{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 134,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 2.3614121675491333,
      "epoch": 0.0075046904315197,
      "grad_norm": 43.25616455078125,
      "learning_rate": 0.0,
      "loss": 8.111627578735352,
      "mean_token_accuracy": 0.15009256824851036,
      "num_tokens": 45107.0,
      "step": 1
    },
    {
      "entropy": 2.371093451976776,
      "epoch": 0.0150093808630394,
      "grad_norm": 43.40286636352539,
      "learning_rate": 4e-05,
      "loss": 8.187829971313477,
      "mean_token_accuracy": 0.14955881237983704,
      "num_tokens": 90056.0,
      "step": 2
    },
    {
      "entropy": 2.339372396469116,
      "epoch": 0.0225140712945591,
      "grad_norm": 41.962520599365234,
      "learning_rate": 8e-05,
      "loss": 7.960697650909424,
      "mean_token_accuracy": 0.1606680415570736,
      "num_tokens": 134197.0,
      "step": 3
    },
    {
      "entropy": 2.590428411960602,
      "epoch": 0.0300187617260788,
      "grad_norm": 40.314186096191406,
      "learning_rate": 0.00012,
      "loss": 7.319167137145996,
      "mean_token_accuracy": 0.17064783722162247,
      "num_tokens": 180050.0,
      "step": 4
    },
    {
      "entropy": 3.0919135212898254,
      "epoch": 0.0375234521575985,
      "grad_norm": 35.18631362915039,
      "learning_rate": 0.00016,
      "loss": 5.797598838806152,
      "mean_token_accuracy": 0.22301814332604408,
      "num_tokens": 228420.0,
      "step": 5
    },
    {
      "entropy": 3.202889919281006,
      "epoch": 0.0450281425891182,
      "grad_norm": 20.162179946899414,
      "learning_rate": 0.0002,
      "loss": 3.864182949066162,
      "mean_token_accuracy": 0.3619539961218834,
      "num_tokens": 264539.0,
      "step": 6
    },
    {
      "entropy": 2.883633553981781,
      "epoch": 0.0525328330206379,
      "grad_norm": 11.903804779052734,
      "learning_rate": 0.00019997331228606256,
      "loss": 2.9499809741973877,
      "mean_token_accuracy": 0.46186020225286484,
      "num_tokens": 312810.0,
      "step": 7
    },
    {
      "entropy": 1.9907608032226562,
      "epoch": 0.0600375234521576,
      "grad_norm": 7.249192714691162,
      "learning_rate": 0.00019989326497167406,
      "loss": 2.2568702697753906,
      "mean_token_accuracy": 0.5792697072029114,
      "num_tokens": 354449.0,
      "step": 8
    },
    {
      "entropy": 1.5780252516269684,
      "epoch": 0.0675422138836773,
      "grad_norm": 4.371347427368164,
      "learning_rate": 0.00019975990552971966,
      "loss": 1.9040088653564453,
      "mean_token_accuracy": 0.6360478699207306,
      "num_tokens": 403860.0,
      "step": 9
    },
    {
      "entropy": 1.5423422455787659,
      "epoch": 0.075046904315197,
      "grad_norm": 3.3591129779815674,
      "learning_rate": 0.00019957331305039123,
      "loss": 1.919539213180542,
      "mean_token_accuracy": 0.6306671798229218,
      "num_tokens": 448604.0,
      "step": 10
    },
    {
      "entropy": 1.5162997543811798,
      "epoch": 0.0825515947467167,
      "grad_norm": 2.5374269485473633,
      "learning_rate": 0.00019933359819428238,
      "loss": 1.8262466192245483,
      "mean_token_accuracy": 0.6332019418478012,
      "num_tokens": 486186.0,
      "step": 11
    },
    {
      "entropy": 1.6451953053474426,
      "epoch": 0.0900562851782364,
      "grad_norm": 2.1963534355163574,
      "learning_rate": 0.00019904090312675995,
      "loss": 1.8073338270187378,
      "mean_token_accuracy": 0.6270972341299057,
      "num_tokens": 526759.0,
      "step": 12
    },
    {
      "entropy": 1.5697188675403595,
      "epoch": 0.0975609756097561,
      "grad_norm": 1.21969735622406,
      "learning_rate": 0.00019869540143365127,
      "loss": 1.574568510055542,
      "mean_token_accuracy": 0.658010259270668,
      "num_tokens": 572472.0,
      "step": 13
    },
    {
      "entropy": 1.5532931089401245,
      "epoch": 0.1050656660412758,
      "grad_norm": 1.0103905200958252,
      "learning_rate": 0.0001982972980182975,
      "loss": 1.5713064670562744,
      "mean_token_accuracy": 0.6637546718120575,
      "num_tokens": 614471.0,
      "step": 14
    },
    {
      "entropy": 1.4622831046581268,
      "epoch": 0.1125703564727955,
      "grad_norm": 0.8437947630882263,
      "learning_rate": 0.00019784682898003347,
      "loss": 1.486205816268921,
      "mean_token_accuracy": 0.6798983067274094,
      "num_tokens": 656953.0,
      "step": 15
    },
    {
      "entropy": 1.533287137746811,
      "epoch": 0.1200750469043152,
      "grad_norm": 0.7793806195259094,
      "learning_rate": 0.00019734426147416647,
      "loss": 1.556842565536499,
      "mean_token_accuracy": 0.6637462079524994,
      "num_tokens": 691655.0,
      "step": 16
    },
    {
      "entropy": 1.325183093547821,
      "epoch": 0.1275797373358349,
      "grad_norm": 0.5686734318733215,
      "learning_rate": 0.00019678989355353726,
      "loss": 1.357476830482483,
      "mean_token_accuracy": 0.70518858730793,
      "num_tokens": 739780.0,
      "step": 17
