{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 1660,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.060350030175015085,
      "grad_norm": 0.2168818861246109,
      "learning_rate": 8.18181818181818e-05,
      "loss": 1.8651,
      "mean_token_accuracy": 0.614892452955246,
      "num_tokens": 157222.0,
      "step": 25
    },
    {
      "epoch": 0.12070006035003017,
      "grad_norm": 0.3058246970176697,
      "learning_rate": 0.00016704545454545452,
      "loss": 1.075,
      "mean_token_accuracy": 0.7393987023830414,
      "num_tokens": 285335.0,
      "step": 50
    },
    {
      "epoch": 0.18105009052504525,
      "grad_norm": 0.21806401014328003,
      "learning_rate": 0.0002522727272727273,
      "loss": 0.6678,
      "mean_token_accuracy": 0.8195570611953735,
      "num_tokens": 443748.0,
      "step": 75
    },
    {
      "epoch": 0.24140012070006034,
      "grad_norm": 0.42833107709884644,
      "learning_rate": 0.0002999887132933212,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8508000493049621,
      "num_tokens": 570813.0,
      "step": 100
    },
    {
      "epoch": 0.30175015087507545,
      "grad_norm": 0.2923378050327301,
      "learning_rate": 0.0002998791256978121,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8748408752679825,
      "num_tokens": 727818.0,
      "step": 125
    },
    {
      "epoch": 0.3621001810500905,
      "grad_norm": 0.5519189238548279,
      "learning_rate": 0.0002996530399366737,
      "loss": 0.353,
      "mean_token_accuracy": 0.9003111439943313,
      "num_tokens": 852739.0,
      "step": 150
    },
    {
      "epoch": 0.4224502112251056,
      "grad_norm": 0.2830982804298401,
      "learning_rate": 0.00029931063174202567,
      "loss": 0.2802,
      "mean_token_accuracy": 0.9183641839027404,
      "num_tokens": 1009253.0,
      "step": 175
    },
    {
      "epoch": 0.4828002414001207,
      "grad_norm": 0.3710537254810333,
      "learning_rate": 0.00029885216726118104,
      "loss": 0.2359,
      "mean_token_accuracy": 0.9340034741163253,
      "num_tokens": 1135711.0,
      "step": 200
    },
    {
      "epoch": 0.5431502715751357,
      "grad_norm": 0.3129195272922516,
      "learning_rate": 0.00029827800284977474,
      "loss": 0.1936,
      "mean_token_accuracy": 0.9437149178981781,
      "num_tokens": 1294974.0,
      "step": 225
    },
    {
      "epoch": 0.6035003017501509,
      "grad_norm": 0.3972657322883606,
      "learning_rate": 0.00029758858479477575,
      "loss": 0.1992,
      "mean_token_accuracy": 0.9438802194595337,
      "num_tokens": 1422454.0,
      "step": 250
    },
    {
      "epoch": 0.663850331925166,
      "grad_norm": 0.3286677300930023,
      "learning_rate": 0.0002967844489675963,
      "loss": 0.1467,
      "mean_token_accuracy": 0.9580748552083969,
      "num_tokens": 1578961.0,
      "step": 275
    },
    {
      "epoch": 0.724200362100181,
      "grad_norm": 0.3204401135444641,
      "learning_rate": 0.00029586622040756957,
      "loss": 0.1492,
      "mean_token_accuracy": 0.9577940207719803,
      "num_tokens": 1704389.0,
      "step": 300
    },
    {
      "epoch": 0.7845503922751962,
      "grad_norm": 0.18775713443756104,
      "learning_rate": 0.0002948346128361186,
      "loss": 0.1163,
      "mean_token_accuracy": 0.9667265516519546,
      "num_tokens": 1863333.0,
      "step": 325
    },
    {
      "epoch": 0.8449004224502112,
      "grad_norm": 0.2705429494380951,
      "learning_rate": 0.00029369042810199416,
      "loss": 0.1007,
      "mean_token_accuracy": 0.9721928060054779,
      "num_tokens": 1990936.0,
      "step": 350
    },
    {
