{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 29709,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.0913017302751542,
      "epoch": 0.003029614481557222,
      "grad_norm": 1.1931798458099365,
      "learning_rate": 3.634894991922456e-07,
      "loss": 1.7418,
      "mean_token_accuracy": 0.7027995571494102,
      "num_tokens": 676413.0,
      "step": 10
    },
    {
      "entropy": 1.0921823412179947,
      "epoch": 0.006059228963114444,
      "grad_norm": 1.1370843648910522,
      "learning_rate": 7.673667205169629e-07,
      "loss": 1.7408,
      "mean_token_accuracy": 0.7032344728708267,
      "num_tokens": 1360739.0,
      "step": 20
    },
    {
      "entropy": 1.0819210216403008,
      "epoch": 0.009088843444671665,
      "grad_norm": 0.9659048914909363,
      "learning_rate": 1.17124394184168e-06,
      "loss": 1.7275,
      "mean_token_accuracy": 0.7082073956727981,
      "num_tokens": 2039094.0,
      "step": 30
    },
    {
      "entropy": 1.0838943645358086,
      "epoch": 0.012118457926228888,
      "grad_norm": 1.0033633708953857,
      "learning_rate": 1.5751211631663977e-06,
      "loss": 1.6918,
      "mean_token_accuracy": 0.709006880223751,
      "num_tokens": 2724293.0,
      "step": 40
    },
    {
      "entropy": 1.0841404393315315,
      "epoch": 0.015148072407786109,
      "grad_norm": 0.976797342300415,
      "learning_rate": 1.978998384491115e-06,
      "loss": 1.628,
      "mean_token_accuracy": 0.7091029062867165,
      "num_tokens": 3410092.0,
      "step": 50
    },
    {
      "entropy": 1.0670536786317826,
      "epoch": 0.01817768688934333,
      "grad_norm": 0.8120208978652954,
      "learning_rate": 2.382875605815832e-06,
      "loss": 1.5408,
      "mean_token_accuracy": 0.7167265444993973,
      "num_tokens": 4097508.0,
      "step": 60
    },
    {
      "entropy": 1.0734323486685753,
      "epoch": 0.02120730137090055,
      "grad_norm": 0.4978794753551483,
      "learning_rate": 2.7867528271405497e-06,
      "loss": 1.4942,
      "mean_token_accuracy": 0.7221754133701325,
      "num_tokens": 4763424.0,
      "step": 70
    },
    {
      "entropy": 1.0327554315328598,
      "epoch": 0.024236915852457776,
      "grad_norm": 0.3606416583061218,
      "learning_rate": 3.1906300484652665e-06,
      "loss": 1.4376,
      "mean_token_accuracy": 0.7297119319438934,
      "num_tokens": 5435268.0,
      "step": 80
    },
    {
      "entropy": 1.024299743771553,
      "epoch": 0.027266530334014997,
      "grad_norm": 0.3630509674549103,
      "learning_rate": 3.5945072697899836e-06,
      "loss": 1.4181,
      "mean_token_accuracy": 0.7321271970868111,
      "num_tokens": 6107570.0,
      "step": 90
    },
    {
      "entropy": 1.0179613307118416,
      "epoch": 0.030296144815572218,
      "grad_norm": 0.31415602564811707,
      "learning_rate": 3.998384491114701e-06,
      "loss": 1.4065,
      "mean_token_accuracy": 0.7293723702430726,
      "num_tokens": 6767478.0,
      "step": 100
    },
    {
      "entropy": 0.9994221851229668,
      "epoch": 0.03332575929712944,
      "grad_norm": 0.34493952989578247,
      "learning_rate": 4.402261712439419e-06,
      "loss": 1.3978,
      "mean_token_accuracy": 0.7380139917135239,
      "num_tokens": 7444132.0,
      "step": 110
    },
    {
      "entropy": 0.9911895662546157,
      "epoch": 0.03635537377868666,
      "grad_norm": 0.33052828907966614,
      "learning_rate": 4.806138933764136e-06,
      "loss": 1.3869,
      "mean_token_accuracy": 0.7341301307082176,
      "num_tokens": 8101986.0,
      "step": 120
    },
    {
      "entropy": 0.9866934418678284,
      "epoch": 0.039384988260243885,
      "grad_norm": 0.4284345805644989,
      "learning_rate": 5.210016155088853e-06,
      "loss": 1.3799,
      "mean_token_accuracy": 0.7413432449102402,
      "num_tokens": 8777186.0,
      "step": 130
    },
    {
      "entropy": 0.9676428496837616,
      "epoch": 0.0424146027418011,
      "grad_norm": 0.35278987884521484,
      "learning_rate": 5.61389337641357e-06,
      "loss": 1.3681,
      "mean_token_accuracy": 0.7491588592529297,
      "num_tokens": 9463550.0,
      "step": 140
    },
    {
      "entropy": 0.9714192464947701,
      "epoch": 0.04544421722335833,
      "grad_norm": 0.4610103666782379,
      "learning_rate": 6.017770597738288e-06,
      "loss": 1.3639,
      "mean_token_accuracy": 0.7401145175099373,
      "num_tokens": 10130428.0,
      "step": 150
    },
    {
      "entropy": 0.95505460947752,
      "epoch": 0.04847383170491555,
      "grad_norm": 0.3843373656272888,
      "learning_rate": 6.421647819063005e-06,
      "loss": 1.3449,
      "mean_token_accuracy": 0.7504446610808373,
      "num_tokens": 10828958.0,
      "step": 160
    },
    {
      "entropy": 0.9737901806831359,
      "epoch": 0.05150344618647277,
      "grad_norm": 0.5503286123275757,
      "learning_rate": 6.825525040387723e-06,
      "loss": 1.3669,
      "mean_token_accuracy": 0.7403925940394401,
      "num_tokens": 11508760.0,
      "step": 170
    },
    {
      "entropy": 0.9536378592252731,
      "epoch": 0.054533060668029994,
      "grad_norm": 0.5051810145378113,
      "learning_rate": 7.22940226171244e-06,
      "loss": 1.3484,
      "mean_token_accuracy": 0.742230424284935,
      "num_tokens": 12165955.0,
      "step": 180
    },
    {
      "entropy": 0.9639706507325172,
      "epoch": 0.05756267514958722,
      "grad_norm": 0.40541717410087585,
      "learning_rate": 7.633279483037158e-06,
      "loss": 1.3539,
      "mean_token_accuracy": 0.7430707395076752,
      "num_tokens": 12853601.0,
      "step": 190
    },
    {
      "entropy": 0.9519797936081886,
      "epoch": 0.060592289631144436,
      "grad_norm": 0.535460352897644,
      "learning_rate": 8.037156704361875e-06,
      "loss": 1.35,
      "mean_token_accuracy": 0.7496364802122116,
      "num_tokens": 13539737.0,
      "step": 200
    },
    {
      "entropy": 0.9520228564739227,
      "epoch": 0.06362190411270166,
      "grad_norm": 0.6223520040512085,
      "learning_rate": 8.441033925686592e-06,
      "loss": 1.3474,
      "mean_token_accuracy": 0.7414789229631424,
      "num_tokens": 14212727.0,
      "step": 210
    },
    {
      "entropy": 0.9600193619728088,
      "epoch": 0.06665151859425889,
      "grad_norm": 0.342178612947464,
      "learning_rate": 8.844911147011308e-06,
      "loss": 1.3543,
      "mean_token_accuracy": 0.7420378759503364,
      "num_tokens": 14883203.0,
      "step": 220
    },
    {
      "entropy": 0.9429377406835556,
      "epoch": 0.0696811330758161,
      "grad_norm": 0.4176937937736511,
      "learning_rate": 9.248788368336027e-06,
      "loss": 1.3472,
      "mean_token_accuracy": 0.7506941795349121,
      "num_tokens": 15569666.0,
      "step": 230
    },
    {
      "entropy": 0.9486132815480233,
      "epoch": 0.07271074755737332,
      "grad_norm": 0.28106167912483215,
      "learning_rate": 9.652665589660744e-06,
      "loss": 1.3432,
      "mean_token_accuracy": 0.743465180695057,
      "num_tokens": 16248620.0,
      "step": 240
    },
    {
      "entropy": 0.9563860490918159,
      "epoch": 0.07574036203893054,
      "grad_norm": 0.5263408422470093,
      "learning_rate": 1.0056542810985462e-05,
      "loss": 1.3492,
      "mean_token_accuracy": 0.7460156157612801,
      "num_tokens": 16936825.0,
      "step": 250
    },
    {
      "entropy": 0.9662684842944145,
      "epoch": 0.07876997652048777,
      "grad_norm": 0.4372612535953522,
      "learning_rate": 1.0460420032310179e-05,
      "loss": 1.3627,
      "mean_token_accuracy": 0.7397747457027435,
      "num_tokens": 17614709.0,
      "step": 260
    },
    {
      "entropy": 0.9495426908135414,
      "epoch": 0.081799591002045,
      "grad_norm": 0.3195766508579254,
      "learning_rate": 1.0864297253634896e-05,
      "loss": 1.3494,
      "mean_token_accuracy": 0.7455736428499222,
      "num_tokens": 18296040.0,
      "step": 270
    },
    {
      "entropy": 0.946918123960495,
      "epoch": 0.0848292054836022,
      "grad_norm": 0.7261886596679688,
      "learning_rate": 1.1268174474959612e-05,
      "loss": 1.3465,
      "mean_token_accuracy": 0.7500029459595681,
      "num_tokens": 18979248.0,
      "step": 280
    },
    {
      "entropy": 0.951644915342331,
      "epoch": 0.08785881996515943,
      "grad_norm": 0.33003976941108704,
      "learning_rate": 1.1672051696284331e-05,
      "loss": 1.357,
      "mean_token_accuracy": 0.7410220310091973,
      "num_tokens": 19635598.0,
      "step": 290
    },
    {
      "entropy": 0.9348028033971787,
      "epoch": 0.09088843444671665,
      "grad_norm": 0.3807949423789978,
      "learning_rate": 1.2075928917609048e-05,
      "loss": 1.3454,
      "mean_token_accuracy": 0.7490146473050118,
      "num_tokens": 20307786.0,
      "step": 300
    },
    {
      "entropy": 0.9427580952644348,
      "epoch": 0.09391804892827388,
      "grad_norm": 0.3040551543235779,
      "learning_rate": 1.2479806138933766e-05,
      "loss": 1.3458,
      "mean_token_accuracy": 0.7427674695849419,
      "num_tokens": 20962399.0,
      "step": 310
    },
    {
      "entropy": 0.9372789472341537,
      "epoch": 0.0969476634098311,
      "grad_norm": 0.30593565106391907,
      "learning_rate": 1.2883683360258483e-05,
      "loss": 1.3477,
      "mean_token_accuracy": 0.744150060415268,
      "num_tokens": 21625782.0,
      "step": 320
    },
    {
      "entropy": 0.9325996667146683,
      "epoch": 0.09997727789138833,
      "grad_norm": 0.3430718779563904,
      "learning_rate": 1.32875605815832e-05,
      "loss": 1.3373,
      "mean_token_accuracy": 0.7520561441779137,
      "num_tokens": 22308600.0,
      "step": 330
    },
    {
      "entropy": 0.936019217967987,
      "epoch": 0.10300689237294554,
      "grad_norm": 0.31659775972366333,
      "learning_rate": 1.3691437802907916e-05,
      "loss": 1.3405,
      "mean_token_accuracy": 0.7451581284403801,
      "num_tokens": 22985019.0,
      "step": 340
    },
    {
      "entropy": 0.9315294235944748,
      "epoch": 0.10603650685450276,
      "grad_norm": 0.3390049636363983,
      "learning_rate": 1.4095315024232633e-05,
      "loss": 1.3404,
      "mean_token_accuracy": 0.7504669681191445,
      "num_tokens": 23663278.0,
      "step": 350
    },
    {
      "entropy": 0.9329710841178894,
      "epoch": 0.10906612133605999,
      "grad_norm": 0.27660778164863586,
      "learning_rate": 1.4499192245557353e-05,
      "loss": 1.3304,
      "mean_token_accuracy": 0.7510873705148697,
      "num_tokens": 24349520.0,
      "step": 360
    },
    {
      "entropy": 0.936093220114708,
      "epoch": 0.11209573581761721,
      "grad_norm": 0.37499693036079407,
      "learning_rate": 1.4903069466882068e-05,
      "loss": 1.3343,
      "mean_token_accuracy": 0.7447810202836991,
      "num_tokens": 25013352.0,
      "step": 370
    },
    {
      "entropy": 0.9364145249128342,
      "epoch": 0.11512535029917444,
      "grad_norm": 0.38497716188430786,
      "learning_rate": 1.5306946688206785e-05,
      "loss": 1.3386,
      "mean_token_accuracy": 0.7500496730208397,
      "num_tokens": 25701865.0,
      "step": 380
    },
    {
      "entropy": 0.9282684996724129,
      "epoch": 0.11815496478073165,
      "grad_norm": 0.3929134011268616,
      "learning_rate": 1.5710823909531502e-05,
      "loss": 1.3294,
      "mean_token_accuracy": 0.756112988293171,
      "num_tokens": 26409204.0,
      "step": 390
    },
    {
      "entropy": 0.9325812742114067,
      "epoch": 0.12118457926228887,
      "grad_norm": 0.2611294686794281,
      "learning_rate": 1.6114701130856222e-05,
      "loss": 1.3342,
      "mean_token_accuracy": 0.7490811735391617,
      "num_tokens": 27089751.0,
      "step": 400
    },
    {
      "entropy": 0.9268250733613967,
      "epoch": 0.1242141937438461,
      "grad_norm": 0.3077852427959442,
      "learning_rate": 1.651857835218094e-05,
      "loss": 1.3269,
      "mean_token_accuracy": 0.7516962170600892,
      "num_tokens": 27770172.0,
      "step": 410
    },
    {
      "entropy": 0.9232483997941017,
      "epoch": 0.12724380822540332,
      "grad_norm": 0.40509089827537537,
      "learning_rate": 1.6922455573505656e-05,
      "loss": 1.3298,
      "mean_token_accuracy": 0.7501018136739731,
      "num_tokens": 28446839.0,
      "step": 420
    },
    {
      "entropy": 0.9298500493168831,
      "epoch": 0.13027342270696055,
      "grad_norm": 0.7046210765838623,
      "learning_rate": 1.7326332794830372e-05,
      "loss": 1.3395,
      "mean_token_accuracy": 0.7426234304904937,
      "num_tokens": 29092374.0,
      "step": 430
    },
    {
      "entropy": 0.9365391865372658,
      "epoch": 0.13330303718851777,
      "grad_norm": 0.30491572618484497,
      "learning_rate": 1.7730210016155093e-05,
      "loss": 1.3322,
      "mean_token_accuracy": 0.7497914135456085,
      "num_tokens": 29782677.0,
      "step": 440
    },
    {
      "entropy": 0.9229448556900024,
      "epoch": 0.136332651670075,
      "grad_norm": 0.2625262141227722,
      "learning_rate": 1.8134087237479806e-05,
      "loss": 1.3253,
      "mean_token_accuracy": 0.7531854346394539,
      "num_tokens": 30465662.0,
      "step": 450
    },
    {
      "entropy": 0.9274683475494385,
      "epoch": 0.1393622661516322,
      "grad_norm": 0.28569865226745605,
      "learning_rate": 1.8537964458804523e-05,
      "loss": 1.3297,
      "mean_token_accuracy": 0.7466325148940086,
      "num_tokens": 31136752.0,
      "step": 460
    },
    {
      "entropy": 0.9268327504396439,
      "epoch": 0.14239188063318942,
      "grad_norm": 0.2649734616279602,
      "learning_rate": 1.894184168012924e-05,
      "loss": 1.326,
      "mean_token_accuracy": 0.7489441350102425,
      "num_tokens": 31817903.0,
      "step": 470
    },
    {
      "entropy": 0.9220950871706008,
      "epoch": 0.14542149511474664,
      "grad_norm": 0.29128533601760864,
      "learning_rate": 1.934571890145396e-05,
      "loss": 1.3292,
      "mean_token_accuracy": 0.7483928576111794,
      "num_tokens": 32484542.0,
      "step": 480
    },
    {
      "entropy": 0.918656262755394,
      "epoch": 0.14845110959630387,
      "grad_norm": 0.23605835437774658,
      "learning_rate": 1.9749596122778676e-05,
      "loss": 1.3282,
      "mean_token_accuracy": 0.7509550377726555,
      "num_tokens": 33167670.0,
      "step": 490
    },
    {
      "entropy": 0.9145197689533233,
      "epoch": 0.1514807240778611,
      "grad_norm": 0.23487012088298798,
      "learning_rate": 2.0153473344103393e-05,
      "loss": 1.3223,
      "mean_token_accuracy": 0.753958448767662,
      "num_tokens": 33848771.0,
      "step": 500
    },
    {
      "entropy": 0.9294292941689491,
      "epoch": 0.15451033855941831,
      "grad_norm": 0.24518702924251556,
      "learning_rate": 2.055735056542811e-05,
      "loss": 1.3245,
      "mean_token_accuracy": 0.7495481982827187,
      "num_tokens": 34530740.0,
      "step": 510
    },
    {
      "entropy": 0.9175413846969604,
      "epoch": 0.15753995304097554,
      "grad_norm": 0.2422870546579361,
      "learning_rate": 2.096122778675283e-05,
      "loss": 1.3182,
      "mean_token_accuracy": 0.754038056731224,
      "num_tokens": 35212668.0,
      "step": 520
    },
    {
      "entropy": 0.9198048174381256,
      "epoch": 0.16056956752253276,
      "grad_norm": 0.29682424664497375,
      "learning_rate": 2.1365105008077547e-05,
      "loss": 1.3243,
      "mean_token_accuracy": 0.7482954531908035,
      "num_tokens": 35884591.0,
      "step": 530
    },
    {
      "entropy": 0.9291597813367843,
      "epoch": 0.16359918200409,
      "grad_norm": 0.25015708804130554,
      "learning_rate": 2.1768982229402264e-05,
      "loss": 1.3351,
      "mean_token_accuracy": 0.7464247986674308,
      "num_tokens": 36549977.0,
      "step": 540
    },
    {
      "entropy": 0.9339838802814484,
      "epoch": 0.1666287964856472,
      "grad_norm": 0.23152834177017212,
      "learning_rate": 2.217285945072698e-05,
      "loss": 1.3308,
      "mean_token_accuracy": 0.7481212452054024,
      "num_tokens": 37225987.0,
      "step": 550
    },
    {
      "entropy": 0.9293303743004799,
      "epoch": 0.1696584109672044,
      "grad_norm": 0.24376530945301056,
      "learning_rate": 2.2576736672051697e-05,
      "loss": 1.329,
      "mean_token_accuracy": 0.750653937458992,
      "num_tokens": 37922786.0,
      "step": 560
    },
    {
      "entropy": 0.9323435664176941,
      "epoch": 0.17268802544876163,
      "grad_norm": 0.22466818988323212,
      "learning_rate": 2.2980613893376414e-05,
      "loss": 1.3318,
      "mean_token_accuracy": 0.7409193903207779,
      "num_tokens": 38583375.0,
      "step": 570
    },
    {
      "entropy": 0.9144039720296859,
      "epoch": 0.17571763993031886,
      "grad_norm": 0.2732308804988861,
      "learning_rate": 2.338449111470113e-05,
      "loss": 1.322,
      "mean_token_accuracy": 0.7507803827524185,
      "num_tokens": 39255897.0,
      "step": 580
    },
    {
      "entropy": 0.9145839855074882,
      "epoch": 0.17874725441187608,
      "grad_norm": 0.2554788589477539,
      "learning_rate": 2.3788368336025848e-05,
      "loss": 1.3256,
      "mean_token_accuracy": 0.7467795327305794,
      "num_tokens": 39920649.0,
      "step": 590
    },
    {
      "entropy": 0.9257472231984138,
      "epoch": 0.1817768688934333,
      "grad_norm": 0.23443762958049774,
      "learning_rate": 2.4192245557350568e-05,
      "loss": 1.3207,
      "mean_token_accuracy": 0.7508188039064407,
      "num_tokens": 40613016.0,
      "step": 600
    },
    {
      "entropy": 0.9184200003743171,
      "epoch": 0.18480648337499053,
      "grad_norm": 0.22434309124946594,
      "learning_rate": 2.4596122778675284e-05,
      "loss": 1.3257,
      "mean_token_accuracy": 0.7488593772053719,
      "num_tokens": 41277901.0,
      "step": 610
    },
    {
      "entropy": 0.9115656569600106,
      "epoch": 0.18783609785654776,
      "grad_norm": 0.27809539437294006,
      "learning_rate": 2.5e-05,
      "loss": 1.326,
      "mean_token_accuracy": 0.7556997299194336,
      "num_tokens": 41970791.0,
      "step": 620
    },
    {
      "entropy": 0.9088801115751266,
      "epoch": 0.19086571233810498,
      "grad_norm": 0.2008364498615265,
      "learning_rate": 2.5403877221324718e-05,
      "loss": 1.3117,
      "mean_token_accuracy": 0.7530135080218315,
      "num_tokens": 42665821.0,
      "step": 630
    },
    {
      "entropy": 0.9266581863164902,
      "epoch": 0.1938953268196622,
      "grad_norm": 0.22976058721542358,
      "learning_rate": 2.5807754442649435e-05,
      "loss": 1.3237,
      "mean_token_accuracy": 0.7488701492547989,
      "num_tokens": 43349418.0,
      "step": 640
    },
    {
      "entropy": 0.915707740187645,
      "epoch": 0.19692494130121943,
      "grad_norm": 0.22756750881671906,
      "learning_rate": 2.621163166397415e-05,
      "loss": 1.3191,
      "mean_token_accuracy": 0.7506552428007126,
      "num_tokens": 44030708.0,
      "step": 650
    },
    {
      "entropy": 0.9139649495482445,
      "epoch": 0.19995455578277666,
      "grad_norm": 0.2183780074119568,
      "learning_rate": 2.6615508885298868e-05,
      "loss": 1.3204,
      "mean_token_accuracy": 0.7497727707028389,
      "num_tokens": 44698948.0,
      "step": 660
    },
    {
      "entropy": 0.9156005099415779,
      "epoch": 0.20298417026433385,
      "grad_norm": 0.25522419810295105,
      "learning_rate": 2.701938610662359e-05,
      "loss": 1.3194,
      "mean_token_accuracy": 0.749407297372818,
      "num_tokens": 45375132.0,
      "step": 670
    },
    {
      "entropy": 0.9204802691936493,
      "epoch": 0.20601378474589108,
      "grad_norm": 0.2535501718521118,
      "learning_rate": 2.7423263327948305e-05,
      "loss": 1.316,
      "mean_token_accuracy": 0.7550273388624191,
      "num_tokens": 46068227.0,
      "step": 680
    },
    {
      "entropy": 0.9193328201770783,
      "epoch": 0.2090433992274483,
      "grad_norm": 0.2618839144706726,
      "learning_rate": 2.7827140549273022e-05,
      "loss": 1.3206,
      "mean_token_accuracy": 0.7545509234070777,
      "num_tokens": 46760521.0,
      "step": 690
    },
    {
      "entropy": 0.9152797266840935,
      "epoch": 0.21207301370900553,
      "grad_norm": 0.21720457077026367,
      "learning_rate": 2.823101777059774e-05,
      "loss": 1.3167,
      "mean_token_accuracy": 0.7576111331582069,
      "num_tokens": 47459871.0,
      "step": 700
    },
    {
      "entropy": 0.9146850362420083,
      "epoch": 0.21510262819056275,
      "grad_norm": 0.20628060400485992,
      "learning_rate": 2.8634894991922456e-05,
      "loss": 1.3168,
      "mean_token_accuracy": 0.7556830614805221,
      "num_tokens": 48155369.0,
      "step": 710
    },
    {
      "entropy": 0.9097549751400947,
      "epoch": 0.21813224267211997,
      "grad_norm": 0.24059303104877472,
      "learning_rate": 2.9038772213247172e-05,
      "loss": 1.316,
      "mean_token_accuracy": 0.756433616578579,
      "num_tokens": 48848798.0,
      "step": 720
    },
    {
      "entropy": 0.9219586238265037,
      "epoch": 0.2211618571536772,
      "grad_norm": 0.2317909449338913,
      "learning_rate": 2.944264943457189e-05,
      "loss": 1.3194,
      "mean_token_accuracy": 0.7525115668773651,
      "num_tokens": 49539969.0,
      "step": 730
    },
    {
      "entropy": 0.9171618938446044,
      "epoch": 0.22419147163523442,
      "grad_norm": 0.22604261338710785,
      "learning_rate": 2.9846526655896606e-05,
      "loss": 1.3294,
      "mean_token_accuracy": 0.7438126534223557,
      "num_tokens": 50202710.0,
      "step": 740
    },
    {
      "entropy": 0.9165790915489197,
      "epoch": 0.22722108611679165,
      "grad_norm": 0.22659213840961456,
      "learning_rate": 3.025040387722133e-05,
      "loss": 1.3258,
      "mean_token_accuracy": 0.7470232158899307,
      "num_tokens": 50865258.0,
      "step": 750
    },
    {
      "entropy": 0.9135280176997185,
      "epoch": 0.23025070059834887,
      "grad_norm": 0.20198464393615723,
      "learning_rate": 3.0654281098546046e-05,
      "loss": 1.3192,
      "mean_token_accuracy": 0.7495270982384682,
      "num_tokens": 51532155.0,
      "step": 760
    },
    {
      "entropy": 0.906398692727089,
      "epoch": 0.23328031507990607,
      "grad_norm": 0.2125513106584549,
      "learning_rate": 3.105815831987076e-05,
      "loss": 1.3137,
      "mean_token_accuracy": 0.7528043925762177,
      "num_tokens": 52213486.0,
      "step": 770
    },
    {
      "entropy": 0.924056950211525,
      "epoch": 0.2363099295614633,
      "grad_norm": 0.1975914090871811,
      "learning_rate": 3.146203554119548e-05,
      "loss": 1.33,
      "mean_token_accuracy": 0.74516611546278,
      "num_tokens": 52880609.0,
      "step": 780
    },
    {
      "entropy": 0.9093386635184288,
      "epoch": 0.23933954404302052,
      "grad_norm": 0.22010082006454468,
      "learning_rate": 3.1865912762520196e-05,
      "loss": 1.3199,
      "mean_token_accuracy": 0.7503703728318214,
      "num_tokens": 53553662.0,
      "step": 790
    },
    {
      "entropy": 0.9259981513023376,
      "epoch": 0.24236915852457774,
      "grad_norm": 0.22695289552211761,
      "learning_rate": 3.226978998384491e-05,
      "loss": 1.3281,
      "mean_token_accuracy": 0.7463536381721496,
      "num_tokens": 54218216.0,
      "step": 800
    },
    {
      "entropy": 0.9152119278907775,
      "epoch": 0.24539877300613497,
      "grad_norm": 0.18423648178577423,
      "learning_rate": 3.267366720516963e-05,
      "loss": 1.3301,
      "mean_token_accuracy": 0.7469505056738853,
      "num_tokens": 54877174.0,
      "step": 810
    },
    {
      "entropy": 0.9253642663359642,
      "epoch": 0.2484283874876922,
      "grad_norm": 0.20384164154529572,
      "learning_rate": 3.307754442649435e-05,
      "loss": 1.3242,
      "mean_token_accuracy": 0.7439268216490745,
      "num_tokens": 55544881.0,
      "step": 820
    },
    {
      "entropy": 0.9119655445218087,
      "epoch": 0.2514580019692494,
      "grad_norm": 0.18420374393463135,
      "learning_rate": 3.3481421647819064e-05,
      "loss": 1.3131,
      "mean_token_accuracy": 0.7535008266568184,
      "num_tokens": 56219809.0,
      "step": 830
    },
    {
      "entropy": 0.914957481622696,
      "epoch": 0.25448761645080664,
      "grad_norm": 0.19027648866176605,