    },
    {
      "entropy": 1.3306374549865723,
      "epoch": 0.1350844277673546,
      "grad_norm": 0.4603631794452667,
      "learning_rate": 0.00019618405399175653,
      "loss": 1.3604191541671753,
      "mean_token_accuracy": 0.7013339400291443,
      "num_tokens": 788169.0,
      "step": 18
    },
    {
      "entropy": 1.4548923969268799,
      "epoch": 0.1425891181988743,
      "grad_norm": 0.46769508719444275,
      "learning_rate": 0.00019552710208822245,
      "loss": 1.4896512031555176,
      "mean_token_accuracy": 0.6769208163022995,
      "num_tokens": 826585.0,
      "step": 19
    },
    {
      "entropy": 1.3101635575294495,
      "epoch": 0.150093808630394,
      "grad_norm": 0.39558693766593933,
      "learning_rate": 0.0001948194274550343,
      "loss": 1.3496911525726318,
      "mean_token_accuracy": 0.7120275795459747,
      "num_tokens": 876680.0,
      "step": 20
    },
    {
      "entropy": 1.3913719058036804,
      "epoch": 0.1575984990619137,
      "grad_norm": 0.3726045489311218,
      "learning_rate": 0.00019406144978592932,
      "loss": 1.3945075273513794,
      "mean_token_accuracy": 0.6928794682025909,
      "num_tokens": 929212.0,
      "step": 21
    },
    {
      "entropy": 1.4546256363391876,
      "epoch": 0.1651031894934334,
      "grad_norm": 0.3835321366786957,
      "learning_rate": 0.0001932536186073787,
      "loss": 1.4543089866638184,
      "mean_token_accuracy": 0.6881714165210724,
      "num_tokens": 968963.0,
      "step": 22
    },
    {
      "entropy": 1.4929974675178528,
      "epoch": 0.1726078799249531,
      "grad_norm": 0.38340237736701965,
      "learning_rate": 0.00019239641301199137,
      "loss": 1.4268107414245605,
      "mean_token_accuracy": 0.6830715984106064,
      "num_tokens": 1014538.0,
      "step": 23
    },
    {
      "entropy": 1.3382728695869446,
      "epoch": 0.1801125703564728,
      "grad_norm": 0.3726077377796173,
      "learning_rate": 0.00019149034137438321,
      "loss": 1.239941120147705,
      "mean_token_accuracy": 0.7238551080226898,
      "num_tokens": 1052897.0,
      "step": 24
    },
    {
      "entropy": 1.414399266242981,
      "epoch": 0.18761726078799248,
      "grad_norm": 0.3222167193889618,
      "learning_rate": 0.00019053594104968027,
      "loss": 1.3135441541671753,
      "mean_token_accuracy": 0.7092919200658798,
      "num_tokens": 1097023.0,
      "step": 25
    },
    {
      "entropy": 1.4715129137039185,
      "epoch": 0.1951219512195122,
      "grad_norm": 0.3475184738636017,
      "learning_rate": 0.00018953377805483476,
      "loss": 1.3715906143188477,
      "mean_token_accuracy": 0.6958068162202835,
      "num_tokens": 1145280.0,
      "step": 26
    },
    {
      "entropy": 1.389900654554367,
      "epoch": 0.2026266416510319,
      "grad_norm": 0.36157509684562683,
      "learning_rate": 0.0001884844467329427,
      "loss": 1.2903525829315186,
      "mean_token_accuracy": 0.7132891714572906,
      "num_tokens": 1192288.0,
      "step": 27
    },
    {
      "entropy": 1.326958566904068,
      "epoch": 0.2101313320825516,
      "grad_norm": 0.3483531177043915,
      "learning_rate": 0.00018738856940076292,
      "loss": 1.251339077949524,
      "mean_token_accuracy": 0.7226762473583221,
      "num_tokens": 1234402.0,
      "step": 28
    },
    {
      "entropy": 1.3374330699443817,
      "epoch": 0.2176360225140713,
      "grad_norm": 0.37691307067871094,
      "learning_rate": 0.00018624679597964541,
      "loss": 1.24509859085083,
      "mean_token_accuracy": 0.7150529026985168,
      "num_tokens": 1273203.0,
      "step": 29
    },
    {
      "entropy": 1.1258195340633392,
      "epoch": 0.225140712945591,
      "grad_norm": 0.2888593077659607,
      "learning_rate": 0.00018505980361008901,
      "loss": 1.093227505683899,
      "mean_token_accuracy": 0.752630427479744,
      "num_tokens": 1320634.0,
      "step": 30
    },
    {
      "entropy": 1.2001722157001495,
      "epoch": 0.2326454033771107,
      "grad_norm": 0.2866954803466797,
      "learning_rate": 0.00018382829625015597,
      "loss": 1.1965141296386719,
      "mean_token_accuracy": 0.7307434678077698,
      "num_tokens": 1365849.0,
      "step": 31
    },
    {
      "entropy": 1.0254056453704834,
      "epoch": 0.2401500938086304,
      "grad_norm": 0.31378430128097534,
      "learning_rate": 0.00018255300425798268,
      "loss": 1.0252065658569336,
      "mean_token_accuracy": 0.7702369093894958,
      "num_tokens": 1413252.0,
      "step": 32
    },
    {
      "entropy": 1.2184112668037415,
      "epoch": 0.24765478424015008,
      "grad_norm": 0.323446124792099,
      "learning_rate": 0.00018123468395863287,
      "loss": 1.2506036758422852,
      "mean_token_accuracy": 0.7172769755125046,
      "num_tokens": 1458782.0,
      "step": 33
    },
    {
      "entropy": 1.1687179207801819,
      "epoch": 0.2551594746716698,
      "grad_norm": 0.3320324718952179,
      "learning_rate": 0.00017987411719555172,
      "loss": 1.1786611080169678,
      "mean_token_accuracy": 0.7287575006484985,
