epoch": 0.3587024329382408,
"grad_norm": 2.703966207158738,
"learning_rate": 6.612177082216744e-06,
"loss": 1.1752,
"mean_token_accuracy": 0.720953693985939,
"num_tokens": 898270643.0,
"step": 6900
},
{
"epoch": 0.3587024329382408,
"eval_entropy": 1.927734375,
"eval_loss": 1.1888833045959473,
"eval_mean_token_accuracy": 0.7133261740207673,
"eval_num_tokens": 898270643.0,
"eval_runtime": 27.0162,
"eval_samples_per_second": 93.795,
"eval_steps_per_second": 1.481,
"step": 6900
},
{
"entropy": 1.94921875,
"epoch": 0.35922229153670204,
"grad_norm": 2.8267809499727026,
"learning_rate": 6.6068174509593745e-06,
"loss": 1.1551,
"mean_token_accuracy": 0.7211970090866089,
"num_tokens": 899573366.0,
"step": 6910
},
{
"entropy": 1.921484375,
"epoch": 0.35974215013516325,
"grad_norm": 2.920654193820593,
"learning_rate": 6.6014578197020055e-06,
"loss": 1.1612,
"mean_token_accuracy": 0.7210631966590881,
"num_tokens": 900876331.0,
"step": 6920
},
{
"entropy": 1.92734375,
"epoch": 0.36026200873362446,
"grad_norm": 3.1613251400632048,
"learning_rate": 6.596098188444635e-06,
"loss": 1.1374,
"mean_token_accuracy": 0.7249650925397872,
"num_tokens": 902178252.0,
"step": 6930
},
{
"entropy": 1.936328125,
"epoch": 0.36078186733208567,
"grad_norm": 2.8085966853611204,
"learning_rate": 6.5907385571872666e-06,
"loss": 1.1293,
"mean_token_accuracy": 0.7276577055454254,
"num_tokens": 903481533.0,
"step": 6940
},
{
"entropy": 1.941796875,
"epoch": 0.3613017259305469,
"grad_norm": 2.706006482727674,
"learning_rate": 6.585378925929896e-06,
"loss": 1.1172,
"mean_token_accuracy": 0.7326745629310608,
"num_tokens": 904782100.0,
"step": 6950
},
{
"entropy": 1.91796875,
"epoch": 0.3618215845290081,
"grad_norm": 2.763208238813126,
"learning_rate": 6.580019294672527e-06,
"loss": 1.1367,
"mean_token_accuracy": 0.723825216293335,
"num_tokens": 906085106.0,
"step": 6960
},
{
"entropy": 1.93125,
"epoch": 0.36234144312746935,
"grad_norm": 2.551734117038396,
"learning_rate": 6.574659663415157e-06,
"loss": 1.1573,
"mean_token_accuracy": 0.7208518415689469,
"num_tokens": 907384393.0,
"step": 6970
},
{
"entropy": 1.9625,
"epoch": 0.36286130172593056,
"grad_norm": 2.692708956537201,
"learning_rate": 6.569300032157788e-06,
"loss": 1.1653,
"mean_token_accuracy": 0.7217450112104415,
"num_tokens": 908687555.0,
"step": 6980
},
{
"entropy": 1.93046875,
"epoch": 0.36338116032439177,
"grad_norm": 2.978012911010852,
"learning_rate": 6.563940400900418e-06,
"loss": 1.1662,
"mean_token_accuracy": 0.7204705119132996,
"num_tokens": 909988983.0,
"step": 6990
},
{
"entropy": 1.958984375,
"epoch": 0.363901018922853,
"grad_norm": 2.9272493401235846,
"learning_rate": 6.558580769643049e-06,
"loss": 1.152,
"mean_token_accuracy": 0.7241693019866944,
"num_tokens": 911292615.0,
"step": 7000
},
{
"epoch": 0.363901018922853,
"eval_entropy": 1.880859375,
"eval_loss": 1.1863093376159668,
"eval_mean_token_accuracy": 0.7136569365859031,
"eval_num_tokens": 911292615.0,
"eval_runtime": 26.9922,
"eval_samples_per_second": 93.879,
"eval_steps_per_second": 1.482,
"step": 7000
},
{
"entropy": 1.9109375,
"epoch": 0.3644208775213142,
"grad_norm": 2.473933057602556,
"learning_rate": 6.553221138385679e-06,
"loss": 1.1854,
"mean_token_accuracy": 0.7157957553863525,
"num_tokens": 912593978.0,
"step": 7010
},
{
"entropy": 1.93984375,
"epoch": 0.3649407361197754,
"grad_norm": 2.700935500823765,
"learning_rate": 6.54786150712831e-06,
"loss": 1.098,
"mean_token_accuracy": 0.7362295836210251,
"num_tokens": 913894157.0,
"step": 7020
},
{
"entropy": 1.93125,
"epoch": 0.36546059471823666,
"grad_norm": 2.50043966444787,
"learning_rate": 6.54250187587094e-06,
"loss": 1.173,
"mean_token_accuracy": 0.7187134027481079,
"num_tokens": 915197550.0,
"step": 7030
},
{
"entropy": 1.93203125,
"epoch": 0.3659804533166979,
"grad_norm": 2.8962438249507843,
"learning_rate": 6.537142244613571e-06,
"loss": 1.1497,
"mean_token_accuracy": 0.7224205017089844,
"num_tokens": 916500275.0,
"step": 7040
},
{
"entropy": 1.93359375,
"epoch": 0.3665003119151591,
"grad_norm": 2.5848234502286007,
"learning_rate": 6.531782613356201e-06,
"loss": 1.1817,
"mean_token_accuracy": 0.7162549763917923,
"num_tokens": 917802840.0,
"step": 7050
},
{
"entropy": 1.937109375,
"epoch": 0.3670201705136203,
"grad_norm": 2.632361116712476,
"learning_rate": 6.526422982098832e-06,
"loss": 1.204,
"mean_token_accuracy": 0.7119734883308411,
"num_tokens": 919107709.0,
"step": 7060
},
{
"entropy": 1.9453125,
"epoch": 0.3675400291120815,
"grad_norm": 2.542345995233883,
"learning_rate": 6.521063350841462e-06,
"loss": 1.1701,
"mean_token_accuracy": 0.7180581241846085,
"num_tokens": 920411208.0,
"step": 7070
},
{
"entropy": 1.91796875,
"epoch": 0.3680598877105427,
"grad_norm": 3.065928600270698,
"learning_rate": 6.515703719584093e-06,
"loss": 1.1268,
"mean_token_accuracy": 0.7270506829023361,
"num_tokens": 921712494.0,
"step": 7080
},
{
"entropy": 1.851953125,
"epoch": 0.368579746309004,
"grad_norm": 2.74005201467313,
"learning_rate": 6.5103440883267235e-06,
"loss": 1.1096,
"mean_token_accuracy": 0.7313259452581405,
"num_tokens": 923014988.0,
"step": 7090
},
{
"entropy": 1.887109375,
"epoch": 0.3690996049074652,
"grad_norm": 2.4824331576792824,
"learning_rate": 6.5049844570693544e-06,
"loss": 1.1539,
"mean_token_accuracy": 0.7215629100799561,
"num_tokens": 924317651.0,
"step": 7100
},
{
"epoch": 0.3690996049074652,
"eval_entropy": 1.92578125,
"eval_loss": 1.1822230815887451,
"eval_mean_token_accuracy": 0.7150372371077538,
"eval_num_tokens": 924317651.0,
"eval_runtime": 26.9906,
"eval_samples_per_second": 93.885,
"eval_steps_per_second": 1.482,
"step": 7100
},
{
"entropy": 1.92578125,
"epoch": 0.3696194635059264,
"grad_norm": 2.98181234167671,
"learning_rate": 6.4996248258119846e-06,
"loss": 1.1239,
"mean_token_accuracy": 0.7315758109092713,
"num_tokens": 925619616.0,
"step": 7110
},
{
"entropy": 1.91640625,
"epoch": 0.3701393221043876,
"grad_norm": 2.4499704012713948,
"learning_rate": 6.4942651945546155e-06,
"loss": 1.1154,
"mean_token_accuracy": 0.7281979411840439,
"num_tokens": 926919113.0,
"step": 7120
},
{
"entropy": 1.9453125,
"epoch": 0.3706591807028488,
"grad_norm": 2.527433938787531,
"learning_rate": 6.488905563297246e-06,
"loss": 1.1595,
"mean_token_accuracy": 0.7217468053102494,
"num_tokens": 928219717.0,
"step": 7130
},
{
"entropy": 1.97734375,
"epoch": 0.37117903930131,
"grad_norm": 3.0076346704481707,
"learning_rate": 6.483545932039877e-06,
"loss": 1.1749,
"mean_token_accuracy": 0.715992221236229,
"num_tokens": 929519323.0,
"step": 7140
},
{
"entropy": 1.920703125,
"epoch": 0.3716988978997713,
"grad_norm": 2.7120482163460085,
"learning_rate": 6.478186300782506e-06,
"loss": 1.1389,
"mean_token_accuracy": 0.7267408281564712,
"num_tokens": 930819613.0,
"step": 7150
},
{
"entropy": 1.93828125,
"epoch": 0.3722187564982325,
"grad_norm": 3.0646418828220563,
"learning_rate": 6.472826669525138e-06,
"loss": 1.1462,
"mean_token_accuracy": 0.7238803923130035,
"num_tokens": 932120580.0,
"step": 7160
},
{
"entropy": 1.940625,
"epoch": 0.3727386150966937,
"grad_norm": 2.4260433578740477,
"learning_rate": 6.467467038267767e-06,
"loss": 1.1295,
"mean_token_accuracy": 0.7292232513427734,
"num_tokens": 933423127.0,
"step": 7170
},
{
"entropy": 1.910546875,
"epoch": 0.3732584736951549,
"grad_norm": 2.4145507802455537,
"learning_rate": 6.462107407010399e-06,
"loss": 1.14,
"mean_token_accuracy": 0.727630203962326,
"num_tokens": 934724924.0,
"step": 7180
},
{
"entropy": 1.93125,
"epoch": 0.3737783322936161,
"grad_norm": 3.0634087980494966,
"learning_rate": 6.456747775753028e-06,
"loss": 1.1968,
"mean_token_accuracy": 0.7128107428550721,
"num_tokens": 936025825.0,
"step": 7190
},
{
"entropy": 1.916015625,
"epoch": 0.37429819089207733,
"grad_norm": 2.2880829480048948,
"learning_rate": 6.451388144495659e-06,
"loss": 1.1556,
"mean_token_accuracy": 0.7233588993549347,
"num_tokens": 937325836.0,
"step": 7200
},
{
"epoch": 0.37429819089207733,
"eval_entropy": 1.9197265625,
"eval_loss": 1.1795099973678589,
"eval_mean_token_accuracy": 0.7156882435083389,
"eval_num_tokens": 937325836.0,
"eval_runtime": 27.051,
"eval_samples_per_second": 93.675,
"eval_steps_per_second": 1.479,
"step": 7200
},
{
"entropy": 1.93046875,
"epoch": 0.3748180494905386,
"grad_norm": 2.9580668572442006,
"learning_rate": 6.446028513238289e-06,
"loss": 1.1308,
"mean_token_accuracy": 0.7290701448917389,
"num_tokens": 938628898.0,
"step": 7210
},
{
"entropy": 1.9109375,
"epoch": 0.3753379080889998,
"grad_norm": 2.735439173800798,
"learning_rate": 6.44066888198092e-06,
"loss": 1.0881,
"mean_token_accuracy": 0.7368822067975997,
"num_tokens": 939929896.0,
"step": 7220
},
{
"entropy": 1.8984375,
"epoch": 0.375857766687461,
"grad_norm": 2.5052510644720676,
"learning_rate": 6.43530925072355e-06,
"loss": 1.1642,
"mean_token_accuracy": 0.7206080019474029,
"num_tokens": 941231732.0,
"step": 7230
},
{
"entropy": 1.959765625,
"epoch": 0.3763776252859222,
"grad_norm": 2.9330497763729637,
"learning_rate": 6.429949619466181e-06,
"loss": 1.1461,
"mean_token_accuracy": 0.723334339261055,
"num_tokens": 942532327.0,
"step": 7240
},
{
"entropy": 1.9265625,
"epoch": 0.37689748388438343,
"grad_norm": 2.608666451303606,
"learning_rate": 6.424589988208811e-06,
"loss": 1.1578,
"mean_token_accuracy": 0.7226617753505706,
"num_tokens": 943832680.0,
"step": 7250
},
{
"entropy": 1.944921875,
"epoch": 0.37741734248284464,
"grad_norm": 2.725553370051137,
"learning_rate": 6.419230356951442e-06,
"loss": 1.1731,
"mean_token_accuracy": 0.7184575766324997,
"num_tokens": 945135245.0,
"step": 7260
},
{
"entropy": 1.926953125,
"epoch": 0.3779372010813059,
"grad_norm": 2.6480535180649927,
"learning_rate": 6.4138707256940725e-06,
"loss": 1.1947,
"mean_token_accuracy": 0.7118302583694458,
"num_tokens": 946439378.0,
"step": 7270
},
{
"entropy": 1.929296875,
"epoch": 0.3784570596797671,
"grad_norm": 2.8482797759937015,
"learning_rate": 6.4085110944367034e-06,
"loss": 1.2059,
"mean_token_accuracy": 0.7115998357534409,
"num_tokens": 947740936.0,
"step": 7280
},
{
"entropy": 1.91796875,
"epoch": 0.3789769182782283,
"grad_norm": 2.430913407079125,
"learning_rate": 6.4031514631793336e-06,
"loss": 1.122,
"mean_token_accuracy": 0.7316271841526032,
"num_tokens": 949042036.0,
"step": 7290
},
{
"entropy": 1.93359375,
"epoch": 0.37949677687668953,
"grad_norm": 2.8511095894981398,
"learning_rate": 6.3977918319219645e-06,
"loss": 1.1452,
"mean_token_accuracy": 0.7255790710449219,
"num_tokens": 950340326.0,
"step": 7300
},
{
"epoch": 0.37949677687668953,
"eval_entropy": 1.9412109375,
"eval_loss": 1.1749075651168823,
"eval_mean_token_accuracy": 0.7166464775800705,
"eval_num_tokens": 950340326.0,
"eval_runtime": 27.0073,
"eval_samples_per_second": 93.827,
"eval_steps_per_second": 1.481,
"step": 7300
},
{
"entropy": 1.93515625,
"epoch": 0.38001663547515074,
"grad_norm": 2.5602902123916693,
"learning_rate": 6.392432200664595e-06,
"loss": 1.136,
"mean_token_accuracy": 0.7283244997262954,
"num_tokens": 951643321.0,
"step": 7310
},
{
"entropy": 1.95546875,
"epoch": 0.38053649407361195,
"grad_norm": 2.8165578039053947,
"learning_rate": 6.387072569407226e-06,
"loss": 1.193,
"mean_token_accuracy": 0.7147913664579392,
"num_tokens": 952944312.0,
"step": 7320
},
{
"entropy": 1.930078125,
"epoch": 0.3810563526720732,
"grad_norm": 2.7474442523819183,
"learning_rate": 6.381712938149856e-06,
"loss": 1.196,
"mean_token_accuracy": 0.7132438510656357,
"num_tokens": 954246960.0,
"step": 7330
},
{
"entropy": 1.9328125,
"epoch": 0.3815762112705344,
"grad_norm": 2.391434523941682,
"learning_rate": 6.376353306892487e-06,
"loss": 1.1685,
"mean_token_accuracy": 0.7192994058132172,
"num_tokens": 955550836.0,
"step": 7340
},
{
"entropy": 1.8953125,
"epoch": 0.38209606986899564,
"grad_norm": 2.799651425904629,
"learning_rate": 6.370993675635117e-06,
"loss": 1.1585,
"mean_token_accuracy": 0.7235007405281066,
"num_tokens": 956853619.0,
"step": 7350
},
{
"entropy": 1.93125,
"epoch": 0.38261592846745685,
"grad_norm": 2.6602091082406663,
"learning_rate": 6.365634044377748e-06,
"loss": 1.1572,
"mean_token_accuracy": 0.7216936737298966,
"num_tokens": 958154709.0,
"step": 7360
},
{
"entropy": 1.8890625,
"epoch": 0.38313578706591805,
"grad_norm": 2.9847063101390447,
"learning_rate": 6.360274413120378e-06,
"loss": 1.1196,
"mean_token_accuracy": 0.7290825963020324,
"num_tokens": 959456986.0,
"step": 7370
},
{
"entropy": 1.904296875,
"epoch": 0.38365564566437926,
"grad_norm": 3.047629440019641,
"learning_rate": 6.354914781863009e-06,
"loss": 1.1251,
"mean_token_accuracy": 0.7301168203353882,
"num_tokens": 960757816.0,
"step": 7380
},
{
"entropy": 1.935546875,
"epoch": 0.38417550426284053,
"grad_norm": 2.740549856918919,
"learning_rate": 6.349555150605638e-06,
"loss": 1.1445,
"mean_token_accuracy": 0.7244645357131958,
"num_tokens": 962060835.0,
"step": 7390
},
{
"entropy": 1.90859375,
"epoch": 0.38469536286130174,
"grad_norm": 2.6612888713726583,
"learning_rate": 6.34419551934827e-06,
"loss": 1.1291,
"mean_token_accuracy": 0.727899506688118,
"num_tokens": 963364306.0,
"step": 7400
},
{
"epoch": 0.38469536286130174,
"eval_entropy": 1.9119140625,
"eval_loss": 1.1737830638885498,
"eval_mean_token_accuracy": 0.7165374115109444,
"eval_num_tokens": 963364306.0,
"eval_runtime": 27.0022,
"eval_samples_per_second": 93.844,
"eval_steps_per_second": 1.481,
"step": 7400
},
{
"entropy": 1.904296875,
"epoch": 0.38521522145976295,
"grad_norm": 2.9281726390849636,
"learning_rate": 6.338835888090899e-06,
"loss": 1.1766,
"mean_token_accuracy": 0.7179235398769379,
"num_tokens": 964668155.0,
"step": 7410
},
{
"entropy": 1.950390625,
"epoch": 0.38573508005822416,
"grad_norm": 3.3147991875802085,
"learning_rate": 6.33347625683353e-06,
"loss": 1.1201,
"mean_token_accuracy": 0.7350755244493484,
"num_tokens": 965968982.0,
"step": 7420
},
{
"entropy": 1.921875,
"epoch": 0.38625493865668536,
"grad_norm": 2.864970628987281,
"learning_rate": 6.32811662557616e-06,
"loss": 1.1171,
"mean_token_accuracy": 0.7304784238338471,
"num_tokens": 967269974.0,
"step": 7430
},
{
"entropy": 1.8953125,
"epoch": 0.3867747972551466,
"grad_norm": 2.7801295754539743,
"learning_rate": 6.322756994318791e-06,
"loss": 1.1195,
"mean_token_accuracy": 0.7327520847320557,
"num_tokens": 968571785.0,
"step": 7440
},
{
"entropy": 1.933984375,
"epoch": 0.38729465585360784,
"grad_norm": 2.5208671807389957,
"learning_rate": 6.3173973630614215e-06,
"loss": 1.1195,
"mean_token_accuracy": 0.7305515229701995,
"num_tokens": 969874978.0,
"step": 7450
},
{
"entropy": 1.909765625,
"epoch": 0.38781451445206905,
"grad_norm": 2.523865324510605,
"learning_rate": 6.3120377318040524e-06,
"loss": 1.135,
"mean_token_accuracy": 0.7251843780279159,
"num_tokens": 971177860.0,
"step": 7460
},
{
"entropy": 1.927734375,
"epoch": 0.38833437305053026,
"grad_norm": 2.9259816594759607,
"learning_rate": 6.3066781005466826e-06,
"loss": 1.1138,
"mean_token_accuracy": 0.7307131439447403,
"num_tokens": 972479992.0,
"step": 7470
},
{
"entropy": 1.90625,
"epoch": 0.38885423164899147,
"grad_norm": 2.9327732317499335,
"learning_rate": 6.3013184692893135e-06,
"loss": 1.1413,
"mean_token_accuracy": 0.726539009809494,
"num_tokens": 973780387.0,
"step": 7480
},
{
"entropy": 1.909375,
"epoch": 0.3893740902474527,
"grad_norm": 2.8970821310137063,
"learning_rate": 6.295958838031944e-06,
"loss": 1.1019,
"mean_token_accuracy": 0.7322548359632493,
"num_tokens": 975082656.0,
"step": 7490
},
{
"entropy": 1.911328125,
"epoch": 0.3898939488459139,
"grad_norm": 2.6434524349881032,
"learning_rate": 6.290599206774575e-06,
"loss": 1.1466,
"mean_token_accuracy": 0.7253625154495239,
"num_tokens": 976382084.0,
"step": 7500
},
{
"epoch": 0.3898939488459139,
"eval_entropy": 1.891796875,
"eval_loss": 1.171810507774353,
"eval_mean_token_accuracy": 0.7167434692382812,
"eval_num_tokens": 976382084.0,
"eval_runtime": 27.0046,
"eval_samples_per_second": 93.836,
"eval_steps_per_second": 1.481,
"step": 7500
},
{
"entropy": 1.95859375,
"epoch": 0.39041380744437515,
"grad_norm": 2.9336386331452684,
"learning_rate": 6.285239575517205e-06,
"loss": 1.1238,
"mean_token_accuracy": 0.7321666300296783,
"num_tokens": 977686355.0,
"step": 7510
},
{
"entropy": 1.93828125,
"epoch": 0.39093366604283636,
"grad_norm": 2.6864851598603945,
"learning_rate": 6.279879944259836e-06,
"loss": 1.1878,
"mean_token_accuracy": 0.7171751588582993,
"num_tokens": 978987058.0,
"step": 7520
},
{
"entropy": 1.94609375,
"epoch": 0.39145352464129757,
"grad_norm": 2.7806686309964865,
"learning_rate": 6.274520313002466e-06,
"loss": 1.1555,
"mean_token_accuracy": 0.7224945217370987,
"num_tokens": 980288217.0,
"step": 7530
},
{
"entropy": 1.89296875,
"epoch": 0.3919733832397588,
"grad_norm": 2.6378463724595966,
"learning_rate": 6.269160681745097e-06,
"loss": 1.1509,
"mean_token_accuracy": 0.7239250987768173,
"num_tokens": 981590552.0,
"step": 7540
},
{
"entropy": 1.97265625,
"epoch": 0.39249324183822,
"grad_norm": 2.4902384769564323,
"learning_rate": 6.263801050487727e-06,
"loss": 1.1425,
"mean_token_accuracy": 0.7230521827936173,
"num_tokens": 982894694.0,
"step": 7550
},
{
"entropy": 1.9203125,
"epoch": 0.3930131004366812,
"grad_norm": 2.8620452620526566,
"learning_rate": 6.258441419230358e-06,
"loss": 1.0668,
"mean_token_accuracy": 0.7429023444652557,
"num_tokens": 984195321.0,
"step": 7560
},
{
"entropy": 1.93515625,
"epoch": 0.39353295903514246,
"grad_norm": 2.4914699164587604,
"learning_rate": 6.253081787972988e-06,
"loss": 1.1292,
"mean_token_accuracy": 0.7267493337392807,
"num_tokens": 985498736.0,
"step": 7570
},
{
"entropy": 1.918359375,
"epoch": 0.39405281763360367,
"grad_norm": 2.9507957869728463,
"learning_rate": 6.247722156715619e-06,
"loss": 1.1691,
"mean_token_accuracy": 0.7174822837114334,
"num_tokens": 986801702.0,
"step": 7580
},
{
"entropy": 1.907421875,
"epoch": 0.3945726762320649,
"grad_norm": 3.1225668153110706,
"learning_rate": 6.242362525458249e-06,
"loss": 1.1477,
"mean_token_accuracy": 0.7229850500822067,
"num_tokens": 988104726.0,
"step": 7590
},
{
"entropy": 1.88671875,
"epoch": 0.3950925348305261,
"grad_norm": 2.610425837813227,
"learning_rate": 6.23700289420088e-06,
"loss": 1.127,
"mean_token_accuracy": 0.726945674419403,
"num_tokens": 989406178.0,
"step": 7600
},
{
"epoch": 0.3950925348305261,
"eval_entropy": 1.9072265625,
"eval_loss": 1.169623613357544,
"eval_mean_token_accuracy": 0.7177217081189156,
"eval_num_tokens": 989406178.0,
"eval_runtime": 27.0691,
"eval_samples_per_second": 93.612,
"eval_steps_per_second": 1.478,
"step": 7600
},
{
"entropy": 1.91171875,
"epoch": 0.3956123934289873,
"grad_norm": 2.683118356428342,
"learning_rate": 6.231643262943509e-06,
"loss": 1.1829,
"mean_token_accuracy": 0.7180872768163681,
"num_tokens": 990707961.0,
"step": 7610
},
{
"entropy": 1.971484375,
"epoch": 0.39613225202744856,
"grad_norm": 2.939497162755304,
"learning_rate": 6.226283631686141e-06,
"loss": 1.1584,
"mean_token_accuracy": 0.7236785620450974,
"num_tokens": 992012116.0,
"step": 7620
},
{
"entropy": 1.9171875,
"epoch": 0.39665211062590977,
"grad_norm": 2.64756720713788,
"learning_rate": 6.2209240004287705e-06,
"loss": 1.1053,
"mean_token_accuracy": 0.731640100479126,
"num_tokens": 993311437.0,
"step": 7630
},
{
"entropy": 1.92734375,
"epoch": 0.397171969224371,
"grad_norm": 2.746869684784606,
"learning_rate": 6.2155643691714014e-06,
"loss": 1.1204,
"mean_token_accuracy": 0.7294285893440247,
"num_tokens": 994610084.0,
"step": 7640
},
{
"entropy": 1.903515625,
"epoch": 0.3976918278228322,
"grad_norm": 2.6033934979032267,
"learning_rate": 6.2102047379140315e-06,
"loss": 1.1834,
"mean_token_accuracy": 0.7151498556137085,
"num_tokens": 995914063.0,
"step": 7650
},
{
"entropy": 1.922265625,
"epoch": 0.3982116864212934,
"grad_norm": 3.0644363319313253,
"learning_rate": 6.2048451066566625e-06,
"loss": 1.1503,
"mean_token_accuracy": 0.7231369256973267,
"num_tokens": 997215684.0,
"step": 7660
},
{
"entropy": 1.93203125,
"epoch": 0.3987315450197546,
"grad_norm": 2.931806007272073,
"learning_rate": 6.199485475399293e-06,
"loss": 1.1684,
"mean_token_accuracy": 0.718231064081192,
"num_tokens": 998516501.0,
"step": 7670
},
{
"entropy": 1.896484375,
"epoch": 0.39925140361821587,
"grad_norm": 2.748134428586604,
"learning_rate": 6.194125844141924e-06,
"loss": 1.1371,
"mean_token_accuracy": 0.7280129939317703,
"num_tokens": 999818515.0,
"step": 7680
},
{
"entropy": 1.939453125,
"epoch": 0.3997712622166771,
"grad_norm": 2.799980280024601,
"learning_rate": 6.188766212884554e-06,
"loss": 1.1598,
"mean_token_accuracy": 0.7200548380613327,
"num_tokens": 1001120978.0,
"step": 7690
},
{
"entropy": 1.886328125,
"epoch": 0.4002911208151383,
"grad_norm": 3.1316004429029665,
"learning_rate": 6.183406581627185e-06,
"loss": 1.0793,
"mean_token_accuracy": 0.7388595819473267,
"num_tokens": 1002421831.0,
"step": 7700
},
{
"epoch": 0.4002911208151383,
"eval_entropy": 1.878515625,
"eval_loss": 1.1672450304031372,
"eval_mean_token_accuracy": 0.7186377570033073,
"eval_num_tokens": 1002421831.0,
"eval_runtime": 26.9572,
"eval_samples_per_second": 94.001,
"eval_steps_per_second": 1.484,
"step": 7700
},
{
"entropy": 1.92578125,
"epoch": 0.4008109794135995,
"grad_norm": 2.629023824973395,
"learning_rate": 6.178046950369815e-06,
"loss": 1.0871,
"mean_token_accuracy": 0.7401099145412445,
"num_tokens": 1003724642.0,
"step": 7710
},
{
"entropy": 1.92109375,
"epoch": 0.4013308380120607,
"grad_norm": 2.2425281577945606,
"learning_rate": 6.172687319112446e-06,
"loss": 1.1269,
"mean_token_accuracy": 0.7282724916934967,
"num_tokens": 1005024788.0,
"step": 7720
},
{
"entropy": 1.90859375,
"epoch": 0.4018506966105219,
"grad_norm": 2.7273217502276066,
"learning_rate": 6.167327687855076e-06,
"loss": 1.1535,
"mean_token_accuracy": 0.7202282726764679,
"num_tokens": 1006326027.0,
"step": 7730
},
{
"entropy": 1.915625,
"epoch": 0.4023705552089832,
"grad_norm": 2.8170596528635694,
"learning_rate": 6.161968056597707e-06,
"loss": 1.0676,
"mean_token_accuracy": 0.7432536721229553,
"num_tokens": 1007624556.0,
"step": 7740
},
{
"entropy": 1.8875,
"epoch": 0.4028904138074444,
"grad_norm": 2.56595444490234,
"learning_rate": 6.156608425340337e-06,
"loss": 1.0989,
"mean_token_accuracy": 0.7342011421918869,
"num_tokens": 1008926971.0,
"step": 7750
},
{
"entropy": 1.90234375,
"epoch": 0.4034102724059056,
"grad_norm": 3.124089108247137,
"learning_rate": 6.151248794082968e-06,
"loss": 1.118,
"mean_token_accuracy": 0.731161630153656,
"num_tokens": 1010229591.0,
"step": 7760
},
{
"entropy": 1.926953125,
"epoch": 0.4039301310043668,
"grad_norm": 3.1260413654677826,
"learning_rate": 6.145889162825598e-06,
"loss": 1.1128,
"mean_token_accuracy": 0.7326441794633866,
"num_tokens": 1011532569.0,
"step": 7770
},
{
"entropy": 1.927734375,
"epoch": 0.404449989602828,
"grad_norm": 2.6696200361313593,
"learning_rate": 6.140529531568229e-06,
"loss": 1.1515,
"mean_token_accuracy": 0.7251476407051086,
"num_tokens": 1012835066.0,
"step": 7780
},
{
"entropy": 1.94375,
"epoch": 0.40496984820128923,
"grad_norm": 2.5479515818627263,
"learning_rate": 6.135169900310859e-06,
"loss": 1.1266,
"mean_token_accuracy": 0.7308172047138214,
"num_tokens": 1014136323.0,
"step": 7790
},
{
"entropy": 1.907421875,
"epoch": 0.4054897067997505,
"grad_norm": 3.146804841054245,
"learning_rate": 6.12981026905349e-06,
"loss": 1.0886,
"mean_token_accuracy": 0.7376264661550522,
"num_tokens": 1015437138.0,
"step": 7800
},
{
"epoch": 0.4054897067997505,
"eval_entropy": 1.909375,
"eval_loss": 1.1622602939605713,
"eval_mean_token_accuracy": 0.7198067858815194,
"eval_num_tokens": 1015437138.0,
"eval_runtime": 26.961,
"eval_samples_per_second": 93.987,
"eval_steps_per_second": 1.484,
"step": 7800
},
{
"entropy": 1.88046875,
"epoch": 0.4060095653982117,
"grad_norm": 3.1794795213366447,
"learning_rate": 6.12445063779612e-06,
"loss": 1.0927,
"mean_token_accuracy": 0.7363271236419677,
"num_tokens": 1016737368.0,
"step": 7810
},
{
"entropy": 1.95234375,
"epoch": 0.4065294239966729,
"grad_norm": 3.0131724167585765,
"learning_rate": 6.119091006538751e-06,
"loss": 1.1398,
"mean_token_accuracy": 0.7293298661708831,
"num_tokens": 1018038818.0,
"step": 7820
},
{
"entropy": 1.883984375,
"epoch": 0.4070492825951341,
"grad_norm": 2.6961946760095348,
"learning_rate": 6.1137313752813805e-06,
"loss": 1.121,
"mean_token_accuracy": 0.7295084029436112,
"num_tokens": 1019338188.0,
"step": 7830
},
{
"entropy": 1.9078125,
"epoch": 0.40756914119359533,
"grad_norm": 2.748537555612983,
"learning_rate": 6.108371744024012e-06,
"loss": 1.1497,
"mean_token_accuracy": 0.7222037255764008,
"num_tokens": 1020641262.0,
"step": 7840
},
{
"entropy": 1.915234375,
"epoch": 0.40808899979205654,
"grad_norm": 2.669061792415386,
"learning_rate": 6.103012112766642e-06,
"loss": 1.1099,
"mean_token_accuracy": 0.734737080335617,
"num_tokens": 1021944033.0,
"step": 7850
},
{
"entropy": 1.920703125,
"epoch": 0.4086088583905178,
"grad_norm": 2.979769705886651,
"learning_rate": 6.0976524815092734e-06,
"loss": 1.1328,
"mean_token_accuracy": 0.7295629769563675,
"num_tokens": 1023245612.0,
"step": 7860
},
{
"entropy": 1.916796875,
"epoch": 0.409128716988979,
"grad_norm": 2.4432260769875795,
"learning_rate": 6.092292850251903e-06,
"loss": 1.1292,
"mean_token_accuracy": 0.7261045336723327,
"num_tokens": 1024546352.0,
"step": 7870
},
{
"entropy": 1.89921875,
"epoch": 0.4096485755874402,
"grad_norm": 2.5036540080510683,
"learning_rate": 6.086933218994534e-06,
"loss": 1.1621,
"mean_token_accuracy": 0.7206638336181641,
"num_tokens": 1025848111.0,
"step": 7880
},
{
"entropy": 1.891796875,
"epoch": 0.41016843418590143,
"grad_norm": 2.7181293529236674,
"learning_rate": 6.081573587737164e-06,
"loss": 1.0735,
"mean_token_accuracy": 0.7402589082717895,
"num_tokens": 1027149499.0,
"step": 7890
},
{
"entropy": 1.908984375,
"epoch": 0.41068829278436264,
"grad_norm": 3.1305059156354353,
"learning_rate": 6.076213956479795e-06,
"loss": 1.1286,
"mean_token_accuracy": 0.7272172003984452,
"num_tokens": 1028448899.0,
"step": 7900
},
{
"epoch": 0.41068829278436264,
"eval_entropy": 1.8625,
"eval_loss": 1.15581476688385,
"eval_mean_token_accuracy": 0.7205261573195457,
"eval_num_tokens": 1028448899.0,
"eval_runtime": 26.9807,
"eval_samples_per_second": 93.919,
"eval_steps_per_second": 1.483,
"step": 7900
},
{
"entropy": 1.88515625,
"epoch": 0.41120815138282385,
"grad_norm": 2.518120365071757,
"learning_rate": 6.070854325222425e-06,
"loss": 1.1373,
"mean_token_accuracy": 0.7300795376300812,
"num_tokens": 1029750106.0,
"step": 7910
},
{
"entropy": 1.935546875,
"epoch": 0.4117280099812851,
"grad_norm": 2.7711904346207024,
"learning_rate": 6.065494693965056e-06,
"loss": 1.1641,
"mean_token_accuracy": 0.7202939063310623,
"num_tokens": 1031052707.0,
"step": 7920
},
{
"entropy": 1.87890625,
"epoch": 0.4122478685797463,
"grad_norm": 3.241621961093104,
"learning_rate": 6.060135062707686e-06,
"loss": 1.1123,
"mean_token_accuracy": 0.7345538824796677,
"num_tokens": 1032354643.0,
"step": 7930
},
{
"entropy": 1.907421875,
"epoch": 0.41276772717820753,
"grad_norm": 3.351780952323983,
"learning_rate": 6.054775431450317e-06,
"loss": 1.0916,
"mean_token_accuracy": 0.7366264194250107,
"num_tokens": 1033657631.0,
"step": 7940
},
{
"entropy": 1.8734375,
"epoch": 0.41328758577666874,
"grad_norm": 2.9137908458136734,
"learning_rate": 6.049415800192947e-06,
"loss": 1.1167,
"mean_token_accuracy": 0.7332431226968765,
"num_tokens": 1034959687.0,
"step": 7950
},
{
"entropy": 1.906640625,
"epoch": 0.41380744437512995,
"grad_norm": 2.780901111646358,
"learning_rate": 6.044056168935578e-06,
"loss": 1.1168,
"mean_token_accuracy": 0.7297006338834763,
"num_tokens": 1036262425.0,
"step": 7960
},
{
"entropy": 1.882421875,
"epoch": 0.41432730297359116,
"grad_norm": 2.4547815337023806,
"learning_rate": 6.038696537678208e-06,
"loss": 1.0941,
"mean_token_accuracy": 0.7350849479436874,
"num_tokens": 1037563417.0,
"step": 7970
},
{
"entropy": 1.890625,
"epoch": 0.4148471615720524,
"grad_norm": 2.542553661850313,
"learning_rate": 6.033336906420839e-06,
"loss": 1.1068,
"mean_token_accuracy": 0.7344323873519898,
"num_tokens": 1038867427.0,
"step": 7980
},
{
"entropy": 1.8875,
"epoch": 0.41536702017051363,
"grad_norm": 2.7175723932350913,
"learning_rate": 6.027977275163469e-06,
"loss": 1.0821,
"mean_token_accuracy": 0.7418496817350387,
"num_tokens": 1040167212.0,
"step": 7990
},
{
"entropy": 1.928125,
"epoch": 0.41588687876897484,
"grad_norm": 2.633994570989676,
"learning_rate": 6.0226176439061e-06,
"loss": 1.1038,
"mean_token_accuracy": 0.7326177150011063,
"num_tokens": 1041468989.0,
"step": 8000
},
{
"epoch": 0.41588687876897484,
"eval_entropy": 1.878125,
"eval_loss": 1.1543986797332764,
"eval_mean_token_accuracy": 0.7215203553438186,
"eval_num_tokens": 1041468989.0,
"eval_runtime": 26.9966,
"eval_samples_per_second": 93.864,
"eval_steps_per_second": 1.482,
"step": 8000
},
{
"entropy": 1.925,
"epoch": 0.41640673736743605,
"grad_norm": 2.800613453906828,
"learning_rate": 6.01725801264873e-06,
"loss": 1.1489,
"mean_token_accuracy": 0.7233373820781708,
"num_tokens": 1042769189.0,
"step": 8010
},
{
"entropy": 1.928125,
"epoch": 0.41692659596589726,
"grad_norm": 2.7982042355815415,
"learning_rate": 6.011898381391361e-06,
"loss": 1.176,
"mean_token_accuracy": 0.7184455931186676,
"num_tokens": 1044070918.0,
"step": 8020
},
{
"entropy": 1.91953125,
"epoch": 0.41744645456435847,
"grad_norm": 2.8840979781555047,
"learning_rate": 6.0065387501339915e-06,
"loss": 1.158,
"mean_token_accuracy": 0.7204338669776916,
"num_tokens": 1045374609.0,
"step": 8030
},
{
"entropy": 1.91328125,
"epoch": 0.41796631316281974,
"grad_norm": 2.9915122101444873,
"learning_rate": 6.0011791188766224e-06,
"loss": 1.1452,
"mean_token_accuracy": 0.7248646497726441,
"num_tokens": 1046675823.0,
"step": 8040
},
{
"entropy": 1.92578125,
"epoch": 0.41848617176128095,
"grad_norm": 2.6711464146270663,
"learning_rate": 5.9958194876192526e-06,
"loss": 1.1717,
"mean_token_accuracy": 0.7211342096328736,
"num_tokens": 1047977777.0,
"step": 8050
},
{
"entropy": 1.9203125,
"epoch": 0.41900603035974215,
"grad_norm": 2.949129160692075,
"learning_rate": 5.9904598563618835e-06,
"loss": 1.0801,
"mean_token_accuracy": 0.7394367784261704,
"num_tokens": 1049277635.0,
"step": 8060
},
{
"entropy": 1.90234375,
"epoch": 0.41952588895820336,
"grad_norm": 2.958342327577349,
"learning_rate": 5.985100225104513e-06,
"loss": 1.1324,
"mean_token_accuracy": 0.729107740521431,
"num_tokens": 1050578301.0,
"step": 8070
},
{
"entropy": 1.910546875,
"epoch": 0.4200457475566646,
"grad_norm": 3.0158277265399525,
"learning_rate": 5.979740593847145e-06,
"loss": 1.0807,
"mean_token_accuracy": 0.736684051156044,
"num_tokens": 1051882253.0,
"step": 8080
},
{
"entropy": 1.892578125,
"epoch": 0.4205656061551258,
"grad_norm": 2.9042064952549564,
"learning_rate": 5.974380962589774e-06,
"loss": 1.0909,
"mean_token_accuracy": 0.7348469465970993,
"num_tokens": 1053181839.0,
"step": 8090
},
{
"entropy": 1.91640625,
"epoch": 0.42108546475358705,
"grad_norm": 2.7047521959452165,
"learning_rate": 5.969021331332405e-06,
"loss": 1.1318,
"mean_token_accuracy": 0.7291948467493057,
"num_tokens": 1054483946.0,
"step": 8100
},
{
"epoch": 0.42108546475358705,
"eval_entropy": 1.93671875,
"eval_loss": 1.1513324975967407,
"eval_mean_token_accuracy": 0.7223552018404007,
"eval_num_tokens": 1054483946.0,
"eval_runtime": 27.0671,
"eval_samples_per_second": 93.619,
"eval_steps_per_second": 1.478,
"step": 8100
},
{
"entropy": 1.928125,
"epoch": 0.42160532335204826,
"grad_norm": 2.6219196768049806,
"learning_rate": 5.963661700075035e-06,
"loss": 1.1856,
"mean_token_accuracy": 0.7169525355100632,
"num_tokens": 1055786315.0,
"step": 8110
},
{
"entropy": 1.909375,
"epoch": 0.42212518195050946,
"grad_norm": 2.4043950643535443,
"learning_rate": 5.958302068817666e-06,
"loss": 1.1148,
"mean_token_accuracy": 0.7340648859739304,
"num_tokens": 1057089851.0,
"step": 8120
},
{
"entropy": 1.946484375,
"epoch": 0.4226450405489707,
"grad_norm": 2.6586993776922334,
"learning_rate": 5.952942437560296e-06,
"loss": 1.1219,
"mean_token_accuracy": 0.7327827632427215,
"num_tokens": 1058392648.0,
"step": 8130
},
{
"entropy": 1.928125,
"epoch": 0.4231648991474319,
"grad_norm": 2.6183073559327856,
"learning_rate": 5.947582806302927e-06,
"loss": 1.0679,
"mean_token_accuracy": 0.7419730395078659,
"num_tokens": 1059695266.0,
"step": 8140
},
{
"entropy": 1.9171875,
"epoch": 0.4236847577458931,
"grad_norm": 3.7266703480236725,
"learning_rate": 5.942223175045557e-06,
"loss": 1.1393,
"mean_token_accuracy": 0.7257071435451508,
"num_tokens": 1060997082.0,
"step": 8150
},
{
"entropy": 1.905859375,
"epoch": 0.42420461634435436,
"grad_norm": 2.587025427426128,
"learning_rate": 5.936863543788188e-06,
"loss": 1.0864,
"mean_token_accuracy": 0.737755274772644,
"num_tokens": 1062298161.0,
"step": 8160
},
{
"entropy": 1.878125,
"epoch": 0.42472447494281557,
"grad_norm": 3.521989806757382,
"learning_rate": 5.931503912530818e-06,
"loss": 1.0928,
"mean_token_accuracy": 0.7353961735963821,
"num_tokens": 1063601825.0,
"step": 8170
},
{
"entropy": 1.904296875,
"epoch": 0.4252443335412768,
"grad_norm": 2.8711447086335404,
"learning_rate": 5.926144281273449e-06,
"loss": 1.1739,
"mean_token_accuracy": 0.7193310886621476,
"num_tokens": 1064904449.0,
"step": 8180
},
{
"entropy": 1.906640625,
"epoch": 0.425764192139738,
"grad_norm": 2.5144808705694737,
"learning_rate": 5.920784650016079e-06,
"loss": 1.1016,
"mean_token_accuracy": 0.7334045797586441,
"num_tokens": 1066206880.0,
"step": 8190
},
{
"entropy": 1.91015625,
"epoch": 0.4262840507381992,
"grad_norm": 2.7849587563051976,
"learning_rate": 5.91542501875871e-06,
"loss": 1.0991,
"mean_token_accuracy": 0.7339576125144959,
"num_tokens": 1067509344.0,
"step": 8200
},
{
"epoch": 0.4262840507381992,
"eval_entropy": 1.8900390625,
"eval_loss": 1.1506166458129883,
"eval_mean_token_accuracy": 0.7225477889180183,
"eval_num_tokens": 1067509344.0,
"eval_runtime": 27.0555,
"eval_samples_per_second": 93.659,
"eval_steps_per_second": 1.478,
"step": 8200
},
{
"entropy": 1.90859375,
"epoch": 0.4268039093366604,
"grad_norm": 2.670940609186009,
"learning_rate": 5.9100653875013405e-06,
"loss": 1.17,
"mean_token_accuracy": 0.718767312169075,
"num_tokens": 1068811889.0,
"step": 8210
},
{
"entropy": 1.86875,
"epoch": 0.42732376793512167,
"grad_norm": 2.944556901820497,
"learning_rate": 5.9047057562439714e-06,
"loss": 1.0979,
"mean_token_accuracy": 0.7336566060781479,
"num_tokens": 1070113151.0,
"step": 8220
},
{
"entropy": 1.91953125,
"epoch": 0.4278436265335829,
"grad_norm": 3.137967334733112,
"learning_rate": 5.8993461249866016e-06,
"loss": 1.1054,
"mean_token_accuracy": 0.7336654782295227,
"num_tokens": 1071414715.0,
"step": 8230
},
{
"entropy": 1.909375,
"epoch": 0.4283634851320441,
"grad_norm": 3.0021922427218177,
"learning_rate": 5.8939864937292325e-06,
"loss": 1.0998,
"mean_token_accuracy": 0.732508972287178,
"num_tokens": 1072713774.0,
"step": 8240
},
{
"entropy": 1.9203125,
"epoch": 0.4288833437305053,
"grad_norm": 2.7080229347771914,
"learning_rate": 5.888626862471863e-06,
"loss": 1.148,
"mean_token_accuracy": 0.7227941036224366,
"num_tokens": 1074017170.0,
"step": 8250
},
{
"entropy": 1.89296875,
"epoch": 0.4294032023289665,
"grad_norm": 2.6781419299253844,
"learning_rate": 5.883267231214494e-06,
"loss": 1.0751,
"mean_token_accuracy": 0.7401218920946121,
"num_tokens": 1075321138.0,
"step": 8260
},
{
"entropy": 1.955078125,
"epoch": 0.4299230609274277,
"grad_norm": 2.7839050060111292,
"learning_rate": 5.877907599957124e-06,
"loss": 1.1515,
"mean_token_accuracy": 0.7235105812549592,
"num_tokens": 1076621301.0,
"step": 8270
},
{
"entropy": 1.896484375,
"epoch": 0.430442919525889,
"grad_norm": 2.7900495153499505,
"learning_rate": 5.872547968699755e-06,
"loss": 1.0474,
"mean_token_accuracy": 0.7449130713939667,
"num_tokens": 1077921223.0,
"step": 8280
},
{
"entropy": 1.94140625,
"epoch": 0.4309627781243502,
"grad_norm": 2.680357949307976,
"learning_rate": 5.867188337442384e-06,
"loss": 1.0993,
"mean_token_accuracy": 0.7347194850444794,
"num_tokens": 1079224217.0,
"step": 8290
},
{
"entropy": 1.912109375,
"epoch": 0.4314826367228114,
"grad_norm": 2.820348386291068,
"learning_rate": 5.861828706185016e-06,
"loss": 1.1092,
"mean_token_accuracy": 0.730591145157814,
"num_tokens": 1080525299.0,
"step": 8300
},
{
"epoch": 0.4314826367228114,
"eval_entropy": 1.9146484375,
"eval_loss": 1.1484886407852173,
"eval_mean_token_accuracy": 0.7226079940795899,
"eval_num_tokens": 1080525299.0,
"eval_runtime": 26.9829,
"eval_samples_per_second": 93.911,
"eval_steps_per_second": 1.482,
"step": 8300
},
{
"entropy": 1.90625,
"epoch": 0.4320024953212726,
"grad_norm": 2.6505904446729676,
"learning_rate": 5.856469074927645e-06,
"loss": 1.1437,
"mean_token_accuracy": 0.7258694559335709,
"num_tokens": 1081827163.0,
"step": 8310
},
{
"entropy": 1.913671875,
"epoch": 0.4325223539197338,
"grad_norm": 3.008664691478876,
"learning_rate": 5.851109443670276e-06,
"loss": 1.1183,
"mean_token_accuracy": 0.7303029835224152,
"num_tokens": 1083127430.0,
"step": 8320
},
{
"entropy": 1.919921875,
"epoch": 0.433042212518195,
"grad_norm": 2.618144037644471,
"learning_rate": 5.845749812412906e-06,
"loss": 1.163,
"mean_token_accuracy": 0.7204117804765702,
"num_tokens": 1084431508.0,
"step": 8330
},
{
"entropy": 1.912109375,
"epoch": 0.4335620711166563,
"grad_norm": 2.4299376355628692,
"learning_rate": 5.840390181155537e-06,
"loss": 1.0646,
"mean_token_accuracy": 0.7440690219402313,
"num_tokens": 1085734521.0,
"step": 8340
},
{
"entropy": 1.9265625,
"epoch": 0.4340819297151175,
"grad_norm": 2.5416322856623355,
"learning_rate": 5.835030549898167e-06,
"loss": 1.156,
"mean_token_accuracy": 0.7208948999643325,
"num_tokens": 1087035509.0,
"step": 8350
},
{
"entropy": 1.88046875,
"epoch": 0.4346017883135787,
"grad_norm": 2.5133308555483427,
"learning_rate": 5.829670918640798e-06,
"loss": 1.096,
"mean_token_accuracy": 0.7358412474393845,
"num_tokens": 1088338867.0,
"step": 8360
},
{
"entropy": 1.887890625,
"epoch": 0.4351216469120399,
"grad_norm": 2.9590225681933706,
"learning_rate": 5.824311287383428e-06,
"loss": 1.1266,
"mean_token_accuracy": 0.7302644103765488,
"num_tokens": 1089642206.0,
"step": 8370
},
{
"entropy": 1.898828125,
"epoch": 0.4356415055105011,
"grad_norm": 2.608811062082287,
"learning_rate": 5.818951656126059e-06,
"loss": 1.1301,
"mean_token_accuracy": 0.729344156384468,
"num_tokens": 1090942738.0,
"step": 8380
},
{
"entropy": 1.919140625,
"epoch": 0.43616136410896233,
"grad_norm": 2.3736153153169948,
"learning_rate": 5.8135920248686895e-06,
"loss": 1.1432,
"mean_token_accuracy": 0.7254176020622254,
"num_tokens": 1092244011.0,
"step": 8390
},
{
"entropy": 1.909765625,
"epoch": 0.4366812227074236,
"grad_norm": 2.459738905306175,
"learning_rate": 5.80823239361132e-06,
"loss": 1.134,
"mean_token_accuracy": 0.7285784602165222,
"num_tokens": 1093546377.0,
"step": 8400
},
{
"epoch": 0.4366812227074236,
"eval_entropy": 1.923046875,
"eval_loss": 1.1468020677566528,
"eval_mean_token_accuracy": 0.7225772097706795,
"eval_num_tokens": 1093546377.0,
"eval_runtime": 27.0925,
"eval_samples_per_second": 93.531,
"eval_steps_per_second": 1.476,
"step": 8400
},
{
"entropy": 1.960546875,
"epoch": 0.4372010813058848,
"grad_norm": 2.6156756767002864,
"learning_rate": 5.8028727623539505e-06,
"loss": 1.2121,
"mean_token_accuracy": 0.7109515428543091,
"num_tokens": 1094848704.0,
"step": 8410
},
{
"entropy": 1.871875,
"epoch": 0.437720939904346,
"grad_norm": 2.7131763622996865,
"learning_rate": 5.7975131310965815e-06,
"loss": 1.0787,
"mean_token_accuracy": 0.738140195608139,
"num_tokens": 1096150370.0,
"step": 8420
},
{
"entropy": 1.885546875,
"epoch": 0.4382407985028072,
"grad_norm": 2.784098979997154,
"learning_rate": 5.792153499839212e-06,
"loss": 1.064,
"mean_token_accuracy": 0.7445976823568344,
"num_tokens": 1097452839.0,
"step": 8430
},
{
"entropy": 1.898046875,
"epoch": 0.43876065710126844,
"grad_norm": 2.4221773410145055,
"learning_rate": 5.786793868581843e-06,
"loss": 1.1325,
"mean_token_accuracy": 0.7268024265766144,
"num_tokens": 1098756203.0,