      "learning_rate": 3.388529886914378e-05,
      "loss": 1.3193,
      "mean_token_accuracy": 0.7487210378050804,
      "num_tokens": 56900537.0,
      "step": 840
    },
    {
      "entropy": 0.9218035861849785,
      "epoch": 0.25751723093236384,
      "grad_norm": 0.16609440743923187,
      "learning_rate": 3.42891760904685e-05,
      "loss": 1.324,
      "mean_token_accuracy": 0.7453356191515923,
      "num_tokens": 57567272.0,
      "step": 850
    },
    {
      "entropy": 0.9182787030935288,
      "epoch": 0.2605468454139211,
      "grad_norm": 0.21082238852977753,
      "learning_rate": 3.4693053311793214e-05,
      "loss": 1.3195,
      "mean_token_accuracy": 0.7457280382514,
      "num_tokens": 58233643.0,
      "step": 860
    },
    {
      "entropy": 0.9092610657215119,
      "epoch": 0.2635764598954783,
      "grad_norm": 0.17639994621276855,
      "learning_rate": 3.509693053311793e-05,
      "loss": 1.3172,
      "mean_token_accuracy": 0.7527112409472465,
      "num_tokens": 58914246.0,
      "step": 870
    },
    {
      "entropy": 0.9199270486831665,
      "epoch": 0.26660607437703554,
      "grad_norm": 0.19245943427085876,
      "learning_rate": 3.550080775444265e-05,
      "loss": 1.3328,
      "mean_token_accuracy": 0.7467374846339225,
      "num_tokens": 59581373.0,
      "step": 880
    },
    {
      "entropy": 0.9137106329202652,
      "epoch": 0.26963568885859274,
      "grad_norm": 0.17330224812030792,
      "learning_rate": 3.5904684975767364e-05,
      "loss": 1.3232,
      "mean_token_accuracy": 0.7464903146028519,
      "num_tokens": 60243225.0,
      "step": 890
    },
    {
      "entropy": 0.9105766281485558,
      "epoch": 0.27266530334015,
      "grad_norm": 0.19774368405342102,
      "learning_rate": 3.630856219709208e-05,
      "loss": 1.3217,
      "mean_token_accuracy": 0.7488515123724937,
      "num_tokens": 60912466.0,
      "step": 900
    },
    {
      "entropy": 0.9119695752859116,
      "epoch": 0.2756949178217072,
      "grad_norm": 0.16941364109516144,
      "learning_rate": 3.6712439418416804e-05,
      "loss": 1.3223,
      "mean_token_accuracy": 0.7479856431484222,
      "num_tokens": 61564122.0,
      "step": 910
    },
    {
      "entropy": 0.914450266957283,
      "epoch": 0.2787245323032644,
      "grad_norm": 0.19449332356452942,
      "learning_rate": 3.711631663974152e-05,
      "loss": 1.3228,
      "mean_token_accuracy": 0.7504349738359452,
      "num_tokens": 62233788.0,
      "step": 920
    },
    {
      "entropy": 0.9095871090888977,
      "epoch": 0.28175414678482164,
      "grad_norm": 0.20824147760868073,
      "learning_rate": 3.752019386106624e-05,
      "loss": 1.3107,
      "mean_token_accuracy": 0.7514926999807358,
      "num_tokens": 62904530.0,
      "step": 930
    },
    {
      "entropy": 0.9049410030245781,
      "epoch": 0.28478376126637883,
      "grad_norm": 0.19165048003196716,
      "learning_rate": 3.7924071082390955e-05,
      "loss": 1.3164,
      "mean_token_accuracy": 0.7510851979255676,
      "num_tokens": 63568085.0,
      "step": 940
    },
    {
      "entropy": 0.9160818889737129,
      "epoch": 0.2878133757479361,
      "grad_norm": 0.1915004700422287,
      "learning_rate": 3.832794830371567e-05,
      "loss": 1.318,
      "mean_token_accuracy": 0.7500975534319878,
      "num_tokens": 64243907.0,
      "step": 950
    },
    {
      "entropy": 0.9147925272583961,
      "epoch": 0.2908429902294933,
      "grad_norm": 0.1861671656370163,
      "learning_rate": 3.873182552504039e-05,
      "loss": 1.3158,
      "mean_token_accuracy": 0.7435453131794929,
      "num_tokens": 64897465.0,
      "step": 960
    },
    {
      "entropy": 0.9032413572072983,
      "epoch": 0.29387260471105053,
      "grad_norm": 0.1919008046388626,
      "learning_rate": 3.9135702746365105e-05,
      "loss": 1.318,
      "mean_token_accuracy": 0.7475991874933243,
      "num_tokens": 65552452.0,
      "step": 970
    },
    {
      "entropy": 0.9074742197990417,
      "epoch": 0.29690221919260773,
      "grad_norm": 0.1770084798336029,
      "learning_rate": 3.953957996768982e-05,
      "loss": 1.3074,
      "mean_token_accuracy": 0.7538325354456902,
      "num_tokens": 66243290.0,
      "step": 980
    },
    {
      "entropy": 0.9117066383361816,
      "epoch": 0.299931833674165,
      "grad_norm": 0.20442838966846466,
      "learning_rate": 3.9943457189014545e-05,
      "loss": 1.3158,
      "mean_token_accuracy": 0.7498932734131813,
      "num_tokens": 66915792.0,
      "step": 990
    },
    {
      "entropy": 0.9175997883081436,
      "epoch": 0.3029614481557222,
      "grad_norm": 0.21906521916389465,
      "learning_rate": 4.034733441033926e-05,
      "loss": 1.3292,
      "mean_token_accuracy": 0.7450326055288314,
      "num_tokens": 67572749.0,
      "step": 1000
    },
    {
      "entropy": 0.9061859205365181,
      "epoch": 0.30599106263727943,
      "grad_norm": 0.18347729742527008,
      "learning_rate": 4.075121163166398e-05,
      "loss": 1.3186,
      "mean_token_accuracy": 0.751769931614399,
      "num_tokens": 68243021.0,
      "step": 1010
    },
    {
      "entropy": 0.9110336869955062,
      "epoch": 0.30902067711883663,
      "grad_norm": 0.1732109934091568,
      "learning_rate": 4.1155088852988696e-05,
      "loss": 1.3261,
      "mean_token_accuracy": 0.7522352248430252,
      "num_tokens": 68916281.0,
      "step": 1020
    },
    {
      "entropy": 0.9066652849316597,
      "epoch": 0.3120502916003938,
      "grad_norm": 0.16653181612491608,
      "learning_rate": 4.155896607431341e-05,
      "loss": 1.3163,
      "mean_token_accuracy": 0.7484516441822052,
      "num_tokens": 69579268.0,
      "step": 1030
    },
    {
      "entropy": 0.9179822281002998,
      "epoch": 0.3150799060819511,
      "grad_norm": 0.16437485814094543,
      "learning_rate": 4.196284329563813e-05,
      "loss": 1.3241,
      "mean_token_accuracy": 0.7510016962885857,
      "num_tokens": 70264079.0,
      "step": 1040
    },
    {
      "entropy": 0.9145135924220085,
      "epoch": 0.3181095205635083,
      "grad_norm": 0.15796828269958496,
      "learning_rate": 4.2366720516962846e-05,
      "loss": 1.3149,
      "mean_token_accuracy": 0.7525526866316795,
      "num_tokens": 70945929.0,
      "step": 1050
    },
    {
      "entropy": 0.9066309273242951,
      "epoch": 0.3211391350450655,
      "grad_norm": 0.15555834770202637,
      "learning_rate": 4.277059773828756e-05,
      "loss": 1.3164,
      "mean_token_accuracy": 0.7513958021998406,
      "num_tokens": 71617250.0,
      "step": 1060
    },
    {
      "entropy": 0.9203481048345565,
      "epoch": 0.3241687495266227,
      "grad_norm": 0.17885833978652954,
      "learning_rate": 4.317447495961228e-05,
      "loss": 1.3168,
      "mean_token_accuracy": 0.7503919199109077,
      "num_tokens": 72294905.0,
      "step": 1070
    },
    {
      "entropy": 0.908906027674675,
      "epoch": 0.32719836400818,
      "grad_norm": 0.15727676451206207,
      "learning_rate": 4.3578352180936996e-05,
      "loss": 1.3187,
      "mean_token_accuracy": 0.7530272096395493,
      "num_tokens": 72977657.0,
      "step": 1080
    },
    {
      "entropy": 0.9122103303670883,
      "epoch": 0.3302279784897372,
      "grad_norm": 0.1925109326839447,
      "learning_rate": 4.398222940226171e-05,
      "loss": 1.3205,
      "mean_token_accuracy": 0.7486942082643508,
      "num_tokens": 73649741.0,
      "step": 1090
    },
    {
      "entropy": 0.9175547957420349,
      "epoch": 0.3332575929712944,
      "grad_norm": 0.1666097491979599,
      "learning_rate": 4.438610662358643e-05,
      "loss": 1.3154,
      "mean_token_accuracy": 0.7506437614560127,
      "num_tokens": 74327124.0,
      "step": 1100
    },
    {
      "entropy": 0.9178244799375535,
      "epoch": 0.3362872074528516,
      "grad_norm": 0.19218212366104126,
      "learning_rate": 4.4789983844911147e-05,
      "loss": 1.3266,
      "mean_token_accuracy": 0.7483278617262841,
      "num_tokens": 74993713.0,
      "step": 1110
    },
    {
      "entropy": 0.9133123695850373,
      "epoch": 0.3393168219344088,
      "grad_norm": 0.21947059035301208,
      "learning_rate": 4.519386106623586e-05,
      "loss": 1.315,
      "mean_token_accuracy": 0.7546236589550972,
      "num_tokens": 75684557.0,
      "step": 1120
    },
    {
      "entropy": 0.9051602393388748,
      "epoch": 0.34234643641596607,
      "grad_norm": 0.17523469030857086,
      "learning_rate": 4.559773828756058e-05,
      "loss": 1.3137,
      "mean_token_accuracy": 0.7507795050740242,
      "num_tokens": 76352021.0,
      "step": 1130
    },
    {
      "entropy": 0.9085398092865944,
      "epoch": 0.34537605089752327,
      "grad_norm": 0.17349685728549957,
      "learning_rate": 4.60016155088853e-05,
      "loss": 1.3122,
      "mean_token_accuracy": 0.7487612262368202,
      "num_tokens": 77019617.0,
      "step": 1140
    },
    {
      "entropy": 0.9007386222481728,
      "epoch": 0.3484056653790805,
      "grad_norm": 0.16674205660820007,
      "learning_rate": 4.640549273021002e-05,
      "loss": 1.3166,
      "mean_token_accuracy": 0.7528904139995575,
      "num_tokens": 77687724.0,
      "step": 1150
    },
    {
      "entropy": 0.9056428268551826,
      "epoch": 0.3514352798606377,
      "grad_norm": 0.17023763060569763,
      "learning_rate": 4.680936995153474e-05,
      "loss": 1.3177,
      "mean_token_accuracy": 0.7535073205828666,
      "num_tokens": 78366654.0,
      "step": 1160
    },
    {
      "entropy": 0.9059616297483444,
      "epoch": 0.35446489434219497,
      "grad_norm": 0.1579790711402893,
      "learning_rate": 4.7213247172859454e-05,
      "loss": 1.3092,
      "mean_token_accuracy": 0.7533615916967392,
      "num_tokens": 79045395.0,
      "step": 1170
    },
    {
      "entropy": 0.9000560775399208,
      "epoch": 0.35749450882375217,
      "grad_norm": 0.17769810557365417,
      "learning_rate": 4.761712439418417e-05,
      "loss": 1.3064,
      "mean_token_accuracy": 0.7528070211410522,
      "num_tokens": 79724818.0,
      "step": 1180
    },
    {
      "entropy": 0.906897796690464,
      "epoch": 0.3605241233053094,
      "grad_norm": 0.1549178808927536,
      "learning_rate": 4.802100161550889e-05,
      "loss": 1.3141,
      "mean_token_accuracy": 0.7501399010419846,
      "num_tokens": 80392394.0,
      "step": 1190
    },
    {
      "entropy": 0.9162770986557007,
      "epoch": 0.3635537377868666,
      "grad_norm": 0.15900464355945587,
      "learning_rate": 4.8424878836833604e-05,
      "loss": 1.3179,
      "mean_token_accuracy": 0.7535986736416816,
      "num_tokens": 81080041.0,
      "step": 1200
    },
    {
      "entropy": 0.9049024254083633,
      "epoch": 0.36658335226842387,
      "grad_norm": 0.1732468456029892,
      "learning_rate": 4.882875605815832e-05,
      "loss": 1.3086,
      "mean_token_accuracy": 0.7500874444842338,
      "num_tokens": 81745731.0,
      "step": 1210
    },
    {
      "entropy": 0.8985488951206207,
      "epoch": 0.36961296674998106,
      "grad_norm": 0.1543564349412918,
      "learning_rate": 4.923263327948304e-05,
      "loss": 1.3099,
      "mean_token_accuracy": 0.7510283410549163,
      "num_tokens": 82410533.0,
      "step": 1220
    },
    {
      "entropy": 0.9009802848100662,
      "epoch": 0.37264258123153826,
      "grad_norm": 0.15668995678424835,
      "learning_rate": 4.963651050080776e-05,
      "loss": 1.3104,
      "mean_token_accuracy": 0.7525711953639984,
      "num_tokens": 83085040.0,
      "step": 1230
    },
    {
      "entropy": 0.9112498596310615,
      "epoch": 0.3756721957130955,
      "grad_norm": 0.19351772964000702,
      "learning_rate": 5.004038772213248e-05,
      "loss": 1.3121,
      "mean_token_accuracy": 0.753445765376091,
      "num_tokens": 83773911.0,
      "step": 1240
    },
    {
      "entropy": 0.9024754449725151,
      "epoch": 0.3787018101946527,
      "grad_norm": 0.15581312775611877,
      "learning_rate": 5.044426494345719e-05,
      "loss": 1.3113,
      "mean_token_accuracy": 0.7532795190811157,
      "num_tokens": 84454497.0,
      "step": 1250
    },
    {
      "entropy": 0.9097319796681405,
      "epoch": 0.38173142467620996,
      "grad_norm": 0.16212372481822968,
      "learning_rate": 5.084814216478191e-05,
      "loss": 1.3199,
      "mean_token_accuracy": 0.7565267696976662,
      "num_tokens": 85151452.0,
      "step": 1260
    },
    {
      "entropy": 0.9046913087368011,
      "epoch": 0.38476103915776716,
      "grad_norm": 0.16039691865444183,
      "learning_rate": 5.125201938610663e-05,
      "loss": 1.3075,
      "mean_token_accuracy": 0.755273811519146,
      "num_tokens": 85831124.0,
      "step": 1270
    },
    {
      "entropy": 0.9037176728248596,
      "epoch": 0.3877906536393244,
      "grad_norm": 0.18313613533973694,
      "learning_rate": 5.1655896607431345e-05,
      "loss": 1.3074,
      "mean_token_accuracy": 0.7516591116786003,
      "num_tokens": 86503499.0,
      "step": 1280
    },
    {
      "entropy": 0.8991178900003434,
      "epoch": 0.3908202681208816,
      "grad_norm": 0.15124350786209106,
      "learning_rate": 5.205977382875606e-05,
      "loss": 1.3132,
      "mean_token_accuracy": 0.7554905250668525,
      "num_tokens": 87183993.0,
      "step": 1290
    },
    {
      "entropy": 0.9112952843308448,
      "epoch": 0.39384988260243886,
      "grad_norm": 0.1569276601076126,
      "learning_rate": 5.246365105008078e-05,
      "loss": 1.3055,
      "mean_token_accuracy": 0.746527048945427,
      "num_tokens": 87860571.0,
      "step": 1300
    },
    {
      "entropy": 0.9121203318238258,
      "epoch": 0.39687949708399606,
      "grad_norm": 0.1368400901556015,
      "learning_rate": 5.2867528271405496e-05,
      "loss": 1.3232,
      "mean_token_accuracy": 0.7461899116635322,
      "num_tokens": 88517537.0,
      "step": 1310
    },
    {
      "entropy": 0.9051359668374062,
      "epoch": 0.3999091115655533,
      "grad_norm": 0.2260213941335678,
      "learning_rate": 5.3271405492730206e-05,
      "loss": 1.3183,
      "mean_token_accuracy": 0.7520864903926849,
      "num_tokens": 89195331.0,
      "step": 1320
    },
    {
      "entropy": 0.9018226325511932,
      "epoch": 0.4029387260471105,
      "grad_norm": 0.16223658621311188,
      "learning_rate": 5.367528271405493e-05,
      "loss": 1.3091,
      "mean_token_accuracy": 0.7538933858275414,
      "num_tokens": 89881354.0,
      "step": 1330
    },
    {
      "entropy": 0.9095069825649261,
      "epoch": 0.4059683405286677,
      "grad_norm": 0.17114722728729248,
      "learning_rate": 5.407915993537964e-05,
      "loss": 1.3234,
      "mean_token_accuracy": 0.7512053146958351,
      "num_tokens": 90556951.0,
      "step": 1340
    },
    {
      "entropy": 0.9174318969249725,
      "epoch": 0.40899795501022496,
      "grad_norm": 0.16955164074897766,
      "learning_rate": 5.448303715670436e-05,
      "loss": 1.3245,
      "mean_token_accuracy": 0.7475357472896575,
      "num_tokens": 91231305.0,
      "step": 1350
    },
    {
      "entropy": 0.9048666179180145,
      "epoch": 0.41202756949178215,
      "grad_norm": 0.13146966695785522,
      "learning_rate": 5.4886914378029086e-05,
      "loss": 1.3152,
      "mean_token_accuracy": 0.7531999439001084,
      "num_tokens": 91905496.0,
      "step": 1360
    },
    {
      "entropy": 0.8951819375157356,
      "epoch": 0.4150571839733394,
      "grad_norm": 0.17518994212150574,
      "learning_rate": 5.5290791599353796e-05,
      "loss": 1.3085,
      "mean_token_accuracy": 0.7539542600512504,
      "num_tokens": 92573582.0,
      "step": 1370
    },
    {
      "entropy": 0.8991360008716583,
      "epoch": 0.4180867984548966,
      "grad_norm": 0.16288205981254578,
      "learning_rate": 5.569466882067852e-05,
      "loss": 1.3105,
      "mean_token_accuracy": 0.749615055322647,
      "num_tokens": 93238965.0,
      "step": 1380
    },
    {
      "entropy": 0.9040171906352044,
      "epoch": 0.42111641293645385,
      "grad_norm": 0.15458066761493683,
      "learning_rate": 5.609854604200323e-05,
      "loss": 1.3143,
      "mean_token_accuracy": 0.74465721398592,
      "num_tokens": 93881702.0,
      "step": 1390
    },
    {
      "entropy": 0.9005857020616531,
      "epoch": 0.42414602741801105,
      "grad_norm": 0.1993013173341751,
      "learning_rate": 5.650242326332795e-05,
      "loss": 1.3075,
      "mean_token_accuracy": 0.7470027223229408,
      "num_tokens": 94547314.0,
      "step": 1400
    },
    {
      "entropy": 0.9084305763244629,
      "epoch": 0.4271756418995683,
      "grad_norm": 0.17728160321712494,
      "learning_rate": 5.690630048465266e-05,
      "loss": 1.3134,
      "mean_token_accuracy": 0.7524486899375915,
      "num_tokens": 95228068.0,
      "step": 1410
    },
    {
      "entropy": 0.906204080581665,
      "epoch": 0.4302052563811255,
      "grad_norm": 0.1374523639678955,
      "learning_rate": 5.731017770597739e-05,
      "loss": 1.3099,
      "mean_token_accuracy": 0.7536841064691544,
      "num_tokens": 95898188.0,
      "step": 1420
    },
    {
      "entropy": 0.8936747655272483,
      "epoch": 0.4332348708626827,
      "grad_norm": 0.22770892083644867,
      "learning_rate": 5.771405492730211e-05,
      "loss": 1.302,
      "mean_token_accuracy": 0.7515007808804512,
      "num_tokens": 96561074.0,
      "step": 1430
    },
    {
      "entropy": 0.9160967841744423,
      "epoch": 0.43626448534423995,
      "grad_norm": 0.1860688477754593,
      "learning_rate": 5.811793214862682e-05,
      "loss": 1.3231,
      "mean_token_accuracy": 0.7463960200548172,
      "num_tokens": 97228420.0,
      "step": 1440
    },
    {
      "entropy": 0.9105578064918518,
      "epoch": 0.43929409982579715,
      "grad_norm": 0.14435435831546783,
      "learning_rate": 5.8521809369951544e-05,
      "loss": 1.3169,
      "mean_token_accuracy": 0.7499482974410057,
      "num_tokens": 97906309.0,
      "step": 1450
    },
    {
      "entropy": 0.9005340814590455,
      "epoch": 0.4423237143073544,
      "grad_norm": 0.14899055659770966,
      "learning_rate": 5.8925686591276254e-05,
      "loss": 1.3051,
      "mean_token_accuracy": 0.7561104699969292,
      "num_tokens": 98587364.0,
      "step": 1460
    },
    {
      "entropy": 0.8977873668074607,
      "epoch": 0.4453533287889116,
      "grad_norm": 0.15377895534038544,
      "learning_rate": 5.932956381260098e-05,
      "loss": 1.3097,
      "mean_token_accuracy": 0.7519298970699311,
      "num_tokens": 99253227.0,
      "step": 1470
    },
    {
      "entropy": 0.9171545535326004,
      "epoch": 0.44838294327046885,
      "grad_norm": 0.25069257616996765,
      "learning_rate": 5.973344103392569e-05,
      "loss": 1.3189,
      "mean_token_accuracy": 0.7501741036772728,
      "num_tokens": 99936394.0,
      "step": 1480
    },
    {
      "entropy": 0.8939554750919342,
      "epoch": 0.45141255775202604,
      "grad_norm": 0.14128153026103973,
      "learning_rate": 6.013731825525041e-05,
      "loss": 1.3056,
      "mean_token_accuracy": 0.7544304430484772,
      "num_tokens": 100609703.0,
      "step": 1490
    },
    {
      "entropy": 0.8988944992423058,
      "epoch": 0.4544421722335833,
      "grad_norm": 0.1412612348794937,
      "learning_rate": 6.054119547657512e-05,
      "loss": 1.3037,
      "mean_token_accuracy": 0.7509754836559296,
      "num_tokens": 101281879.0,
      "step": 1500
    },
    {
      "entropy": 0.9070074841380119,
      "epoch": 0.4574717867151405,
      "grad_norm": 0.16410954296588898,
      "learning_rate": 6.094507269789984e-05,
      "loss": 1.3068,
      "mean_token_accuracy": 0.7516612082719802,
      "num_tokens": 101960642.0,
      "step": 1510
    },
    {
      "entropy": 0.9051597341895103,
      "epoch": 0.46050140119669775,
      "grad_norm": 0.14085623621940613,
      "learning_rate": 6.134894991922456e-05,
      "loss": 1.3162,
      "mean_token_accuracy": 0.753963851928711,
      "num_tokens": 102642682.0,
      "step": 1520
    },
    {
      "entropy": 0.9033859834074974,
      "epoch": 0.46353101567825494,
      "grad_norm": 0.15599344670772552,
      "learning_rate": 6.175282714054928e-05,
      "loss": 1.3099,
      "mean_token_accuracy": 0.7503072485327721,
      "num_tokens": 103309851.0,
      "step": 1530
    },
    {
      "entropy": 0.8905642539262771,
      "epoch": 0.46656063015981214,
      "grad_norm": 0.15819436311721802,
      "learning_rate": 6.2156704361874e-05,
      "loss": 1.3062,
      "mean_token_accuracy": 0.7543276980519295,
      "num_tokens": 103977447.0,
      "step": 1540
    },
    {
      "entropy": 0.9092970073223114,
      "epoch": 0.4695902446413694,
      "grad_norm": 0.14411026239395142,
      "learning_rate": 6.256058158319871e-05,
      "loss": 1.3011,
      "mean_token_accuracy": 0.7539940506219864,
      "num_tokens": 104673098.0,
      "step": 1550
    },
    {
      "entropy": 0.9055310070514679,
      "epoch": 0.4726198591229266,
      "grad_norm": 0.13895118236541748,
      "learning_rate": 6.296445880452343e-05,
      "loss": 1.3045,
      "mean_token_accuracy": 0.7529991418123245,
      "num_tokens": 105355415.0,
      "step": 1560
    },
    {
      "entropy": 0.9107632115483284,
      "epoch": 0.47564947360448384,
      "grad_norm": 0.1816311925649643,
      "learning_rate": 6.336833602584815e-05,
      "loss": 1.3155,
      "mean_token_accuracy": 0.7511478364467621,
      "num_tokens": 106022321.0,
      "step": 1570
    },
    {
      "entropy": 0.901921109855175,
      "epoch": 0.47867908808604104,
      "grad_norm": 0.12530627846717834,
      "learning_rate": 6.377221324717286e-05,
      "loss": 1.3099,
      "mean_token_accuracy": 0.7471093505620956,
      "num_tokens": 106678615.0,
      "step": 1580
    },
    {
      "entropy": 0.9076029598712921,
      "epoch": 0.4817087025675983,
      "grad_norm": 0.13566669821739197,
      "learning_rate": 6.417609046849758e-05,
      "loss": 1.3133,
      "mean_token_accuracy": 0.7520950257778167,
      "num_tokens": 107359476.0,
      "step": 1590
    },
    {
      "entropy": 0.9045285671949387,
      "epoch": 0.4847383170491555,
      "grad_norm": 0.13298770785331726,
      "learning_rate": 6.45799676898223e-05,
      "loss": 1.3057,
      "mean_token_accuracy": 0.7559781327843667,
      "num_tokens": 108039664.0,
      "step": 1600
    },
    {
      "entropy": 0.9136948719620704,
      "epoch": 0.48776793153071274,
      "grad_norm": 0.16770410537719727,
      "learning_rate": 6.498384491114701e-05,
      "loss": 1.3199,
      "mean_token_accuracy": 0.750070121884346,
      "num_tokens": 108709395.0,
      "step": 1610
    },
    {
      "entropy": 0.9032595708966256,
      "epoch": 0.49079754601226994,
      "grad_norm": 0.1383184939622879,
      "learning_rate": 6.538772213247173e-05,
      "loss": 1.3076,
      "mean_token_accuracy": 0.7500165730714798,
      "num_tokens": 109385993.0,
      "step": 1620
    },
    {
      "entropy": 0.8895248159766197,
      "epoch": 0.49382716049382713,
      "grad_norm": 0.14151626825332642,
      "learning_rate": 6.579159935379645e-05,
      "loss": 1.2997,
      "mean_token_accuracy": 0.75679482370615,
      "num_tokens": 110060837.0,
      "step": 1630
    },
    {
      "entropy": 0.9049818530678749,
      "epoch": 0.4968567749753844,
      "grad_norm": 0.13891007006168365,
      "learning_rate": 6.619547657512116e-05,
      "loss": 1.3137,
      "mean_token_accuracy": 0.7519566610455513,
      "num_tokens": 110739371.0,
      "step": 1640
    },
    {
      "entropy": 0.9043860971927643,
      "epoch": 0.4998863894569416,
      "grad_norm": 0.17376911640167236,
      "learning_rate": 6.659935379644588e-05,
      "loss": 1.3066,
      "mean_token_accuracy": 0.754023851454258,
      "num_tokens": 111429428.0,
      "step": 1650
    },
    {
      "entropy": 0.8931585773825645,
      "epoch": 0.5029160039384988,
      "grad_norm": 0.14392395317554474,
      "learning_rate": 6.70032310177706e-05,
      "loss": 1.3051,
      "mean_token_accuracy": 0.7539272055029869,
      "num_tokens": 112101071.0,
      "step": 1660
    },
    {
      "entropy": 0.8960565209388733,
      "epoch": 0.505945618420056,
      "grad_norm": 0.12862016260623932,
      "learning_rate": 6.740710823909531e-05,
      "loss": 1.3062,