      "num_tokens": 1505304.0,
      "step": 34
    },
    {
      "entropy": 1.0758436769247055,
      "epoch": 0.2626641651031895,
      "grad_norm": 0.31610196828842163,
      "learning_rate": 0.00017847211086688503,
      "loss": 1.083768367767334,
      "mean_token_accuracy": 0.7550794929265976,
      "num_tokens": 1556382.0,
      "step": 35
    },
    {
      "entropy": 1.2227874994277954,
      "epoch": 0.2701688555347092,
      "grad_norm": 0.3684009611606598,
      "learning_rate": 0.00017702949644694038,
      "loss": 1.202125072479248,
      "mean_token_accuracy": 0.7233716994524002,
      "num_tokens": 1591634.0,
      "step": 36
    },
    {
      "entropy": 1.0862102061510086,
      "epoch": 0.2776735459662289,
      "grad_norm": 0.3123534321784973,
      "learning_rate": 0.00017554712949307243,
      "loss": 1.0474365949630737,
      "mean_token_accuracy": 0.7583552747964859,
      "num_tokens": 1641330.0,
      "step": 37
    },
    {
      "entropy": 1.1538784801959991,
      "epoch": 0.2851782363977486,
      "grad_norm": 0.33295923471450806,
      "learning_rate": 0.00017402588913828602,
      "loss": 1.1146290302276611,
      "mean_token_accuracy": 0.7485005110502243,
      "num_tokens": 1682964.0,
      "step": 38
    },
    {
      "entropy": 1.1429019272327423,
      "epoch": 0.2926829268292683,
      "grad_norm": 0.3762746751308441,
      "learning_rate": 0.00017246667756985707,
      "loss": 1.0793370008468628,
      "mean_token_accuracy": 0.7509189248085022,
      "num_tokens": 1726656.0,
      "step": 39
    },
    {
      "entropy": 1.0788290947675705,
      "epoch": 0.300187617260788,
      "grad_norm": 0.35338789224624634,
      "learning_rate": 0.00017087041949428124,
      "loss": 1.0463558435440063,
      "mean_token_accuracy": 0.7557903528213501,
      "num_tokens": 1772848.0,
      "step": 40
    },
    {
      "entropy": 1.0233246088027954,
      "epoch": 0.3076923076923077,
      "grad_norm": 0.38455283641815186,
      "learning_rate": 0.00016923806158886733,
      "loss": 0.9958292841911316,
      "mean_token_accuracy": 0.7716485261917114,
      "num_tokens": 1813929.0,
      "step": 41
    },
    {
      "entropy": 0.9923187047243118,
      "epoch": 0.3151969981238274,
      "grad_norm": 0.3728187680244446,
      "learning_rate": 0.0001675705719403002,
      "loss": 0.9894548654556274,
      "mean_token_accuracy": 0.7702084630727768,
      "num_tokens": 1860559.0,
      "step": 42
    },
    {
      "entropy": 0.9995710849761963,
      "epoch": 0.3227016885553471,
      "grad_norm": 0.37269216775894165,
      "learning_rate": 0.00016586893947050736,
      "loss": 1.025907039642334,
      "mean_token_accuracy": 0.7622115463018417,
      "num_tokens": 1905263.0,
      "step": 43
    },
    {
      "entropy": 1.0477020889520645,
      "epoch": 0.3302063789868668,
      "grad_norm": 0.41793540120124817,
      "learning_rate": 0.00016413417335016834,
      "loss": 1.042026162147522,
      "mean_token_accuracy": 0.7561753541231155,
      "num_tokens": 1949817.0,
      "step": 44
    },
    {
      "entropy": 0.9699642211198807,
      "epoch": 0.33771106941838647,
      "grad_norm": 0.41920262575149536,
      "learning_rate": 0.00016236730240021577,
      "loss": 0.9484196901321411,
      "mean_token_accuracy": 0.7777801603078842,
      "num_tokens": 1987323.0,
      "step": 45
    },
    {
      "entropy": 0.9605746865272522,
      "epoch": 0.3452157598499062,
      "grad_norm": 0.42247241735458374,
      "learning_rate": 0.00016056937448168243,
      "loss": 0.9348800778388977,
      "mean_token_accuracy": 0.7801952809095383,
      "num_tokens": 2035301.0,
      "step": 46
    },
    {
      "entropy": 1.0968843400478363,
      "epoch": 0.3527204502814259,
      "grad_norm": 0.5121493339538574,
      "learning_rate": 0.00015874145587425661,
      "loss": 1.1251085996627808,
      "mean_token_accuracy": 0.7408822476863861,
      "num_tokens": 2075276.0,
      "step": 47
    },
    {
      "entropy": 0.93604576587677,
      "epoch": 0.3602251407129456,
      "grad_norm": 0.4335438907146454,
      "learning_rate": 0.00015688463064391373,
      "loss": 0.9258860945701599,
      "mean_token_accuracy": 0.7860793620347977,
      "num_tokens": 2123516.0,
      "step": 48
    },
    {
      "entropy": 0.9970589131116867,
      "epoch": 0.3677298311444653,
      "grad_norm": 0.4164389371871948,
      "learning_rate": 0.000155,
      "loss": 0.9580228328704834,
      "mean_token_accuracy": 0.7780613005161285,
      "num_tokens": 2169918.0,
      "step": 49
    },
    {
      "entropy": 0.8539141863584518,
      "epoch": 0.37523452157598497,
      "grad_norm": 0.37810465693473816,
      "learning_rate": 0.00015308868164214885,
      "loss": 0.8277876377105713,
      "mean_token_accuracy": 0.8033854067325592,
      "num_tokens": 2218158.0,
      "step": 50
    },
    {
      "entropy": 1.0844905823469162,
      "epoch": 0.3827392120075047,
      "grad_norm": 0.4299311339855194,
      "learning_rate": 0.00015115180909741749,