"step": 8440
},
{
"entropy": 1.855078125,
"epoch": 0.4392805156997297,
"grad_norm": 3.002023651832749,
"learning_rate": 5.781434237324473e-06,
"loss": 1.0651,
"mean_token_accuracy": 0.7422190129756927,
"num_tokens": 1100056839.0,
"step": 8450
},
{
"entropy": 1.89296875,
"epoch": 0.4398003742981909,
"grad_norm": 2.787282873460978,
"learning_rate": 5.776074606067104e-06,
"loss": 1.1136,
"mean_token_accuracy": 0.731579253077507,
"num_tokens": 1101358827.0,
"step": 8460
},
{
"entropy": 1.89140625,
"epoch": 0.4403202328966521,
"grad_norm": 2.3810501955138204,
"learning_rate": 5.770714974809734e-06,
"loss": 1.1314,
"mean_token_accuracy": 0.7286120533943177,
"num_tokens": 1102660010.0,
"step": 8470
},
{
"entropy": 1.864453125,
"epoch": 0.44084009149511333,
"grad_norm": 2.8815671959825795,
"learning_rate": 5.765355343552365e-06,
"loss": 1.1284,
"mean_token_accuracy": 0.7314709663391114,
"num_tokens": 1103961581.0,
"step": 8480
},
{
"entropy": 1.88671875,
"epoch": 0.44135995009357454,
"grad_norm": 2.940905040461145,
"learning_rate": 5.759995712294995e-06,
"loss": 1.1094,
"mean_token_accuracy": 0.7335751861333847,
"num_tokens": 1105265441.0,
"step": 8490
},
{
"entropy": 1.862890625,
"epoch": 0.44187980869203575,
"grad_norm": 2.7405728079981615,
"learning_rate": 5.754636081037626e-06,
"loss": 1.0913,
"mean_token_accuracy": 0.7365264147520065,
"num_tokens": 1106565630.0,
"step": 8500
},
{
"epoch": 0.44187980869203575,
"eval_entropy": 1.8544921875,
"eval_loss": 1.1418663263320923,
"eval_mean_token_accuracy": 0.7242832064628602,
"eval_num_tokens": 1106565630.0,
"eval_runtime": 26.9761,
"eval_samples_per_second": 93.935,
"eval_steps_per_second": 1.483,
"step": 8500
},
{
"entropy": 1.85546875,
"epoch": 0.442399667290497,
"grad_norm": 2.8611947945979703,
"learning_rate": 5.749276449780255e-06,
"loss": 1.0505,
"mean_token_accuracy": 0.7446595638990402,
"num_tokens": 1107870032.0,
"step": 8510
},
{
"entropy": 1.917578125,
"epoch": 0.4429195258889582,
"grad_norm": 2.8354767148785935,
"learning_rate": 5.743916818522887e-06,
"loss": 1.0871,
"mean_token_accuracy": 0.7378126084804535,
"num_tokens": 1109170571.0,
"step": 8520
},
{
"entropy": 1.914453125,
"epoch": 0.44343938448741943,
"grad_norm": 2.817867424837466,
"learning_rate": 5.738557187265516e-06,
"loss": 1.0874,
"mean_token_accuracy": 0.736152121424675,
"num_tokens": 1110469243.0,
"step": 8530
},
{
"entropy": 1.86328125,
"epoch": 0.44395924308588064,
"grad_norm": 2.7433707335911865,
"learning_rate": 5.733197556008147e-06,
"loss": 1.0678,
"mean_token_accuracy": 0.7465791553258896,
"num_tokens": 1111771141.0,
"step": 8540
},
{
"entropy": 1.905078125,
"epoch": 0.44447910168434185,
"grad_norm": 2.821038526880317,
"learning_rate": 5.727837924750777e-06,
"loss": 1.1648,
"mean_token_accuracy": 0.7224083334207535,
"num_tokens": 1113070955.0,
"step": 8550
},
{
"entropy": 1.9125,
"epoch": 0.44499896028280306,
"grad_norm": 2.739565729779149,
"learning_rate": 5.722478293493408e-06,
"loss": 1.0728,
"mean_token_accuracy": 0.7424873322248459,
"num_tokens": 1114373199.0,
"step": 8560
},
{
"entropy": 1.901171875,
"epoch": 0.4455188188812643,
"grad_norm": 3.3987385619965815,
"learning_rate": 5.7171186622360384e-06,
"loss": 1.1172,
"mean_token_accuracy": 0.7311230093240738,
"num_tokens": 1115673196.0,
"step": 8570
},
{
"entropy": 1.905859375,
"epoch": 0.44603867747972553,
"grad_norm": 2.89617385624965,
"learning_rate": 5.711759030978669e-06,
"loss": 1.0679,
"mean_token_accuracy": 0.7431687206029892,
"num_tokens": 1116976995.0,
"step": 8580
},
{
"entropy": 1.8921875,
"epoch": 0.44655853607818674,
"grad_norm": 2.7237297834562684,
"learning_rate": 5.7063993997212995e-06,
"loss": 1.0662,
"mean_token_accuracy": 0.7442143797874451,
"num_tokens": 1118277870.0,
"step": 8590
},
{
"entropy": 1.900390625,
"epoch": 0.44707839467664795,
"grad_norm": 2.6048070651863937,
"learning_rate": 5.7010397684639305e-06,
"loss": 1.1116,
"mean_token_accuracy": 0.7316889643669129,
"num_tokens": 1119579938.0,
"step": 8600
},
{
"epoch": 0.44707839467664795,
"eval_entropy": 1.8873046875,
"eval_loss": 1.1386899948120117,
"eval_mean_token_accuracy": 0.7249342352151871,
"eval_num_tokens": 1119579938.0,
"eval_runtime": 27.0379,
"eval_samples_per_second": 93.72,
"eval_steps_per_second": 1.479,
"step": 8600
},
{
"entropy": 1.892578125,
"epoch": 0.44759825327510916,
"grad_norm": 2.9361458550020125,
"learning_rate": 5.695680137206561e-06,
"loss": 1.1403,
"mean_token_accuracy": 0.7280385792255402,
"num_tokens": 1120881637.0,
"step": 8610
},
{
"entropy": 1.91328125,
"epoch": 0.44811811187357037,
"grad_norm": 3.03932628732968,
"learning_rate": 5.690320505949191e-06,
"loss": 1.0952,
"mean_token_accuracy": 0.7367443531751633,
"num_tokens": 1122183748.0,
"step": 8620
},
{
"entropy": 1.90859375,
"epoch": 0.44863797047203163,
"grad_norm": 2.828454216251391,
"learning_rate": 5.684960874691822e-06,
"loss": 1.1107,
"mean_token_accuracy": 0.7343203157186509,
"num_tokens": 1123485679.0,
"step": 8630
},
{
"entropy": 1.9171875,
"epoch": 0.44915782907049284,
"grad_norm": 2.5904996140615406,
"learning_rate": 5.679601243434452e-06,
"loss": 1.1133,
"mean_token_accuracy": 0.7332333266735077,
"num_tokens": 1124789807.0,
"step": 8640
},
{
"entropy": 1.850390625,
"epoch": 0.44967768766895405,
"grad_norm": 2.927296329277988,
"learning_rate": 5.674241612177083e-06,
"loss": 1.1428,
"mean_token_accuracy": 0.7263437777757644,
"num_tokens": 1126089908.0,
"step": 8650
},
{
"entropy": 1.895703125,
"epoch": 0.45019754626741526,
"grad_norm": 2.6858977783142413,
"learning_rate": 5.668881980919713e-06,
"loss": 1.0892,
"mean_token_accuracy": 0.7363338857889176,
"num_tokens": 1127389960.0,
"step": 8660
},
{
"entropy": 1.8859375,
"epoch": 0.45071740486587647,
"grad_norm": 3.322572176946994,
"learning_rate": 5.663522349662344e-06,
"loss": 1.0748,
"mean_token_accuracy": 0.7403390973806381,
"num_tokens": 1128691626.0,
"step": 8670
},
{
"entropy": 1.90859375,
"epoch": 0.4512372634643377,
"grad_norm": 2.6194966609347943,
"learning_rate": 5.658162718404974e-06,
"loss": 1.0466,
"mean_token_accuracy": 0.7473376870155335,
"num_tokens": 1129994759.0,
"step": 8680
},
{
"entropy": 1.86640625,
"epoch": 0.45175712206279894,
"grad_norm": 2.861976882288161,
"learning_rate": 5.652803087147605e-06,
"loss": 1.0664,
"mean_token_accuracy": 0.7413408368825912,
"num_tokens": 1131298573.0,
"step": 8690
},
{
"entropy": 1.925,
"epoch": 0.45227698066126015,
"grad_norm": 2.690947168299841,
"learning_rate": 5.647443455890234e-06,
"loss": 1.0883,
"mean_token_accuracy": 0.7382152915000916,
"num_tokens": 1132600894.0,
"step": 8700
},
{
"epoch": 0.45227698066126015,
"eval_entropy": 1.8875,
"eval_loss": 1.1386017799377441,
"eval_mean_token_accuracy": 0.7251568585634232,
"eval_num_tokens": 1132600894.0,
"eval_runtime": 27.0989,
"eval_samples_per_second": 93.509,
"eval_steps_per_second": 1.476,
"step": 8700
},
{
"entropy": 1.89609375,
"epoch": 0.45279683925972136,
"grad_norm": 2.7583171426459865,
"learning_rate": 5.642083824632866e-06,
"loss": 1.09,
"mean_token_accuracy": 0.7392143100500107,
"num_tokens": 1133902456.0,
"step": 8710
},
{
"entropy": 1.909375,
"epoch": 0.45331669785818257,
"grad_norm": 2.7225747913805702,
"learning_rate": 5.636724193375495e-06,
"loss": 1.1087,
"mean_token_accuracy": 0.7316530883312226,
"num_tokens": 1135204054.0,
"step": 8720
},
{
"entropy": 1.875,
"epoch": 0.4538365564566438,
"grad_norm": 2.7988884350702907,
"learning_rate": 5.631364562118126e-06,
"loss": 1.0827,
"mean_token_accuracy": 0.7425318211317062,
"num_tokens": 1136507614.0,
"step": 8730
},
{
"entropy": 1.88046875,
"epoch": 0.454356415055105,
"grad_norm": 2.963139824325591,
"learning_rate": 5.6260049308607565e-06,
"loss": 1.0829,
"mean_token_accuracy": 0.7388422340154648,
"num_tokens": 1137809078.0,
"step": 8740
},
{
"entropy": 1.89765625,
"epoch": 0.45487627365356625,
"grad_norm": 2.6468684253990973,
"learning_rate": 5.6206452996033874e-06,
"loss": 1.1547,
"mean_token_accuracy": 0.7250247955322265,
"num_tokens": 1139110712.0,
"step": 8750
},
{
"entropy": 1.875,
"epoch": 0.45539613225202746,
"grad_norm": 2.43840505893127,
"learning_rate": 5.6152856683460176e-06,
"loss": 1.0778,
"mean_token_accuracy": 0.7380663812160492,
"num_tokens": 1140413278.0,
"step": 8760
},
{
"entropy": 1.889453125,
"epoch": 0.4559159908504887,
"grad_norm": 2.9566769114795814,
"learning_rate": 5.6099260370886485e-06,
"loss": 1.0825,
"mean_token_accuracy": 0.7405662268400193,
"num_tokens": 1141713771.0,
"step": 8770
},
{
"entropy": 1.858203125,
"epoch": 0.4564358494489499,
"grad_norm": 2.683948885845185,
"learning_rate": 5.604566405831279e-06,
"loss": 1.0742,
"mean_token_accuracy": 0.7393934458494187,
"num_tokens": 1143016246.0,
"step": 8780
},
{
"entropy": 1.848828125,
"epoch": 0.4569557080474111,
"grad_norm": 2.530156770690808,
"learning_rate": 5.59920677457391e-06,
"loss": 1.0773,
"mean_token_accuracy": 0.7391411364078522,
"num_tokens": 1144318027.0,
"step": 8790
},
{
"entropy": 1.8890625,
"epoch": 0.4574755666458723,
"grad_norm": 3.0171417358423542,
"learning_rate": 5.59384714331654e-06,
"loss": 1.1639,
"mean_token_accuracy": 0.7201961368322373,
"num_tokens": 1145619241.0,
"step": 8800
},
{
"epoch": 0.4574755666458723,
"eval_entropy": 1.861328125,
"eval_loss": 1.1357858180999756,
"eval_mean_token_accuracy": 0.725797027349472,
"eval_num_tokens": 1145619241.0,
"eval_runtime": 26.9867,
"eval_samples_per_second": 93.898,
"eval_steps_per_second": 1.482,
"step": 8800
},
{
"entropy": 1.904296875,
"epoch": 0.45799542524433356,
"grad_norm": 2.7553778724216498,
"learning_rate": 5.588487512059171e-06,
"loss": 1.1307,
"mean_token_accuracy": 0.7270528495311737,
"num_tokens": 1146921844.0,
"step": 8810
},
{
"entropy": 1.890625,
"epoch": 0.4585152838427948,
"grad_norm": 2.914642064241222,
"learning_rate": 5.583127880801801e-06,
"loss": 1.1238,
"mean_token_accuracy": 0.7301478147506714,
"num_tokens": 1148223884.0,
"step": 8820
},
{
"entropy": 1.911328125,
"epoch": 0.459035142441256,
"grad_norm": 2.7206523215329823,
"learning_rate": 5.577768249544432e-06,
"loss": 1.1184,
"mean_token_accuracy": 0.7357638269662857,
"num_tokens": 1149524400.0,
"step": 8830
},
{
"entropy": 1.871875,
"epoch": 0.4595550010397172,
"grad_norm": 2.7080430608928845,
"learning_rate": 5.572408618287062e-06,
"loss": 1.1192,
"mean_token_accuracy": 0.7309505313634872,
"num_tokens": 1150825179.0,
"step": 8840
},
{
"entropy": 1.8875,
"epoch": 0.4600748596381784,
"grad_norm": 2.929240919771855,
"learning_rate": 5.567048987029693e-06,
"loss": 1.1065,
"mean_token_accuracy": 0.7326941132545471,
"num_tokens": 1152127125.0,
"step": 8850
},
{
"entropy": 1.881640625,
"epoch": 0.4605947182366396,
"grad_norm": 2.762146332727214,
"learning_rate": 5.561689355772323e-06,
"loss": 1.0515,
"mean_token_accuracy": 0.7456770986318588,
"num_tokens": 1153428232.0,
"step": 8860
},
{
"entropy": 1.86953125,
"epoch": 0.4611145768351009,
"grad_norm": 2.830565872449354,
"learning_rate": 5.556329724514954e-06,
"loss": 1.0945,
"mean_token_accuracy": 0.7357078909873962,
"num_tokens": 1154730739.0,
"step": 8870
},
{
"entropy": 1.894921875,
"epoch": 0.4616344354335621,
"grad_norm": 3.0608290083678624,
"learning_rate": 5.550970093257584e-06,
"loss": 1.1238,
"mean_token_accuracy": 0.730123245716095,
"num_tokens": 1156034786.0,
"step": 8880
},
{
"entropy": 1.88125,
"epoch": 0.4621542940320233,
"grad_norm": 2.9522854718613813,
"learning_rate": 5.545610462000215e-06,
"loss": 1.0802,
"mean_token_accuracy": 0.7388280689716339,
"num_tokens": 1157335508.0,
"step": 8890
},
{
"entropy": 1.869921875,
"epoch": 0.4626741526304845,
"grad_norm": 2.5157481234120045,
"learning_rate": 5.540250830742845e-06,
"loss": 1.1067,
"mean_token_accuracy": 0.7326318830251694,
"num_tokens": 1158632425.0,
"step": 8900
},
{
"epoch": 0.4626741526304845,
"eval_entropy": 1.855078125,
"eval_loss": 1.1329712867736816,
"eval_mean_token_accuracy": 0.7262461245059967,
"eval_num_tokens": 1158632425.0,
"eval_runtime": 27.041,
"eval_samples_per_second": 93.71,
"eval_steps_per_second": 1.479,
"step": 8900
},
{
"entropy": 1.8640625,
"epoch": 0.4631940112289457,
"grad_norm": 2.6736842476360314,
"learning_rate": 5.534891199485476e-06,
"loss": 1.0715,
"mean_token_accuracy": 0.7405211359262467,
"num_tokens": 1159931876.0,
"step": 8910
},
{
"entropy": 1.891015625,
"epoch": 0.4637138698274069,
"grad_norm": 2.6204246980470995,
"learning_rate": 5.529531568228106e-06,
"loss": 1.0709,
"mean_token_accuracy": 0.7426177740097046,
"num_tokens": 1161233904.0,
"step": 8920
},
{
"entropy": 1.88359375,
"epoch": 0.4642337284258682,
"grad_norm": 2.4793123476442207,
"learning_rate": 5.524171936970737e-06,
"loss": 1.1006,
"mean_token_accuracy": 0.7344273328781128,
"num_tokens": 1162534397.0,
"step": 8930
},
{
"entropy": 1.887890625,
"epoch": 0.4647535870243294,
"grad_norm": 3.0013299901940007,
"learning_rate": 5.5188123057133665e-06,
"loss": 1.107,
"mean_token_accuracy": 0.7322743833065033,
"num_tokens": 1163835838.0,
"step": 8940
},
{
"entropy": 1.8703125,
"epoch": 0.4652734456227906,
"grad_norm": 2.958448584934749,
"learning_rate": 5.513452674455998e-06,
"loss": 1.117,
"mean_token_accuracy": 0.7290244549512863,
"num_tokens": 1165136979.0,
"step": 8950
},
{
"entropy": 1.860546875,
"epoch": 0.4657933042212518,
"grad_norm": 3.1338024230772485,
"learning_rate": 5.508093043198628e-06,
"loss": 1.0488,
"mean_token_accuracy": 0.7449048221111297,
"num_tokens": 1166438340.0,
"step": 8960
},
{
"entropy": 1.8671875,
"epoch": 0.466313162819713,
"grad_norm": 2.8598713404109737,
"learning_rate": 5.502733411941259e-06,
"loss": 1.0754,
"mean_token_accuracy": 0.7409269511699677,
"num_tokens": 1167739604.0,
"step": 8970
},
{
"entropy": 1.8578125,
"epoch": 0.46683302141817423,
"grad_norm": 2.5923379460911113,
"learning_rate": 5.497373780683889e-06,
"loss": 1.0375,
"mean_token_accuracy": 0.7494511336088181,
"num_tokens": 1169042671.0,
"step": 8980
},
{
"entropy": 1.8921875,
"epoch": 0.4673528800166355,
"grad_norm": 2.872761847893193,
"learning_rate": 5.49201414942652e-06,
"loss": 1.1082,
"mean_token_accuracy": 0.7347300261259079,
"num_tokens": 1170342551.0,
"step": 8990
},
{
"entropy": 1.902734375,
"epoch": 0.4678727386150967,
"grad_norm": 2.6518040919643626,
"learning_rate": 5.48665451816915e-06,
"loss": 1.1317,
"mean_token_accuracy": 0.7319387316703796,
"num_tokens": 1171644146.0,
"step": 9000
},
{
"epoch": 0.4678727386150967,
"eval_entropy": 1.897265625,
"eval_loss": 1.1262856721878052,
"eval_mean_token_accuracy": 0.7278810277581215,
"eval_num_tokens": 1171644146.0,
"eval_runtime": 26.9972,
"eval_samples_per_second": 93.862,
"eval_steps_per_second": 1.482,
"step": 9000
},
{
"entropy": 1.8828125,
"epoch": 0.4683925972135579,
"grad_norm": 2.6024842930555816,
"learning_rate": 5.481294886911781e-06,
"loss": 1.0746,
"mean_token_accuracy": 0.7422453105449677,
"num_tokens": 1172944362.0,
"step": 9010
},
{
"entropy": 1.91171875,
"epoch": 0.4689124558120191,
"grad_norm": 2.6055210413050487,
"learning_rate": 5.475935255654411e-06,
"loss": 1.0627,
"mean_token_accuracy": 0.7425989866256714,
"num_tokens": 1174247171.0,
"step": 9020
},
{
"entropy": 1.8859375,
"epoch": 0.46943231441048033,
"grad_norm": 3.3472925388310286,
"learning_rate": 5.470575624397042e-06,
"loss": 1.0474,
"mean_token_accuracy": 0.7477692246437073,
"num_tokens": 1175547859.0,
"step": 9030
},
{
"entropy": 1.9046875,
"epoch": 0.46995217300894154,
"grad_norm": 2.596341432826683,
"learning_rate": 5.465215993139672e-06,
"loss": 1.1209,
"mean_token_accuracy": 0.7295790702104569,
"num_tokens": 1176849985.0,
"step": 9040
},
{
"entropy": 1.8796875,
"epoch": 0.4704720316074028,
"grad_norm": 2.650432695139415,
"learning_rate": 5.459856361882303e-06,
"loss": 1.0663,
"mean_token_accuracy": 0.7452544689178466,
"num_tokens": 1178150079.0,
"step": 9050
},
{
"entropy": 1.856640625,
"epoch": 0.470991890205864,
"grad_norm": 2.7252031979273337,
"learning_rate": 5.454496730624933e-06,
"loss": 1.081,
"mean_token_accuracy": 0.738840925693512,
"num_tokens": 1179451552.0,
"step": 9060
},
{
"entropy": 1.878515625,
"epoch": 0.4715117488043252,
"grad_norm": 2.4651616506532705,
"learning_rate": 5.449137099367564e-06,
"loss": 1.0988,
"mean_token_accuracy": 0.7372615188360214,
"num_tokens": 1180753893.0,
"step": 9070
},
{
"entropy": 1.8546875,
"epoch": 0.47203160740278644,
"grad_norm": 2.799394478480939,
"learning_rate": 5.443777468110194e-06,
"loss": 1.1006,
"mean_token_accuracy": 0.7333727449178695,
"num_tokens": 1182056889.0,
"step": 9080
},
{
"entropy": 1.86796875,
"epoch": 0.47255146600124764,
"grad_norm": 2.8407750768383853,
"learning_rate": 5.438417836852825e-06,
"loss": 1.0753,
"mean_token_accuracy": 0.7414343118667602,
"num_tokens": 1183359402.0,
"step": 9090
},
{
"entropy": 1.8765625,
"epoch": 0.47307132459970885,
"grad_norm": 3.1851996177342383,
"learning_rate": 5.433058205595455e-06,
"loss": 1.0779,
"mean_token_accuracy": 0.7409386664628983,
"num_tokens": 1184659575.0,
"step": 9100
},
{
"epoch": 0.47307132459970885,
"eval_entropy": 1.86484375,
"eval_loss": 1.1217617988586426,
"eval_mean_token_accuracy": 0.7289512321352959,
"eval_num_tokens": 1184659575.0,
"eval_runtime": 27.1043,
"eval_samples_per_second": 93.491,
"eval_steps_per_second": 1.476,
"step": 9100
},
{
"entropy": 1.925390625,
"epoch": 0.4735911831981701,
"grad_norm": 2.511777645932209,
"learning_rate": 5.427698574338086e-06,
"loss": 1.1498,
"mean_token_accuracy": 0.7260004043579101,
"num_tokens": 1185959919.0,
"step": 9110
},
{
"entropy": 1.840234375,
"epoch": 0.4741110417966313,
"grad_norm": 2.725272780875534,
"learning_rate": 5.422338943080716e-06,
"loss": 1.0642,
"mean_token_accuracy": 0.7436506986618042,
"num_tokens": 1187261935.0,
"step": 9120
},
{
"entropy": 1.858984375,
"epoch": 0.47463090039509254,
"grad_norm": 2.726446728658268,
"learning_rate": 5.416979311823347e-06,
"loss": 1.0531,
"mean_token_accuracy": 0.7445475876331329,
"num_tokens": 1188565113.0,
"step": 9130
},
{
"entropy": 1.871484375,
"epoch": 0.47515075899355375,
"grad_norm": 2.921464533869169,
"learning_rate": 5.4116196805659775e-06,
"loss": 1.0562,
"mean_token_accuracy": 0.7454096645116806,
"num_tokens": 1189865630.0,
"step": 9140
},
{
"entropy": 1.921484375,
"epoch": 0.47567061759201495,
"grad_norm": 2.4510269267415787,
"learning_rate": 5.4062600493086085e-06,
"loss": 1.0329,
"mean_token_accuracy": 0.7495617687702179,
"num_tokens": 1191166331.0,
"step": 9150
},
{
"entropy": 1.880859375,
"epoch": 0.47619047619047616,
"grad_norm": 2.6654803026565466,
"learning_rate": 5.400900418051238e-06,
"loss": 1.0686,
"mean_token_accuracy": 0.7424628496170044,
"num_tokens": 1192467713.0,
"step": 9160
},
{
"entropy": 1.903515625,
"epoch": 0.47671033478893743,
"grad_norm": 2.970102619465702,
"learning_rate": 5.3955407867938695e-06,
"loss": 1.1061,
"mean_token_accuracy": 0.7352486044168473,
"num_tokens": 1193767842.0,
"step": 9170
},
{
"entropy": 1.882421875,
"epoch": 0.47723019338739864,
"grad_norm": 2.8736673182975125,
"learning_rate": 5.390181155536499e-06,
"loss": 1.0923,
"mean_token_accuracy": 0.7389247953891754,
"num_tokens": 1195064385.0,
"step": 9180
},
{
"entropy": 1.90390625,
"epoch": 0.47775005198585985,
"grad_norm": 2.732090071016337,
"learning_rate": 5.38482152427913e-06,
"loss": 1.1055,
"mean_token_accuracy": 0.7357033848762512,
"num_tokens": 1196367298.0,
"step": 9190
},
{
"entropy": 1.8859375,
"epoch": 0.47826991058432106,
"grad_norm": 2.5871032344885516,
"learning_rate": 5.37946189302176e-06,
"loss": 1.0913,
"mean_token_accuracy": 0.7380078971385956,
"num_tokens": 1197666806.0,
"step": 9200
},
{
"epoch": 0.47826991058432106,
"eval_entropy": 1.8828125,
"eval_loss": 1.1190465688705444,
"eval_mean_token_accuracy": 0.7297381073236465,
"eval_num_tokens": 1197666806.0,
"eval_runtime": 27.084,
"eval_samples_per_second": 93.561,
"eval_steps_per_second": 1.477,
"step": 9200
},
{
"entropy": 1.86875,
"epoch": 0.47878976918278227,
"grad_norm": 2.487137936796052,
"learning_rate": 5.374102261764391e-06,
"loss": 1.1369,
"mean_token_accuracy": 0.7257491528987885,
"num_tokens": 1198969248.0,
"step": 9210
},
{
"entropy": 1.883984375,
"epoch": 0.4793096277812435,
"grad_norm": 2.449006786598289,
"learning_rate": 5.368742630507021e-06,
"loss": 1.1215,
"mean_token_accuracy": 0.7301113814115524,
"num_tokens": 1200269722.0,
"step": 9220
},
{
"entropy": 1.893359375,
"epoch": 0.47982948637970474,
"grad_norm": 3.1194020885143297,
"learning_rate": 5.363382999249652e-06,
"loss": 1.1058,
"mean_token_accuracy": 0.7333852171897888,
"num_tokens": 1201572300.0,
"step": 9230
},
{
"entropy": 1.866015625,
"epoch": 0.48034934497816595,
"grad_norm": 2.8078703793409274,
"learning_rate": 5.358023367992282e-06,
"loss": 1.0566,
"mean_token_accuracy": 0.7465455591678619,
"num_tokens": 1202874710.0,
"step": 9240
},
{
"entropy": 1.9109375,
"epoch": 0.48086920357662716,
"grad_norm": 2.6461084334311646,
"learning_rate": 5.352663736734913e-06,
"loss": 1.1423,
"mean_token_accuracy": 0.7270054429769516,
"num_tokens": 1204176845.0,
"step": 9250
},
{
"entropy": 1.9078125,
"epoch": 0.48138906217508837,
"grad_norm": 2.949440227380189,
"learning_rate": 5.347304105477543e-06,
"loss": 1.093,
"mean_token_accuracy": 0.7371095836162567,
"num_tokens": 1205478171.0,
"step": 9260
},
{
"entropy": 1.898046875,
"epoch": 0.4819089207735496,
"grad_norm": 2.5329907223626917,
"learning_rate": 5.341944474220174e-06,
"loss": 1.0446,
"mean_token_accuracy": 0.7460516124963761,
"num_tokens": 1206778037.0,
"step": 9270
},
{
"entropy": 1.907421875,
"epoch": 0.4824287793720108,
"grad_norm": 3.214593533549903,
"learning_rate": 5.336584842962804e-06,
"loss": 1.0724,
"mean_token_accuracy": 0.7417277246713638,
"num_tokens": 1208082624.0,
"step": 9280
},
{
"entropy": 1.865625,
"epoch": 0.48294863797047205,
"grad_norm": 2.4920086495061136,
"learning_rate": 5.331225211705435e-06,
"loss": 1.1136,
"mean_token_accuracy": 0.7318115472793579,
"num_tokens": 1209386167.0,
"step": 9290
},
{
"entropy": 1.900390625,
"epoch": 0.48346849656893326,
"grad_norm": 2.615640020338249,
"learning_rate": 5.325865580448065e-06,
"loss": 1.1382,
"mean_token_accuracy": 0.731728020310402,
"num_tokens": 1210689180.0,
"step": 9300
},
{
"epoch": 0.48346849656893326,
"eval_entropy": 1.9119140625,
"eval_loss": 1.1174408197402954,
"eval_mean_token_accuracy": 0.7298210740089417,
"eval_num_tokens": 1210689180.0,
"eval_runtime": 27.0117,
"eval_samples_per_second": 93.811,
"eval_steps_per_second": 1.481,
"step": 9300
},
{
"entropy": 1.876171875,
"epoch": 0.48398835516739447,
"grad_norm": 3.1727585568271937,
"learning_rate": 5.320505949190696e-06,
"loss": 1.0046,
"mean_token_accuracy": 0.7578696608543396,
"num_tokens": 1211991106.0,
"step": 9310
},
{
"entropy": 1.85234375,
"epoch": 0.4845082137658557,
"grad_norm": 2.498056543088736,
"learning_rate": 5.3151463179333265e-06,
"loss": 1.0976,
"mean_token_accuracy": 0.7353385120630265,
"num_tokens": 1213292550.0,
"step": 9320
},
{
"entropy": 1.842578125,
"epoch": 0.4850280723643169,
"grad_norm": 2.4674853739006912,
"learning_rate": 5.3097866866759574e-06,
"loss": 1.0514,
"mean_token_accuracy": 0.7426634758710862,
"num_tokens": 1214594641.0,
"step": 9330
},
{
"entropy": 1.87734375,
"epoch": 0.48554793096277815,
"grad_norm": 3.199665989736237,
"learning_rate": 5.3044270554185876e-06,
"loss": 1.133,
"mean_token_accuracy": 0.7311118483543396,
"num_tokens": 1215894610.0,
"step": 9340
},
{
"entropy": 1.859765625,
"epoch": 0.48606778956123936,
"grad_norm": 2.865781401386921,
"learning_rate": 5.2990674241612185e-06,
"loss": 1.0231,
"mean_token_accuracy": 0.7523659318685532,
"num_tokens": 1217197010.0,
"step": 9350
},
{
"entropy": 1.8609375,
"epoch": 0.48658764815970057,
"grad_norm": 2.603884259449824,
"learning_rate": 5.293707792903849e-06,
"loss": 1.0368,
"mean_token_accuracy": 0.7495386511087417,
"num_tokens": 1218499922.0,
"step": 9360
},
{
"entropy": 1.846484375,
"epoch": 0.4871075067581618,
"grad_norm": 2.5902357602513613,
"learning_rate": 5.28834816164648e-06,
"loss": 1.0369,
"mean_token_accuracy": 0.748224288225174,
"num_tokens": 1219804569.0,
"step": 9370
},
{
"entropy": 1.84453125,
"epoch": 0.487627365356623,
"grad_norm": 2.724503790563799,
"learning_rate": 5.282988530389109e-06,
"loss": 1.0658,
"mean_token_accuracy": 0.7434933692216873,
"num_tokens": 1221107494.0,
"step": 9380
},
{
"entropy": 1.880078125,
"epoch": 0.4881472239550842,
"grad_norm": 2.523483294150873,
"learning_rate": 5.277628899131741e-06,
"loss": 1.096,
"mean_token_accuracy": 0.7368319272994995,
"num_tokens": 1222408058.0,
"step": 9390
},
{
"entropy": 1.88515625,
"epoch": 0.48866708255354546,
"grad_norm": 2.9886656899939177,
"learning_rate": 5.27226926787437e-06,
"loss": 1.0961,
"mean_token_accuracy": 0.7347585290670395,
"num_tokens": 1223709586.0,
"step": 9400
},
{
"epoch": 0.48866708255354546,
"eval_entropy": 1.8501953125,
"eval_loss": 1.1141390800476074,
"eval_mean_token_accuracy": 0.7304085105657577,
"eval_num_tokens": 1223709586.0,
"eval_runtime": 26.9903,
"eval_samples_per_second": 93.886,
"eval_steps_per_second": 1.482,
"step": 9400
},
{
"entropy": 1.895703125,
"epoch": 0.48918694115200667,
"grad_norm": 2.7250589032843107,
"learning_rate": 5.266909636617001e-06,
"loss": 1.0937,
"mean_token_accuracy": 0.7398485511541366,
"num_tokens": 1225010617.0,
"step": 9410
},
{
"entropy": 1.882421875,
"epoch": 0.4897067997504679,
"grad_norm": 2.9511712757538513,
"learning_rate": 5.261550005359631e-06,
"loss": 1.0609,
"mean_token_accuracy": 0.7452236950397492,
"num_tokens": 1226312907.0,
"step": 9420
},
{
"entropy": 1.898828125,
"epoch": 0.4902266583489291,
"grad_norm": 2.7184140513705017,
"learning_rate": 5.256190374102262e-06,
"loss": 1.1269,
"mean_token_accuracy": 0.7291383266448974,
"num_tokens": 1227615037.0,
"step": 9430
},
{
"entropy": 1.880859375,
"epoch": 0.4907465169473903,
"grad_norm": 2.604642101326909,
"learning_rate": 5.250830742844892e-06,
"loss": 1.0531,
"mean_token_accuracy": 0.7482659310102463,
"num_tokens": 1228916934.0,
"step": 9440
},
{
"entropy": 1.89296875,
"epoch": 0.4912663755458515,
"grad_norm": 2.987854990383074,
"learning_rate": 5.245471111587523e-06,
"loss": 1.0964,
"mean_token_accuracy": 0.7370205968618393,
"num_tokens": 1230219002.0,
"step": 9450
},
{
"entropy": 1.869921875,
"epoch": 0.4917862341443128,
"grad_norm": 2.5353102470459885,
"learning_rate": 5.240111480330153e-06,
"loss": 1.086,
"mean_token_accuracy": 0.7366574048995972,
"num_tokens": 1231519318.0,
"step": 9460
},
{
"entropy": 1.834375,
"epoch": 0.492306092742774,
"grad_norm": 3.139892664999347,
"learning_rate": 5.234751849072784e-06,
"loss": 1.0747,
"mean_token_accuracy": 0.7405396401882172,
"num_tokens": 1232820577.0,
"step": 9470
},
{
"entropy": 1.890625,
"epoch": 0.4928259513412352,
"grad_norm": 2.522879643327786,
"learning_rate": 5.229392217815414e-06,
"loss": 1.007,
"mean_token_accuracy": 0.7557831406593323,
"num_tokens": 1234123295.0,
"step": 9480
},
{
"entropy": 1.915625,
"epoch": 0.4933458099396964,
"grad_norm": 2.7311046312791505,
"learning_rate": 5.224032586558045e-06,
"loss": 1.0584,
"mean_token_accuracy": 0.7465008527040482,
"num_tokens": 1235423531.0,
"step": 9490
},
{
"entropy": 1.885546875,
"epoch": 0.4938656685381576,
"grad_norm": 2.9940432896883493,
"learning_rate": 5.2186729553006755e-06,
"loss": 1.0364,
"mean_token_accuracy": 0.7498366504907608,
"num_tokens": 1236727252.0,
"step": 9500
},
{
"epoch": 0.4938656685381576,
"eval_entropy": 1.8802734375,
"eval_loss": 1.1127192974090576,
"eval_mean_token_accuracy": 0.731123349070549,
"eval_num_tokens": 1236727252.0,
"eval_runtime": 27.0248,
"eval_samples_per_second": 93.766,
"eval_steps_per_second": 1.48,
"step": 9500
},
{
"entropy": 1.9140625,
"epoch": 0.4943855271366188,
"grad_norm": 3.1050786275101347,
"learning_rate": 5.2133133240433064e-06,
"loss": 1.0779,
"mean_token_accuracy": 0.7422169059514999,
"num_tokens": 1238029181.0,
"step": 9510
},
{
"entropy": 1.863671875,
"epoch": 0.4949053857350801,
"grad_norm": 2.9894852842088278,
"learning_rate": 5.2079536927859366e-06,
"loss": 1.0606,
"mean_token_accuracy": 0.7439016401767731,
"num_tokens": 1239330975.0,
"step": 9520
},
{
"entropy": 1.8890625,
"epoch": 0.4954252443335413,
"grad_norm": 2.8218744991433593,
"learning_rate": 5.2025940615285675e-06,
"loss": 1.0649,
"mean_token_accuracy": 0.7414388447999954,
"num_tokens": 1240633213.0,
"step": 9530
},
{
"entropy": 1.883984375,
"epoch": 0.4959451029320025,
"grad_norm": 2.645308197533684,
"learning_rate": 5.197234430271198e-06,
"loss": 1.0458,
"mean_token_accuracy": 0.7496142297983169,
"num_tokens": 1241936636.0,
"step": 9540
},
{
"entropy": 1.871875,
"epoch": 0.4964649615304637,
"grad_norm": 2.831323307487516,
"learning_rate": 5.191874799013829e-06,
"loss": 1.0093,
"mean_token_accuracy": 0.7588496834039689,
"num_tokens": 1243240058.0,
"step": 9550
},
{
"entropy": 1.886328125,
"epoch": 0.4969848201289249,
"grad_norm": 2.7229572866215337,
"learning_rate": 5.186515167756459e-06,
"loss": 1.1247,
"mean_token_accuracy": 0.7299771398305893,
"num_tokens": 1244542996.0,
"step": 9560
},
{
"entropy": 1.854296875,
"epoch": 0.49750467872738613,
"grad_norm": 2.7838833448556644,
"learning_rate": 5.18115553649909e-06,
"loss": 1.1111,
"mean_token_accuracy": 0.7318018555641175,
"num_tokens": 1245845768.0,
"step": 9570
},
{
"entropy": 1.873046875,
"epoch": 0.4980245373258474,
"grad_norm": 2.770661899287635,
"learning_rate": 5.17579590524172e-06,
"loss": 1.0604,
"mean_token_accuracy": 0.7468955099582673,
"num_tokens": 1247148961.0,
"step": 9580
},
{
"entropy": 1.859375,
"epoch": 0.4985443959243086,
"grad_norm": 2.764986123488908,
"learning_rate": 5.170436273984351e-06,
"loss": 1.0673,
"mean_token_accuracy": 0.7396160453557968,
"num_tokens": 1248445965.0,
"step": 9590
},
{
"entropy": 1.845703125,
"epoch": 0.4990642545227698,
"grad_norm": 2.3040032526006557,
"learning_rate": 5.16507664272698e-06,
"loss": 1.0638,
"mean_token_accuracy": 0.7420497059822082,
"num_tokens": 1249749211.0,
"step": 9600
},
{
"epoch": 0.4990642545227698,
"eval_entropy": 1.841796875,
"eval_loss": 1.1083897352218628,
"eval_mean_token_accuracy": 0.7318060398101807,
"eval_num_tokens": 1249749211.0,
"eval_runtime": 27.0344,
"eval_samples_per_second": 93.732,
"eval_steps_per_second": 1.48,
"step": 9600
},
{
"entropy": 1.843359375,
"epoch": 0.499584113121231,
"grad_norm": 2.7118235200022327,
"learning_rate": 5.159717011469612e-06,
"loss": 1.1534,
"mean_token_accuracy": 0.7262647598981857,
"num_tokens": 1251049326.0,
"step": 9610
},
{
"entropy": 1.907421875,
"epoch": 0.5001039717196922,
"grad_norm": 2.44533231243564,
"learning_rate": 5.154357380212241e-06,
"loss": 1.133,
"mean_token_accuracy": 0.7246289312839508,
"num_tokens": 1252351638.0,
"step": 9620
},
{
"entropy": 1.860546875,
"epoch": 0.5006238303181535,
"grad_norm": 2.383639218943792,
"learning_rate": 5.148997748954873e-06,
"loss": 1.0438,
"mean_token_accuracy": 0.7470720946788788,
"num_tokens": 1253655038.0,
"step": 9630
},
{
"entropy": 1.849609375,
"epoch": 0.5011436889166146,
"grad_norm": 2.851578889773281,
"learning_rate": 5.143638117697502e-06,
"loss": 1.0381,
"mean_token_accuracy": 0.7484759896993637,
"num_tokens": 1254956996.0,
"step": 9640
},
{
"entropy": 1.88515625,
"epoch": 0.5016635475150759,
"grad_norm": 2.9561742350464644,
"learning_rate": 5.138278486440133e-06,
"loss": 1.1235,
"mean_token_accuracy": 0.7301020979881286,
"num_tokens": 1256260967.0,
"step": 9650
},
{
"entropy": 1.91015625,
"epoch": 0.5021834061135371,
"grad_norm": 2.2441057943361558,
"learning_rate": 5.132918855182763e-06,
"loss": 1.1032,
"mean_token_accuracy": 0.7350785136222839,
"num_tokens": 1257561819.0,
"step": 9660
},
{
"entropy": 1.9109375,
"epoch": 0.5027032647119983,
"grad_norm": 2.7342485363425557,
"learning_rate": 5.127559223925394e-06,
"loss": 1.0742,
"mean_token_accuracy": 0.7422219038009643,
"num_tokens": 1258864212.0,
"step": 9670
},
{
"entropy": 1.902734375,
"epoch": 0.5032231233104596,
"grad_norm": 2.6246106206055297,
"learning_rate": 5.1221995926680245e-06,
"loss": 1.1009,
"mean_token_accuracy": 0.7352589666843414,
"num_tokens": 1260166950.0,
"step": 9680
},
{
"entropy": 1.90234375,
"epoch": 0.5037429819089208,
"grad_norm": 2.9340152792557594,
"learning_rate": 5.1168399614106554e-06,
"loss": 1.1056,
"mean_token_accuracy": 0.7340193420648575,
"num_tokens": 1261467835.0,
"step": 9690
},
{
"entropy": 1.867578125,
"epoch": 0.504262840507382,
"grad_norm": 3.1639373469531265,
"learning_rate": 5.1114803301532855e-06,
"loss": 1.0344,
"mean_token_accuracy": 0.7483770072460174,
"num_tokens": 1262769973.0,
"step": 9700
},
{
"epoch": 0.504262840507382,
"eval_entropy": 1.845703125,
"eval_loss": 1.1049330234527588,
"eval_mean_token_accuracy": 0.7331295818090439,
"eval_num_tokens": 1262769973.0,
"eval_runtime": 26.9874,
"eval_samples_per_second": 93.896,
"eval_steps_per_second": 1.482,
"step": 9700
},
{
"entropy": 1.8671875,
"epoch": 0.5047826991058432,
"grad_norm": 2.891930996347582,
"learning_rate": 5.1061206988959165e-06,
"loss": 1.0821,
"mean_token_accuracy": 0.738443922996521,
"num_tokens": 1264073879.0,
"step": 9710
},
{
"entropy": 1.896875,
"epoch": 0.5053025577043044,
"grad_norm": 2.895496055777251,
"learning_rate": 5.100761067638547e-06,
"loss": 1.0572,
"mean_token_accuracy": 0.7475954711437225,
"num_tokens": 1265375579.0,
"step": 9720
},
{
"entropy": 1.927734375,
"epoch": 0.5058224163027657,
"grad_norm": 2.67493303806759,
"learning_rate": 5.095401436381178e-06,
"loss": 1.0895,
"mean_token_accuracy": 0.7376600503921509,
"num_tokens": 1266676298.0,
"step": 9730
},
{
"entropy": 1.875390625,
"epoch": 0.5063422749012269,
"grad_norm": 2.9860280137332595,
"learning_rate": 5.090041805123808e-06,
"loss": 1.0752,
"mean_token_accuracy": 0.7407987654209137,
"num_tokens": 1267979377.0,
"step": 9740
},
{
"entropy": 1.89296875,
"epoch": 0.5068621334996881,
"grad_norm": 2.8619496127980995,
"learning_rate": 5.084682173866439e-06,
"loss": 1.0907,
"mean_token_accuracy": 0.7380967974662781,
"num_tokens": 1269281351.0,
"step": 9750
},
{
"entropy": 1.894140625,
"epoch": 0.5073819920981493,
"grad_norm": 3.1755048134341592,
"learning_rate": 5.079322542609069e-06,
"loss": 1.0625,
"mean_token_accuracy": 0.7458548188209534,
"num_tokens": 1270584869.0,
"step": 9760
},
{
"entropy": 1.874609375,
"epoch": 0.5079018506966105,
"grad_norm": 3.4632331937424508,
"learning_rate": 5.0739629113517e-06,
"loss": 1.04,
"mean_token_accuracy": 0.7478562742471695,
"num_tokens": 1271888812.0,
"step": 9770
},
{
"entropy": 1.859375,
"epoch": 0.5084217092950717,
"grad_norm": 2.91049529866568,
"learning_rate": 5.06860328009433e-06,
"loss": 1.0597,
"mean_token_accuracy": 0.7440079778432847,
"num_tokens": 1273190002.0,
"step": 9780
},
{
"entropy": 1.883984375,
"epoch": 0.508941567893533,
"grad_norm": 2.437818093980988,
"learning_rate": 5.063243648836961e-06,
"loss": 1.0703,
"mean_token_accuracy": 0.7416574239730835,
"num_tokens": 1274494495.0,
"step": 9790
},
{
"entropy": 1.882421875,
"epoch": 0.5094614264919942,
"grad_norm": 2.9343610327432135,
"learning_rate": 5.057884017579591e-06,
"loss": 1.1067,
"mean_token_accuracy": 0.7345606207847595,
"num_tokens": 1275795937.0,
"step": 9800
},
{
"epoch": 0.5094614264919942,
"eval_entropy": 1.8763671875,
"eval_loss": 1.0978033542633057,
"eval_mean_token_accuracy": 0.7342393159866333,
"eval_num_tokens": 1275795937.0,
"eval_runtime": 27.0839,
"eval_samples_per_second": 93.561,
"eval_steps_per_second": 1.477,
"step": 9800
},
{
"entropy": 1.88125,
"epoch": 0.5099812850904554,
"grad_norm": 2.432489840990661,
"learning_rate": 5.052524386322222e-06,
"loss": 1.075,
"mean_token_accuracy": 0.7435490131378174,
"num_tokens": 1277099784.0,
"step": 9810
},
{
"entropy": 1.91328125,
"epoch": 0.5105011436889166,
"grad_norm": 2.8053196219949967,
"learning_rate": 5.047164755064852e-06,
"loss": 1.1114,
"mean_token_accuracy": 0.7344515919685364,
"num_tokens": 1278400970.0,
"step": 9820
},
{
"entropy": 1.855078125,
"epoch": 0.5110210022873778,
"grad_norm": 2.873474571557467,
"learning_rate": 5.041805123807483e-06,
"loss": 1.0867,
"mean_token_accuracy": 0.7396327584981919,
"num_tokens": 1279702085.0,
"step": 9830
},
{
"entropy": 1.869140625,
"epoch": 0.511540860885839,
"grad_norm": 2.5771791715952195,
"learning_rate": 5.036445492550112e-06,
"loss": 1.145,
"mean_token_accuracy": 0.7267734229564666,
"num_tokens": 1281002314.0,
"step": 9840
},
{
"entropy": 1.82421875,
"epoch": 0.5120607194843003,
"grad_norm": 3.1087838984742957,
"learning_rate": 5.031085861292744e-06,
"loss": 1.0064,
"mean_token_accuracy": 0.7582368403673172,
"num_tokens": 1282305251.0,
"step": 9850
},
{
"entropy": 1.846875,
"epoch": 0.5125805780827615,
"grad_norm": 2.881228227821757,
"learning_rate": 5.0257262300353734e-06,
"loss": 1.0666,
"mean_token_accuracy": 0.7460797697305679,
"num_tokens": 1283604572.0,
"step": 9860
},
{
"entropy": 1.83359375,
"epoch": 0.5131004366812227,
"grad_norm": 2.4090978026742413,
"learning_rate": 5.020366598778004e-06,
"loss": 1.119,
"mean_token_accuracy": 0.7320625841617584,
"num_tokens": 1284907573.0,
"step": 9870
},
{
"entropy": 1.859765625,
"epoch": 0.5136202952796839,
"grad_norm": 2.6679114307463454,
"learning_rate": 5.0150069675206345e-06,
"loss": 1.0798,
"mean_token_accuracy": 0.7406738579273224,
"num_tokens": 1286211693.0,
"step": 9880
},
{
"entropy": 1.85546875,
"epoch": 0.5141401538781452,
"grad_norm": 2.666136092955274,
"learning_rate": 5.0096473362632655e-06,
"loss": 1.0264,
"mean_token_accuracy": 0.7535203576087952,
"num_tokens": 1287514717.0,
"step": 9890
},
{
"entropy": 1.880078125,
"epoch": 0.5146600124766063,
"grad_norm": 2.752361418590538,
"learning_rate": 5.004287705005896e-06,
"loss": 1.0657,
"mean_token_accuracy": 0.7446679979562759,
"num_tokens": 1288816123.0,
"step": 9900
},
{
"epoch": 0.5146600124766063,
"eval_entropy": 1.8322265625,
"eval_loss": 1.0906286239624023,
"eval_mean_token_accuracy": 0.7360314100980758,
"eval_num_tokens": 1288816123.0,
"eval_runtime": 27.0287,
"eval_samples_per_second": 93.752,
"eval_steps_per_second": 1.48,
"step": 9900
},
{
"entropy": 1.82890625,
"epoch": 0.5151798710750676,
"grad_norm": 2.5583162827556873,
"learning_rate": 4.998928073748527e-06,
"loss": 1.0959,
"mean_token_accuracy": 0.7381640583276748,
"num_tokens": 1290117761.0,
"step": 9910
},
{
"entropy": 1.847265625,
"epoch": 0.5156997296735288,
"grad_norm": 2.8628246291707877,
"learning_rate": 4.993568442491157e-06,
"loss": 1.0548,
"mean_token_accuracy": 0.7430138289928436,
"num_tokens": 1291418870.0,
"step": 9920
},
{
"entropy": 1.847265625,
"epoch": 0.51621958827199,
"grad_norm": 2.6007106215067273,
"learning_rate": 4.988208811233788e-06,
"loss": 1.0726,
"mean_token_accuracy": 0.7418247818946838,
"num_tokens": 1292720638.0,
"step": 9930
},
{
"entropy": 1.852734375,
"epoch": 0.5167394468704513,
"grad_norm": 2.869194480121915,
"learning_rate": 4.982849179976418e-06,
"loss": 1.0688,
"mean_token_accuracy": 0.7421867460012436,
"num_tokens": 1294022245.0,
"step": 9940
},
{
"entropy": 1.845703125,
"epoch": 0.5172593054689124,
"grad_norm": 2.9575491841071906,
"learning_rate": 4.977489548719048e-06,
"loss": 1.1006,
"mean_token_accuracy": 0.735472172498703,
"num_tokens": 1295323620.0,
"step": 9950
},
{
"entropy": 1.8515625,
"epoch": 0.5177791640673737,
"grad_norm": 2.7747101306407567,
"learning_rate": 4.972129917461679e-06,
"loss": 1.02,
"mean_token_accuracy": 0.7543217658996582,
"num_tokens": 1296626366.0,
"step": 9960
},
{
"entropy": 1.849609375,
"epoch": 0.5182990226658349,
"grad_norm": 2.8927682031908035,
"learning_rate": 4.966770286204309e-06,
"loss": 1.0948,
"mean_token_accuracy": 0.7377075403928757,
"num_tokens": 1297930259.0,
"step": 9970
},
{
"entropy": 1.8359375,
"epoch": 0.5188188812642961,
"grad_norm": 3.105845062440033,
"learning_rate": 4.96141065494694e-06,
"loss": 1.0181,
"mean_token_accuracy": 0.7536915183067322,
"num_tokens": 1299232032.0,
"step": 9980
},
{
"entropy": 1.82265625,
"epoch": 0.5193387398627574,
"grad_norm": 2.6286488626612874,
"learning_rate": 4.95605102368957e-06,
"loss": 1.0514,
"mean_token_accuracy": 0.7464523851871491,
"num_tokens": 1300534199.0,
"step": 9990
},
{
"entropy": 1.83046875,
"epoch": 0.5198585984612185,
"grad_norm": 2.9426611396596885,
"learning_rate": 4.950691392432201e-06,