      "epoch": 0.9052504526252263,
      "grad_norm": 0.19949601590633392,
      "learning_rate": 0.0002924345555580135,
      "loss": 0.0936,
      "mean_token_accuracy": 0.9747320550680161,
      "num_tokens": 2147408.0,
      "step": 375
    },
    {
      "epoch": 0.9656004828002414,
      "grad_norm": 0.30361101031303406,
      "learning_rate": 0.000291067971369783,
      "loss": 0.0975,
      "mean_token_accuracy": 0.972884624004364,
      "num_tokens": 2273014.0,
      "step": 400
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.08620278537273407,
      "eval_mean_token_accuracy": 0.9761469618694202,
      "eval_num_tokens": 2354180.0,
      "eval_runtime": 61.6733,
      "eval_samples_per_second": 5.983,
      "eval_steps_per_second": 3.0,
      "step": 415
    },
    {
      "epoch": 1.024140012070006,
      "grad_norm": 0.09305181354284286,
      "learning_rate": 0.0002895917377569438,
      "loss": 0.0907,
      "mean_token_accuracy": 0.974765161878055,
      "num_tokens": 2423410.0,
      "step": 425
    },
    {
      "epoch": 1.0844900422450212,
      "grad_norm": 0.3788456916809082,
      "learning_rate": 0.00028800700216752875,
      "loss": 0.0593,
      "mean_token_accuracy": 0.9832174813747406,
      "num_tokens": 2566972.0,
      "step": 450
    },
    {
      "epoch": 1.1448400724200363,
      "grad_norm": 0.14668315649032593,
      "learning_rate": 0.00028631499638607285,
      "loss": 0.0762,
      "mean_token_accuracy": 0.9784989726543426,
      "num_tokens": 2708394.0,
      "step": 475
    },
    {
      "epoch": 1.2051901025950513,
      "grad_norm": 0.10731452703475952,
      "learning_rate": 0.0002845170355761712,
      "loss": 0.0532,
      "mean_token_accuracy": 0.9849696922302246,
      "num_tokens": 2850550.0,
      "step": 500
    },
    {
      "epoch": 1.2655401327700664,
      "grad_norm": 0.12523867189884186,
      "learning_rate": 0.0002826145172582274,
      "loss": 0.0664,
      "mean_token_accuracy": 0.9817911142110824,
      "num_tokens": 2991599.0,
      "step": 525
    },
    {
      "epoch": 1.3258901629450814,
      "grad_norm": 0.12591515481472015,
      "learning_rate": 0.00028060892022318764,
      "loss": 0.0483,
      "mean_token_accuracy": 0.9862601357698441,
      "num_tokens": 3132335.0,
      "step": 550
    },
    {
      "epoch": 1.3862401931200965,
      "grad_norm": 0.1305922567844391,
      "learning_rate": 0.0002785018033831051,
      "loss": 0.0681,
      "mean_token_accuracy": 0.9813536697626114,
      "num_tokens": 3272427.0,
      "step": 575
    },
    {
      "epoch": 1.4465902232951118,
      "grad_norm": 0.15813934803009033,
      "learning_rate": 0.0002762948045594276,
      "loss": 0.0493,
      "mean_token_accuracy": 0.9858457553386688,
      "num_tokens": 3413641.0,
      "step": 600
    },
    {
      "epoch": 1.5069402534701268,
      "grad_norm": 0.13504762947559357,
      "learning_rate": 0.0002739896392099502,
      "loss": 0.061,
      "mean_token_accuracy": 0.98312191426754,
      "num_tokens": 3553593.0,
      "step": 625
    },
    {
      "epoch": 1.567290283645142,
      "grad_norm": 0.11668980866670609,
      "learning_rate": 0.00027158809909542307,
      "loss": 0.0452,
      "mean_token_accuracy": 0.9869613242149353,
      "num_tokens": 3696166.0,
      "step": 650
    },
    {
      "epoch": 1.627640313820157,
      "grad_norm": 0.1464385837316513,
      "learning_rate": 0.00026909205088685,
      "loss": 0.0613,
      "mean_token_accuracy": 0.9831733548641205,
      "num_tokens": 3839563.0,
      "step": 675
    },
    {
      "epoch": 1.687990343995172,
      "grad_norm": 0.1125669926404953,