      "loss": 1.0677986145019531,
      "mean_token_accuracy": 0.7502051144838333,
      "num_tokens": 2257094.0,
      "step": 51
    },
    {
      "entropy": 0.9411480873823166,
      "epoch": 0.3902439024390244,
      "grad_norm": 0.4616101384162903,
      "learning_rate": 0.00014919053104803722,
      "loss": 0.897161066532135,
      "mean_token_accuracy": 0.7910304218530655,
      "num_tokens": 2301904.0,
      "step": 52
    },
    {
      "entropy": 0.8943736851215363,
      "epoch": 0.3977485928705441,
      "grad_norm": 0.5220033526420593,
      "learning_rate": 0.0001472060106501755,
      "loss": 0.9061237573623657,
      "mean_token_accuracy": 0.7858951538801193,
      "num_tokens": 2344732.0,
      "step": 53
    },
    {
      "entropy": 0.8761279582977295,
      "epoch": 0.4052532833020638,
      "grad_norm": 0.5949953198432922,
      "learning_rate": 0.00014519942484411403,
      "loss": 0.9096457958221436,
      "mean_token_accuracy": 0.7874524891376495,
      "num_tokens": 2389681.0,
      "step": 54
    },
    {
      "entropy": 0.8406584709882736,
      "epoch": 0.41275797373358347,
      "grad_norm": 0.468824565410614,
      "learning_rate": 0.00014317196365625255,
      "loss": 0.8100703954696655,
      "mean_token_accuracy": 0.8096983134746552,
      "num_tokens": 2435706.0,
      "step": 55
    },
    {
      "entropy": 0.9034422188997269,
      "epoch": 0.4202626641651032,
      "grad_norm": 0.4798317551612854,
      "learning_rate": 0.00014112482949335105,
      "loss": 0.8664395809173584,
      "mean_token_accuracy": 0.7980868816375732,
      "num_tokens": 2482015.0,
      "step": 56
    },
    {
      "entropy": 0.7600571662187576,
      "epoch": 0.4277673545966229,
      "grad_norm": 0.48762235045433044,
      "learning_rate": 0.00013905923642943018,
      "loss": 0.7243920564651489,
      "mean_token_accuracy": 0.8297153562307358,
      "num_tokens": 2529345.0,
      "step": 57
    },
    {
      "entropy": 0.8553325086832047,
      "epoch": 0.4352720450281426,
      "grad_norm": 0.5713600516319275,
      "learning_rate": 0.0001369764094857522,
      "loss": 0.8530070781707764,
      "mean_token_accuracy": 0.7953900098800659,
      "num_tokens": 2568227.0,
      "step": 58
    },
    {
      "entropy": 0.6945742964744568,
      "epoch": 0.4427767354596623,
      "grad_norm": 0.542955219745636,
      "learning_rate": 0.00013487758390430944,
      "loss": 0.6961327195167542,
      "mean_token_accuracy": 0.8368848413228989,
      "num_tokens": 2616056.0,
      "step": 59
    },
    {
      "entropy": 0.7480716407299042,
      "epoch": 0.450281425891182,
      "grad_norm": 0.5016547441482544,
      "learning_rate": 0.00013276400441525126,
      "loss": 0.6955996155738831,
      "mean_token_accuracy": 0.8333006799221039,
      "num_tokens": 2658729.0,
      "step": 60
    },
    {
      "entropy": 0.734462708234787,
      "epoch": 0.45778611632270166,
      "grad_norm": 0.4354403614997864,
      "learning_rate": 0.00013063692449868412,
      "loss": 0.6925354599952698,
      "mean_token_accuracy": 0.8338971585035324,
      "num_tokens": 2703348.0,
      "step": 61
    },
    {
      "entropy": 0.8301516175270081,
      "epoch": 0.4652908067542214,
      "grad_norm": 0.43768566846847534,
      "learning_rate": 0.00012849760564128207,
      "loss": 0.7932766675949097,
      "mean_token_accuracy": 0.8125782161951065,
      "num_tokens": 2750948.0,
      "step": 62
    },
    {
      "entropy": 0.7254582941532135,
      "epoch": 0.4727954971857411,
      "grad_norm": 0.5280841588973999,
      "learning_rate": 0.0001263473165881493,
      "loss": 0.6793215870857239,
      "mean_token_accuracy": 0.8378380239009857,
      "num_tokens": 2796457.0,
      "step": 63
    },
    {
      "entropy": 0.6487044766545296,
      "epoch": 0.4803001876172608,
      "grad_norm": 0.5871987342834473,
      "learning_rate": 0.00012418733259037762,
      "loss": 0.6333906054496765,
      "mean_token_accuracy": 0.8444840759038925,
      "num_tokens": 2843701.0,
      "step": 64
    },
    {
      "entropy": 0.6695453971624374,
      "epoch": 0.4878048780487805,
      "grad_norm": 0.4749222695827484,
      "learning_rate": 0.0001220189346487458,
      "loss": 0.6295377612113953,
      "mean_token_accuracy": 0.8430429995059967,
      "num_tokens": 2885585.0,
      "step": 65
    },
    {
      "entropy": 0.8910400867462158,
      "epoch": 0.49530956848030017,
      "grad_norm": 0.5251490473747253,
      "learning_rate": 0.00011984340875400872,
      "loss": 0.8795643448829651,
      "mean_token_accuracy": 0.7943810224533081,
      "num_tokens": 2930659.0,
      "step": 66
    },
    {
      "entropy": 0.7800526767969131,
      "epoch": 0.5028142589118199,
      "grad_norm": 0.6017885208129883,
      "learning_rate": 0.00011766204512422723,
      "loss": 0.7206172943115234,
      "mean_token_accuracy": 0.8300265371799469,
      "num_tokens": 2981674.0,
      "step": 67
    },
    {
      "entropy": 0.879515990614891,
      "epoch": 0.5103189493433395,
      "grad_norm": 0.6215891242027283,