"loss": 1.0198,
"mean_token_accuracy": 0.7532619804143905,
"num_tokens": 1301834740.0,
"step": 10000
},
{
"epoch": 0.5198585984612185,
"eval_entropy": 1.8083984375,
"eval_loss": 1.085893154144287,
"eval_mean_token_accuracy": 0.7377725154161453,
"eval_num_tokens": 1301834740.0,
"eval_runtime": 27.0011,
"eval_samples_per_second": 93.848,
"eval_steps_per_second": 1.481,
"step": 10000
},
{
"entropy": 1.83984375,
"epoch": 0.5203784570596798,
"grad_norm": 2.8335578906425734,
"learning_rate": 4.945331761174831e-06,
"loss": 1.0457,
"mean_token_accuracy": 0.7479885935783386,
"num_tokens": 1303137908.0,
"step": 10010
},
{
"entropy": 1.845703125,
"epoch": 0.5208983156581409,
"grad_norm": 2.657401639254849,
"learning_rate": 4.939972129917462e-06,
"loss": 1.0723,
"mean_token_accuracy": 0.7418494105339051,
"num_tokens": 1304439770.0,
"step": 10020
},
{
"entropy": 1.8375,
"epoch": 0.5214181742566022,
"grad_norm": 2.6658244111598974,
"learning_rate": 4.934612498660092e-06,
"loss": 1.0651,
"mean_token_accuracy": 0.7436996906995773,
"num_tokens": 1305742344.0,
"step": 10030
},
{
"entropy": 1.856640625,
"epoch": 0.5219380328550635,
"grad_norm": 3.067380445558592,
"learning_rate": 4.929252867402723e-06,
"loss": 1.018,
"mean_token_accuracy": 0.7566038787364959,
"num_tokens": 1307042635.0,
"step": 10040
},
{
"entropy": 1.809375,
"epoch": 0.5224578914535246,
"grad_norm": 3.601246640202431,
"learning_rate": 4.923893236145353e-06,
"loss": 1.066,
"mean_token_accuracy": 0.7429820388555527,
"num_tokens": 1308346693.0,
"step": 10050
},
{
"entropy": 1.849609375,
"epoch": 0.5229777500519859,
"grad_norm": 2.650136793586911,
"learning_rate": 4.9185336048879835e-06,
"loss": 1.0378,
"mean_token_accuracy": 0.7523560136556625,
"num_tokens": 1309648865.0,
"step": 10060
},
{
"entropy": 1.828125,
"epoch": 0.523497608650447,
"grad_norm": 2.915699056393652,
"learning_rate": 4.9131739736306145e-06,
"loss": 1.0299,
"mean_token_accuracy": 0.7548382937908172,
"num_tokens": 1310950461.0,
"step": 10070
},
{
"entropy": 1.863671875,
"epoch": 0.5240174672489083,
"grad_norm": 3.0883738630035116,
"learning_rate": 4.907814342373245e-06,
"loss": 1.0787,
"mean_token_accuracy": 0.7414929419755936,
"num_tokens": 1312253062.0,
"step": 10080
},
{
"entropy": 1.84296875,
"epoch": 0.5245373258473696,
"grad_norm": 3.393790152369483,
"learning_rate": 4.902454711115876e-06,
"loss": 1.0567,
"mean_token_accuracy": 0.7448720216751099,
"num_tokens": 1313554979.0,
"step": 10090
},
{
"entropy": 1.836328125,
"epoch": 0.5250571844458307,
"grad_norm": 3.0216851076651206,
"learning_rate": 4.897095079858506e-06,
"loss": 0.9675,
"mean_token_accuracy": 0.7689126849174499,
"num_tokens": 1314855764.0,
"step": 10100
},
{
"epoch": 0.5250571844458307,
"eval_entropy": 1.839453125,
"eval_loss": 1.0819180011749268,
"eval_mean_token_accuracy": 0.7386129289865494,
"eval_num_tokens": 1314855764.0,
"eval_runtime": 27.0139,
"eval_samples_per_second": 93.804,
"eval_steps_per_second": 1.481,
"step": 10100
},
{
"entropy": 1.87109375,
"epoch": 0.525577043044292,
"grad_norm": 2.494358170664954,
"learning_rate": 4.891735448601137e-06,
"loss": 1.0863,
"mean_token_accuracy": 0.7397398114204407,
"num_tokens": 1316158991.0,
"step": 10110
},
{
"entropy": 1.874609375,
"epoch": 0.5260969016427531,
"grad_norm": 2.553270755986068,
"learning_rate": 4.886375817343767e-06,
"loss": 1.0652,
"mean_token_accuracy": 0.7432066857814789,
"num_tokens": 1317460394.0,
"step": 10120
},
{
"entropy": 1.87265625,
"epoch": 0.5266167602412144,
"grad_norm": 2.762950626993889,
"learning_rate": 4.881016186086398e-06,
"loss": 1.0526,
"mean_token_accuracy": 0.7484016954898834,
"num_tokens": 1318763559.0,
"step": 10130
},
{
"entropy": 1.86015625,
"epoch": 0.5271366188396757,
"grad_norm": 3.2719973063033785,
"learning_rate": 4.875656554829028e-06,
"loss": 1.0952,
"mean_token_accuracy": 0.7338238030672073,
"num_tokens": 1320066023.0,
"step": 10140
},
{
"entropy": 1.86015625,
"epoch": 0.5276564774381368,
"grad_norm": 2.824325103431716,
"learning_rate": 4.870296923571659e-06,
"loss": 1.0231,
"mean_token_accuracy": 0.7528125405311584,
"num_tokens": 1321367579.0,
"step": 10150
},
{
"entropy": 1.842578125,
"epoch": 0.5281763360365981,
"grad_norm": 2.8869643794269524,
"learning_rate": 4.864937292314289e-06,
"loss": 1.0777,
"mean_token_accuracy": 0.7404178589582443,
"num_tokens": 1322670566.0,
"step": 10160
},
{
"entropy": 1.852734375,
"epoch": 0.5286961946350592,
"grad_norm": 2.6498991787052697,
"learning_rate": 4.859577661056919e-06,
"loss": 1.0603,
"mean_token_accuracy": 0.7458404392004013,
"num_tokens": 1323971386.0,
"step": 10170
},
{
"entropy": 1.86015625,
"epoch": 0.5292160532335205,
"grad_norm": 2.656070707741719,
"learning_rate": 4.85421802979955e-06,
"loss": 1.136,
"mean_token_accuracy": 0.7285775899887085,
"num_tokens": 1325272922.0,
"step": 10180
},
{
"entropy": 1.85859375,
"epoch": 0.5297359118319817,
"grad_norm": 3.2806475879201478,
"learning_rate": 4.84885839854218e-06,
"loss": 1.0587,
"mean_token_accuracy": 0.745150575041771,
"num_tokens": 1326574601.0,
"step": 10190
},
{
"entropy": 1.8375,
"epoch": 0.5302557704304429,
"grad_norm": 2.781384718565076,
"learning_rate": 4.843498767284811e-06,
"loss": 1.0874,
"mean_token_accuracy": 0.7381851553916932,
"num_tokens": 1327875001.0,
"step": 10200
},
{
"epoch": 0.5302557704304429,
"eval_entropy": 1.814453125,
"eval_loss": 1.0820246934890747,
"eval_mean_token_accuracy": 0.7382693722844124,
"eval_num_tokens": 1327875001.0,
"eval_runtime": 27.0813,
"eval_samples_per_second": 93.57,
"eval_steps_per_second": 1.477,
"step": 10200
},
{
"entropy": 1.8140625,
"epoch": 0.5307756290289042,
"grad_norm": 3.0959926168355114,
"learning_rate": 4.838139136027441e-06,
"loss": 1.0858,
"mean_token_accuracy": 0.739691111445427,
"num_tokens": 1329178538.0,
"step": 10210
},
{
"entropy": 1.849609375,
"epoch": 0.5312954876273653,
"grad_norm": 3.0761125875740336,
"learning_rate": 4.832779504770072e-06,
"loss": 1.045,
"mean_token_accuracy": 0.7482000052928924,
"num_tokens": 1330481256.0,
"step": 10220
},
{
"entropy": 1.853515625,
"epoch": 0.5318153462258266,
"grad_norm": 2.8355309744394934,
"learning_rate": 4.827419873512702e-06,
"loss": 1.0749,
"mean_token_accuracy": 0.7416262596845626,
"num_tokens": 1331783632.0,
"step": 10230
},
{
"entropy": 1.80703125,
"epoch": 0.5323352048242878,
"grad_norm": 2.614168182982355,
"learning_rate": 4.822060242255333e-06,
"loss": 0.9735,
"mean_token_accuracy": 0.7636407673358917,
"num_tokens": 1333084718.0,
"step": 10240
},
{
"entropy": 1.82890625,
"epoch": 0.532855063422749,
"grad_norm": 2.8533411369687838,
"learning_rate": 4.8167006109979635e-06,
"loss": 1.1303,
"mean_token_accuracy": 0.7290268927812577,
"num_tokens": 1334384337.0,
"step": 10250
},
{
"entropy": 1.826171875,
"epoch": 0.5333749220212103,
"grad_norm": 2.616397930404055,
"learning_rate": 4.8113409797405945e-06,
"loss": 1.0675,
"mean_token_accuracy": 0.7416959553956985,
"num_tokens": 1335685552.0,
"step": 10260
},
{
"entropy": 1.819921875,
"epoch": 0.5338947806196714,
"grad_norm": 2.5511515610300934,
"learning_rate": 4.805981348483225e-06,
"loss": 1.0789,
"mean_token_accuracy": 0.7388208746910095,
"num_tokens": 1336988709.0,
"step": 10270
},
{
"entropy": 1.829296875,
"epoch": 0.5344146392181327,
"grad_norm": 2.9464346192307356,
"learning_rate": 4.800621717225855e-06,
"loss": 1.0598,
"mean_token_accuracy": 0.745069146156311,
"num_tokens": 1338287957.0,
"step": 10280
},
{
"entropy": 1.803515625,
"epoch": 0.5349344978165939,
"grad_norm": 2.5426565168312267,
"learning_rate": 4.795262085968486e-06,
"loss": 1.0627,
"mean_token_accuracy": 0.7460123866796493,
"num_tokens": 1339589044.0,
"step": 10290
},
{
"entropy": 1.838671875,
"epoch": 0.5354543564150551,
"grad_norm": 2.634055735547378,
"learning_rate": 4.789902454711116e-06,
"loss": 1.1011,
"mean_token_accuracy": 0.735072934627533,
"num_tokens": 1340893302.0,
"step": 10300
},
{
"epoch": 0.5354543564150551,
"eval_entropy": 1.83046875,
"eval_loss": 1.0785558223724365,
"eval_mean_token_accuracy": 0.739104999601841,
"eval_num_tokens": 1340893302.0,
"eval_runtime": 27.0304,
"eval_samples_per_second": 93.746,
"eval_steps_per_second": 1.48,
"step": 10300
},
{
"entropy": 1.825390625,
"epoch": 0.5359742150135163,
"grad_norm": 2.602447787417166,
"learning_rate": 4.784542823453747e-06,
"loss": 1.0811,
"mean_token_accuracy": 0.7381346583366394,
"num_tokens": 1342194017.0,
"step": 10310
},
{
"entropy": 1.821484375,
"epoch": 0.5364940736119775,
"grad_norm": 2.705651172703253,
"learning_rate": 4.779183192196377e-06,
"loss": 1.0524,
"mean_token_accuracy": 0.7468776375055313,
"num_tokens": 1343495641.0,
"step": 10320
},
{
"entropy": 1.857421875,
"epoch": 0.5370139322104388,
"grad_norm": 3.0371110828424253,
"learning_rate": 4.773823560939008e-06,
"loss": 1.0784,
"mean_token_accuracy": 0.7395465075969696,
"num_tokens": 1344796750.0,
"step": 10330
},
{
"entropy": 1.8484375,
"epoch": 0.5375337908089,
"grad_norm": 3.3349111592236222,
"learning_rate": 4.768463929681638e-06,
"loss": 1.0513,
"mean_token_accuracy": 0.7472383499145507,
"num_tokens": 1346097223.0,
"step": 10340
},
{
"entropy": 1.846875,
"epoch": 0.5380536494073612,
"grad_norm": 3.1008549706495234,
"learning_rate": 4.763104298424269e-06,
"loss": 1.078,
"mean_token_accuracy": 0.7430384755134583,
"num_tokens": 1347400853.0,
"step": 10350
},
{
"entropy": 1.837109375,
"epoch": 0.5385735080058224,
"grad_norm": 2.789665238314933,
"learning_rate": 4.757744667166899e-06,
"loss": 1.0506,
"mean_token_accuracy": 0.7464760959148407,
"num_tokens": 1348702504.0,
"step": 10360
},
{
"entropy": 1.85703125,
"epoch": 0.5390933666042836,
"grad_norm": 3.0040725446967755,
"learning_rate": 4.75238503590953e-06,
"loss": 1.022,
"mean_token_accuracy": 0.7516975790262223,
"num_tokens": 1350006526.0,
"step": 10370
},
{
"entropy": 1.8171875,
"epoch": 0.5396132252027449,
"grad_norm": 2.6381649853514446,
"learning_rate": 4.74702540465216e-06,
"loss": 1.0261,
"mean_token_accuracy": 0.7523701399564743,
"num_tokens": 1351308569.0,
"step": 10380
},
{
"entropy": 1.8546875,
"epoch": 0.5401330838012061,
"grad_norm": 2.9957058075034975,
"learning_rate": 4.741665773394791e-06,
"loss": 1.082,
"mean_token_accuracy": 0.7403950065374374,
"num_tokens": 1352609379.0,
"step": 10390
},
{
"entropy": 1.8359375,
"epoch": 0.5406529423996673,
"grad_norm": 2.6036888803224474,
"learning_rate": 4.736306142137421e-06,
"loss": 1.0148,
"mean_token_accuracy": 0.7546595484018326,
"num_tokens": 1353912504.0,
"step": 10400
},
{
"epoch": 0.5406529423996673,
"eval_entropy": 1.8283203125,
"eval_loss": 1.0741336345672607,
"eval_mean_token_accuracy": 0.7405048072338104,
"eval_num_tokens": 1353912504.0,
"eval_runtime": 27.0728,
"eval_samples_per_second": 93.6,
"eval_steps_per_second": 1.477,
"step": 10400
},
{
"entropy": 1.847265625,
"epoch": 0.5411728009981285,
"grad_norm": 3.057139335981563,
"learning_rate": 4.730946510880051e-06,
"loss": 1.0004,
"mean_token_accuracy": 0.7604137390851975,
"num_tokens": 1355212032.0,
"step": 10410
},
{
"entropy": 1.801171875,
"epoch": 0.5416926595965897,
"grad_norm": 2.8459662849797893,
"learning_rate": 4.725586879622682e-06,
"loss": 1.0204,
"mean_token_accuracy": 0.7563480496406555,
"num_tokens": 1356510623.0,
"step": 10420
},
{
"entropy": 1.844921875,
"epoch": 0.5422125181950509,
"grad_norm": 3.2166301862941866,
"learning_rate": 4.7202272483653125e-06,
"loss": 1.0507,
"mean_token_accuracy": 0.7471937805414199,
"num_tokens": 1357812782.0,
"step": 10430
},
{
"entropy": 1.81015625,
"epoch": 0.5427323767935122,
"grad_norm": 2.8717269262560614,
"learning_rate": 4.7148676171079435e-06,
"loss": 1.0007,
"mean_token_accuracy": 0.7549396932125092,
"num_tokens": 1359112725.0,
"step": 10440
},
{
"entropy": 1.848828125,
"epoch": 0.5432522353919734,
"grad_norm": 2.571421118218861,
"learning_rate": 4.709507985850574e-06,
"loss": 1.0284,
"mean_token_accuracy": 0.7496370553970337,
"num_tokens": 1360412694.0,
"step": 10450
},
{
"entropy": 1.81796875,
"epoch": 0.5437720939904346,
"grad_norm": 2.33743526366292,
"learning_rate": 4.7041483545932045e-06,
"loss": 1.0097,
"mean_token_accuracy": 0.7546673536300659,
"num_tokens": 1361714443.0,
"step": 10460
},
{
"entropy": 1.819140625,
"epoch": 0.5442919525888958,
"grad_norm": 3.390788938548033,
"learning_rate": 4.698788723335835e-06,
"loss": 1.0269,
"mean_token_accuracy": 0.7501760631799698,
"num_tokens": 1363016198.0,
"step": 10470
},
{
"entropy": 1.86015625,
"epoch": 0.544811811187357,
"grad_norm": 2.4733318388943992,
"learning_rate": 4.693429092078466e-06,
"loss": 1.0252,
"mean_token_accuracy": 0.7509783715009689,
"num_tokens": 1364318959.0,
"step": 10480
},
{
"entropy": 1.8375,
"epoch": 0.5453316697858183,
"grad_norm": 2.6639615403582755,
"learning_rate": 4.688069460821096e-06,
"loss": 1.0576,
"mean_token_accuracy": 0.7474205434322357,
"num_tokens": 1365617925.0,
"step": 10490
},
{
"entropy": 1.84375,
"epoch": 0.5458515283842795,
"grad_norm": 2.679779279080176,
"learning_rate": 4.682709829563727e-06,
"loss": 1.0583,
"mean_token_accuracy": 0.744410040974617,
"num_tokens": 1366920343.0,
"step": 10500
},
{
"epoch": 0.5458515283842795,
"eval_entropy": 1.8376953125,
"eval_loss": 1.0683139562606812,
"eval_mean_token_accuracy": 0.7415347665548324,
"eval_num_tokens": 1366920343.0,
"eval_runtime": 27.003,
"eval_samples_per_second": 93.841,
"eval_steps_per_second": 1.481,
"step": 10500
},
{
"entropy": 1.840234375,
"epoch": 0.5463713869827407,
"grad_norm": 2.8661450383948344,
"learning_rate": 4.677350198306357e-06,
"loss": 1.0778,
"mean_token_accuracy": 0.7430825382471085,
"num_tokens": 1368223271.0,
"step": 10510
},
{
"entropy": 1.818359375,
"epoch": 0.546891245581202,
"grad_norm": 2.835474702271471,
"learning_rate": 4.671990567048987e-06,
"loss": 0.9999,
"mean_token_accuracy": 0.7560867697000504,
"num_tokens": 1369525160.0,
"step": 10520
},
{
"entropy": 1.849609375,
"epoch": 0.5474111041796631,
"grad_norm": 2.9613411470580493,
"learning_rate": 4.666630935791618e-06,
"loss": 1.1025,
"mean_token_accuracy": 0.7362652361392975,
"num_tokens": 1370824666.0,
"step": 10530
},
{
"entropy": 1.800390625,
"epoch": 0.5479309627781244,
"grad_norm": 2.634684062852215,
"learning_rate": 4.661271304534248e-06,
"loss": 1.0023,
"mean_token_accuracy": 0.7579319536685943,
"num_tokens": 1372126002.0,
"step": 10540
},
{
"entropy": 1.851171875,
"epoch": 0.5484508213765855,
"grad_norm": 3.659955996106995,
"learning_rate": 4.655911673276879e-06,
"loss": 1.0456,
"mean_token_accuracy": 0.7478381425142289,
"num_tokens": 1373427276.0,
"step": 10550
},
{
"entropy": 1.8625,
"epoch": 0.5489706799750468,
"grad_norm": 2.663732217961801,
"learning_rate": 4.650552042019509e-06,
"loss": 1.037,
"mean_token_accuracy": 0.7476789325475692,
"num_tokens": 1374729945.0,
"step": 10560
},
{
"entropy": 1.8125,
"epoch": 0.549490538573508,
"grad_norm": 2.5687930823092278,
"learning_rate": 4.64519241076214e-06,
"loss": 1.0265,
"mean_token_accuracy": 0.7542361050844193,
"num_tokens": 1376033075.0,
"step": 10570
},
{
"entropy": 1.84921875,
"epoch": 0.5500103971719692,
"grad_norm": 2.494747371310881,
"learning_rate": 4.63983277950477e-06,
"loss": 1.0588,
"mean_token_accuracy": 0.7447778552770614,
"num_tokens": 1377333036.0,
"step": 10580
},
{
"entropy": 1.83671875,
"epoch": 0.5505302557704305,
"grad_norm": 3.0174934045522277,
"learning_rate": 4.634473148247401e-06,
"loss": 1.035,
"mean_token_accuracy": 0.7523740351200103,
"num_tokens": 1378634008.0,
"step": 10590
},
{
"entropy": 1.838671875,
"epoch": 0.5510501143688916,
"grad_norm": 2.7886088062647625,
"learning_rate": 4.629113516990031e-06,
"loss": 1.0056,
"mean_token_accuracy": 0.7553982168436051,
"num_tokens": 1379936286.0,
"step": 10600
},
{
"epoch": 0.5510501143688916,
"eval_entropy": 1.8138671875,
"eval_loss": 1.0685162544250488,
"eval_mean_token_accuracy": 0.7414503991603851,
"eval_num_tokens": 1379936286.0,
"eval_runtime": 27.0591,
"eval_samples_per_second": 93.647,
"eval_steps_per_second": 1.478,
"step": 10600
},
{
"entropy": 1.847265625,
"epoch": 0.5515699729673529,
"grad_norm": 2.634582388088716,
"learning_rate": 4.623753885732662e-06,
"loss": 1.0295,
"mean_token_accuracy": 0.750268229842186,
"num_tokens": 1381238980.0,
"step": 10610
},
{
"entropy": 1.841015625,
"epoch": 0.5520898315658141,
"grad_norm": 2.779146260932326,
"learning_rate": 4.6183942544752924e-06,
"loss": 1.0575,
"mean_token_accuracy": 0.7466258227825164,
"num_tokens": 1382539048.0,
"step": 10620
},
{
"entropy": 1.812109375,
"epoch": 0.5526096901642753,
"grad_norm": 3.2224668020479457,
"learning_rate": 4.6130346232179226e-06,
"loss": 1.0328,
"mean_token_accuracy": 0.7502933651208877,
"num_tokens": 1383839339.0,
"step": 10630
},
{
"entropy": 1.809765625,
"epoch": 0.5531295487627366,
"grad_norm": 3.0098587173814972,
"learning_rate": 4.6076749919605535e-06,
"loss": 1.0641,
"mean_token_accuracy": 0.7449207782745362,
"num_tokens": 1385141559.0,
"step": 10640
},
{
"entropy": 1.8078125,
"epoch": 0.5536494073611977,
"grad_norm": 2.725603203398412,
"learning_rate": 4.602315360703184e-06,
"loss": 1.0597,
"mean_token_accuracy": 0.7456376820802688,
"num_tokens": 1386443282.0,
"step": 10650
},
{
"entropy": 1.820703125,
"epoch": 0.554169265959659,
"grad_norm": 3.0116575879353324,
"learning_rate": 4.596955729445815e-06,
"loss": 1.0048,
"mean_token_accuracy": 0.7576094299554825,
"num_tokens": 1387747257.0,
"step": 10660
},
{
"entropy": 1.786328125,
"epoch": 0.5546891245581201,
"grad_norm": 3.405905587680715,
"learning_rate": 4.591596098188445e-06,
"loss": 1.0549,
"mean_token_accuracy": 0.7461033761501312,
"num_tokens": 1389050870.0,
"step": 10670
},
{
"entropy": 1.822265625,
"epoch": 0.5552089831565814,
"grad_norm": 2.7943932116326637,
"learning_rate": 4.586236466931076e-06,
"loss": 1.042,
"mean_token_accuracy": 0.7499489068984986,
"num_tokens": 1390351745.0,
"step": 10680
},
{
"entropy": 1.850390625,
"epoch": 0.5557288417550427,
"grad_norm": 2.965111585987488,
"learning_rate": 4.580876835673706e-06,
"loss": 1.0112,
"mean_token_accuracy": 0.7563900977373124,
"num_tokens": 1391654004.0,
"step": 10690
},
{
"entropy": 1.81171875,
"epoch": 0.5562487003535038,
"grad_norm": 3.224369158811251,
"learning_rate": 4.575517204416337e-06,
"loss": 1.0756,
"mean_token_accuracy": 0.7403017222881317,
"num_tokens": 1392955975.0,
"step": 10700
},
{
"epoch": 0.5562487003535038,
"eval_entropy": 1.8232421875,
"eval_loss": 1.0640748739242554,
"eval_mean_token_accuracy": 0.7428328186273575,
"eval_num_tokens": 1392955975.0,
"eval_runtime": 26.9953,
"eval_samples_per_second": 93.868,
"eval_steps_per_second": 1.482,
"step": 10700
},
{
"entropy": 1.815625,
"epoch": 0.5567685589519651,
"grad_norm": 2.7176497479175974,
"learning_rate": 4.570157573158967e-06,
"loss": 1.0113,
"mean_token_accuracy": 0.7531196564435959,
"num_tokens": 1394258232.0,
"step": 10710
},
{
"entropy": 1.830859375,
"epoch": 0.5572884175504262,
"grad_norm": 2.3120831540679316,
"learning_rate": 4.564797941901598e-06,
"loss": 1.0813,
"mean_token_accuracy": 0.7416792064905167,
"num_tokens": 1395559597.0,
"step": 10720
},
{
"entropy": 1.8375,
"epoch": 0.5578082761488875,
"grad_norm": 2.93661716185051,
"learning_rate": 4.559438310644228e-06,
"loss": 1.027,
"mean_token_accuracy": 0.7535325318574906,
"num_tokens": 1396863106.0,
"step": 10730
},
{
"entropy": 1.808203125,
"epoch": 0.5583281347473488,
"grad_norm": 2.7288366775811066,
"learning_rate": 4.554078679386858e-06,
"loss": 1.0907,
"mean_token_accuracy": 0.7375077068805694,
"num_tokens": 1398163178.0,
"step": 10740
},
{
"entropy": 1.834765625,
"epoch": 0.5588479933458099,
"grad_norm": 3.098413642524893,
"learning_rate": 4.548719048129489e-06,
"loss": 1.0233,
"mean_token_accuracy": 0.7524217307567597,
"num_tokens": 1399467152.0,
"step": 10750
},
{
"entropy": 1.81015625,
"epoch": 0.5593678519442712,
"grad_norm": 2.806725920718146,
"learning_rate": 4.543359416872119e-06,
"loss": 1.099,
"mean_token_accuracy": 0.7372580885887146,
"num_tokens": 1400769111.0,
"step": 10760
},
{
"entropy": 1.805859375,
"epoch": 0.5598877105427323,
"grad_norm": 3.0718902610706813,
"learning_rate": 4.53799978561475e-06,
"loss": 1.0398,
"mean_token_accuracy": 0.7500275552272797,
"num_tokens": 1402071635.0,
"step": 10770
},
{
"entropy": 1.849609375,
"epoch": 0.5604075691411936,
"grad_norm": 2.951931047364739,
"learning_rate": 4.53264015435738e-06,
"loss": 1.0499,
"mean_token_accuracy": 0.7470193892717362,
"num_tokens": 1403373382.0,
"step": 10780
},
{
"entropy": 1.85703125,
"epoch": 0.5609274277396548,
"grad_norm": 2.7874274882227303,
"learning_rate": 4.527280523100011e-06,
"loss": 1.0716,
"mean_token_accuracy": 0.7422095656394958,
"num_tokens": 1404674491.0,
"step": 10790
},
{
"entropy": 1.83125,
"epoch": 0.561447286338116,
"grad_norm": 3.0280502548533246,
"learning_rate": 4.5219208918426414e-06,
"loss": 0.9605,
"mean_token_accuracy": 0.7652639031410218,
"num_tokens": 1405977493.0,
"step": 10800
},
{
"epoch": 0.561447286338116,
"eval_entropy": 1.7673828125,
"eval_loss": 1.0627888441085815,
"eval_mean_token_accuracy": 0.7433530449867248,
"eval_num_tokens": 1405977493.0,
"eval_runtime": 27.0426,
"eval_samples_per_second": 93.704,
"eval_steps_per_second": 1.479,
"step": 10800
},
{
"entropy": 1.78515625,
"epoch": 0.5619671449365773,
"grad_norm": 2.764600561979598,
"learning_rate": 4.516561260585272e-06,
"loss": 1.0222,
"mean_token_accuracy": 0.7516841471195221,
"num_tokens": 1407279178.0,
"step": 10810
},
{
"entropy": 1.8484375,
"epoch": 0.5624870035350384,
"grad_norm": 2.6274439409558856,
"learning_rate": 4.5112016293279025e-06,
"loss": 1.0244,
"mean_token_accuracy": 0.7550068736076355,
"num_tokens": 1408580607.0,
"step": 10820
},
{
"entropy": 1.833203125,
"epoch": 0.5630068621334997,
"grad_norm": 2.5384505998172884,
"learning_rate": 4.5058419980705335e-06,
"loss": 0.9945,
"mean_token_accuracy": 0.7624350160360336,
"num_tokens": 1409882692.0,
"step": 10830
},
{
"entropy": 1.81796875,
"epoch": 0.5635267207319609,
"grad_norm": 3.130034840988354,
"learning_rate": 4.500482366813164e-06,
"loss": 1.0614,
"mean_token_accuracy": 0.743175408244133,
"num_tokens": 1411184706.0,
"step": 10840
},
{
"entropy": 1.839453125,
"epoch": 0.5640465793304221,
"grad_norm": 3.4003055576879535,
"learning_rate": 4.495122735555794e-06,
"loss": 1.0459,
"mean_token_accuracy": 0.7461069792509079,
"num_tokens": 1412484969.0,
"step": 10850
},
{
"entropy": 1.85078125,
"epoch": 0.5645664379288834,
"grad_norm": 2.9471903574823943,
"learning_rate": 4.489763104298425e-06,
"loss": 1.0091,
"mean_token_accuracy": 0.7551805168390274,
"num_tokens": 1413788283.0,
"step": 10860
},
{
"entropy": 1.830078125,
"epoch": 0.5650862965273445,
"grad_norm": 2.445923389869576,
"learning_rate": 4.484403473041055e-06,
"loss": 1.0169,
"mean_token_accuracy": 0.7529974430799484,
"num_tokens": 1415091014.0,
"step": 10870
},
{
"entropy": 1.831640625,
"epoch": 0.5656061551258058,
"grad_norm": 3.0485597529985076,
"learning_rate": 4.479043841783686e-06,
"loss": 1.0398,
"mean_token_accuracy": 0.7496816605329514,
"num_tokens": 1416393419.0,
"step": 10880
},
{
"entropy": 1.83203125,
"epoch": 0.566126013724267,
"grad_norm": 2.6110909986032813,
"learning_rate": 4.473684210526316e-06,
"loss": 1.0183,
"mean_token_accuracy": 0.7518593400716782,
"num_tokens": 1417694695.0,
"step": 10890
},
{
"entropy": 1.823828125,
"epoch": 0.5666458723227282,
"grad_norm": 2.565862447116262,
"learning_rate": 4.468324579268947e-06,
"loss": 0.9941,
"mean_token_accuracy": 0.7576651304960251,
"num_tokens": 1418995497.0,
"step": 10900
},
{
"epoch": 0.5666458723227282,
"eval_entropy": 1.7943359375,
"eval_loss": 1.0589063167572021,
"eval_mean_token_accuracy": 0.7439443483948708,
"eval_num_tokens": 1418995497.0,
"eval_runtime": 27.0974,
"eval_samples_per_second": 93.515,
"eval_steps_per_second": 1.476,
"step": 10900
},
{
"entropy": 1.793359375,
"epoch": 0.5671657309211894,
"grad_norm": 2.557336905491837,
"learning_rate": 4.462964948011577e-06,
"loss": 1.0231,
"mean_token_accuracy": 0.7562372565269471,
"num_tokens": 1420298891.0,
"step": 10910
},
{
"entropy": 1.83359375,
"epoch": 0.5676855895196506,
"grad_norm": 2.7683847438246625,
"learning_rate": 4.457605316754208e-06,
"loss": 1.0652,
"mean_token_accuracy": 0.7443037003278732,
"num_tokens": 1421601042.0,
"step": 10920
},
{
"entropy": 1.84609375,
"epoch": 0.5682054481181119,
"grad_norm": 3.1118138527783805,
"learning_rate": 4.452245685496838e-06,
"loss": 0.9831,
"mean_token_accuracy": 0.7623260080814361,
"num_tokens": 1422903267.0,
"step": 10930
},
{
"entropy": 1.799609375,
"epoch": 0.5687253067165731,
"grad_norm": 2.666802229985345,
"learning_rate": 4.446886054239469e-06,
"loss": 1.0087,
"mean_token_accuracy": 0.755416652560234,
"num_tokens": 1424204727.0,
"step": 10940
},
{
"entropy": 1.841796875,
"epoch": 0.5692451653150343,
"grad_norm": 2.794349669202065,
"learning_rate": 4.441526422982099e-06,
"loss": 1.0715,
"mean_token_accuracy": 0.7419457286596298,
"num_tokens": 1425505547.0,
"step": 10950
},
{
"entropy": 1.787890625,
"epoch": 0.5697650239134955,
"grad_norm": 2.88136377582001,
"learning_rate": 4.436166791724729e-06,
"loss": 0.9654,
"mean_token_accuracy": 0.7666537404060364,
"num_tokens": 1426809110.0,
"step": 10960
},
{
"entropy": 1.821484375,
"epoch": 0.5702848825119567,
"grad_norm": 2.5226918374624088,
"learning_rate": 4.43080716046736e-06,
"loss": 0.9969,
"mean_token_accuracy": 0.7600201427936554,
"num_tokens": 1428112307.0,
"step": 10970
},
{
"entropy": 1.840625,
"epoch": 0.570804741110418,
"grad_norm": 2.695122701707645,
"learning_rate": 4.4254475292099904e-06,
"loss": 1.0072,
"mean_token_accuracy": 0.7564296513795853,
"num_tokens": 1429414533.0,
"step": 10980
},
{
"entropy": 1.7859375,
"epoch": 0.5713245997088792,
"grad_norm": 2.7885099670446656,
"learning_rate": 4.420087897952621e-06,
"loss": 1.0017,
"mean_token_accuracy": 0.7564855188131332,
"num_tokens": 1430714862.0,
"step": 10990
},
{
"entropy": 1.817578125,
"epoch": 0.5718444583073404,
"grad_norm": 2.724635743272623,
"learning_rate": 4.4147282666952515e-06,
"loss": 1.018,
"mean_token_accuracy": 0.7516950726509094,
"num_tokens": 1432016242.0,
"step": 11000
},
{
"epoch": 0.5718444583073404,
"eval_entropy": 1.827734375,
"eval_loss": 1.0549143552780151,
"eval_mean_token_accuracy": 0.7446142882108688,
"eval_num_tokens": 1432016242.0,
"eval_runtime": 27.0421,
"eval_samples_per_second": 93.706,
"eval_steps_per_second": 1.479,
"step": 11000
},
{
"entropy": 1.84375,
"epoch": 0.5723643169058016,
"grad_norm": 2.842672918161374,
"learning_rate": 4.4093686354378825e-06,
"loss": 1.0338,
"mean_token_accuracy": 0.7535577833652496,
"num_tokens": 1433320136.0,
"step": 11010
},
{
"entropy": 1.823046875,
"epoch": 0.5728841755042628,
"grad_norm": 2.7634302334734926,
"learning_rate": 4.404009004180513e-06,
"loss": 1.1159,
"mean_token_accuracy": 0.7348658710718154,
"num_tokens": 1434623607.0,
"step": 11020
},
{
"entropy": 1.836328125,
"epoch": 0.5734040341027241,
"grad_norm": 2.8235644176856542,
"learning_rate": 4.398649372923144e-06,
"loss": 1.0634,
"mean_token_accuracy": 0.7454641371965408,
"num_tokens": 1435926359.0,
"step": 11030
},
{
"entropy": 1.790234375,
"epoch": 0.5739238927011853,
"grad_norm": 2.735341897347149,
"learning_rate": 4.393289741665774e-06,
"loss": 0.9625,
"mean_token_accuracy": 0.7672121524810791,
"num_tokens": 1437228807.0,
"step": 11040
},
{
"entropy": 1.820703125,
"epoch": 0.5744437512996465,
"grad_norm": 2.957769896826195,
"learning_rate": 4.387930110408405e-06,
"loss": 1.0048,
"mean_token_accuracy": 0.7586895823478699,
"num_tokens": 1438530743.0,
"step": 11050
},
{
"entropy": 1.812109375,
"epoch": 0.5749636098981077,
"grad_norm": 3.0146006297517616,
"learning_rate": 4.382570479151035e-06,
"loss": 1.0786,
"mean_token_accuracy": 0.7413741409778595,
"num_tokens": 1439833132.0,
"step": 11060
},
{
"entropy": 1.86796875,
"epoch": 0.575483468496569,
"grad_norm": 2.71191633364465,
"learning_rate": 4.377210847893666e-06,
"loss": 1.0134,
"mean_token_accuracy": 0.7565578937530517,
"num_tokens": 1441136054.0,
"step": 11070
},
{
"entropy": 1.828515625,
"epoch": 0.5760033270950301,
"grad_norm": 3.1067091870524974,
"learning_rate": 4.371851216636296e-06,
"loss": 1.0604,
"mean_token_accuracy": 0.7461466014385223,
"num_tokens": 1442437427.0,
"step": 11080
},
{
"entropy": 1.812890625,
"epoch": 0.5765231856934914,
"grad_norm": 2.407926465102427,
"learning_rate": 4.366491585378926e-06,
"loss": 1.0473,
"mean_token_accuracy": 0.751253855228424,
"num_tokens": 1443739677.0,
"step": 11090
},
{
"entropy": 1.816015625,
"epoch": 0.5770430442919526,
"grad_norm": 2.7000233526540245,
"learning_rate": 4.361131954121557e-06,
"loss": 1.0033,
"mean_token_accuracy": 0.7593827843666077,
"num_tokens": 1445042710.0,
"step": 11100
},
{
"epoch": 0.5770430442919526,
"eval_entropy": 1.8287109375,
"eval_loss": 1.0482763051986694,
"eval_mean_token_accuracy": 0.7467321798205375,
"eval_num_tokens": 1445042710.0,
"eval_runtime": 27.0996,
"eval_samples_per_second": 93.507,
"eval_steps_per_second": 1.476,
"step": 11100
},
{
"entropy": 1.829296875,
"epoch": 0.5775629028904138,
"grad_norm": 3.0938891288827137,
"learning_rate": 4.355772322864187e-06,
"loss": 1.0162,
"mean_token_accuracy": 0.7544076889753342,
"num_tokens": 1446345651.0,
"step": 11110
},
{
"entropy": 1.801953125,
"epoch": 0.578082761488875,
"grad_norm": 2.787219279686718,
"learning_rate": 4.350412691606818e-06,
"loss": 0.9957,
"mean_token_accuracy": 0.7588774114847183,
"num_tokens": 1447645824.0,
"step": 11120
},
{
"entropy": 1.81875,
"epoch": 0.5786026200873362,
"grad_norm": 2.751434135348773,
"learning_rate": 4.345053060349448e-06,
"loss": 1.0652,
"mean_token_accuracy": 0.7396714657545089,
"num_tokens": 1448946903.0,
"step": 11130
},
{
"entropy": 1.8203125,
"epoch": 0.5791224786857975,
"grad_norm": 2.6727405585449358,
"learning_rate": 4.339693429092079e-06,
"loss": 1.0123,
"mean_token_accuracy": 0.7557833731174469,
"num_tokens": 1450248337.0,
"step": 11140
},
{
"entropy": 1.810546875,
"epoch": 0.5796423372842587,
"grad_norm": 2.78256238910243,
"learning_rate": 4.334333797834709e-06,
"loss": 1.0486,
"mean_token_accuracy": 0.7509361416101455,
"num_tokens": 1451551323.0,
"step": 11150
},
{
"entropy": 1.785546875,
"epoch": 0.5801621958827199,
"grad_norm": 2.7594673549975517,
"learning_rate": 4.32897416657734e-06,
"loss": 0.9963,
"mean_token_accuracy": 0.7593833267688751,
"num_tokens": 1452853674.0,
"step": 11160
},
{
"entropy": 1.80703125,
"epoch": 0.5806820544811812,
"grad_norm": 3.269480668942356,
"learning_rate": 4.32361453531997e-06,
"loss": 1.0397,
"mean_token_accuracy": 0.7504145979881287,
"num_tokens": 1454155087.0,
"step": 11170
},
{
"entropy": 1.7609375,
"epoch": 0.5812019130796423,
"grad_norm": 3.2769067474592006,
"learning_rate": 4.318254904062601e-06,
"loss": 1.0076,
"mean_token_accuracy": 0.7571777641773224,
"num_tokens": 1455457488.0,
"step": 11180
},
{
"entropy": 1.8359375,
"epoch": 0.5817217716781036,
"grad_norm": 2.748939465256801,
"learning_rate": 4.3128952728052315e-06,
"loss": 1.0385,
"mean_token_accuracy": 0.7514304906129837,
"num_tokens": 1456759416.0,
"step": 11190
},
{
"entropy": 1.824609375,
"epoch": 0.5822416302765647,
"grad_norm": 3.1287977243420375,
"learning_rate": 4.307535641547862e-06,
"loss": 1.0042,
"mean_token_accuracy": 0.7568298161029816,
"num_tokens": 1458061571.0,
"step": 11200
},
{
"epoch": 0.5822416302765647,
"eval_entropy": 1.7875,
"eval_loss": 1.0476077795028687,
"eval_mean_token_accuracy": 0.746434785425663,
"eval_num_tokens": 1458061571.0,
"eval_runtime": 27.0368,
"eval_samples_per_second": 93.724,
"eval_steps_per_second": 1.479,
"step": 11200
},
{
"entropy": 1.799609375,
"epoch": 0.582761488875026,
"grad_norm": 3.06077561758173,
"learning_rate": 4.3021760102904926e-06,
"loss": 1.0001,
"mean_token_accuracy": 0.7597445189952851,
"num_tokens": 1459364165.0,
"step": 11210
},
{
"entropy": 1.796875,
"epoch": 0.5832813474734873,
"grad_norm": 2.904613379840475,
"learning_rate": 4.296816379033123e-06,
"loss": 0.956,
"mean_token_accuracy": 0.7689589917659759,
"num_tokens": 1460665209.0,
"step": 11220
},
{
"entropy": 1.79453125,
"epoch": 0.5838012060719484,
"grad_norm": 2.5378003405692255,
"learning_rate": 4.291456747775754e-06,
"loss": 1.0034,
"mean_token_accuracy": 0.757875970005989,
"num_tokens": 1461966443.0,
"step": 11230
},
{
"entropy": 1.8171875,
"epoch": 0.5843210646704097,
"grad_norm": 2.6742241787334105,
"learning_rate": 4.286097116518384e-06,
"loss": 1.0442,
"mean_token_accuracy": 0.7484052300453186,
"num_tokens": 1463269239.0,
"step": 11240
},
{
"entropy": 1.8609375,
"epoch": 0.5848409232688708,
"grad_norm": 3.1816584314733913,
"learning_rate": 4.280737485261015e-06,
"loss": 1.0805,
"mean_token_accuracy": 0.7405004471540451,
"num_tokens": 1464569130.0,
"step": 11250
},
{
"entropy": 1.83046875,
"epoch": 0.5853607818673321,
"grad_norm": 2.8564349561903297,
"learning_rate": 4.275377854003645e-06,
"loss": 1.1184,
"mean_token_accuracy": 0.7312965095043182,
"num_tokens": 1465872363.0,
"step": 11260
},
{
"entropy": 1.82890625,
"epoch": 0.5858806404657934,
"grad_norm": 2.9065543083421255,
"learning_rate": 4.270018222746276e-06,
"loss": 1.0388,
"mean_token_accuracy": 0.7494263142347336,
"num_tokens": 1467175069.0,
"step": 11270
},
{
"entropy": 1.801953125,
"epoch": 0.5864004990642545,
"grad_norm": 2.609095939413824,
"learning_rate": 4.264658591488906e-06,
"loss": 1.0291,
"mean_token_accuracy": 0.7508643001317978,
"num_tokens": 1468475960.0,
"step": 11280
},
{
"entropy": 1.822265625,
"epoch": 0.5869203576627158,
"grad_norm": 2.5615094655664903,
"learning_rate": 4.259298960231537e-06,
"loss": 1.0644,
"mean_token_accuracy": 0.7468615204095841,
"num_tokens": 1469777428.0,
"step": 11290
},
{
"entropy": 1.79765625,
"epoch": 0.5874402162611769,
"grad_norm": 2.9345868872786967,
"learning_rate": 4.253939328974167e-06,
"loss": 1.0856,
"mean_token_accuracy": 0.7422370672225952,
"num_tokens": 1471078994.0,
"step": 11300
},
{
"epoch": 0.5874402162611769,
"eval_entropy": 1.7791015625,
"eval_loss": 1.0470556020736694,
"eval_mean_token_accuracy": 0.7465546175837516,
"eval_num_tokens": 1471078994.0,
"eval_runtime": 27.1017,
"eval_samples_per_second": 93.5,
"eval_steps_per_second": 1.476,
"step": 11300
},
{
"entropy": 1.818359375,
"epoch": 0.5879600748596382,
"grad_norm": 2.5675545599958576,
"learning_rate": 4.248579697716797e-06,
"loss": 1.0372,
"mean_token_accuracy": 0.7524241745471955,
"num_tokens": 1472383281.0,
"step": 11310
},
{
"entropy": 1.796484375,
"epoch": 0.5884799334580993,
"grad_norm": 2.4775098589396296,
"learning_rate": 4.243220066459428e-06,
"loss": 0.9699,
"mean_token_accuracy": 0.7644401967525483,
"num_tokens": 1473684218.0,
"step": 11320
},
{
"entropy": 1.808984375,
"epoch": 0.5889997920565606,
"grad_norm": 3.2713595274426948,
"learning_rate": 4.237860435202058e-06,
"loss": 1.0145,
"mean_token_accuracy": 0.7542434215545655,
"num_tokens": 1474986476.0,
"step": 11330
},
{
"entropy": 1.814453125,
"epoch": 0.5895196506550219,
"grad_norm": 2.7270772505633047,
"learning_rate": 4.232500803944689e-06,
"loss": 0.9972,
"mean_token_accuracy": 0.7605729520320892,
"num_tokens": 1476286114.0,
"step": 11340
},
{
"entropy": 1.808203125,
"epoch": 0.590039509253483,
"grad_norm": 2.6599590427886395,
"learning_rate": 4.227141172687319e-06,
"loss": 1.0414,
"mean_token_accuracy": 0.7463055342435837,
"num_tokens": 1477589337.0,
"step": 11350
},
{
"entropy": 1.78359375,
"epoch": 0.5905593678519443,
"grad_norm": 3.0476465708183533,
"learning_rate": 4.22178154142995e-06,
"loss": 0.9747,
"mean_token_accuracy": 0.7639808356761932,
"num_tokens": 1478892771.0,
"step": 11360
},
{
"entropy": 1.821875,
"epoch": 0.5910792264504054,
"grad_norm": 2.8193951129574377,
"learning_rate": 4.2164219101725805e-06,
"loss": 0.9656,
"mean_token_accuracy": 0.7664701163768768,
"num_tokens": 1480195033.0,
"step": 11370
},
{
"entropy": 1.82265625,
"epoch": 0.5915990850488667,
"grad_norm": 2.794414997702922,
"learning_rate": 4.2110622789152114e-06,
"loss": 0.9748,
"mean_token_accuracy": 0.7648434281349182,
"num_tokens": 1481496618.0,
"step": 11380
},
{
"entropy": 1.821484375,
"epoch": 0.592118943647328,
"grad_norm": 2.6278117212801657,
"learning_rate": 4.2057026476578416e-06,
"loss": 0.9949,
"mean_token_accuracy": 0.7615526139736175,
"num_tokens": 1482797206.0,
"step": 11390
},
{
"entropy": 1.824609375,
"epoch": 0.5926388022457891,
"grad_norm": 2.8296865887371094,
"learning_rate": 4.2003430164004725e-06,
"loss": 1.0321,
"mean_token_accuracy": 0.7477697223424912,
"num_tokens": 1484097702.0,
"step": 11400
},
{
"epoch": 0.5926388022457891,
"eval_entropy": 1.7869140625,
"eval_loss": 1.044844627380371,
"eval_mean_token_accuracy": 0.7470950469374656,
"eval_num_tokens": 1484097702.0,
"eval_runtime": 27.0422,
"eval_samples_per_second": 93.705,
"eval_steps_per_second": 1.479,
"step": 11400
},
{
"entropy": 1.828125,
"epoch": 0.5931586608442504,
"grad_norm": 2.749569868632866,
"learning_rate": 4.194983385143103e-06,
"loss": 1.052,
"mean_token_accuracy": 0.7460355401039124,
"num_tokens": 1485398305.0,
"step": 11410
},
{
"entropy": 1.820703125,
"epoch": 0.5936785194427115,
"grad_norm": 2.543814699407546,
"learning_rate": 4.189623753885733e-06,
"loss": 1.0595,
"mean_token_accuracy": 0.7430806964635849,
"num_tokens": 1486697669.0,
"step": 11420
},
{
"entropy": 1.821484375,
"epoch": 0.5941983780411728,
"grad_norm": 3.067390872763037,
"learning_rate": 4.184264122628364e-06,
"loss": 0.9526,
"mean_token_accuracy": 0.7687475472688675,
"num_tokens": 1487999785.0,
"step": 11430
},
{
"entropy": 1.80703125,
"epoch": 0.594718236639634,
"grad_norm": 2.5867144747622643,
"learning_rate": 4.178904491370994e-06,
"loss": 1.073,
"mean_token_accuracy": 0.7418835282325744,
"num_tokens": 1489300966.0,
"step": 11440
},
{
"entropy": 1.834375,
"epoch": 0.5952380952380952,
"grad_norm": 2.4903607402435552,
"learning_rate": 4.173544860113625e-06,
"loss": 1.0501,
"mean_token_accuracy": 0.7467874974012375,
"num_tokens": 1490601848.0,
"step": 11450
},
{
"entropy": 1.805078125,
"epoch": 0.5957579538365565,
"grad_norm": 2.8055602295848927,
"learning_rate": 4.168185228856255e-06,
"loss": 1.012,
"mean_token_accuracy": 0.7552043467760086,
"num_tokens": 1491903955.0,
"step": 11460
},
{
"entropy": 1.846875,
"epoch": 0.5962778124350177,
"grad_norm": 2.8674072868848652,
"learning_rate": 4.162825597598886e-06,
"loss": 1.0363,
"mean_token_accuracy": 0.7490568488836289,
"num_tokens": 1493206356.0,
"step": 11470
},
{
"entropy": 1.801171875,
"epoch": 0.5967976710334789,
"grad_norm": 2.84576557593928,
"learning_rate": 4.157465966341516e-06,
"loss": 0.999,
"mean_token_accuracy": 0.7602721244096756,
"num_tokens": 1494505949.0,
"step": 11480
},
{
"entropy": 1.787109375,
"epoch": 0.5973175296319401,
"grad_norm": 2.69395637282585,
"learning_rate": 4.152106335084147e-06,
"loss": 0.9945,
"mean_token_accuracy": 0.7617682188749313,
"num_tokens": 1495809091.0,
"step": 11490
},
{
"entropy": 1.823828125,
"epoch": 0.5978373882304013,
"grad_norm": 2.676867185398759,
"learning_rate": 4.146746703826777e-06,
"loss": 1.0339,
"mean_token_accuracy": 0.750063705444336,
"num_tokens": 1497107983.0,
"step": 11500
},
{
"epoch": 0.5978373882304013,
"eval_entropy": 1.7724609375,
"eval_loss": 1.040482759475708,
"eval_mean_token_accuracy": 0.7479965060949325,
"eval_num_tokens": 1497107983.0,
"eval_runtime": 27.0284,
"eval_samples_per_second": 93.753,
"eval_steps_per_second": 1.48,
"step": 11500
},
{
"entropy": 1.794921875,
"epoch": 0.5983572468288626,
"grad_norm": 2.7976708322150583,
"learning_rate": 4.141387072569408e-06,
"loss": 0.9979,
"mean_token_accuracy": 0.7591456085443496,
"num_tokens": 1498411357.0,
"step": 11510
},
{
"entropy": 1.783984375,
"epoch": 0.5988771054273238,
"grad_norm": 2.7994922763742256,
"learning_rate": 4.136027441312038e-06,
"loss": 1.0175,
"mean_token_accuracy": 0.7550407528877259,
"num_tokens": 1499710902.0,
"step": 11520
},
{
"entropy": 1.81484375,
"epoch": 0.599396964025785,
"grad_norm": 2.834504454337439,
"learning_rate": 4.130667810054668e-06,
"loss": 1.027,
"mean_token_accuracy": 0.7554367780685425,
"num_tokens": 1501013043.0,
"step": 11530
},
{
"entropy": 1.8234375,
"epoch": 0.5999168226242462,
"grad_norm": 2.7872337315233735,