      "learning_rate": 0.0002665034347145612,
      "loss": 0.0435,
      "mean_token_accuracy": 0.9872170311212539,
      "num_tokens": 3984309.0,
      "step": 700
    },
    {
      "epoch": 1.748340374170187,
      "grad_norm": 0.10721296817064285,
      "learning_rate": 0.000263824262660187,
      "loss": 0.0622,
      "mean_token_accuracy": 0.9823342782258987,
      "num_tokens": 4127163.0,
      "step": 725
    },
    {
      "epoch": 1.8086904043452021,
      "grad_norm": 0.09685028344392776,
      "learning_rate": 0.0002610566171927056,
      "loss": 0.043,
      "mean_token_accuracy": 0.9875605070590973,
      "num_tokens": 4270459.0,
      "step": 750
    },
    {
      "epoch": 1.8690404345202172,
      "grad_norm": 0.09882082790136337,
      "learning_rate": 0.00025820264954977976,
      "loss": 0.056,
      "mean_token_accuracy": 0.984576758146286,
      "num_tokens": 4412568.0,
      "step": 775
    },
    {
      "epoch": 1.9293904646952322,
      "grad_norm": 0.07612240314483643,
      "learning_rate": 0.00025526457806564136,
      "loss": 0.041,
      "mean_token_accuracy": 0.9879619979858398,
      "num_tokens": 4553943.0,
      "step": 800
    },
    {
      "epoch": 1.9897404948702473,
      "grad_norm": 0.10256911814212799,
      "learning_rate": 0.00025224468644682245,
      "loss": 0.0496,
      "mean_token_accuracy": 0.9865266931056976,
      "num_tokens": 4687925.0,
      "step": 825
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.052460573613643646,
      "eval_mean_token_accuracy": 0.9856111246186334,
      "eval_num_tokens": 4708360.0,
      "eval_runtime": 61.6647,
      "eval_samples_per_second": 5.984,
      "eval_steps_per_second": 3.0,
      "step": 830
    },
    {
      "epoch": 2.048280024140012,
      "grad_norm": 0.05109108239412308,
      "learning_rate": 0.00024914532199707444,
      "loss": 0.0471,
      "mean_token_accuracy": 0.9859890218862554,
      "num_tokens": 4837792.0,
      "step": 850
    },
    {
      "epoch": 2.1086300543150274,
      "grad_norm": 0.12332670390605927,
      "learning_rate": 0.00024596889379285353,
      "loss": 0.0331,
      "mean_token_accuracy": 0.9905285978317261,
      "num_tokens": 4971259.0,
      "step": 875
    },
    {
      "epoch": 2.1689800844900424,
      "grad_norm": 0.0654679387807846,
      "learning_rate": 0.00024271787081079228,
      "loss": 0.0486,
      "mean_token_accuracy": 0.9860249239206315,
      "num_tokens": 5121366.0,
      "step": 900
    },
    {
      "epoch": 2.2293301146650575,
      "grad_norm": 0.08594338595867157,
      "learning_rate": 0.00023939478000861117,
      "loss": 0.031,
      "mean_token_accuracy": 0.9908674705028534,
      "num_tokens": 5253015.0,
      "step": 925
    },
    {
      "epoch": 2.2896801448400725,
      "grad_norm": 0.08495251834392548,
      "learning_rate": 0.00023600220436096318,
      "loss": 0.0424,
      "mean_token_accuracy": 0.9873432791233063,
      "num_tokens": 5403783.0,
      "step": 950
    },
    {
      "epoch": 2.3500301750150876,
      "grad_norm": 0.11103720963001251,
      "learning_rate": 0.00023254278085173684,
      "loss": 0.0294,
      "mean_token_accuracy": 0.9907593816518784,
      "num_tokens": 5536898.0,
      "step": 975
    },
    {
      "epoch": 2.4103802051901027,
      "grad_norm": 0.12526410818099976,
      "learning_rate": 0.00022901919842437972,
      "loss": 0.0508,
      "mean_token_accuracy": 0.9854990404844284,
      "num_tokens": 5690565.0,
      "step": 1000
    },
    {
      "epoch": 2.4707302353651177,
      "grad_norm": 0.1827692687511444,
      "learning_rate": 0.00022543419589183397,