      "learning_rate": 0.00011547613743959127,
      "loss": 0.8367047309875488,
      "mean_token_accuracy": 0.8011380583047867,
      "num_tokens": 3018919.0,
      "step": 68
    },
    {
      "entropy": 0.8410104364156723,
      "epoch": 0.5178236397748592,
      "grad_norm": 0.572844386100769,
      "learning_rate": 0.00011328698207518933,
      "loss": 0.8201464414596558,
      "mean_token_accuracy": 0.8056236356496811,
      "num_tokens": 3062912.0,
      "step": 69
    },
    {
      "entropy": 0.6954026967287064,
      "epoch": 0.525328330206379,
      "grad_norm": 0.7804688215255737,
      "learning_rate": 0.00011109587733217978,
      "loss": 0.6999291181564331,
      "mean_token_accuracy": 0.8319266140460968,
      "num_tokens": 3113330.0,
      "step": 70
    },
    {
      "entropy": 0.8273082971572876,
      "epoch": 0.5328330206378987,
      "grad_norm": 0.4803156554698944,
      "learning_rate": 0.00010890412266782024,
      "loss": 0.7979245185852051,
      "mean_token_accuracy": 0.8105114996433258,
      "num_tokens": 3162764.0,
      "step": 71
    },
    {
      "entropy": 0.7039017826318741,
      "epoch": 0.5403377110694184,
      "grad_norm": 0.6295526027679443,
      "learning_rate": 0.0001067130179248107,
      "loss": 0.6436035633087158,
      "mean_token_accuracy": 0.8462854474782944,
      "num_tokens": 3205071.0,
      "step": 72
    },
    {
      "entropy": 0.7428467124700546,
      "epoch": 0.5478424015009381,
      "grad_norm": 0.5865572094917297,
      "learning_rate": 0.00010452386256040876,
      "loss": 0.6983019113540649,
      "mean_token_accuracy": 0.8351266235113144,
      "num_tokens": 3251136.0,
      "step": 73
    },
    {
      "entropy": 0.7354733347892761,
      "epoch": 0.5553470919324578,
      "grad_norm": 0.4852830469608307,
      "learning_rate": 0.00010233795487577278,
      "loss": 0.7100962400436401,
      "mean_token_accuracy": 0.8305991142988205,
      "num_tokens": 3295114.0,
      "step": 74
    },
    {
      "entropy": 0.7458101883530617,
      "epoch": 0.5628517823639775,
      "grad_norm": 0.6101520657539368,
      "learning_rate": 0.00010015659124599133,
      "loss": 0.7333742380142212,
      "mean_token_accuracy": 0.8243828564882278,
      "num_tokens": 3344361.0,
      "step": 75
    },
    {
      "entropy": 0.7685392051935196,
      "epoch": 0.5703564727954972,
      "grad_norm": 0.728769063949585,
      "learning_rate": 9.79810653512542e-05,
      "loss": 0.7456347942352295,
      "mean_token_accuracy": 0.8238406628370285,
      "num_tokens": 3386423.0,
      "step": 76
    },
    {
      "entropy": 0.6607794463634491,
      "epoch": 0.5778611632270169,
      "grad_norm": 0.49787014722824097,
      "learning_rate": 9.58126674096224e-05,
      "loss": 0.5854030251502991,
      "mean_token_accuracy": 0.8605410158634186,
      "num_tokens": 3435485.0,
      "step": 77
    },
    {
      "entropy": 0.594479039311409,
      "epoch": 0.5853658536585366,
      "grad_norm": 0.5602482557296753,
      "learning_rate": 9.365268341185074e-05,
      "loss": 0.5215998291969299,
      "mean_token_accuracy": 0.8757767528295517,
      "num_tokens": 3484391.0,
      "step": 78
    },
    {
      "entropy": 0.9014105945825577,
      "epoch": 0.5928705440900562,
      "grad_norm": 0.5909271240234375,
      "learning_rate": 9.150239435871794e-05,
      "loss": 0.8536403179168701,
      "mean_token_accuracy": 0.798549085855484,
      "num_tokens": 3534914.0,
      "step": 79
    },
    {
      "entropy": 0.5203250125050545,
      "epoch": 0.600375234521576,
      "grad_norm": 0.4599478542804718,
      "learning_rate": 8.936307550131591e-05,
      "loss": 0.4753674268722534,
      "mean_token_accuracy": 0.8840101659297943,
      "num_tokens": 3580514.0,
      "step": 80
    },
    {
      "entropy": 0.6973117589950562,
      "epoch": 0.6078799249530957,
      "grad_norm": 0.6743919849395752,
      "learning_rate": 8.723599558474873e-05,
      "loss": 0.6727331280708313,
      "mean_token_accuracy": 0.8355155885219574,
      "num_tokens": 3621253.0,
      "step": 81
    },
    {
      "entropy": 0.5571995303034782,
      "epoch": 0.6153846153846154,
      "grad_norm": 0.46017736196517944,
      "learning_rate": 8.512241609569057e-05,
      "loss": 0.5408527255058289,
      "mean_token_accuracy": 0.873713806271553,
      "num_tokens": 3673208.0,
      "step": 82
    },
    {
      "entropy": 0.6034152880311012,
      "epoch": 0.6228893058161351,
      "grad_norm": 0.5736773014068604,
      "learning_rate": 8.302359051424785e-05,
      "loss": 0.5289053320884705,
      "mean_token_accuracy": 0.8750683069229126,
      "num_tokens": 3717242.0,
      "step": 83
    },
    {
      "entropy": 0.7445124536752701,
      "epoch": 0.6303939962476548,
      "grad_norm": 0.6038132905960083,
      "learning_rate": 8.094076357056987e-05,
      "loss": 0.6798694133758545,
      "mean_token_accuracy": 0.8354571461677551,
      "num_tokens": 3758932.0,
      "step": 84
    },
    {
      "entropy": 0.7136811465024948,