      "mean_token_accuracy": 0.752163028717041,
      "num_tokens": 112776625.0,
      "step": 1670
    },
    {
      "entropy": 0.8961879700422287,
      "epoch": 0.5089752329016133,
      "grad_norm": 0.13162676990032196,
      "learning_rate": 6.781098546042004e-05,
      "loss": 1.304,
      "mean_token_accuracy": 0.7529030278325081,
      "num_tokens": 113445247.0,
      "step": 1680
    },
    {
      "entropy": 0.8980554938316345,
      "epoch": 0.5120048473831705,
      "grad_norm": 0.173239603638649,
      "learning_rate": 6.821486268174475e-05,
      "loss": 1.3073,
      "mean_token_accuracy": 0.7578601375222206,
      "num_tokens": 114139450.0,
      "step": 1690
    },
    {
      "entropy": 0.9074166044592857,
      "epoch": 0.5150344618647277,
      "grad_norm": 0.14682774245738983,
      "learning_rate": 6.861873990306948e-05,
      "loss": 1.3123,
      "mean_token_accuracy": 0.7464081138372421,
      "num_tokens": 114794323.0,
      "step": 1700
    },
    {
      "entropy": 0.910780593752861,
      "epoch": 0.5180640763462849,
      "grad_norm": 0.14194752275943756,
      "learning_rate": 6.902261712439418e-05,
      "loss": 1.3149,
      "mean_token_accuracy": 0.7536340922117233,
      "num_tokens": 115491317.0,
      "step": 1710
    },
    {
      "entropy": 0.9062231466174125,
      "epoch": 0.5210936908278422,
      "grad_norm": 0.15659962594509125,
      "learning_rate": 6.942649434571891e-05,
      "loss": 1.3211,
      "mean_token_accuracy": 0.7496401265263557,
      "num_tokens": 116154826.0,
      "step": 1720
    },
    {
      "entropy": 0.9074418857693672,
      "epoch": 0.5241233053093994,
      "grad_norm": 0.14031372964382172,
      "learning_rate": 6.983037156704361e-05,
      "loss": 1.3116,
      "mean_token_accuracy": 0.7487875491380691,
      "num_tokens": 116817268.0,
      "step": 1730
    },
    {
      "entropy": 0.8963802754878998,
      "epoch": 0.5271529197909566,
      "grad_norm": 0.12388686835765839,
      "learning_rate": 7.023424878836834e-05,
      "loss": 1.3083,
      "mean_token_accuracy": 0.7544260993599892,
      "num_tokens": 117498412.0,
      "step": 1740
    },
    {
      "entropy": 0.9051859691739083,
      "epoch": 0.5301825342725138,
      "grad_norm": 0.12408740818500519,
      "learning_rate": 7.063812600969306e-05,
      "loss": 1.3167,
      "mean_token_accuracy": 0.7449529379606247,
      "num_tokens": 118154932.0,
      "step": 1750
    },
    {
      "entropy": 0.9028876557946205,
      "epoch": 0.5332121487540711,
      "grad_norm": 0.12819243967533112,
      "learning_rate": 7.104200323101778e-05,
      "loss": 1.3099,
      "mean_token_accuracy": 0.7520097956061363,
      "num_tokens": 118825179.0,
      "step": 1760
    },
    {
      "entropy": 0.9011894449591636,
      "epoch": 0.5362417632356282,
      "grad_norm": 0.12917159497737885,
      "learning_rate": 7.14458804523425e-05,
      "loss": 1.3023,
      "mean_token_accuracy": 0.750635813176632,
      "num_tokens": 119503544.0,
      "step": 1770
    },
    {
      "entropy": 0.9058702141046524,
      "epoch": 0.5392713777171855,
      "grad_norm": 0.13154388964176178,
      "learning_rate": 7.184975767366721e-05,
      "loss": 1.3037,
      "mean_token_accuracy": 0.7499114170670509,
      "num_tokens": 120175476.0,
      "step": 1780
    },
    {
      "entropy": 0.8858574673533439,
      "epoch": 0.5423009921987427,
      "grad_norm": 0.12256129086017609,
      "learning_rate": 7.225363489499193e-05,
      "loss": 1.2895,
      "mean_token_accuracy": 0.7617484778165817,
      "num_tokens": 120871896.0,
      "step": 1790
    },
    {
      "entropy": 0.9071286633610726,
      "epoch": 0.5453306066803,
      "grad_norm": 0.12982648611068726,
      "learning_rate": 7.265751211631664e-05,
      "loss": 1.3137,
      "mean_token_accuracy": 0.7497906401753426,
      "num_tokens": 121548848.0,
      "step": 1800
    },
    {
      "entropy": 0.9043298572301864,
      "epoch": 0.5483602211618571,
      "grad_norm": 0.1469738930463791,
      "learning_rate": 7.306138933764136e-05,
      "loss": 1.3026,
      "mean_token_accuracy": 0.7541209653019905,
      "num_tokens": 122244302.0,
      "step": 1810
    },
    {
      "entropy": 0.9004949659109116,
      "epoch": 0.5513898356434144,
      "grad_norm": 0.1332157850265503,
      "learning_rate": 7.346526655896608e-05,
      "loss": 1.308,
      "mean_token_accuracy": 0.7553296402096749,
      "num_tokens": 122930654.0,
      "step": 1820
    },
    {
      "entropy": 0.8950658679008484,
      "epoch": 0.5544194501249716,
      "grad_norm": 0.14082078635692596,
      "learning_rate": 7.38691437802908e-05,
      "loss": 1.3041,
      "mean_token_accuracy": 0.751467551290989,
      "num_tokens": 123603708.0,
      "step": 1830
    },
    {
      "entropy": 0.8976402968168259,
      "epoch": 0.5574490646065288,
      "grad_norm": 0.13623163104057312,
      "learning_rate": 7.427302100161551e-05,
      "loss": 1.307,
      "mean_token_accuracy": 0.7543731585144997,
      "num_tokens": 124285203.0,
      "step": 1840
    },
    {
      "entropy": 0.9020175158977508,
      "epoch": 0.560478679088086,
      "grad_norm": 0.13262920081615448,
      "learning_rate": 7.467689822294023e-05,
      "loss": 1.3155,
      "mean_token_accuracy": 0.7512376889586448,
      "num_tokens": 124955188.0,
      "step": 1850
    },
    {
      "entropy": 0.8993414729833603,
      "epoch": 0.5635082935696433,
      "grad_norm": 0.12388888746500015,
      "learning_rate": 7.508077544426494e-05,
      "loss": 1.299,
      "mean_token_accuracy": 0.7550491631031037,
      "num_tokens": 125638409.0,
      "step": 1860
    },
    {
      "entropy": 0.8926667675375939,
      "epoch": 0.5665379080512005,
      "grad_norm": 0.10712610930204391,
      "learning_rate": 7.548465266558966e-05,
      "loss": 1.2996,
      "mean_token_accuracy": 0.7553495973348617,
      "num_tokens": 126318498.0,
      "step": 1870
    },
    {
      "entropy": 0.9032000929117203,
      "epoch": 0.5695675225327577,
      "grad_norm": 0.1535160094499588,
      "learning_rate": 7.588852988691438e-05,
      "loss": 1.3046,
      "mean_token_accuracy": 0.7536028608679771,
      "num_tokens": 126996280.0,
      "step": 1880
    },
    {
      "entropy": 0.8911816298961639,
      "epoch": 0.5725971370143149,
      "grad_norm": 0.13242319226264954,
      "learning_rate": 7.62924071082391e-05,
      "loss": 1.3041,
      "mean_token_accuracy": 0.755043214559555,
      "num_tokens": 127670506.0,
      "step": 1890
    },
    {
      "entropy": 0.8888354226946831,
      "epoch": 0.5756267514958722,
      "grad_norm": 0.12417840957641602,
      "learning_rate": 7.669628432956381e-05,
      "loss": 1.2999,
      "mean_token_accuracy": 0.7529826268553734,
      "num_tokens": 128339321.0,
      "step": 1900
    },
    {
      "entropy": 0.8958159863948822,
      "epoch": 0.5786563659774294,
      "grad_norm": 0.12293798476457596,
      "learning_rate": 7.710016155088854e-05,
      "loss": 1.3022,
      "mean_token_accuracy": 0.7521115452051162,
      "num_tokens": 129008376.0,
      "step": 1910
    },
    {
      "entropy": 0.8976372435688973,
      "epoch": 0.5816859804589866,
      "grad_norm": 0.12593239545822144,
      "learning_rate": 7.750403877221325e-05,
      "loss": 1.306,
      "mean_token_accuracy": 0.7565735295414925,
      "num_tokens": 129695657.0,
      "step": 1920
    },
    {
      "entropy": 0.8995207667350769,
      "epoch": 0.5847155949405438,
      "grad_norm": 0.1241973489522934,
      "learning_rate": 7.790791599353798e-05,
      "loss": 1.304,
      "mean_token_accuracy": 0.7579465642571449,
      "num_tokens": 130385609.0,
      "step": 1930
    },
    {
      "entropy": 0.8977545350790024,
      "epoch": 0.5877452094221011,
      "grad_norm": 0.12729386985301971,
      "learning_rate": 7.831179321486268e-05,
      "loss": 1.3057,
      "mean_token_accuracy": 0.7563092142343522,
      "num_tokens": 131069212.0,
      "step": 1940
    },
    {
      "entropy": 0.8979323357343674,
      "epoch": 0.5907748239036582,
      "grad_norm": 0.11168866604566574,
      "learning_rate": 7.871567043618741e-05,
      "loss": 1.2993,
      "mean_token_accuracy": 0.7524091944098472,
      "num_tokens": 131746856.0,
      "step": 1950
    },
    {
      "entropy": 0.8882772088050842,
      "epoch": 0.5938044383852155,
      "grad_norm": 0.1279069185256958,
      "learning_rate": 7.911954765751211e-05,
      "loss": 1.2985,
      "mean_token_accuracy": 0.7544113993644714,
      "num_tokens": 132423575.0,
      "step": 1960
    },
    {
      "entropy": 0.8982036307454109,
      "epoch": 0.5968340528667727,
      "grad_norm": 0.133626326918602,
      "learning_rate": 7.952342487883684e-05,
      "loss": 1.3017,
      "mean_token_accuracy": 0.7510138586163521,
      "num_tokens": 133085251.0,
      "step": 1970
    },
    {
      "entropy": 0.9008032843470574,
      "epoch": 0.59986366734833,
      "grad_norm": 0.129901722073555,
      "learning_rate": 7.992730210016155e-05,
      "loss": 1.3118,
      "mean_token_accuracy": 0.7510970681905746,
      "num_tokens": 133756971.0,
      "step": 1980
    },
    {
      "entropy": 0.8944846227765083,
      "epoch": 0.6028932818298871,
      "grad_norm": 0.12799471616744995,
      "learning_rate": 8.033117932148628e-05,
      "loss": 1.308,
      "mean_token_accuracy": 0.7527639657258988,
      "num_tokens": 134432445.0,
      "step": 1990
    },
    {
      "entropy": 0.9071986898779869,
      "epoch": 0.6059228963114444,
      "grad_norm": 0.12767554819583893,
      "learning_rate": 8.0735056542811e-05,
      "loss": 1.3109,
      "mean_token_accuracy": 0.7490269899368286,
      "num_tokens": 135107011.0,
      "step": 2000
    },
    {
      "entropy": 0.9076018154621124,
      "epoch": 0.6089525107930016,
      "grad_norm": 0.1324162483215332,
      "learning_rate": 8.113893376413571e-05,
      "loss": 1.3129,
      "mean_token_accuracy": 0.7502601996064187,
      "num_tokens": 135783615.0,
      "step": 2010
    },
    {
      "entropy": 0.9079612717032433,
      "epoch": 0.6119821252745589,
      "grad_norm": 0.13695330917835236,
      "learning_rate": 8.154281098546043e-05,
      "loss": 1.3148,
      "mean_token_accuracy": 0.7501401528716087,
      "num_tokens": 136461898.0,
      "step": 2020
    },
    {
      "entropy": 0.9103457570075989,
      "epoch": 0.615011739756116,
      "grad_norm": 0.14059525728225708,
      "learning_rate": 8.194668820678514e-05,
      "loss": 1.316,
      "mean_token_accuracy": 0.7511693567037583,
      "num_tokens": 137136732.0,
      "step": 2030
    },
    {
      "entropy": 0.895035968720913,
      "epoch": 0.6180413542376733,
      "grad_norm": 0.12641917169094086,
      "learning_rate": 8.235056542810986e-05,
      "loss": 1.3015,
      "mean_token_accuracy": 0.7535883888602257,
      "num_tokens": 137808825.0,
      "step": 2040
    },
    {
      "entropy": 0.8961443632841111,
      "epoch": 0.6210709687192305,
      "grad_norm": 0.14057329297065735,
      "learning_rate": 8.275444264943458e-05,
      "loss": 1.3004,
      "mean_token_accuracy": 0.7532713904976844,
      "num_tokens": 138491788.0,
      "step": 2050
    },
    {
      "entropy": 0.8971175700426102,
      "epoch": 0.6241005832007877,
      "grad_norm": 0.12069716304540634,
      "learning_rate": 8.31583198707593e-05,
      "loss": 1.2959,
      "mean_token_accuracy": 0.759382213652134,
      "num_tokens": 139184469.0,
      "step": 2060
    },
    {
      "entropy": 0.9055205196142196,
      "epoch": 0.6271301976823449,
      "grad_norm": 0.12682268023490906,
      "learning_rate": 8.356219709208401e-05,
      "loss": 1.3218,
      "mean_token_accuracy": 0.7479860588908196,
      "num_tokens": 139850956.0,
      "step": 2070
    },
    {
      "entropy": 0.9075116515159607,
      "epoch": 0.6301598121639022,
      "grad_norm": 0.1375780701637268,
      "learning_rate": 8.396607431340873e-05,
      "loss": 1.3146,
      "mean_token_accuracy": 0.7482119768857955,
      "num_tokens": 140521012.0,
      "step": 2080
    },
    {
      "entropy": 0.8973190426826477,
      "epoch": 0.6331894266454594,
      "grad_norm": 0.11789047718048096,
      "learning_rate": 8.436995153473344e-05,
      "loss": 1.2971,
      "mean_token_accuracy": 0.7563918709754944,
      "num_tokens": 141204334.0,
      "step": 2090
    },
    {
      "entropy": 0.9005673974752426,
      "epoch": 0.6362190411270165,
      "grad_norm": 0.12826324999332428,
      "learning_rate": 8.477382875605816e-05,
      "loss": 1.3046,
      "mean_token_accuracy": 0.7507783487439156,
      "num_tokens": 141873785.0,
      "step": 2100
    },
    {
      "entropy": 0.903111070394516,
      "epoch": 0.6392486556085738,
      "grad_norm": 0.1388614922761917,
      "learning_rate": 8.517770597738288e-05,
      "loss": 1.3102,
      "mean_token_accuracy": 0.7523079171776772,
      "num_tokens": 142549820.0,
      "step": 2110
    },
    {
      "entropy": 0.9010299742221832,
      "epoch": 0.642278270090131,
      "grad_norm": 0.12743902206420898,
      "learning_rate": 8.55815831987076e-05,
      "loss": 1.3158,
      "mean_token_accuracy": 0.7477226942777634,
      "num_tokens": 143205717.0,
      "step": 2120
    },
    {
      "entropy": 0.8998202800750732,
      "epoch": 0.6453078845716883,
      "grad_norm": 0.14212191104888916,
      "learning_rate": 8.598546042003231e-05,
      "loss": 1.3056,
      "mean_token_accuracy": 0.7463966920971871,
      "num_tokens": 143861299.0,
      "step": 2130
    },
    {
      "entropy": 0.8835368379950523,
      "epoch": 0.6483374990532454,
      "grad_norm": 0.1122928187251091,
      "learning_rate": 8.638933764135703e-05,
      "loss": 1.2886,
      "mean_token_accuracy": 0.7610997334122658,
      "num_tokens": 144557885.0,
      "step": 2140
    },
    {
      "entropy": 0.8946089386940003,
      "epoch": 0.6513671135348027,
      "grad_norm": 0.13596343994140625,
      "learning_rate": 8.679321486268174e-05,
      "loss": 1.2981,
      "mean_token_accuracy": 0.7543591871857643,
      "num_tokens": 145240703.0,
      "step": 2150
    },
    {
      "entropy": 0.9056951612234115,
      "epoch": 0.65439672801636,
      "grad_norm": 0.1226610541343689,
      "learning_rate": 8.719709208400648e-05,
      "loss": 1.3086,
      "mean_token_accuracy": 0.7488695859909058,
      "num_tokens": 145913481.0,
      "step": 2160
    },
    {
      "entropy": 0.906532245874405,
      "epoch": 0.6574263424979171,
      "grad_norm": 0.13818389177322388,
      "learning_rate": 8.760096930533118e-05,
      "loss": 1.3156,
      "mean_token_accuracy": 0.7476009607315064,
      "num_tokens": 146577442.0,
      "step": 2170
    },
    {
      "entropy": 0.8970832392573357,
      "epoch": 0.6604559569794743,
      "grad_norm": 0.11767213046550751,
      "learning_rate": 8.800484652665591e-05,
      "loss": 1.3108,
      "mean_token_accuracy": 0.7520548105239868,
      "num_tokens": 147246470.0,
      "step": 2180
    },
    {
      "entropy": 0.898816491663456,
      "epoch": 0.6634855714610316,
      "grad_norm": 0.12864157557487488,
      "learning_rate": 8.840872374798061e-05,
      "loss": 1.3081,
      "mean_token_accuracy": 0.7583793714642525,
      "num_tokens": 147928598.0,
      "step": 2190
    },
    {
      "entropy": 0.8935007572174072,
      "epoch": 0.6665151859425889,
      "grad_norm": 0.12095800042152405,
      "learning_rate": 8.881260096930534e-05,
      "loss": 1.2998,
      "mean_token_accuracy": 0.7564240217208862,
      "num_tokens": 148610120.0,
      "step": 2200
    },
    {
      "entropy": 0.9021251410245895,
      "epoch": 0.669544800424146,
      "grad_norm": 0.1266978681087494,
      "learning_rate": 8.921647819063005e-05,
      "loss": 1.311,
      "mean_token_accuracy": 0.7462113335728645,
      "num_tokens": 149256959.0,
      "step": 2210
    },
    {
      "entropy": 0.8901038557291031,
      "epoch": 0.6725744149057032,
      "grad_norm": 0.12045814841985703,
      "learning_rate": 8.962035541195478e-05,
      "loss": 1.2956,
      "mean_token_accuracy": 0.7490550830960274,
      "num_tokens": 149916961.0,
      "step": 2220
    },
    {
      "entropy": 0.8851861774921417,
      "epoch": 0.6756040293872605,
      "grad_norm": 0.1233411505818367,
      "learning_rate": 9.002423263327949e-05,
      "loss": 1.2895,
      "mean_token_accuracy": 0.757862888276577,
      "num_tokens": 150597520.0,
      "step": 2230
    },
    {
      "entropy": 0.8945758268237114,
      "epoch": 0.6786336438688176,
      "grad_norm": 0.11606448888778687,
      "learning_rate": 9.042810985460421e-05,
      "loss": 1.3073,
      "mean_token_accuracy": 0.751634182035923,
      "num_tokens": 151268277.0,
      "step": 2240
    },
    {
      "entropy": 0.9055110797286033,
      "epoch": 0.6816632583503749,
      "grad_norm": 0.1208593100309372,
      "learning_rate": 9.083198707592893e-05,
      "loss": 1.3085,
      "mean_token_accuracy": 0.7520596653223037,
      "num_tokens": 151937122.0,
      "step": 2250
    },
    {
      "entropy": 0.8921197906136513,
      "epoch": 0.6846928728319321,
      "grad_norm": 0.13769711554050446,
      "learning_rate": 9.123586429725364e-05,
      "loss": 1.2993,
      "mean_token_accuracy": 0.7518444448709488,
      "num_tokens": 152611330.0,
      "step": 2260
    },
    {
      "entropy": 0.9008217513561249,
      "epoch": 0.6877224873134894,
      "grad_norm": 0.13109052181243896,
      "learning_rate": 9.163974151857836e-05,
      "loss": 1.3106,
      "mean_token_accuracy": 0.7524908572435379,
      "num_tokens": 153288510.0,
      "step": 2270
    },
    {
      "entropy": 0.9004259362816811,
      "epoch": 0.6907521017950465,
      "grad_norm": 0.11402872204780579,
      "learning_rate": 9.204361873990308e-05,
      "loss": 1.2996,
      "mean_token_accuracy": 0.7515971258282661,
      "num_tokens": 153966417.0,
      "step": 2280
    },
    {
      "entropy": 0.8992137014865875,
      "epoch": 0.6937817162766038,
      "grad_norm": 0.13402849435806274,
      "learning_rate": 9.244749596122779e-05,
      "loss": 1.3007,
      "mean_token_accuracy": 0.7534859776496887,
      "num_tokens": 154654259.0,
      "step": 2290
    },
    {
      "entropy": 0.9095842137932777,
      "epoch": 0.696811330758161,
      "grad_norm": 0.12530536949634552,
      "learning_rate": 9.285137318255251e-05,
      "loss": 1.316,
      "mean_token_accuracy": 0.7539889082312584,
      "num_tokens": 155328311.0,
      "step": 2300
    },
    {
      "entropy": 0.8930700272321701,
      "epoch": 0.6998409452397183,
      "grad_norm": 0.12575875222682953,
      "learning_rate": 9.325525040387723e-05,
      "loss": 1.2997,
      "mean_token_accuracy": 0.7537848263978958,
      "num_tokens": 156003757.0,
      "step": 2310
    },
    {
      "entropy": 0.9096755936741829,
      "epoch": 0.7028705597212754,
      "grad_norm": 0.12259205430746078,
      "learning_rate": 9.365912762520194e-05,
      "loss": 1.3162,
      "mean_token_accuracy": 0.7465154975652695,
      "num_tokens": 156672017.0,
      "step": 2320
    },
    {
      "entropy": 0.8987697646021843,
      "epoch": 0.7059001742028327,
      "grad_norm": 0.14718621969223022,
      "learning_rate": 9.406300484652666e-05,
      "loss": 1.3027,
      "mean_token_accuracy": 0.757715892791748,
      "num_tokens": 157369100.0,
      "step": 2330
    },
    {
      "entropy": 0.8899315431714058,
      "epoch": 0.7089297886843899,
      "grad_norm": 0.12225979566574097,
      "learning_rate": 9.446688206785138e-05,
      "loss": 1.2981,
      "mean_token_accuracy": 0.7547729939222336,
      "num_tokens": 158036981.0,
      "step": 2340
    },
    {
      "entropy": 0.8941576257348061,
      "epoch": 0.7119594031659471,
      "grad_norm": 0.11984486132860184,
      "learning_rate": 9.48707592891761e-05,
      "loss": 1.295,
      "mean_token_accuracy": 0.7545301422476769,
      "num_tokens": 158723158.0,
      "step": 2350
    },
    {
      "entropy": 0.8968440085649491,
      "epoch": 0.7149890176475043,
      "grad_norm": 0.1378418505191803,
      "learning_rate": 9.527463651050081e-05,
      "loss": 1.2987,
      "mean_token_accuracy": 0.754326069355011,
      "num_tokens": 159407156.0,
      "step": 2360
    },
    {
      "entropy": 0.9001185387372971,
      "epoch": 0.7180186321290616,
      "grad_norm": 0.1220158115029335,
      "learning_rate": 9.567851373182553e-05,
      "loss": 1.2997,
      "mean_token_accuracy": 0.7525874629616738,
      "num_tokens": 160085714.0,
      "step": 2370
    },
    {
      "entropy": 0.898030799627304,
      "epoch": 0.7210482466106188,
      "grad_norm": 0.1130693331360817,
      "learning_rate": 9.608239095315024e-05,
      "loss": 1.298,
      "mean_token_accuracy": 0.7585792228579521,
      "num_tokens": 160784292.0,
      "step": 2380
    },
    {
      "entropy": 0.8831940025091172,
      "epoch": 0.724077861092176,
      "grad_norm": 0.12520240247249603,
      "learning_rate": 9.648626817447497e-05,
      "loss": 1.2882,
      "mean_token_accuracy": 0.7544678583741188,
      "num_tokens": 161464712.0,
      "step": 2390
    },
    {
      "entropy": 0.8926929473876953,
      "epoch": 0.7271074755737332,
      "grad_norm": 0.13716495037078857,
      "learning_rate": 9.689014539579968e-05,
      "loss": 1.2953,
      "mean_token_accuracy": 0.7539828538894653,
      "num_tokens": 162152296.0,
      "step": 2400
    },
    {
      "entropy": 0.8940726727247238,
      "epoch": 0.7301370900552905,
      "grad_norm": 0.11432312428951263,
      "learning_rate": 9.729402261712441e-05,
      "loss": 1.2963,
      "mean_token_accuracy": 0.7570353090763092,
      "num_tokens": 162838234.0,
      "step": 2410
    },
    {
      "entropy": 0.8960199370980263,
      "epoch": 0.7331667045368477,
      "grad_norm": 0.14145903289318085,
      "learning_rate": 9.769789983844911e-05,
      "loss": 1.3043,
      "mean_token_accuracy": 0.7553196415305138,
      "num_tokens": 163511723.0,
      "step": 2420
    },
    {
      "entropy": 0.890855559706688,
      "epoch": 0.7361963190184049,
      "grad_norm": 0.10976195335388184,
      "learning_rate": 9.810177705977384e-05,
      "loss": 1.305,
      "mean_token_accuracy": 0.7565831497311593,
      "num_tokens": 164191370.0,
      "step": 2430
    },
    {
      "entropy": 0.9013715043663979,
      "epoch": 0.7392259334999621,
      "grad_norm": 0.1118873581290245,
      "learning_rate": 9.850565428109854e-05,
      "loss": 1.304,
      "mean_token_accuracy": 0.7534702464938163,
      "num_tokens": 164872192.0,
      "step": 2440
    },
    {
      "entropy": 0.896777106821537,
      "epoch": 0.7422555479815194,
      "grad_norm": 0.1089012399315834,
      "learning_rate": 9.890953150242327e-05,
      "loss": 1.3014,
      "mean_token_accuracy": 0.7518857836723327,
      "num_tokens": 165552101.0,
      "step": 2450
    },
    {
      "entropy": 0.900914216041565,
      "epoch": 0.7452851624630765,
      "grad_norm": 0.11077170819044113,
      "learning_rate": 9.931340872374798e-05,
      "loss": 1.3123,
      "mean_token_accuracy": 0.7488637402653694,
      "num_tokens": 166209905.0,
      "step": 2460
    },
    {
      "entropy": 0.8996037840843201,
      "epoch": 0.7483147769446338,
      "grad_norm": 0.1135358139872551,
      "learning_rate": 9.971728594507271e-05,
      "loss": 1.3063,
      "mean_token_accuracy": 0.7565658450126648,
      "num_tokens": 166896674.0,
      "step": 2470
    },
    {
      "entropy": 0.8926998823881149,
      "epoch": 0.751344391426191,
      "grad_norm": 0.13807424902915955,
      "learning_rate": 9.99999989963874e-05,
      "loss": 1.2951,
      "mean_token_accuracy": 0.7562185510993004,
      "num_tokens": 167580000.0,
      "step": 2480
    },
    {
      "entropy": 0.895380625128746,
      "epoch": 0.7543740059077483,
      "grad_norm": 0.10803461819887161,
      "learning_rate": 9.999998115438692e-05,
      "loss": 1.3077,
      "mean_token_accuracy": 0.7538283810019493,
      "num_tokens": 168253653.0,
      "step": 2490
    },
    {
      "entropy": 0.8799598172307015,
      "epoch": 0.7574036203893054,
      "grad_norm": 0.11616635322570801,
      "learning_rate": 9.999994100989359e-05,
      "loss": 1.2896,
      "mean_token_accuracy": 0.756775687634945,
      "num_tokens": 168934649.0,
      "step": 2500
    },