"learning_rate": 4.125308178797299e-06,
"loss": 1.0571,
"mean_token_accuracy": 0.7449495941400528,
"num_tokens": 1502315799.0,
"step": 11540
},
{
"entropy": 1.798828125,
"epoch": 0.6004366812227074,
"grad_norm": 2.5610000096497982,
"learning_rate": 4.1199485475399295e-06,
"loss": 0.9964,
"mean_token_accuracy": 0.7592524856328964,
"num_tokens": 1503616627.0,
"step": 11550
},
{
"entropy": 1.81953125,
"epoch": 0.6009565398211686,
"grad_norm": 2.7489255290200765,
"learning_rate": 4.1145889162825604e-06,
"loss": 1.0283,
"mean_token_accuracy": 0.7499420672655106,
"num_tokens": 1504918577.0,
"step": 11560
},
{
"entropy": 1.81953125,
"epoch": 0.6014763984196299,
"grad_norm": 2.5323635994369984,
"learning_rate": 4.1092292850251906e-06,
"loss": 1.0003,
"mean_token_accuracy": 0.7576062381267548,
"num_tokens": 1506221163.0,
"step": 11570
},
{
"entropy": 1.829296875,
"epoch": 0.6019962570180911,
"grad_norm": 2.8072190102104115,
"learning_rate": 4.1038696537678215e-06,
"loss": 1.0161,
"mean_token_accuracy": 0.7567219614982605,
"num_tokens": 1507518213.0,
"step": 11580
},
{
"entropy": 1.81015625,
"epoch": 0.6025161156165523,
"grad_norm": 3.0609476772028015,
"learning_rate": 4.098510022510452e-06,
"loss": 1.0052,
"mean_token_accuracy": 0.7570147007703781,
"num_tokens": 1508820716.0,
"step": 11590
},
{
"entropy": 1.83046875,
"epoch": 0.6030359742150135,
"grad_norm": 3.568789573427951,
"learning_rate": 4.093150391253083e-06,
"loss": 1.0156,
"mean_token_accuracy": 0.7530331879854202,
"num_tokens": 1510119755.0,
"step": 11600
},
{
"epoch": 0.6030359742150135,
"eval_entropy": 1.8013671875,
"eval_loss": 1.0382553339004517,
"eval_mean_token_accuracy": 0.7489291131496429,
"eval_num_tokens": 1510119755.0,
"eval_runtime": 27.0043,
"eval_samples_per_second": 93.837,
"eval_steps_per_second": 1.481,
"step": 11600
},
{
"entropy": 1.812109375,
"epoch": 0.6035558328134747,
"grad_norm": 2.4004705098416523,
"learning_rate": 4.087790759995713e-06,
"loss": 1.0382,
"mean_token_accuracy": 0.7521683245897293,
"num_tokens": 1511420895.0,
"step": 11610
},
{
"entropy": 1.773828125,
"epoch": 0.604075691411936,
"grad_norm": 2.5521756518865106,
"learning_rate": 4.082431128738344e-06,
"loss": 0.9673,
"mean_token_accuracy": 0.7645098686218261,
"num_tokens": 1512723798.0,
"step": 11620
},
{
"entropy": 1.8140625,
"epoch": 0.6045955500103972,
"grad_norm": 2.9218662112929343,
"learning_rate": 4.077071497480974e-06,
"loss": 1.0169,
"mean_token_accuracy": 0.7542355537414551,
"num_tokens": 1514026812.0,
"step": 11630
},
{
"entropy": 1.8453125,
"epoch": 0.6051154086088584,
"grad_norm": 2.619416958553604,
"learning_rate": 4.071711866223604e-06,
"loss": 1.033,
"mean_token_accuracy": 0.7523566812276841,
"num_tokens": 1515328481.0,
"step": 11640
},
{
"entropy": 1.85078125,
"epoch": 0.6056352672073196,
"grad_norm": 3.2301958708048817,
"learning_rate": 4.066352234966235e-06,
"loss": 1.0095,
"mean_token_accuracy": 0.7579028367996216,
"num_tokens": 1516628855.0,
"step": 11650
},
{
"entropy": 1.834375,
"epoch": 0.6061551258057808,
"grad_norm": 2.6812535961818935,
"learning_rate": 4.060992603708865e-06,
"loss": 1.029,
"mean_token_accuracy": 0.7516258001327515,
"num_tokens": 1517930429.0,
"step": 11660
},
{
"entropy": 1.848046875,
"epoch": 0.6066749844042421,
"grad_norm": 2.879065311710356,
"learning_rate": 4.055632972451496e-06,
"loss": 0.9557,
"mean_token_accuracy": 0.7689045190811157,
"num_tokens": 1519230296.0,
"step": 11670
},
{
"entropy": 1.801171875,
"epoch": 0.6071948430027032,
"grad_norm": 2.308372108530545,
"learning_rate": 4.050273341194126e-06,
"loss": 1.0151,
"mean_token_accuracy": 0.754237613081932,
"num_tokens": 1520533955.0,
"step": 11680
},
{
"entropy": 1.833203125,
"epoch": 0.6077147016011645,
"grad_norm": 2.9705362923874827,
"learning_rate": 4.044913709936757e-06,
"loss": 1.0234,
"mean_token_accuracy": 0.7530678808689117,
"num_tokens": 1521835034.0,
"step": 11690
},
{
"entropy": 1.8203125,
"epoch": 0.6082345601996257,
"grad_norm": 2.8389347072550586,
"learning_rate": 4.039554078679387e-06,
"loss": 0.9901,
"mean_token_accuracy": 0.7652809739112854,
"num_tokens": 1523136681.0,
"step": 11700
},
{
"epoch": 0.6082345601996257,
"eval_entropy": 1.815625,
"eval_loss": 1.0361155271530151,
"eval_mean_token_accuracy": 0.7493843644857406,
"eval_num_tokens": 1523136681.0,
"eval_runtime": 27.014,
"eval_samples_per_second": 93.803,
"eval_steps_per_second": 1.481,
"step": 11700
},
{
"entropy": 1.8140625,
"epoch": 0.6087544187980869,
"grad_norm": 2.931335759690872,
"learning_rate": 4.034194447422018e-06,
"loss": 1.0288,
"mean_token_accuracy": 0.753569358587265,
"num_tokens": 1524439171.0,
"step": 11710
},
{
"entropy": 1.8421875,
"epoch": 0.6092742773965482,
"grad_norm": 2.5947781749931282,
"learning_rate": 4.028834816164648e-06,
"loss": 1.0884,
"mean_token_accuracy": 0.7390716820955276,
"num_tokens": 1525742099.0,
"step": 11720
},
{
"entropy": 1.798046875,
"epoch": 0.6097941359950093,
"grad_norm": 3.226487221166174,
"learning_rate": 4.023475184907279e-06,
"loss": 0.9929,
"mean_token_accuracy": 0.7615871816873551,
"num_tokens": 1527040829.0,
"step": 11730
},
{
"entropy": 1.824609375,
"epoch": 0.6103139945934706,
"grad_norm": 3.157667525042087,
"learning_rate": 4.0181155536499094e-06,
"loss": 1.0482,
"mean_token_accuracy": 0.7485737860202789,
"num_tokens": 1528344373.0,
"step": 11740
},
{
"entropy": 1.83828125,
"epoch": 0.6108338531919318,
"grad_norm": 2.8792599003289943,
"learning_rate": 4.0127559223925396e-06,
"loss": 0.9815,
"mean_token_accuracy": 0.7651452869176865,
"num_tokens": 1529644641.0,
"step": 11750
},
{
"entropy": 1.832421875,
"epoch": 0.611353711790393,
"grad_norm": 3.2182632973451115,
"learning_rate": 4.0073962911351705e-06,
"loss": 0.9793,
"mean_token_accuracy": 0.7655253291130066,
"num_tokens": 1530945775.0,
"step": 11760
},
{
"entropy": 1.79765625,
"epoch": 0.6118735703888543,
"grad_norm": 2.586873627246423,
"learning_rate": 4.002036659877801e-06,
"loss": 0.9806,
"mean_token_accuracy": 0.7631507635116577,
"num_tokens": 1532248263.0,
"step": 11770
},
{
"entropy": 1.814453125,
"epoch": 0.6123934289873154,
"grad_norm": 2.3057324772325845,
"learning_rate": 3.996677028620432e-06,
"loss": 1.0159,
"mean_token_accuracy": 0.7552026063203812,
"num_tokens": 1533552348.0,
"step": 11780
},
{
"entropy": 1.77890625,
"epoch": 0.6129132875857767,
"grad_norm": 4.257541022857633,
"learning_rate": 3.991317397363062e-06,
"loss": 0.9838,
"mean_token_accuracy": 0.7618939578533173,
"num_tokens": 1534853821.0,
"step": 11790
},
{
"entropy": 1.78828125,
"epoch": 0.6134331461842378,
"grad_norm": 2.7539279125729053,
"learning_rate": 3.985957766105693e-06,
"loss": 0.9966,
"mean_token_accuracy": 0.7596998423337936,
"num_tokens": 1536154451.0,
"step": 11800
},
{
"epoch": 0.6134331461842378,
"eval_entropy": 1.7927734375,
"eval_loss": 1.0311816930770874,
"eval_mean_token_accuracy": 0.7506178691983223,
"eval_num_tokens": 1536154451.0,
"eval_runtime": 27.0426,
"eval_samples_per_second": 93.704,
"eval_steps_per_second": 1.479,
"step": 11800
},
{
"entropy": 1.83515625,
"epoch": 0.6139530047826991,
"grad_norm": 3.220859830305546,
"learning_rate": 3.980598134848323e-06,
"loss": 1.035,
"mean_token_accuracy": 0.7513429075479507,
"num_tokens": 1537456831.0,
"step": 11810
},
{
"entropy": 1.81484375,
"epoch": 0.6144728633811604,
"grad_norm": 3.078338400972501,
"learning_rate": 3.975238503590954e-06,
"loss": 0.9503,
"mean_token_accuracy": 0.7722801983356475,
"num_tokens": 1538758847.0,
"step": 11820
},
{
"entropy": 1.812890625,
"epoch": 0.6149927219796215,
"grad_norm": 2.4500815035723917,
"learning_rate": 3.969878872333584e-06,
"loss": 1.0493,
"mean_token_accuracy": 0.7480691939592361,
"num_tokens": 1540060941.0,
"step": 11830
},
{
"entropy": 1.840625,
"epoch": 0.6155125805780828,
"grad_norm": 2.60669442282315,
"learning_rate": 3.964519241076215e-06,
"loss": 0.9835,
"mean_token_accuracy": 0.7605772107839585,
"num_tokens": 1541362621.0,
"step": 11840
},
{
"entropy": 1.821484375,
"epoch": 0.6160324391765439,
"grad_norm": 2.948475096054445,
"learning_rate": 3.959159609818845e-06,
"loss": 1.008,
"mean_token_accuracy": 0.7614253491163254,
"num_tokens": 1542665663.0,
"step": 11850
},
{
"entropy": 1.816015625,
"epoch": 0.6165522977750052,
"grad_norm": 3.0225874080606276,
"learning_rate": 3.953799978561476e-06,
"loss": 1.0088,
"mean_token_accuracy": 0.7550907373428345,
"num_tokens": 1543967306.0,
"step": 11860
},
{
"entropy": 1.807421875,
"epoch": 0.6170721563734665,
"grad_norm": 2.98625039312156,
"learning_rate": 3.948440347304105e-06,
"loss": 1.036,
"mean_token_accuracy": 0.7532598793506622,
"num_tokens": 1545268956.0,
"step": 11870
},
{
"entropy": 1.806640625,
"epoch": 0.6175920149719276,
"grad_norm": 3.0358869066587078,
"learning_rate": 3.943080716046736e-06,
"loss": 1.0298,
"mean_token_accuracy": 0.7508641660213471,
"num_tokens": 1546572034.0,
"step": 11880
},
{
"entropy": 1.79453125,
"epoch": 0.6181118735703889,
"grad_norm": 3.1415383630533578,
"learning_rate": 3.937721084789366e-06,
"loss": 1.0253,
"mean_token_accuracy": 0.7551880568265915,
"num_tokens": 1547876322.0,
"step": 11890
},
{
"entropy": 1.77265625,
"epoch": 0.61863173216885,
"grad_norm": 3.0191400308693086,
"learning_rate": 3.932361453531997e-06,
"loss": 0.9874,
"mean_token_accuracy": 0.7612119674682617,
"num_tokens": 1549178129.0,
"step": 11900
},
{
"epoch": 0.61863173216885,
"eval_entropy": 1.7669921875,
"eval_loss": 1.0256494283676147,
"eval_mean_token_accuracy": 0.7516138896346092,
"eval_num_tokens": 1549178129.0,
"eval_runtime": 27.0812,
"eval_samples_per_second": 93.57,
"eval_steps_per_second": 1.477,
"step": 11900
},
{
"entropy": 1.82109375,
"epoch": 0.6191515907673113,
"grad_norm": 2.889761533198714,
"learning_rate": 3.9270018222746274e-06,
"loss": 1.0821,
"mean_token_accuracy": 0.7425061732530593,
"num_tokens": 1550481450.0,
"step": 11910
},
{
"entropy": 1.81953125,
"epoch": 0.6196714493657726,
"grad_norm": 2.908263122501541,
"learning_rate": 3.921642191017258e-06,
"loss": 0.9678,
"mean_token_accuracy": 0.7677363365888595,
"num_tokens": 1551784077.0,
"step": 11920
},
{
"entropy": 1.788671875,
"epoch": 0.6201913079642337,
"grad_norm": 2.484165991595491,
"learning_rate": 3.9162825597598885e-06,
"loss": 1.009,
"mean_token_accuracy": 0.7555081486701966,
"num_tokens": 1553086045.0,
"step": 11930
},
{
"entropy": 1.8140625,
"epoch": 0.620711166562695,
"grad_norm": 3.123326764170852,
"learning_rate": 3.9109229285025195e-06,
"loss": 1.0175,
"mean_token_accuracy": 0.754459798336029,
"num_tokens": 1554387201.0,
"step": 11940
},
{
"entropy": 1.7890625,
"epoch": 0.6212310251611561,
"grad_norm": 2.560583293013844,
"learning_rate": 3.90556329724515e-06,
"loss": 0.9898,
"mean_token_accuracy": 0.7611349940299987,
"num_tokens": 1555691119.0,
"step": 11950
},
{
"entropy": 1.76953125,
"epoch": 0.6217508837596174,
"grad_norm": 2.511427657064107,
"learning_rate": 3.90020366598778e-06,
"loss": 0.9849,
"mean_token_accuracy": 0.7629994601011276,
"num_tokens": 1556992330.0,
"step": 11960
},
{
"entropy": 1.797265625,
"epoch": 0.6222707423580786,
"grad_norm": 2.7274216505008586,
"learning_rate": 3.894844034730411e-06,
"loss": 1.0188,
"mean_token_accuracy": 0.7521152317523956,
"num_tokens": 1558295673.0,
"step": 11970
},
{
"entropy": 1.745703125,
"epoch": 0.6227906009565398,
"grad_norm": 2.8183358103195957,
"learning_rate": 3.889484403473041e-06,
"loss": 0.9499,
"mean_token_accuracy": 0.7687581509351731,
"num_tokens": 1559600264.0,
"step": 11980
},
{
"entropy": 1.776953125,
"epoch": 0.6233104595550011,
"grad_norm": 2.62466060277136,
"learning_rate": 3.884124772215672e-06,
"loss": 1.0168,
"mean_token_accuracy": 0.756759199500084,
"num_tokens": 1560900037.0,
"step": 11990
},
{
"entropy": 1.772265625,
"epoch": 0.6238303181534622,
"grad_norm": 2.6979014543038993,
"learning_rate": 3.878765140958302e-06,
"loss": 1.0236,
"mean_token_accuracy": 0.7543421059846878,
"num_tokens": 1562200543.0,
"step": 12000
},
{
"epoch": 0.6238303181534622,
"eval_entropy": 1.7888671875,
"eval_loss": 1.0241879224777222,
"eval_mean_token_accuracy": 0.752562215924263,
"eval_num_tokens": 1562200543.0,
"eval_runtime": 27.0253,
"eval_samples_per_second": 93.764,
"eval_steps_per_second": 1.48,
"step": 12000
},
{
"entropy": 1.79453125,
"epoch": 0.6243501767519235,
"grad_norm": 2.811134872250385,
"learning_rate": 3.873405509700933e-06,
"loss": 0.9569,
"mean_token_accuracy": 0.7690840661525726,
"num_tokens": 1563501773.0,
"step": 12010
},
{
"entropy": 1.79296875,
"epoch": 0.6248700353503847,
"grad_norm": 2.595757392904427,
"learning_rate": 3.868045878443563e-06,
"loss": 1.0167,
"mean_token_accuracy": 0.7571112632751464,
"num_tokens": 1564803739.0,
"step": 12020
},
{
"entropy": 1.811328125,
"epoch": 0.6253898939488459,
"grad_norm": 2.6619772715556294,
"learning_rate": 3.862686247186194e-06,
"loss": 1.014,
"mean_token_accuracy": 0.7586792945861817,
"num_tokens": 1566106430.0,
"step": 12030
},
{
"entropy": 1.784765625,
"epoch": 0.6259097525473072,
"grad_norm": 2.8769393672800287,
"learning_rate": 3.857326615928824e-06,
"loss": 1.0298,
"mean_token_accuracy": 0.7508581846952438,
"num_tokens": 1567406552.0,
"step": 12040
},
{
"entropy": 1.781640625,
"epoch": 0.6264296111457683,
"grad_norm": 2.5773077802117745,
"learning_rate": 3.851966984671455e-06,
"loss": 0.9637,
"mean_token_accuracy": 0.7663023442029953,
"num_tokens": 1568705928.0,
"step": 12050
},
{
"entropy": 1.76875,
"epoch": 0.6269494697442296,
"grad_norm": 2.9748628703167106,
"learning_rate": 3.846607353414085e-06,
"loss": 1.0253,
"mean_token_accuracy": 0.7541603684425354,
"num_tokens": 1570005462.0,
"step": 12060
},
{
"entropy": 1.762109375,
"epoch": 0.6274693283426908,
"grad_norm": 2.8099876930970478,
"learning_rate": 3.841247722156715e-06,
"loss": 0.9612,
"mean_token_accuracy": 0.770962941646576,
"num_tokens": 1571308763.0,
"step": 12070
},
{
"entropy": 1.824609375,
"epoch": 0.627989186941152,
"grad_norm": 3.0171639965506767,
"learning_rate": 3.835888090899346e-06,
"loss": 1.0069,
"mean_token_accuracy": 0.7548595041036605,
"num_tokens": 1572611267.0,
"step": 12080
},
{
"entropy": 1.80390625,
"epoch": 0.6285090455396132,
"grad_norm": 2.859641440014087,
"learning_rate": 3.8305284596419764e-06,
"loss": 0.9722,
"mean_token_accuracy": 0.7676572561264038,
"num_tokens": 1573912961.0,
"step": 12090
},
{
"entropy": 1.782421875,
"epoch": 0.6290289041380744,
"grad_norm": 3.004514877715107,
"learning_rate": 3.825168828384607e-06,
"loss": 0.9766,
"mean_token_accuracy": 0.7679377377033234,
"num_tokens": 1575213147.0,
"step": 12100
},
{
"epoch": 0.6290289041380744,
"eval_entropy": 1.8033203125,
"eval_loss": 1.0208802223205566,
"eval_mean_token_accuracy": 0.7528571411967278,
"eval_num_tokens": 1575213147.0,
"eval_runtime": 27.0536,
"eval_samples_per_second": 93.666,
"eval_steps_per_second": 1.479,
"step": 12100
},
{
"entropy": 1.804296875,
"epoch": 0.6295487627365357,
"grad_norm": 2.5508888995901735,
"learning_rate": 3.8198091971272375e-06,
"loss": 0.9781,
"mean_token_accuracy": 0.7654912531375885,
"num_tokens": 1576516065.0,
"step": 12110
},
{
"entropy": 1.79609375,
"epoch": 0.6300686213349969,
"grad_norm": 2.6023695620236396,
"learning_rate": 3.814449565869868e-06,
"loss": 1.0218,
"mean_token_accuracy": 0.7538674443960189,
"num_tokens": 1577817818.0,
"step": 12120
},
{
"entropy": 1.777734375,
"epoch": 0.6305884799334581,
"grad_norm": 2.8305679654145024,
"learning_rate": 3.8090899346124986e-06,
"loss": 1.014,
"mean_token_accuracy": 0.7566417217254638,
"num_tokens": 1579121508.0,
"step": 12130
},
{
"entropy": 1.80234375,
"epoch": 0.6311083385319193,
"grad_norm": 2.6957701807516865,
"learning_rate": 3.803730303355129e-06,
"loss": 1.0538,
"mean_token_accuracy": 0.7495754212141037,
"num_tokens": 1580424234.0,
"step": 12140
},
{
"entropy": 1.769140625,
"epoch": 0.6316281971303805,
"grad_norm": 2.7237032004355424,
"learning_rate": 3.7983706720977597e-06,
"loss": 0.962,
"mean_token_accuracy": 0.7671535640954972,
"num_tokens": 1581727325.0,
"step": 12150
},
{
"entropy": 1.79765625,
"epoch": 0.6321480557288418,
"grad_norm": 3.3685074305930844,
"learning_rate": 3.7930110408403903e-06,
"loss": 1.0752,
"mean_token_accuracy": 0.7420872747898102,
"num_tokens": 1583027515.0,
"step": 12160
},
{
"entropy": 1.79609375,
"epoch": 0.632667914327303,
"grad_norm": 2.635075798264179,
"learning_rate": 3.787651409583021e-06,
"loss": 1.0729,
"mean_token_accuracy": 0.7421532332897186,
"num_tokens": 1584329032.0,
"step": 12170
},
{
"entropy": 1.814453125,
"epoch": 0.6331877729257642,
"grad_norm": 2.8358309335929452,
"learning_rate": 3.7822917783256514e-06,
"loss": 1.0205,
"mean_token_accuracy": 0.7549308776855469,
"num_tokens": 1585632143.0,
"step": 12180
},
{
"entropy": 1.793359375,
"epoch": 0.6337076315242254,
"grad_norm": 3.2053944222933826,
"learning_rate": 3.776932147068282e-06,
"loss": 1.0055,
"mean_token_accuracy": 0.7593362092971802,
"num_tokens": 1586934539.0,
"step": 12190
},
{
"entropy": 1.80546875,
"epoch": 0.6342274901226866,
"grad_norm": 3.093873124045496,
"learning_rate": 3.7715725158109125e-06,
"loss": 0.9941,
"mean_token_accuracy": 0.7608711153268815,
"num_tokens": 1588238334.0,
"step": 12200
},
{
"epoch": 0.6342274901226866,
"eval_entropy": 1.785546875,
"eval_loss": 1.020053505897522,
"eval_mean_token_accuracy": 0.7530313611030579,
"eval_num_tokens": 1588238334.0,
"eval_runtime": 26.9808,
"eval_samples_per_second": 93.919,
"eval_steps_per_second": 1.483,
"step": 12200
},
{
"entropy": 1.801171875,
"epoch": 0.6347473487211478,
"grad_norm": 2.858373685041135,
"learning_rate": 3.766212884553543e-06,
"loss": 1.056,
"mean_token_accuracy": 0.746307498216629,
"num_tokens": 1589541092.0,
"step": 12210
},
{
"entropy": 1.794140625,
"epoch": 0.6352672073196091,
"grad_norm": 2.4538193536346147,
"learning_rate": 3.7608532532961735e-06,
"loss": 1.0136,
"mean_token_accuracy": 0.7579671651124954,
"num_tokens": 1590839595.0,
"step": 12220
},
{
"entropy": 1.7890625,
"epoch": 0.6357870659180703,
"grad_norm": 2.7134255702897905,
"learning_rate": 3.7554936220388037e-06,
"loss": 0.9981,
"mean_token_accuracy": 0.7607898503541947,
"num_tokens": 1592143321.0,
"step": 12230
},
{
"entropy": 1.81015625,
"epoch": 0.6363069245165315,
"grad_norm": 2.89873910137971,
"learning_rate": 3.7501339907814342e-06,
"loss": 1.0052,
"mean_token_accuracy": 0.7577779799699783,
"num_tokens": 1593444073.0,
"step": 12240
},
{
"entropy": 1.837109375,
"epoch": 0.6368267831149927,
"grad_norm": 3.1937758721499074,
"learning_rate": 3.7447743595240648e-06,
"loss": 1.0514,
"mean_token_accuracy": 0.7463254243135452,
"num_tokens": 1594747787.0,
"step": 12250
},
{
"entropy": 1.832421875,
"epoch": 0.6373466417134539,
"grad_norm": 2.7634746052111088,
"learning_rate": 3.7394147282666953e-06,
"loss": 0.9678,
"mean_token_accuracy": 0.764824566245079,
"num_tokens": 1596049498.0,
"step": 12260
},
{
"entropy": 1.80390625,
"epoch": 0.6378665003119152,
"grad_norm": 2.4698086688236547,
"learning_rate": 3.734055097009326e-06,
"loss": 1.06,
"mean_token_accuracy": 0.7463922053575516,
"num_tokens": 1597353030.0,
"step": 12270
},
{
"entropy": 1.79765625,
"epoch": 0.6383863589103764,
"grad_norm": 2.6007948096791305,
"learning_rate": 3.7286954657519564e-06,
"loss": 1.0816,
"mean_token_accuracy": 0.7414164394140244,
"num_tokens": 1598654132.0,
"step": 12280
},
{
"entropy": 1.81484375,
"epoch": 0.6389062175088376,
"grad_norm": 2.554947571524958,
"learning_rate": 3.723335834494587e-06,
"loss": 0.9675,
"mean_token_accuracy": 0.7665542811155319,
"num_tokens": 1599956307.0,
"step": 12290
},
{
"entropy": 1.766015625,
"epoch": 0.6394260761072988,
"grad_norm": 3.0258144851802657,
"learning_rate": 3.7179762032372175e-06,
"loss": 0.9896,
"mean_token_accuracy": 0.7645464599132538,
"num_tokens": 1601257659.0,
"step": 12300
},
{
"epoch": 0.6394260761072988,
"eval_entropy": 1.7427734375,
"eval_loss": 1.0198103189468384,
"eval_mean_token_accuracy": 0.7532563760876656,
"eval_num_tokens": 1601257659.0,
"eval_runtime": 27.0776,
"eval_samples_per_second": 93.583,
"eval_steps_per_second": 1.477,
"step": 12300
},
{
"entropy": 1.796875,
"epoch": 0.63994593470576,
"grad_norm": 2.52854936242033,
"learning_rate": 3.712616571979848e-06,
"loss": 1.0086,
"mean_token_accuracy": 0.7559391587972641,
"num_tokens": 1602557231.0,
"step": 12310
},
{
"entropy": 1.8140625,
"epoch": 0.6404657933042213,
"grad_norm": 2.706858383831779,
"learning_rate": 3.7072569407224786e-06,
"loss": 0.9931,
"mean_token_accuracy": 0.7612469613552093,
"num_tokens": 1603858191.0,
"step": 12320
},
{
"entropy": 1.75703125,
"epoch": 0.6409856519026824,
"grad_norm": 2.9293271971284196,
"learning_rate": 3.701897309465109e-06,
"loss": 1.0265,
"mean_token_accuracy": 0.7521074116230011,
"num_tokens": 1605159827.0,
"step": 12330
},
{
"entropy": 1.790625,
"epoch": 0.6415055105011437,
"grad_norm": 3.049616363621405,
"learning_rate": 3.6965376782077393e-06,
"loss": 1.0173,
"mean_token_accuracy": 0.7562534123659134,
"num_tokens": 1606462889.0,
"step": 12340
},
{
"entropy": 1.81875,
"epoch": 0.642025369099605,
"grad_norm": 2.841792037089896,
"learning_rate": 3.69117804695037e-06,
"loss": 0.9232,
"mean_token_accuracy": 0.7766727089881897,
"num_tokens": 1607762616.0,
"step": 12350
},
{
"entropy": 1.834375,
"epoch": 0.6425452276980661,
"grad_norm": 2.607386246460394,
"learning_rate": 3.6858184156930004e-06,
"loss": 0.9834,
"mean_token_accuracy": 0.7649342387914657,
"num_tokens": 1609066200.0,
"step": 12360
},
{
"entropy": 1.82734375,
"epoch": 0.6430650862965274,
"grad_norm": 2.9785227979892603,
"learning_rate": 3.680458784435631e-06,
"loss": 1.0242,
"mean_token_accuracy": 0.7556058466434479,
"num_tokens": 1610365072.0,
"step": 12370
},
{
"entropy": 1.828515625,
"epoch": 0.6435849448949885,
"grad_norm": 3.026666404110663,
"learning_rate": 3.6750991531782614e-06,
"loss": 1.0111,
"mean_token_accuracy": 0.7580647945404053,
"num_tokens": 1611668675.0,
"step": 12380
},
{
"entropy": 1.7984375,
"epoch": 0.6441048034934498,
"grad_norm": 3.488912238659851,
"learning_rate": 3.669739521920892e-06,
"loss": 1.025,
"mean_token_accuracy": 0.7543676316738128,
"num_tokens": 1612967297.0,
"step": 12390
},
{
"entropy": 1.8265625,
"epoch": 0.644624662091911,
"grad_norm": 3.410607581558229,
"learning_rate": 3.6643798906635225e-06,
"loss": 0.9546,
"mean_token_accuracy": 0.7708420187234879,
"num_tokens": 1614267423.0,
"step": 12400
},
{
"epoch": 0.644624662091911,
"eval_entropy": 1.791015625,
"eval_loss": 1.0182867050170898,
"eval_mean_token_accuracy": 0.7532566413283348,
"eval_num_tokens": 1614267423.0,
"eval_runtime": 27.0071,
"eval_samples_per_second": 93.827,
"eval_steps_per_second": 1.481,
"step": 12400
},
{
"entropy": 1.7765625,
"epoch": 0.6451445206903722,
"grad_norm": 2.6958936957194326,
"learning_rate": 3.659020259406153e-06,
"loss": 1.0409,
"mean_token_accuracy": 0.7499320805072784,
"num_tokens": 1615569735.0,
"step": 12410
},
{
"entropy": 1.80546875,
"epoch": 0.6456643792888335,
"grad_norm": 2.434041451690604,
"learning_rate": 3.6536606281487836e-06,
"loss": 1.0801,
"mean_token_accuracy": 0.7437429428100586,
"num_tokens": 1616873176.0,
"step": 12420
},
{
"entropy": 1.815234375,
"epoch": 0.6461842378872946,
"grad_norm": 3.0262905530473168,
"learning_rate": 3.648300996891414e-06,
"loss": 0.9982,
"mean_token_accuracy": 0.7602374374866485,
"num_tokens": 1618175934.0,
"step": 12430
},
{
"entropy": 1.78828125,
"epoch": 0.6467040964857559,
"grad_norm": 2.614404560277402,
"learning_rate": 3.6429413656340447e-06,
"loss": 1.0267,
"mean_token_accuracy": 0.7531673222780227,
"num_tokens": 1619478883.0,
"step": 12440
},
{
"entropy": 1.755859375,
"epoch": 0.647223955084217,
"grad_norm": 2.704970906144797,
"learning_rate": 3.637581734376675e-06,
"loss": 1.0132,
"mean_token_accuracy": 0.7564902305603027,
"num_tokens": 1620780400.0,
"step": 12450
},
{
"entropy": 1.776171875,
"epoch": 0.6477438136826783,
"grad_norm": 2.747534027220895,
"learning_rate": 3.6322221031193054e-06,
"loss": 1.0027,
"mean_token_accuracy": 0.75719433426857,
"num_tokens": 1622080255.0,
"step": 12460
},
{
"entropy": 1.771484375,
"epoch": 0.6482636722811396,
"grad_norm": 3.010639892455914,
"learning_rate": 3.626862471861936e-06,
"loss": 1.0164,
"mean_token_accuracy": 0.7553359061479569,
"num_tokens": 1623383004.0,
"step": 12470
},
{
"entropy": 1.76875,
"epoch": 0.6487835308796007,
"grad_norm": 3.2821696184317775,
"learning_rate": 3.6215028406045665e-06,
"loss": 1.0134,
"mean_token_accuracy": 0.7558949142694473,
"num_tokens": 1624683531.0,
"step": 12480
},
{
"entropy": 1.773046875,
"epoch": 0.649303389478062,
"grad_norm": 2.3485632912557053,
"learning_rate": 3.616143209347197e-06,
"loss": 0.9853,
"mean_token_accuracy": 0.763113534450531,
"num_tokens": 1625986507.0,
"step": 12490
},
{
"entropy": 1.755078125,
"epoch": 0.6498232480765231,
"grad_norm": 2.6196083565306383,
"learning_rate": 3.6107835780898276e-06,
"loss": 0.9945,
"mean_token_accuracy": 0.7614129215478898,
"num_tokens": 1627285430.0,
"step": 12500
},
{
"epoch": 0.6498232480765231,
"eval_entropy": 1.7650390625,
"eval_loss": 1.0163224935531616,
"eval_mean_token_accuracy": 0.7537347123026847,
"eval_num_tokens": 1627285430.0,
"eval_runtime": 27.0704,
"eval_samples_per_second": 93.608,
"eval_steps_per_second": 1.478,
"step": 12500
},
{
"entropy": 1.821484375,
"epoch": 0.6503431066749844,
"grad_norm": 2.9526619706085997,
"learning_rate": 3.605423946832458e-06,
"loss": 1.0198,
"mean_token_accuracy": 0.7547599822282791,
"num_tokens": 1628589366.0,
"step": 12510
},
{
"entropy": 1.787890625,
"epoch": 0.6508629652734457,
"grad_norm": 2.8656203876814095,
"learning_rate": 3.6000643155750887e-06,
"loss": 0.969,
"mean_token_accuracy": 0.7655649513006211,
"num_tokens": 1629890429.0,
"step": 12520
},
{
"entropy": 1.785546875,
"epoch": 0.6513828238719068,
"grad_norm": 2.5374855923283164,
"learning_rate": 3.5947046843177192e-06,
"loss": 1.0359,
"mean_token_accuracy": 0.7498370200395584,
"num_tokens": 1631193024.0,
"step": 12530
},
{
"entropy": 1.751171875,
"epoch": 0.6519026824703681,
"grad_norm": 2.7659433180874644,
"learning_rate": 3.5893450530603498e-06,
"loss": 1.0094,
"mean_token_accuracy": 0.7576611459255218,
"num_tokens": 1632494928.0,
"step": 12540
},
{
"entropy": 1.789453125,
"epoch": 0.6524225410688292,
"grad_norm": 2.8978781513694636,
"learning_rate": 3.5839854218029803e-06,
"loss": 0.97,
"mean_token_accuracy": 0.7657379180192947,
"num_tokens": 1633794955.0,
"step": 12550
},
{
"entropy": 1.773046875,
"epoch": 0.6529423996672905,
"grad_norm": 2.4674061224227173,
"learning_rate": 3.578625790545611e-06,
"loss": 0.952,
"mean_token_accuracy": 0.7688183009624481,
"num_tokens": 1635097458.0,
"step": 12560
},
{
"entropy": 1.777734375,
"epoch": 0.6534622582657517,
"grad_norm": 3.0342969624024714,
"learning_rate": 3.573266159288241e-06,
"loss": 1.006,
"mean_token_accuracy": 0.7587312906980515,
"num_tokens": 1636399402.0,
"step": 12570
},
{
"entropy": 1.8296875,
"epoch": 0.6539821168642129,
"grad_norm": 3.246987632008159,
"learning_rate": 3.5679065280308715e-06,
"loss": 1.0401,
"mean_token_accuracy": 0.7515109479427338,
"num_tokens": 1637700904.0,
"step": 12580
},
{
"entropy": 1.79140625,
"epoch": 0.6545019754626742,
"grad_norm": 2.888820702278454,
"learning_rate": 3.562546896773502e-06,
"loss": 0.9441,
"mean_token_accuracy": 0.7755965083837509,
"num_tokens": 1639003664.0,
"step": 12590
},
{
"entropy": 1.80546875,
"epoch": 0.6550218340611353,
"grad_norm": 2.923825204589134,
"learning_rate": 3.5571872655161326e-06,
"loss": 0.9751,
"mean_token_accuracy": 0.7632757097482681,
"num_tokens": 1640305426.0,
"step": 12600
},
{
"epoch": 0.6550218340611353,
"eval_entropy": 1.7732421875,
"eval_loss": 1.0136444568634033,
"eval_mean_token_accuracy": 0.754615193605423,
"eval_num_tokens": 1640305426.0,
"eval_runtime": 27.0386,
"eval_samples_per_second": 93.718,
"eval_steps_per_second": 1.479,
"step": 12600
},
{
"entropy": 1.7734375,
"epoch": 0.6555416926595966,
"grad_norm": 3.277395312890085,
"learning_rate": 3.551827634258763e-06,
"loss": 0.9497,
"mean_token_accuracy": 0.7716007739305496,
"num_tokens": 1641607340.0,
"step": 12610
},
{
"entropy": 1.758984375,
"epoch": 0.6560615512580578,
"grad_norm": 2.4683529056797164,
"learning_rate": 3.5464680030013937e-06,
"loss": 0.9974,
"mean_token_accuracy": 0.7625887840986252,
"num_tokens": 1642910392.0,
"step": 12620
},
{
"entropy": 1.79296875,
"epoch": 0.656581409856519,
"grad_norm": 2.9279900304671953,
"learning_rate": 3.5411083717440243e-06,
"loss": 0.9905,
"mean_token_accuracy": 0.7615129023790359,
"num_tokens": 1644213463.0,
"step": 12630
},
{
"entropy": 1.79609375,
"epoch": 0.6571012684549803,
"grad_norm": 3.1947150829975812,
"learning_rate": 3.535748740486655e-06,
"loss": 0.9893,
"mean_token_accuracy": 0.7608874708414077,
"num_tokens": 1645513495.0,
"step": 12640
},
{
"entropy": 1.809765625,
"epoch": 0.6576211270534414,
"grad_norm": 2.4248940500672287,
"learning_rate": 3.5303891092292854e-06,
"loss": 1.0139,
"mean_token_accuracy": 0.7542876422405242,
"num_tokens": 1646814182.0,
"step": 12650
},
{
"entropy": 1.784375,
"epoch": 0.6581409856519027,
"grad_norm": 2.7366117846969407,
"learning_rate": 3.525029477971916e-06,
"loss": 0.9577,
"mean_token_accuracy": 0.7702014178037644,
"num_tokens": 1648115775.0,
"step": 12660
},
{
"entropy": 1.767578125,
"epoch": 0.6586608442503639,
"grad_norm": 3.003590922479666,
"learning_rate": 3.5196698467145464e-06,
"loss": 0.9842,
"mean_token_accuracy": 0.7629594922065734,
"num_tokens": 1649418464.0,
"step": 12670
},
{
"entropy": 1.80546875,
"epoch": 0.6591807028488251,
"grad_norm": 2.6124311560214766,
"learning_rate": 3.5143102154571766e-06,
"loss": 1.0509,
"mean_token_accuracy": 0.7496615529060364,
"num_tokens": 1650720502.0,
"step": 12680
},
{
"entropy": 1.805078125,
"epoch": 0.6597005614472864,
"grad_norm": 2.4807185817262414,
"learning_rate": 3.508950584199807e-06,
"loss": 1.0127,
"mean_token_accuracy": 0.7566198647022248,
"num_tokens": 1652022580.0,
"step": 12690
},
{
"entropy": 1.792578125,
"epoch": 0.6602204200457475,
"grad_norm": 3.146416166150857,
"learning_rate": 3.5035909529424377e-06,
"loss": 0.9994,
"mean_token_accuracy": 0.759733310341835,
"num_tokens": 1653323070.0,
"step": 12700
},
{
"epoch": 0.6602204200457475,
"eval_entropy": 1.7654296875,
"eval_loss": 1.0121417045593262,
"eval_mean_token_accuracy": 0.7551995113492012,
"eval_num_tokens": 1653323070.0,
"eval_runtime": 27.0329,
"eval_samples_per_second": 93.738,
"eval_steps_per_second": 1.48,
"step": 12700
},
{
"entropy": 1.790625,
"epoch": 0.6607402786442088,
"grad_norm": 2.7429798250146367,
"learning_rate": 3.498231321685068e-06,
"loss": 1.0083,
"mean_token_accuracy": 0.7574114143848419,
"num_tokens": 1654624133.0,
"step": 12710
},
{
"entropy": 1.795703125,
"epoch": 0.66126013724267,
"grad_norm": 3.2311438466480165,
"learning_rate": 3.4928716904276988e-06,
"loss": 0.9579,
"mean_token_accuracy": 0.769022011756897,
"num_tokens": 1655925702.0,
"step": 12720
},
{
"entropy": 1.760546875,
"epoch": 0.6617799958411312,
"grad_norm": 2.8356650201544262,
"learning_rate": 3.4875120591703293e-06,
"loss": 1.0099,
"mean_token_accuracy": 0.7556679487228394,
"num_tokens": 1657229369.0,
"step": 12730
},
{
"entropy": 1.78671875,
"epoch": 0.6622998544395924,
"grad_norm": 2.931601903697856,
"learning_rate": 3.48215242791296e-06,
"loss": 0.9811,
"mean_token_accuracy": 0.7638945281505585,
"num_tokens": 1658531237.0,
"step": 12740
},
{
"entropy": 1.766796875,
"epoch": 0.6628197130380536,
"grad_norm": 3.0512320028000604,
"learning_rate": 3.4767927966555904e-06,
"loss": 0.9408,
"mean_token_accuracy": 0.769646617770195,
"num_tokens": 1659832850.0,
"step": 12750
},
{
"entropy": 1.7734375,
"epoch": 0.6633395716365149,
"grad_norm": 3.197430410786029,
"learning_rate": 3.471433165398221e-06,
"loss": 1.0627,
"mean_token_accuracy": 0.7447639048099518,
"num_tokens": 1661133311.0,
"step": 12760
},
{
"entropy": 1.8015625,
"epoch": 0.6638594302349761,
"grad_norm": 3.1308425389784187,
"learning_rate": 3.4660735341408515e-06,
"loss": 0.9584,
"mean_token_accuracy": 0.767105957865715,
"num_tokens": 1662436122.0,
"step": 12770
},
{
"entropy": 1.770703125,
"epoch": 0.6643792888334373,
"grad_norm": 2.8749963262567726,
"learning_rate": 3.460713902883482e-06,
"loss": 0.9239,
"mean_token_accuracy": 0.7775328665971756,
"num_tokens": 1663735728.0,
"step": 12780
},
{
"entropy": 1.81796875,
"epoch": 0.6648991474318985,
"grad_norm": 3.4796132062540948,
"learning_rate": 3.455354271626112e-06,
"loss": 0.9316,
"mean_token_accuracy": 0.7762039065361023,
"num_tokens": 1665038515.0,
"step": 12790
},
{
"entropy": 1.772265625,
"epoch": 0.6654190060303597,
"grad_norm": 2.5259496033661573,
"learning_rate": 3.4499946403687427e-06,
"loss": 0.9617,
"mean_token_accuracy": 0.7649107873439789,
"num_tokens": 1666342099.0,
"step": 12800
},
{
"epoch": 0.6654190060303597,
"eval_entropy": 1.7435546875,
"eval_loss": 1.0114469528198242,
"eval_mean_token_accuracy": 0.7552617222070694,
"eval_num_tokens": 1666342099.0,
"eval_runtime": 27.0038,
"eval_samples_per_second": 93.839,
"eval_steps_per_second": 1.481,
"step": 12800
},
{
"entropy": 1.769140625,
"epoch": 0.665938864628821,
"grad_norm": 3.166739087082146,
"learning_rate": 3.4446350091113733e-06,
"loss": 0.9925,
"mean_token_accuracy": 0.7607234507799149,
"num_tokens": 1667641349.0,
"step": 12810
},
{
"entropy": 1.7953125,
"epoch": 0.6664587232272822,
"grad_norm": 2.7571139655863477,
"learning_rate": 3.439275377854004e-06,
"loss": 1.0123,
"mean_token_accuracy": 0.759319058060646,
"num_tokens": 1668943892.0,
"step": 12820
},
{
"entropy": 1.743359375,
"epoch": 0.6669785818257434,
"grad_norm": 2.995189330919771,
"learning_rate": 3.4339157465966343e-06,
"loss": 0.9921,
"mean_token_accuracy": 0.7610234975814819,
"num_tokens": 1670245789.0,
"step": 12830
},
{
"entropy": 1.791796875,
"epoch": 0.6674984404242046,
"grad_norm": 2.957331120472829,
"learning_rate": 3.428556115339265e-06,
"loss": 0.9994,
"mean_token_accuracy": 0.7603914439678192,
"num_tokens": 1671548272.0,
"step": 12840
},
{
"entropy": 1.776953125,
"epoch": 0.6680182990226659,
"grad_norm": 3.1211388580533868,
"learning_rate": 3.4231964840818954e-06,
"loss": 0.9285,
"mean_token_accuracy": 0.7739948779344559,
"num_tokens": 1672851019.0,
"step": 12850
},
{
"entropy": 1.74296875,
"epoch": 0.668538157621127,
"grad_norm": 2.7404516981385956,
"learning_rate": 3.417836852824526e-06,
"loss": 0.9605,
"mean_token_accuracy": 0.7688549607992172,
"num_tokens": 1674155484.0,
"step": 12860
},
{
"entropy": 1.759765625,
"epoch": 0.6690580162195883,
"grad_norm": 2.7459598158889307,
"learning_rate": 3.4124772215671565e-06,
"loss": 1.0188,
"mean_token_accuracy": 0.7559636861085892,
"num_tokens": 1675459150.0,
"step": 12870
},
{
"entropy": 1.789453125,
"epoch": 0.6695778748180495,
"grad_norm": 2.9250242863931555,
"learning_rate": 3.407117590309787e-06,
"loss": 1.0122,
"mean_token_accuracy": 0.7556913822889328,
"num_tokens": 1676761432.0,
"step": 12880
},
{
"entropy": 1.784375,
"epoch": 0.6700977334165107,
"grad_norm": 2.707235937681867,
"learning_rate": 3.4017579590524176e-06,
"loss": 1.0168,
"mean_token_accuracy": 0.7559157639741898,
"num_tokens": 1678064210.0,
"step": 12890
},
{
"entropy": 1.814453125,
"epoch": 0.670617592014972,
"grad_norm": 3.063238794391883,
"learning_rate": 3.396398327795048e-06,
"loss": 1.0048,
"mean_token_accuracy": 0.7594418108463288,
"num_tokens": 1679365374.0,
"step": 12900
},
{
"epoch": 0.670617592014972,
"eval_entropy": 1.7689453125,
"eval_loss": 1.0096248388290405,
"eval_mean_token_accuracy": 0.7557065472006798,
"eval_num_tokens": 1679365374.0,
"eval_runtime": 27.0186,
"eval_samples_per_second": 93.787,
"eval_steps_per_second": 1.48,
"step": 12900
},
{
"entropy": 1.794921875,
"epoch": 0.6711374506134331,
"grad_norm": 2.581142173285352,
"learning_rate": 3.3910386965376783e-06,
"loss": 1.038,
"mean_token_accuracy": 0.7489736109972001,
"num_tokens": 1680665144.0,
"step": 12910
},
{
"entropy": 1.786328125,
"epoch": 0.6716573092118944,
"grad_norm": 3.255077398656469,
"learning_rate": 3.385679065280309e-06,
"loss": 1.0103,
"mean_token_accuracy": 0.7577112555503845,
"num_tokens": 1681967014.0,
"step": 12920
},
{
"entropy": 1.772265625,
"epoch": 0.6721771678103556,
"grad_norm": 2.8259588218007177,
"learning_rate": 3.3803194340229394e-06,
"loss": 0.9071,
"mean_token_accuracy": 0.779361230134964,
"num_tokens": 1683268440.0,
"step": 12930
},
{
"entropy": 1.776171875,
"epoch": 0.6726970264088168,
"grad_norm": 2.966646490772424,
"learning_rate": 3.37495980276557e-06,
"loss": 0.9531,
"mean_token_accuracy": 0.7699274063110352,
"num_tokens": 1684568477.0,
"step": 12940
},
{
"entropy": 1.781640625,
"epoch": 0.673216885007278,
"grad_norm": 2.6693233274300234,
"learning_rate": 3.3696001715082005e-06,
"loss": 1.0133,
"mean_token_accuracy": 0.7550119251012802,
"num_tokens": 1685870409.0,
"step": 12950
},
{
"entropy": 1.79296875,
"epoch": 0.6737367436057392,
"grad_norm": 2.705197325182112,
"learning_rate": 3.364240540250831e-06,
"loss": 0.9879,
"mean_token_accuracy": 0.7618043690919876,
"num_tokens": 1687172658.0,
"step": 12960
},
{
"entropy": 1.78984375,
"epoch": 0.6742566022042005,
"grad_norm": 2.9644499643670716,
"learning_rate": 3.3588809089934616e-06,
"loss": 0.9857,
"mean_token_accuracy": 0.7637134939432144,
"num_tokens": 1688474403.0,
"step": 12970
},
{
"entropy": 1.775390625,
"epoch": 0.6747764608026616,
"grad_norm": 2.7687345375288053,
"learning_rate": 3.353521277736092e-06,
"loss": 0.9077,
"mean_token_accuracy": 0.7780576944351196,
"num_tokens": 1689776645.0,
"step": 12980
},
{
"entropy": 1.787109375,
"epoch": 0.6752963194011229,
"grad_norm": 3.2250558234939817,
"learning_rate": 3.3481616464787227e-06,
"loss": 1.0147,
"mean_token_accuracy": 0.7581279784440994,
"num_tokens": 1691076924.0,
"step": 12990
},
{
"entropy": 1.74609375,
"epoch": 0.6758161779995842,
"grad_norm": 3.416549158441902,
"learning_rate": 3.3428020152213532e-06,
"loss": 0.9432,
"mean_token_accuracy": 0.7734142482280731,
"num_tokens": 1692378420.0,
"step": 13000
},
{
"epoch": 0.6758161779995842,
"eval_entropy": 1.7544921875,
"eval_loss": 1.0061640739440918,
"eval_mean_token_accuracy": 0.7560686692595482,
"eval_num_tokens": 1692378420.0,
"eval_runtime": 26.9682,
"eval_samples_per_second": 93.963,
"eval_steps_per_second": 1.483,
"step": 13000
},
{
"entropy": 1.75859375,
"epoch": 0.6763360365980453,
"grad_norm": 2.781383490774206,
"learning_rate": 3.3374423839639838e-06,
"loss": 0.9383,
"mean_token_accuracy": 0.7719745188951492,
"num_tokens": 1693680954.0,
"step": 13010
},
{
"entropy": 1.776953125,
"epoch": 0.6768558951965066,
"grad_norm": 2.3815048803387446,
"learning_rate": 3.332082752706614e-06,
"loss": 0.9523,
"mean_token_accuracy": 0.7709066241979599,
"num_tokens": 1694983800.0,
"step": 13020
},
{
"entropy": 1.7515625,
"epoch": 0.6773757537949677,
"grad_norm": 2.849821879859067,
"learning_rate": 3.3267231214492444e-06,
"loss": 1.0699,
"mean_token_accuracy": 0.7445438712835312,
"num_tokens": 1696287111.0,
"step": 13030
},
{
"entropy": 1.7890625,
"epoch": 0.677895612393429,
"grad_norm": 2.98381926218252,
"learning_rate": 3.321363490191875e-06,
"loss": 0.9922,
"mean_token_accuracy": 0.762668851017952,
"num_tokens": 1697588185.0,
"step": 13040
},
{
"entropy": 1.772265625,
"epoch": 0.6784154709918903,
"grad_norm": 3.2306964828384843,
"learning_rate": 3.3160038589345055e-06,
"loss": 1.0488,
"mean_token_accuracy": 0.7523745864629745,
"num_tokens": 1698891091.0,
"step": 13050
},
{
"entropy": 1.782421875,
"epoch": 0.6789353295903514,
"grad_norm": 2.856939831775827,
"learning_rate": 3.310644227677136e-06,
"loss": 1.0226,
"mean_token_accuracy": 0.7568733483552933,
"num_tokens": 1700191133.0,
"step": 13060
},
{
"entropy": 1.808984375,
"epoch": 0.6794551881888127,
"grad_norm": 2.682148805584655,
"learning_rate": 3.3052845964197666e-06,
"loss": 1.044,
"mean_token_accuracy": 0.7518608838319778,
"num_tokens": 1701492643.0,
"step": 13070
},
{
"entropy": 1.8125,
"epoch": 0.6799750467872738,
"grad_norm": 2.4940478205235626,
"learning_rate": 3.299924965162397e-06,
"loss": 0.9438,
"mean_token_accuracy": 0.7732564002275467,
"num_tokens": 1702796342.0,
"step": 13080
},
{
"entropy": 1.798828125,
"epoch": 0.6804949053857351,