      "epoch": 0.6378986866791745,
      "grad_norm": 0.5773375034332275,
      "learning_rate": 7.887517050664898e-05,
      "loss": 0.6740403175354004,
      "mean_token_accuracy": 0.8374431282281876,
      "num_tokens": 3804722.0,
      "step": 85
    },
    {
      "entropy": 0.6839980036020279,
      "epoch": 0.6454033771106942,
      "grad_norm": 0.5740060806274414,
      "learning_rate": 7.682803634374745e-05,
      "loss": 0.6391246318817139,
      "mean_token_accuracy": 0.8439541906118393,
      "num_tokens": 3847775.0,
      "step": 86
    },
    {
      "entropy": 0.8194562047719955,
      "epoch": 0.6529080675422139,
      "grad_norm": 0.6135064959526062,
      "learning_rate": 7.480057515588599e-05,
      "loss": 0.799869179725647,
      "mean_token_accuracy": 0.8095971494913101,
      "num_tokens": 3891056.0,
      "step": 87
    },
    {
      "entropy": 0.6449038982391357,
      "epoch": 0.6604127579737336,
      "grad_norm": 0.5352879762649536,
      "learning_rate": 7.279398934982455e-05,
      "loss": 0.5862640738487244,
      "mean_token_accuracy": 0.8591591566801071,
      "num_tokens": 3936990.0,
      "step": 88
    },
    {
      "entropy": 0.7616558074951172,
      "epoch": 0.6679174484052532,
      "grad_norm": 0.5597410798072815,
      "learning_rate": 7.080946895196281e-05,
      "loss": 0.7417187094688416,
      "mean_token_accuracy": 0.8248282074928284,
      "num_tokens": 3987539.0,
      "step": 89
    },
    {
      "entropy": 0.7138193696737289,
      "epoch": 0.6754221388367729,
      "grad_norm": 0.6417974233627319,
      "learning_rate": 6.884819090258255e-05,
      "loss": 0.6560176610946655,
      "mean_token_accuracy": 0.8427004814147949,
      "num_tokens": 4033306.0,
      "step": 90
    },
    {
      "entropy": 0.6953269094228745,
      "epoch": 0.6829268292682927,
      "grad_norm": 0.5494018793106079,
      "learning_rate": 6.691131835785121e-05,
      "loss": 0.663178563117981,
      "mean_token_accuracy": 0.8422030210494995,
      "num_tokens": 4080129.0,
      "step": 91
    },
    {
      "entropy": 0.6544447913765907,
      "epoch": 0.6904315196998124,
      "grad_norm": 0.5673516988754272,
      "learning_rate": 6.500000000000002e-05,
      "loss": 0.6048790812492371,
      "mean_token_accuracy": 0.8534903079271317,
      "num_tokens": 4123651.0,
      "step": 92
    },
    {
      "entropy": 0.5819362550973892,
      "epoch": 0.6979362101313321,
      "grad_norm": 0.5087871551513672,
      "learning_rate": 6.311536935608628e-05,
      "loss": 0.5172831416130066,
      "mean_token_accuracy": 0.8726725280284882,
      "num_tokens": 4164004.0,
      "step": 93
    },
    {
      "entropy": 0.6534986197948456,
      "epoch": 0.7054409005628518,
      "grad_norm": 0.5322282314300537,
      "learning_rate": 6.125854412574339e-05,
      "loss": 0.6321709752082825,
      "mean_token_accuracy": 0.8517012149095535,
      "num_tokens": 4213716.0,
      "step": 94
    },
    {
      "entropy": 0.6258772611618042,
      "epoch": 0.7129455909943715,
      "grad_norm": 0.520404040813446,
      "learning_rate": 5.943062551831755e-05,
      "loss": 0.5752474069595337,
      "mean_token_accuracy": 0.8636364340782166,
      "num_tokens": 4261566.0,
      "step": 95
    },
    {
      "entropy": 0.5919274315237999,
      "epoch": 0.7204502814258912,
      "grad_norm": 0.6311873197555542,
      "learning_rate": 5.763269759978427e-05,
      "loss": 0.5671862959861755,
      "mean_token_accuracy": 0.8643747568130493,
      "num_tokens": 4305966.0,
      "step": 96
    },
    {
      "entropy": 0.6080443188548088,
      "epoch": 0.7279549718574109,
      "grad_norm": 0.5083109736442566,
      "learning_rate": 5.586582664983168e-05,
      "loss": 0.561866819858551,
      "mean_token_accuracy": 0.8662776947021484,
      "num_tokens": 4351780.0,
      "step": 97
    },
    {
      "entropy": 0.5952275395393372,
      "epoch": 0.7354596622889306,
      "grad_norm": 0.6353148221969604,
      "learning_rate": 5.413106052949264e-05,
      "loss": 0.5392923355102539,
      "mean_token_accuracy": 0.8713220357894897,
      "num_tokens": 4401689.0,
      "step": 98
    },
    {
      "entropy": 0.6480003744363785,
      "epoch": 0.7429643527204502,
      "grad_norm": 0.5440754294395447,
      "learning_rate": 5.242942805969982e-05,
      "loss": 0.6002803444862366,
      "mean_token_accuracy": 0.8569887280464172,
      "num_tokens": 4443532.0,
      "step": 99
    },
    {
      "entropy": 0.5362399220466614,
      "epoch": 0.7504690431519699,
      "grad_norm": 0.7220810055732727,
      "learning_rate": 5.0761938411132706e-05,
      "loss": 0.5190533399581909,
      "mean_token_accuracy": 0.8753415942192078,
      "num_tokens": 4489816.0,
      "step": 100
    },
    {
      "entropy": 0.6086771860718727,
      "epoch": 0.7579737335834896,
      "grad_norm": 0.6053838729858398,
      "learning_rate": 4.9129580505718754e-05,
      "loss": 0.5690251588821411,
      "mean_token_accuracy": 0.8591867834329605,
      "num_tokens": 4536503.0,
      "step": 101
    },