    {
      "entropy": 0.8974808976054192,
      "epoch": 0.7604332348708627,
      "grad_norm": 0.10577427595853806,
      "learning_rate": 9.99998785629253e-05,
      "loss": 1.302,
      "mean_token_accuracy": 0.7530675888061523,
      "num_tokens": 169623256.0,
      "step": 2510
    },
    {
      "entropy": 0.8934375584125519,
      "epoch": 0.7634628493524199,
      "grad_norm": 0.11860733479261398,
      "learning_rate": 9.999979381350992e-05,
      "loss": 1.2971,
      "mean_token_accuracy": 0.7546120777726173,
      "num_tokens": 170305389.0,
      "step": 2520
    },
    {
      "entropy": 0.8967828616499901,
      "epoch": 0.7664924638339772,
      "grad_norm": 0.10974089056253433,
      "learning_rate": 9.999968676168526e-05,
      "loss": 1.308,
      "mean_token_accuracy": 0.7479175120592118,
      "num_tokens": 170971417.0,
      "step": 2530
    },
    {
      "entropy": 0.9001986250281334,
      "epoch": 0.7695220783155343,
      "grad_norm": 0.13815833628177643,
      "learning_rate": 9.999955740749904e-05,
      "loss": 1.3073,
      "mean_token_accuracy": 0.7483186662197113,
      "num_tokens": 171641890.0,
      "step": 2540
    },
    {
      "entropy": 0.8926223456859589,
      "epoch": 0.7725516927970916,
      "grad_norm": 0.128375843167305,
      "learning_rate": 9.999940575100901e-05,
      "loss": 1.2979,
      "mean_token_accuracy": 0.7589505150914192,
      "num_tokens": 172338796.0,
      "step": 2550
    },
    {
      "entropy": 0.8888041824102402,
      "epoch": 0.7755813072786488,
      "grad_norm": 0.12634579837322235,
      "learning_rate": 9.999923179228277e-05,
      "loss": 1.3083,
      "mean_token_accuracy": 0.7520693674683571,
      "num_tokens": 173007762.0,
      "step": 2560
    },
    {
      "entropy": 0.8981804877519608,
      "epoch": 0.778610921760206,
      "grad_norm": 0.11115466803312302,
      "learning_rate": 9.999903553139792e-05,
      "loss": 1.3026,
      "mean_token_accuracy": 0.7505076140165329,
      "num_tokens": 173673680.0,
      "step": 2570
    },
    {
      "entropy": 0.9000903144478798,
      "epoch": 0.7816405362417632,
      "grad_norm": 0.11153290420770645,
      "learning_rate": 9.999881696844202e-05,
      "loss": 1.3021,
      "mean_token_accuracy": 0.752273240685463,
      "num_tokens": 174344011.0,
      "step": 2580
    },
    {
      "entropy": 0.8882798403501511,
      "epoch": 0.7846701507233205,
      "grad_norm": 0.1012105643749237,
      "learning_rate": 9.999857610351258e-05,
      "loss": 1.2892,
      "mean_token_accuracy": 0.754874874651432,
      "num_tokens": 175027770.0,
      "step": 2590
    },
    {
      "entropy": 0.8896907702088356,
      "epoch": 0.7876997652048777,
      "grad_norm": 0.12526828050613403,
      "learning_rate": 9.999831293671697e-05,
      "loss": 1.2868,
      "mean_token_accuracy": 0.7556891471147538,
      "num_tokens": 175708636.0,
      "step": 2600
    },
    {
      "entropy": 0.8958791553974151,
      "epoch": 0.7907293796864349,
      "grad_norm": 0.1167873814702034,
      "learning_rate": 9.999802746817263e-05,
      "loss": 1.2985,
      "mean_token_accuracy": 0.7518065556883812,
      "num_tokens": 176389737.0,
      "step": 2610
    },
    {
      "entropy": 0.8975002259016037,
      "epoch": 0.7937589941679921,
      "grad_norm": 0.12051751464605331,
      "learning_rate": 9.99977196980069e-05,
      "loss": 1.2944,
      "mean_token_accuracy": 0.7527152106165886,
      "num_tokens": 177074799.0,
      "step": 2620
    },
    {
      "entropy": 0.892160353064537,
      "epoch": 0.7967886086495494,
      "grad_norm": 0.12530362606048584,
      "learning_rate": 9.999738962635703e-05,
      "loss": 1.2975,
      "mean_token_accuracy": 0.7520674377679825,
      "num_tokens": 177748277.0,
      "step": 2630
    },
    {
      "entropy": 0.8813617646694183,
      "epoch": 0.7998182231311066,
      "grad_norm": 0.1015031561255455,
      "learning_rate": 9.999703725337024e-05,
      "loss": 1.2999,
      "mean_token_accuracy": 0.7577098682522774,
      "num_tokens": 178415853.0,
      "step": 2640
    },
    {
      "entropy": 0.8927841350436211,
      "epoch": 0.8028478376126638,
      "grad_norm": 0.12152383476495743,
      "learning_rate": 9.999666257920375e-05,
      "loss": 1.2957,
      "mean_token_accuracy": 0.757025220990181,
      "num_tokens": 179100762.0,
      "step": 2650
    },
    {
      "entropy": 0.8956415086984635,
      "epoch": 0.805877452094221,
      "grad_norm": 0.11789049953222275,
      "learning_rate": 9.999626560402464e-05,
      "loss": 1.2948,
      "mean_token_accuracy": 0.7551279291510582,
      "num_tokens": 179787685.0,
      "step": 2660
    },
    {
      "entropy": 0.8890686973929405,
      "epoch": 0.8089070665757783,
      "grad_norm": 0.10717087239027023,
      "learning_rate": 9.999584632801001e-05,
      "loss": 1.2958,
      "mean_token_accuracy": 0.7556095629930496,
      "num_tokens": 180464759.0,
      "step": 2670
    },
    {
      "entropy": 0.8760408699512482,
      "epoch": 0.8119366810573354,
      "grad_norm": 0.12731553614139557,
      "learning_rate": 9.999540475134687e-05,
      "loss": 1.2881,
      "mean_token_accuracy": 0.7579726591706276,
      "num_tokens": 181146753.0,
      "step": 2680
    },
    {
      "entropy": 0.8964680597186089,
      "epoch": 0.8149662955388927,
      "grad_norm": 0.11903461068868637,
      "learning_rate": 9.999494087423218e-05,
      "loss": 1.2968,
      "mean_token_accuracy": 0.7534425109624863,
      "num_tokens": 181818721.0,
      "step": 2690
    },
    {
      "entropy": 0.899785116314888,
      "epoch": 0.8179959100204499,
      "grad_norm": 0.0975867435336113,
      "learning_rate": 9.999445469687285e-05,
      "loss": 1.3025,
      "mean_token_accuracy": 0.7537168964743615,
      "num_tokens": 182501311.0,
      "step": 2700
    },
    {
      "entropy": 0.8866352945566177,
      "epoch": 0.8210255245020072,
      "grad_norm": 0.1379888355731964,
      "learning_rate": 9.999394621948574e-05,
      "loss": 1.3041,
      "mean_token_accuracy": 0.7531851753592491,
      "num_tokens": 183176822.0,
      "step": 2710
    },
    {
      "entropy": 0.8942874193191528,
      "epoch": 0.8240551389835643,
      "grad_norm": 0.1034228503704071,
      "learning_rate": 9.999341544229767e-05,
      "loss": 1.2937,
      "mean_token_accuracy": 0.7550044193863868,
      "num_tokens": 183859190.0,
      "step": 2720
    },
    {
      "entropy": 0.8920839637517929,
      "epoch": 0.8270847534651216,
      "grad_norm": 0.10949212312698364,
      "learning_rate": 9.999286236554537e-05,
      "loss": 1.298,
      "mean_token_accuracy": 0.7496090158820152,
      "num_tokens": 184526568.0,
      "step": 2730
    },
    {
      "entropy": 0.8993422657251358,
      "epoch": 0.8301143679466788,
      "grad_norm": 0.11065369844436646,
      "learning_rate": 9.999228698947557e-05,
      "loss": 1.2983,
      "mean_token_accuracy": 0.749603633582592,
      "num_tokens": 185197242.0,
      "step": 2740
    },
    {
      "entropy": 0.8873072430491448,
      "epoch": 0.833143982428236,
      "grad_norm": 0.11458907276391983,
      "learning_rate": 9.99916893143449e-05,
      "loss": 1.3013,
      "mean_token_accuracy": 0.7548268765211106,
      "num_tokens": 185872647.0,
      "step": 2750
    },
    {
      "entropy": 0.9006134569644928,
      "epoch": 0.8361735969097932,
      "grad_norm": 0.12513191998004913,
      "learning_rate": 9.999106934041995e-05,
      "loss": 1.3071,
      "mean_token_accuracy": 0.7464882954955101,
      "num_tokens": 186533201.0,
      "step": 2760
    },
    {
      "entropy": 0.8941516593098641,
      "epoch": 0.8392032113913505,
      "grad_norm": 0.12465955317020416,
      "learning_rate": 9.999042706797725e-05,
      "loss": 1.3016,
      "mean_token_accuracy": 0.7457407161593437,
      "num_tokens": 187187516.0,
      "step": 2770
    },
    {
      "entropy": 0.8858958646655083,
      "epoch": 0.8422328258729077,
      "grad_norm": 0.13242392241954803,
      "learning_rate": 9.998976249730331e-05,
      "loss": 1.2983,
      "mean_token_accuracy": 0.7558173567056656,
      "num_tokens": 187861549.0,
      "step": 2780
    },
    {
      "entropy": 0.8848733186721802,
      "epoch": 0.8452624403544649,
      "grad_norm": 0.11963985860347748,
      "learning_rate": 9.998907562869456e-05,
      "loss": 1.2959,
      "mean_token_accuracy": 0.7539189040660859,
      "num_tokens": 188526055.0,
      "step": 2790
    },
    {
      "entropy": 0.8775170937180519,
      "epoch": 0.8482920548360221,
      "grad_norm": 0.11567117273807526,
      "learning_rate": 9.998836646245735e-05,
      "loss": 1.2749,
      "mean_token_accuracy": 0.7605938255786896,
      "num_tokens": 189222345.0,
      "step": 2800
    },
    {
      "entropy": 0.8879165351390839,
      "epoch": 0.8513216693175794,
      "grad_norm": 0.11390013247728348,
      "learning_rate": 9.998763499890805e-05,
      "loss": 1.293,
      "mean_token_accuracy": 0.7548198118805886,
      "num_tokens": 189896560.0,
      "step": 2810
    },
    {
      "entropy": 0.8948941051959991,
      "epoch": 0.8543512837991366,
      "grad_norm": 0.11404690891504288,
      "learning_rate": 9.998688123837287e-05,
      "loss": 1.303,
      "mean_token_accuracy": 0.7508691176772118,
      "num_tokens": 190566008.0,
      "step": 2820
    },
    {
      "entropy": 0.8890450417995452,
      "epoch": 0.8573808982806937,
      "grad_norm": 0.09506414085626602,
      "learning_rate": 9.998610518118806e-05,
      "loss": 1.3019,
      "mean_token_accuracy": 0.753497801721096,
      "num_tokens": 191232821.0,
      "step": 2830
    },
    {
      "entropy": 0.8918934434652328,
      "epoch": 0.860410512762251,
      "grad_norm": 0.10562010109424591,
      "learning_rate": 9.998530682769978e-05,
      "loss": 1.2944,
      "mean_token_accuracy": 0.7505758956074715,
      "num_tokens": 191906407.0,
      "step": 2840
    },
    {
      "entropy": 0.8885342419147492,
      "epoch": 0.8634401272438083,
      "grad_norm": 0.11452876776456833,
      "learning_rate": 9.998448617826413e-05,
      "loss": 1.2858,
      "mean_token_accuracy": 0.7563617274165153,
      "num_tokens": 192593029.0,
      "step": 2850
    },
    {
      "entropy": 0.8962948873639107,
      "epoch": 0.8664697417253654,
      "grad_norm": 0.10869462043046951,
      "learning_rate": 9.998364323324716e-05,
      "loss": 1.2989,
      "mean_token_accuracy": 0.7537289977073669,
      "num_tokens": 193271425.0,
      "step": 2860
    },
    {
      "entropy": 0.89390327334404,
      "epoch": 0.8694993562069226,
      "grad_norm": 0.12010079622268677,
      "learning_rate": 9.99827779930249e-05,
      "loss": 1.3072,
      "mean_token_accuracy": 0.7536683186888695,
      "num_tokens": 193948533.0,
      "step": 2870
    },
    {
      "entropy": 0.9021015837788582,
      "epoch": 0.8725289706884799,
      "grad_norm": 0.103371761739254,
      "learning_rate": 9.998189045798322e-05,
      "loss": 1.2982,
      "mean_token_accuracy": 0.7517461240291595,
      "num_tokens": 194625948.0,
      "step": 2880
    },
    {
      "entropy": 0.8892711877822876,
      "epoch": 0.8755585851700372,
      "grad_norm": 0.09570302069187164,
      "learning_rate": 9.998098062851805e-05,
      "loss": 1.3046,
      "mean_token_accuracy": 0.7510400608181953,
      "num_tokens": 195288381.0,
      "step": 2890
    },
    {
      "entropy": 0.8950867876410484,
      "epoch": 0.8785881996515943,
      "grad_norm": 0.11434589326381683,
      "learning_rate": 9.998004850503522e-05,
      "loss": 1.2964,
      "mean_token_accuracy": 0.7557195812463761,
      "num_tokens": 195976557.0,
      "step": 2900
    },
    {
      "entropy": 0.8812711849808693,
      "epoch": 0.8816178141331515,
      "grad_norm": 0.1001354306936264,
      "learning_rate": 9.99790940879505e-05,
      "loss": 1.2874,
      "mean_token_accuracy": 0.7547182634472847,
      "num_tokens": 196643437.0,
      "step": 2910
    },
    {
      "entropy": 0.8998318642377854,
      "epoch": 0.8846474286147088,
      "grad_norm": 0.10829000174999237,
      "learning_rate": 9.99781173776896e-05,
      "loss": 1.308,
      "mean_token_accuracy": 0.7503742709755897,
      "num_tokens": 197309401.0,
      "step": 2920
    },
    {
      "entropy": 0.888218542933464,
      "epoch": 0.887677043096266,
      "grad_norm": 0.1064077690243721,
      "learning_rate": 9.997711837468818e-05,
      "loss": 1.2828,
      "mean_token_accuracy": 0.7523164615035057,
      "num_tokens": 197975801.0,
      "step": 2930
    },
    {
      "entropy": 0.9008451148867607,
      "epoch": 0.8907066575778232,
      "grad_norm": 0.11014664173126221,
      "learning_rate": 9.997609707939187e-05,
      "loss": 1.2933,
      "mean_token_accuracy": 0.7497159570455552,
      "num_tokens": 198644684.0,
      "step": 2940
    },
    {
      "entropy": 0.8830799236893654,
      "epoch": 0.8937362720593804,
      "grad_norm": 0.12068645656108856,
      "learning_rate": 9.997505349225617e-05,
      "loss": 1.2909,
      "mean_token_accuracy": 0.7548075884580612,
      "num_tokens": 199317340.0,
      "step": 2950
    },
    {
      "entropy": 0.8823300585150718,
      "epoch": 0.8967658865409377,
      "grad_norm": 0.10538719594478607,
      "learning_rate": 9.997398761374663e-05,
      "loss": 1.2899,
      "mean_token_accuracy": 0.7563262611627579,
      "num_tokens": 199999874.0,
      "step": 2960
    },
    {
      "entropy": 0.8856275871396064,
      "epoch": 0.8997955010224948,
      "grad_norm": 0.1214800700545311,
      "learning_rate": 9.997289944433864e-05,
      "loss": 1.2885,
      "mean_token_accuracy": 0.7522132501006127,
      "num_tokens": 200669314.0,
      "step": 2970
    },
    {
      "entropy": 0.8952860370278358,
      "epoch": 0.9028251155040521,
      "grad_norm": 0.1363692432641983,
      "learning_rate": 9.997178898451761e-05,
      "loss": 1.3036,
      "mean_token_accuracy": 0.7530059188604354,
      "num_tokens": 201345860.0,
      "step": 2980
    },
    {
      "entropy": 0.8916691571474076,
      "epoch": 0.9058547299856093,
      "grad_norm": 0.11812197417020798,
      "learning_rate": 9.997065623477883e-05,
      "loss": 1.3088,
      "mean_token_accuracy": 0.7496755793690681,
      "num_tokens": 202008085.0,
      "step": 2990
    },
    {
      "entropy": 0.8959484532475471,
      "epoch": 0.9088843444671666,
      "grad_norm": 0.10616706311702728,
      "learning_rate": 9.996950119562758e-05,
      "loss": 1.301,
      "mean_token_accuracy": 0.7539889186620712,
      "num_tokens": 202686534.0,
      "step": 3000
    },
    {
      "entropy": 0.8973194777965545,
      "epoch": 0.9119139589487237,
      "grad_norm": 0.11548098176717758,
      "learning_rate": 9.996832386757908e-05,
      "loss": 1.2957,
      "mean_token_accuracy": 0.7508701875805854,
      "num_tokens": 203366790.0,
      "step": 3010
    },
    {
      "entropy": 0.8855000004172325,
      "epoch": 0.914943573430281,
      "grad_norm": 0.11258887499570847,
      "learning_rate": 9.996712425115845e-05,
      "loss": 1.2915,
      "mean_token_accuracy": 0.753988440334797,
      "num_tokens": 204037411.0,
      "step": 3020
    },
    {
      "entropy": 0.8910848885774613,
      "epoch": 0.9179731879118382,
      "grad_norm": 0.10067538172006607,
      "learning_rate": 9.99659023469008e-05,
      "loss": 1.2913,
      "mean_token_accuracy": 0.7553201824426651,
      "num_tokens": 204727212.0,
      "step": 3030
    },
    {
      "entropy": 0.8893559783697128,
      "epoch": 0.9210028023933955,
      "grad_norm": 0.1173923909664154,
      "learning_rate": 9.996465815535114e-05,
      "loss": 1.2951,
      "mean_token_accuracy": 0.7548139676451683,
      "num_tokens": 205410370.0,
      "step": 3040
    },
    {
      "entropy": 0.890807381272316,
      "epoch": 0.9240324168749526,
      "grad_norm": 0.10941410809755325,
      "learning_rate": 9.996339167706446e-05,
      "loss": 1.3003,
      "mean_token_accuracy": 0.7539852261543274,
      "num_tokens": 206082283.0,
      "step": 3050
    },
    {
      "entropy": 0.8826736882328987,
      "epoch": 0.9270620313565099,
      "grad_norm": 0.10980314761400223,
      "learning_rate": 9.996210291260565e-05,
      "loss": 1.2864,
      "mean_token_accuracy": 0.753743426501751,
      "num_tokens": 206761589.0,
      "step": 3060
    },
    {
      "entropy": 0.8879106298089028,
      "epoch": 0.9300916458380671,
      "grad_norm": 0.11397937685251236,
      "learning_rate": 9.996079186254958e-05,
      "loss": 1.3003,
      "mean_token_accuracy": 0.754043634235859,
      "num_tokens": 207439302.0,
      "step": 3070
    },
    {
      "entropy": 0.8939175412058831,
      "epoch": 0.9331212603196243,
      "grad_norm": 0.09899991005659103,
      "learning_rate": 9.995945852748106e-05,
      "loss": 1.3063,
      "mean_token_accuracy": 0.7534767419099808,
      "num_tokens": 208114788.0,
      "step": 3080
    },
    {
      "entropy": 0.8829623639583588,
      "epoch": 0.9361508748011815,
      "grad_norm": 0.11054675281047821,
      "learning_rate": 9.99581029079948e-05,
      "loss": 1.2891,
      "mean_token_accuracy": 0.7545877754688263,
      "num_tokens": 208789598.0,
      "step": 3090
    },
    {
      "entropy": 0.8849677100777626,
      "epoch": 0.9391804892827388,
      "grad_norm": 0.09774205088615417,
      "learning_rate": 9.995672500469548e-05,
      "loss": 1.2924,
      "mean_token_accuracy": 0.7535423532128334,
      "num_tokens": 209459851.0,
      "step": 3100
    },
    {
      "entropy": 0.8895751640200615,
      "epoch": 0.942210103764296,
      "grad_norm": 0.12433359771966934,
      "learning_rate": 9.995532481819772e-05,
      "loss": 1.2913,
      "mean_token_accuracy": 0.7532431796193123,
      "num_tokens": 210127795.0,
      "step": 3110
    },
    {
      "entropy": 0.881451603770256,
      "epoch": 0.9452397182458532,
      "grad_norm": 0.10106968879699707,
      "learning_rate": 9.995390234912605e-05,
      "loss": 1.2983,
      "mean_token_accuracy": 0.7505646288394928,
      "num_tokens": 210784502.0,
      "step": 3120
    },
    {
      "entropy": 0.8905109122395516,
      "epoch": 0.9482693327274104,
      "grad_norm": 0.10910733789205551,
      "learning_rate": 9.995245759811501e-05,
      "loss": 1.2959,
      "mean_token_accuracy": 0.7532315716147423,
      "num_tokens": 211464326.0,
      "step": 3130
    },
    {
      "entropy": 0.8880620300769806,
      "epoch": 0.9512989472089677,
      "grad_norm": 0.11476880311965942,
      "learning_rate": 9.995099056580896e-05,
      "loss": 1.2973,
      "mean_token_accuracy": 0.755242583155632,
      "num_tokens": 212137201.0,
      "step": 3140
    },
    {
      "entropy": 0.8845109969377518,
      "epoch": 0.9543285616905249,
      "grad_norm": 0.11044140160083771,
      "learning_rate": 9.994950125286234e-05,
      "loss": 1.2874,
      "mean_token_accuracy": 0.751310071349144,
      "num_tokens": 212796724.0,
      "step": 3150
    },
    {
      "entropy": 0.890080326795578,
      "epoch": 0.9573581761720821,
      "grad_norm": 0.1296996921300888,
      "learning_rate": 9.994798965993943e-05,
      "loss": 1.2968,
      "mean_token_accuracy": 0.7520108208060264,
      "num_tokens": 213457695.0,
      "step": 3160
    },
    {
      "entropy": 0.8923839345574379,
      "epoch": 0.9603877906536393,
      "grad_norm": 0.10951551795005798,
      "learning_rate": 9.994645578771446e-05,
      "loss": 1.2965,
      "mean_token_accuracy": 0.7527667090296746,
      "num_tokens": 214126561.0,
      "step": 3170
    },
    {
      "entropy": 0.8904188767075538,
      "epoch": 0.9634174051351966,
      "grad_norm": 0.10850483179092407,
      "learning_rate": 9.994489963687163e-05,
      "loss": 1.2898,
      "mean_token_accuracy": 0.7545693084597588,
      "num_tokens": 214813344.0,
      "step": 3180
    },
    {
      "entropy": 0.8910052701830864,
      "epoch": 0.9664470196167537,
      "grad_norm": 0.11159470677375793,
      "learning_rate": 9.994332120810508e-05,
      "loss": 1.2982,
      "mean_token_accuracy": 0.7537946432828904,
      "num_tokens": 215498268.0,
      "step": 3190
    },
    {
      "entropy": 0.896054194867611,
      "epoch": 0.969476634098311,
      "grad_norm": 0.11511316150426865,
      "learning_rate": 9.994172050211883e-05,
      "loss": 1.3002,
      "mean_token_accuracy": 0.7577290192246438,
      "num_tokens": 216189675.0,
      "step": 3200
    },
    {
      "entropy": 0.8780703529715538,
      "epoch": 0.9725062485798682,
      "grad_norm": 0.10506260395050049,
      "learning_rate": 9.99400975196269e-05,
      "loss": 1.2887,
      "mean_token_accuracy": 0.7529979765415191,
      "num_tokens": 216852550.0,
      "step": 3210
    },
    {
      "entropy": 0.8913513898849488,
      "epoch": 0.9755358630614255,
      "grad_norm": 0.10629456490278244,
      "learning_rate": 9.993845226135322e-05,
      "loss": 1.2926,
      "mean_token_accuracy": 0.7555703178048134,
      "num_tokens": 217539292.0,
      "step": 3220
    },
    {
      "entropy": 0.8858451247215271,
      "epoch": 0.9785654775429826,
      "grad_norm": 0.12247616052627563,
      "learning_rate": 9.993678472803165e-05,
      "loss": 1.2882,
      "mean_token_accuracy": 0.7532886505126953,
      "num_tokens": 218206220.0,
      "step": 3230
    },
    {
      "entropy": 0.894928352534771,
      "epoch": 0.9815950920245399,
      "grad_norm": 0.10783626139163971,
      "learning_rate": 9.993509492040598e-05,
      "loss": 1.2993,
      "mean_token_accuracy": 0.7458206444978714,
      "num_tokens": 218859599.0,
      "step": 3240
    },
    {
      "entropy": 0.8780856564640999,
      "epoch": 0.9846247065060971,
      "grad_norm": 0.11459079384803772,
      "learning_rate": 9.993338283922998e-05,
      "loss": 1.2793,
      "mean_token_accuracy": 0.7525880768895149,
      "num_tokens": 219529665.0,
      "step": 3250
    },
    {
      "entropy": 0.8861164033412934,
      "epoch": 0.9876543209876543,
      "grad_norm": 0.11810479313135147,
      "learning_rate": 9.99316484852673e-05,
      "loss": 1.2921,
      "mean_token_accuracy": 0.7580062180757523,
      "num_tokens": 220212376.0,
      "step": 3260
    },
    {
      "entropy": 0.8857352554798126,
      "epoch": 0.9906839354692115,
      "grad_norm": 0.12247594445943832,
      "learning_rate": 9.992989185929156e-05,
      "loss": 1.2957,
      "mean_token_accuracy": 0.7553387239575386,
      "num_tokens": 220892023.0,
      "step": 3270
    },
    {
      "entropy": 0.8915065169334412,
      "epoch": 0.9937135499507688,
      "grad_norm": 0.10942309349775314,
      "learning_rate": 9.992811296208631e-05,
      "loss": 1.2972,
      "mean_token_accuracy": 0.7555397480726243,
      "num_tokens": 221575451.0,
      "step": 3280
    },
    {
      "entropy": 0.8895631492137909,
      "epoch": 0.996743164432326,
      "grad_norm": 0.10511741042137146,
      "learning_rate": 9.9926311794445e-05,
      "loss": 1.3013,
      "mean_token_accuracy": 0.7511893466114998,
      "num_tokens": 222243998.0,
      "step": 3290
    },
    {
      "entropy": 0.8838838443160058,
      "epoch": 0.9997727789138832,
      "grad_norm": 0.11750248819589615,
      "learning_rate": 9.992448835717108e-05,
      "loss": 1.2903,
      "mean_token_accuracy": 0.7565528631210328,
      "num_tokens": 222926827.0,
      "step": 3300
    },
    {
      "entropy": 0.8911170256443512,
      "epoch": 1.0027266530334016,
      "grad_norm": 0.10560097545385361,
      "learning_rate": 9.992264265107784e-05,
      "loss": 1.3018,
      "mean_token_accuracy": 0.7522377738585839,
      "num_tokens": 223577278.0,
      "step": 3310
    },
    {
      "entropy": 0.8920167312026024,
      "epoch": 1.0057562675149587,
      "grad_norm": 0.1127898320555687,
      "learning_rate": 9.992077467698862e-05,
      "loss": 1.2945,
      "mean_token_accuracy": 0.7525437578558922,
      "num_tokens": 224251920.0,
      "step": 3320
    },
    {
      "entropy": 0.8868375703692436,
      "epoch": 1.0087858819965159,
      "grad_norm": 0.10572829842567444,
      "learning_rate": 9.991888443573658e-05,
      "loss": 1.2903,
      "mean_token_accuracy": 0.7567735984921455,
      "num_tokens": 224929501.0,
      "step": 3330
    },
    {
      "entropy": 0.8888847678899765,
      "epoch": 1.0118154964780732,