      "loss": 0.0308,
      "mean_token_accuracy": 0.9908353638648987,
      "num_tokens": 5822868.0,
      "step": 1025
    },
    {
      "epoch": 2.5310802655401328,
      "grad_norm": 0.08467936515808105,
      "learning_rate": 0.00022179055980770993,
      "loss": 0.0447,
      "mean_token_accuracy": 0.9870854699611664,
      "num_tokens": 5974120.0,
      "step": 1050
    },
    {
      "epoch": 2.591430295715148,
      "grad_norm": 0.08671411126852036,
      "learning_rate": 0.0002180911223003513,
      "loss": 0.0336,
      "mean_token_accuracy": 0.9898967409133911,
      "num_tokens": 6106399.0,
      "step": 1075
    },
    {
      "epoch": 2.651780325890163,
      "grad_norm": 0.10606463998556137,
      "learning_rate": 0.00021433875887147627,
      "loss": 0.0365,
      "mean_token_accuracy": 0.9888207441568375,
      "num_tokens": 6256139.0,
      "step": 1100
    },
    {
      "epoch": 2.712130356065178,
      "grad_norm": 0.08840422332286835,
      "learning_rate": 0.00021053638616110525,
      "loss": 0.0295,
      "mean_token_accuracy": 0.9906065487861633,
      "num_tokens": 6388996.0,
      "step": 1125
    },
    {
      "epoch": 2.772480386240193,
      "grad_norm": 0.1253369003534317,
      "learning_rate": 0.00020668695968051274,
      "loss": 0.0415,
      "mean_token_accuracy": 0.9876212304830552,
      "num_tokens": 6538751.0,
      "step": 1150
    },
    {
      "epoch": 2.832830416415208,
      "grad_norm": 0.10428694635629654,
      "learning_rate": 0.00020279347151496482,
      "loss": 0.0285,
      "mean_token_accuracy": 0.9914796513319015,
      "num_tokens": 6671509.0,
      "step": 1175
    },
    {
      "epoch": 2.8931804465902236,
      "grad_norm": 0.09738153964281082,
      "learning_rate": 0.00019885894799802922,
      "loss": 0.0441,
      "mean_token_accuracy": 0.986878205537796,
      "num_tokens": 6824932.0,
      "step": 1200
    },
    {
      "epoch": 2.9535304767652386,
      "grad_norm": 0.043908387422561646,
      "learning_rate": 0.00019488644735926396,
      "loss": 0.0299,
      "mean_token_accuracy": 0.9910147470235825,
      "num_tokens": 6959350.0,
      "step": 1225
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.04450182244181633,
      "eval_mean_token_accuracy": 0.9876940347052909,
      "eval_num_tokens": 7062540.0,
      "eval_runtime": 61.71,
      "eval_samples_per_second": 5.98,
      "eval_steps_per_second": 2.998,
      "step": 1245
    },
    {
      "epoch": 3.012070006035003,
      "grad_norm": 0.11456421762704849,
      "learning_rate": 0.00019087905734711452,
      "loss": 0.0441,
      "mean_token_accuracy": 0.9872288673194414,
      "num_tokens": 7099425.0,
      "step": 1250
    },
    {
      "epoch": 3.0724200362100182,
      "grad_norm": 0.12781904637813568,
      "learning_rate": 0.00018683989282886613,
      "loss": 0.0283,
      "mean_token_accuracy": 0.9909683948755265,
      "num_tokens": 7249165.0,
      "step": 1275
    },
    {
      "epoch": 3.1327700663850333,
      "grad_norm": 0.10656405240297318,
      "learning_rate": 0.0001827720933695173,
      "loss": 0.0328,
      "mean_token_accuracy": 0.9903161740303039,
      "num_tokens": 7386640.0,
      "step": 1300
    },
    {
      "epoch": 3.1931200965600484,
      "grad_norm": 0.04701218381524086,
      "learning_rate": 0.00017867882079145627,
      "loss": 0.0268,
      "mean_token_accuracy": 0.9914222890138626,
      "num_tokens": 7535281.0,
      "step": 1325
    },
    {
      "epoch": 3.2534701267350634,
      "grad_norm": 0.10875603556632996,
      "learning_rate": 0.00017456325671683724,
      "loss": 0.0297,