    {
      "entropy": 0.6956367194652557,
      "epoch": 0.7654784240150094,
      "grad_norm": 0.7699829936027527,
      "learning_rate": 4.753332243014295e-05,
      "loss": 0.6582237482070923,
      "mean_token_accuracy": 0.8379938304424286,
      "num_tokens": 4576669.0,
      "step": 102
    },
    {
      "entropy": 0.5546053946018219,
      "epoch": 0.7729831144465291,
      "grad_norm": 0.5455949306488037,
      "learning_rate": 4.597411086171402e-05,
      "loss": 0.4713042676448822,
      "mean_token_accuracy": 0.8811607509851456,
      "num_tokens": 4618405.0,
      "step": 103
    },
    {
      "entropy": 0.5784935429692268,
      "epoch": 0.7804878048780488,
      "grad_norm": 0.7412635087966919,
      "learning_rate": 4.4452870506927576e-05,
      "loss": 0.485187828540802,
      "mean_token_accuracy": 0.881045863032341,
      "num_tokens": 4659278.0,
      "step": 104
    },
    {
      "entropy": 0.6371504589915276,
      "epoch": 0.7879924953095685,
      "grad_norm": 0.6781969666481018,
      "learning_rate": 4.297050355305964e-05,
      "loss": 0.5490198731422424,
      "mean_token_accuracy": 0.867992639541626,
      "num_tokens": 4704241.0,
      "step": 105
    },
    {
      "entropy": 0.6112045347690582,
      "epoch": 0.7954971857410882,
      "grad_norm": 0.5682294964790344,
      "learning_rate": 4.152788913311498e-05,
      "loss": 0.5587624311447144,
      "mean_token_accuracy": 0.8684162348508835,
      "num_tokens": 4748939.0,
      "step": 106
    },
    {
      "entropy": 0.49986153095960617,
      "epoch": 0.8030018761726079,
      "grad_norm": 0.5927514433860779,
      "learning_rate": 4.0125882804448324e-05,
      "loss": 0.45605650544166565,
      "mean_token_accuracy": 0.8914675712585449,
      "num_tokens": 4790949.0,
      "step": 107
    },
    {
      "entropy": 0.5914800465106964,
      "epoch": 0.8105065666041276,
      "grad_norm": 0.5249906182289124,
      "learning_rate": 3.876531604136714e-05,
      "loss": 0.5487126708030701,
      "mean_token_accuracy": 0.8693995326757431,
      "num_tokens": 4838082.0,
      "step": 108
    },
    {
      "entropy": 0.5456953942775726,
      "epoch": 0.8180112570356473,
      "grad_norm": 0.5577149391174316,
      "learning_rate": 3.744699574201736e-05,
      "loss": 0.523871898651123,
      "mean_token_accuracy": 0.8754665851593018,
      "num_tokens": 4890292.0,
      "step": 109
    },
    {
      "entropy": 0.584635928273201,
      "epoch": 0.8255159474671669,
      "grad_norm": 0.5176509618759155,
      "learning_rate": 3.617170374984402e-05,
      "loss": 0.5098640322685242,
      "mean_token_accuracy": 0.8694778680801392,
      "num_tokens": 4937857.0,
      "step": 110
    },
    {
      "entropy": 0.5596759244799614,
      "epoch": 0.8330206378986866,
      "grad_norm": 0.4887101948261261,
      "learning_rate": 3.4940196389911e-05,
      "loss": 0.5138705968856812,
      "mean_token_accuracy": 0.8753239661455154,
      "num_tokens": 4983594.0,
      "step": 111
    },
    {
      "entropy": 0.5349718183279037,
      "epoch": 0.8405253283302064,
      "grad_norm": 0.5526368021965027,
      "learning_rate": 3.375320402035458e-05,
      "loss": 0.48363634943962097,
      "mean_token_accuracy": 0.8848558515310287,
      "num_tokens": 5026617.0,
      "step": 112
    },
    {
      "entropy": 0.621735468506813,
      "epoch": 0.8480300187617261,
      "grad_norm": 0.6342886090278625,
      "learning_rate": 3.261143059923709e-05,
      "loss": 0.5607931017875671,
      "mean_token_accuracy": 0.8654929846525192,
      "num_tokens": 5069910.0,
      "step": 113
    },
    {
      "entropy": 0.4928681254386902,
      "epoch": 0.8555347091932458,
      "grad_norm": 0.5040988326072693,
      "learning_rate": 3.151555326705731e-05,
      "loss": 0.425799697637558,
      "mean_token_accuracy": 0.8913719058036804,
      "num_tokens": 5112597.0,
      "step": 114
    },
    {
      "entropy": 0.5078200325369835,
      "epoch": 0.8630393996247655,
      "grad_norm": 0.4822465479373932,
      "learning_rate": 3.046622194516529e-05,
      "loss": 0.4364722669124603,
      "mean_token_accuracy": 0.8885586261749268,
      "num_tokens": 5158641.0,
      "step": 115
    },
    {
      "entropy": 0.48565153032541275,
      "epoch": 0.8705440900562852,
      "grad_norm": 0.4889708459377289,
      "learning_rate": 2.9464058950319738e-05,
      "loss": 0.4391924738883972,
      "mean_token_accuracy": 0.8953022360801697,
      "num_tokens": 5207131.0,
      "step": 116
    },
    {
      "entropy": 0.6985752284526825,
      "epoch": 0.8780487804878049,
      "grad_norm": 0.6654022932052612,
      "learning_rate": 2.8509658625616808e-05,
      "loss": 0.6480734944343567,
      "mean_token_accuracy": 0.8425026684999466,
      "num_tokens": 5249784.0,
      "step": 117
    },
    {
      "entropy": 0.5020286291837692,
      "epoch": 0.8855534709193246,
      "grad_norm": 0.4819194972515106,
      "learning_rate": 2.7603586988008672e-05,
      "loss": 0.4483570456504822,
      "mean_token_accuracy": 0.8907182961702347,