      "grad_norm": 0.11192059516906738,
      "learning_rate": 9.991697192816489e-05,
      "loss": 1.2928,
      "mean_token_accuracy": 0.7525388732552528,
      "num_tokens": 225603557.0,
      "step": 3340
    },
    {
      "entropy": 0.8859956607222557,
      "epoch": 1.0148451109596304,
      "grad_norm": 0.1040537878870964,
      "learning_rate": 9.99150371551266e-05,
      "loss": 1.2912,
      "mean_token_accuracy": 0.7513513550162315,
      "num_tokens": 226274502.0,
      "step": 3350
    },
    {
      "entropy": 0.8881052494049072,
      "epoch": 1.0178747254411875,
      "grad_norm": 0.10165461897850037,
      "learning_rate": 9.991308011748475e-05,
      "loss": 1.2961,
      "mean_token_accuracy": 0.7474104389548302,
      "num_tokens": 226935109.0,
      "step": 3360
    },
    {
      "entropy": 0.871713387966156,
      "epoch": 1.0209043399227449,
      "grad_norm": 0.11055523157119751,
      "learning_rate": 9.991110081611225e-05,
      "loss": 1.2756,
      "mean_token_accuracy": 0.7542689144611359,
      "num_tokens": 227608243.0,
      "step": 3370
    },
    {
      "entropy": 0.8877625584602356,
      "epoch": 1.023933954404302,
      "grad_norm": 0.10797322541475296,
      "learning_rate": 9.990909925189196e-05,
      "loss": 1.2931,
      "mean_token_accuracy": 0.7551879778504371,
      "num_tokens": 228289728.0,
      "step": 3380
    },
    {
      "entropy": 0.8882021844387055,
      "epoch": 1.0269635688858594,
      "grad_norm": 0.11925064772367477,
      "learning_rate": 9.990707542571672e-05,
      "loss": 1.2914,
      "mean_token_accuracy": 0.7551611572504043,
      "num_tokens": 228977061.0,
      "step": 3390
    },
    {
      "entropy": 0.8877963319420814,
      "epoch": 1.0299931833674165,
      "grad_norm": 0.0958029180765152,
      "learning_rate": 9.990502933848923e-05,
      "loss": 1.2821,
      "mean_token_accuracy": 0.7543014630675315,
      "num_tokens": 229661460.0,
      "step": 3400
    },
    {
      "entropy": 0.8862324357032776,
      "epoch": 1.0330227978489737,
      "grad_norm": 0.10870645195245743,
      "learning_rate": 9.990296099112213e-05,
      "loss": 1.2945,
      "mean_token_accuracy": 0.7560849383473396,
      "num_tokens": 230351187.0,
      "step": 3410
    },
    {
      "entropy": 0.8828345820307731,
      "epoch": 1.036052412330531,
      "grad_norm": 0.10174642503261566,
      "learning_rate": 9.990087038453802e-05,
      "loss": 1.2788,
      "mean_token_accuracy": 0.7653154522180557,
      "num_tokens": 231066749.0,
      "step": 3420
    },
    {
      "entropy": 0.8851967990398407,
      "epoch": 1.0390820268120882,
      "grad_norm": 0.11057489365339279,
      "learning_rate": 9.989875751966944e-05,
      "loss": 1.2966,
      "mean_token_accuracy": 0.754563856124878,
      "num_tokens": 231741660.0,
      "step": 3430
    },
    {
      "entropy": 0.8901623770594597,
      "epoch": 1.0421116412936453,
      "grad_norm": 0.1161167174577713,
      "learning_rate": 9.98966223974588e-05,
      "loss": 1.3022,
      "mean_token_accuracy": 0.7524377182126045,
      "num_tokens": 232405078.0,
      "step": 3440
    },
    {
      "entropy": 0.8907693952322007,
      "epoch": 1.0451412557752027,
      "grad_norm": 0.09815775603055954,
      "learning_rate": 9.989446501885848e-05,
      "loss": 1.298,
      "mean_token_accuracy": 0.7495367169380188,
      "num_tokens": 233070740.0,
      "step": 3450
    },
    {
      "entropy": 0.8830181643366813,
      "epoch": 1.0481708702567598,
      "grad_norm": 0.09803644567728043,
      "learning_rate": 9.989228538483078e-05,
      "loss": 1.2939,
      "mean_token_accuracy": 0.7502244040369987,
      "num_tokens": 233727328.0,
      "step": 3460
    },
    {
      "entropy": 0.8811929568648338,
      "epoch": 1.051200484738317,
      "grad_norm": 0.09551838785409927,
      "learning_rate": 9.989008349634795e-05,
      "loss": 1.2815,
      "mean_token_accuracy": 0.7551064103841781,
      "num_tokens": 234399755.0,
      "step": 3470
    },
    {
      "entropy": 0.8824715837836266,
      "epoch": 1.0542300992198743,
      "grad_norm": 0.10890055447816849,
      "learning_rate": 9.98878593543921e-05,
      "loss": 1.2858,
      "mean_token_accuracy": 0.7526574030518531,
      "num_tokens": 235078669.0,
      "step": 3480
    },
    {
      "entropy": 0.8932882875204087,
      "epoch": 1.0572597137014315,
      "grad_norm": 0.10177361220121384,
      "learning_rate": 9.988561295995535e-05,
      "loss": 1.2986,
      "mean_token_accuracy": 0.7547635450959206,
      "num_tokens": 235772791.0,
      "step": 3490
    },
    {
      "entropy": 0.8782318502664566,
      "epoch": 1.0602893281829888,
      "grad_norm": 0.10841168463230133,
      "learning_rate": 9.988334431403966e-05,
      "loss": 1.2894,
      "mean_token_accuracy": 0.7566312372684478,
      "num_tokens": 236441005.0,
      "step": 3500
    },
    {
      "entropy": 0.8738766640424729,
      "epoch": 1.063318942664546,
      "grad_norm": 0.13350026309490204,
      "learning_rate": 9.9881053417657e-05,
      "loss": 1.2844,
      "mean_token_accuracy": 0.7570654019713402,
      "num_tokens": 237117858.0,
      "step": 3510
    },
    {
      "entropy": 0.8770181387662888,
      "epoch": 1.066348557146103,
      "grad_norm": 0.10082123428583145,
      "learning_rate": 9.98787402718292e-05,
      "loss": 1.2854,
      "mean_token_accuracy": 0.7538925632834435,
      "num_tokens": 237790207.0,
      "step": 3520
    },
    {
      "entropy": 0.8860610723495483,
      "epoch": 1.0693781716276605,
      "grad_norm": 0.12682177126407623,
      "learning_rate": 9.987640487758804e-05,
      "loss": 1.2883,
      "mean_token_accuracy": 0.7573604539036751,
      "num_tokens": 238473396.0,
      "step": 3530
    },
    {
      "entropy": 0.8924138516187667,
      "epoch": 1.0724077861092176,
      "grad_norm": 0.10974256694316864,
      "learning_rate": 9.987404723597525e-05,
      "loss": 1.3031,
      "mean_token_accuracy": 0.7491044074296951,
      "num_tokens": 239126306.0,
      "step": 3540
    },
    {
      "entropy": 0.8796985596418381,
      "epoch": 1.0754374005907748,
      "grad_norm": 0.12989287078380585,
      "learning_rate": 9.987166734804243e-05,
      "loss": 1.2775,
      "mean_token_accuracy": 0.7557509571313858,
      "num_tokens": 239799162.0,
      "step": 3550
    },
    {
      "entropy": 0.8759955242276192,
      "epoch": 1.0784670150723321,
      "grad_norm": 0.10880783945322037,
      "learning_rate": 9.986926521485111e-05,
      "loss": 1.2795,
      "mean_token_accuracy": 0.7552510201931,
      "num_tokens": 240477949.0,
      "step": 3560
    },
    {
      "entropy": 0.8844213604927063,
      "epoch": 1.0814966295538893,
      "grad_norm": 0.10586912930011749,
      "learning_rate": 9.986684083747282e-05,
      "loss": 1.2892,
      "mean_token_accuracy": 0.749414810538292,
      "num_tokens": 241136134.0,
      "step": 3570
    },
    {
      "entropy": 0.8810238495469094,
      "epoch": 1.0845262440354464,
      "grad_norm": 0.11063455790281296,
      "learning_rate": 9.98643942169889e-05,
      "loss": 1.2898,
      "mean_token_accuracy": 0.7572943776845932,
      "num_tokens": 241814428.0,
      "step": 3580
    },
    {
      "entropy": 0.8795574530959129,
      "epoch": 1.0875558585170038,
      "grad_norm": 0.17099305987358093,
      "learning_rate": 9.986192535449068e-05,
      "loss": 1.2942,
      "mean_token_accuracy": 0.7528563812375069,
      "num_tokens": 242486061.0,
      "step": 3590
    },
    {
      "entropy": 0.8812819436192513,
      "epoch": 1.090585472998561,
      "grad_norm": 0.11202829331159592,
      "learning_rate": 9.985943425107943e-05,
      "loss": 1.297,
      "mean_token_accuracy": 0.7580265551805496,
      "num_tokens": 243170535.0,
      "step": 3600
    },
    {
      "entropy": 0.8764916747808457,
      "epoch": 1.093615087480118,
      "grad_norm": 0.1257563978433609,
      "learning_rate": 9.985692090786624e-05,
      "loss": 1.2758,
      "mean_token_accuracy": 0.7582903057336807,
      "num_tokens": 243865679.0,
      "step": 3610
    },
    {
      "entropy": 0.8862913087010383,
      "epoch": 1.0966447019616754,
      "grad_norm": 0.11346849799156189,
      "learning_rate": 9.985438532597224e-05,
      "loss": 1.2847,
      "mean_token_accuracy": 0.7537027999758721,
      "num_tokens": 244537231.0,
      "step": 3620
    },
    {
      "entropy": 0.8843289449810982,
      "epoch": 1.0996743164432325,
      "grad_norm": 0.10032348334789276,
      "learning_rate": 9.985182750652842e-05,
      "loss": 1.2861,
      "mean_token_accuracy": 0.755133393406868,
      "num_tokens": 245219847.0,
      "step": 3630
    },
    {
      "entropy": 0.887570683658123,
      "epoch": 1.10270393092479,
      "grad_norm": 0.11875443160533905,
      "learning_rate": 9.984924745067567e-05,
      "loss": 1.29,
      "mean_token_accuracy": 0.7504461675882339,
      "num_tokens": 245884554.0,
      "step": 3640
    },
    {
      "entropy": 0.8764506101608276,
      "epoch": 1.105733545406347,
      "grad_norm": 0.10695013403892517,
      "learning_rate": 9.984664515956486e-05,
      "loss": 1.2879,
      "mean_token_accuracy": 0.7540923699736595,
      "num_tokens": 246553745.0,
      "step": 3650
    },
    {
      "entropy": 0.889299499988556,
      "epoch": 1.1087631598879042,
      "grad_norm": 0.12081286311149597,
      "learning_rate": 9.98440206343567e-05,
      "loss": 1.2925,
      "mean_token_accuracy": 0.7518796786665917,
      "num_tokens": 247222648.0,
      "step": 3660
    },
    {
      "entropy": 0.874249455332756,
      "epoch": 1.1117927743694616,
      "grad_norm": 0.10921664535999298,
      "learning_rate": 9.98413738762219e-05,
      "loss": 1.2799,
      "mean_token_accuracy": 0.7637021541595459,
      "num_tokens": 247926515.0,
      "step": 3670
    },
    {
      "entropy": 0.8825855419039726,
      "epoch": 1.1148223888510187,
      "grad_norm": 0.10977496951818466,
      "learning_rate": 9.983870488634102e-05,
      "loss": 1.2902,
      "mean_token_accuracy": 0.7545716062188148,
      "num_tokens": 248605455.0,
      "step": 3680
    },
    {
      "entropy": 0.8877398714423179,
      "epoch": 1.1178520033325758,
      "grad_norm": 0.12345528602600098,
      "learning_rate": 9.983601366590457e-05,
      "loss": 1.2936,
      "mean_token_accuracy": 0.7516496270895004,
      "num_tokens": 249286066.0,
      "step": 3690
    },
    {
      "entropy": 0.878480176627636,
      "epoch": 1.1208816178141332,
      "grad_norm": 0.11568955332040787,
      "learning_rate": 9.983330021611297e-05,
      "loss": 1.2833,
      "mean_token_accuracy": 0.7629318952560424,
      "num_tokens": 249967062.0,
      "step": 3700
    },
    {
      "entropy": 0.8880110561847687,
      "epoch": 1.1239112322956903,
      "grad_norm": 0.125409334897995,
      "learning_rate": 9.983056453817657e-05,
      "loss": 1.2921,
      "mean_token_accuracy": 0.7574349761009216,
      "num_tokens": 250657458.0,
      "step": 3710
    },
    {
      "entropy": 0.8835234329104423,
      "epoch": 1.1269408467772477,
      "grad_norm": 0.10620824247598648,
      "learning_rate": 9.98278066333156e-05,
      "loss": 1.2856,
      "mean_token_accuracy": 0.7525221645832062,
      "num_tokens": 251325229.0,
      "step": 3720
    },
    {
      "entropy": 0.8757663741707802,
      "epoch": 1.1299704612588048,
      "grad_norm": 0.1087966337800026,
      "learning_rate": 9.98250265027602e-05,
      "loss": 1.2781,
      "mean_token_accuracy": 0.7562583640217782,
      "num_tokens": 252005769.0,
      "step": 3730
    },
    {
      "entropy": 0.8799458980560303,
      "epoch": 1.133000075740362,
      "grad_norm": 0.09769339114427567,
      "learning_rate": 9.982222414775048e-05,
      "loss": 1.2847,
      "mean_token_accuracy": 0.7568388402462005,
      "num_tokens": 252684613.0,
      "step": 3740
    },
    {
      "entropy": 0.8838448315858841,
      "epoch": 1.1360296902219194,
      "grad_norm": 0.12134331464767456,
      "learning_rate": 9.981939956953645e-05,
      "loss": 1.2864,
      "mean_token_accuracy": 0.7574134305119514,
      "num_tokens": 253374054.0,
      "step": 3750
    },
    {
      "entropy": 0.8901113018393516,
      "epoch": 1.1390593047034765,
      "grad_norm": 0.10419149696826935,
      "learning_rate": 9.981655276937798e-05,
      "loss": 1.2892,
      "mean_token_accuracy": 0.753422400355339,
      "num_tokens": 254052098.0,
      "step": 3760
    },
    {
      "entropy": 0.8802848294377327,
      "epoch": 1.1420889191850336,
      "grad_norm": 0.09308606386184692,
      "learning_rate": 9.98136837485449e-05,
      "loss": 1.2817,
      "mean_token_accuracy": 0.7608528643846512,
      "num_tokens": 254741940.0,
      "step": 3770
    },
    {
      "entropy": 0.8828628584742546,
      "epoch": 1.145118533666591,
      "grad_norm": 0.10295186936855316,
      "learning_rate": 9.981079250831691e-05,
      "loss": 1.2906,
      "mean_token_accuracy": 0.7555951252579689,
      "num_tokens": 255423370.0,
      "step": 3780
    },
    {
      "entropy": 0.8731147646903992,
      "epoch": 1.1481481481481481,
      "grad_norm": 0.11505525559186935,
      "learning_rate": 9.980787904998367e-05,
      "loss": 1.2839,
      "mean_token_accuracy": 0.7563813909888267,
      "num_tokens": 256098019.0,
      "step": 3790
    },
    {
      "entropy": 0.8898579612374306,
      "epoch": 1.1511777626297053,
      "grad_norm": 0.11030592024326324,
      "learning_rate": 9.980494337484474e-05,
      "loss": 1.2995,
      "mean_token_accuracy": 0.757056450843811,
      "num_tokens": 256790873.0,
      "step": 3800
    },
    {
      "entropy": 0.8818782925605774,
      "epoch": 1.1542073771112626,
      "grad_norm": 0.10960029810667038,
      "learning_rate": 9.980198548420957e-05,
      "loss": 1.2926,
      "mean_token_accuracy": 0.7521349757909774,
      "num_tokens": 257462265.0,
      "step": 3810
    },
    {
      "entropy": 0.881763118505478,
      "epoch": 1.1572369915928198,
      "grad_norm": 0.10717824846506119,
      "learning_rate": 9.97990053793975e-05,
      "loss": 1.2885,
      "mean_token_accuracy": 0.7590022325515747,
      "num_tokens": 258140456.0,
      "step": 3820
    },
    {
      "entropy": 0.872139398753643,
      "epoch": 1.160266606074377,
      "grad_norm": 0.10643764585256577,
      "learning_rate": 9.979600306173784e-05,
      "loss": 1.2766,
      "mean_token_accuracy": 0.7581851229071617,
      "num_tokens": 258824709.0,
      "step": 3830
    },
    {
      "entropy": 0.8823701068758965,
      "epoch": 1.1632962205559343,
      "grad_norm": 0.10628998279571533,
      "learning_rate": 9.979297853256976e-05,
      "loss": 1.2849,
      "mean_token_accuracy": 0.7554057717323304,
      "num_tokens": 259502340.0,
      "step": 3840
    },
    {
      "entropy": 0.8798398494720459,
      "epoch": 1.1663258350374914,
      "grad_norm": 0.10866077244281769,
      "learning_rate": 9.978993179324235e-05,
      "loss": 1.2877,
      "mean_token_accuracy": 0.7594596952199936,
      "num_tokens": 260184864.0,
      "step": 3850
    },
    {
      "entropy": 0.8884417340159416,
      "epoch": 1.1693554495190488,
      "grad_norm": 0.11306183785200119,
      "learning_rate": 9.978686284511461e-05,
      "loss": 1.2887,
      "mean_token_accuracy": 0.7554182320833206,
      "num_tokens": 260872291.0,
      "step": 3860
    },
    {
      "entropy": 0.8777225404977799,
      "epoch": 1.172385064000606,
      "grad_norm": 0.10495192557573318,
      "learning_rate": 9.978377168955545e-05,
      "loss": 1.2848,
      "mean_token_accuracy": 0.754628412425518,
      "num_tokens": 261526593.0,
      "step": 3870
    },
    {
      "entropy": 0.876430393755436,
      "epoch": 1.175414678482163,
      "grad_norm": 0.10036977380514145,
      "learning_rate": 9.978065832794368e-05,
      "loss": 1.2938,
      "mean_token_accuracy": 0.755861134827137,
      "num_tokens": 262189896.0,
      "step": 3880
    },
    {
      "entropy": 0.883771327137947,
      "epoch": 1.1784442929637204,
      "grad_norm": 0.10100783407688141,
      "learning_rate": 9.9777522761668e-05,
      "loss": 1.2998,
      "mean_token_accuracy": 0.7537850499153137,
      "num_tokens": 262854808.0,
      "step": 3890
    },
    {
      "entropy": 0.8863219231367111,
      "epoch": 1.1814739074452776,
      "grad_norm": 0.10015401989221573,
      "learning_rate": 9.977436499212704e-05,
      "loss": 1.2849,
      "mean_token_accuracy": 0.7554245010018349,
      "num_tokens": 263541282.0,
      "step": 3900
    },
    {
      "entropy": 0.8819337621331215,
      "epoch": 1.1845035219268347,
      "grad_norm": 0.1085055023431778,
      "learning_rate": 9.977118502072933e-05,
      "loss": 1.2887,
      "mean_token_accuracy": 0.7516910418868065,
      "num_tokens": 264209728.0,
      "step": 3910
    },
    {
      "entropy": 0.8842660367488862,
      "epoch": 1.187533136408392,
      "grad_norm": 0.13629150390625,
      "learning_rate": 9.976798284889328e-05,
      "loss": 1.2934,
      "mean_token_accuracy": 0.7530283167958259,
      "num_tokens": 264878445.0,
      "step": 3920
    },
    {
      "entropy": 0.8812041372060776,
      "epoch": 1.1905627508899492,
      "grad_norm": 0.12887629866600037,
      "learning_rate": 9.976475847804723e-05,
      "loss": 1.2819,
      "mean_token_accuracy": 0.7571926325559616,
      "num_tokens": 265569481.0,
      "step": 3930
    },
    {
      "entropy": 0.8909408152103424,
      "epoch": 1.1935923653715066,
      "grad_norm": 0.10945051908493042,
      "learning_rate": 9.976151190962941e-05,
      "loss": 1.3019,
      "mean_token_accuracy": 0.7494542822241783,
      "num_tokens": 266234478.0,
      "step": 3940
    },
    {
      "entropy": 0.8925550684332848,
      "epoch": 1.1966219798530637,
      "grad_norm": 0.10394002497196198,
      "learning_rate": 9.975824314508794e-05,
      "loss": 1.2984,
      "mean_token_accuracy": 0.7525490179657937,
      "num_tokens": 266902064.0,
      "step": 3950
    },
    {
      "entropy": 0.8868040114641189,
      "epoch": 1.1996515943346209,
      "grad_norm": 0.10101783275604248,
      "learning_rate": 9.975495218588089e-05,
      "loss": 1.2952,
      "mean_token_accuracy": 0.7574144750833511,
      "num_tokens": 267586908.0,
      "step": 3960
    },
    {
      "entropy": 0.8726276144385338,
      "epoch": 1.2026812088161782,
      "grad_norm": 0.1203441247344017,
      "learning_rate": 9.975163903347615e-05,
      "loss": 1.286,
      "mean_token_accuracy": 0.7547452345490455,
      "num_tokens": 268257457.0,
      "step": 3970
    },
    {
      "entropy": 0.8927853986620903,
      "epoch": 1.2057108232977354,
      "grad_norm": 0.11372381448745728,
      "learning_rate": 9.974830368935156e-05,
      "loss": 1.2943,
      "mean_token_accuracy": 0.7555569022893905,
      "num_tokens": 268943971.0,
      "step": 3980
    },
    {
      "entropy": 0.8818222165107727,
      "epoch": 1.2087404377792925,
      "grad_norm": 0.12832969427108765,
      "learning_rate": 9.974494615499487e-05,
      "loss": 1.2887,
      "mean_token_accuracy": 0.7539437264204025,
      "num_tokens": 269618169.0,
      "step": 3990
    },
    {
      "entropy": 0.8911607503890991,
      "epoch": 1.2117700522608499,
      "grad_norm": 0.10649841278791428,
      "learning_rate": 9.974156643190369e-05,
      "loss": 1.2929,
      "mean_token_accuracy": 0.7534996464848518,
      "num_tokens": 270296734.0,
      "step": 4000
    },
    {
      "entropy": 0.8867155745625496,
      "epoch": 1.214799666742407,
      "grad_norm": 0.09304763376712799,
      "learning_rate": 9.973816452158555e-05,
      "loss": 1.2863,
      "mean_token_accuracy": 0.7531761646270752,
      "num_tokens": 270979344.0,
      "step": 4010
    },
    {
      "entropy": 0.8814450964331627,
      "epoch": 1.2178292812239642,
      "grad_norm": 0.11045564711093903,
      "learning_rate": 9.973474042555787e-05,
      "loss": 1.289,
      "mean_token_accuracy": 0.7529393047094345,
      "num_tokens": 271652049.0,
      "step": 4020
    },
    {
      "entropy": 0.8878124937415123,
      "epoch": 1.2208588957055215,
      "grad_norm": 0.10780147463083267,
      "learning_rate": 9.973129414534797e-05,
      "loss": 1.2897,
      "mean_token_accuracy": 0.7520148739218712,
      "num_tokens": 272325049.0,
      "step": 4030
    },
    {
      "entropy": 0.8735245853662491,
      "epoch": 1.2238885101870787,
      "grad_norm": 0.12177599221467972,
      "learning_rate": 9.972782568249308e-05,
      "loss": 1.2847,
      "mean_token_accuracy": 0.7628583461046219,
      "num_tokens": 273024307.0,
      "step": 4040
    },
    {
      "entropy": 0.8731584176421165,
      "epoch": 1.2269181246686358,
      "grad_norm": 0.11406262218952179,
      "learning_rate": 9.972433503854028e-05,
      "loss": 1.2906,
      "mean_token_accuracy": 0.7569164022803306,
      "num_tokens": 273691121.0,
      "step": 4050
    },
    {
      "entropy": 0.8802512750029564,
      "epoch": 1.2299477391501932,
      "grad_norm": 0.12157581746578217,
      "learning_rate": 9.972082221504659e-05,
      "loss": 1.2819,
      "mean_token_accuracy": 0.754756960272789,
      "num_tokens": 274362828.0,
      "step": 4060
    },
    {
      "entropy": 0.8889210850000382,
      "epoch": 1.2329773536317503,
      "grad_norm": 0.10852083563804626,
      "learning_rate": 9.97172872135789e-05,
      "loss": 1.2944,
      "mean_token_accuracy": 0.749279560148716,
      "num_tokens": 275024532.0,
      "step": 4070
    },
    {
      "entropy": 0.8845911890268325,
      "epoch": 1.2360069681133075,
      "grad_norm": 0.11005109548568726,
      "learning_rate": 9.971373003571401e-05,
      "loss": 1.2904,
      "mean_token_accuracy": 0.753645335137844,
      "num_tokens": 275696277.0,
      "step": 4080
    },
    {
      "entropy": 0.8746917456388473,
      "epoch": 1.2390365825948648,
      "grad_norm": 0.12576720118522644,
      "learning_rate": 9.971015068303856e-05,
      "loss": 1.2795,
      "mean_token_accuracy": 0.7542210564017295,
      "num_tokens": 276361925.0,
      "step": 4090
    },
    {
      "entropy": 0.8848961919546128,
      "epoch": 1.242066197076422,
      "grad_norm": 0.10003882646560669,
      "learning_rate": 9.970654915714917e-05,
      "loss": 1.2927,
      "mean_token_accuracy": 0.7483493506908416,
      "num_tokens": 277026539.0,
      "step": 4100
    },
    {
      "entropy": 0.8785070508718491,
      "epoch": 1.2450958115579793,
      "grad_norm": 0.09673545509576797,
      "learning_rate": 9.970292545965225e-05,
      "loss": 1.2905,
      "mean_token_accuracy": 0.7515549525618553,
      "num_tokens": 277699565.0,
      "step": 4110
    },
    {
      "entropy": 0.8770564898848534,
      "epoch": 1.2481254260395365,
      "grad_norm": 0.10010041296482086,
      "learning_rate": 9.969927959216421e-05,
      "loss": 1.2867,
      "mean_token_accuracy": 0.7587835326790809,
      "num_tokens": 278376956.0,
      "step": 4120
    },
    {
      "entropy": 0.8782692402601242,
      "epoch": 1.2511550405210936,
      "grad_norm": 0.11118870228528976,
      "learning_rate": 9.969561155631123e-05,
      "loss": 1.2828,
      "mean_token_accuracy": 0.757876068353653,
      "num_tokens": 279052134.0,
      "step": 4130
    },
    {
      "entropy": 0.8728073760867119,
      "epoch": 1.254184655002651,
      "grad_norm": 0.1019531860947609,
      "learning_rate": 9.969192135372948e-05,
      "loss": 1.2771,
      "mean_token_accuracy": 0.7562713280320168,
      "num_tokens": 279720145.0,
      "step": 4140
    },
    {
      "entropy": 0.8769888460636139,
      "epoch": 1.2572142694842081,
      "grad_norm": 0.10588269680738449,
      "learning_rate": 9.968820898606496e-05,
      "loss": 1.2882,
      "mean_token_accuracy": 0.7562667936086654,
      "num_tokens": 280396688.0,
      "step": 4150
    },
    {
      "entropy": 0.881463372707367,
      "epoch": 1.2602438839657655,
      "grad_norm": 0.11158733814954758,
      "learning_rate": 9.968447445497356e-05,
      "loss": 1.2923,
      "mean_token_accuracy": 0.7556000202894211,
      "num_tokens": 281078455.0,
      "step": 4160
    },
    {
      "entropy": 0.884491217136383,
      "epoch": 1.2632734984473226,
      "grad_norm": 0.10962716490030289,
      "learning_rate": 9.96807177621211e-05,
      "loss": 1.2855,