"grad_norm": 2.5200883970200247,
"learning_rate": 3.2945653339050277e-06,
"loss": 0.9702,
"mean_token_accuracy": 0.7643312066793442,
"num_tokens": 1704098075.0,
"step": 13090
},
{
"entropy": 1.79921875,
"epoch": 0.6810147639841962,
"grad_norm": 2.985025285447716,
"learning_rate": 3.2892057026476583e-06,
"loss": 1.009,
"mean_token_accuracy": 0.7562459111213684,
"num_tokens": 1705398960.0,
"step": 13100
},
{
"epoch": 0.6810147639841962,
"eval_entropy": 1.758203125,
"eval_loss": 1.0056779384613037,
"eval_mean_token_accuracy": 0.7563122555613517,
"eval_num_tokens": 1705398960.0,
"eval_runtime": 27.0288,
"eval_samples_per_second": 93.752,
"eval_steps_per_second": 1.48,
"step": 13100
},
{
"entropy": 1.78671875,
"epoch": 0.6815346225826575,
"grad_norm": 2.8469386765567086,
"learning_rate": 3.283846071390289e-06,
"loss": 1.0346,
"mean_token_accuracy": 0.7541121751070022,
"num_tokens": 1706698230.0,
"step": 13110
},
{
"entropy": 1.7921875,
"epoch": 0.6820544811811188,
"grad_norm": 2.7092987045539525,
"learning_rate": 3.2784864401329193e-06,
"loss": 0.9453,
"mean_token_accuracy": 0.7723611682653427,
"num_tokens": 1708000575.0,
"step": 13120
},
{
"entropy": 1.7640625,
"epoch": 0.6825743397795799,
"grad_norm": 3.514660775597032,
"learning_rate": 3.2731268088755495e-06,
"loss": 0.9335,
"mean_token_accuracy": 0.7735237210988999,
"num_tokens": 1709303325.0,
"step": 13130
},
{
"entropy": 1.8234375,
"epoch": 0.6830941983780412,
"grad_norm": 2.3840023832560013,
"learning_rate": 3.26776717761818e-06,
"loss": 0.9649,
"mean_token_accuracy": 0.7672140568494796,
"num_tokens": 1710604478.0,
"step": 13140
},
{
"entropy": 1.808203125,
"epoch": 0.6836140569765023,
"grad_norm": 2.839192714976561,
"learning_rate": 3.2624075463608106e-06,
"loss": 0.9764,
"mean_token_accuracy": 0.7646079808473587,
"num_tokens": 1711905916.0,
"step": 13150
},
{
"entropy": 1.792578125,
"epoch": 0.6841339155749636,
"grad_norm": 2.781305747387911,
"learning_rate": 3.257047915103441e-06,
"loss": 1.0484,
"mean_token_accuracy": 0.74795863032341,
"num_tokens": 1713206757.0,
"step": 13160
},
{
"entropy": 1.804296875,
"epoch": 0.6846537741734249,
"grad_norm": 2.6217601818155476,
"learning_rate": 3.2516882838460717e-06,
"loss": 0.9272,
"mean_token_accuracy": 0.779311153292656,
"num_tokens": 1714509626.0,
"step": 13170
},
{
"entropy": 1.783984375,
"epoch": 0.685173632771886,
"grad_norm": 3.721280552166602,
"learning_rate": 3.246328652588702e-06,
"loss": 0.9245,
"mean_token_accuracy": 0.775540879368782,
"num_tokens": 1715812017.0,
"step": 13180
},
{
"entropy": 1.769921875,
"epoch": 0.6856934913703473,
"grad_norm": 3.022709699924594,
"learning_rate": 3.2409690213313328e-06,
"loss": 0.9929,
"mean_token_accuracy": 0.7655958384275436,
"num_tokens": 1717112676.0,
"step": 13190
},
{
"entropy": 1.798046875,
"epoch": 0.6862133499688085,
"grad_norm": 2.6647166564443165,
"learning_rate": 3.2356093900739633e-06,
"loss": 1.0321,
"mean_token_accuracy": 0.7517904877662659,
"num_tokens": 1718413503.0,
"step": 13200
},
{
"epoch": 0.6862133499688085,
"eval_entropy": 1.762890625,
"eval_loss": 1.0025354623794556,
"eval_mean_token_accuracy": 0.7570832967758179,
"eval_num_tokens": 1718413503.0,
"eval_runtime": 27.0363,
"eval_samples_per_second": 93.726,
"eval_steps_per_second": 1.479,
"step": 13200
},
{
"entropy": 1.759375,
"epoch": 0.6867332085672697,
"grad_norm": 2.213741072107955,
"learning_rate": 3.230249758816594e-06,
"loss": 0.9438,
"mean_token_accuracy": 0.7708381950855255,
"num_tokens": 1719712871.0,
"step": 13210
},
{
"entropy": 1.750390625,
"epoch": 0.6872530671657309,
"grad_norm": 2.9779357669494435,
"learning_rate": 3.2248901275592244e-06,
"loss": 0.9119,
"mean_token_accuracy": 0.7803842961788178,
"num_tokens": 1721014685.0,
"step": 13220
},
{
"entropy": 1.76484375,
"epoch": 0.6877729257641921,
"grad_norm": 2.722135730986077,
"learning_rate": 3.219530496301855e-06,
"loss": 1.0369,
"mean_token_accuracy": 0.7516238868236542,
"num_tokens": 1722318023.0,
"step": 13230
},
{
"entropy": 1.78984375,
"epoch": 0.6882927843626534,
"grad_norm": 3.0317458582876693,
"learning_rate": 3.2141708650444855e-06,
"loss": 0.9635,
"mean_token_accuracy": 0.7701661884784698,
"num_tokens": 1723619866.0,
"step": 13240
},
{
"entropy": 1.7578125,
"epoch": 0.6888126429611146,
"grad_norm": 2.9715592844217853,
"learning_rate": 3.2088112337871156e-06,
"loss": 1.004,
"mean_token_accuracy": 0.7579945117235184,
"num_tokens": 1724921560.0,
"step": 13250
},
{
"entropy": 1.776171875,
"epoch": 0.6893325015595758,
"grad_norm": 3.4028358553903915,
"learning_rate": 3.203451602529746e-06,
"loss": 1.0311,
"mean_token_accuracy": 0.750464791059494,
"num_tokens": 1726225601.0,
"step": 13260
},
{
"entropy": 1.77890625,
"epoch": 0.689852360158037,
"grad_norm": 2.8532671633517293,
"learning_rate": 3.1980919712723767e-06,
"loss": 0.9397,
"mean_token_accuracy": 0.7735306054353714,
"num_tokens": 1727529974.0,
"step": 13270
},
{
"entropy": 1.772265625,
"epoch": 0.6903722187564982,
"grad_norm": 2.4070567819346493,
"learning_rate": 3.1927323400150072e-06,
"loss": 1.0169,
"mean_token_accuracy": 0.7547536939382553,
"num_tokens": 1728831254.0,
"step": 13280
},
{
"entropy": 1.75703125,
"epoch": 0.6908920773549595,
"grad_norm": 3.0506336788474644,
"learning_rate": 3.187372708757638e-06,
"loss": 0.9953,
"mean_token_accuracy": 0.7575918108224868,
"num_tokens": 1730131259.0,
"step": 13290
},
{
"entropy": 1.753515625,
"epoch": 0.6914119359534207,
"grad_norm": 2.780346115997765,
"learning_rate": 3.1820130775002683e-06,
"loss": 1.0094,
"mean_token_accuracy": 0.7575051188468933,
"num_tokens": 1731431714.0,
"step": 13300
},
{
"epoch": 0.6914119359534207,
"eval_entropy": 1.7630859375,
"eval_loss": 1.0001399517059326,
"eval_mean_token_accuracy": 0.7577238604426384,
"eval_num_tokens": 1731431714.0,
"eval_runtime": 27.0173,
"eval_samples_per_second": 93.792,
"eval_steps_per_second": 1.481,
"step": 13300
},
{
"entropy": 1.7640625,
"epoch": 0.6919317945518819,
"grad_norm": 2.855555292788476,
"learning_rate": 3.176653446242899e-06,
"loss": 0.9812,
"mean_token_accuracy": 0.762491625547409,
"num_tokens": 1732733996.0,
"step": 13310
},
{
"entropy": 1.749609375,
"epoch": 0.6924516531503431,
"grad_norm": 2.610440185709792,
"learning_rate": 3.1712938149855294e-06,
"loss": 0.9575,
"mean_token_accuracy": 0.7711554527282715,
"num_tokens": 1734034241.0,
"step": 13320
},
{
"entropy": 1.76953125,
"epoch": 0.6929715117488043,
"grad_norm": 2.6428264937168913,
"learning_rate": 3.16593418372816e-06,
"loss": 0.954,
"mean_token_accuracy": 0.768805456161499,
"num_tokens": 1735335665.0,
"step": 13330
},
{
"entropy": 1.748828125,
"epoch": 0.6934913703472655,
"grad_norm": 3.0202071281522755,
"learning_rate": 3.1605745524707905e-06,
"loss": 0.9894,
"mean_token_accuracy": 0.7636685878038406,
"num_tokens": 1736634338.0,
"step": 13340
},
{
"entropy": 1.7296875,
"epoch": 0.6940112289457268,
"grad_norm": 2.8131195874749073,
"learning_rate": 3.155214921213421e-06,
"loss": 0.9341,
"mean_token_accuracy": 0.7729353904724121,
"num_tokens": 1737934852.0,
"step": 13350
},
{
"entropy": 1.71953125,
"epoch": 0.694531087544188,
"grad_norm": 2.955224341180588,
"learning_rate": 3.149855289956051e-06,
"loss": 0.9126,
"mean_token_accuracy": 0.7797635644674301,
"num_tokens": 1739235622.0,
"step": 13360
},
{
"entropy": 1.73828125,
"epoch": 0.6950509461426492,
"grad_norm": 2.733503765307635,
"learning_rate": 3.1444956586986817e-06,
"loss": 0.9752,
"mean_token_accuracy": 0.7676136583089829,
"num_tokens": 1740538005.0,
"step": 13370
},
{
"entropy": 1.720703125,
"epoch": 0.6955708047411104,
"grad_norm": 3.106542457945049,
"learning_rate": 3.1391360274413123e-06,
"loss": 0.9851,
"mean_token_accuracy": 0.7629240572452545,
"num_tokens": 1741840475.0,
"step": 13380
},
{
"entropy": 1.74453125,
"epoch": 0.6960906633395716,
"grad_norm": 2.7037298868937873,
"learning_rate": 3.133776396183943e-06,
"loss": 0.9778,
"mean_token_accuracy": 0.7650624126195907,
"num_tokens": 1743143268.0,
"step": 13390
},
{
"entropy": 1.773828125,
"epoch": 0.6966105219380329,
"grad_norm": 2.8457453403612574,
"learning_rate": 3.1284167649265734e-06,
"loss": 0.9413,
"mean_token_accuracy": 0.7752017498016357,
"num_tokens": 1744445260.0,
"step": 13400
},
{
"epoch": 0.6966105219380329,
"eval_entropy": 1.7458984375,
"eval_loss": 0.9991111159324646,
"eval_mean_token_accuracy": 0.7583109423518181,
"eval_num_tokens": 1744445260.0,
"eval_runtime": 27.0684,
"eval_samples_per_second": 93.615,
"eval_steps_per_second": 1.478,
"step": 13400
},
{
"entropy": 1.76328125,
"epoch": 0.6971303805364941,
"grad_norm": 2.590598058323976,
"learning_rate": 3.123057133669204e-06,
"loss": 0.9813,
"mean_token_accuracy": 0.7661974191665649,
"num_tokens": 1745748289.0,
"step": 13410
},
{
"entropy": 1.7828125,
"epoch": 0.6976502391349553,
"grad_norm": 2.561087376678174,
"learning_rate": 3.1176975024118345e-06,
"loss": 0.982,
"mean_token_accuracy": 0.7615082949399948,
"num_tokens": 1747047863.0,
"step": 13420
},
{
"entropy": 1.740625,
"epoch": 0.6981700977334165,
"grad_norm": 3.07022610190316,
"learning_rate": 3.112337871154465e-06,
"loss": 0.9808,
"mean_token_accuracy": 0.7592459410429001,
"num_tokens": 1748350288.0,
"step": 13430
},
{
"entropy": 1.75,
"epoch": 0.6986899563318777,
"grad_norm": 2.833066577351817,
"learning_rate": 3.1069782398970956e-06,
"loss": 0.9468,
"mean_token_accuracy": 0.7737169414758682,
"num_tokens": 1749652968.0,
"step": 13440
},
{
"entropy": 1.762109375,
"epoch": 0.699209814930339,
"grad_norm": 2.922534349006682,
"learning_rate": 3.101618608639726e-06,
"loss": 0.9542,
"mean_token_accuracy": 0.7702479600906372,
"num_tokens": 1750956486.0,
"step": 13450
},
{
"entropy": 1.76171875,
"epoch": 0.6997296735288001,
"grad_norm": 3.3477914861251903,
"learning_rate": 3.0962589773823567e-06,
"loss": 0.9633,
"mean_token_accuracy": 0.7697488069534302,
"num_tokens": 1752260304.0,
"step": 13460
},
{
"entropy": 1.766796875,
"epoch": 0.7002495321272614,
"grad_norm": 2.6060557226182324,
"learning_rate": 3.0908993461249868e-06,
"loss": 0.9381,
"mean_token_accuracy": 0.7730059564113617,
"num_tokens": 1753563997.0,
"step": 13470
},
{
"entropy": 1.785546875,
"epoch": 0.7007693907257226,
"grad_norm": 2.686781556071249,
"learning_rate": 3.0855397148676173e-06,
"loss": 0.9398,
"mean_token_accuracy": 0.7724482238292694,
"num_tokens": 1754865987.0,
"step": 13480
},
{
"entropy": 1.777734375,
"epoch": 0.7012892493241838,
"grad_norm": 3.3499145587845987,
"learning_rate": 3.080180083610248e-06,
"loss": 0.8982,
"mean_token_accuracy": 0.7814147979021072,
"num_tokens": 1756168543.0,
"step": 13490
},
{
"entropy": 1.757421875,
"epoch": 0.7018091079226451,
"grad_norm": 2.5720952774646126,
"learning_rate": 3.0748204523528784e-06,
"loss": 0.976,
"mean_token_accuracy": 0.7635124921798706,
"num_tokens": 1757470137.0,
"step": 13500
},
{
"epoch": 0.7018091079226451,
"eval_entropy": 1.765234375,
"eval_loss": 0.9970366954803467,
"eval_mean_token_accuracy": 0.7585760354995728,
"eval_num_tokens": 1757470137.0,
"eval_runtime": 27.0459,
"eval_samples_per_second": 93.692,
"eval_steps_per_second": 1.479,
"step": 13500
},
{
"entropy": 1.789453125,
"epoch": 0.7023289665211062,
"grad_norm": 2.919749648798687,
"learning_rate": 3.069460821095509e-06,
"loss": 0.9515,
"mean_token_accuracy": 0.7701379537582398,
"num_tokens": 1758771558.0,
"step": 13510
},
{
"entropy": 1.77421875,
"epoch": 0.7028488251195675,
"grad_norm": 3.053134582263807,
"learning_rate": 3.0641011898381395e-06,
"loss": 0.9763,
"mean_token_accuracy": 0.7667567372322083,
"num_tokens": 1760074116.0,
"step": 13520
},
{
"entropy": 1.7671875,
"epoch": 0.7033686837180287,
"grad_norm": 3.1944364441137387,
"learning_rate": 3.05874155858077e-06,
"loss": 0.9886,
"mean_token_accuracy": 0.7619622349739075,
"num_tokens": 1761374801.0,
"step": 13530
},
{
"entropy": 1.760546875,
"epoch": 0.7038885423164899,
"grad_norm": 2.688375089399267,
"learning_rate": 3.0533819273234006e-06,
"loss": 1.0179,
"mean_token_accuracy": 0.7540708363056183,
"num_tokens": 1762677188.0,
"step": 13540
},
{
"entropy": 1.733203125,
"epoch": 0.7044084009149512,
"grad_norm": 2.7594174000338043,
"learning_rate": 3.048022296066031e-06,
"loss": 0.9591,
"mean_token_accuracy": 0.7691460311412811,
"num_tokens": 1763978473.0,
"step": 13550
},
{
"entropy": 1.75390625,
"epoch": 0.7049282595134123,
"grad_norm": 3.2484459769278473,
"learning_rate": 3.0426626648086617e-06,
"loss": 0.9796,
"mean_token_accuracy": 0.7638642698526382,
"num_tokens": 1765280525.0,
"step": 13560
},
{
"entropy": 1.756640625,
"epoch": 0.7054481181118736,
"grad_norm": 2.7376599798707684,
"learning_rate": 3.0373030335512923e-06,
"loss": 0.9711,
"mean_token_accuracy": 0.7654487431049347,
"num_tokens": 1766583036.0,
"step": 13570
},
{
"entropy": 1.740625,
"epoch": 0.7059679767103348,
"grad_norm": 2.719717708229424,
"learning_rate": 3.031943402293923e-06,
"loss": 0.9592,
"mean_token_accuracy": 0.7672193884849549,
"num_tokens": 1767884669.0,
"step": 13580
},
{
"entropy": 1.724609375,
"epoch": 0.706487835308796,
"grad_norm": 3.1279621864604388,
"learning_rate": 3.026583771036553e-06,
"loss": 0.9478,
"mean_token_accuracy": 0.7713881701231002,
"num_tokens": 1769185708.0,
"step": 13590
},
{
"entropy": 1.74375,
"epoch": 0.7070076939072573,
"grad_norm": 2.611354160772395,
"learning_rate": 3.0212241397791835e-06,
"loss": 0.9644,
"mean_token_accuracy": 0.769554403424263,
"num_tokens": 1770487754.0,
"step": 13600
},
{
"epoch": 0.7070076939072573,
"eval_entropy": 1.7603515625,
"eval_loss": 0.9953738451004028,
"eval_mean_token_accuracy": 0.7593453705310822,
"eval_num_tokens": 1770487754.0,
"eval_runtime": 27.0791,
"eval_samples_per_second": 93.578,
"eval_steps_per_second": 1.477,
"step": 13600
},
{
"entropy": 1.753125,
"epoch": 0.7075275525057184,
"grad_norm": 2.6020748844335295,
"learning_rate": 3.015864508521814e-06,
"loss": 0.9965,
"mean_token_accuracy": 0.7611087441444397,
"num_tokens": 1771789142.0,
"step": 13610
},
{
"entropy": 1.72421875,
"epoch": 0.7080474111041797,
"grad_norm": 2.7654914714748093,
"learning_rate": 3.0105048772644446e-06,
"loss": 0.9362,
"mean_token_accuracy": 0.7737610787153244,
"num_tokens": 1773092328.0,
"step": 13620
},
{
"entropy": 1.740234375,
"epoch": 0.7085672697026408,
"grad_norm": 2.563512195494997,
"learning_rate": 3.005145246007075e-06,
"loss": 0.9771,
"mean_token_accuracy": 0.7645963579416275,
"num_tokens": 1774395794.0,
"step": 13630
},
{
"entropy": 1.758984375,
"epoch": 0.7090871283011021,
"grad_norm": 2.8218458222775418,
"learning_rate": 2.9997856147497057e-06,
"loss": 0.9428,
"mean_token_accuracy": 0.7706138789653778,
"num_tokens": 1775698590.0,
"step": 13640
},
{
"entropy": 1.740625,
"epoch": 0.7096069868995634,
"grad_norm": 2.195533142931692,
"learning_rate": 2.994425983492336e-06,
"loss": 0.9235,
"mean_token_accuracy": 0.776863619685173,
"num_tokens": 1776998059.0,
"step": 13650
},
{
"entropy": 1.78359375,
"epoch": 0.7101268454980245,
"grad_norm": 2.3413831147530124,
"learning_rate": 2.9890663522349667e-06,
"loss": 0.926,
"mean_token_accuracy": 0.7763805508613586,
"num_tokens": 1778299240.0,
"step": 13660
},
{
"entropy": 1.778125,
"epoch": 0.7106467040964858,
"grad_norm": 2.9671583819559206,
"learning_rate": 2.9837067209775973e-06,
"loss": 0.9587,
"mean_token_accuracy": 0.7690684467554092,
"num_tokens": 1779601398.0,
"step": 13670
},
{
"entropy": 1.754296875,
"epoch": 0.7111665626949469,
"grad_norm": 2.8050198974458964,
"learning_rate": 2.978347089720228e-06,
"loss": 0.9488,
"mean_token_accuracy": 0.7696071565151215,
"num_tokens": 1780901783.0,
"step": 13680
},
{
"entropy": 1.774609375,
"epoch": 0.7116864212934082,
"grad_norm": 2.3157673658941618,
"learning_rate": 2.9729874584628584e-06,
"loss": 0.9535,
"mean_token_accuracy": 0.7690162986516953,
"num_tokens": 1782206850.0,
"step": 13690
},
{
"entropy": 1.752734375,
"epoch": 0.7122062798918695,
"grad_norm": 2.7144453016907684,
"learning_rate": 2.9676278272054885e-06,
"loss": 0.9094,
"mean_token_accuracy": 0.7795255213975907,
"num_tokens": 1783507609.0,
"step": 13700
},
{
"epoch": 0.7122062798918695,
"eval_entropy": 1.74765625,
"eval_loss": 0.9925302863121033,
"eval_mean_token_accuracy": 0.7598472476005554,
"eval_num_tokens": 1783507609.0,
"eval_runtime": 27.1009,
"eval_samples_per_second": 93.502,
"eval_steps_per_second": 1.476,
"step": 13700
},
{
"entropy": 1.759765625,
"epoch": 0.7127261384903306,
"grad_norm": 2.7517001308654447,
"learning_rate": 2.962268195948119e-06,
"loss": 0.9342,
"mean_token_accuracy": 0.7740807861089707,
"num_tokens": 1784807575.0,
"step": 13710
},
{
"entropy": 1.75078125,
"epoch": 0.7132459970887919,
"grad_norm": 2.654878018846942,
"learning_rate": 2.9569085646907496e-06,
"loss": 0.9719,
"mean_token_accuracy": 0.7662523984909058,
"num_tokens": 1786107915.0,
"step": 13720
},
{
"entropy": 1.751953125,
"epoch": 0.713765855687253,
"grad_norm": 2.815020020845423,
"learning_rate": 2.95154893343338e-06,
"loss": 0.9664,
"mean_token_accuracy": 0.7690186381340027,
"num_tokens": 1787410287.0,
"step": 13730
},
{
"entropy": 1.778125,
"epoch": 0.7142857142857143,
"grad_norm": 2.678053763525448,
"learning_rate": 2.9461893021760107e-06,
"loss": 0.9651,
"mean_token_accuracy": 0.7670602679252625,
"num_tokens": 1788712437.0,
"step": 13740
},
{
"entropy": 1.751171875,
"epoch": 0.7148055728841755,
"grad_norm": 2.8114696917573387,
"learning_rate": 2.9408296709186412e-06,
"loss": 0.9007,
"mean_token_accuracy": 0.7829609632492065,
"num_tokens": 1790015781.0,
"step": 13750
},
{
"entropy": 1.7875,
"epoch": 0.7153254314826367,
"grad_norm": 2.8904252442258493,
"learning_rate": 2.935470039661272e-06,
"loss": 0.9408,
"mean_token_accuracy": 0.7738532781600952,
"num_tokens": 1791314610.0,
"step": 13760
},
{
"entropy": 1.7578125,
"epoch": 0.715845290081098,
"grad_norm": 2.995342844517079,
"learning_rate": 2.9301104084039023e-06,
"loss": 0.9548,
"mean_token_accuracy": 0.7727679222822189,
"num_tokens": 1792617741.0,
"step": 13770
},
{
"entropy": 1.75703125,
"epoch": 0.7163651486795591,
"grad_norm": 2.643557101230092,
"learning_rate": 2.924750777146533e-06,
"loss": 0.925,
"mean_token_accuracy": 0.7778384238481522,
"num_tokens": 1793921368.0,
"step": 13780
},
{
"entropy": 1.786328125,
"epoch": 0.7168850072780204,
"grad_norm": 2.4197933055613188,
"learning_rate": 2.9193911458891634e-06,
"loss": 1.0094,
"mean_token_accuracy": 0.7586355239152909,
"num_tokens": 1795221802.0,
"step": 13790
},
{
"entropy": 1.783984375,
"epoch": 0.7174048658764816,
"grad_norm": 2.9904657652502533,
"learning_rate": 2.914031514631794e-06,
"loss": 0.9472,
"mean_token_accuracy": 0.77179713845253,
"num_tokens": 1796521880.0,
"step": 13800
},
{
"epoch": 0.7174048658764816,
"eval_entropy": 1.74765625,
"eval_loss": 0.9913745522499084,
"eval_mean_token_accuracy": 0.7601203098893166,
"eval_num_tokens": 1796521880.0,
"eval_runtime": 27.0768,
"eval_samples_per_second": 93.586,
"eval_steps_per_second": 1.477,
"step": 13800
},
{
"entropy": 1.7578125,
"epoch": 0.7179247244749428,
"grad_norm": 2.875111007823221,
"learning_rate": 2.908671883374424e-06,
"loss": 0.9483,
"mean_token_accuracy": 0.7726662278175354,
"num_tokens": 1797825531.0,
"step": 13810
},
{
"entropy": 1.76171875,
"epoch": 0.7184445830734041,
"grad_norm": 2.94059930496427,
"learning_rate": 2.9033122521170546e-06,
"loss": 1.0337,
"mean_token_accuracy": 0.7524798929691314,
"num_tokens": 1799127173.0,
"step": 13820
},
{
"entropy": 1.76484375,
"epoch": 0.7189644416718652,
"grad_norm": 2.6367424687217103,
"learning_rate": 2.897952620859685e-06,
"loss": 0.9449,
"mean_token_accuracy": 0.7727033644914627,
"num_tokens": 1800430422.0,
"step": 13830
},
{
"entropy": 1.737109375,
"epoch": 0.7194843002703265,
"grad_norm": 2.2786397055483847,
"learning_rate": 2.8925929896023157e-06,
"loss": 0.9965,
"mean_token_accuracy": 0.761152395606041,
"num_tokens": 1801729298.0,
"step": 13840
},
{
"entropy": 1.767578125,
"epoch": 0.7200041588687877,
"grad_norm": 2.6002627272196612,
"learning_rate": 2.8872333583449463e-06,
"loss": 0.9733,
"mean_token_accuracy": 0.7665502429008484,
"num_tokens": 1803030810.0,
"step": 13850
},
{
"entropy": 1.78046875,
"epoch": 0.7205240174672489,
"grad_norm": 3.914490841464164,
"learning_rate": 2.881873727087577e-06,
"loss": 0.952,
"mean_token_accuracy": 0.7704506903886795,
"num_tokens": 1804334087.0,
"step": 13860
},
{
"entropy": 1.77265625,
"epoch": 0.7210438760657101,
"grad_norm": 2.613511012827387,
"learning_rate": 2.8765140958302074e-06,
"loss": 0.9741,
"mean_token_accuracy": 0.7657042175531388,
"num_tokens": 1805637391.0,
"step": 13870
},
{
"entropy": 1.76484375,
"epoch": 0.7215637346641713,
"grad_norm": 2.9628383904505378,
"learning_rate": 2.871154464572838e-06,
"loss": 1.0063,
"mean_token_accuracy": 0.7587782114744186,
"num_tokens": 1806939089.0,
"step": 13880
},
{
"entropy": 1.73984375,
"epoch": 0.7220835932626326,
"grad_norm": 2.8090142798874767,
"learning_rate": 2.8657948333154685e-06,
"loss": 0.9565,
"mean_token_accuracy": 0.7692507475614547,
"num_tokens": 1808242438.0,
"step": 13890
},
{
"entropy": 1.7234375,
"epoch": 0.7226034518610938,
"grad_norm": 3.0572746195158818,
"learning_rate": 2.860435202058099e-06,
"loss": 0.8394,
"mean_token_accuracy": 0.796851959824562,
"num_tokens": 1809546083.0,
"step": 13900
},
{
"epoch": 0.7226034518610938,
"eval_entropy": 1.7087890625,
"eval_loss": 0.9909153580665588,
"eval_mean_token_accuracy": 0.7600336998701096,
"eval_num_tokens": 1809546083.0,
"eval_runtime": 27.0632,
"eval_samples_per_second": 93.633,
"eval_steps_per_second": 1.478,
"step": 13900
},
{
"entropy": 1.740625,
"epoch": 0.723123310459555,
"grad_norm": 3.305388047351689,
"learning_rate": 2.8550755708007296e-06,
"loss": 0.9262,
"mean_token_accuracy": 0.7777227580547332,
"num_tokens": 1810849442.0,
"step": 13910
},
{
"entropy": 1.752734375,
"epoch": 0.7236431690580162,
"grad_norm": 3.2182192755869026,
"learning_rate": 2.84971593954336e-06,
"loss": 0.9556,
"mean_token_accuracy": 0.771889865398407,
"num_tokens": 1812152265.0,
"step": 13920
},
{
"entropy": 1.76875,
"epoch": 0.7241630276564774,
"grad_norm": 3.234013318587429,
"learning_rate": 2.8443563082859902e-06,
"loss": 0.9478,
"mean_token_accuracy": 0.7687838941812515,
"num_tokens": 1813455160.0,
"step": 13930
},
{
"entropy": 1.734375,
"epoch": 0.7246828862549387,
"grad_norm": 2.477299333981701,
"learning_rate": 2.8389966770286208e-06,
"loss": 0.9309,
"mean_token_accuracy": 0.7785810112953186,
"num_tokens": 1814757925.0,
"step": 13940
},
{
"entropy": 1.735546875,
"epoch": 0.7252027448533999,
"grad_norm": 2.4910629163825604,
"learning_rate": 2.8336370457712513e-06,
"loss": 0.9518,
"mean_token_accuracy": 0.7710670292377472,
"num_tokens": 1816061726.0,
"step": 13950
},
{
"entropy": 1.769140625,
"epoch": 0.7257226034518611,
"grad_norm": 3.3477540504156025,
"learning_rate": 2.828277414513882e-06,
"loss": 0.9191,
"mean_token_accuracy": 0.7778213739395141,
"num_tokens": 1817362791.0,
"step": 13960
},
{
"entropy": 1.758203125,
"epoch": 0.7262424620503223,
"grad_norm": 2.83239569381125,
"learning_rate": 2.8229177832565124e-06,
"loss": 1.0108,
"mean_token_accuracy": 0.7563022643327713,
"num_tokens": 1818664165.0,
"step": 13970
},
{
"entropy": 1.76015625,
"epoch": 0.7267623206487835,
"grad_norm": 3.0437901707418247,
"learning_rate": 2.817558151999143e-06,
"loss": 0.8818,
"mean_token_accuracy": 0.7856209456920624,
"num_tokens": 1819965591.0,
"step": 13980
},
{
"entropy": 1.75859375,
"epoch": 0.7272821792472447,
"grad_norm": 2.6175678473009856,
"learning_rate": 2.8121985207417735e-06,
"loss": 0.9441,
"mean_token_accuracy": 0.7753179371356964,
"num_tokens": 1821270208.0,
"step": 13990
},
{
"entropy": 1.746875,
"epoch": 0.727802037845706,
"grad_norm": 2.891023844450965,
"learning_rate": 2.806838889484404e-06,
"loss": 0.9581,
"mean_token_accuracy": 0.7680196613073349,
"num_tokens": 1822574609.0,
"step": 14000
},
{
"epoch": 0.727802037845706,
"eval_entropy": 1.7451171875,
"eval_loss": 0.9876113533973694,
"eval_mean_token_accuracy": 0.7610298991203308,
"eval_num_tokens": 1822574609.0,
"eval_runtime": 27.0667,
"eval_samples_per_second": 93.621,
"eval_steps_per_second": 1.478,
"step": 14000
},
{
"entropy": 1.74375,
"epoch": 0.7283218964441672,
"grad_norm": 2.610020998913199,
"learning_rate": 2.8014792582270346e-06,
"loss": 1.0287,
"mean_token_accuracy": 0.75540771484375,
"num_tokens": 1823876093.0,
"step": 14010
},
{
"entropy": 1.754296875,
"epoch": 0.7288417550426284,
"grad_norm": 2.749511198380044,
"learning_rate": 2.796119626969665e-06,
"loss": 0.9981,
"mean_token_accuracy": 0.7600598752498626,
"num_tokens": 1825176846.0,
"step": 14020
},
{
"entropy": 1.7484375,
"epoch": 0.7293616136410896,
"grad_norm": 2.7127526736905687,
"learning_rate": 2.7907599957122957e-06,
"loss": 0.9781,
"mean_token_accuracy": 0.7641529619693757,
"num_tokens": 1826477152.0,
"step": 14030
},
{
"entropy": 1.771875,
"epoch": 0.7298814722395508,
"grad_norm": 3.380853569025533,
"learning_rate": 2.7854003644549254e-06,
"loss": 0.948,
"mean_token_accuracy": 0.7706679195165634,
"num_tokens": 1827779847.0,
"step": 14040
},
{
"entropy": 1.778515625,
"epoch": 0.7304013308380121,
"grad_norm": 2.481720115276413,
"learning_rate": 2.780040733197556e-06,
"loss": 0.9585,
"mean_token_accuracy": 0.7695966243743897,
"num_tokens": 1829082294.0,
"step": 14050
},
{
"entropy": 1.770703125,
"epoch": 0.7309211894364733,
"grad_norm": 2.3352105524351296,
"learning_rate": 2.7746811019401865e-06,
"loss": 0.9575,
"mean_token_accuracy": 0.770712086558342,
"num_tokens": 1830385062.0,
"step": 14060
},
{
"entropy": 1.758984375,
"epoch": 0.7314410480349345,
"grad_norm": 2.910603150115711,
"learning_rate": 2.769321470682817e-06,
"loss": 0.9667,
"mean_token_accuracy": 0.7656277120113373,
"num_tokens": 1831684713.0,
"step": 14070
},
{
"entropy": 1.747265625,
"epoch": 0.7319609066333957,
"grad_norm": 2.3071422275198645,
"learning_rate": 2.7639618394254476e-06,
"loss": 0.9361,
"mean_token_accuracy": 0.7733233451843262,
"num_tokens": 1832985494.0,
"step": 14080
},
{
"entropy": 1.7328125,
"epoch": 0.7324807652318569,
"grad_norm": 2.759064935063,
"learning_rate": 2.758602208168078e-06,
"loss": 0.9413,
"mean_token_accuracy": 0.773416268825531,
"num_tokens": 1834288248.0,
"step": 14090
},
{
"entropy": 1.751171875,
"epoch": 0.7330006238303182,
"grad_norm": 2.798160873205504,
"learning_rate": 2.7532425769107087e-06,
"loss": 0.9612,
"mean_token_accuracy": 0.7727220565080642,
"num_tokens": 1835590125.0,
"step": 14100
},
{
"epoch": 0.7330006238303182,
"eval_entropy": 1.7419921875,
"eval_loss": 0.987047553062439,
"eval_mean_token_accuracy": 0.7613460645079613,
"eval_num_tokens": 1835590125.0,
"eval_runtime": 27.0066,
"eval_samples_per_second": 93.829,
"eval_steps_per_second": 1.481,
"step": 14100
},
{
"entropy": 1.75859375,
"epoch": 0.7335204824287793,
"grad_norm": 2.574413825807132,
"learning_rate": 2.7478829456533392e-06,
"loss": 1.0188,
"mean_token_accuracy": 0.7583661079406738,
"num_tokens": 1836892613.0,
"step": 14110
},
{
"entropy": 1.746875,
"epoch": 0.7340403410272406,
"grad_norm": 2.6189888477441716,
"learning_rate": 2.7425233143959693e-06,
"loss": 0.9662,
"mean_token_accuracy": 0.7686508387327194,
"num_tokens": 1838195233.0,
"step": 14120
},
{
"entropy": 1.76640625,
"epoch": 0.7345601996257018,
"grad_norm": 2.8552696172083616,
"learning_rate": 2.7371636831386e-06,
"loss": 1.0054,
"mean_token_accuracy": 0.7585200190544128,
"num_tokens": 1839496003.0,
"step": 14130
},
{
"entropy": 1.784765625,
"epoch": 0.735080058224163,
"grad_norm": 3.1476718296188486,
"learning_rate": 2.7318040518812304e-06,
"loss": 0.9695,
"mean_token_accuracy": 0.7674199789762497,
"num_tokens": 1840795775.0,
"step": 14140
},
{
"entropy": 1.74140625,
"epoch": 0.7355999168226243,
"grad_norm": 3.1236727828783524,
"learning_rate": 2.726444420623861e-06,
"loss": 0.9469,
"mean_token_accuracy": 0.7719689577817916,
"num_tokens": 1842099384.0,
"step": 14150
},
{
"entropy": 1.733203125,
"epoch": 0.7361197754210854,
"grad_norm": 2.146103253301842,
"learning_rate": 2.7210847893664915e-06,
"loss": 0.9499,
"mean_token_accuracy": 0.7720874100923538,
"num_tokens": 1843401551.0,
"step": 14160
},
{
"entropy": 1.71484375,
"epoch": 0.7366396340195467,
"grad_norm": 2.849759551708361,
"learning_rate": 2.715725158109122e-06,
"loss": 0.9274,
"mean_token_accuracy": 0.7776429146528244,
"num_tokens": 1844703248.0,
"step": 14170
},
{
"entropy": 1.721875,
"epoch": 0.737159492618008,
"grad_norm": 3.2359196054740686,
"learning_rate": 2.7103655268517526e-06,
"loss": 0.9151,
"mean_token_accuracy": 0.7770067691802979,
"num_tokens": 1846005383.0,
"step": 14180
},
{
"entropy": 1.72890625,
"epoch": 0.7376793512164691,
"grad_norm": 3.464316417971457,
"learning_rate": 2.705005895594383e-06,
"loss": 0.9228,
"mean_token_accuracy": 0.7771593689918518,
"num_tokens": 1847306992.0,
"step": 14190
},
{
"entropy": 1.75703125,
"epoch": 0.7381992098149304,
"grad_norm": 3.06472268710951,
"learning_rate": 2.6996462643370137e-06,
"loss": 0.9185,
"mean_token_accuracy": 0.7766084969043732,
"num_tokens": 1848609086.0,
"step": 14200
},
{
"epoch": 0.7381992098149304,
"eval_entropy": 1.716015625,
"eval_loss": 0.9863733053207397,
"eval_mean_token_accuracy": 0.7610517725348472,
"eval_num_tokens": 1848609086.0,
"eval_runtime": 27.0343,
"eval_samples_per_second": 93.733,
"eval_steps_per_second": 1.48,
"step": 14200
},
{
"entropy": 1.748828125,
"epoch": 0.7387190684133915,
"grad_norm": 2.974639960300524,
"learning_rate": 2.6942866330796443e-06,
"loss": 0.9996,
"mean_token_accuracy": 0.7618665128946305,
"num_tokens": 1849912610.0,
"step": 14210
},
{
"entropy": 1.755859375,
"epoch": 0.7392389270118528,
"grad_norm": 2.622776313972694,
"learning_rate": 2.688927001822275e-06,
"loss": 0.9568,
"mean_token_accuracy": 0.7700879126787186,
"num_tokens": 1851216071.0,
"step": 14220
},
{
"entropy": 1.746484375,
"epoch": 0.7397587856103139,
"grad_norm": 2.7969552939615463,
"learning_rate": 2.683567370564905e-06,
"loss": 0.8902,
"mean_token_accuracy": 0.7854810804128647,
"num_tokens": 1852519403.0,
"step": 14230
},
{
"entropy": 1.75078125,
"epoch": 0.7402786442087752,
"grad_norm": 2.4789662559074395,
"learning_rate": 2.6782077393075355e-06,
"loss": 0.9351,
"mean_token_accuracy": 0.7755362391471863,
"num_tokens": 1853818810.0,
"step": 14240
},
{
"entropy": 1.74609375,
"epoch": 0.7407985028072365,
"grad_norm": 2.962240355633116,
"learning_rate": 2.672848108050166e-06,
"loss": 1.0197,
"mean_token_accuracy": 0.7538286507129669,
"num_tokens": 1855121181.0,
"step": 14250
},
{
"entropy": 1.75859375,
"epoch": 0.7413183614056976,
"grad_norm": 2.7846072613683344,
"learning_rate": 2.6674884767927966e-06,
"loss": 0.9783,
"mean_token_accuracy": 0.7648535251617432,
"num_tokens": 1856423668.0,
"step": 14260
},
{
"entropy": 1.74375,
"epoch": 0.7418382200041589,
"grad_norm": 2.8689275858545566,
"learning_rate": 2.662128845535427e-06,
"loss": 0.9505,
"mean_token_accuracy": 0.7692068099975586,
"num_tokens": 1857725897.0,
"step": 14270
},
{
"entropy": 1.751171875,
"epoch": 0.74235807860262,
"grad_norm": 3.154383843886253,
"learning_rate": 2.6567692142780577e-06,
"loss": 0.9705,
"mean_token_accuracy": 0.7653627902269363,
"num_tokens": 1859027981.0,
"step": 14280
},
{
"entropy": 1.77109375,
"epoch": 0.7428779372010813,
"grad_norm": 2.413439834577098,
"learning_rate": 2.6514095830206882e-06,
"loss": 0.9174,
"mean_token_accuracy": 0.778501570224762,
"num_tokens": 1860331159.0,
"step": 14290
},
{
"entropy": 1.7640625,
"epoch": 0.7433977957995426,
"grad_norm": 2.669149719289135,
"learning_rate": 2.6460499517633188e-06,
"loss": 0.9382,
"mean_token_accuracy": 0.7742870658636093,
"num_tokens": 1861634978.0,
"step": 14300
},
{
"epoch": 0.7433977957995426,
"eval_entropy": 1.7375,
"eval_loss": 0.9841892719268799,
"eval_mean_token_accuracy": 0.7618602320551873,
"eval_num_tokens": 1861634978.0,
"eval_runtime": 27.0822,
"eval_samples_per_second": 93.567,
"eval_steps_per_second": 1.477,
"step": 14300
},
{
"entropy": 1.753515625,
"epoch": 0.7439176543980037,
"grad_norm": 2.8895546539458428,
"learning_rate": 2.6406903205059493e-06,
"loss": 1.0353,
"mean_token_accuracy": 0.7535191029310226,
"num_tokens": 1862937484.0,
"step": 14310
},
{
"entropy": 1.76328125,
"epoch": 0.744437512996465,
"grad_norm": 2.402638888182589,
"learning_rate": 2.63533068924858e-06,
"loss": 0.9789,
"mean_token_accuracy": 0.7645485490560532,
"num_tokens": 1864234385.0,
"step": 14320
},
{
"entropy": 1.73671875,
"epoch": 0.7449573715949261,
"grad_norm": 2.638955357206824,
"learning_rate": 2.6299710579912104e-06,
"loss": 0.9583,
"mean_token_accuracy": 0.7705558061599731,
"num_tokens": 1865537058.0,
"step": 14330
},
{
"entropy": 1.74140625,
"epoch": 0.7454772301933874,
"grad_norm": 2.957947972442456,
"learning_rate": 2.6246114267338405e-06,
"loss": 0.9163,
"mean_token_accuracy": 0.7790726155042649,
"num_tokens": 1866840228.0,
"step": 14340
},
{
"entropy": 1.762890625,
"epoch": 0.7459970887918487,
"grad_norm": 3.036774333309838,
"learning_rate": 2.619251795476471e-06,
"loss": 0.9378,
"mean_token_accuracy": 0.7727710843086243,
"num_tokens": 1868141548.0,
"step": 14350
},
{
"entropy": 1.766015625,
"epoch": 0.7465169473903098,
"grad_norm": 2.706976369510526,
"learning_rate": 2.6138921642191016e-06,
"loss": 0.9805,
"mean_token_accuracy": 0.7652006387710572,
"num_tokens": 1869444535.0,
"step": 14360
},
{
"entropy": 1.76796875,
"epoch": 0.7470368059887711,
"grad_norm": 2.829111089562102,
"learning_rate": 2.608532532961732e-06,
"loss": 0.9673,
"mean_token_accuracy": 0.7677335619926453,
"num_tokens": 1870747317.0,
"step": 14370
},
{
"entropy": 1.77890625,
"epoch": 0.7475566645872322,
"grad_norm": 3.1976307058890976,
"learning_rate": 2.6031729017043627e-06,
"loss": 0.9026,
"mean_token_accuracy": 0.7835441559553147,
"num_tokens": 1872048533.0,
"step": 14380
},
{
"entropy": 1.753125,
"epoch": 0.7480765231856935,
"grad_norm": 3.4636342735427075,
"learning_rate": 2.5978132704469933e-06,
"loss": 0.9287,
"mean_token_accuracy": 0.7768224209547043,
"num_tokens": 1873349215.0,
"step": 14390
},
{
"entropy": 1.750390625,
"epoch": 0.7485963817841547,
"grad_norm": 2.305124367470247,
"learning_rate": 2.592453639189624e-06,
"loss": 0.9257,
"mean_token_accuracy": 0.7771239370107651,
"num_tokens": 1874649979.0,
"step": 14400
},
{
"epoch": 0.7485963817841547,
"eval_entropy": 1.752734375,
"eval_loss": 0.9823573231697083,
"eval_mean_token_accuracy": 0.7627713352441787,
"eval_num_tokens": 1874649979.0,
"eval_runtime": 27.1029,
"eval_samples_per_second": 93.495,
"eval_steps_per_second": 1.476,
"step": 14400
},
{
"entropy": 1.746484375,
"epoch": 0.7491162403826159,
"grad_norm": 2.4717986866962574,
"learning_rate": 2.5870940079322544e-06,
"loss": 0.8958,
"mean_token_accuracy": 0.7811079531908035,
"num_tokens": 1875951980.0,
"step": 14410
},
{
"entropy": 1.741796875,
"epoch": 0.7496360989810772,
"grad_norm": 2.470882240687234,
"learning_rate": 2.581734376674885e-06,
"loss": 0.9976,
"mean_token_accuracy": 0.7585900634527206,
"num_tokens": 1877253945.0,
"step": 14420
},
{
"entropy": 1.790625,
"epoch": 0.7501559575795383,
"grad_norm": 2.7543294224399233,
"learning_rate": 2.5763747454175154e-06,
"loss": 0.9473,
"mean_token_accuracy": 0.772933629155159,
"num_tokens": 1878556295.0,
"step": 14430
},
{
"entropy": 1.76015625,
"epoch": 0.7506758161779996,
"grad_norm": 2.5138041059831733,
"learning_rate": 2.571015114160146e-06,
"loss": 1.0016,
"mean_token_accuracy": 0.7614201605319977,
"num_tokens": 1879859291.0,
"step": 14440
},
{
"entropy": 1.756640625,
"epoch": 0.7511956747764608,
"grad_norm": 2.776425096180524,
"learning_rate": 2.5656554829027765e-06,
"loss": 0.9706,
"mean_token_accuracy": 0.7681435078382493,
"num_tokens": 1881160950.0,
"step": 14450
},
{
"entropy": 1.74296875,
"epoch": 0.751715533374922,
"grad_norm": 2.2071088188936137,
"learning_rate": 2.5602958516454067e-06,
"loss": 0.9581,
"mean_token_accuracy": 0.7712674498558044,
"num_tokens": 1882461997.0,
"step": 14460
},
{
"entropy": 1.757421875,
"epoch": 0.7522353919733833,
"grad_norm": 2.8712932971776883,
"learning_rate": 2.554936220388037e-06,
"loss": 0.9045,
"mean_token_accuracy": 0.7812943756580353,
"num_tokens": 1883763898.0,
"step": 14470
},
{
"entropy": 1.753125,
"epoch": 0.7527552505718444,
"grad_norm": 2.721164168469557,
"learning_rate": 2.5495765891306678e-06,
"loss": 0.9156,
"mean_token_accuracy": 0.7766843229532242,
"num_tokens": 1885062458.0,
"step": 14480
},
{
"entropy": 1.74453125,
"epoch": 0.7532751091703057,
"grad_norm": 2.499696502058153,
"learning_rate": 2.5442169578732983e-06,
"loss": 0.9008,
"mean_token_accuracy": 0.7847321599721908,
"num_tokens": 1886364534.0,
"step": 14490
},
{
"entropy": 1.755078125,
"epoch": 0.7537949677687669,
"grad_norm": 3.2336968396730357,
"learning_rate": 2.538857326615929e-06,
"loss": 0.9743,
"mean_token_accuracy": 0.764453512430191,
"num_tokens": 1887667028.0,
"step": 14500
},
{
"epoch": 0.7537949677687669,
"eval_entropy": 1.7435546875,
"eval_loss": 0.9812038540840149,
"eval_mean_token_accuracy": 0.7626863971352578,
"eval_num_tokens": 1887667028.0,
"eval_runtime": 26.97,
"eval_samples_per_second": 93.956,
"eval_steps_per_second": 1.483,
"step": 14500
},
{
"entropy": 1.73671875,
"epoch": 0.7543148263672281,
"grad_norm": 3.0007364890247517,
"learning_rate": 2.5334976953585594e-06,
"loss": 0.9216,
"mean_token_accuracy": 0.7778298050165177,
"num_tokens": 1888969713.0,
"step": 14510
},
{
"entropy": 1.7390625,
"epoch": 0.7548346849656893,
"grad_norm": 2.658935237982246,
"learning_rate": 2.52813806410119e-06,
"loss": 0.9034,
"mean_token_accuracy": 0.7837458163499832,
"num_tokens": 1890271464.0,
"step": 14520
},
{
"entropy": 1.755078125,
"epoch": 0.7553545435641505,
"grad_norm": 3.0202539650743443,
"learning_rate": 2.5227784328438205e-06,
"loss": 0.8903,
"mean_token_accuracy": 0.7838906139135361,
"num_tokens": 1891574884.0,
"step": 14530
},
{
"entropy": 1.773828125,
"epoch": 0.7558744021626118,
"grad_norm": 2.765605001342347,
"learning_rate": 2.517418801586451e-06,
"loss": 0.9868,
"mean_token_accuracy": 0.7649797558784485,
"num_tokens": 1892873894.0,
"step": 14540
},
{
"entropy": 1.79296875,
"epoch": 0.756394260761073,
"grad_norm": 2.5782021761957252,
"learning_rate": 2.5120591703290816e-06,
"loss": 0.9807,
"mean_token_accuracy": 0.7656522989273071,
"num_tokens": 1894176783.0,
"step": 14550
},
{
"entropy": 1.771875,
"epoch": 0.7569141193595342,
"grad_norm": 2.6750066625672444,
"learning_rate": 2.506699539071712e-06,
"loss": 0.992,
"mean_token_accuracy": 0.7638985246419907,
"num_tokens": 1895477308.0,
"step": 14560
},
{
"entropy": 1.748046875,
"epoch": 0.7574339779579954,
"grad_norm": 2.696383282650331,
"learning_rate": 2.5013399078143423e-06,
"loss": 0.9283,
"mean_token_accuracy": 0.7767240077257156,
"num_tokens": 1896777439.0,
"step": 14570
},
{
"entropy": 1.777734375,
"epoch": 0.7579538365564567,
"grad_norm": 2.983440285120926,
"learning_rate": 2.495980276556973e-06,
"loss": 0.9937,
"mean_token_accuracy": 0.7616151422262192,
"num_tokens": 1898077333.0,
"step": 14580
},
{
"entropy": 1.76328125,
"epoch": 0.7584736951549179,
"grad_norm": 2.7268527802502116,
"learning_rate": 2.4906206452996033e-06,
"loss": 0.9968,
"mean_token_accuracy": 0.7631165444850921,
"num_tokens": 1899380590.0,
"step": 14590
},
{
"entropy": 1.762109375,
"epoch": 0.7589935537533791,
"grad_norm": 2.825802552627355,
"learning_rate": 2.485261014042234e-06,
"loss": 0.9722,
"mean_token_accuracy": 0.7664202004671097,
"num_tokens": 1900683400.0,
"step": 14600
},
{
"epoch": 0.7589935537533791,
"eval_entropy": 1.754296875,
"eval_loss": 0.9802638292312622,
"eval_mean_token_accuracy": 0.7632532313466072,
"eval_num_tokens": 1900683400.0,
"eval_runtime": 27.0349,
"eval_samples_per_second": 93.731,
"eval_steps_per_second": 1.48,
"step": 14600
},
{
"entropy": 1.776171875,
"epoch": 0.7595134123518403,
"grad_norm": 2.7813406064962543,
"learning_rate": 2.4799013827848644e-06,
"loss": 0.9298,
"mean_token_accuracy": 0.7762244075536728,
"num_tokens": 1901985012.0,
"step": 14610
},
{
"entropy": 1.7359375,
"epoch": 0.7600332709503015,
"grad_norm": 3.201005442959932,
"learning_rate": 2.474541751527495e-06,
"loss": 0.9324,
"mean_token_accuracy": 0.7778979957103729,
"num_tokens": 1903287249.0,