      "num_tokens": 5293867.0,
      "step": 118
    },
    {
      "entropy": 0.5373758226633072,
      "epoch": 0.8930581613508443,
      "grad_norm": 0.533804178237915,
      "learning_rate": 2.674638139262133e-05,
      "loss": 0.4949713349342346,
      "mean_token_accuracy": 0.8807879090309143,
      "num_tokens": 5335271.0,
      "step": 119
    },
    {
      "entropy": 0.6371819153428078,
      "epoch": 0.900562851782364,
      "grad_norm": 0.6009949445724487,
      "learning_rate": 2.593855021407069e-05,
      "loss": 0.5801702737808228,
      "mean_token_accuracy": 0.8532385528087616,
      "num_tokens": 5378357.0,
      "step": 120
    },
    {
      "entropy": 0.6247220039367676,
      "epoch": 0.9080675422138836,
      "grad_norm": 0.6049973368644714,
      "learning_rate": 2.5180572544965702e-05,
      "loss": 0.5846748948097229,
      "mean_token_accuracy": 0.8602764010429382,
      "num_tokens": 5419093.0,
      "step": 121
    },
    {
      "entropy": 0.5582565367221832,
      "epoch": 0.9155722326454033,
      "grad_norm": 0.5421613454818726,
      "learning_rate": 2.4472897911777584e-05,
      "loss": 0.5074694752693176,
      "mean_token_accuracy": 0.8764741271734238,
      "num_tokens": 5468026.0,
      "step": 122
    },
    {
      "entropy": 0.5597152858972549,
      "epoch": 0.9230769230769231,
      "grad_norm": 0.5344402194023132,
      "learning_rate": 2.3815946008243476e-05,
      "loss": 0.530523955821991,
      "mean_token_accuracy": 0.8788765370845795,
      "num_tokens": 5513888.0,
      "step": 123
    },
    {
      "entropy": 0.6477740779519081,
      "epoch": 0.9305816135084428,
      "grad_norm": 0.6750009059906006,
      "learning_rate": 2.321010644646273e-05,
      "loss": 0.6173692941665649,
      "mean_token_accuracy": 0.8617543280124664,
      "num_tokens": 5554946.0,
      "step": 124
    },
    {
      "entropy": 0.4598197713494301,
      "epoch": 0.9380863039399625,
      "grad_norm": 0.5382166504859924,
      "learning_rate": 2.265573852583353e-05,
      "loss": 0.41298604011535645,
      "mean_token_accuracy": 0.900470107793808,
      "num_tokens": 5602571.0,
      "step": 125
    },
    {
      "entropy": 0.5343818813562393,
      "epoch": 0.9455909943714822,
      "grad_norm": 0.5444672703742981,
      "learning_rate": 2.2153171019966566e-05,
      "loss": 0.49157363176345825,
      "mean_token_accuracy": 0.8833711892366409,
      "num_tokens": 5643872.0,
      "step": 126
    },
    {
      "entropy": 0.5449606254696846,
      "epoch": 0.9530956848030019,
      "grad_norm": 0.5583885312080383,
      "learning_rate": 2.1702701981702516e-05,
      "loss": 0.48061904311180115,
      "mean_token_accuracy": 0.8823899626731873,
      "num_tokens": 5687028.0,
      "step": 127
    },
    {
      "entropy": 0.52015720307827,
      "epoch": 0.9606003752345216,
      "grad_norm": 0.5448853969573975,
      "learning_rate": 2.1304598566348757e-05,
      "loss": 0.47128140926361084,
      "mean_token_accuracy": 0.8836861997842789,
      "num_tokens": 5730812.0,
      "step": 128
    },
    {
      "entropy": 0.5044189319014549,
      "epoch": 0.9681050656660413,
      "grad_norm": 0.6111878752708435,
      "learning_rate": 2.0959096873240082e-05,
      "loss": 0.44628357887268066,
      "mean_token_accuracy": 0.8896998018026352,
      "num_tokens": 5773131.0,
      "step": 129
    },
    {
      "entropy": 0.5925156995654106,
      "epoch": 0.975609756097561,
      "grad_norm": 0.7184950113296509,
      "learning_rate": 2.0666401805717626e-05,
      "loss": 0.5486721396446228,
      "mean_token_accuracy": 0.8632885962724686,
      "num_tokens": 5813935.0,
      "step": 130
    },
    {
      "entropy": 0.5575186833739281,
      "epoch": 0.9831144465290806,
      "grad_norm": 0.5785585641860962,
      "learning_rate": 2.042668694960878e-05,
      "loss": 0.5043210983276367,
      "mean_token_accuracy": 0.879310816526413,
      "num_tokens": 5855219.0,
      "step": 131
    },
    {
      "entropy": 0.5164650678634644,
      "epoch": 0.9906191369606003,
      "grad_norm": 0.8453945517539978,
      "learning_rate": 2.0240094470280356e-05,
      "loss": 0.45579153299331665,
      "mean_token_accuracy": 0.8882085084915161,
      "num_tokens": 5898072.0,
      "step": 132
    },
    {
      "entropy": 0.47110557556152344,
      "epoch": 0.99812382739212,
      "grad_norm": 0.5557724237442017,
      "learning_rate": 2.010673502832593e-05,
      "loss": 0.41308286786079407,
      "mean_token_accuracy": 0.9016138017177582,
      "num_tokens": 5938844.0,
      "step": 133
    },
    {
      "entropy": 0.7972898483276367,
      "epoch": 1.0,
      "grad_norm": 2.5536210536956787,
      "learning_rate": 2.0026687713937458e-05,
      "loss": 0.7482244968414307,
      "mean_token_accuracy": 0.8238636255264282,
      "num_tokens": 5940957.0,
      "step": 134
    }
  ],
  "logging_steps": 1,
  "max_steps": 134,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.621709527537919e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}