      "mean_token_accuracy": 0.7545291304588317,
      "num_tokens": 281752436.0,
      "step": 4170
    },
    {
      "entropy": 0.876974818110466,
      "epoch": 1.2663031129288798,
      "grad_norm": 0.11437703669071198,
      "learning_rate": 9.967693890918322e-05,
      "loss": 1.2814,
      "mean_token_accuracy": 0.7614217966794967,
      "num_tokens": 282448782.0,
      "step": 4180
    },
    {
      "entropy": 0.8695633113384247,
      "epoch": 1.2693327274104371,
      "grad_norm": 0.11076894402503967,
      "learning_rate": 9.967313789784548e-05,
      "loss": 1.2777,
      "mean_token_accuracy": 0.7562600076198578,
      "num_tokens": 283113429.0,
      "step": 4190
    },
    {
      "entropy": 0.8868823811411858,
      "epoch": 1.2723623418919943,
      "grad_norm": 0.12090124189853668,
      "learning_rate": 9.966931472980335e-05,
      "loss": 1.2951,
      "mean_token_accuracy": 0.758067575097084,
      "num_tokens": 283803378.0,
      "step": 4200
    },
    {
      "entropy": 0.8870190247893334,
      "epoch": 1.2753919563735514,
      "grad_norm": 0.1082521453499794,
      "learning_rate": 9.966546940676212e-05,
      "loss": 1.2852,
      "mean_token_accuracy": 0.756389519572258,
      "num_tokens": 284487935.0,
      "step": 4210
    },
    {
      "entropy": 0.8746539890766144,
      "epoch": 1.2784215708551088,
      "grad_norm": 0.10147053748369217,
      "learning_rate": 9.9661601930437e-05,
      "loss": 1.2789,
      "mean_token_accuracy": 0.7614048615097999,
      "num_tokens": 285170657.0,
      "step": 4220
    },
    {
      "entropy": 0.8725147917866707,
      "epoch": 1.281451185336666,
      "grad_norm": 0.11726569384336472,
      "learning_rate": 9.965771230255312e-05,
      "loss": 1.2829,
      "mean_token_accuracy": 0.7540886506438256,
      "num_tokens": 285844852.0,
      "step": 4230
    },
    {
      "entropy": 0.8828619286417961,
      "epoch": 1.284480799818223,
      "grad_norm": 0.0989094227552414,
      "learning_rate": 9.96538005248454e-05,
      "loss": 1.2855,
      "mean_token_accuracy": 0.7537649929523468,
      "num_tokens": 286515351.0,
      "step": 4240
    },
    {
      "entropy": 0.8782873898744583,
      "epoch": 1.2875104142997804,
      "grad_norm": 0.1214325875043869,
      "learning_rate": 9.96498665990587e-05,
      "loss": 1.2769,
      "mean_token_accuracy": 0.7535983622074127,
      "num_tokens": 287195468.0,
      "step": 4250
    },
    {
      "entropy": 0.8750748768448829,
      "epoch": 1.2905400287813376,
      "grad_norm": 0.10766669362783432,
      "learning_rate": 9.964591052694776e-05,
      "loss": 1.2815,
      "mean_token_accuracy": 0.7573293700814248,
      "num_tokens": 287864110.0,
      "step": 4260
    },
    {
      "entropy": 0.8821496844291687,
      "epoch": 1.2935696432628947,
      "grad_norm": 0.12105648219585419,
      "learning_rate": 9.964193231027719e-05,
      "loss": 1.2869,
      "mean_token_accuracy": 0.7547226190567017,
      "num_tokens": 288540926.0,
      "step": 4270
    },
    {
      "entropy": 0.8781195849180221,
      "epoch": 1.296599257744452,
      "grad_norm": 0.1090843677520752,
      "learning_rate": 9.963793195082145e-05,
      "loss": 1.2869,
      "mean_token_accuracy": 0.7584155231714249,
      "num_tokens": 289231734.0,
      "step": 4280
    },
    {
      "entropy": 0.8835205435752869,
      "epoch": 1.2996288722260092,
      "grad_norm": 0.10426436364650726,
      "learning_rate": 9.963390945036492e-05,
      "loss": 1.2809,
      "mean_token_accuracy": 0.7557179808616639,
      "num_tokens": 289919399.0,
      "step": 4290
    },
    {
      "entropy": 0.8799811050295829,
      "epoch": 1.3026584867075663,
      "grad_norm": 0.1049380972981453,
      "learning_rate": 9.962986481070184e-05,
      "loss": 1.2834,
      "mean_token_accuracy": 0.757496677339077,
      "num_tokens": 290605907.0,
      "step": 4300
    },
    {
      "entropy": 0.8923942729830742,
      "epoch": 1.3056881011891237,
      "grad_norm": 0.10135234892368317,
      "learning_rate": 9.96257980336363e-05,
      "loss": 1.2983,
      "mean_token_accuracy": 0.7495782405138016,
      "num_tokens": 291271083.0,
      "step": 4310
    },
    {
      "entropy": 0.8736655041575432,
      "epoch": 1.3087177156706808,
      "grad_norm": 0.1003749743103981,
      "learning_rate": 9.96217091209823e-05,
      "loss": 1.2764,
      "mean_token_accuracy": 0.7608644172549248,
      "num_tokens": 291954576.0,
      "step": 4320
    },
    {
      "entropy": 0.877230028808117,
      "epoch": 1.311747330152238,
      "grad_norm": 0.10136937350034714,
      "learning_rate": 9.96175980745637e-05,
      "loss": 1.291,
      "mean_token_accuracy": 0.7531885996460914,
      "num_tokens": 292622310.0,
      "step": 4330
    },
    {
      "entropy": 0.8699756860733032,
      "epoch": 1.3147769446337954,
      "grad_norm": 0.1041715145111084,
      "learning_rate": 9.961346489621424e-05,
      "loss": 1.2745,
      "mean_token_accuracy": 0.7576002046465874,
      "num_tokens": 293298177.0,
      "step": 4340
    },
    {
      "entropy": 0.8843838930130005,
      "epoch": 1.3178065591153525,
      "grad_norm": 0.14830972254276276,
      "learning_rate": 9.96093095877775e-05,
      "loss": 1.2964,
      "mean_token_accuracy": 0.756398618221283,
      "num_tokens": 293969251.0,
      "step": 4350
    },
    {
      "entropy": 0.885055086016655,
      "epoch": 1.3208361735969099,
      "grad_norm": 0.10774227231740952,
      "learning_rate": 9.960513215110699e-05,
      "loss": 1.2935,
      "mean_token_accuracy": 0.7495698869228363,
      "num_tokens": 294628233.0,
      "step": 4360
    },
    {
      "entropy": 0.8791931942105293,
      "epoch": 1.323865788078467,
      "grad_norm": 0.12053106725215912,
      "learning_rate": 9.960093258806601e-05,
      "loss": 1.281,
      "mean_token_accuracy": 0.7571017786860466,
      "num_tokens": 295306528.0,
      "step": 4370
    },
    {
      "entropy": 0.8699711933732033,
      "epoch": 1.3268954025600244,
      "grad_norm": 0.09862420707941055,
      "learning_rate": 9.95967109005278e-05,
      "loss": 1.2909,
      "mean_token_accuracy": 0.7579729691147804,
      "num_tokens": 295972866.0,
      "step": 4380
    },
    {
      "entropy": 0.8723306894302368,
      "epoch": 1.3299250170415815,
      "grad_norm": 0.10990869998931885,
      "learning_rate": 9.959246709037546e-05,
      "loss": 1.2867,
      "mean_token_accuracy": 0.7555100724101067,
      "num_tokens": 296645546.0,
      "step": 4390
    },
    {
      "entropy": 0.8749506160616874,
      "epoch": 1.3329546315231386,
      "grad_norm": 0.10895519703626633,
      "learning_rate": 9.958820115950192e-05,
      "loss": 1.2835,
      "mean_token_accuracy": 0.7593660697340965,
      "num_tokens": 297319750.0,
      "step": 4400
    },
    {
      "entropy": 0.8869535252451897,
      "epoch": 1.335984246004696,
      "grad_norm": 0.09852419793605804,
      "learning_rate": 9.958391310981001e-05,
      "loss": 1.2935,
      "mean_token_accuracy": 0.7517649352550506,
      "num_tokens": 297983484.0,
      "step": 4410
    },
    {
      "entropy": 0.8746039196848869,
      "epoch": 1.3390138604862531,
      "grad_norm": 0.09832970052957535,
      "learning_rate": 9.957960294321241e-05,
      "loss": 1.2792,
      "mean_token_accuracy": 0.757124112546444,
      "num_tokens": 298666438.0,
      "step": 4420
    },
    {
      "entropy": 0.8672180339694023,
      "epoch": 1.3420434749678103,
      "grad_norm": 0.11954855918884277,
      "learning_rate": 9.957527066163167e-05,
      "loss": 1.2817,
      "mean_token_accuracy": 0.7571039840579032,
      "num_tokens": 299332287.0,
      "step": 4430
    },
    {
      "entropy": 0.8617986917495728,
      "epoch": 1.3450730894493677,
      "grad_norm": 0.13852055370807648,
      "learning_rate": 9.95709162670002e-05,
      "loss": 1.2695,
      "mean_token_accuracy": 0.7625701040029526,
      "num_tokens": 300033099.0,
      "step": 4440
    },
    {
      "entropy": 0.8819914147257805,
      "epoch": 1.3481027039309248,
      "grad_norm": 0.1199556440114975,
      "learning_rate": 9.956653976126028e-05,
      "loss": 1.2931,
      "mean_token_accuracy": 0.7540413573384285,
      "num_tokens": 300698440.0,
      "step": 4450
    },
    {
      "entropy": 0.8740648403763771,
      "epoch": 1.351132318412482,
      "grad_norm": 0.12838856875896454,
      "learning_rate": 9.956214114636407e-05,
      "loss": 1.2729,
      "mean_token_accuracy": 0.7602340117096901,
      "num_tokens": 301384029.0,
      "step": 4460
    },
    {
      "entropy": 0.8644291788339615,
      "epoch": 1.3541619328940393,
      "grad_norm": 0.14665523171424866,
      "learning_rate": 9.955772042427352e-05,
      "loss": 1.2688,
      "mean_token_accuracy": 0.7580505087971687,
      "num_tokens": 302053693.0,
      "step": 4470
    },
    {
      "entropy": 0.8759150877594948,
      "epoch": 1.3571915473755964,
      "grad_norm": 0.110257588326931,
      "learning_rate": 9.955327759696056e-05,
      "loss": 1.282,
      "mean_token_accuracy": 0.7561334684491158,
      "num_tokens": 302738677.0,
      "step": 4480
    },
    {
      "entropy": 0.8849847927689553,
      "epoch": 1.3602211618571536,
      "grad_norm": 0.09808912873268127,
      "learning_rate": 9.954881266640686e-05,
      "loss": 1.2843,
      "mean_token_accuracy": 0.754899476468563,
      "num_tokens": 303421514.0,
      "step": 4490
    },
    {
      "entropy": 0.8758756816387177,
      "epoch": 1.363250776338711,
      "grad_norm": 0.1278056800365448,
      "learning_rate": 9.954432563460403e-05,
      "loss": 1.2892,
      "mean_token_accuracy": 0.7498972356319428,
      "num_tokens": 304069800.0,
      "step": 4500
    },
    {
      "entropy": 0.8744087040424346,
      "epoch": 1.366280390820268,
      "grad_norm": 0.1052425354719162,
      "learning_rate": 9.95398165035535e-05,
      "loss": 1.2871,
      "mean_token_accuracy": 0.753123264014721,
      "num_tokens": 304726982.0,
      "step": 4510
    },
    {
      "entropy": 0.8700909554958344,
      "epoch": 1.3693100053018252,
      "grad_norm": 0.10656082630157471,
      "learning_rate": 9.953528527526659e-05,
      "loss": 1.2771,
      "mean_token_accuracy": 0.7586307719349861,
      "num_tokens": 305405130.0,
      "step": 4520
    },
    {
      "entropy": 0.8599863409996032,
      "epoch": 1.3723396197833826,
      "grad_norm": 0.1053885743021965,
      "learning_rate": 9.953073195176441e-05,
      "loss": 1.2689,
      "mean_token_accuracy": 0.7555717900395393,
      "num_tokens": 306073303.0,
      "step": 4530
    },
    {
      "entropy": 0.8755203872919083,
      "epoch": 1.3753692342649397,
      "grad_norm": 0.10587162524461746,
      "learning_rate": 9.952615653507801e-05,
      "loss": 1.2903,
      "mean_token_accuracy": 0.7531731277704239,
      "num_tokens": 306727301.0,
      "step": 4540
    },
    {
      "entropy": 0.8776930332183838,
      "epoch": 1.3783988487464969,
      "grad_norm": 0.10954715311527252,
      "learning_rate": 9.952155902724825e-05,
      "loss": 1.2838,
      "mean_token_accuracy": 0.7542294219136239,
      "num_tokens": 307402679.0,
      "step": 4550
    },
    {
      "entropy": 0.8587560087442399,
      "epoch": 1.3814284632280542,
      "grad_norm": 0.09825403243303299,
      "learning_rate": 9.951693943032584e-05,
      "loss": 1.2702,
      "mean_token_accuracy": 0.7581495806574822,
      "num_tokens": 308078088.0,
      "step": 4560
    },
    {
      "entropy": 0.878232404589653,
      "epoch": 1.3844580777096114,
      "grad_norm": 0.11433161795139313,
      "learning_rate": 9.951229774637133e-05,
      "loss": 1.2885,
      "mean_token_accuracy": 0.753068993985653,
      "num_tokens": 308738906.0,
      "step": 4570
    },
    {
      "entropy": 0.8618496775627136,
      "epoch": 1.3874876921911687,
      "grad_norm": 0.09896930307149887,
      "learning_rate": 9.950763397745518e-05,
      "loss": 1.2758,
      "mean_token_accuracy": 0.7582163274288177,
      "num_tokens": 309410743.0,
      "step": 4580
    },
    {
      "entropy": 0.8812419295310974,
      "epoch": 1.3905173066727259,
      "grad_norm": 0.10881632566452026,
      "learning_rate": 9.950294812565764e-05,
      "loss": 1.2879,
      "mean_token_accuracy": 0.7559777319431304,
      "num_tokens": 310083429.0,
      "step": 4590
    },
    {
      "entropy": 0.8856163591146469,
      "epoch": 1.3935469211542832,
      "grad_norm": 0.09903638064861298,
      "learning_rate": 9.949824019306883e-05,
      "loss": 1.2821,
      "mean_token_accuracy": 0.7560224905610085,
      "num_tokens": 310766017.0,
      "step": 4600
    },
    {
      "entropy": 0.8799282774329186,
      "epoch": 1.3965765356358404,
      "grad_norm": 0.1002153754234314,
      "learning_rate": 9.949351018178876e-05,
      "loss": 1.2837,
      "mean_token_accuracy": 0.7538231521844864,
      "num_tokens": 311438137.0,
      "step": 4610
    },
    {
      "entropy": 0.8874617710709571,
      "epoch": 1.3996061501173975,
      "grad_norm": 0.1033317819237709,
      "learning_rate": 9.948875809392724e-05,
      "loss": 1.2923,
      "mean_token_accuracy": 0.7539964064955711,
      "num_tokens": 312103399.0,
      "step": 4620
    },
    {
      "entropy": 0.882491098344326,
      "epoch": 1.402635764598955,
      "grad_norm": 0.1603907346725464,
      "learning_rate": 9.94839839316039e-05,
      "loss": 1.2886,
      "mean_token_accuracy": 0.7503088295459748,
      "num_tokens": 312761854.0,
      "step": 4630
    },
    {
      "entropy": 0.8864075720310212,
      "epoch": 1.405665379080512,
      "grad_norm": 0.13780377805233002,
      "learning_rate": 9.947918769694829e-05,
      "loss": 1.2861,
      "mean_token_accuracy": 0.7553250983357429,
      "num_tokens": 313446543.0,
      "step": 4640
    },
    {
      "entropy": 0.8821283102035522,
      "epoch": 1.4086949935620692,
      "grad_norm": 0.10473109781742096,
      "learning_rate": 9.947436939209976e-05,
      "loss": 1.2914,
      "mean_token_accuracy": 0.7523757264018058,
      "num_tokens": 314119299.0,
      "step": 4650
    },
    {
      "entropy": 0.8727343067526817,
      "epoch": 1.4117246080436265,
      "grad_norm": 0.12804046273231506,
      "learning_rate": 9.946952901920752e-05,
      "loss": 1.282,
      "mean_token_accuracy": 0.7526518985629082,
      "num_tokens": 314776989.0,
      "step": 4660
    },
    {
      "entropy": 0.8780082523822784,
      "epoch": 1.4147542225251837,
      "grad_norm": 0.09967619180679321,
      "learning_rate": 9.94646665804306e-05,
      "loss": 1.2964,
      "mean_token_accuracy": 0.7571787014603615,
      "num_tokens": 315452116.0,
      "step": 4670
    },
    {
      "entropy": 0.8719828292727471,
      "epoch": 1.4177838370067408,
      "grad_norm": 0.11002221703529358,
      "learning_rate": 9.945978207793793e-05,
      "loss": 1.2773,
      "mean_token_accuracy": 0.7546962603926659,
      "num_tokens": 316123195.0,
      "step": 4680
    },
    {
      "entropy": 0.8698706135153771,
      "epoch": 1.4208134514882982,
      "grad_norm": 0.11689720302820206,
      "learning_rate": 9.945487551390823e-05,
      "loss": 1.2822,
      "mean_token_accuracy": 0.7534810394048691,
      "num_tokens": 316793218.0,
      "step": 4690
    },
    {
      "entropy": 0.8843429014086723,
      "epoch": 1.4238430659698553,
      "grad_norm": 0.10962754487991333,
      "learning_rate": 9.944994689053004e-05,
      "loss": 1.2842,
      "mean_token_accuracy": 0.7533178478479385,
      "num_tokens": 317477092.0,
      "step": 4700
    },
    {
      "entropy": 0.8758864805102349,
      "epoch": 1.4268726804514125,
      "grad_norm": 0.11763538420200348,
      "learning_rate": 9.944499621000181e-05,
      "loss": 1.2844,
      "mean_token_accuracy": 0.7536002263426781,
      "num_tokens": 318143077.0,
      "step": 4710
    },
    {
      "entropy": 0.8750483646988869,
      "epoch": 1.4299022949329698,
      "grad_norm": 0.11927127838134766,
      "learning_rate": 9.944002347453176e-05,
      "loss": 1.2764,
      "mean_token_accuracy": 0.7574943244457245,
      "num_tokens": 318837687.0,
      "step": 4720
    },
    {
      "entropy": 0.8810449033975601,
      "epoch": 1.432931909414527,
      "grad_norm": 0.09647629410028458,
      "learning_rate": 9.9435028686338e-05,
      "loss": 1.2795,
      "mean_token_accuracy": 0.7602642744779586,
      "num_tokens": 319526456.0,
      "step": 4730
    },
    {
      "entropy": 0.8667797297239304,
      "epoch": 1.4359615238960841,
      "grad_norm": 0.08604594320058823,
      "learning_rate": 9.943001184764845e-05,
      "loss": 1.2777,
      "mean_token_accuracy": 0.7617032259702683,
      "num_tokens": 320213783.0,
      "step": 4740
    },
    {
      "entropy": 0.8809635192155838,
      "epoch": 1.4389911383776415,
      "grad_norm": 0.10920163244009018,
      "learning_rate": 9.942497296070087e-05,
      "loss": 1.2754,
      "mean_token_accuracy": 0.759989570081234,
      "num_tokens": 320914323.0,
      "step": 4750
    },
    {
      "entropy": 0.8780781880021096,
      "epoch": 1.4420207528591986,
      "grad_norm": 0.09842902421951294,
      "learning_rate": 9.941991202774287e-05,
      "loss": 1.2828,
      "mean_token_accuracy": 0.756574122607708,
      "num_tokens": 321586894.0,
      "step": 4760
    },
    {
      "entropy": 0.8765702039003372,
      "epoch": 1.4450503673407558,
      "grad_norm": 0.11967124789953232,
      "learning_rate": 9.941482905103185e-05,
      "loss": 1.2817,
      "mean_token_accuracy": 0.7565953239798546,
      "num_tokens": 322257869.0,
      "step": 4770
    },
    {
      "entropy": 0.8695339649915695,
      "epoch": 1.4480799818223131,
      "grad_norm": 0.09481111913919449,
      "learning_rate": 9.940972403283511e-05,
      "loss": 1.279,
      "mean_token_accuracy": 0.7593210890889168,
      "num_tokens": 322939452.0,
      "step": 4780
    },
    {
      "entropy": 0.8763134732842446,
      "epoch": 1.4511095963038703,
      "grad_norm": 0.09376729279756546,
      "learning_rate": 9.94045969754297e-05,
      "loss": 1.2791,
      "mean_token_accuracy": 0.7569221302866935,
      "num_tokens": 323614509.0,
      "step": 4790
    },
    {
      "entropy": 0.8825785338878631,
      "epoch": 1.4541392107854276,
      "grad_norm": 0.10248862951993942,
      "learning_rate": 9.939944788110258e-05,
      "loss": 1.2803,
      "mean_token_accuracy": 0.7533236041665077,
      "num_tokens": 324295414.0,
      "step": 4800
    },
    {
      "entropy": 0.8729474455118179,
      "epoch": 1.4571688252669848,
      "grad_norm": 0.09469485282897949,
      "learning_rate": 9.939427675215047e-05,
      "loss": 1.2836,
      "mean_token_accuracy": 0.7562346294522285,
      "num_tokens": 324964806.0,
      "step": 4810
    },
    {
      "entropy": 0.8822784170508384,
      "epoch": 1.4601984397485421,
      "grad_norm": 0.10746327042579651,
      "learning_rate": 9.938908359087999e-05,
      "loss": 1.286,
      "mean_token_accuracy": 0.7533738493919373,
      "num_tokens": 325633215.0,
      "step": 4820
    },
    {
      "entropy": 0.8711674973368645,
      "epoch": 1.4632280542300993,
      "grad_norm": 0.10797962546348572,
      "learning_rate": 9.938386839960753e-05,
      "loss": 1.2763,
      "mean_token_accuracy": 0.7570240288972855,
      "num_tokens": 326307140.0,
      "step": 4830
    },
    {
      "entropy": 0.8763237610459328,
      "epoch": 1.4662576687116564,
      "grad_norm": 0.09768716990947723,
      "learning_rate": 9.937863118065932e-05,
      "loss": 1.2865,
      "mean_token_accuracy": 0.7576685190200806,
      "num_tokens": 326997033.0,
      "step": 4840
    },
    {
      "entropy": 0.8790414869785309,
      "epoch": 1.4692872831932138,
      "grad_norm": 0.11268474161624908,
      "learning_rate": 9.937337193637144e-05,
      "loss": 1.2874,
      "mean_token_accuracy": 0.7569172218441963,
      "num_tokens": 327672611.0,
      "step": 4850
    },
    {
      "entropy": 0.873789718747139,
      "epoch": 1.472316897674771,
      "grad_norm": 0.1072414219379425,
      "learning_rate": 9.936809066908975e-05,
      "loss": 1.2877,
      "mean_token_accuracy": 0.7574461817741394,
      "num_tokens": 328348125.0,
      "step": 4860
    },
    {
      "entropy": 0.869526818394661,
      "epoch": 1.475346512156328,
      "grad_norm": 0.13900327682495117,
      "learning_rate": 9.936278738116999e-05,
      "loss": 1.2797,
      "mean_token_accuracy": 0.7563824996352195,
      "num_tokens": 329021041.0,
      "step": 4870
    },
    {
      "entropy": 0.8688310131430625,
      "epoch": 1.4783761266378854,
      "grad_norm": 0.10042646527290344,
      "learning_rate": 9.935746207497766e-05,
      "loss": 1.2685,
      "mean_token_accuracy": 0.764211705327034,
      "num_tokens": 329722877.0,
      "step": 4880
    },
    {
      "entropy": 0.8855464920401573,
      "epoch": 1.4814057411194426,
      "grad_norm": 0.1466985046863556,
      "learning_rate": 9.935211475288815e-05,
      "loss": 1.2944,
      "mean_token_accuracy": 0.7504862293601036,
      "num_tokens": 330382084.0,
      "step": 4890
    },
    {
      "entropy": 0.8752141341567039,
      "epoch": 1.4844353556009997,
      "grad_norm": 0.13028596341609955,
      "learning_rate": 9.934674541728659e-05,
      "loss": 1.2804,
      "mean_token_accuracy": 0.758796738088131,
      "num_tokens": 331071319.0,
      "step": 4900
    },
    {
      "entropy": 0.8620499595999718,
      "epoch": 1.487464970082557,
      "grad_norm": 0.0955832302570343,
      "learning_rate": 9.934135407056801e-05,
      "loss": 1.2652,
      "mean_token_accuracy": 0.7580614849925041,
      "num_tokens": 331749844.0,
      "step": 4910
    },
    {
      "entropy": 0.8735622838139534,
      "epoch": 1.4904945845641142,
      "grad_norm": 0.11174575239419937,
      "learning_rate": 9.933594071513721e-05,
      "loss": 1.2812,
      "mean_token_accuracy": 0.7544201269745827,
      "num_tokens": 332417067.0,
      "step": 4920
    },
    {
      "entropy": 0.8812348023056984,
      "epoch": 1.4935241990456714,
      "grad_norm": 0.10573475062847137,
      "learning_rate": 9.93305053534088e-05,
      "loss": 1.2843,
      "mean_token_accuracy": 0.7559713959693909,
      "num_tokens": 333094908.0,
      "step": 4930
    },
    {
      "entropy": 0.868513534963131,
      "epoch": 1.4965538135272287,
      "grad_norm": 0.11965668201446533,
      "learning_rate": 9.932504798780724e-05,
      "loss": 1.2743,
      "mean_token_accuracy": 0.7537699833512306,
      "num_tokens": 333758618.0,
      "step": 4940
    },
    {
      "entropy": 0.884345480799675,
      "epoch": 1.4995834280087859,
      "grad_norm": 0.11149779707193375,
      "learning_rate": 9.93195686207668e-05,
      "loss": 1.2941,
      "mean_token_accuracy": 0.7486284375190735,
      "num_tokens": 334415616.0,
      "step": 4950
    },
    {
      "entropy": 0.8693102568387985,
      "epoch": 1.502613042490343,
      "grad_norm": 0.09893301129341125,
      "learning_rate": 9.93140672547315e-05,
      "loss": 1.2728,
      "mean_token_accuracy": 0.7564949616789818,
      "num_tokens": 335083027.0,
      "step": 4960
    },
    {
      "entropy": 0.8698922634124756,
      "epoch": 1.5056426569719004,
      "grad_norm": 0.09813852608203888,
      "learning_rate": 9.930854389215528e-05,
      "loss": 1.2804,
      "mean_token_accuracy": 0.7566551297903061,
      "num_tokens": 335757782.0,
      "step": 4970
    },
    {
      "entropy": 0.8766652792692184,
      "epoch": 1.5086722714534575,
      "grad_norm": 0.10962329804897308,
      "learning_rate": 9.930299853550182e-05,
      "loss": 1.2879,
      "mean_token_accuracy": 0.7540410026907921,
      "num_tokens": 336413450.0,
      "step": 4980
    },
    {
      "entropy": 0.883516064286232,
      "epoch": 1.5117018859350146,
      "grad_norm": 0.12260321527719498,
      "learning_rate": 9.929743118724462e-05,
      "loss": 1.2956,
      "mean_token_accuracy": 0.7535830676555634,
      "num_tokens": 337090592.0,
      "step": 4990
    },
    {
      "entropy": 0.870100949704647,
      "epoch": 1.514731500416572,
      "grad_norm": 0.10302906483411789,
      "learning_rate": 9.9291841849867e-05,
      "loss": 1.2792,
      "mean_token_accuracy": 0.7576045885682106,
      "num_tokens": 337766068.0,
      "step": 5000