"step": 14620
},
{
"entropy": 1.75859375,
"epoch": 0.7605531295487628,
"grad_norm": 2.856155537851972,
"learning_rate": 2.4691821202701255e-06,
"loss": 1.0034,
"mean_token_accuracy": 0.7624159514904022,
"num_tokens": 1904588533.0,
"step": 14630
},
{
"entropy": 1.758203125,
"epoch": 0.7610729881472239,
"grad_norm": 2.46153293934386,
"learning_rate": 2.463822489012756e-06,
"loss": 0.9304,
"mean_token_accuracy": 0.7745770931243896,
"num_tokens": 1905890505.0,
"step": 14640
},
{
"entropy": 1.76640625,
"epoch": 0.7615928467456852,
"grad_norm": 2.6000807109426347,
"learning_rate": 2.4584628577553866e-06,
"loss": 0.9209,
"mean_token_accuracy": 0.7798342823982238,
"num_tokens": 1907191207.0,
"step": 14650
},
{
"entropy": 1.773046875,
"epoch": 0.7621127053441464,
"grad_norm": 2.7093794203854493,
"learning_rate": 2.453103226498017e-06,
"loss": 0.8838,
"mean_token_accuracy": 0.7862056702375412,
"num_tokens": 1908492866.0,
"step": 14660
},
{
"entropy": 1.780078125,
"epoch": 0.7626325639426076,
"grad_norm": 2.5962686105124493,
"learning_rate": 2.4477435952406477e-06,
"loss": 0.922,
"mean_token_accuracy": 0.778702712059021,
"num_tokens": 1909794864.0,
"step": 14670
},
{
"entropy": 1.7296875,
"epoch": 0.7631524225410689,
"grad_norm": 2.7421249117881077,
"learning_rate": 2.442383963983278e-06,
"loss": 0.911,
"mean_token_accuracy": 0.7803871929645538,
"num_tokens": 1911097493.0,
"step": 14680
},
{
"entropy": 1.75390625,
"epoch": 0.76367228113953,
"grad_norm": 2.9638961348986923,
"learning_rate": 2.4370243327259084e-06,
"loss": 0.9635,
"mean_token_accuracy": 0.768519964814186,
"num_tokens": 1912396996.0,
"step": 14690
},
{
"entropy": 1.758984375,
"epoch": 0.7641921397379913,
"grad_norm": 2.732115846724963,
"learning_rate": 2.431664701468539e-06,
"loss": 0.9394,
"mean_token_accuracy": 0.774222981929779,
"num_tokens": 1913698499.0,
"step": 14700
},
{
"epoch": 0.7641921397379913,
"eval_entropy": 1.751171875,
"eval_loss": 0.9791877269744873,
"eval_mean_token_accuracy": 0.7631065994501114,
"eval_num_tokens": 1913698499.0,
"eval_runtime": 27.0378,
"eval_samples_per_second": 93.721,
"eval_steps_per_second": 1.479,
"step": 14700
},
{
"entropy": 1.762109375,
"epoch": 0.7647119983364525,
"grad_norm": 2.6314534121279376,
"learning_rate": 2.4263050702111695e-06,
"loss": 0.9178,
"mean_token_accuracy": 0.7809780955314636,
"num_tokens": 1914998578.0,
"step": 14710
},
{
"entropy": 1.749609375,
"epoch": 0.7652318569349137,
"grad_norm": 2.575669557229566,
"learning_rate": 2.4209454389538e-06,
"loss": 0.9519,
"mean_token_accuracy": 0.767992052435875,
"num_tokens": 1916299244.0,
"step": 14720
},
{
"entropy": 1.749609375,
"epoch": 0.765751715533375,
"grad_norm": 2.4568713195795944,
"learning_rate": 2.4155858076964306e-06,
"loss": 0.9267,
"mean_token_accuracy": 0.7737519264221191,
"num_tokens": 1917601006.0,
"step": 14730
},
{
"entropy": 1.7375,
"epoch": 0.7662715741318361,
"grad_norm": 2.759679606584365,
"learning_rate": 2.410226176439061e-06,
"loss": 0.9849,
"mean_token_accuracy": 0.7613626897335053,
"num_tokens": 1918903109.0,
"step": 14740
},
{
"entropy": 1.739453125,
"epoch": 0.7667914327302974,
"grad_norm": 3.7263271686353314,
"learning_rate": 2.4048665451816917e-06,
"loss": 0.9366,
"mean_token_accuracy": 0.7787836968898774,
"num_tokens": 1920205438.0,
"step": 14750
},
{
"entropy": 1.754296875,
"epoch": 0.7673112913287585,
"grad_norm": 2.398539433059064,
"learning_rate": 2.399506913924322e-06,
"loss": 0.9841,
"mean_token_accuracy": 0.7651033908128738,
"num_tokens": 1921507784.0,
"step": 14760
},
{
"entropy": 1.753515625,
"epoch": 0.7678311499272198,
"grad_norm": 2.441292691711973,
"learning_rate": 2.3941472826669528e-06,
"loss": 0.9645,
"mean_token_accuracy": 0.7672067821025849,
"num_tokens": 1922809415.0,
"step": 14770
},
{
"entropy": 1.7609375,
"epoch": 0.7683510085256811,
"grad_norm": 2.712664712199468,
"learning_rate": 2.3887876514095833e-06,
"loss": 0.8988,
"mean_token_accuracy": 0.785337170958519,
"num_tokens": 1924110872.0,
"step": 14780
},
{
"entropy": 1.7546875,
"epoch": 0.7688708671241422,
"grad_norm": 2.900067445579784,
"learning_rate": 2.3834280201522134e-06,
"loss": 0.9426,
"mean_token_accuracy": 0.7728705704212189,
"num_tokens": 1925413304.0,
"step": 14790
},
{
"entropy": 1.72890625,
"epoch": 0.7693907257226035,
"grad_norm": 3.439483060382314,
"learning_rate": 2.378068388894844e-06,
"loss": 0.9448,
"mean_token_accuracy": 0.7749762803316116,
"num_tokens": 1926716961.0,
"step": 14800
},
{
"epoch": 0.7693907257226035,
"eval_entropy": 1.722265625,
"eval_loss": 0.9771442413330078,
"eval_mean_token_accuracy": 0.7633682399988174,
"eval_num_tokens": 1926716961.0,
"eval_runtime": 27.0588,
"eval_samples_per_second": 93.648,
"eval_steps_per_second": 1.478,
"step": 14800
},
{
"entropy": 1.74296875,
"epoch": 0.7699105843210646,
"grad_norm": 2.977861817555228,
"learning_rate": 2.3727087576374745e-06,
"loss": 0.9502,
"mean_token_accuracy": 0.771796602010727,
"num_tokens": 1928020508.0,
"step": 14810
},
{
"entropy": 1.7734375,
"epoch": 0.7704304429195259,
"grad_norm": 2.748446068898822,
"learning_rate": 2.367349126380105e-06,
"loss": 0.944,
"mean_token_accuracy": 0.7746872305870056,
"num_tokens": 1929322372.0,
"step": 14820
},
{
"entropy": 1.72734375,
"epoch": 0.7709503015179872,
"grad_norm": 2.804915477891189,
"learning_rate": 2.3619894951227356e-06,
"loss": 0.9264,
"mean_token_accuracy": 0.7783137500286103,
"num_tokens": 1930624217.0,
"step": 14830
},
{
"entropy": 1.746875,
"epoch": 0.7714701601164483,
"grad_norm": 2.8852237207311653,
"learning_rate": 2.356629863865366e-06,
"loss": 0.9779,
"mean_token_accuracy": 0.7663089752197265,
"num_tokens": 1931926402.0,
"step": 14840
},
{
"entropy": 1.745703125,
"epoch": 0.7719900187149096,
"grad_norm": 2.626323399721019,
"learning_rate": 2.3512702326079967e-06,
"loss": 0.978,
"mean_token_accuracy": 0.7651987224817276,
"num_tokens": 1933229934.0,
"step": 14850
},
{
"entropy": 1.750390625,
"epoch": 0.7725098773133707,
"grad_norm": 2.626771932440381,
"learning_rate": 2.3459106013506273e-06,
"loss": 0.9245,
"mean_token_accuracy": 0.7739851325750351,
"num_tokens": 1934533407.0,
"step": 14860
},
{
"entropy": 1.696484375,
"epoch": 0.773029735911832,
"grad_norm": 2.868884278750296,
"learning_rate": 2.340550970093258e-06,
"loss": 0.8908,
"mean_token_accuracy": 0.7863349348306656,
"num_tokens": 1935835992.0,
"step": 14870
},
{
"entropy": 1.750390625,
"epoch": 0.7735495945102931,
"grad_norm": 2.8213891013938346,
"learning_rate": 2.3351913388358883e-06,
"loss": 0.9568,
"mean_token_accuracy": 0.7706453293561936,
"num_tokens": 1937137535.0,
"step": 14880
},
{
"entropy": 1.74921875,
"epoch": 0.7740694531087544,
"grad_norm": 2.62060061427741,
"learning_rate": 2.329831707578519e-06,
"loss": 0.9106,
"mean_token_accuracy": 0.7796021461486816,
"num_tokens": 1938438366.0,
"step": 14890
},
{
"entropy": 1.723046875,
"epoch": 0.7745893117072157,
"grad_norm": 3.4297222769433082,
"learning_rate": 2.3244720763211494e-06,
"loss": 0.9798,
"mean_token_accuracy": 0.7633107393980026,
"num_tokens": 1939740466.0,
"step": 14900
},
{
"epoch": 0.7745893117072157,
"eval_entropy": 1.73125,
"eval_loss": 0.9752053618431091,
"eval_mean_token_accuracy": 0.7644504889845848,
"eval_num_tokens": 1939740466.0,
"eval_runtime": 27.0061,
"eval_samples_per_second": 93.831,
"eval_steps_per_second": 1.481,
"step": 14900
},
{
"entropy": 1.74921875,
"epoch": 0.7751091703056768,
"grad_norm": 2.432355226720451,
"learning_rate": 2.3191124450637796e-06,
"loss": 0.9498,
"mean_token_accuracy": 0.7734127372503281,
"num_tokens": 1941040248.0,
"step": 14910
},
{
"entropy": 1.74609375,
"epoch": 0.7756290289041381,
"grad_norm": 2.324559033255044,
"learning_rate": 2.31375281380641e-06,
"loss": 0.8861,
"mean_token_accuracy": 0.7826823979616165,
"num_tokens": 1942341347.0,
"step": 14920
},
{
"entropy": 1.771484375,
"epoch": 0.7761488875025992,
"grad_norm": 3.583863033240035,
"learning_rate": 2.3083931825490407e-06,
"loss": 0.9984,
"mean_token_accuracy": 0.7609005004167557,
"num_tokens": 1943644581.0,
"step": 14930
},
{
"entropy": 1.744921875,
"epoch": 0.7766687461010605,
"grad_norm": 2.562874996191587,
"learning_rate": 2.303033551291671e-06,
"loss": 0.9605,
"mean_token_accuracy": 0.7692199498414993,
"num_tokens": 1944947532.0,
"step": 14940
},
{
"entropy": 1.74609375,
"epoch": 0.7771886046995218,
"grad_norm": 3.0791089731107006,
"learning_rate": 2.2976739200343017e-06,
"loss": 0.98,
"mean_token_accuracy": 0.7654582172632217,
"num_tokens": 1946247924.0,
"step": 14950
},
{
"entropy": 1.7296875,
"epoch": 0.7777084632979829,
"grad_norm": 2.8341861861787967,
"learning_rate": 2.2923142887769323e-06,
"loss": 0.9414,
"mean_token_accuracy": 0.7723586022853851,
"num_tokens": 1947549991.0,
"step": 14960
},
{
"entropy": 1.7609375,
"epoch": 0.7782283218964442,
"grad_norm": 2.2642406243198314,
"learning_rate": 2.286954657519563e-06,
"loss": 0.99,
"mean_token_accuracy": 0.7605105251073837,
"num_tokens": 1948852171.0,
"step": 14970
},
{
"entropy": 1.737109375,
"epoch": 0.7787481804949054,
"grad_norm": 3.042187481848504,
"learning_rate": 2.2815950262621934e-06,
"loss": 0.9725,
"mean_token_accuracy": 0.7655585139989853,
"num_tokens": 1950151302.0,
"step": 14980
},
{
"entropy": 1.733984375,
"epoch": 0.7792680390933666,
"grad_norm": 2.7106854033440655,
"learning_rate": 2.276235395004824e-06,
"loss": 0.9455,
"mean_token_accuracy": 0.7754229664802551,
"num_tokens": 1951454375.0,
"step": 14990
},
{
"entropy": 1.7640625,
"epoch": 0.7797878976918278,
"grad_norm": 2.7183367538834,
"learning_rate": 2.2708757637474545e-06,
"loss": 0.9658,
"mean_token_accuracy": 0.7675707906484603,
"num_tokens": 1952757321.0,
"step": 15000
},
{
"epoch": 0.7797878976918278,
"eval_entropy": 1.7140625,
"eval_loss": 0.9733641743659973,
"eval_mean_token_accuracy": 0.7642887860536576,
"eval_num_tokens": 1952757321.0,
"eval_runtime": 27.0548,
"eval_samples_per_second": 93.662,
"eval_steps_per_second": 1.478,
"step": 15000
},
{
"entropy": 1.72421875,
"epoch": 0.780307756290289,
"grad_norm": 3.0636816701277994,
"learning_rate": 2.265516132490085e-06,
"loss": 0.8817,
"mean_token_accuracy": 0.7853486031293869,
"num_tokens": 1954059748.0,
"step": 15010
},
{
"entropy": 1.739453125,
"epoch": 0.7808276148887503,
"grad_norm": 2.8112908967335346,
"learning_rate": 2.260156501232715e-06,
"loss": 0.9034,
"mean_token_accuracy": 0.7803995221853256,
"num_tokens": 1955363260.0,
"step": 15020
},
{
"entropy": 1.758203125,
"epoch": 0.7813474734872115,
"grad_norm": 2.527355657474679,
"learning_rate": 2.2547968699753457e-06,
"loss": 0.9977,
"mean_token_accuracy": 0.758691418170929,
"num_tokens": 1956662472.0,
"step": 15030
},
{
"entropy": 1.75,
"epoch": 0.7818673320856727,
"grad_norm": 2.6627286431115182,
"learning_rate": 2.2494372387179762e-06,
"loss": 0.9781,
"mean_token_accuracy": 0.7647781103849411,
"num_tokens": 1957963455.0,
"step": 15040
},
{
"entropy": 1.719140625,
"epoch": 0.7823871906841339,
"grad_norm": 3.848381985446338,
"learning_rate": 2.244077607460607e-06,
"loss": 0.9257,
"mean_token_accuracy": 0.7762020200490951,
"num_tokens": 1959267224.0,
"step": 15050
},
{
"entropy": 1.728125,
"epoch": 0.7829070492825951,
"grad_norm": 2.5605748817436274,
"learning_rate": 2.2387179762032373e-06,
"loss": 0.9066,
"mean_token_accuracy": 0.781126856803894,
"num_tokens": 1960570351.0,
"step": 15060
},
{
"entropy": 1.73984375,
"epoch": 0.7834269078810564,
"grad_norm": 2.384364939785793,
"learning_rate": 2.233358344945868e-06,
"loss": 0.9742,
"mean_token_accuracy": 0.7631551921367645,
"num_tokens": 1961872276.0,
"step": 15070
},
{
"entropy": 1.7359375,
"epoch": 0.7839467664795176,
"grad_norm": 2.54777503650705,
"learning_rate": 2.2279987136884984e-06,
"loss": 0.9486,
"mean_token_accuracy": 0.773741552233696,
"num_tokens": 1963174865.0,
"step": 15080
},
{
"entropy": 1.7515625,
"epoch": 0.7844666250779788,
"grad_norm": 2.770987790174176,
"learning_rate": 2.222639082431129e-06,
"loss": 0.9164,
"mean_token_accuracy": 0.781196528673172,
"num_tokens": 1964475055.0,
"step": 15090
},
{
"entropy": 1.713671875,
"epoch": 0.78498648367644,
"grad_norm": 2.3253061174071155,
"learning_rate": 2.2172794511737595e-06,
"loss": 0.9157,
"mean_token_accuracy": 0.7780691504478454,
"num_tokens": 1965775444.0,
"step": 15100
},
{
"epoch": 0.78498648367644,
"eval_entropy": 1.705078125,
"eval_loss": 0.9718566536903381,
"eval_mean_token_accuracy": 0.7651963368058204,
"eval_num_tokens": 1965775444.0,
"eval_runtime": 27.0822,
"eval_samples_per_second": 93.567,
"eval_steps_per_second": 1.477,
"step": 15100
},
{
"entropy": 1.756640625,
"epoch": 0.7855063422749012,
"grad_norm": 2.905644773490105,
"learning_rate": 2.21191981991639e-06,
"loss": 1.0075,
"mean_token_accuracy": 0.7573029071092605,
"num_tokens": 1967077589.0,
"step": 15110
},
{
"entropy": 1.74375,
"epoch": 0.7860262008733624,
"grad_norm": 2.6199609829729797,
"learning_rate": 2.2065601886590206e-06,
"loss": 0.9633,
"mean_token_accuracy": 0.7687530845403672,
"num_tokens": 1968377739.0,
"step": 15120
},
{
"entropy": 1.740625,
"epoch": 0.7865460594718237,
"grad_norm": 2.637648097066736,
"learning_rate": 2.2012005574016507e-06,
"loss": 0.954,
"mean_token_accuracy": 0.7695484906435013,
"num_tokens": 1969681539.0,
"step": 15130
},
{
"entropy": 1.748046875,
"epoch": 0.7870659180702849,
"grad_norm": 2.3189469171769654,
"learning_rate": 2.1958409261442813e-06,
"loss": 0.9521,
"mean_token_accuracy": 0.7725260972976684,
"num_tokens": 1970983103.0,
"step": 15140
},
{
"entropy": 1.769921875,
"epoch": 0.7875857766687461,
"grad_norm": 2.7286031555323933,
"learning_rate": 2.190481294886912e-06,
"loss": 0.8166,
"mean_token_accuracy": 0.8028953939676284,
"num_tokens": 1972286404.0,
"step": 15150
},
{
"entropy": 1.73046875,
"epoch": 0.7881056352672073,
"grad_norm": 2.894429243182871,
"learning_rate": 2.1851216636295424e-06,
"loss": 0.9499,
"mean_token_accuracy": 0.7742726594209671,
"num_tokens": 1973589041.0,
"step": 15160
},
{
"entropy": 1.725,
"epoch": 0.7886254938656685,
"grad_norm": 2.873773463294939,
"learning_rate": 2.179762032372173e-06,
"loss": 0.8842,
"mean_token_accuracy": 0.7874167323112488,
"num_tokens": 1974888234.0,
"step": 15170
},
{
"entropy": 1.7390625,
"epoch": 0.7891453524641298,
"grad_norm": 2.945109431760653,
"learning_rate": 2.1744024011148035e-06,
"loss": 0.935,
"mean_token_accuracy": 0.7764123320579529,
"num_tokens": 1976189731.0,
"step": 15180
},
{
"entropy": 1.717578125,
"epoch": 0.789665211062591,
"grad_norm": 2.797759135113352,
"learning_rate": 2.169042769857434e-06,
"loss": 0.9524,
"mean_token_accuracy": 0.7691715806722641,
"num_tokens": 1977490923.0,
"step": 15190
},
{
"entropy": 1.755078125,
"epoch": 0.7901850696610522,
"grad_norm": 2.6853465229273255,
"learning_rate": 2.1636831386000646e-06,
"loss": 0.9086,
"mean_token_accuracy": 0.7805203646421432,
"num_tokens": 1978789049.0,
"step": 15200
},
{
"epoch": 0.7901850696610522,
"eval_entropy": 1.7482421875,
"eval_loss": 0.9693311452865601,
"eval_mean_token_accuracy": 0.7656849324703217,
"eval_num_tokens": 1978789049.0,
"eval_runtime": 26.9997,
"eval_samples_per_second": 93.853,
"eval_steps_per_second": 1.481,
"step": 15200
},
{
"entropy": 1.737890625,
"epoch": 0.7907049282595134,
"grad_norm": 2.786664593741998,
"learning_rate": 2.158323507342695e-06,
"loss": 0.9338,
"mean_token_accuracy": 0.7734246492385864,
"num_tokens": 1980090606.0,
"step": 15210
},
{
"entropy": 1.7359375,
"epoch": 0.7912247868579746,
"grad_norm": 2.8092153615270785,
"learning_rate": 2.1529638760853257e-06,
"loss": 0.9497,
"mean_token_accuracy": 0.7723321884870529,
"num_tokens": 1981390338.0,
"step": 15220
},
{
"entropy": 1.741796875,
"epoch": 0.7917446454564359,
"grad_norm": 3.153948268662094,
"learning_rate": 2.147604244827956e-06,
"loss": 0.9286,
"mean_token_accuracy": 0.7764100968837738,
"num_tokens": 1982692967.0,
"step": 15230
},
{
"entropy": 1.72890625,
"epoch": 0.7922645040548971,
"grad_norm": 2.699088242856241,
"learning_rate": 2.1422446135705868e-06,
"loss": 0.8741,
"mean_token_accuracy": 0.7915118038654327,
"num_tokens": 1983994050.0,
"step": 15240
},
{
"entropy": 1.746484375,
"epoch": 0.7927843626533583,
"grad_norm": 2.8332881909388643,
"learning_rate": 2.136884982313217e-06,
"loss": 0.9786,
"mean_token_accuracy": 0.7646807372570038,
"num_tokens": 1985296791.0,
"step": 15250
},
{
"entropy": 1.727734375,
"epoch": 0.7933042212518195,
"grad_norm": 3.0759918575347065,
"learning_rate": 2.1315253510558474e-06,
"loss": 0.9124,
"mean_token_accuracy": 0.7775000929832458,
"num_tokens": 1986598902.0,
"step": 15260
},
{
"entropy": 1.761328125,
"epoch": 0.7938240798502807,
"grad_norm": 3.0013252616824193,
"learning_rate": 2.126165719798478e-06,
"loss": 0.948,
"mean_token_accuracy": 0.7696346253156662,
"num_tokens": 1987901329.0,
"step": 15270
},
{
"entropy": 1.753125,
"epoch": 0.794343938448742,
"grad_norm": 3.0349963345811704,
"learning_rate": 2.1208060885411085e-06,
"loss": 0.9681,
"mean_token_accuracy": 0.7671276599168777,
"num_tokens": 1989203460.0,
"step": 15280
},
{
"entropy": 1.783984375,
"epoch": 0.7948637970472031,
"grad_norm": 2.4849149108266477,
"learning_rate": 2.115446457283739e-06,
"loss": 0.9923,
"mean_token_accuracy": 0.7621104061603546,
"num_tokens": 1990502768.0,
"step": 15290
},
{
"entropy": 1.73828125,
"epoch": 0.7953836556456644,
"grad_norm": 2.9942425594457336,
"learning_rate": 2.1100868260263696e-06,
"loss": 0.9221,
"mean_token_accuracy": 0.7778311401605607,
"num_tokens": 1991804595.0,
"step": 15300
},
{
"epoch": 0.7953836556456644,
"eval_entropy": 1.74765625,
"eval_loss": 0.9686875939369202,
"eval_mean_token_accuracy": 0.7656173691153526,
"eval_num_tokens": 1991804595.0,
"eval_runtime": 26.9942,
"eval_samples_per_second": 93.872,
"eval_steps_per_second": 1.482,
"step": 15300
},
{
"entropy": 1.774609375,
"epoch": 0.7959035142441256,
"grad_norm": 2.7633264795827763,
"learning_rate": 2.104727194769e-06,
"loss": 0.9049,
"mean_token_accuracy": 0.7802119046449661,
"num_tokens": 1993106650.0,
"step": 15310
},
{
"entropy": 1.7578125,
"epoch": 0.7964233728425868,
"grad_norm": 3.0716880536481432,
"learning_rate": 2.0993675635116307e-06,
"loss": 0.976,
"mean_token_accuracy": 0.7664124220609665,
"num_tokens": 1994410015.0,
"step": 15320
},
{
"entropy": 1.75234375,
"epoch": 0.7969432314410481,
"grad_norm": 3.165626901595579,
"learning_rate": 2.0940079322542612e-06,
"loss": 0.9455,
"mean_token_accuracy": 0.7759915173053742,
"num_tokens": 1995711393.0,
"step": 15330
},
{
"entropy": 1.756640625,
"epoch": 0.7974630900395092,
"grad_norm": 2.2383527957717377,
"learning_rate": 2.088648300996892e-06,
"loss": 0.8636,
"mean_token_accuracy": 0.7921938061714172,
"num_tokens": 1997012956.0,
"step": 15340
},
{
"entropy": 1.7484375,
"epoch": 0.7979829486379705,
"grad_norm": 2.9875137428446426,
"learning_rate": 2.0832886697395223e-06,
"loss": 0.9845,
"mean_token_accuracy": 0.7651865869760514,
"num_tokens": 1998315819.0,
"step": 15350
},
{
"entropy": 1.7578125,
"epoch": 0.7985028072364317,
"grad_norm": 2.9156010553824627,
"learning_rate": 2.0779290384821525e-06,
"loss": 0.9861,
"mean_token_accuracy": 0.7607543677091598,
"num_tokens": 1999617055.0,
"step": 15360
},
{
"entropy": 1.7484375,
"epoch": 0.7990226658348929,
"grad_norm": 2.826460290431747,
"learning_rate": 2.072569407224783e-06,
"loss": 0.9908,
"mean_token_accuracy": 0.7638754487037659,
"num_tokens": 2000918301.0,
"step": 15370
},
{
"entropy": 1.749609375,
"epoch": 0.7995425244333542,
"grad_norm": 2.8295502757739976,
"learning_rate": 2.0672097759674136e-06,
"loss": 0.9388,
"mean_token_accuracy": 0.7760955929756165,
"num_tokens": 2002218237.0,
"step": 15380
},
{
"entropy": 1.748828125,
"epoch": 0.8000623830318153,
"grad_norm": 2.4526586219429523,
"learning_rate": 2.061850144710044e-06,
"loss": 0.9143,
"mean_token_accuracy": 0.7790937304496766,
"num_tokens": 2003521633.0,
"step": 15390
},
{
"entropy": 1.745703125,
"epoch": 0.8005822416302766,
"grad_norm": 2.9900840896364964,
"learning_rate": 2.0564905134526747e-06,
"loss": 0.9094,
"mean_token_accuracy": 0.7817790478467941,
"num_tokens": 2004821714.0,
"step": 15400
},
{
"epoch": 0.8005822416302766,
"eval_entropy": 1.7263671875,
"eval_loss": 0.9676682949066162,
"eval_mean_token_accuracy": 0.7661436215043068,
"eval_num_tokens": 2004821714.0,
"eval_runtime": 27.0134,
"eval_samples_per_second": 93.805,
"eval_steps_per_second": 1.481,
"step": 15400
},
{
"entropy": 1.741015625,
"epoch": 0.8011021002287377,
"grad_norm": 2.64294930828231,
"learning_rate": 2.051130882195305e-06,
"loss": 0.9528,
"mean_token_accuracy": 0.7709811985492706,
"num_tokens": 2006122734.0,
"step": 15410
},
{
"entropy": 1.77734375,
"epoch": 0.801621958827199,
"grad_norm": 2.778233467863379,
"learning_rate": 2.0457712509379357e-06,
"loss": 0.9504,
"mean_token_accuracy": 0.7710093140602112,
"num_tokens": 2007422888.0,
"step": 15420
},
{
"entropy": 1.770703125,
"epoch": 0.8021418174256603,
"grad_norm": 2.6545896514968006,
"learning_rate": 2.0404116196805663e-06,
"loss": 0.8716,
"mean_token_accuracy": 0.7870096385478973,
"num_tokens": 2008724680.0,
"step": 15430
},
{
"entropy": 1.75859375,
"epoch": 0.8026616760241214,
"grad_norm": 2.9590807246015665,
"learning_rate": 2.035051988423197e-06,
"loss": 1.0011,
"mean_token_accuracy": 0.7606434732675552,
"num_tokens": 2010027848.0,
"step": 15440
},
{
"entropy": 1.76015625,
"epoch": 0.8031815346225827,
"grad_norm": 2.6596843240463452,
"learning_rate": 2.0296923571658274e-06,
"loss": 0.9653,
"mean_token_accuracy": 0.769119480252266,
"num_tokens": 2011332009.0,
"step": 15450
},
{
"entropy": 1.75234375,
"epoch": 0.8037013932210438,
"grad_norm": 2.5608686062533943,
"learning_rate": 2.024332725908458e-06,
"loss": 0.9671,
"mean_token_accuracy": 0.7702234059572219,
"num_tokens": 2012634862.0,
"step": 15460
},
{
"entropy": 1.755078125,
"epoch": 0.8042212518195051,
"grad_norm": 2.75600097090661,
"learning_rate": 2.018973094651088e-06,
"loss": 0.9281,
"mean_token_accuracy": 0.7724601805210114,
"num_tokens": 2013937251.0,
"step": 15470
},
{
"entropy": 1.71796875,
"epoch": 0.8047411104179664,
"grad_norm": 2.8231982933198574,
"learning_rate": 2.0136134633937186e-06,
"loss": 0.8722,
"mean_token_accuracy": 0.7887499511241913,
"num_tokens": 2015238229.0,
"step": 15480
},
{
"entropy": 1.73984375,
"epoch": 0.8052609690164275,
"grad_norm": 2.7008721141806236,
"learning_rate": 2.008253832136349e-06,
"loss": 0.9258,
"mean_token_accuracy": 0.7752217233181,
"num_tokens": 2016541027.0,
"step": 15490
},
{
"entropy": 1.732421875,
"epoch": 0.8057808276148888,
"grad_norm": 2.981531924289679,
"learning_rate": 2.0028942008789797e-06,
"loss": 0.929,
"mean_token_accuracy": 0.7744055479764939,
"num_tokens": 2017843134.0,
"step": 15500
},
{
"epoch": 0.8057808276148888,
"eval_entropy": 1.7265625,
"eval_loss": 0.9664068818092346,
"eval_mean_token_accuracy": 0.7661684975028038,
"eval_num_tokens": 2017843134.0,
"eval_runtime": 26.9187,
"eval_samples_per_second": 94.135,
"eval_steps_per_second": 1.486,
"step": 15500
},
{
"entropy": 1.73671875,
"epoch": 0.8063006862133499,
"grad_norm": 2.389615878959754,
"learning_rate": 1.9975345696216102e-06,
"loss": 0.9276,
"mean_token_accuracy": 0.7758827984333039,
"num_tokens": 2019146045.0,
"step": 15510
},
{
"entropy": 1.7640625,
"epoch": 0.8068205448118112,
"grad_norm": 2.403639390727456,
"learning_rate": 1.9921749383642408e-06,
"loss": 0.9285,
"mean_token_accuracy": 0.7759013503789902,
"num_tokens": 2020447931.0,
"step": 15520
},
{
"entropy": 1.731640625,
"epoch": 0.8073404034102724,
"grad_norm": 2.6940339861876126,
"learning_rate": 1.9868153071068713e-06,
"loss": 0.9334,
"mean_token_accuracy": 0.7741329878568649,
"num_tokens": 2021750806.0,
"step": 15530
},
{
"entropy": 1.725390625,
"epoch": 0.8078602620087336,
"grad_norm": 2.7592349178169098,
"learning_rate": 1.981455675849502e-06,
"loss": 0.9086,
"mean_token_accuracy": 0.7807505369186402,
"num_tokens": 2023052041.0,
"step": 15540
},
{
"entropy": 1.72890625,
"epoch": 0.8083801206071949,
"grad_norm": 2.349710499840235,
"learning_rate": 1.9760960445921324e-06,
"loss": 0.9873,
"mean_token_accuracy": 0.7655806928873062,
"num_tokens": 2024354938.0,
"step": 15550
},
{
"entropy": 1.74375,
"epoch": 0.808899979205656,
"grad_norm": 2.6638257819236526,
"learning_rate": 1.970736413334763e-06,
"loss": 0.9215,
"mean_token_accuracy": 0.7790184497833252,
"num_tokens": 2025659113.0,
"step": 15560
},
{
"entropy": 1.768359375,
"epoch": 0.8094198378041173,
"grad_norm": 2.6305365155333633,
"learning_rate": 1.9653767820773935e-06,
"loss": 0.8849,
"mean_token_accuracy": 0.7838692307472229,
"num_tokens": 2026962671.0,
"step": 15570
},
{
"entropy": 1.75,
"epoch": 0.8099396964025785,
"grad_norm": 2.4615726744061455,
"learning_rate": 1.960017150820024e-06,
"loss": 0.9602,
"mean_token_accuracy": 0.769368925690651,
"num_tokens": 2028262855.0,
"step": 15580
},
{
"entropy": 1.73984375,
"epoch": 0.8104595550010397,
"grad_norm": 3.412973023975296,
"learning_rate": 1.954657519562654e-06,
"loss": 0.9569,
"mean_token_accuracy": 0.7714107632637024,
"num_tokens": 2029566130.0,
"step": 15590
},
{
"entropy": 1.78046875,
"epoch": 0.810979413599501,
"grad_norm": 2.58656933339121,
"learning_rate": 1.9492978883052847e-06,
"loss": 0.9658,
"mean_token_accuracy": 0.7667077034711838,
"num_tokens": 2030864847.0,
"step": 15600
},
{
"epoch": 0.810979413599501,
"eval_entropy": 1.7404296875,
"eval_loss": 0.9655699133872986,
"eval_mean_token_accuracy": 0.7665741622447968,
"eval_num_tokens": 2030864847.0,
"eval_runtime": 27.0271,
"eval_samples_per_second": 93.758,
"eval_steps_per_second": 1.48,
"step": 15600
},
{
"entropy": 1.730078125,
"epoch": 0.8114992721979621,
"grad_norm": 2.6205361502183906,
"learning_rate": 1.9439382570479153e-06,
"loss": 0.9286,
"mean_token_accuracy": 0.7780615568161011,
"num_tokens": 2032168507.0,
"step": 15610
},
{
"entropy": 1.733984375,
"epoch": 0.8120191307964234,
"grad_norm": 2.623219311905963,
"learning_rate": 1.938578625790546e-06,
"loss": 0.9576,
"mean_token_accuracy": 0.7680694162845612,
"num_tokens": 2033471151.0,
"step": 15620
},
{
"entropy": 1.73984375,
"epoch": 0.8125389893948846,
"grad_norm": 2.5043747024917464,
"learning_rate": 1.9332189945331764e-06,
"loss": 0.9269,
"mean_token_accuracy": 0.7779242634773255,
"num_tokens": 2034773685.0,
"step": 15630
},
{
"entropy": 1.71484375,
"epoch": 0.8130588479933458,
"grad_norm": 3.0001331823954254,
"learning_rate": 1.927859363275807e-06,
"loss": 0.913,
"mean_token_accuracy": 0.7791012465953827,
"num_tokens": 2036076915.0,
"step": 15640
},
{
"entropy": 1.719921875,
"epoch": 0.813578706591807,
"grad_norm": 2.6428433025924476,
"learning_rate": 1.9224997320184375e-06,
"loss": 0.8758,
"mean_token_accuracy": 0.7873733162879943,
"num_tokens": 2037379131.0,
"step": 15650
},
{
"entropy": 1.749609375,
"epoch": 0.8140985651902682,
"grad_norm": 2.1860733020874616,
"learning_rate": 1.9171401007610676e-06,
"loss": 0.9551,
"mean_token_accuracy": 0.769901230931282,
"num_tokens": 2038680859.0,
"step": 15660
},
{
"entropy": 1.755078125,
"epoch": 0.8146184237887295,
"grad_norm": 2.7067500924378742,
"learning_rate": 1.911780469503698e-06,
"loss": 0.9362,
"mean_token_accuracy": 0.7747596085071564,
"num_tokens": 2039982617.0,
"step": 15670
},
{
"entropy": 1.74609375,
"epoch": 0.8151382823871907,
"grad_norm": 2.7093956296743396,
"learning_rate": 1.9064208382463287e-06,
"loss": 0.9235,
"mean_token_accuracy": 0.7800243735313416,
"num_tokens": 2041287218.0,
"step": 15680
},
{
"entropy": 1.726171875,
"epoch": 0.8156581409856519,
"grad_norm": 2.4701129991802007,
"learning_rate": 1.9010612069889592e-06,
"loss": 0.9293,
"mean_token_accuracy": 0.7752142995595932,
"num_tokens": 2042589796.0,
"step": 15690
},
{
"entropy": 1.743359375,
"epoch": 0.8161779995841131,
"grad_norm": 2.5310284591819032,
"learning_rate": 1.8957015757315898e-06,
"loss": 0.8883,
"mean_token_accuracy": 0.7871850520372391,
"num_tokens": 2043891283.0,
"step": 15700
},
{
"epoch": 0.8161779995841131,
"eval_entropy": 1.72578125,
"eval_loss": 0.9634828567504883,
"eval_mean_token_accuracy": 0.7672611489892006,
"eval_num_tokens": 2043891283.0,
"eval_runtime": 26.9874,
"eval_samples_per_second": 93.896,
"eval_steps_per_second": 1.482,
"step": 15700
},
{
"entropy": 1.72578125,
"epoch": 0.8166978581825743,
"grad_norm": 2.698086196401316,
"learning_rate": 1.8903419444742201e-06,
"loss": 0.9313,
"mean_token_accuracy": 0.778390496969223,
"num_tokens": 2045193061.0,
"step": 15710
},
{
"entropy": 1.74296875,
"epoch": 0.8172177167810356,
"grad_norm": 2.6865858008854344,
"learning_rate": 1.8849823132168507e-06,
"loss": 0.9399,
"mean_token_accuracy": 0.7737501919269562,
"num_tokens": 2046493482.0,
"step": 15720
},
{
"entropy": 1.75390625,
"epoch": 0.8177375753794968,
"grad_norm": 2.4129025861546975,
"learning_rate": 1.8796226819594812e-06,
"loss": 0.9548,
"mean_token_accuracy": 0.7721105575561523,
"num_tokens": 2047795663.0,
"step": 15730
},
{
"entropy": 1.737109375,
"epoch": 0.818257433977958,
"grad_norm": 2.5099103605493065,
"learning_rate": 1.8742630507021118e-06,
"loss": 0.9293,
"mean_token_accuracy": 0.7754753738641739,
"num_tokens": 2049096942.0,
"step": 15740
},
{
"entropy": 1.7265625,
"epoch": 0.8187772925764192,
"grad_norm": 2.9514356795629473,
"learning_rate": 1.8689034194447423e-06,
"loss": 0.9408,
"mean_token_accuracy": 0.773339468240738,
"num_tokens": 2050398699.0,
"step": 15750
},
{
"entropy": 1.721484375,
"epoch": 0.8192971511748804,
"grad_norm": 3.712402823028229,
"learning_rate": 1.8635437881873728e-06,
"loss": 0.9374,
"mean_token_accuracy": 0.7783810555934906,
"num_tokens": 2051700064.0,
"step": 15760
},
{
"entropy": 1.755078125,
"epoch": 0.8198170097733416,
"grad_norm": 2.754622417655119,
"learning_rate": 1.8581841569300032e-06,
"loss": 0.917,
"mean_token_accuracy": 0.7803317815065384,
"num_tokens": 2053003789.0,
"step": 15770
},
{
"entropy": 1.740625,
"epoch": 0.8203368683718029,
"grad_norm": 2.7169359034724345,
"learning_rate": 1.8528245256726337e-06,
"loss": 0.9194,
"mean_token_accuracy": 0.7763634890317916,
"num_tokens": 2054307501.0,
"step": 15780
},
{
"entropy": 1.713671875,
"epoch": 0.8208567269702641,
"grad_norm": 2.923546293647611,
"learning_rate": 1.8474648944152643e-06,
"loss": 0.9434,
"mean_token_accuracy": 0.7746605515480042,
"num_tokens": 2055611018.0,
"step": 15790
},
{
"entropy": 1.713671875,
"epoch": 0.8213765855687253,
"grad_norm": 2.159840626147234,
"learning_rate": 1.8421052631578948e-06,
"loss": 0.9728,
"mean_token_accuracy": 0.7704523235559464,
"num_tokens": 2056914610.0,
"step": 15800
},
{
"epoch": 0.8213765855687253,
"eval_entropy": 1.7263671875,
"eval_loss": 0.9622150659561157,
"eval_mean_token_accuracy": 0.7671663194894791,
"eval_num_tokens": 2056914610.0,
"eval_runtime": 26.9749,
"eval_samples_per_second": 93.939,
"eval_steps_per_second": 1.483,
"step": 15800
},
{
"entropy": 1.759375,
"epoch": 0.8218964441671865,
"grad_norm": 2.4561843620985853,
"learning_rate": 1.8367456319005254e-06,
"loss": 0.9078,
"mean_token_accuracy": 0.7833277344703674,
"num_tokens": 2058215961.0,
"step": 15810
},
{
"entropy": 1.716015625,
"epoch": 0.8224163027656477,
"grad_norm": 2.8168158441736364,
"learning_rate": 1.8313860006431557e-06,
"loss": 0.9633,
"mean_token_accuracy": 0.7729220271110535,
"num_tokens": 2059515936.0,
"step": 15820
},
{
"entropy": 1.70859375,
"epoch": 0.822936161364109,
"grad_norm": 2.8891757537345977,
"learning_rate": 1.8260263693857862e-06,
"loss": 0.932,
"mean_token_accuracy": 0.7764736473560333,
"num_tokens": 2060817775.0,
"step": 15830
},
{
"entropy": 1.709765625,
"epoch": 0.8234560199625702,
"grad_norm": 2.6228733255982193,
"learning_rate": 1.8206667381284168e-06,
"loss": 0.9489,
"mean_token_accuracy": 0.7705067306756973,
"num_tokens": 2062120940.0,
"step": 15840
},
{
"entropy": 1.740234375,
"epoch": 0.8239758785610314,
"grad_norm": 2.473137229102988,
"learning_rate": 1.8153071068710473e-06,
"loss": 0.9325,
"mean_token_accuracy": 0.7753347307443619,
"num_tokens": 2063422705.0,
"step": 15850
},
{
"entropy": 1.734375,
"epoch": 0.8244957371594926,
"grad_norm": 3.0200395602077124,
"learning_rate": 1.8099474756136779e-06,
"loss": 0.9078,
"mean_token_accuracy": 0.7792269110679626,
"num_tokens": 2064721584.0,
"step": 15860
},
{
"entropy": 1.748828125,
"epoch": 0.8250155957579538,
"grad_norm": 3.071489930667252,
"learning_rate": 1.8045878443563084e-06,
"loss": 0.9575,
"mean_token_accuracy": 0.772504699230194,
"num_tokens": 2066025305.0,
"step": 15870
},
{
"entropy": 1.744140625,
"epoch": 0.8255354543564151,
"grad_norm": 2.5627443451569394,
"learning_rate": 1.7992282130989388e-06,
"loss": 0.9012,
"mean_token_accuracy": 0.7840084612369538,
"num_tokens": 2067326574.0,
"step": 15880
},
{
"entropy": 1.728125,
"epoch": 0.8260553129548762,
"grad_norm": 3.439022584665469,
"learning_rate": 1.7938685818415693e-06,
"loss": 0.8851,
"mean_token_accuracy": 0.7896751374006271,
"num_tokens": 2068628516.0,
"step": 15890
},
{
"entropy": 1.715234375,
"epoch": 0.8265751715533375,
"grad_norm": 3.008791740307406,
"learning_rate": 1.7885089505841999e-06,
"loss": 0.9455,
"mean_token_accuracy": 0.7746002614498139,
"num_tokens": 2069930150.0,
"step": 15900
},
{
"epoch": 0.8265751715533375,
"eval_entropy": 1.726953125,
"eval_loss": 0.9613496661186218,
"eval_mean_token_accuracy": 0.7673721894621849,
"eval_num_tokens": 2069930150.0,
"eval_runtime": 27.0286,
"eval_samples_per_second": 93.753,
"eval_steps_per_second": 1.48,
"step": 15900
},
{
"entropy": 1.745703125,
"epoch": 0.8270950301517987,
"grad_norm": 2.5469827898597783,
"learning_rate": 1.7831493193268304e-06,
"loss": 0.9058,
"mean_token_accuracy": 0.7809749066829681,
"num_tokens": 2071231183.0,
"step": 15910
},
{
"entropy": 1.71640625,
"epoch": 0.8276148887502599,
"grad_norm": 2.7252218973824363,
"learning_rate": 1.777789688069461e-06,
"loss": 0.969,
"mean_token_accuracy": 0.7695471167564392,
"num_tokens": 2072535179.0,
"step": 15920
},
{
"entropy": 1.755859375,
"epoch": 0.8281347473487212,
"grad_norm": 2.4769244655810176,
"learning_rate": 1.7724300568120915e-06,
"loss": 0.9601,
"mean_token_accuracy": 0.7689635992050171,
"num_tokens": 2073837164.0,
"step": 15930
},
{
"entropy": 1.73515625,
"epoch": 0.8286546059471823,
"grad_norm": 2.345844142788363,
"learning_rate": 1.7670704255547218e-06,
"loss": 0.9209,
"mean_token_accuracy": 0.7817540794610978,
"num_tokens": 2075138956.0,
"step": 15940
},
{
"entropy": 1.77578125,
"epoch": 0.8291744645456436,
"grad_norm": 3.1043363511587465,
"learning_rate": 1.7617107942973524e-06,
"loss": 0.9481,
"mean_token_accuracy": 0.773292264342308,
"num_tokens": 2076440590.0,
"step": 15950
},
{
"entropy": 1.759765625,
"epoch": 0.8296943231441049,
"grad_norm": 3.0213695396458617,
"learning_rate": 1.756351163039983e-06,
"loss": 0.9224,
"mean_token_accuracy": 0.7792001992464066,
"num_tokens": 2077743917.0,
"step": 15960
},
{
"entropy": 1.7734375,
"epoch": 0.830214181742566,
"grad_norm": 3.168815312755156,
"learning_rate": 1.7509915317826135e-06,
"loss": 0.928,
"mean_token_accuracy": 0.7741633504629135,
"num_tokens": 2079045304.0,
"step": 15970
},
{
"entropy": 1.74609375,
"epoch": 0.8307340403410273,
"grad_norm": 2.6894626463865445,
"learning_rate": 1.745631900525244e-06,
"loss": 0.9642,
"mean_token_accuracy": 0.7661105632781983,
"num_tokens": 2080348810.0,
"step": 15980
},
{
"entropy": 1.743359375,
"epoch": 0.8312538989394884,
"grad_norm": 2.504516333258791,
"learning_rate": 1.7402722692678744e-06,
"loss": 0.8983,
"mean_token_accuracy": 0.7846140921115875,
"num_tokens": 2081650860.0,
"step": 15990
},
{
"entropy": 1.734375,
"epoch": 0.8317737575379497,
"grad_norm": 3.3670984750442425,
"learning_rate": 1.734912638010505e-06,
"loss": 0.9035,
"mean_token_accuracy": 0.7824330866336823,
"num_tokens": 2082954446.0,
"step": 16000
},
{
"epoch": 0.8317737575379497,
"eval_entropy": 1.7357421875,
"eval_loss": 0.9585598707199097,
"eval_mean_token_accuracy": 0.7681140899658203,
"eval_num_tokens": 2082954446.0,
"eval_runtime": 26.9722,
"eval_samples_per_second": 93.949,
"eval_steps_per_second": 1.483,
"step": 16000
},
{
"entropy": 1.7453125,
"epoch": 0.8322936161364108,
"grad_norm": 2.423004510101209,
"learning_rate": 1.7295530067531355e-06,
"loss": 0.9638,
"mean_token_accuracy": 0.7691928565502166,
"num_tokens": 2084255666.0,
"step": 16010
},
{
"entropy": 1.76171875,
"epoch": 0.8328134747348721,
"grad_norm": 3.024090721468171,
"learning_rate": 1.724193375495766e-06,
"loss": 0.9217,
"mean_token_accuracy": 0.7786274492740631,
"num_tokens": 2085557382.0,
"step": 16020
},
{
"entropy": 1.751171875,
"epoch": 0.8333333333333334,
"grad_norm": 2.441476272684357,
"learning_rate": 1.7188337442383965e-06,
"loss": 0.8835,
"mean_token_accuracy": 0.7877138197422028,
"num_tokens": 2086857044.0,
"step": 16030
},
{
"entropy": 1.715234375,
"epoch": 0.8338531919317945,
"grad_norm": 2.6692996764672565,
"learning_rate": 1.713474112981027e-06,
"loss": 0.9334,
"mean_token_accuracy": 0.7767039626836777,
"num_tokens": 2088161492.0,
"step": 16040
},
{
"entropy": 1.744140625,
"epoch": 0.8343730505302558,
"grad_norm": 2.654836468343714,
"learning_rate": 1.7081144817236574e-06,
"loss": 0.9796,
"mean_token_accuracy": 0.7646795302629471,
"num_tokens": 2089464649.0,
"step": 16050
},
{
"entropy": 1.759375,
"epoch": 0.8348929091287169,
"grad_norm": 3.5385183890782104,
"learning_rate": 1.702754850466288e-06,
"loss": 0.9046,
"mean_token_accuracy": 0.782479801774025,
"num_tokens": 2090763338.0,
"step": 16060
},
{
"entropy": 1.7515625,
"epoch": 0.8354127677271782,
"grad_norm": 2.6996812778780535,
"learning_rate": 1.6973952192089185e-06,
"loss": 0.9542,
"mean_token_accuracy": 0.7699927508831024,
"num_tokens": 2092066301.0,
"step": 16070
},
{
"entropy": 1.7421875,
"epoch": 0.8359326263256395,
"grad_norm": 2.83873423724771,
"learning_rate": 1.692035587951549e-06,
"loss": 0.9327,
"mean_token_accuracy": 0.7752880096435547,
"num_tokens": 2093367427.0,
"step": 16080
},
{
"entropy": 1.717578125,
"epoch": 0.8364524849241006,
"grad_norm": 3.0215346876655444,
"learning_rate": 1.6866759566941796e-06,
"loss": 0.8896,
"mean_token_accuracy": 0.7859819173812866,
"num_tokens": 2094671021.0,
"step": 16090
},
{
"entropy": 1.72265625,
"epoch": 0.8369723435225619,
"grad_norm": 2.9023201436233,
"learning_rate": 1.6813163254368102e-06,
"loss": 0.966,
"mean_token_accuracy": 0.767346304655075,
"num_tokens": 2095972173.0,
"step": 16100
},
{
"epoch": 0.8369723435225619,
"eval_entropy": 1.7255859375,
"eval_loss": 0.9566917419433594,
"eval_mean_token_accuracy": 0.7686152294278145,
"eval_num_tokens": 2095972173.0,
"eval_runtime": 26.9841,
"eval_samples_per_second": 93.907,
"eval_steps_per_second": 1.482,
"step": 16100
},
{
"entropy": 1.753125,
"epoch": 0.837492202121023,
"grad_norm": 2.5208830062322733,
"learning_rate": 1.6759566941794405e-06,
"loss": 0.9739,
"mean_token_accuracy": 0.7665610790252686,
"num_tokens": 2097273035.0,
"step": 16110
},
{
"entropy": 1.737890625,
"epoch": 0.8380120607194843,
"grad_norm": 2.7100679762499187,
"learning_rate": 1.670597062922071e-06,
"loss": 0.9105,
"mean_token_accuracy": 0.7814458221197128,
"num_tokens": 2098571516.0,
"step": 16120
},
{
"entropy": 1.740625,
"epoch": 0.8385319193179456,
"grad_norm": 2.8737118056656294,
"learning_rate": 1.6652374316647016e-06,
"loss": 0.9163,
"mean_token_accuracy": 0.780165022611618,
"num_tokens": 2099873187.0,
"step": 16130
},
{
"entropy": 1.729296875,
"epoch": 0.8390517779164067,
"grad_norm": 3.0293187792800755,
"learning_rate": 1.6598778004073321e-06,
"loss": 0.9109,
"mean_token_accuracy": 0.7806250125169754,
"num_tokens": 2101175365.0,
"step": 16140
},
{
"entropy": 1.727734375,
"epoch": 0.839571636514868,
"grad_norm": 3.1188730583625977,
"learning_rate": 1.6545181691499627e-06,
"loss": 0.9277,
"mean_token_accuracy": 0.7775994211435318,
"num_tokens": 2102477889.0,
"step": 16150
},
{
"entropy": 1.71875,
"epoch": 0.8400914951133291,
"grad_norm": 3.059703992178224,
"learning_rate": 1.649158537892593e-06,
"loss": 0.9375,
"mean_token_accuracy": 0.7773182272911072,
"num_tokens": 2103780930.0,
"step": 16160
},
{
"entropy": 1.733984375,
"epoch": 0.8406113537117904,
"grad_norm": 2.5837490997291352,
"learning_rate": 1.6437989066352236e-06,
"loss": 0.998,
"mean_token_accuracy": 0.759396106004715,