      "mean_token_accuracy": 0.9906892210245133,
      "num_tokens": 7672221.0,
      "step": 1350
    },
    {
      "epoch": 3.3138201569100785,
      "grad_norm": 0.09782923758029938,
      "learning_rate": 0.00017042860009456638,
      "loss": 0.0264,
      "mean_token_accuracy": 0.991526146531105,
      "num_tokens": 7821001.0,
      "step": 1375
    },
    {
      "epoch": 3.3741701870850935,
      "grad_norm": 0.05467990040779114,
      "learning_rate": 0.00016627806471382066,
      "loss": 0.0266,
      "mean_token_accuracy": 0.9914808225631714,
      "num_tokens": 7955346.0,
      "step": 1400
    },
    {
      "epoch": 3.4345202172601086,
      "grad_norm": 0.057833656668663025,
      "learning_rate": 0.00016211487670603078,
      "loss": 0.0271,
      "mean_token_accuracy": 0.9915524727106094,
      "num_tokens": 8102107.0,
      "step": 1425
    },
    {
      "epoch": 3.4948702474351236,
      "grad_norm": 0.1365857869386673,
      "learning_rate": 0.0001579422720372715,
      "loss": 0.0296,
      "mean_token_accuracy": 0.9906032651662826,
      "num_tokens": 8237816.0,
      "step": 1450
    },
    {
      "epoch": 3.5552202776101387,
      "grad_norm": 0.04984680563211441,
      "learning_rate": 0.00015376349399300745,
      "loss": 0.0268,
      "mean_token_accuracy": 0.9916295224428177,
      "num_tokens": 8386557.0,
      "step": 1475
    },
    {
      "epoch": 3.6155703077851538,
      "grad_norm": 0.08074755221605301,
      "learning_rate": 0.0001495817906571492,
      "loss": 0.0302,
      "mean_token_accuracy": 0.9905178165435791,
      "num_tokens": 8522898.0,
      "step": 1500
    },
    {
      "epoch": 3.675920337960169,
      "grad_norm": 0.03389836475253105,
      "learning_rate": 0.00014540041238738055,
      "loss": 0.0256,
      "mean_token_accuracy": 0.9916897600889206,
      "num_tokens": 8671388.0,
      "step": 1525
    },
    {
      "epoch": 3.736270368135184,
      "grad_norm": 0.06772787123918533,
      "learning_rate": 0.00014122260928871734,
      "loss": 0.0307,
      "mean_token_accuracy": 0.9902506107091904,
      "num_tokens": 8807614.0,
      "step": 1550
    },
    {
      "epoch": 3.796620398310199,
      "grad_norm": 0.04803523048758507,
      "learning_rate": 0.00013705162868726396,
      "loss": 0.0252,
      "mean_token_accuracy": 0.9921325802803039,
      "num_tokens": 8953964.0,
      "step": 1575
    },
    {
      "epoch": 3.856970428485214,
      "grad_norm": 0.0687125101685524,
      "learning_rate": 0.00013289071260612855,
      "loss": 0.0284,
      "mean_token_accuracy": 0.9912543642520905,
      "num_tokens": 9088576.0,
      "step": 1600
    },
    {
      "epoch": 3.9173204586602295,
      "grad_norm": 0.07590165734291077,
      "learning_rate": 0.00012874309524546083,
      "loss": 0.0262,
      "mean_token_accuracy": 0.9915939372777939,
      "num_tokens": 9237259.0,
      "step": 1625
    },
    {
      "epoch": 3.9776704888352445,
      "grad_norm": 0.06971056014299393,
      "learning_rate": 0.00012461200046857084,
      "loss": 0.0252,
      "mean_token_accuracy": 0.9921457231044769,
      "num_tokens": 9369029.0,
      "step": 1650
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.04080882668495178,
      "eval_mean_token_accuracy": 0.9887693820772945,
      "eval_num_tokens": 9416720.0,
      "eval_runtime": 61.7793,
      "eval_samples_per_second": 5.973,
      "eval_steps_per_second": 2.995,
      "step": 1660
    }
  ],
  "logging_steps": 25,
  "max_steps": 2905,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.045453549957048e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}