    },
    {
      "entropy": 0.8732549920678139,
      "epoch": 1.5177611148981294,
      "grad_norm": 0.10612724721431732,
      "learning_rate": 9.928623052586207e-05,
      "loss": 1.2834,
      "mean_token_accuracy": 0.7551342651247979,
      "num_tokens": 338434072.0,
      "step": 5010
    },
    {
      "entropy": 0.8764806777238846,
      "epoch": 1.5207907293796863,
      "grad_norm": 0.11097440123558044,
      "learning_rate": 9.928059721773277e-05,
      "loss": 1.2809,
      "mean_token_accuracy": 0.7545226275920868,
      "num_tokens": 339109284.0,
      "step": 5020
    },
    {
      "entropy": 0.8647286728024483,
      "epoch": 1.5238203438612437,
      "grad_norm": 0.1085539236664772,
      "learning_rate": 9.927494192799187e-05,
      "loss": 1.275,
      "mean_token_accuracy": 0.7621150195598603,
      "num_tokens": 339787535.0,
      "step": 5030
    },
    {
      "entropy": 0.8686896696686744,
      "epoch": 1.526849958342801,
      "grad_norm": 0.24225415289402008,
      "learning_rate": 9.926926465916187e-05,
      "loss": 1.2769,
      "mean_token_accuracy": 0.7608293399214745,
      "num_tokens": 340480025.0,
      "step": 5040
    },
    {
      "entropy": 0.8659367457032203,
      "epoch": 1.529879572824358,
      "grad_norm": 0.1000758707523346,
      "learning_rate": 9.926356541377511e-05,
      "loss": 1.2812,
      "mean_token_accuracy": 0.7570604413747788,
      "num_tokens": 341152337.0,
      "step": 5050
    },
    {
      "entropy": 0.8760500758886337,
      "epoch": 1.5329091873059153,
      "grad_norm": 0.11556119471788406,
      "learning_rate": 9.925784419437378e-05,
      "loss": 1.284,
      "mean_token_accuracy": 0.7534583285450935,
      "num_tokens": 341827001.0,
      "step": 5060
    },
    {
      "entropy": 0.8637175768613815,
      "epoch": 1.5359388017874727,
      "grad_norm": 0.10243014991283417,
      "learning_rate": 9.92521010035098e-05,
      "loss": 1.2713,
      "mean_token_accuracy": 0.7606059044599534,
      "num_tokens": 342514084.0,
      "step": 5070
    },
    {
      "entropy": 0.8739731296896934,
      "epoch": 1.5389684162690298,
      "grad_norm": 0.10937733948230743,
      "learning_rate": 9.924633584374492e-05,
      "loss": 1.2856,
      "mean_token_accuracy": 0.7547189772129059,
      "num_tokens": 343186510.0,
      "step": 5080
    },
    {
      "entropy": 0.8762591108679771,
      "epoch": 1.541998030750587,
      "grad_norm": 0.09622039645910263,
      "learning_rate": 9.924054871765072e-05,
      "loss": 1.2744,
      "mean_token_accuracy": 0.756362409889698,
      "num_tokens": 343857408.0,
      "step": 5090
    },
    {
      "entropy": 0.8718797892332077,
      "epoch": 1.5450276452321443,
      "grad_norm": 0.09628882259130478,
      "learning_rate": 9.923473962780851e-05,
      "loss": 1.2841,
      "mean_token_accuracy": 0.75397929251194,
      "num_tokens": 344514362.0,
      "step": 5100
    },
    {
      "entropy": 0.8766556069254875,
      "epoch": 1.5480572597137015,
      "grad_norm": 0.1028362438082695,
      "learning_rate": 9.922890857680944e-05,
      "loss": 1.2815,
      "mean_token_accuracy": 0.7548733547329902,
      "num_tokens": 345187744.0,
      "step": 5110
    },
    {
      "entropy": 0.8707106709480286,
      "epoch": 1.5510868741952586,
      "grad_norm": 0.1039656549692154,
      "learning_rate": 9.922305556725447e-05,
      "loss": 1.2733,
      "mean_token_accuracy": 0.7616032481193542,
      "num_tokens": 345870292.0,
      "step": 5120
    },
    {
      "entropy": 0.8699567258358002,
      "epoch": 1.554116488676816,
      "grad_norm": 0.11757346242666245,
      "learning_rate": 9.921718060175433e-05,
      "loss": 1.2763,
      "mean_token_accuracy": 0.7593500182032585,
      "num_tokens": 346551306.0,
      "step": 5130
    },
    {
      "entropy": 0.8759250968694687,
      "epoch": 1.557146103158373,
      "grad_norm": 0.10635928809642792,
      "learning_rate": 9.921128368292955e-05,
      "loss": 1.2822,
      "mean_token_accuracy": 0.7576028853654861,
      "num_tokens": 347232359.0,
      "step": 5140
    },
    {
      "entropy": 0.868228904902935,
      "epoch": 1.5601757176399302,
      "grad_norm": 0.10289059579372406,
      "learning_rate": 9.920536481341043e-05,
      "loss": 1.2826,
      "mean_token_accuracy": 0.7575011759996414,
      "num_tokens": 347904580.0,
      "step": 5150
    },
    {
      "entropy": 0.8698439791798591,
      "epoch": 1.5632053321214876,
      "grad_norm": 0.1073436439037323,
      "learning_rate": 9.919942399583711e-05,
      "loss": 1.2731,
      "mean_token_accuracy": 0.7596629023551941,
      "num_tokens": 348592926.0,
      "step": 5160
    },
    {
      "entropy": 0.8713696137070656,
      "epoch": 1.5662349466030447,
      "grad_norm": 0.09829632937908173,
      "learning_rate": 9.919346123285947e-05,
      "loss": 1.279,
      "mean_token_accuracy": 0.7549123540520668,
      "num_tokens": 349259226.0,
      "step": 5170
    },
    {
      "entropy": 0.87287005931139,
      "epoch": 1.5692645610846019,
      "grad_norm": 0.09961492568254471,
      "learning_rate": 9.91874765271372e-05,
      "loss": 1.2739,
      "mean_token_accuracy": 0.7540394902229309,
      "num_tokens": 349934237.0,
      "step": 5180
    },
    {
      "entropy": 0.8709146901965141,
      "epoch": 1.5722941755661592,
      "grad_norm": 0.09700673073530197,
      "learning_rate": 9.918146988133981e-05,
      "loss": 1.2676,
      "mean_token_accuracy": 0.7552365481853485,
      "num_tokens": 350599993.0,
      "step": 5190
    },
    {
      "entropy": 0.8723388150334358,
      "epoch": 1.5753237900477164,
      "grad_norm": 0.10038121044635773,
      "learning_rate": 9.917544129814653e-05,
      "loss": 1.285,
      "mean_token_accuracy": 0.7541535750031472,
      "num_tokens": 351265711.0,
      "step": 5200
    },
    {
      "entropy": 0.8757157295942306,
      "epoch": 1.5783534045292735,
      "grad_norm": 0.10745465010404587,
      "learning_rate": 9.916939078024642e-05,
      "loss": 1.2792,
      "mean_token_accuracy": 0.7533563077449799,
      "num_tokens": 351929991.0,
      "step": 5210
    },
    {
      "entropy": 0.863342608511448,
      "epoch": 1.581383019010831,
      "grad_norm": 0.1044878140091896,
      "learning_rate": 9.916331833033831e-05,
      "loss": 1.269,
      "mean_token_accuracy": 0.7579909726977349,
      "num_tokens": 352606117.0,
      "step": 5220
    },
    {
      "entropy": 0.8738278642296791,
      "epoch": 1.5844126334923883,
      "grad_norm": 0.1057114228606224,
      "learning_rate": 9.915722395113083e-05,
      "loss": 1.2857,
      "mean_token_accuracy": 0.749761326611042,
      "num_tokens": 353258900.0,
      "step": 5230
    },
    {
      "entropy": 0.8743241295218468,
      "epoch": 1.5874422479739452,
      "grad_norm": 0.1229625940322876,
      "learning_rate": 9.915110764534236e-05,
      "loss": 1.2848,
      "mean_token_accuracy": 0.7557677298784256,
      "num_tokens": 353926805.0,
      "step": 5240
    },
    {
      "entropy": 0.8771678328514099,
      "epoch": 1.5904718624555025,
      "grad_norm": 0.105762779712677,
      "learning_rate": 9.91449694157011e-05,
      "loss": 1.2931,
      "mean_token_accuracy": 0.7478478208184243,
      "num_tokens": 354596448.0,
      "step": 5250
    },
    {
      "entropy": 0.8795763000845909,
      "epoch": 1.59350147693706,
      "grad_norm": 0.11217822134494781,
      "learning_rate": 9.9138809264945e-05,
      "loss": 1.2869,
      "mean_token_accuracy": 0.7559655770659447,
      "num_tokens": 355274131.0,
      "step": 5260
    },
    {
      "entropy": 0.8697790130972862,
      "epoch": 1.5965310914186168,
      "grad_norm": 0.09957357496023178,
      "learning_rate": 9.913262719582177e-05,
      "loss": 1.2781,
      "mean_token_accuracy": 0.7564894899725914,
      "num_tokens": 355944901.0,
      "step": 5270
    },
    {
      "entropy": 0.8686492651700973,
      "epoch": 1.5995607059001742,
      "grad_norm": 0.12873774766921997,
      "learning_rate": 9.912642321108896e-05,
      "loss": 1.2842,
      "mean_token_accuracy": 0.753792816400528,
      "num_tokens": 356611500.0,
      "step": 5280
    },
    {
      "entropy": 0.8777525410056114,
      "epoch": 1.6025903203817315,
      "grad_norm": 0.09984857589006424,
      "learning_rate": 9.912019731351383e-05,
      "loss": 1.2792,
      "mean_token_accuracy": 0.7551235914230346,
      "num_tokens": 357287167.0,
      "step": 5290
    },
    {
      "entropy": 0.8676513060927391,
      "epoch": 1.6056199348632887,
      "grad_norm": 0.10700773447751999,
      "learning_rate": 9.911394950587347e-05,
      "loss": 1.2853,
      "mean_token_accuracy": 0.7546106904745102,
      "num_tokens": 357959149.0,
      "step": 5300
    },
    {
      "entropy": 0.8721698507666588,
      "epoch": 1.6086495493448458,
      "grad_norm": 0.13279834389686584,
      "learning_rate": 9.910767979095467e-05,
      "loss": 1.2741,
      "mean_token_accuracy": 0.7623576447367668,
      "num_tokens": 358645791.0,
      "step": 5310
    },
    {
      "entropy": 0.8680669054389,
      "epoch": 1.6116791638264032,
      "grad_norm": 0.11023905873298645,
      "learning_rate": 9.910138817155409e-05,
      "loss": 1.2647,
      "mean_token_accuracy": 0.7610991448163986,
      "num_tokens": 359338240.0,
      "step": 5320
    },
    {
      "entropy": 0.8698973953723907,
      "epoch": 1.6147087783079603,
      "grad_norm": 0.10102608799934387,
      "learning_rate": 9.909507465047807e-05,
      "loss": 1.2784,
      "mean_token_accuracy": 0.7560726568102837,
      "num_tokens": 360016274.0,
      "step": 5330
    },
    {
      "entropy": 0.873685722053051,
      "epoch": 1.6177383927895175,
      "grad_norm": 0.10017687827348709,
      "learning_rate": 9.908873923054276e-05,
      "loss": 1.2784,
      "mean_token_accuracy": 0.7573044985532761,
      "num_tokens": 360700736.0,
      "step": 5340
    },
    {
      "entropy": 0.866089777648449,
      "epoch": 1.6207680072710748,
      "grad_norm": 0.10463502258062363,
      "learning_rate": 9.908238191457409e-05,
      "loss": 1.2654,
      "mean_token_accuracy": 0.7583584994077682,
      "num_tokens": 361365977.0,
      "step": 5350
    },
    {
      "entropy": 0.8682324588298798,
      "epoch": 1.623797621752632,
      "grad_norm": 0.12103241682052612,
      "learning_rate": 9.907600270540773e-05,
      "loss": 1.2726,
      "mean_token_accuracy": 0.7610218763351441,
      "num_tokens": 362051727.0,
      "step": 5360
    },
    {
      "entropy": 0.8686078056693077,
      "epoch": 1.6268272362341891,
      "grad_norm": 0.10196422785520554,
      "learning_rate": 9.906960160588911e-05,
      "loss": 1.272,
      "mean_token_accuracy": 0.760060878098011,
      "num_tokens": 362738689.0,
      "step": 5370
    },
    {
      "entropy": 0.8809151217341423,
      "epoch": 1.6298568507157465,
      "grad_norm": 0.10597776621580124,
      "learning_rate": 9.906317861887349e-05,
      "loss": 1.2887,
      "mean_token_accuracy": 0.7559880495071412,
      "num_tokens": 363411332.0,
      "step": 5380
    },
    {
      "entropy": 0.870364136993885,
      "epoch": 1.6328864651973036,
      "grad_norm": 0.10422533750534058,
      "learning_rate": 9.905673374722579e-05,
      "loss": 1.2691,
      "mean_token_accuracy": 0.7573687374591828,
      "num_tokens": 364083141.0,
      "step": 5390
    },
    {
      "entropy": 0.8716929346323014,
      "epoch": 1.6359160796788608,
      "grad_norm": 0.09048692882061005,
      "learning_rate": 9.905026699382078e-05,
      "loss": 1.275,
      "mean_token_accuracy": 0.759492652118206,
      "num_tokens": 364770249.0,
      "step": 5400
    },
    {
      "entropy": 0.8717505663633347,
      "epoch": 1.6389456941604181,
      "grad_norm": 0.12246933579444885,
      "learning_rate": 9.904377836154294e-05,
      "loss": 1.2753,
      "mean_token_accuracy": 0.7598855897784234,
      "num_tokens": 365456562.0,
      "step": 5410
    },
    {
      "entropy": 0.8657186686992645,
      "epoch": 1.6419753086419753,
      "grad_norm": 0.11263216286897659,
      "learning_rate": 9.903726785328651e-05,
      "loss": 1.2785,
      "mean_token_accuracy": 0.757922874391079,
      "num_tokens": 366132737.0,
      "step": 5420
    },
    {
      "entropy": 0.8811621457338333,
      "epoch": 1.6450049231235324,
      "grad_norm": 0.10388202965259552,
      "learning_rate": 9.903073547195555e-05,
      "loss": 1.2846,
      "mean_token_accuracy": 0.7510637819766999,
      "num_tokens": 366807208.0,
      "step": 5430
    },
    {
      "entropy": 0.8589653491973877,
      "epoch": 1.6480345376050898,
      "grad_norm": 0.0990695059299469,
      "learning_rate": 9.902418122046377e-05,
      "loss": 1.258,
      "mean_token_accuracy": 0.7583812937140465,
      "num_tokens": 367495342.0,
      "step": 5440
    },
    {
      "entropy": 0.8624016270041466,
      "epoch": 1.651064152086647,
      "grad_norm": 0.11094211786985397,
      "learning_rate": 9.901760510173474e-05,
      "loss": 1.2738,
      "mean_token_accuracy": 0.759155260026455,
      "num_tokens": 368163766.0,
      "step": 5450
    },
    {
      "entropy": 0.8742574378848076,
      "epoch": 1.654093766568204,
      "grad_norm": 0.09800291806459427,
      "learning_rate": 9.901100711870172e-05,
      "loss": 1.2763,
      "mean_token_accuracy": 0.7553526028990746,
      "num_tokens": 368834733.0,
      "step": 5460
    },
    {
      "entropy": 0.8638382017612457,
      "epoch": 1.6571233810497614,
      "grad_norm": 0.11040802299976349,
      "learning_rate": 9.900438727430775e-05,
      "loss": 1.2682,
      "mean_token_accuracy": 0.7588236540555954,
      "num_tokens": 369510419.0,
      "step": 5470
    },
    {
      "entropy": 0.8756036445498466,
      "epoch": 1.6601529955313188,
      "grad_norm": 0.09599775075912476,
      "learning_rate": 9.899774557150559e-05,
      "loss": 1.2818,
      "mean_token_accuracy": 0.7558170795440674,
      "num_tokens": 370180297.0,
      "step": 5480
    },
    {
      "entropy": 0.8640083447098732,
      "epoch": 1.6631826100128757,
      "grad_norm": 0.09472623467445374,
      "learning_rate": 9.89910820132578e-05,
      "loss": 1.2703,
      "mean_token_accuracy": 0.7586445361375809,
      "num_tokens": 370858685.0,
      "step": 5490
    },
    {
      "entropy": 0.8590528950095176,
      "epoch": 1.666212224494433,
      "grad_norm": 0.09686709195375443,
      "learning_rate": 9.898439660253662e-05,
      "loss": 1.2683,
      "mean_token_accuracy": 0.7555449813604355,
      "num_tokens": 371520228.0,
      "step": 5500
    },
    {
      "entropy": 0.8739757299423218,
      "epoch": 1.6692418389759904,
      "grad_norm": 0.10918669402599335,
      "learning_rate": 9.897768934232412e-05,
      "loss": 1.2768,
      "mean_token_accuracy": 0.7539754524827004,
      "num_tokens": 372190366.0,
      "step": 5510
    },
    {
      "entropy": 0.8738907009363175,
      "epoch": 1.6722714534575476,
      "grad_norm": 0.10441508144140244,
      "learning_rate": 9.897096023561205e-05,
      "loss": 1.2876,
      "mean_token_accuracy": 0.7504386335611344,
      "num_tokens": 372838175.0,
      "step": 5520
    },
    {
      "entropy": 0.8645015090703965,
      "epoch": 1.6753010679391047,
      "grad_norm": 0.11962274461984634,
      "learning_rate": 9.896420928540193e-05,
      "loss": 1.262,
      "mean_token_accuracy": 0.7578808009624481,
      "num_tokens": 373513090.0,
      "step": 5530
    },
    {
      "entropy": 0.872953699529171,
      "epoch": 1.678330682420662,
      "grad_norm": 0.11019662022590637,
      "learning_rate": 9.895743649470504e-05,
      "loss": 1.2797,
      "mean_token_accuracy": 0.7557148829102516,
      "num_tokens": 374188731.0,
      "step": 5540
    },
    {
      "entropy": 0.8694672659039497,
      "epoch": 1.6813602969022192,
      "grad_norm": 0.10090140253305435,
      "learning_rate": 9.895064186654236e-05,
      "loss": 1.2718,
      "mean_token_accuracy": 0.7555843487381935,
      "num_tokens": 374866416.0,
      "step": 5550
    },
    {
      "entropy": 0.8638262122869491,
      "epoch": 1.6843899113837764,
      "grad_norm": 0.10722874104976654,
      "learning_rate": 9.894382540394465e-05,
      "loss": 1.2715,
      "mean_token_accuracy": 0.7578088521957398,
      "num_tokens": 375542152.0,
      "step": 5560
    },
    {
      "entropy": 0.8736162975430488,
      "epoch": 1.6874195258653337,
      "grad_norm": 0.09556689113378525,
      "learning_rate": 9.893698710995239e-05,
      "loss": 1.2794,
      "mean_token_accuracy": 0.7577393084764481,
      "num_tokens": 376215629.0,
      "step": 5570
    },
    {
      "entropy": 0.8686207786202431,
      "epoch": 1.6904491403468909,
      "grad_norm": 0.11578402668237686,
      "learning_rate": 9.893012698761578e-05,
      "loss": 1.2785,
      "mean_token_accuracy": 0.7590609088540077,
      "num_tokens": 376884318.0,
      "step": 5580
    },
    {
      "entropy": 0.8811026304960251,
      "epoch": 1.693478754828448,
      "grad_norm": 0.14174573123455048,
      "learning_rate": 9.892324503999481e-05,
      "loss": 1.2932,
      "mean_token_accuracy": 0.7513189151883125,
      "num_tokens": 377543265.0,
      "step": 5590
    },
    {
      "entropy": 0.8696400627493859,
      "epoch": 1.6965083693100054,
      "grad_norm": 0.11313126981258392,
      "learning_rate": 9.891634127015916e-05,
      "loss": 1.2786,
      "mean_token_accuracy": 0.7602464750409126,
      "num_tokens": 378234862.0,
      "step": 5600
    },
    {
      "entropy": 0.8803420409560203,
      "epoch": 1.6995379837915625,
      "grad_norm": 0.11448783427476883,
      "learning_rate": 9.890941568118824e-05,
      "loss": 1.2881,
      "mean_token_accuracy": 0.7523095175623894,
      "num_tokens": 378902936.0,
      "step": 5610
    },
    {
      "entropy": 0.8576599538326264,
      "epoch": 1.7025675982731197,
      "grad_norm": 0.12137230485677719,
      "learning_rate": 9.890246827617124e-05,
      "loss": 1.2718,
      "mean_token_accuracy": 0.7621455058455467,
      "num_tokens": 379586990.0,
      "step": 5620
    },
    {
      "entropy": 0.8703350991010665,
      "epoch": 1.705597212754677,
      "grad_norm": 0.09922225028276443,
      "learning_rate": 9.889549905820703e-05,
      "loss": 1.2741,
      "mean_token_accuracy": 0.7533906310796737,
      "num_tokens": 380262953.0,
      "step": 5630
    },
    {
      "entropy": 0.8703566938638687,
      "epoch": 1.7086268272362342,
      "grad_norm": 0.11044721305370331,
      "learning_rate": 9.888850803040424e-05,
      "loss": 1.2706,
      "mean_token_accuracy": 0.754063467681408,
      "num_tokens": 380923283.0,
      "step": 5640
    },
    {
      "entropy": 0.8689834162592888,
      "epoch": 1.7116564417177913,
      "grad_norm": 0.10516627877950668,
      "learning_rate": 9.88814951958812e-05,
      "loss": 1.2813,
      "mean_token_accuracy": 0.7555178359150887,
      "num_tokens": 381596056.0,
      "step": 5650
    },
    {
      "entropy": 0.8661510065197945,
      "epoch": 1.7146860561993487,
      "grad_norm": 0.10870422422885895,
      "learning_rate": 9.887446055776601e-05,
      "loss": 1.275,
      "mean_token_accuracy": 0.7583819463849067,
      "num_tokens": 382272509.0,
      "step": 5660
    },
    {
      "entropy": 0.8740237966179848,
      "epoch": 1.7177156706809058,
      "grad_norm": 0.11483550816774368,
      "learning_rate": 9.886740411919645e-05,
      "loss": 1.2796,
      "mean_token_accuracy": 0.7544118940830231,
      "num_tokens": 382931957.0,
      "step": 5670
    },
    {
      "entropy": 0.8701524794101715,
      "epoch": 1.720745285162463,
      "grad_norm": 0.1000395193696022,
      "learning_rate": 9.886032588332006e-05,
      "loss": 1.2738,
      "mean_token_accuracy": 0.7561038956046104,
      "num_tokens": 383601391.0,
      "step": 5680
    },
    {
      "entropy": 0.8631006270647049,
      "epoch": 1.7237748996440203,
      "grad_norm": 0.16434259712696075,
      "learning_rate": 9.885322585329409e-05,
      "loss": 1.2664,
      "mean_token_accuracy": 0.7609700664877892,
      "num_tokens": 384284931.0,
      "step": 5690
    },
    {
      "entropy": 0.8777032285928726,
      "epoch": 1.7268045141255777,
      "grad_norm": 0.10967931151390076,
      "learning_rate": 9.884610403228547e-05,
      "loss": 1.2756,
      "mean_token_accuracy": 0.7540204703807831,
      "num_tokens": 384974180.0,
      "step": 5700
    },
    {
      "entropy": 0.8755529925227166,
      "epoch": 1.7298341286071346,
      "grad_norm": 0.10465622693300247,
      "learning_rate": 9.883896042347094e-05,
      "loss": 1.2796,
      "mean_token_accuracy": 0.7556664749979973,
      "num_tokens": 385658132.0,
      "step": 5710
    },
    {
      "entropy": 0.877837872505188,
      "epoch": 1.732863743088692,
      "grad_norm": 0.10969074815511703,
      "learning_rate": 9.883179503003688e-05,
      "loss": 1.279,
      "mean_token_accuracy": 0.7578695207834244,
      "num_tokens": 386345994.0,
      "step": 5720
    },
    {
      "entropy": 0.8727489277720452,
      "epoch": 1.7358933575702493,
      "grad_norm": 0.09947185963392258,
      "learning_rate": 9.882460785517941e-05,
      "loss": 1.279,
      "mean_token_accuracy": 0.7556466355919838,
      "num_tokens": 387024607.0,
      "step": 5730
    },
    {
      "entropy": 0.8709679767489433,
      "epoch": 1.7389229720518065,
      "grad_norm": 0.11316635459661484,
      "learning_rate": 9.881739890210437e-05,
      "loss": 1.2806,
      "mean_token_accuracy": 0.7591634154319763,
      "num_tokens": 387700729.0,
      "step": 5740
    },
    {
      "entropy": 0.8856248810887337,
      "epoch": 1.7419525865333636,
      "grad_norm": 0.1003992035984993,
      "learning_rate": 9.881016817402735e-05,
      "loss": 1.2825,
      "mean_token_accuracy": 0.7522728234529495,
      "num_tokens": 388377310.0,
      "step": 5750
    },
    {
      "entropy": 0.867271040380001,
      "epoch": 1.744982201014921,
      "grad_norm": 0.10750974714756012,
      "learning_rate": 9.880291567417356e-05,
      "loss": 1.2743,
      "mean_token_accuracy": 0.7556522607803344,
      "num_tokens": 389049038.0,
      "step": 5760
    },
    {
      "entropy": 0.8712126925587654,
      "epoch": 1.748011815496478,
      "grad_norm": 0.11678169667720795,
      "learning_rate": 9.879564140577801e-05,
      "loss": 1.2777,
      "mean_token_accuracy": 0.7550342082977295,
      "num_tokens": 389720842.0,
      "step": 5770
    },
    {
      "entropy": 0.8805740848183632,
      "epoch": 1.7510414299780352,
      "grad_norm": 0.11123267561197281,
      "learning_rate": 9.878834537208541e-05,
      "loss": 1.2924,
      "mean_token_accuracy": 0.7502569913864136,
      "num_tokens": 390386713.0,
      "step": 5780
    },
    {
      "entropy": 0.8717425882816314,
      "epoch": 1.7540710444595926,
      "grad_norm": 0.10272000730037689,
      "learning_rate": 9.87810275763501e-05,
      "loss": 1.2851,
      "mean_token_accuracy": 0.7530350014567375,
      "num_tokens": 391053621.0,
      "step": 5790
    },
    {
      "entropy": 0.8708827048540115,
      "epoch": 1.7571006589411498,
      "grad_norm": 0.1100660189986229,
      "learning_rate": 9.877368802183622e-05,
      "loss": 1.2695,
      "mean_token_accuracy": 0.7565293610095978,
      "num_tokens": 391734227.0,
      "step": 5800
    },
    {
      "entropy": 0.867878869175911,
      "epoch": 1.760130273422707,
      "grad_norm": 0.124241404235363,
      "learning_rate": 9.876632671181758e-05,
      "loss": 1.2696,
      "mean_token_accuracy": 0.7570357009768486,
      "num_tokens": 392407776.0,
      "step": 5810
    },
    {
      "entropy": 0.8709599047899246,
      "epoch": 1.7631598879042643,
      "grad_norm": 0.11583121865987778,
      "learning_rate": 9.875894364957767e-05,
      "loss": 1.2754,
      "mean_token_accuracy": 0.7543368205428124,
      "num_tokens": 393079411.0,
      "step": 5820
    },
    {
      "entropy": 0.861686784029007,
      "epoch": 1.7661895023858214,
      "grad_norm": 0.10723171383142471,
      "learning_rate": 9.875153883840973e-05,
      "loss": 1.2806,
      "mean_token_accuracy": 0.7571526080369949,
      "num_tokens": 393748704.0,
      "step": 5830
    },
    {
      "entropy": 0.8687817886471748,
      "epoch": 1.7692191168673785,