"num_tokens": 2105083256.0,
"step": 16170
},
{
"entropy": 1.734375,
"epoch": 0.8411312123102516,
"grad_norm": 2.6812884876270977,
"learning_rate": 1.6384392753778541e-06,
"loss": 0.9765,
"mean_token_accuracy": 0.7652894645929337,
"num_tokens": 2106386922.0,
"step": 16180
},
{
"entropy": 1.730078125,
"epoch": 0.8416510709087128,
"grad_norm": 2.9639717448742884,
"learning_rate": 1.6330796441204847e-06,
"loss": 0.9143,
"mean_token_accuracy": 0.7797675400972366,
"num_tokens": 2107687417.0,
"step": 16190
},
{
"entropy": 1.742578125,
"epoch": 0.8421709295071741,
"grad_norm": 2.5440409665021284,
"learning_rate": 1.6277200128631152e-06,
"loss": 0.8666,
"mean_token_accuracy": 0.7922940731048584,
"num_tokens": 2108989390.0,
"step": 16200
},
{
"epoch": 0.8421709295071741,
"eval_entropy": 1.7189453125,
"eval_loss": 0.9552922248840332,
"eval_mean_token_accuracy": 0.7691604793071747,
"eval_num_tokens": 2108989390.0,
"eval_runtime": 26.9505,
"eval_samples_per_second": 94.024,
"eval_steps_per_second": 1.484,
"step": 16200
},
{
"entropy": 1.74296875,
"epoch": 0.8426907881056352,
"grad_norm": 2.855139695153378,
"learning_rate": 1.6223603816057457e-06,
"loss": 0.9737,
"mean_token_accuracy": 0.7651296943426132,
"num_tokens": 2110289835.0,
"step": 16210
},
{
"entropy": 1.728515625,
"epoch": 0.8432106467040965,
"grad_norm": 2.636205026729025,
"learning_rate": 1.617000750348376e-06,
"loss": 0.9721,
"mean_token_accuracy": 0.7689586490392685,
"num_tokens": 2111592862.0,
"step": 16220
},
{
"entropy": 1.70234375,
"epoch": 0.8437305053025577,
"grad_norm": 2.5433147434918393,
"learning_rate": 1.6116411190910066e-06,
"loss": 0.9227,
"mean_token_accuracy": 0.77628353536129,
"num_tokens": 2112895643.0,
"step": 16230
},
{
"entropy": 1.734765625,
"epoch": 0.8442503639010189,
"grad_norm": 2.5416545503103514,
"learning_rate": 1.6062814878336372e-06,
"loss": 0.9017,
"mean_token_accuracy": 0.784505033493042,
"num_tokens": 2114196009.0,
"step": 16240
},
{
"entropy": 1.735546875,
"epoch": 0.8447702224994802,
"grad_norm": 3.0373292209739677,
"learning_rate": 1.6009218565762677e-06,
"loss": 0.9704,
"mean_token_accuracy": 0.7696331709623336,
"num_tokens": 2115500681.0,
"step": 16250
},
{
"entropy": 1.734765625,
"epoch": 0.8452900810979413,
"grad_norm": 2.6031354138899614,
"learning_rate": 1.5955622253188983e-06,
"loss": 0.9087,
"mean_token_accuracy": 0.7787926286458969,
"num_tokens": 2116803550.0,
"step": 16260
},
{
"entropy": 1.7375,
"epoch": 0.8458099396964026,
"grad_norm": 2.6488803016177025,
"learning_rate": 1.5902025940615288e-06,
"loss": 0.9417,
"mean_token_accuracy": 0.7740314453840256,
"num_tokens": 2118104254.0,
"step": 16270
},
{
"entropy": 1.751171875,
"epoch": 0.8463297982948638,
"grad_norm": 2.637268795406206,
"learning_rate": 1.5848429628041592e-06,
"loss": 0.9286,
"mean_token_accuracy": 0.7767329603433609,
"num_tokens": 2119406303.0,
"step": 16280
},
{
"entropy": 1.732421875,
"epoch": 0.846849656893325,
"grad_norm": 2.394615909771607,
"learning_rate": 1.5794833315467897e-06,
"loss": 0.9479,
"mean_token_accuracy": 0.7728170096874237,
"num_tokens": 2120710623.0,
"step": 16290
},
{
"entropy": 1.738671875,
"epoch": 0.8473695154917862,
"grad_norm": 2.2436601622340433,
"learning_rate": 1.5741237002894202e-06,
"loss": 0.9602,
"mean_token_accuracy": 0.7680613219738006,
"num_tokens": 2122012807.0,
"step": 16300
},
{
"epoch": 0.8473695154917862,
"eval_entropy": 1.7228515625,
"eval_loss": 0.9543511867523193,
"eval_mean_token_accuracy": 0.7691166132688523,
"eval_num_tokens": 2122012807.0,
"eval_runtime": 27.073,
"eval_samples_per_second": 93.599,
"eval_steps_per_second": 1.477,
"step": 16300
},
{
"entropy": 1.74453125,
"epoch": 0.8478893740902474,
"grad_norm": 2.6923787969888555,
"learning_rate": 1.5687640690320508e-06,
"loss": 0.8979,
"mean_token_accuracy": 0.7819160938262939,
"num_tokens": 2123315484.0,
"step": 16310
},
{
"entropy": 1.741015625,
"epoch": 0.8484092326887087,
"grad_norm": 2.9134036495243008,
"learning_rate": 1.5634044377746813e-06,
"loss": 1.035,
"mean_token_accuracy": 0.7523791760206222,
"num_tokens": 2124616787.0,
"step": 16320
},
{
"entropy": 1.7140625,
"epoch": 0.8489290912871699,
"grad_norm": 2.3481678499304097,
"learning_rate": 1.5580448065173117e-06,
"loss": 0.9424,
"mean_token_accuracy": 0.7720117598772049,
"num_tokens": 2125918773.0,
"step": 16330
},
{
"entropy": 1.721484375,
"epoch": 0.8494489498856311,
"grad_norm": 3.9328588759706777,
"learning_rate": 1.5526851752599422e-06,
"loss": 0.9122,
"mean_token_accuracy": 0.7787463366985321,
"num_tokens": 2127221915.0,
"step": 16340
},
{
"entropy": 1.75234375,
"epoch": 0.8499688084840923,
"grad_norm": 2.828748650881853,
"learning_rate": 1.5473255440025728e-06,
"loss": 0.948,
"mean_token_accuracy": 0.7708906382322311,
"num_tokens": 2128524726.0,
"step": 16350
},
{
"entropy": 1.712890625,
"epoch": 0.8504886670825536,
"grad_norm": 2.528558949778028,
"learning_rate": 1.5419659127452033e-06,
"loss": 0.9274,
"mean_token_accuracy": 0.7765229880809784,
"num_tokens": 2129828666.0,
"step": 16360
},
{
"entropy": 1.741015625,
"epoch": 0.8510085256810148,
"grad_norm": 2.602462912902093,
"learning_rate": 1.5366062814878339e-06,
"loss": 0.9226,
"mean_token_accuracy": 0.7817203342914582,
"num_tokens": 2131131232.0,
"step": 16370
},
{
"entropy": 1.7265625,
"epoch": 0.851528384279476,
"grad_norm": 3.0287872571556953,
"learning_rate": 1.5312466502304644e-06,
"loss": 0.9611,
"mean_token_accuracy": 0.7717675298452378,
"num_tokens": 2132429718.0,
"step": 16380
},
{
"entropy": 1.7046875,
"epoch": 0.8520482428779372,
"grad_norm": 2.4556728671838526,
"learning_rate": 1.5258870189730947e-06,
"loss": 0.9031,
"mean_token_accuracy": 0.7825746595859527,
"num_tokens": 2133732941.0,
"step": 16390
},
{
"entropy": 1.734765625,
"epoch": 0.8525681014763984,
"grad_norm": 2.7558061842577932,
"learning_rate": 1.5205273877157253e-06,
"loss": 0.9091,
"mean_token_accuracy": 0.7813933610916137,
"num_tokens": 2135033508.0,
"step": 16400
},
{
"epoch": 0.8525681014763984,
"eval_entropy": 1.719140625,
"eval_loss": 0.9531128406524658,
"eval_mean_token_accuracy": 0.7696149244904518,
"eval_num_tokens": 2135033508.0,
"eval_runtime": 27.4799,
"eval_samples_per_second": 92.213,
"eval_steps_per_second": 1.456,
"step": 16400
},
{
"entropy": 1.743359375,
"epoch": 0.8530879600748597,
"grad_norm": 2.647292092819134,
"learning_rate": 1.5151677564583558e-06,
"loss": 0.9258,
"mean_token_accuracy": 0.7781663179397583,
"num_tokens": 2136334760.0,
"step": 16410
},
{
"entropy": 1.72421875,
"epoch": 0.8536078186733208,
"grad_norm": 2.9115234784360458,
"learning_rate": 1.5098081252009864e-06,
"loss": 0.93,
"mean_token_accuracy": 0.7775302052497863,
"num_tokens": 2137636096.0,
"step": 16420
},
{
"entropy": 1.711328125,
"epoch": 0.8541276772717821,
"grad_norm": 2.6345666844065385,
"learning_rate": 1.504448493943617e-06,
"loss": 0.8662,
"mean_token_accuracy": 0.7918179124593735,
"num_tokens": 2138935836.0,
"step": 16430
},
{
"entropy": 1.696875,
"epoch": 0.8546475358702433,
"grad_norm": 2.9727785993297453,
"learning_rate": 1.4990888626862475e-06,
"loss": 0.8714,
"mean_token_accuracy": 0.7898651868104934,
"num_tokens": 2140234446.0,
"step": 16440
},
{
"entropy": 1.7234375,
"epoch": 0.8551673944687045,
"grad_norm": 3.067584904425998,
"learning_rate": 1.4937292314288778e-06,
"loss": 0.9449,
"mean_token_accuracy": 0.7728902637958527,
"num_tokens": 2141535319.0,
"step": 16450
},
{
"entropy": 1.721484375,
"epoch": 0.8556872530671658,
"grad_norm": 2.829859454861725,
"learning_rate": 1.4883696001715084e-06,
"loss": 0.9327,
"mean_token_accuracy": 0.7761115252971649,
"num_tokens": 2142838080.0,
"step": 16460
},
{
"entropy": 1.75,
"epoch": 0.8562071116656269,
"grad_norm": 2.469919946754567,
"learning_rate": 1.483009968914139e-06,
"loss": 0.951,
"mean_token_accuracy": 0.7722401797771454,
"num_tokens": 2144136970.0,
"step": 16470
},
{
"entropy": 1.7265625,
"epoch": 0.8567269702640882,
"grad_norm": 2.7399119624912456,
"learning_rate": 1.4776503376567694e-06,
"loss": 0.8978,
"mean_token_accuracy": 0.783169737458229,
"num_tokens": 2145439210.0,
"step": 16480
},
{
"entropy": 1.708984375,
"epoch": 0.8572468288625494,
"grad_norm": 2.6092275340702415,
"learning_rate": 1.4722907063994e-06,
"loss": 0.9573,
"mean_token_accuracy": 0.7666968882083893,
"num_tokens": 2146741766.0,
"step": 16490
},
{
"entropy": 1.70625,
"epoch": 0.8577666874610106,
"grad_norm": 2.749883959887991,
"learning_rate": 1.4669310751420303e-06,
"loss": 0.9052,
"mean_token_accuracy": 0.779526287317276,
"num_tokens": 2148043704.0,
"step": 16500
},
{
"epoch": 0.8577666874610106,
"eval_entropy": 1.7025390625,
"eval_loss": 0.9526218771934509,
"eval_mean_token_accuracy": 0.7698836922645569,
"eval_num_tokens": 2148043704.0,
"eval_runtime": 26.9777,
"eval_samples_per_second": 93.929,
"eval_steps_per_second": 1.483,
"step": 16500
},
{
"entropy": 1.705859375,
"epoch": 0.8582865460594719,
"grad_norm": 2.777597963931321,
"learning_rate": 1.4615714438846609e-06,
"loss": 0.9121,
"mean_token_accuracy": 0.7795868694782258,
"num_tokens": 2149345925.0,
"step": 16510
},
{
"entropy": 1.723046875,
"epoch": 0.858806404657933,
"grad_norm": 2.823640887296501,
"learning_rate": 1.4562118126272914e-06,
"loss": 0.8582,
"mean_token_accuracy": 0.7893594294786453,
"num_tokens": 2150648269.0,
"step": 16520
},
{
"entropy": 1.72109375,
"epoch": 0.8593262632563943,
"grad_norm": 2.8486683664488717,
"learning_rate": 1.450852181369922e-06,
"loss": 0.8864,
"mean_token_accuracy": 0.7860481113195419,
"num_tokens": 2151949603.0,
"step": 16530
},
{
"entropy": 1.708984375,
"epoch": 0.8598461218548554,
"grad_norm": 2.689855866901763,
"learning_rate": 1.4454925501125525e-06,
"loss": 0.9244,
"mean_token_accuracy": 0.7751286506652832,
"num_tokens": 2153249159.0,
"step": 16540
},
{
"entropy": 1.744921875,
"epoch": 0.8603659804533167,
"grad_norm": 2.6572149530359637,
"learning_rate": 1.440132918855183e-06,
"loss": 0.8809,
"mean_token_accuracy": 0.7885504633188247,
"num_tokens": 2154551249.0,
"step": 16550
},
{
"entropy": 1.73671875,
"epoch": 0.860885839051778,
"grad_norm": 2.771161450314403,
"learning_rate": 1.4347732875978134e-06,
"loss": 1.0101,
"mean_token_accuracy": 0.7597829103469849,
"num_tokens": 2155854943.0,
"step": 16560
},
{
"entropy": 1.7390625,
"epoch": 0.8614056976502391,
"grad_norm": 2.677817204375468,
"learning_rate": 1.429413656340444e-06,
"loss": 0.8666,
"mean_token_accuracy": 0.7896337062120438,
"num_tokens": 2157157496.0,
"step": 16570
},
{
"entropy": 1.710546875,
"epoch": 0.8619255562487004,
"grad_norm": 2.9356463061337426,
"learning_rate": 1.4240540250830745e-06,
"loss": 0.9395,
"mean_token_accuracy": 0.775232857465744,
"num_tokens": 2158458076.0,
"step": 16580
},
{
"entropy": 1.714453125,
"epoch": 0.8624454148471615,
"grad_norm": 2.6110604337009793,
"learning_rate": 1.418694393825705e-06,
"loss": 0.8623,
"mean_token_accuracy": 0.7934818625450134,
"num_tokens": 2159757957.0,
"step": 16590
},
{
"entropy": 1.74140625,
"epoch": 0.8629652734456228,
"grad_norm": 2.9874469768693483,
"learning_rate": 1.4133347625683356e-06,
"loss": 0.9419,
"mean_token_accuracy": 0.7718954682350159,
"num_tokens": 2161056900.0,
"step": 16600
},
{
"epoch": 0.8629652734456228,
"eval_entropy": 1.7078125,
"eval_loss": 0.9509727954864502,
"eval_mean_token_accuracy": 0.7703008681535721,
"eval_num_tokens": 2161056900.0,
"eval_runtime": 26.9723,
"eval_samples_per_second": 93.948,
"eval_steps_per_second": 1.483,
"step": 16600
},
{
"entropy": 1.725,
"epoch": 0.8634851320440841,
"grad_norm": 2.4856763806561455,
"learning_rate": 1.4079751313109661e-06,
"loss": 0.9639,
"mean_token_accuracy": 0.7730087131261826,
"num_tokens": 2162359467.0,
"step": 16610
},
{
"entropy": 1.70234375,
"epoch": 0.8640049906425452,
"grad_norm": 3.106968870467817,
"learning_rate": 1.4026155000535965e-06,
"loss": 0.859,
"mean_token_accuracy": 0.7933248221874237,
"num_tokens": 2163659973.0,
"step": 16620
},
{
"entropy": 1.697265625,
"epoch": 0.8645248492410065,
"grad_norm": 2.999819935757903,
"learning_rate": 1.397255868796227e-06,
"loss": 0.9573,
"mean_token_accuracy": 0.7732995897531509,
"num_tokens": 2164962404.0,
"step": 16630
},
{
"entropy": 1.70625,
"epoch": 0.8650447078394676,
"grad_norm": 3.121448215468768,
"learning_rate": 1.3918962375388576e-06,
"loss": 0.8477,
"mean_token_accuracy": 0.7923937380313874,
"num_tokens": 2166261902.0,
"step": 16640
},
{
"entropy": 1.707421875,
"epoch": 0.8655645664379289,
"grad_norm": 2.649891292518444,
"learning_rate": 1.386536606281488e-06,
"loss": 0.8474,
"mean_token_accuracy": 0.7965181291103363,
"num_tokens": 2167559920.0,
"step": 16650
},
{
"entropy": 1.73984375,
"epoch": 0.86608442503639,
"grad_norm": 2.5034502961546425,
"learning_rate": 1.3811769750241187e-06,
"loss": 0.9187,
"mean_token_accuracy": 0.778735563158989,
"num_tokens": 2168861923.0,
"step": 16660
},
{
"entropy": 1.706640625,
"epoch": 0.8666042836348513,
"grad_norm": 2.985761711526429,
"learning_rate": 1.375817343766749e-06,
"loss": 0.9535,
"mean_token_accuracy": 0.774389672279358,
"num_tokens": 2170164388.0,
"step": 16670
},
{
"entropy": 1.715234375,
"epoch": 0.8671241422333126,
"grad_norm": 2.6521664093918793,
"learning_rate": 1.3704577125093795e-06,
"loss": 0.8949,
"mean_token_accuracy": 0.7816339761018753,
"num_tokens": 2171467633.0,
"step": 16680
},
{
"entropy": 1.728515625,
"epoch": 0.8676440008317737,
"grad_norm": 2.573022426104598,
"learning_rate": 1.36509808125201e-06,
"loss": 0.9094,
"mean_token_accuracy": 0.7817778795957565,
"num_tokens": 2172768438.0,
"step": 16690
},
{
"entropy": 1.719921875,
"epoch": 0.868163859430235,
"grad_norm": 2.1897321387224116,
"learning_rate": 1.3597384499946406e-06,
"loss": 0.8623,
"mean_token_accuracy": 0.7896959483623505,
"num_tokens": 2174071833.0,
"step": 16700
},
{
"epoch": 0.868163859430235,
"eval_entropy": 1.6908203125,
"eval_loss": 0.9495359063148499,
"eval_mean_token_accuracy": 0.7706428378820419,
"eval_num_tokens": 2174071833.0,
"eval_runtime": 26.9575,
"eval_samples_per_second": 94.0,
"eval_steps_per_second": 1.484,
"step": 16700
},
{
"entropy": 1.711328125,
"epoch": 0.8686837180286962,
"grad_norm": 2.641187001225974,
"learning_rate": 1.3543788187372712e-06,
"loss": 0.934,
"mean_token_accuracy": 0.7752001971006394,
"num_tokens": 2175374594.0,
"step": 16710
},
{
"entropy": 1.7109375,
"epoch": 0.8692035766271574,
"grad_norm": 2.862857483989528,
"learning_rate": 1.3490191874799017e-06,
"loss": 0.9412,
"mean_token_accuracy": 0.7744590282440186,
"num_tokens": 2176676845.0,
"step": 16720
},
{
"entropy": 1.74765625,
"epoch": 0.8697234352256187,
"grad_norm": 3.006461348418482,
"learning_rate": 1.343659556222532e-06,
"loss": 0.95,
"mean_token_accuracy": 0.7718720018863678,
"num_tokens": 2177978564.0,
"step": 16730
},
{
"entropy": 1.719140625,
"epoch": 0.8702432938240798,
"grad_norm": 2.42093116763303,
"learning_rate": 1.3382999249651626e-06,
"loss": 0.9095,
"mean_token_accuracy": 0.7811307013034821,
"num_tokens": 2179280010.0,
"step": 16740
},
{
"entropy": 1.72734375,
"epoch": 0.8707631524225411,
"grad_norm": 2.6724560053591735,
"learning_rate": 1.332940293707793e-06,
"loss": 0.9369,
"mean_token_accuracy": 0.77666075527668,
"num_tokens": 2180579632.0,
"step": 16750
},
{
"entropy": 1.72421875,
"epoch": 0.8712830110210023,
"grad_norm": 2.640021069950666,
"learning_rate": 1.3275806624504235e-06,
"loss": 0.9499,
"mean_token_accuracy": 0.7743670821189881,
"num_tokens": 2181881510.0,
"step": 16760
},
{
"entropy": 1.701953125,
"epoch": 0.8718028696194635,
"grad_norm": 2.788284205610377,
"learning_rate": 1.3222210311930538e-06,
"loss": 0.9729,
"mean_token_accuracy": 0.7686333626508712,
"num_tokens": 2183183854.0,
"step": 16770
},
{
"entropy": 1.7171875,
"epoch": 0.8723227282179247,
"grad_norm": 3.4213010975163276,
"learning_rate": 1.3168613999356844e-06,
"loss": 0.909,
"mean_token_accuracy": 0.7834930688142776,
"num_tokens": 2184484216.0,
"step": 16780
},
{
"entropy": 1.702734375,
"epoch": 0.8728425868163859,
"grad_norm": 2.434562882531212,
"learning_rate": 1.311501768678315e-06,
"loss": 0.9156,
"mean_token_accuracy": 0.779088693857193,
"num_tokens": 2185786415.0,
"step": 16790
},
{
"entropy": 1.704296875,
"epoch": 0.8733624454148472,
"grad_norm": 2.9600234264061873,
"learning_rate": 1.3061421374209455e-06,
"loss": 0.7995,
"mean_token_accuracy": 0.8053589582443237,
"num_tokens": 2187088820.0,
"step": 16800
},
{
"epoch": 0.8733624454148472,
"eval_entropy": 1.6908203125,
"eval_loss": 0.9489900469779968,
"eval_mean_token_accuracy": 0.7707233056426048,
"eval_num_tokens": 2187088820.0,
"eval_runtime": 27.0578,
"eval_samples_per_second": 93.651,
"eval_steps_per_second": 1.478,
"step": 16800
},
{
"entropy": 1.72890625,
"epoch": 0.8738823040133084,
"grad_norm": 2.5469650885573913,
"learning_rate": 1.300782506163576e-06,
"loss": 0.9529,
"mean_token_accuracy": 0.7715460687875748,
"num_tokens": 2188388899.0,
"step": 16810
},
{
"entropy": 1.719140625,
"epoch": 0.8744021626117696,
"grad_norm": 2.5655028356434904,
"learning_rate": 1.2954228749062063e-06,
"loss": 0.9013,
"mean_token_accuracy": 0.7854694962501526,
"num_tokens": 2189688617.0,
"step": 16820
},
{
"entropy": 1.714453125,
"epoch": 0.8749220212102308,
"grad_norm": 2.496794303022197,
"learning_rate": 1.2900632436488369e-06,
"loss": 0.9577,
"mean_token_accuracy": 0.7703644394874573,
"num_tokens": 2190990179.0,
"step": 16830
},
{
"entropy": 1.7296875,
"epoch": 0.875441879808692,
"grad_norm": 2.428340477097752,
"learning_rate": 1.2847036123914674e-06,
"loss": 0.9498,
"mean_token_accuracy": 0.7711494535207748,
"num_tokens": 2192289493.0,
"step": 16840
},
{
"entropy": 1.748046875,
"epoch": 0.8759617384071533,
"grad_norm": 2.8878097776098777,
"learning_rate": 1.279343981134098e-06,
"loss": 0.933,
"mean_token_accuracy": 0.7762864798307418,
"num_tokens": 2193591120.0,
"step": 16850
},
{
"entropy": 1.71875,
"epoch": 0.8764815970056145,
"grad_norm": 2.4098241680027708,
"learning_rate": 1.2739843498767285e-06,
"loss": 0.9106,
"mean_token_accuracy": 0.7838896721601486,
"num_tokens": 2194891080.0,
"step": 16860
},
{
"entropy": 1.70078125,
"epoch": 0.8770014556040757,
"grad_norm": 2.4795243203319393,
"learning_rate": 1.268624718619359e-06,
"loss": 0.899,
"mean_token_accuracy": 0.7853422135114669,
"num_tokens": 2196193847.0,
"step": 16870
},
{
"entropy": 1.72421875,
"epoch": 0.8775213142025369,
"grad_norm": 2.509410635088686,
"learning_rate": 1.2632650873619894e-06,
"loss": 0.9143,
"mean_token_accuracy": 0.7773025572299957,
"num_tokens": 2197498331.0,
"step": 16880
},
{
"entropy": 1.712890625,
"epoch": 0.8780411728009981,
"grad_norm": 2.589267926384853,
"learning_rate": 1.25790545610462e-06,
"loss": 0.9117,
"mean_token_accuracy": 0.7811739712953567,
"num_tokens": 2198801318.0,
"step": 16890
},
{
"entropy": 1.71640625,
"epoch": 0.8785610313994594,
"grad_norm": 2.7192744762482013,
"learning_rate": 1.2525458248472505e-06,
"loss": 0.8801,
"mean_token_accuracy": 0.7873202979564666,
"num_tokens": 2200101966.0,
"step": 16900
},
{
"epoch": 0.8785610313994594,
"eval_entropy": 1.70390625,
"eval_loss": 0.9472702741622925,
"eval_mean_token_accuracy": 0.7711299672722817,
"eval_num_tokens": 2200101966.0,
"eval_runtime": 27.0596,
"eval_samples_per_second": 93.645,
"eval_steps_per_second": 1.478,
"step": 16900
},
{
"entropy": 1.73203125,
"epoch": 0.8790808899979206,
"grad_norm": 3.233207152990379,
"learning_rate": 1.2471861935898813e-06,
"loss": 0.9282,
"mean_token_accuracy": 0.7763419300317764,
"num_tokens": 2201405338.0,
"step": 16910
},
{
"entropy": 1.719140625,
"epoch": 0.8796007485963818,
"grad_norm": 2.4977890214739698,
"learning_rate": 1.2418265623325116e-06,
"loss": 0.9266,
"mean_token_accuracy": 0.775205773115158,
"num_tokens": 2202707630.0,
"step": 16920
},
{
"entropy": 1.737109375,
"epoch": 0.880120607194843,
"grad_norm": 2.310963182558919,
"learning_rate": 1.2364669310751421e-06,
"loss": 0.9395,
"mean_token_accuracy": 0.7774498581886291,
"num_tokens": 2204007677.0,
"step": 16930
},
{
"entropy": 1.705859375,
"epoch": 0.8806404657933042,
"grad_norm": 3.863632486798441,
"learning_rate": 1.2311072998177727e-06,
"loss": 0.986,
"mean_token_accuracy": 0.7657640516757965,
"num_tokens": 2205310191.0,
"step": 16940
},
{
"entropy": 1.716796875,
"epoch": 0.8811603243917654,
"grad_norm": 2.88662980668808,
"learning_rate": 1.2257476685604032e-06,
"loss": 0.9581,
"mean_token_accuracy": 0.7702996164560318,
"num_tokens": 2206613310.0,
"step": 16950
},
{
"entropy": 1.7265625,
"epoch": 0.8816801829902267,
"grad_norm": 3.039370137190877,
"learning_rate": 1.2203880373030338e-06,
"loss": 0.9029,
"mean_token_accuracy": 0.7842733591794968,
"num_tokens": 2207914663.0,
"step": 16960
},
{
"entropy": 1.723046875,
"epoch": 0.8822000415886879,
"grad_norm": 2.4536212604897236,
"learning_rate": 1.2150284060456641e-06,
"loss": 0.9807,
"mean_token_accuracy": 0.7632170081138611,
"num_tokens": 2209214465.0,
"step": 16970
},
{
"entropy": 1.74765625,
"epoch": 0.8827199001871491,
"grad_norm": 2.4854719349517165,
"learning_rate": 1.2096687747882947e-06,
"loss": 0.955,
"mean_token_accuracy": 0.771897703409195,
"num_tokens": 2210516988.0,
"step": 16980
},
{
"entropy": 1.712890625,
"epoch": 0.8832397587856103,
"grad_norm": 2.637606640299538,
"learning_rate": 1.2043091435309252e-06,
"loss": 0.8767,
"mean_token_accuracy": 0.7875324815511704,
"num_tokens": 2211818673.0,
"step": 16990
},
{
"entropy": 1.735546875,
"epoch": 0.8837596173840715,
"grad_norm": 3.1702137063722313,
"learning_rate": 1.1989495122735558e-06,
"loss": 0.9094,
"mean_token_accuracy": 0.7821446478366851,
"num_tokens": 2213118910.0,
"step": 17000
},
{
"epoch": 0.8837596173840715,
"eval_entropy": 1.7294921875,
"eval_loss": 0.9464492797851562,
"eval_mean_token_accuracy": 0.7712777078151702,
"eval_num_tokens": 2213118910.0,
"eval_runtime": 26.9615,
"eval_samples_per_second": 93.986,
"eval_steps_per_second": 1.484,
"step": 17000
},
{
"entropy": 1.735546875,
"epoch": 0.8842794759825328,
"grad_norm": 2.48825558882404,
"learning_rate": 1.1935898810161863e-06,
"loss": 0.9096,
"mean_token_accuracy": 0.7824355334043502,
"num_tokens": 2214420689.0,
"step": 17010
},
{
"entropy": 1.697265625,
"epoch": 0.884799334580994,
"grad_norm": 2.5259368751917926,
"learning_rate": 1.1882302497588166e-06,
"loss": 0.873,
"mean_token_accuracy": 0.7920572012662888,
"num_tokens": 2215721892.0,
"step": 17020
},
{
"entropy": 1.727734375,
"epoch": 0.8853191931794552,
"grad_norm": 2.3477327892257494,
"learning_rate": 1.1828706185014472e-06,
"loss": 0.924,
"mean_token_accuracy": 0.7783131778240204,
"num_tokens": 2217021672.0,
"step": 17030
},
{
"entropy": 1.73046875,
"epoch": 0.8858390517779164,
"grad_norm": 2.9327329808598135,
"learning_rate": 1.1775109872440777e-06,
"loss": 0.9404,
"mean_token_accuracy": 0.7744867652654648,
"num_tokens": 2218322000.0,
"step": 17040
},
{
"entropy": 1.716015625,
"epoch": 0.8863589103763776,
"grad_norm": 2.495718968036555,
"learning_rate": 1.172151355986708e-06,
"loss": 0.905,
"mean_token_accuracy": 0.7838757127523422,
"num_tokens": 2219622027.0,
"step": 17050
},
{
"entropy": 1.734765625,
"epoch": 0.8868787689748389,
"grad_norm": 2.786417709196235,
"learning_rate": 1.1667917247293386e-06,
"loss": 0.9257,
"mean_token_accuracy": 0.7765525490045547,
"num_tokens": 2220924892.0,
"step": 17060
},
{
"entropy": 1.69140625,
"epoch": 0.8873986275733,
"grad_norm": 3.325873719413414,
"learning_rate": 1.1614320934719692e-06,
"loss": 0.8905,
"mean_token_accuracy": 0.78846774995327,
"num_tokens": 2222228433.0,
"step": 17070
},
{
"entropy": 1.71796875,
"epoch": 0.8879184861717613,
"grad_norm": 2.4696074763325204,
"learning_rate": 1.1560724622145997e-06,
"loss": 0.9289,
"mean_token_accuracy": 0.7742702394723893,
"num_tokens": 2223531652.0,
"step": 17080
},
{
"entropy": 1.70625,
"epoch": 0.8884383447702225,
"grad_norm": 2.9281372223183357,
"learning_rate": 1.1507128309572302e-06,
"loss": 0.8388,
"mean_token_accuracy": 0.797678542137146,
"num_tokens": 2224835045.0,
"step": 17090
},
{
"entropy": 1.71640625,
"epoch": 0.8889582033686837,
"grad_norm": 2.596531284641399,
"learning_rate": 1.1453531996998608e-06,
"loss": 0.8564,
"mean_token_accuracy": 0.7959716618061066,
"num_tokens": 2226135591.0,
"step": 17100
},
{
"epoch": 0.8889582033686837,
"eval_entropy": 1.707421875,
"eval_loss": 0.9461068511009216,
"eval_mean_token_accuracy": 0.7713882640004158,
"eval_num_tokens": 2226135591.0,
"eval_runtime": 26.9578,
"eval_samples_per_second": 93.999,
"eval_steps_per_second": 1.484,
"step": 17100
},
{
"entropy": 1.719140625,
"epoch": 0.889478061967145,
"grad_norm": 2.409531070695101,
"learning_rate": 1.1399935684424911e-06,
"loss": 0.9422,
"mean_token_accuracy": 0.7755474865436554,
"num_tokens": 2227438593.0,
"step": 17110
},
{
"entropy": 1.73984375,
"epoch": 0.8899979205656061,
"grad_norm": 2.7010645985671964,
"learning_rate": 1.1346339371851217e-06,
"loss": 0.9278,
"mean_token_accuracy": 0.7782766997814179,
"num_tokens": 2228740793.0,
"step": 17120
},
{
"entropy": 1.7265625,
"epoch": 0.8905177791640674,
"grad_norm": 2.353429592598051,
"learning_rate": 1.1292743059277522e-06,
"loss": 0.869,
"mean_token_accuracy": 0.7913944810628891,
"num_tokens": 2230043684.0,
"step": 17130
},
{
"entropy": 1.698046875,
"epoch": 0.8910376377625286,
"grad_norm": 2.5271458723700397,
"learning_rate": 1.1239146746703828e-06,
"loss": 0.9052,
"mean_token_accuracy": 0.7812930643558502,
"num_tokens": 2231345260.0,
"step": 17140
},
{
"entropy": 1.733203125,
"epoch": 0.8915574963609898,
"grad_norm": 2.5710241189741847,
"learning_rate": 1.1185550434130133e-06,
"loss": 0.9225,
"mean_token_accuracy": 0.7815763622522354,
"num_tokens": 2232647792.0,
"step": 17150
},
{
"entropy": 1.716796875,
"epoch": 0.8920773549594511,
"grad_norm": 2.742326532049174,
"learning_rate": 1.1131954121556436e-06,
"loss": 0.8671,
"mean_token_accuracy": 0.791019856929779,
"num_tokens": 2233949412.0,
"step": 17160
},
{
"entropy": 1.719921875,
"epoch": 0.8925972135579122,
"grad_norm": 3.0166986549964103,
"learning_rate": 1.1078357808982742e-06,
"loss": 0.9877,
"mean_token_accuracy": 0.7625998437404633,
"num_tokens": 2235250251.0,
"step": 17170
},
{
"entropy": 1.70703125,
"epoch": 0.8931170721563735,
"grad_norm": 2.735187882521676,
"learning_rate": 1.1024761496409047e-06,
"loss": 0.8767,
"mean_token_accuracy": 0.7899385988712311,
"num_tokens": 2236552619.0,
"step": 17180
},
{
"entropy": 1.70390625,
"epoch": 0.8936369307548346,
"grad_norm": 3.1300638540188683,
"learning_rate": 1.0971165183835353e-06,
"loss": 0.9431,
"mean_token_accuracy": 0.773919427394867,
"num_tokens": 2237853090.0,
"step": 17190
},
{
"entropy": 1.719140625,
"epoch": 0.8941567893532959,
"grad_norm": 2.8739853375997515,
"learning_rate": 1.0917568871261658e-06,
"loss": 0.9027,
"mean_token_accuracy": 0.7813316076993942,
"num_tokens": 2239156492.0,
"step": 17200
},
{
"epoch": 0.8941567893532959,
"eval_entropy": 1.7041015625,
"eval_loss": 0.9447429180145264,
"eval_mean_token_accuracy": 0.7716016322374344,
"eval_num_tokens": 2239156492.0,
"eval_runtime": 26.9372,
"eval_samples_per_second": 94.071,
"eval_steps_per_second": 1.485,
"step": 17200
},
{
"entropy": 1.70703125,
"epoch": 0.8946766479517572,
"grad_norm": 2.8070816656417175,
"learning_rate": 1.0863972558687964e-06,
"loss": 0.9279,
"mean_token_accuracy": 0.7804146021604538,
"num_tokens": 2240460567.0,
"step": 17210
},
{
"entropy": 1.68046875,
"epoch": 0.8951965065502183,
"grad_norm": 2.7773860181227596,
"learning_rate": 1.0810376246114267e-06,
"loss": 0.8802,
"mean_token_accuracy": 0.7850437968969345,
"num_tokens": 2241760843.0,
"step": 17220
},
{
"entropy": 1.67265625,
"epoch": 0.8957163651486796,
"grad_norm": 2.565265564684608,
"learning_rate": 1.0756779933540573e-06,
"loss": 0.8588,
"mean_token_accuracy": 0.7939569532871247,
"num_tokens": 2243062576.0,
"step": 17230
},
{
"entropy": 1.71484375,
"epoch": 0.8962362237471407,
"grad_norm": 3.418704730557707,
"learning_rate": 1.0703183620966878e-06,
"loss": 0.8856,
"mean_token_accuracy": 0.7870086461305619,
"num_tokens": 2244364974.0,
"step": 17240
},
{
"entropy": 1.727734375,
"epoch": 0.896756082345602,
"grad_norm": 2.3722214861270206,
"learning_rate": 1.0649587308393184e-06,
"loss": 0.9653,
"mean_token_accuracy": 0.766468209028244,
"num_tokens": 2245665028.0,
"step": 17250
},
{
"entropy": 1.721875,
"epoch": 0.8972759409440633,
"grad_norm": 2.9408186963725846,
"learning_rate": 1.059599099581949e-06,
"loss": 0.8668,
"mean_token_accuracy": 0.7921581953763962,
"num_tokens": 2246968819.0,
"step": 17260
},
{
"entropy": 1.70625,
"epoch": 0.8977957995425244,
"grad_norm": 3.2040602083756267,
"learning_rate": 1.0542394683245795e-06,
"loss": 0.9446,
"mean_token_accuracy": 0.7708188474178315,
"num_tokens": 2248272251.0,
"step": 17270
},
{
"entropy": 1.73359375,
"epoch": 0.8983156581409857,
"grad_norm": 3.0617235039590884,
"learning_rate": 1.0488798370672098e-06,
"loss": 0.8876,
"mean_token_accuracy": 0.7847558498382569,
"num_tokens": 2249574035.0,
"step": 17280
},
{
"entropy": 1.72421875,
"epoch": 0.8988355167394468,
"grad_norm": 3.088148818484128,
"learning_rate": 1.0435202058098403e-06,
"loss": 0.883,
"mean_token_accuracy": 0.7865015596151352,
"num_tokens": 2250876342.0,
"step": 17290
},
{
"entropy": 1.688671875,
"epoch": 0.8993553753379081,
"grad_norm": 2.2283919237418557,
"learning_rate": 1.0381605745524709e-06,
"loss": 0.8435,
"mean_token_accuracy": 0.7962318360805511,
"num_tokens": 2252179761.0,
"step": 17300
},
{
"epoch": 0.8993553753379081,
"eval_entropy": 1.7033203125,
"eval_loss": 0.943964421749115,
"eval_mean_token_accuracy": 0.7720164701342582,
"eval_num_tokens": 2252179761.0,
"eval_runtime": 27.0626,
"eval_samples_per_second": 93.635,
"eval_steps_per_second": 1.478,
"step": 17300
},
{
"entropy": 1.719921875,
"epoch": 0.8998752339363693,
"grad_norm": 2.195845673845418,
"learning_rate": 1.0328009432951014e-06,
"loss": 0.9377,
"mean_token_accuracy": 0.7737020134925843,
"num_tokens": 2253482042.0,
"step": 17310
},
{
"entropy": 1.709765625,
"epoch": 0.9003950925348305,
"grad_norm": 2.6519702062041346,
"learning_rate": 1.027441312037732e-06,
"loss": 0.8687,
"mean_token_accuracy": 0.7917429208755493,
"num_tokens": 2254783467.0,
"step": 17320
},
{
"entropy": 1.721484375,
"epoch": 0.9009149511332918,
"grad_norm": 2.5770620159479196,
"learning_rate": 1.0220816807803623e-06,
"loss": 0.8223,
"mean_token_accuracy": 0.8054763108491898,
"num_tokens": 2256086384.0,
"step": 17330
},
{
"entropy": 1.72578125,
"epoch": 0.9014348097317529,
"grad_norm": 2.255695339103547,
"learning_rate": 1.0167220495229929e-06,
"loss": 0.9249,
"mean_token_accuracy": 0.7785108149051666,
"num_tokens": 2257389568.0,
"step": 17340
},
{
"entropy": 1.69609375,
"epoch": 0.9019546683302142,
"grad_norm": 2.4451266079286103,
"learning_rate": 1.0113624182656234e-06,
"loss": 0.9097,
"mean_token_accuracy": 0.7798655241727829,
"num_tokens": 2258693128.0,
"step": 17350
},
{
"entropy": 1.716796875,
"epoch": 0.9024745269286754,
"grad_norm": 3.009176073357241,
"learning_rate": 1.006002787008254e-06,
"loss": 0.9421,
"mean_token_accuracy": 0.7752411454916001,
"num_tokens": 2259993794.0,
"step": 17360
},
{
"entropy": 1.70859375,
"epoch": 0.9029943855271366,
"grad_norm": 2.827615611273616,
"learning_rate": 1.0006431557508845e-06,
"loss": 0.8804,
"mean_token_accuracy": 0.7873235315084457,
"num_tokens": 2261296385.0,
"step": 17370
},
{
"entropy": 1.71328125,
"epoch": 0.9035142441255979,
"grad_norm": 2.7699600891265823,
"learning_rate": 9.95283524493515e-07,
"loss": 0.8726,
"mean_token_accuracy": 0.7903792887926102,
"num_tokens": 2262597793.0,
"step": 17380
},
{
"entropy": 1.698828125,
"epoch": 0.904034102724059,
"grad_norm": 2.599440209991957,
"learning_rate": 9.899238932361454e-07,
"loss": 0.9141,
"mean_token_accuracy": 0.7799760162830353,
"num_tokens": 2263898666.0,
"step": 17390
},
{
"entropy": 1.725390625,
"epoch": 0.9045539613225203,
"grad_norm": 2.7458657612720225,
"learning_rate": 9.84564261978776e-07,
"loss": 0.9131,
"mean_token_accuracy": 0.7794664114713669,
"num_tokens": 2265203412.0,
"step": 17400
},
{
"epoch": 0.9045539613225203,
"eval_entropy": 1.7033203125,
"eval_loss": 0.9433298707008362,
"eval_mean_token_accuracy": 0.772235244512558,
"eval_num_tokens": 2265203412.0,
"eval_runtime": 27.0302,
"eval_samples_per_second": 93.747,
"eval_steps_per_second": 1.48,
"step": 17400
},
{
"entropy": 1.713671875,
"epoch": 0.9050738199209815,
"grad_norm": 2.5833045268240156,
"learning_rate": 9.792046307214065e-07,
"loss": 0.8921,
"mean_token_accuracy": 0.7821447908878326,
"num_tokens": 2266504068.0,
"step": 17410
},
{
"entropy": 1.707421875,
"epoch": 0.9055936785194427,
"grad_norm": 3.4641337111011086,
"learning_rate": 9.73844999464037e-07,
"loss": 0.9242,
"mean_token_accuracy": 0.7764649510383606,
"num_tokens": 2267805879.0,
"step": 17420
},
{
"entropy": 1.698828125,
"epoch": 0.9061135371179039,
"grad_norm": 3.211371934641073,
"learning_rate": 9.684853682066676e-07,
"loss": 0.928,
"mean_token_accuracy": 0.7754935950040818,
"num_tokens": 2269108912.0,
"step": 17430
},
{
"entropy": 1.72890625,
"epoch": 0.9066333957163651,
"grad_norm": 2.6497703964537678,
"learning_rate": 9.631257369492981e-07,
"loss": 0.9091,
"mean_token_accuracy": 0.7843220621347428,
"num_tokens": 2270408619.0,
"step": 17440
},
{
"entropy": 1.73125,
"epoch": 0.9071532543148264,
"grad_norm": 2.979073879941843,
"learning_rate": 9.577661056919284e-07,
"loss": 0.928,
"mean_token_accuracy": 0.7779957234859467,
"num_tokens": 2271710111.0,
"step": 17450
},
{
"entropy": 1.694921875,
"epoch": 0.9076731129132876,
"grad_norm": 2.539324981092276,
"learning_rate": 9.52406474434559e-07,
"loss": 0.8504,
"mean_token_accuracy": 0.7935720205307006,
"num_tokens": 2273010972.0,
"step": 17460
},
{
"entropy": 1.72421875,
"epoch": 0.9081929715117488,
"grad_norm": 3.585801630986423,
"learning_rate": 9.470468431771895e-07,
"loss": 0.936,
"mean_token_accuracy": 0.7747984647750854,
"num_tokens": 2274315012.0,
"step": 17470
},
{
"entropy": 1.728125,
"epoch": 0.90871283011021,
"grad_norm": 2.6261699472273468,
"learning_rate": 9.416872119198201e-07,
"loss": 0.9103,
"mean_token_accuracy": 0.7797531455755233,
"num_tokens": 2275617623.0,
"step": 17480
},
{
"entropy": 1.69921875,
"epoch": 0.9092326887086712,
"grad_norm": 2.7144327038410494,
"learning_rate": 9.363275806624505e-07,
"loss": 0.8875,
"mean_token_accuracy": 0.7863741964101791,
"num_tokens": 2276919349.0,
"step": 17490
},
{
"entropy": 1.719140625,
"epoch": 0.9097525473071325,
"grad_norm": 2.479986252492752,
"learning_rate": 9.309679494050811e-07,
"loss": 0.8823,
"mean_token_accuracy": 0.7868152827024459,
"num_tokens": 2278219351.0,
"step": 17500
},
{
"epoch": 0.9097525473071325,
"eval_entropy": 1.702734375,
"eval_loss": 0.9430034756660461,
"eval_mean_token_accuracy": 0.7721793368458748,
"eval_num_tokens": 2278219351.0,
"eval_runtime": 27.0303,
"eval_samples_per_second": 93.747,
"eval_steps_per_second": 1.48,
"step": 17500
},
{
"entropy": 1.72578125,
"epoch": 0.9102724059055937,
"grad_norm": 2.603788092143054,
"learning_rate": 9.256083181477116e-07,
"loss": 0.9321,
"mean_token_accuracy": 0.7802274167537689,
"num_tokens": 2279522977.0,
"step": 17510
},
{
"entropy": 1.69140625,
"epoch": 0.9107922645040549,
"grad_norm": 2.602338010673515,
"learning_rate": 9.202486868903421e-07,
"loss": 0.8839,
"mean_token_accuracy": 0.7898648619651795,
"num_tokens": 2280827133.0,
"step": 17520
},
{
"entropy": 1.732421875,
"epoch": 0.9113121231025161,
"grad_norm": 2.7366273353721473,
"learning_rate": 9.148890556329726e-07,
"loss": 0.9791,
"mean_token_accuracy": 0.7656744956970215,
"num_tokens": 2282130553.0,
"step": 17530
},
{
"entropy": 1.73671875,
"epoch": 0.9118319817009773,
"grad_norm": 2.516661290210401,
"learning_rate": 9.095294243756031e-07,
"loss": 0.9074,
"mean_token_accuracy": 0.7813127785921097,
"num_tokens": 2283432868.0,
"step": 17540
},
{
"entropy": 1.715625,
"epoch": 0.9123518402994385,
"grad_norm": 2.860208777660273,
"learning_rate": 9.041697931182336e-07,
"loss": 0.9775,
"mean_token_accuracy": 0.764993017911911,
"num_tokens": 2284733347.0,
"step": 17550
},
{
"entropy": 1.715625,
"epoch": 0.9128716988978998,
"grad_norm": 3.1125329261975536,
"learning_rate": 8.98810161860864e-07,
"loss": 0.9352,
"mean_token_accuracy": 0.7755971252918243,
"num_tokens": 2286037087.0,
"step": 17560
},
{
"entropy": 1.70625,
"epoch": 0.913391557496361,
"grad_norm": 2.6657122874382337,
"learning_rate": 8.934505306034945e-07,
"loss": 0.8867,
"mean_token_accuracy": 0.7875713557004929,
"num_tokens": 2287337885.0,
"step": 17570
},
{
"entropy": 1.71875,
"epoch": 0.9139114160948222,
"grad_norm": 3.4703854191078145,
"learning_rate": 8.88090899346125e-07,
"loss": 0.9579,
"mean_token_accuracy": 0.772272664308548,
"num_tokens": 2288639501.0,
"step": 17580
},
{
"entropy": 1.726171875,
"epoch": 0.9144312746932834,
"grad_norm": 2.5435144737663777,
"learning_rate": 8.827312680887555e-07,
"loss": 0.9076,
"mean_token_accuracy": 0.7839017242193222,
"num_tokens": 2289943326.0,
"step": 17590
},
{
"entropy": 1.7125,
"epoch": 0.9149511332917446,
"grad_norm": 2.3996366113976704,
"learning_rate": 8.77371636831386e-07,
"loss": 0.9641,
"mean_token_accuracy": 0.7694620430469513,
"num_tokens": 2291244110.0,
"step": 17600
},
{
"epoch": 0.9149511332917446,
"eval_entropy": 1.6951171875,
"eval_loss": 0.9421440958976746,
"eval_mean_token_accuracy": 0.7724334865808486,
"eval_num_tokens": 2291244110.0,
"eval_runtime": 26.9828,
"eval_samples_per_second": 93.912,
"eval_steps_per_second": 1.482,
"step": 17600
},
{
"entropy": 1.695703125,
"epoch": 0.9154709918902059,
"grad_norm": 2.6078739851608828,
"learning_rate": 8.720120055740166e-07,
"loss": 0.8904,
"mean_token_accuracy": 0.7871272295713425,
"num_tokens": 2292544974.0,
"step": 17610
},
{
"entropy": 1.71953125,
"epoch": 0.9159908504886671,
"grad_norm": 2.3213763292784195,
"learning_rate": 8.66652374316647e-07,
"loss": 0.8995,
"mean_token_accuracy": 0.7820333391427994,
"num_tokens": 2293847279.0,
"step": 17620
},
{
"entropy": 1.72578125,
"epoch": 0.9165107090871283,
"grad_norm": 2.7780816827132377,
"learning_rate": 8.612927430592775e-07,
"loss": 0.9109,
"mean_token_accuracy": 0.7834825724363327,
"num_tokens": 2295149131.0,
"step": 17630
},
{
"entropy": 1.726171875,
"epoch": 0.9170305676855895,
"grad_norm": 2.7764002641826653,
"learning_rate": 8.559331118019081e-07,
"loss": 0.9329,
"mean_token_accuracy": 0.7784050792455673,
"num_tokens": 2296451177.0,
"step": 17640
},
{
"entropy": 1.721875,
"epoch": 0.9175504262840507,
"grad_norm": 2.387246120243253,
"learning_rate": 8.505734805445385e-07,
"loss": 0.9233,
"mean_token_accuracy": 0.7772431135177612,
"num_tokens": 2297751120.0,
"step": 17650
},
{
"entropy": 1.708984375,
"epoch": 0.918070284882512,
"grad_norm": 2.9368578355190165,
"learning_rate": 8.452138492871691e-07,
"loss": 0.9168,
"mean_token_accuracy": 0.7799305468797684,
"num_tokens": 2299053798.0,
"step": 17660
},
{
"entropy": 1.72265625,
"epoch": 0.9185901434809731,
"grad_norm": 2.491616034177155,
"learning_rate": 8.398542180297996e-07,
"loss": 0.9416,
"mean_token_accuracy": 0.7715413421392441,
"num_tokens": 2300355063.0,
"step": 17670
},
{
"entropy": 1.6875,
"epoch": 0.9191100020794344,
"grad_norm": 2.5438409946613634,
"learning_rate": 8.344945867724301e-07,
"loss": 0.8771,
"mean_token_accuracy": 0.7873465567827225,
"num_tokens": 2301658285.0,
"step": 17680
},
{
"entropy": 1.728125,
"epoch": 0.9196298606778956,
"grad_norm": 2.4860699998631888,
"learning_rate": 8.291349555150606e-07,
"loss": 0.8504,
"mean_token_accuracy": 0.7953920751810074,
"num_tokens": 2302956288.0,
"step": 17690
},
{
"entropy": 1.682421875,
"epoch": 0.9201497192763568,
"grad_norm": 2.6743670047456027,
"learning_rate": 8.237753242576912e-07,
"loss": 0.8732,
"mean_token_accuracy": 0.785806754231453,
"num_tokens": 2304255445.0,
"step": 17700
},
{
"epoch": 0.9201497192763568,
"eval_entropy": 1.6912109375,
"eval_loss": 0.9414154291152954,
"eval_mean_token_accuracy": 0.7726485714316368,
"eval_num_tokens": 2304255445.0,
"eval_runtime": 27.0888,
"eval_samples_per_second": 93.544,
"eval_steps_per_second": 1.477,
"step": 17700
},
{
"entropy": 1.720703125,
"epoch": 0.9206695778748181,
"grad_norm": 2.884106162989069,
"learning_rate": 8.184156930003216e-07,
"loss": 0.9383,
"mean_token_accuracy": 0.7747535407543182,
"num_tokens": 2305557395.0,
"step": 17710
},
{
"entropy": 1.731640625,
"epoch": 0.9211894364732792,
"grad_norm": 3.1110946669254016,
"learning_rate": 8.130560617429521e-07,
"loss": 0.9836,
"mean_token_accuracy": 0.7641501188278198,
"num_tokens": 2306859582.0,
"step": 17720
},
{
"entropy": 1.721484375,
"epoch": 0.9217092950717405,
"grad_norm": 2.810503323233352,
"learning_rate": 8.076964304855826e-07,
"loss": 0.8573,
"mean_token_accuracy": 0.7907409578561783,
"num_tokens": 2308162308.0,
"step": 17730
},
{
"entropy": 1.694140625,
"epoch": 0.9222291536702018,
"grad_norm": 2.7683044114348947,
"learning_rate": 8.023367992282131e-07,
"loss": 0.8695,
"mean_token_accuracy": 0.7887108832597732,
"num_tokens": 2309464898.0,
"step": 17740
},
{
"entropy": 1.718359375,
"epoch": 0.9227490122686629,
"grad_norm": 2.838498188453563,
"learning_rate": 7.969771679708437e-07,
"loss": 0.9356,
"mean_token_accuracy": 0.7756464421749115,
"num_tokens": 2310767874.0,
"step": 17750
},
{
"entropy": 1.7,
"epoch": 0.9232688708671242,
"grad_norm": 2.1756652207099934,
"learning_rate": 7.916175367134741e-07,
"loss": 0.8934,
"mean_token_accuracy": 0.7871183961629867,
"num_tokens": 2312069834.0,
"step": 17760
},
{
"entropy": 1.713671875,
"epoch": 0.9237887294655853,
"grad_norm": 2.7275522031678623,
"learning_rate": 7.862579054561047e-07,
"loss": 0.8736,
"mean_token_accuracy": 0.7900153934955597,
"num_tokens": 2313370451.0,
"step": 17770
},
{
"entropy": 1.726171875,
"epoch": 0.9243085880640466,
"grad_norm": 2.6194561539893617,
"learning_rate": 7.808982741987352e-07,
"loss": 0.8734,
"mean_token_accuracy": 0.7897523701190948,
"num_tokens": 2314673898.0,
"step": 17780
},
{
"entropy": 1.718359375,
"epoch": 0.9248284466625079,
"grad_norm": 2.310987836951584,
"learning_rate": 7.755386429413656e-07,
"loss": 0.9242,
"mean_token_accuracy": 0.7761760890483856,
"num_tokens": 2315974676.0,
"step": 17790
},
{
"entropy": 1.694921875,
"epoch": 0.925348305260969,
"grad_norm": 2.8505695856451214,
"learning_rate": 7.701790116839962e-07,
"loss": 0.9376,
"mean_token_accuracy": 0.7763515889644623,
"num_tokens": 2317276588.0,
"step": 17800
},
{
"epoch": 0.925348305260969,
"eval_entropy": 1.69375,
"eval_loss": 0.939235270023346,
"eval_mean_token_accuracy": 0.773291589319706,
"eval_num_tokens": 2317276588.0,
"eval_runtime": 27.0206,
"eval_samples_per_second": 93.78,
"eval_steps_per_second": 1.48,
"step": 17800
},
{
"entropy": 1.703125,
"epoch": 0.9258681638594303,
"grad_norm": 2.8152785892827517,
"learning_rate": 7.648193804266267e-07,
"loss": 0.9853,
"mean_token_accuracy": 0.7663014709949494,
"num_tokens": 2318580097.0,
"step": 17810
},
{
"entropy": 1.71796875,
"epoch": 0.9263880224578914,
"grad_norm": 3.2243675761474986,
"learning_rate": 7.594597491692572e-07,
"loss": 0.9676,
"mean_token_accuracy": 0.765509432554245,
"num_tokens": 2319880100.0,
"step": 17820
},
{
"entropy": 1.730078125,
"epoch": 0.9269078810563527,
"grad_norm": 2.2705076544496854,
"learning_rate": 7.541001179118877e-07,
"loss": 0.9535,
"mean_token_accuracy": 0.7697697907686234,
"num_tokens": 2321182485.0,
"step": 17830
},
{
"entropy": 1.71015625,
"epoch": 0.9274277396548138,
"grad_norm": 2.6684443372520485,
"learning_rate": 7.487404866545183e-07,
"loss": 0.943,
"mean_token_accuracy": 0.7748803436756134,
"num_tokens": 2322485265.0,
"step": 17840
},
{
"entropy": 1.7265625,
"epoch": 0.9279475982532751,
"grad_norm": 2.801572538024007,
"learning_rate": 7.433808553971487e-07,
"loss": 0.895,
"mean_token_accuracy": 0.785537713766098,
"num_tokens": 2323786917.0,
"step": 17850
},
{
"entropy": 1.7125,
"epoch": 0.9284674568517364,
"grad_norm": 2.430738555487872,
"learning_rate": 7.380212241397793e-07,
"loss": 0.9068,
"mean_token_accuracy": 0.7807348072528839,
"num_tokens": 2325089854.0,
"step": 17860
},
{
"entropy": 1.71484375,
"epoch": 0.9289873154501975,
"grad_norm": 2.7740515804986106,
"learning_rate": 7.326615928824098e-07,
"loss": 0.9534,
"mean_token_accuracy": 0.7729498475790024,
"num_tokens": 2326392120.0,
"step": 17870
},
{
"entropy": 1.705078125,
"epoch": 0.9295071740486588,
"grad_norm": 2.6522415738913825,
"learning_rate": 7.273019616250403e-07,
"loss": 0.9041,
"mean_token_accuracy": 0.7818612903356552,
"num_tokens": 2327694955.0,
"step": 17880
},
{
"entropy": 1.6984375,
"epoch": 0.9300270326471199,
"grad_norm": 2.4697058932446123,
"learning_rate": 7.219423303676708e-07,
"loss": 0.9396,
"mean_token_accuracy": 0.7752180218696594,
"num_tokens": 2328996598.0,
"step": 17890
},
{
"entropy": 1.7109375,
"epoch": 0.9305468912455812,
"grad_norm": 3.7338157631214357,
"learning_rate": 7.165826991103012e-07,
"loss": 0.8259,
"mean_token_accuracy": 0.7991910308599472,
"num_tokens": 2330297960.0,
"step": 17900
},
{
"epoch": 0.9305468912455812,
"eval_entropy": 1.6876953125,
"eval_loss": 0.9386802315711975,
"eval_mean_token_accuracy": 0.7734104990959167,
"eval_num_tokens": 2330297960.0,
"eval_runtime": 27.0238,
"eval_samples_per_second": 93.769,
"eval_steps_per_second": 1.48,
"step": 17900
},
{
"entropy": 1.69453125,
"epoch": 0.9310667498440425,
"grad_norm": 2.9660945826118783,
"learning_rate": 7.112230678529318e-07,
"loss": 0.8982,
"mean_token_accuracy": 0.7834569126367569,
"num_tokens": 2331598003.0,
"step": 17910
},
{
"entropy": 1.735546875,
"epoch": 0.9315866084425036,
"grad_norm": 2.522985749542941,
"learning_rate": 7.058634365955623e-07,
"loss": 0.9374,
"mean_token_accuracy": 0.7744401007890701,
"num_tokens": 2332899394.0,
"step": 17920
},
{
"entropy": 1.727734375,
"epoch": 0.9321064670409649,
"grad_norm": 2.73947775945392,
"learning_rate": 7.005038053381928e-07,
"loss": 0.9268,
"mean_token_accuracy": 0.7759706169366837,
"num_tokens": 2334201986.0,
"step": 17930
},
{
"entropy": 1.70234375,
"epoch": 0.932626325639426,
"grad_norm": 2.979213443205554,
"learning_rate": 6.951441740808233e-07,
"loss": 0.9015,
"mean_token_accuracy": 0.7848176747560501,
"num_tokens": 2335503580.0,
"step": 17940
},
{
"entropy": 1.724609375,
"epoch": 0.9331461842378873,
"grad_norm": 2.3534775119433347,
"learning_rate": 6.897845428234539e-07,
"loss": 0.937,
"mean_token_accuracy": 0.7760796934366226,
"num_tokens": 2336804187.0,
"step": 17950
},
{
"entropy": 1.708203125,
"epoch": 0.9336660428363485,
"grad_norm": 3.1775385022248352,
"learning_rate": 6.844249115660843e-07,
"loss": 0.9573,
"mean_token_accuracy": 0.7716828197240829,
"num_tokens": 2338106866.0,
"step": 17960
},
{
"entropy": 1.7015625,
"epoch": 0.9341859014348097,
"grad_norm": 2.797098432368521,
"learning_rate": 6.790652803087149e-07,
"loss": 0.8785,
"mean_token_accuracy": 0.7885834217071533,
"num_tokens": 2339409626.0,
"step": 17970
},
{
"entropy": 1.727734375,
"epoch": 0.934705760033271,
"grad_norm": 2.249655899999818,
"learning_rate": 6.737056490513454e-07,
"loss": 0.9716,
"mean_token_accuracy": 0.7680929601192474,
"num_tokens": 2340711985.0,
"step": 17980
},
{
"entropy": 1.71953125,
"epoch": 0.9352256186317321,
"grad_norm": 2.464126513458493,
"learning_rate": 6.683460177939758e-07,
"loss": 0.8962,
"mean_token_accuracy": 0.7843928962945939,
"num_tokens": 2342013095.0,
"step": 17990
},
{
"entropy": 1.708984375,
"epoch": 0.9357454772301934,
"grad_norm": 2.5766147783205726,
"learning_rate": 6.629863865366064e-07,
"loss": 0.9086,
"mean_token_accuracy": 0.7816471189260483,
"num_tokens": 2343312891.0,
"step": 18000
},
{
"epoch": 0.9357454772301934,
"eval_entropy": 1.6990234375,
"eval_loss": 0.9378655552864075,
"eval_mean_token_accuracy": 0.7736992001533508,
"eval_num_tokens": 2343312891.0,
"eval_runtime": 26.9847,
"eval_samples_per_second": 93.905,
"eval_steps_per_second": 1.482,
"step": 18000
},
{
"entropy": 1.71328125,
"epoch": 0.9362653358286546,
"grad_norm": 3.8173545588384394,
"learning_rate": 6.576267552792369e-07,
"loss": 0.9019,
"mean_token_accuracy": 0.7821660578250885,
"num_tokens": 2344613716.0,
"step": 18010
},
{
"entropy": 1.71875,
"epoch": 0.9367851944271158,
"grad_norm": 2.559403768497008,
"learning_rate": 6.522671240218674e-07,
"loss": 0.9071,
"mean_token_accuracy": 0.7829224199056626,
"num_tokens": 2345916095.0,
"step": 18020
},
{
"entropy": 1.70703125,
"epoch": 0.9373050530255771,
"grad_norm": 2.832401092090321,
"learning_rate": 6.469074927644979e-07,
"loss": 0.9251,
"mean_token_accuracy": 0.7782054781913758,
"num_tokens": 2347219346.0,
"step": 18030
},
{
"entropy": 1.711328125,
"epoch": 0.9378249116240382,
"grad_norm": 2.958090045209552,
"learning_rate": 6.415478615071285e-07,
"loss": 0.9617,
"mean_token_accuracy": 0.7694919466972351,
"num_tokens": 2348522579.0,
"step": 18040
},
{
"entropy": 1.73671875,
"epoch": 0.9383447702224995,
"grad_norm": 2.8805336259483205,
"learning_rate": 6.361882302497589e-07,
"loss": 0.9953,
"mean_token_accuracy": 0.762589517235756,
"num_tokens": 2349821708.0,
"step": 18050
},
{
"entropy": 1.721875,
"epoch": 0.9388646288209607,
"grad_norm": 2.6178814051559898,
"learning_rate": 6.308285989923895e-07,
"loss": 0.8964,
"mean_token_accuracy": 0.7832231611013413,
"num_tokens": 2351124995.0,
"step": 18060
},
{
"entropy": 1.705859375,
"epoch": 0.9393844874194219,
"grad_norm": 3.1425557398660526,
"learning_rate": 6.254689677350199e-07,
"loss": 0.8988,
"mean_token_accuracy": 0.7814945936203003,
"num_tokens": 2352424010.0,
"step": 18070
},
{
"entropy": 1.70546875,
"epoch": 0.9399043460178831,
"grad_norm": 2.7373784653600124,
"learning_rate": 6.201093364776503e-07,
"loss": 0.9468,
"mean_token_accuracy": 0.7743207067251205,
"num_tokens": 2353725536.0,
"step": 18080
},
{
"entropy": 1.70625,
"epoch": 0.9404242046163444,
"grad_norm": 2.673001314079472,
"learning_rate": 6.147497052202809e-07,
"loss": 0.9483,
"mean_token_accuracy": 0.7750268816947937,
"num_tokens": 2355023440.0,
"step": 18090
},
{
"entropy": 1.71875,
"epoch": 0.9409440632148056,
"grad_norm": 2.8978384228101004,
"learning_rate": 6.093900739629114e-07,
"loss": 0.9116,
"mean_token_accuracy": 0.7818041294813156,
"num_tokens": 2356326090.0,
"step": 18100
},
{
"epoch": 0.9409440632148056,
"eval_entropy": 1.6919921875,
"eval_loss": 0.9373577237129211,
"eval_mean_token_accuracy": 0.7737729415297508,
"eval_num_tokens": 2356326090.0,
"eval_runtime": 27.0154,
"eval_samples_per_second": 93.798,
"eval_steps_per_second": 1.481,
"step": 18100
},
{
"entropy": 1.71171875,
"epoch": 0.9414639218132668,
"grad_norm": 2.721990254495139,
"learning_rate": 6.040304427055419e-07,
"loss": 0.9342,
"mean_token_accuracy": 0.7786854475736618,
"num_tokens": 2357625659.0,
"step": 18110
},
{
"entropy": 1.742578125,
"epoch": 0.941983780411728,
"grad_norm": 2.5667956692743883,
"learning_rate": 5.986708114481724e-07,
"loss": 0.9389,
"mean_token_accuracy": 0.7736960768699646,
"num_tokens": 2358926558.0,
"step": 18120
},
{
"entropy": 1.7203125,
"epoch": 0.9425036390101892,
"grad_norm": 2.8957508079232506,
"learning_rate": 5.93311180190803e-07,
"loss": 0.8701,
"mean_token_accuracy": 0.7908826410770416,
"num_tokens": 2360227682.0,
"step": 18130
},
{
"entropy": 1.72734375,
"epoch": 0.9430234976086505,
"grad_norm": 2.304530266769058,
"learning_rate": 5.879515489334334e-07,
"loss": 0.9521,
"mean_token_accuracy": 0.7695625275373459,
"num_tokens": 2361529112.0,
"step": 18140
},
{
"entropy": 1.715625,
"epoch": 0.9435433562071117,
"grad_norm": 2.9012724704515884,
"learning_rate": 5.82591917676064e-07,
"loss": 0.8492,
"mean_token_accuracy": 0.7948765844106674,
"num_tokens": 2362830856.0,
"step": 18150
},
{
"entropy": 1.708203125,
"epoch": 0.9440632148055729,
"grad_norm": 2.972102163460491,
"learning_rate": 5.772322864186945e-07,
"loss": 0.95,
"mean_token_accuracy": 0.7732912987470627,
"num_tokens": 2364133037.0,
"step": 18160
},
{
"entropy": 1.697265625,
"epoch": 0.9445830734040341,
"grad_norm": 2.858435097480405,
"learning_rate": 5.718726551613249e-07,
"loss": 0.876,
"mean_token_accuracy": 0.791174441576004,
"num_tokens": 2365435608.0,
"step": 18170
},
{
"entropy": 1.70078125,
"epoch": 0.9451029320024953,
"grad_norm": 2.600838938576883,
"learning_rate": 5.665130239039555e-07,
"loss": 0.8875,
"mean_token_accuracy": 0.7903335362672805,
"num_tokens": 2366737431.0,
"step": 18180
},
{
"entropy": 1.71484375,
"epoch": 0.9456227906009566,
"grad_norm": 2.3340505816697545,
"learning_rate": 5.61153392646586e-07,
"loss": 0.9416,
"mean_token_accuracy": 0.7767238467931747,
"num_tokens": 2368040852.0,
"step": 18190
},
{
"entropy": 1.697265625,
"epoch": 0.9461426491994177,
"grad_norm": 2.5107466130045144,
"learning_rate": 5.557937613892165e-07,
"loss": 0.8652,
"mean_token_accuracy": 0.7920861512422561,
"num_tokens": 2369344309.0,
"step": 18200
},
{
"epoch": 0.9461426491994177,
"eval_entropy": 1.701953125,
"eval_loss": 0.9362885355949402,
"eval_mean_token_accuracy": 0.77409558147192,
"eval_num_tokens": 2369344309.0,
"eval_runtime": 26.988,
"eval_samples_per_second": 93.893,
"eval_steps_per_second": 1.482,
"step": 18200
},
{
"entropy": 1.709375,
"epoch": 0.946662507797879,
"grad_norm": 2.430227735472621,
"learning_rate": 5.50434130131847e-07,
"loss": 0.9559,
"mean_token_accuracy": 0.7686403065919876,
"num_tokens": 2370646514.0,
"step": 18210
},
{
"entropy": 1.70078125,
"epoch": 0.9471823663963402,
"grad_norm": 2.477476878549962,
"learning_rate": 5.450744988744776e-07,
"loss": 0.8956,
"mean_token_accuracy": 0.7845601290464401,
"num_tokens": 2371948550.0,
"step": 18220
},
{
"entropy": 1.704296875,
"epoch": 0.9477022249948014,
"grad_norm": 2.9836890732912815,
"learning_rate": 5.39714867617108e-07,
"loss": 0.8824,
"mean_token_accuracy": 0.7854431182146072,
"num_tokens": 2373251925.0,
"step": 18230
},
{
"entropy": 1.707421875,
"epoch": 0.9482220835932627,
"grad_norm": 2.9290745745798445,
"learning_rate": 5.343552363597384e-07,
"loss": 0.8905,
"mean_token_accuracy": 0.7864167779684067,
"num_tokens": 2374552792.0,
"step": 18240
},
{
"entropy": 1.6921875,
"epoch": 0.9487419421917238,
"grad_norm": 2.8899129654630538,
"learning_rate": 5.28995605102369e-07,
"loss": 0.9031,
"mean_token_accuracy": 0.7832641273736953,
"num_tokens": 2375854564.0,
"step": 18250
},
{
"entropy": 1.70625,
"epoch": 0.9492618007901851,
"grad_norm": 2.5753258280373066,
"learning_rate": 5.236359738449994e-07,
"loss": 0.8882,
"mean_token_accuracy": 0.7851995527744293,
"num_tokens": 2377156018.0,
"step": 18260
},
{
"entropy": 1.707421875,
"epoch": 0.9497816593886463,
"grad_norm": 2.7633311164171115,
"learning_rate": 5.1827634258763e-07,
"loss": 0.8863,
"mean_token_accuracy": 0.7868488073348999,
"num_tokens": 2378458173.0,
"step": 18270
},
{
"entropy": 1.71875,
"epoch": 0.9503015179871075,
"grad_norm": 2.453708918189624,
"learning_rate": 5.129167113302605e-07,
"loss": 0.8809,
"mean_token_accuracy": 0.7863835990428925,
"num_tokens": 2379759408.0,
"step": 18280
},
{
"entropy": 1.69453125,
"epoch": 0.9508213765855688,
"grad_norm": 2.494399172091332,
"learning_rate": 5.07557080072891e-07,
"loss": 0.8879,
"mean_token_accuracy": 0.7849009603261947,
"num_tokens": 2381058233.0,
"step": 18290
},
{
"entropy": 1.703515625,
"epoch": 0.9513412351840299,
"grad_norm": 2.532843877991443,
"learning_rate": 5.021974488155215e-07,
"loss": 0.9203,
"mean_token_accuracy": 0.7828258693218231,
"num_tokens": 2382358266.0,
"step": 18300
},
{
"epoch": 0.9513412351840299,
"eval_entropy": 1.684765625,
"eval_loss": 0.9346003532409668,
"eval_mean_token_accuracy": 0.7742012709379196,
"eval_num_tokens": 2382358266.0,
"eval_runtime": 27.0461,
"eval_samples_per_second": 93.692,
"eval_steps_per_second": 1.479,
"step": 18300
},
{
"entropy": 1.70234375,
"epoch": 0.9518610937824912,
"grad_norm": 2.6611283325072637,
"learning_rate": 4.968378175581521e-07,
"loss": 0.9038,
"mean_token_accuracy": 0.7826411038637161,
"num_tokens": 2383661718.0,
"step": 18310
},
{
"entropy": 1.70078125,
"epoch": 0.9523809523809523,
"grad_norm": 2.505914353694984,
"learning_rate": 4.914781863007825e-07,
"loss": 0.9028,
"mean_token_accuracy": 0.7839677035808563,
"num_tokens": 2384965109.0,
"step": 18320
},
{
"entropy": 1.720703125,
"epoch": 0.9529008109794136,
"grad_norm": 3.1119846898642725,
"learning_rate": 4.86118555043413e-07,
"loss": 0.9466,
"mean_token_accuracy": 0.7718929171562194,
"num_tokens": 2386263952.0,
"step": 18330
},
{
"entropy": 1.712890625,
"epoch": 0.9534206695778749,
"grad_norm": 2.590660553231583,
"learning_rate": 4.807589237860436e-07,
"loss": 0.8944,
"mean_token_accuracy": 0.787765109539032,
"num_tokens": 2387564504.0,
"step": 18340
},
{
"entropy": 1.687109375,
"epoch": 0.953940528176336,
"grad_norm": 2.882189015757662,
"learning_rate": 4.753992925286741e-07,
"loss": 0.8926,
"mean_token_accuracy": 0.786284226179123,
"num_tokens": 2388867172.0,
"step": 18350
},
{
"entropy": 1.687890625,
"epoch": 0.9544603867747973,
"grad_norm": 2.4899592829001036,
"learning_rate": 4.700396612713046e-07,
"loss": 0.9117,
"mean_token_accuracy": 0.7829731374979019,
"num_tokens": 2390170084.0,
"step": 18360
},
{
"entropy": 1.70390625,
"epoch": 0.9549802453732584,
"grad_norm": 3.6931538633843117,
"learning_rate": 4.646800300139351e-07,
"loss": 0.8938,
"mean_token_accuracy": 0.7853878140449524,
"num_tokens": 2391471812.0,
"step": 18370
},
{
"entropy": 1.70546875,
"epoch": 0.9555001039717197,
"grad_norm": 3.2893562205865985,
"learning_rate": 4.5932039875656557e-07,
"loss": 0.9738,
"mean_token_accuracy": 0.7703835904598236,
"num_tokens": 2392772427.0,
"step": 18380
},
{
"entropy": 1.71484375,
"epoch": 0.956019962570181,
"grad_norm": 2.3989918782412856,
"learning_rate": 4.539607674991961e-07,
"loss": 0.8742,
"mean_token_accuracy": 0.7906468451023102,
"num_tokens": 2394075243.0,
"step": 18390
},
{
"entropy": 1.703125,
"epoch": 0.9565398211686421,
"grad_norm": 2.9394021839914863,
"learning_rate": 4.486011362418266e-07,
"loss": 0.8622,
"mean_token_accuracy": 0.7922106266021729,
"num_tokens": 2395379290.0,
"step": 18400
},
{
"epoch": 0.9565398211686421,
"eval_entropy": 1.699609375,
"eval_loss": 0.9338641166687012,
"eval_mean_token_accuracy": 0.7741296887397766,
"eval_num_tokens": 2395379290.0,
"eval_runtime": 26.9993,
"eval_samples_per_second": 93.854,
"eval_steps_per_second": 1.482,
"step": 18400
},
{
"entropy": 1.704296875,
"epoch": 0.9570596797671034,
"grad_norm": 2.2413689325829256,
"learning_rate": 4.432415049844571e-07,
"loss": 0.9238,
"mean_token_accuracy": 0.7773810118436814,
"num_tokens": 2396682739.0,
"step": 18410
},
{
"entropy": 1.71640625,
"epoch": 0.9575795383655645,
"grad_norm": 2.4445483187953156,
"learning_rate": 4.3788187372708765e-07,
"loss": 0.898,
"mean_token_accuracy": 0.7847286909818649,
"num_tokens": 2397986735.0,
"step": 18420
},
{
"entropy": 1.733203125,
"epoch": 0.9580993969640258,
"grad_norm": 2.878321554703064,
"learning_rate": 4.3252224246971814e-07,
"loss": 0.971,
"mean_token_accuracy": 0.7672819674015046,
"num_tokens": 2399288636.0,
"step": 18430
},
{
"entropy": 1.708984375,
"epoch": 0.958619255562487,
"grad_norm": 2.4694757354185892,
"learning_rate": 4.2716261121234864e-07,
"loss": 0.9181,
"mean_token_accuracy": 0.7779204994440079,
"num_tokens": 2400591488.0,
"step": 18440
},
{
"entropy": 1.690625,
"epoch": 0.9591391141609482,
"grad_norm": 2.5714181093931727,
"learning_rate": 4.2180297995497913e-07,
"loss": 0.871,
"mean_token_accuracy": 0.7889610588550567,
"num_tokens": 2401895128.0,
"step": 18450
},
{
"entropy": 1.715625,
"epoch": 0.9596589727594095,
"grad_norm": 2.7074175632662856,
"learning_rate": 4.164433486976097e-07,
"loss": 0.96,
"mean_token_accuracy": 0.771069410443306,
"num_tokens": 2403198863.0,
"step": 18460
},
{
"entropy": 1.704296875,
"epoch": 0.9601788313578706,
"grad_norm": 2.6030193353868696,
"learning_rate": 4.1108371744024017e-07,
"loss": 0.8829,
"mean_token_accuracy": 0.7889763325452804,
"num_tokens": 2404499445.0,
"step": 18470
},
{
"entropy": 1.6890625,
"epoch": 0.9606986899563319,
"grad_norm": 2.534988347803682,
"learning_rate": 4.0572408618287066e-07,
"loss": 0.9687,
"mean_token_accuracy": 0.7657503515481949,
"num_tokens": 2405800615.0,
"step": 18480
},
{
"entropy": 1.6921875,
"epoch": 0.961218548554793,
"grad_norm": 2.143650818625468,
"learning_rate": 4.003644549255012e-07,
"loss": 0.8805,
"mean_token_accuracy": 0.7895686358213425,
"num_tokens": 2407103560.0,
"step": 18490
},
{
"entropy": 1.69296875,
"epoch": 0.9617384071532543,
"grad_norm": 2.5344926766447355,
"learning_rate": 3.950048236681317e-07,
"loss": 0.888,
"mean_token_accuracy": 0.7855937719345093,
"num_tokens": 2408404550.0,
"step": 18500
},
{
"epoch": 0.9617384071532543,
"eval_entropy": 1.6951171875,
"eval_loss": 0.9330964088439941,
"eval_mean_token_accuracy": 0.7741666674613953,
"eval_num_tokens": 2408404550.0,
"eval_runtime": 27.0047,
"eval_samples_per_second": 93.836,
"eval_steps_per_second": 1.481,
"step": 18500
},
{
"entropy": 1.705078125,
"epoch": 0.9622582657517156,
"grad_norm": 2.450121778441867,
"learning_rate": 3.8964519241076214e-07,
"loss": 0.9392,
"mean_token_accuracy": 0.7768229216337204,
"num_tokens": 2409705548.0,
"step": 18510
},
{
"entropy": 1.69765625,
"epoch": 0.9627781243501767,
"grad_norm": 2.8767263803446155,
"learning_rate": 3.8428556115339264e-07,
"loss": 0.9092,
"mean_token_accuracy": 0.7811383306980133,
"num_tokens": 2411006903.0,
"step": 18520
},
{
"entropy": 1.7,
"epoch": 0.963297982948638,
"grad_norm": 2.5683421275363334,
"learning_rate": 3.7892592989602313e-07,
"loss": 0.838,
"mean_token_accuracy": 0.7989502549171448,
"num_tokens": 2412309001.0,
"step": 18530
},
{
"entropy": 1.697265625,
"epoch": 0.9638178415470992,
"grad_norm": 2.2359370770057923,
"learning_rate": 3.735662986386537e-07,
"loss": 0.9032,
"mean_token_accuracy": 0.782451406121254,
"num_tokens": 2413609095.0,
"step": 18540
},
{
"entropy": 1.69296875,
"epoch": 0.9643377001455604,
"grad_norm": 2.4970426815760343,
"learning_rate": 3.6820666738128417e-07,
"loss": 0.858,
"mean_token_accuracy": 0.7932672381401062,
"num_tokens": 2414910939.0,
"step": 18550
},
{
"entropy": 1.6828125,
"epoch": 0.9648575587440216,
"grad_norm": 3.180042900712498,
"learning_rate": 3.6284703612391466e-07,
"loss": 0.9062,
"mean_token_accuracy": 0.7794785290956497,
"num_tokens": 2416209667.0,
"step": 18560
},
{
"entropy": 1.682421875,
"epoch": 0.9653774173424828,
"grad_norm": 2.4794721590068716,
"learning_rate": 3.574874048665452e-07,
"loss": 0.8978,
"mean_token_accuracy": 0.7835963577032089,
"num_tokens": 2417512475.0,
"step": 18570
},
{
"entropy": 1.68984375,
"epoch": 0.9658972759409441,
"grad_norm": 2.371723752991273,
"learning_rate": 3.521277736091757e-07,
"loss": 0.8954,
"mean_token_accuracy": 0.78342886865139,
"num_tokens": 2418812236.0,
"step": 18580
},
{
"entropy": 1.69296875,
"epoch": 0.9664171345394053,
"grad_norm": 3.300236537566366,
"learning_rate": 3.467681423518062e-07,
"loss": 0.91,
"mean_token_accuracy": 0.780864617228508,
"num_tokens": 2420113315.0,
"step": 18590
},
{
"entropy": 1.698828125,
"epoch": 0.9669369931378665,
"grad_norm": 2.3903514034450777,
"learning_rate": 3.4140851109443675e-07,
"loss": 0.9063,
"mean_token_accuracy": 0.7834517866373062,
"num_tokens": 2421414041.0,
"step": 18600
},
{
"epoch": 0.9669369931378665,
"eval_entropy": 1.6837890625,
"eval_loss": 0.9327428340911865,
"eval_mean_token_accuracy": 0.7741594478487969,
"eval_num_tokens": 2421414041.0,
"eval_runtime": 27.0702,
"eval_samples_per_second": 93.608,
"eval_steps_per_second": 1.478,
"step": 18600
},
{
"entropy": 1.690625,
"epoch": 0.9674568517363277,
"grad_norm": 3.1142091837932657,
"learning_rate": 3.3604887983706724e-07,
"loss": 0.8922,
"mean_token_accuracy": 0.7869615674018859,
"num_tokens": 2422713855.0,
"step": 18610
},
{
"entropy": 1.673046875,
"epoch": 0.9679767103347889,
"grad_norm": 2.5438506632592035,
"learning_rate": 3.3068924857969773e-07,
"loss": 0.8362,
"mean_token_accuracy": 0.7963182508945466,
"num_tokens": 2424014830.0,
"step": 18620
},
{
"entropy": 1.686328125,
"epoch": 0.9684965689332502,
"grad_norm": 2.5280088176316204,
"learning_rate": 3.253296173223282e-07,
"loss": 0.868,
"mean_token_accuracy": 0.790490847826004,
"num_tokens": 2425317921.0,
"step": 18630
},
{
"entropy": 1.680078125,
"epoch": 0.9690164275317114,
"grad_norm": 3.081924563054614,
"learning_rate": 3.1996998606495877e-07,
"loss": 0.8344,
"mean_token_accuracy": 0.797171813249588,
"num_tokens": 2426620072.0,
"step": 18640
},
{
"entropy": 1.677734375,
"epoch": 0.9695362861301726,
"grad_norm": 3.072533022834567,
"learning_rate": 3.1461035480758927e-07,
"loss": 0.8568,
"mean_token_accuracy": 0.7923459321260452,
"num_tokens": 2427919292.0,
"step": 18650
},
{
"entropy": 1.691015625,
"epoch": 0.9700561447286338,
"grad_norm": 2.4221227304027826,
"learning_rate": 3.0925072355021976e-07,
"loss": 0.8945,
"mean_token_accuracy": 0.7850315511226654,
"num_tokens": 2429221629.0,
"step": 18660
},
{
"entropy": 1.683203125,
"epoch": 0.970576003327095,
"grad_norm": 2.459875221325511,
"learning_rate": 3.038910922928503e-07,
"loss": 0.863,
"mean_token_accuracy": 0.7943198204040527,
"num_tokens": 2430525527.0,
"step": 18670
},
{
"entropy": 1.68671875,
"epoch": 0.9710958619255563,
"grad_norm": 2.6693227963345096,
"learning_rate": 2.985314610354808e-07,
"loss": 0.9007,
"mean_token_accuracy": 0.7844446778297425,
"num_tokens": 2431827920.0,
"step": 18680
},
{
"entropy": 1.67890625,
"epoch": 0.9716157205240175,
"grad_norm": 2.3015359080649866,
"learning_rate": 2.931718297781113e-07,
"loss": 0.775,
"mean_token_accuracy": 0.8164342522621155,
"num_tokens": 2433132535.0,
"step": 18690
},
{
"entropy": 1.6890625,
"epoch": 0.9721355791224787,
"grad_norm": 2.7496417220370706,
"learning_rate": 2.878121985207418e-07,
"loss": 0.8459,
"mean_token_accuracy": 0.7971536666154861,
"num_tokens": 2434434468.0,
"step": 18700
},
{
"epoch": 0.9721355791224787,
"eval_entropy": 1.678515625,
"eval_loss": 0.9319614768028259,
"eval_mean_token_accuracy": 0.774788984656334,
"eval_num_tokens": 2434434468.0,
"eval_runtime": 27.0296,
"eval_samples_per_second": 93.749,
"eval_steps_per_second": 1.48,
"step": 18700
},
{
"entropy": 1.676953125,
"epoch": 0.9726554377209399,
"grad_norm": 3.1770404870627655,
"learning_rate": 2.824525672633723e-07,
"loss": 0.8654,
"mean_token_accuracy": 0.7907682448625565,
"num_tokens": 2435736898.0,
"step": 18710
},
{
"entropy": 1.66875,
"epoch": 0.9731752963194011,
"grad_norm": 3.593683031407824,
"learning_rate": 2.770929360060028e-07,
"loss": 0.8728,
"mean_token_accuracy": 0.7884618639945984,
"num_tokens": 2437040550.0,
"step": 18720
},
{
"entropy": 1.707421875,
"epoch": 0.9736951549178623,
"grad_norm": 3.0311978017080765,
"learning_rate": 2.717333047486333e-07,
"loss": 0.9173,
"mean_token_accuracy": 0.7793467909097671,
"num_tokens": 2438340690.0,
"step": 18730
},
{
"entropy": 1.670703125,
"epoch": 0.9742150135163236,
"grad_norm": 2.387226343596206,
"learning_rate": 2.663736734912638e-07,
"loss": 0.8839,
"mean_token_accuracy": 0.7881101727485657,
"num_tokens": 2439642505.0,
"step": 18740
},
{
"entropy": 1.69609375,
"epoch": 0.9747348721147848,
"grad_norm": 2.3944772717819376,
"learning_rate": 2.610140422338943e-07,
"loss": 0.8881,
"mean_token_accuracy": 0.785685008764267,
"num_tokens": 2440945591.0,
"step": 18750
},
{
"entropy": 1.693359375,
"epoch": 0.975254730713246,
"grad_norm": 2.6352727051188363,
"learning_rate": 2.5565441097652486e-07,
"loss": 0.9129,
"mean_token_accuracy": 0.7785031735897064,
"num_tokens": 2442245215.0,
"step": 18760
},
{
"entropy": 1.690625,
"epoch": 0.9757745893117072,
"grad_norm": 2.66521042737604,
"learning_rate": 2.5029477971915535e-07,
"loss": 0.9835,
"mean_token_accuracy": 0.7662282586097717,
"num_tokens": 2443547399.0,
"step": 18770
},
{
"entropy": 1.680859375,
"epoch": 0.9762944479101684,
"grad_norm": 2.7115842240661143,
"learning_rate": 2.4493514846178584e-07,
"loss": 0.8827,
"mean_token_accuracy": 0.7884075313806533,
"num_tokens": 2444850499.0,
"step": 18780
},
{
"entropy": 1.698046875,
"epoch": 0.9768143065086297,
"grad_norm": 3.402367908886023,
"learning_rate": 2.3957551720441634e-07,
"loss": 0.8936,
"mean_token_accuracy": 0.7865809917449951,
"num_tokens": 2446150149.0,
"step": 18790
},
{
"entropy": 1.69921875,
"epoch": 0.9773341651070909,
"grad_norm": 3.165976991096047,
"learning_rate": 2.3421588594704686e-07,
"loss": 0.9301,
"mean_token_accuracy": 0.7765855729579926,
"num_tokens": 2447451236.0,
"step": 18800
},
{
"epoch": 0.9773341651070909,
"eval_entropy": 1.6810546875,
"eval_loss": 0.9315959215164185,
"eval_mean_token_accuracy": 0.7747406706213951,
"eval_num_tokens": 2447451236.0,
"eval_runtime": 27.065,
"eval_samples_per_second": 93.627,
"eval_steps_per_second": 1.478,
"step": 18800
},
{
"entropy": 1.6953125,
"epoch": 0.9778540237055521,
"grad_norm": 2.3436093024990705,
"learning_rate": 2.2885625468967738e-07,
"loss": 0.8485,
"mean_token_accuracy": 0.7953097254037857,
"num_tokens": 2448751474.0,
"step": 18810
},
{
"entropy": 1.69140625,
"epoch": 0.9783738823040133,
"grad_norm": 2.5182373854493862,
"learning_rate": 2.234966234323079e-07,
"loss": 0.848,
"mean_token_accuracy": 0.7940034896135331,
"num_tokens": 2450054501.0,
"step": 18820
},
{
"entropy": 1.6859375,
"epoch": 0.9788937409024745,
"grad_norm": 2.665064125808606,
"learning_rate": 2.181369921749384e-07,
"loss": 0.9,
"mean_token_accuracy": 0.7836889952421189,
"num_tokens": 2451358321.0,
"step": 18830
},
{
"entropy": 1.677734375,
"epoch": 0.9794135995009358,
"grad_norm": 2.925937667193729,
"learning_rate": 2.127773609175689e-07,
"loss": 0.8781,
"mean_token_accuracy": 0.787773722410202,
"num_tokens": 2452659742.0,
"step": 18840
},
{
"entropy": 1.709375,
"epoch": 0.9799334580993969,
"grad_norm": 2.3373593779837516,
"learning_rate": 2.0741772966019938e-07,
"loss": 0.927,
"mean_token_accuracy": 0.7762221962213516,
"num_tokens": 2453964028.0,
"step": 18850
},
{
"entropy": 1.705859375,
"epoch": 0.9804533166978582,
"grad_norm": 3.2982403007087275,
"learning_rate": 2.020580984028299e-07,
"loss": 0.9078,
"mean_token_accuracy": 0.7820105463266372,
"num_tokens": 2455267358.0,
"step": 18860
},
{
"entropy": 1.6890625,
"epoch": 0.9809731752963194,
"grad_norm": 3.330345014988728,
"learning_rate": 1.966984671454604e-07,
"loss": 0.9082,
"mean_token_accuracy": 0.7807829558849335,
"num_tokens": 2456568645.0,
"step": 18870
},
{
"entropy": 1.6921875,
"epoch": 0.9814930338947806,
"grad_norm": 2.8137089074308634,
"learning_rate": 1.913388358880909e-07,
"loss": 0.8544,
"mean_token_accuracy": 0.7938121765851974,
"num_tokens": 2457870131.0,
"step": 18880
},
{
"entropy": 1.694921875,
"epoch": 0.9820128924932419,
"grad_norm": 3.1419059789387402,
"learning_rate": 1.859792046307214e-07,
"loss": 0.9438,
"mean_token_accuracy": 0.7730648577213287,
"num_tokens": 2459170827.0,
"step": 18890
},
{
"entropy": 1.69375,
"epoch": 0.982532751091703,
"grad_norm": 3.0190823232561335,
"learning_rate": 1.8061957337335192e-07,
"loss": 0.8431,
"mean_token_accuracy": 0.7977301478385925,
"num_tokens": 2460474349.0,
"step": 18900
},
{
"epoch": 0.982532751091703,
"eval_entropy": 1.6791015625,
"eval_loss": 0.9312637448310852,
"eval_mean_token_accuracy": 0.7747837781906128,
"eval_num_tokens": 2460474349.0,
"eval_runtime": 27.0384,
"eval_samples_per_second": 93.719,
"eval_steps_per_second": 1.479,
"step": 18900
},
{
"entropy": 1.686328125,
"epoch": 0.9830526096901643,
"grad_norm": 2.982073953789415,
"learning_rate": 1.7525994211598245e-07,
"loss": 0.8894,
"mean_token_accuracy": 0.7862489014863968,
"num_tokens": 2461777441.0,
"step": 18910
},
{
"entropy": 1.702734375,
"epoch": 0.9835724682886255,
"grad_norm": 2.208776605795735,
"learning_rate": 1.6990031085861294e-07,
"loss": 0.9616,
"mean_token_accuracy": 0.7683578997850418,
"num_tokens": 2463076701.0,
"step": 18920
},
{
"entropy": 1.683203125,
"epoch": 0.9840923268870867,
"grad_norm": 2.654125493772313,
"learning_rate": 1.6454067960124346e-07,
"loss": 0.9104,
"mean_token_accuracy": 0.7820166140794754,
"num_tokens": 2464377527.0,
"step": 18930
},
{
"entropy": 1.707421875,
"epoch": 0.984612185485548,
"grad_norm": 2.489495822093012,
"learning_rate": 1.5918104834387395e-07,
"loss": 0.8753,
"mean_token_accuracy": 0.787780550122261,
"num_tokens": 2465680413.0,
"step": 18940
},
{
"entropy": 1.680859375,
"epoch": 0.9851320440840091,
"grad_norm": 3.1388298891881945,
"learning_rate": 1.5382141708650445e-07,
"loss": 0.8613,
"mean_token_accuracy": 0.7904051959514617,
"num_tokens": 2466981607.0,
"step": 18950
},
{
"entropy": 1.7015625,
"epoch": 0.9856519026824704,
"grad_norm": 2.5481635771960454,
"learning_rate": 1.4846178582913497e-07,
"loss": 0.9688,
"mean_token_accuracy": 0.7667078703641892,
"num_tokens": 2468284551.0,
"step": 18960
},
{
"entropy": 1.7078125,
"epoch": 0.9861717612809315,
"grad_norm": 3.178238248840821,
"learning_rate": 1.4310215457176546e-07,
"loss": 0.8764,
"mean_token_accuracy": 0.7897622525691986,
"num_tokens": 2469587852.0,
"step": 18970
},
{
"entropy": 1.698046875,
"epoch": 0.9866916198793928,
"grad_norm": 2.400598746534951,
"learning_rate": 1.3774252331439598e-07,
"loss": 0.8374,
"mean_token_accuracy": 0.8000957489013671,
"num_tokens": 2470889388.0,
"step": 18980
},
{
"entropy": 1.687890625,
"epoch": 0.9872114784778541,
"grad_norm": 2.692549832592766,
"learning_rate": 1.323828920570265e-07,
"loss": 0.8685,
"mean_token_accuracy": 0.7907023280858994,
"num_tokens": 2472189071.0,
"step": 18990
},
{
"entropy": 1.694921875,
"epoch": 0.9877313370763152,
"grad_norm": 2.6933063556344683,
"learning_rate": 1.27023260799657e-07,
"loss": 0.8864,
"mean_token_accuracy": 0.7878907710313797,
"num_tokens": 2473489988.0,
"step": 19000
},
{
"epoch": 0.9877313370763152,
"eval_entropy": 1.6853515625,
"eval_loss": 0.9311267137527466,
"eval_mean_token_accuracy": 0.7748992249369622,
"eval_num_tokens": 2473489988.0,
"eval_runtime": 27.0716,
"eval_samples_per_second": 93.604,
"eval_steps_per_second": 1.478,
"step": 19000
},
{
"entropy": 1.709765625,
"epoch": 0.9882511956747765,
"grad_norm": 2.9293687939219133,
"learning_rate": 1.2166362954228751e-07,
"loss": 0.9088,
"mean_token_accuracy": 0.7844247817993164,
"num_tokens": 2474793672.0,
"step": 19010
},
{
"entropy": 1.71484375,
"epoch": 0.9887710542732376,
"grad_norm": 4.016326418894613,
"learning_rate": 1.16303998284918e-07,
"loss": 0.903,
"mean_token_accuracy": 0.7803129762411117,
"num_tokens": 2476095711.0,
"step": 19020
},
{
"entropy": 1.69921875,
"epoch": 0.9892909128716989,
"grad_norm": 2.6975740619212383,
"learning_rate": 1.1094436702754851e-07,
"loss": 0.948,
"mean_token_accuracy": 0.7713859885931015,
"num_tokens": 2477399265.0,
"step": 19030
},
{
"entropy": 1.705859375,
"epoch": 0.9898107714701602,
"grad_norm": 2.6784085683444205,
"learning_rate": 1.0558473577017902e-07,
"loss": 0.9136,
"mean_token_accuracy": 0.7810904264450074,
"num_tokens": 2478702211.0,
"step": 19040
},
{
"entropy": 1.69375,
"epoch": 0.9903306300686213,
"grad_norm": 2.336027665351406,
"learning_rate": 1.0022510451280953e-07,
"loss": 0.8791,
"mean_token_accuracy": 0.7893612116575242,
"num_tokens": 2480004949.0,
"step": 19050
},
{
"entropy": 1.683984375,
"epoch": 0.9908504886670826,
"grad_norm": 2.271903875457208,
"learning_rate": 9.486547325544003e-08,
"loss": 0.8626,
"mean_token_accuracy": 0.794873270392418,
"num_tokens": 2481309107.0,
"step": 19060
},
{
"entropy": 1.718359375,
"epoch": 0.9913703472655437,
"grad_norm": 2.54711099524349,
"learning_rate": 8.950584199807054e-08,
"loss": 0.9141,
"mean_token_accuracy": 0.7807758837938309,
"num_tokens": 2482608312.0,
"step": 19070
},
{
"entropy": 1.714453125,
"epoch": 0.991890205864005,
"grad_norm": 3.103356782439801,
"learning_rate": 8.414621074070105e-08,
"loss": 0.9063,
"mean_token_accuracy": 0.7810538589954377,
"num_tokens": 2483908925.0,
"step": 19080
},
{
"entropy": 1.688671875,
"epoch": 0.9924100644624662,
"grad_norm": 3.1140689643725485,
"learning_rate": 7.878657948333154e-08,
"loss": 0.909,
"mean_token_accuracy": 0.7818417578935624,
"num_tokens": 2485212728.0,
"step": 19090
},
{
"entropy": 1.702734375,
"epoch": 0.9929299230609274,
"grad_norm": 2.5297429493885333,
"learning_rate": 7.342694822596206e-08,
"loss": 0.9797,
"mean_token_accuracy": 0.766809344291687,
"num_tokens": 2486517117.0,
"step": 19100
},
{
"epoch": 0.9929299230609274,
"eval_entropy": 1.6794921875,
"eval_loss": 0.9309945106506348,
"eval_mean_token_accuracy": 0.7749229103326798,
"eval_num_tokens": 2486517117.0,
"eval_runtime": 27.0017,
"eval_samples_per_second": 93.846,
"eval_steps_per_second": 1.481,
"step": 19100
},
{
"entropy": 1.69921875,
"epoch": 0.9934497816593887,
"grad_norm": 2.514164395691452,
"learning_rate": 6.806731696859257e-08,
"loss": 0.8399,
"mean_token_accuracy": 0.7947330474853516,
"num_tokens": 2487819441.0,
"step": 19110
},
{
"entropy": 1.68984375,
"epoch": 0.9939696402578498,
"grad_norm": 2.599941327956738,
"learning_rate": 6.270768571122308e-08,
"loss": 0.9255,
"mean_token_accuracy": 0.7784787833690643,
"num_tokens": 2489118339.0,
"step": 19120
},
{
"entropy": 1.684765625,
"epoch": 0.9944894988563111,
"grad_norm": 2.4650581446572724,
"learning_rate": 5.734805445385358e-08,
"loss": 0.9304,
"mean_token_accuracy": 0.7770549833774567,
"num_tokens": 2490421255.0,
"step": 19130
},
{
"entropy": 1.682421875,
"epoch": 0.9950093574547723,
"grad_norm": 2.794362035065077,
"learning_rate": 5.1988423196484083e-08,
"loss": 0.8395,
"mean_token_accuracy": 0.7976394891738892,
"num_tokens": 2491721750.0,
"step": 19140
},
{
"entropy": 1.69375,
"epoch": 0.9955292160532335,
"grad_norm": 2.7587915943060057,
"learning_rate": 4.662879193911459e-08,
"loss": 0.886,
"mean_token_accuracy": 0.7875616699457169,
"num_tokens": 2493024389.0,
"step": 19150
},
{
"entropy": 1.683203125,
"epoch": 0.9960490746516948,
"grad_norm": 2.1834535751247754,
"learning_rate": 4.1269160681745104e-08,
"loss": 0.8854,
"mean_token_accuracy": 0.7831056147813797,
"num_tokens": 2494325675.0,
"step": 19160
},
{
"entropy": 1.69375,
"epoch": 0.9965689332501559,
"grad_norm": 2.734910804116641,
"learning_rate": 3.5909529424375604e-08,
"loss": 0.8702,
"mean_token_accuracy": 0.7911708027124404,
"num_tokens": 2495626396.0,
"step": 19170
},
{
"entropy": 1.70234375,
"epoch": 0.9970887918486172,
"grad_norm": 2.6895310555741188,
"learning_rate": 3.054989816700611e-08,
"loss": 0.9195,
"mean_token_accuracy": 0.780265873670578,
"num_tokens": 2496929751.0,
"step": 19180
},
{
"entropy": 1.694140625,
"epoch": 0.9976086504470784,
"grad_norm": 2.27287323464454,
"learning_rate": 2.519026690963662e-08,
"loss": 0.8908,
"mean_token_accuracy": 0.7883580178022385,
"num_tokens": 2498229781.0,
"step": 19190
},
{
"entropy": 1.686328125,
"epoch": 0.9981285090455396,
"grad_norm": 2.538521479557618,
"learning_rate": 1.9830635652267125e-08,
"loss": 0.9023,
"mean_token_accuracy": 0.783353590965271,
"num_tokens": 2499528308.0,
"step": 19200
},
{
"epoch": 0.9981285090455396,
"eval_entropy": 1.6822265625,
"eval_loss": 0.9309091567993164,
"eval_mean_token_accuracy": 0.7749084889888763,
"eval_num_tokens": 2499528308.0,
"eval_runtime": 26.9938,
"eval_samples_per_second": 93.873,
"eval_steps_per_second": 1.482,
"step": 19200
},
{
"entropy": 1.70859375,
"epoch": 0.9986483676440008,
"grad_norm": 3.0391651375707296,
"learning_rate": 1.4471004394897633e-08,
"loss": 0.8911,
"mean_token_accuracy": 0.7850496351718903,
"num_tokens": 2500828911.0,
"step": 19210
},
{
"entropy": 1.684765625,
"epoch": 0.999168226242462,
"grad_norm": 2.7163701476522024,
"learning_rate": 9.111373137528138e-09,
"loss": 0.8661,
"mean_token_accuracy": 0.7926260650157928,
"num_tokens": 2502132253.0,
"step": 19220
},
{
"entropy": 1.714453125,
"epoch": 0.9996880848409233,
"grad_norm": 2.3876050494787733,
"learning_rate": 3.751741880158645e-09,
"loss": 0.91,
"mean_token_accuracy": 0.7799148887395859,
"num_tokens": 2503434782.0,
"step": 19230
},
{
"entropy": 1.693359375,
"epoch": 1.0,
"mean_token_accuracy": 0.7858945379654566,
"num_tokens": 2504216233.0,
"step": 19236,
"total_flos": 4000231083835392.0,
"train_loss": 1.1146164414391346,
"train_runtime": 104826.0086,
"train_samples_per_second": 23.488,
"train_steps_per_second": 0.184
}
],
"logging_steps": 10,
"max_steps": 19236,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1924,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4000231083835392.0,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}
Reference in New Issue View Git Blame Copy Permalink