      "grad_norm": 0.10481414943933487,
      "learning_rate": 9.874411228161665e-05,
      "loss": 1.2746,
      "mean_token_accuracy": 0.7538354933261872,
      "num_tokens": 394408888.0,
      "step": 5840
    },
    {
      "entropy": 0.8641815066337586,
      "epoch": 1.772248731348936,
      "grad_norm": 0.11235295236110687,
      "learning_rate": 9.873666398251107e-05,
      "loss": 1.2682,
      "mean_token_accuracy": 0.7587895512580871,
      "num_tokens": 395085340.0,
      "step": 5850
    },
    {
      "entropy": 0.8639056503772735,
      "epoch": 1.775278345830493,
      "grad_norm": 0.0975707545876503,
      "learning_rate": 9.872919394441529e-05,
      "loss": 1.2667,
      "mean_token_accuracy": 0.7623942896723748,
      "num_tokens": 395765695.0,
      "step": 5860
    },
    {
      "entropy": 0.8668558150529861,
      "epoch": 1.7783079603120502,
      "grad_norm": 0.11737079918384552,
      "learning_rate": 9.872170217066133e-05,
      "loss": 1.2751,
      "mean_token_accuracy": 0.7637094333767891,
      "num_tokens": 396457097.0,
      "step": 5870
    },
    {
      "entropy": 0.8787486582994462,
      "epoch": 1.7813375747936075,
      "grad_norm": 0.13197804987430573,
      "learning_rate": 9.871418866459088e-05,
      "loss": 1.277,
      "mean_token_accuracy": 0.7568523600697518,
      "num_tokens": 397144826.0,
      "step": 5880
    },
    {
      "entropy": 0.8611941158771514,
      "epoch": 1.7843671892751647,
      "grad_norm": 0.10854239761829376,
      "learning_rate": 9.870665342955536e-05,
      "loss": 1.2709,
      "mean_token_accuracy": 0.7587135076522827,
      "num_tokens": 397823258.0,
      "step": 5890
    },
    {
      "entropy": 0.8542108252644539,
      "epoch": 1.7873968037567218,
      "grad_norm": 0.10155311226844788,
      "learning_rate": 9.869909646891585e-05,
      "loss": 1.2561,
      "mean_token_accuracy": 0.7667166635394096,
      "num_tokens": 398526408.0,
      "step": 5900
    },
    {
      "entropy": 0.8701602831482887,
      "epoch": 1.7904264182382792,
      "grad_norm": 0.10298395156860352,
      "learning_rate": 9.869151778604313e-05,
      "loss": 1.268,
      "mean_token_accuracy": 0.7556657701730728,
      "num_tokens": 399200079.0,
      "step": 5910
    },
    {
      "entropy": 0.8665944993495941,
      "epoch": 1.7934560327198366,
      "grad_norm": 0.132856547832489,
      "learning_rate": 9.868391738431768e-05,
      "loss": 1.2739,
      "mean_token_accuracy": 0.7654621496796608,
      "num_tokens": 399900518.0,
      "step": 5920
    },
    {
      "entropy": 0.8721923619508744,
      "epoch": 1.7964856472013935,
      "grad_norm": 0.10510390251874924,
      "learning_rate": 9.867629526712966e-05,
      "loss": 1.2749,
      "mean_token_accuracy": 0.7579918444156647,
      "num_tokens": 400575779.0,
      "step": 5930
    },
    {
      "entropy": 0.8664195269346238,
      "epoch": 1.7995152616829508,
      "grad_norm": 0.11647294461727142,
      "learning_rate": 9.866865143787891e-05,
      "loss": 1.2722,
      "mean_token_accuracy": 0.7604745402932167,
      "num_tokens": 401263785.0,
      "step": 5940
    },
    {
      "entropy": 0.8546414092183113,
      "epoch": 1.8025448761645082,
      "grad_norm": 0.12280753999948502,
      "learning_rate": 9.866098589997496e-05,
      "loss": 1.2619,
      "mean_token_accuracy": 0.7601353496313095,
      "num_tokens": 401933892.0,
      "step": 5950
    },
    {
      "entropy": 0.8621708378195763,
      "epoch": 1.8055744906460651,
      "grad_norm": 0.10126500576734543,
      "learning_rate": 9.8653298656837e-05,
      "loss": 1.2738,
      "mean_token_accuracy": 0.760628679394722,
      "num_tokens": 402617173.0,
      "step": 5960
    },
    {
      "entropy": 0.861925731599331,
      "epoch": 1.8086041051276225,
      "grad_norm": 0.1042354628443718,
      "learning_rate": 9.8645589711894e-05,
      "loss": 1.2728,
      "mean_token_accuracy": 0.7555099457502366,
      "num_tokens": 403278315.0,
      "step": 5970
    },
    {
      "entropy": 0.8682982236146927,
      "epoch": 1.8116337196091798,
      "grad_norm": 0.10307774692773819,
      "learning_rate": 9.863785906858446e-05,
      "loss": 1.2789,
      "mean_token_accuracy": 0.7566445082426071,
      "num_tokens": 403941185.0,
      "step": 5980
    },
    {
      "entropy": 0.865261273086071,
      "epoch": 1.814663334090737,
      "grad_norm": 0.11925513297319412,
      "learning_rate": 9.863010673035664e-05,
      "loss": 1.2796,
      "mean_token_accuracy": 0.7551749140024185,
      "num_tokens": 404617320.0,
      "step": 5990
    },
    {
      "entropy": 0.8638197466731071,
      "epoch": 1.8176929485722941,
      "grad_norm": 0.1083996593952179,
      "learning_rate": 9.862233270066852e-05,
      "loss": 1.2717,
      "mean_token_accuracy": 0.7606314837932586,
      "num_tokens": 405289238.0,
      "step": 6000
    },
    {
      "entropy": 0.8722123250365257,
      "epoch": 1.8207225630538515,
      "grad_norm": 0.10418243706226349,
      "learning_rate": 9.861453698298765e-05,
      "loss": 1.2815,
      "mean_token_accuracy": 0.7577072277665138,
      "num_tokens": 405976804.0,
      "step": 6010
    },
    {
      "entropy": 0.8649901524186134,
      "epoch": 1.8237521775354086,
      "grad_norm": 0.09926159679889679,
      "learning_rate": 9.860671958079135e-05,
      "loss": 1.2711,
      "mean_token_accuracy": 0.7606795608997345,
      "num_tokens": 406667602.0,
      "step": 6020
    },
    {
      "entropy": 0.8635318517684937,
      "epoch": 1.8267817920169658,
      "grad_norm": 0.11913318932056427,
      "learning_rate": 9.859888049756656e-05,
      "loss": 1.2767,
      "mean_token_accuracy": 0.7582060411572457,
      "num_tokens": 407334010.0,
      "step": 6030
    },
    {
      "entropy": 0.8809746861457824,
      "epoch": 1.8298114064985231,
      "grad_norm": 0.10927748680114746,
      "learning_rate": 9.859101973680989e-05,
      "loss": 1.2828,
      "mean_token_accuracy": 0.7558536440134048,
      "num_tokens": 408019021.0,
      "step": 6040
    },
    {
      "entropy": 0.8621488586068153,
      "epoch": 1.8328410209800803,
      "grad_norm": 0.12048788368701935,
      "learning_rate": 9.858313730202765e-05,
      "loss": 1.2671,
      "mean_token_accuracy": 0.7627221181988716,
      "num_tokens": 408701807.0,
      "step": 6050
    },
    {
      "entropy": 0.8674041420221329,
      "epoch": 1.8358706354616374,
      "grad_norm": 0.09827356040477753,
      "learning_rate": 9.857523319673578e-05,
      "loss": 1.2668,
      "mean_token_accuracy": 0.7573729902505875,
      "num_tokens": 409377487.0,
      "step": 6060
    },
    {
      "entropy": 0.8690124318003655,
      "epoch": 1.8389002499431948,
      "grad_norm": 0.10867494344711304,
      "learning_rate": 9.856730742445995e-05,
      "loss": 1.2706,
      "mean_token_accuracy": 0.7620067074894905,
      "num_tokens": 410067582.0,
      "step": 6070
    },
    {
      "entropy": 0.8586040124297142,
      "epoch": 1.841929864424752,
      "grad_norm": 0.11834122985601425,
      "learning_rate": 9.85593599887354e-05,
      "loss": 1.2565,
      "mean_token_accuracy": 0.7621797949075699,
      "num_tokens": 410754726.0,
      "step": 6080
    },
    {
      "entropy": 0.8774945259094238,
      "epoch": 1.844959478906309,
      "grad_norm": 0.10356058925390244,
      "learning_rate": 9.85513908931071e-05,
      "loss": 1.2828,
      "mean_token_accuracy": 0.7524344474077225,
      "num_tokens": 411423839.0,
      "step": 6090
    },
    {
      "entropy": 0.876988162100315,
      "epoch": 1.8479890933878664,
      "grad_norm": 0.1258077472448349,
      "learning_rate": 9.854340014112968e-05,
      "loss": 1.2793,
      "mean_token_accuracy": 0.7554637297987938,
      "num_tokens": 412089967.0,
      "step": 6100
    },
    {
      "entropy": 0.8628374233841896,
      "epoch": 1.8510187078694236,
      "grad_norm": 0.1431972235441208,
      "learning_rate": 9.853538773636742e-05,
      "loss": 1.272,
      "mean_token_accuracy": 0.7536155939102173,
      "num_tokens": 412746947.0,
      "step": 6110
    },
    {
      "entropy": 0.8653267577290535,
      "epoch": 1.8540483223509807,
      "grad_norm": 0.10213130712509155,
      "learning_rate": 9.85273536823942e-05,
      "loss": 1.2693,
      "mean_token_accuracy": 0.763349962234497,
      "num_tokens": 413448245.0,
      "step": 6120
    },
    {
      "entropy": 0.8732479408383369,
      "epoch": 1.857077936832538,
      "grad_norm": 0.10812725126743317,
      "learning_rate": 9.85192979827937e-05,
      "loss": 1.2776,
      "mean_token_accuracy": 0.7534717291593551,
      "num_tokens": 414118443.0,
      "step": 6130
    },
    {
      "entropy": 0.8676200777292251,
      "epoch": 1.8601075513140954,
      "grad_norm": 0.12057381868362427,
      "learning_rate": 9.851122064115908e-05,
      "loss": 1.2706,
      "mean_token_accuracy": 0.7561385720968247,
      "num_tokens": 414785476.0,
      "step": 6140
    },
    {
      "entropy": 0.8682653442025184,
      "epoch": 1.8631371657956524,
      "grad_norm": 0.1044098436832428,
      "learning_rate": 9.850312166109327e-05,
      "loss": 1.2822,
      "mean_token_accuracy": 0.7545553788542747,
      "num_tokens": 415450110.0,
      "step": 6150
    },
    {
      "entropy": 0.8706517443060875,
      "epoch": 1.8661667802772097,
      "grad_norm": 0.10405302792787552,
      "learning_rate": 9.849500104620884e-05,
      "loss": 1.2768,
      "mean_token_accuracy": 0.7562116101384163,
      "num_tokens": 416129549.0,
      "step": 6160
    },
    {
      "entropy": 0.8670205220580101,
      "epoch": 1.869196394758767,
      "grad_norm": 0.11802846193313599,
      "learning_rate": 9.848685880012795e-05,
      "loss": 1.2727,
      "mean_token_accuracy": 0.7545148894190788,
      "num_tokens": 416791373.0,
      "step": 6170
    },
    {
      "entropy": 0.8728636398911476,
      "epoch": 1.872226009240324,
      "grad_norm": 0.12189928442239761,
      "learning_rate": 9.847869492648249e-05,
      "loss": 1.2808,
      "mean_token_accuracy": 0.7547622472047806,
      "num_tokens": 417470032.0,
      "step": 6180
    },
    {
      "entropy": 0.8675832554697991,
      "epoch": 1.8752556237218814,
      "grad_norm": 0.10060123354196548,
      "learning_rate": 9.847050942891394e-05,
      "loss": 1.2728,
      "mean_token_accuracy": 0.7568170398473739,
      "num_tokens": 418145871.0,
      "step": 6190
    },
    {
      "entropy": 0.871724471449852,
      "epoch": 1.8782852382034387,
      "grad_norm": 0.11063071340322495,
      "learning_rate": 9.846230231107343e-05,
      "loss": 1.281,
      "mean_token_accuracy": 0.7572342157363892,
      "num_tokens": 418822283.0,
      "step": 6200
    },
    {
      "entropy": 0.8718278899788856,
      "epoch": 1.8813148526849959,
      "grad_norm": 0.10236962884664536,
      "learning_rate": 9.845407357662175e-05,
      "loss": 1.2695,
      "mean_token_accuracy": 0.7589347824454308,
      "num_tokens": 419513609.0,
      "step": 6210
    },
    {
      "entropy": 0.8620535895228386,
      "epoch": 1.884344467166553,
      "grad_norm": 0.10928516834974289,
      "learning_rate": 9.844582322922936e-05,
      "loss": 1.2776,
      "mean_token_accuracy": 0.7593270123004914,
      "num_tokens": 420182206.0,
      "step": 6220
    },
    {
      "entropy": 0.8648419260978699,
      "epoch": 1.8873740816481104,
      "grad_norm": 0.09895563870668411,
      "learning_rate": 9.843755127257627e-05,
      "loss": 1.27,
      "mean_token_accuracy": 0.7572863906621933,
      "num_tokens": 420852032.0,
      "step": 6230
    },
    {
      "entropy": 0.8686920776963234,
      "epoch": 1.8904036961296675,
      "grad_norm": 0.10020595043897629,
      "learning_rate": 9.842925771035223e-05,
      "loss": 1.2659,
      "mean_token_accuracy": 0.7583027169108391,
      "num_tokens": 421539380.0,
      "step": 6240
    },
    {
      "entropy": 0.8602744072675705,
      "epoch": 1.8934333106112247,
      "grad_norm": 0.1016506552696228,
      "learning_rate": 9.842094254625656e-05,
      "loss": 1.2645,
      "mean_token_accuracy": 0.7595287203788758,
      "num_tokens": 422213057.0,
      "step": 6250
    },
    {
      "entropy": 0.8802652597427368,
      "epoch": 1.896462925092782,
      "grad_norm": 0.12578418850898743,
      "learning_rate": 9.841260578399825e-05,
      "loss": 1.2827,
      "mean_token_accuracy": 0.7529232487082481,
      "num_tokens": 422881960.0,
      "step": 6260
    },
    {
      "entropy": 0.8586903229355812,
      "epoch": 1.8994925395743392,
      "grad_norm": 0.09964416176080704,
      "learning_rate": 9.840424742729591e-05,
      "loss": 1.27,
      "mean_token_accuracy": 0.7611197888851166,
      "num_tokens": 423560249.0,
      "step": 6270
    },
    {
      "entropy": 0.8650608524680138,
      "epoch": 1.9025221540558963,
      "grad_norm": 0.12132469564676285,
      "learning_rate": 9.839586747987779e-05,
      "loss": 1.2701,
      "mean_token_accuracy": 0.7577430590987205,
      "num_tokens": 424229675.0,
      "step": 6280
    },
    {
      "entropy": 0.8667382508516311,
      "epoch": 1.9055517685374537,
      "grad_norm": 0.10416774451732635,
      "learning_rate": 9.838746594548177e-05,
      "loss": 1.2724,
      "mean_token_accuracy": 0.7566452905535698,
      "num_tokens": 424901997.0,
      "step": 6290
    },
    {
      "entropy": 0.8674384728074074,
      "epoch": 1.9085813830190108,
      "grad_norm": 0.11140474677085876,
      "learning_rate": 9.837904282785534e-05,
      "loss": 1.2666,
      "mean_token_accuracy": 0.7615894317626953,
      "num_tokens": 425595505.0,
      "step": 6300
    },
    {
      "entropy": 0.8667259678244591,
      "epoch": 1.911610997500568,
      "grad_norm": 0.11175062507390976,
      "learning_rate": 9.837059813075564e-05,
      "loss": 1.2755,
      "mean_token_accuracy": 0.7553184553980827,
      "num_tokens": 426270325.0,
      "step": 6310
    },
    {
      "entropy": 0.871880267560482,
      "epoch": 1.9146406119821253,
      "grad_norm": 0.10434798151254654,
      "learning_rate": 9.836213185794944e-05,
      "loss": 1.2819,
      "mean_token_accuracy": 0.75400460511446,
      "num_tokens": 426936170.0,
      "step": 6320
    },
    {
      "entropy": 0.8698974281549454,
      "epoch": 1.9176702264636825,
      "grad_norm": 0.11283425986766815,
      "learning_rate": 9.83536440132131e-05,
      "loss": 1.2738,
      "mean_token_accuracy": 0.7545256331562996,
      "num_tokens": 427599913.0,
      "step": 6330
    },
    {
      "entropy": 0.856814657151699,
      "epoch": 1.9206998409452396,
      "grad_norm": 0.10942177474498749,
      "learning_rate": 9.834513460033262e-05,
      "loss": 1.2643,
      "mean_token_accuracy": 0.7606557041406632,
      "num_tokens": 428288763.0,
      "step": 6340
    },
    {
      "entropy": 0.8598751842975616,
      "epoch": 1.923729455426797,
      "grad_norm": 0.1006433367729187,
      "learning_rate": 9.833660362310364e-05,
      "loss": 1.2758,
      "mean_token_accuracy": 0.7584156692028046,
      "num_tokens": 428972887.0,
      "step": 6350
    },
    {
      "entropy": 0.8560686483979225,
      "epoch": 1.9267590699083543,
      "grad_norm": 0.12851066887378693,
      "learning_rate": 9.832805108533141e-05,
      "loss": 1.2726,
      "mean_token_accuracy": 0.757812787592411,
      "num_tokens": 429641614.0,
      "step": 6360
    },
    {
      "entropy": 0.8634830087423324,
      "epoch": 1.9297886843899112,
      "grad_norm": 0.13254991173744202,
      "learning_rate": 9.831947699083076e-05,
      "loss": 1.2875,
      "mean_token_accuracy": 0.7525684475898743,
      "num_tokens": 430292249.0,
      "step": 6370
    },
    {
      "entropy": 0.868242746591568,
      "epoch": 1.9328182988714686,
      "grad_norm": 0.10359557718038559,
      "learning_rate": 9.831088134342619e-05,
      "loss": 1.272,
      "mean_token_accuracy": 0.759606520831585,
      "num_tokens": 430974530.0,
      "step": 6380
    },
    {
      "entropy": 0.8602611169219017,
      "epoch": 1.935847913353026,
      "grad_norm": 0.10770441591739655,
      "learning_rate": 9.830226414695178e-05,
      "loss": 1.2738,
      "mean_token_accuracy": 0.7602724403142929,
      "num_tokens": 431642824.0,
      "step": 6390
    },
    {
      "entropy": 0.8656735271215439,
      "epoch": 1.938877527834583,
      "grad_norm": 0.10923300683498383,
      "learning_rate": 9.829362540525123e-05,
      "loss": 1.2768,
      "mean_token_accuracy": 0.7595035210251808,
      "num_tokens": 432319865.0,
      "step": 6400
    },
    {
      "entropy": 0.8671105921268463,
      "epoch": 1.9419071423161403,
      "grad_norm": 0.10808205604553223,
      "learning_rate": 9.828496512217786e-05,
      "loss": 1.2689,
      "mean_token_accuracy": 0.7590477406978607,
      "num_tokens": 433004538.0,
      "step": 6410
    },
    {
      "entropy": 0.8676582247018814,
      "epoch": 1.9449367567976976,
      "grad_norm": 0.10780012607574463,
      "learning_rate": 9.827628330159458e-05,
      "loss": 1.2595,
      "mean_token_accuracy": 0.7648343667387962,
      "num_tokens": 433702296.0,
      "step": 6420
    },
    {
      "entropy": 0.8603417366743088,
      "epoch": 1.9479663712792548,
      "grad_norm": 0.1039799153804779,
      "learning_rate": 9.826757994737391e-05,
      "loss": 1.2676,
      "mean_token_accuracy": 0.7605415299534798,
      "num_tokens": 434391934.0,
      "step": 6430
    },
    {
      "entropy": 0.8720110803842545,
      "epoch": 1.950995985760812,
      "grad_norm": 0.1017688512802124,
      "learning_rate": 9.8258855063398e-05,
      "loss": 1.2795,
      "mean_token_accuracy": 0.7516930446028709,
      "num_tokens": 435055466.0,
      "step": 6440
    },
    {
      "entropy": 0.8747083127498627,
      "epoch": 1.9540256002423693,
      "grad_norm": 0.10512080043554306,
      "learning_rate": 9.825010865355857e-05,
      "loss": 1.2804,
      "mean_token_accuracy": 0.7554742082953453,
      "num_tokens": 435723328.0,
      "step": 6450
    },
    {
      "entropy": 0.8562737569212914,
      "epoch": 1.9570552147239264,
      "grad_norm": 0.11978275328874588,
      "learning_rate": 9.824134072175696e-05,
      "loss": 1.2598,
      "mean_token_accuracy": 0.7656181454658508,
      "num_tokens": 436423933.0,
      "step": 6460
    },
    {
      "entropy": 0.8638344198465348,
      "epoch": 1.9600848292054835,
      "grad_norm": 0.10439760982990265,
      "learning_rate": 9.82325512719041e-05,
      "loss": 1.2712,
      "mean_token_accuracy": 0.7559848457574845,
      "num_tokens": 437104622.0,
      "step": 6470
    },
    {
      "entropy": 0.8664592877030373,
      "epoch": 1.963114443687041,
      "grad_norm": 0.1122550442814827,
      "learning_rate": 9.822374030792054e-05,
      "loss": 1.2749,
      "mean_token_accuracy": 0.7568196907639504,
      "num_tokens": 437779579.0,
      "step": 6480
    },
    {
      "entropy": 0.8707536339759827,
      "epoch": 1.966144058168598,
      "grad_norm": 0.11339100450277328,
      "learning_rate": 9.82149078337364e-05,
      "loss": 1.2747,
      "mean_token_accuracy": 0.7549351662397384,
      "num_tokens": 438451449.0,
      "step": 6490
    },
    {
      "entropy": 0.861909331381321,
      "epoch": 1.9691736726501552,
      "grad_norm": 0.10671177506446838,
      "learning_rate": 9.820605385329142e-05,
      "loss": 1.2706,
      "mean_token_accuracy": 0.7554570451378823,
      "num_tokens": 439108629.0,
      "step": 6500
    },
    {
      "entropy": 0.867060911655426,
      "epoch": 1.9722032871317126,
      "grad_norm": 0.10488517582416534,
      "learning_rate": 9.819717837053488e-05,
      "loss": 1.2837,
      "mean_token_accuracy": 0.7523362159729003,
      "num_tokens": 439778874.0,
      "step": 6510
    },
    {
      "entropy": 0.8727175980806351,
      "epoch": 1.9752329016132697,
      "grad_norm": 0.10052948445081711,
      "learning_rate": 9.818828138942575e-05,
      "loss": 1.2708,
      "mean_token_accuracy": 0.7553378149867058,
      "num_tokens": 440458837.0,
      "step": 6520
    },
    {
      "entropy": 0.860368287563324,
      "epoch": 1.9782625160948268,
      "grad_norm": 0.09840691834688187,
      "learning_rate": 9.817936291393246e-05,
      "loss": 1.2791,
      "mean_token_accuracy": 0.7558258727192879,
      "num_tokens": 441114532.0,
      "step": 6530
    },
    {
      "entropy": 0.8804644227027894,
      "epoch": 1.9812921305763842,
      "grad_norm": 0.10521312057971954,
      "learning_rate": 9.817042294803314e-05,
      "loss": 1.2823,
      "mean_token_accuracy": 0.7543828547000885,
      "num_tokens": 441787728.0,
      "step": 6540
    },
    {
      "entropy": 0.8673185363411904,
      "epoch": 1.9843217450579413,
      "grad_norm": 0.11936893314123154,
      "learning_rate": 9.816146149571546e-05,
      "loss": 1.283,
      "mean_token_accuracy": 0.7550018146634102,
      "num_tokens": 442442350.0,
      "step": 6550
    },
    {
      "entropy": 0.8756032794713974,
      "epoch": 1.9873513595394985,
      "grad_norm": 0.10171148180961609,
      "learning_rate": 9.815247856097667e-05,
      "loss": 1.2727,
      "mean_token_accuracy": 0.7548216253519058,
      "num_tokens": 443113248.0,
      "step": 6560
    },
    {
      "entropy": 0.8717926159501076,
      "epoch": 1.9903809740210558,
      "grad_norm": 0.11027976870536804,
      "learning_rate": 9.814347414782358e-05,
      "loss": 1.272,
      "mean_token_accuracy": 0.7534134685993195,
      "num_tokens": 443778988.0,
      "step": 6570
    },
    {
      "entropy": 0.8772870868444442,
      "epoch": 1.993410588502613,
      "grad_norm": 0.10239473730325699,
      "learning_rate": 9.813444826027267e-05,
      "loss": 1.274,
      "mean_token_accuracy": 0.7549547031521797,
      "num_tokens": 444464540.0,
      "step": 6580
    },
    {
      "entropy": 0.866197295486927,
      "epoch": 1.9964402029841701,
      "grad_norm": 0.10646392405033112,
      "learning_rate": 9.812540090234988e-05,
      "loss": 1.2734,
      "mean_token_accuracy": 0.7553907632827759,
      "num_tokens": 445139655.0,
      "step": 6590
    },
    {
      "entropy": 0.8781222715973854,
      "epoch": 1.9994698174657275,
      "grad_norm": 0.10427211970090866,
      "learning_rate": 9.811633207809081e-05,
      "loss": 1.2805,
      "mean_token_accuracy": 0.7585296750068664,
      "num_tokens": 445824421.0,
      "step": 6600
    },
    {
      "entropy": 0.855379995627281,
      "epoch": 2.002423691585246,
      "grad_norm": 0.12197387963533401,
      "learning_rate": 9.810724179154061e-05,
      "loss": 1.2648,
      "mean_token_accuracy": 0.7599545411574535,
      "num_tokens": 446481716.0,
      "step": 6610
    },
    {
      "entropy": 0.8597342357039451,
      "epoch": 2.005453306066803,
      "grad_norm": 0.10399507731199265,
      "learning_rate": 9.809813004675399e-05,
      "loss": 1.2643,
      "mean_token_accuracy": 0.7580909013748169,
      "num_tokens": 447160728.0,
      "step": 6620
    },
    {
      "entropy": 0.8628667056560516,
      "epoch": 2.00848292054836,
      "grad_norm": 0.10651499032974243,
      "learning_rate": 9.808899684779527e-05,
      "loss": 1.2694,
      "mean_token_accuracy": 0.7572256475687027,
      "num_tokens": 447827882.0,
      "step": 6630
    },
    {
      "entropy": 0.8654381215572358,
      "epoch": 2.0115125350299174,
      "grad_norm": 0.11701236665248871,
      "learning_rate": 9.807984219873829e-05,
      "loss": 1.2749,
      "mean_token_accuracy": 0.7546640053391457,
      "num_tokens": 448495014.0,
      "step": 6640
    },
    {
      "entropy": 0.8621604233980179,
      "epoch": 2.014542149511475,
      "grad_norm": 0.11036840826272964,
      "learning_rate": 9.807066610366646e-05,
      "loss": 1.2636,
      "mean_token_accuracy": 0.7617029055953026,
      "num_tokens": 449183934.0,
      "step": 6650
    },
    {
      "entropy": 0.8624231189489364,
      "epoch": 2.0175717639930317,
      "grad_norm": 0.10649905353784561,
      "learning_rate": 9.806146856667284e-05,
      "loss": 1.2709,
      "mean_token_accuracy": 0.7637782678008079,
      "num_tokens": 449866563.0,
      "step": 6660
    },
    {
      "entropy": 0.8569630533456802,
      "epoch": 2.020601378474589,
      "grad_norm": 0.10633864998817444,
      "learning_rate": 9.805224959185995e-05,
      "loss": 1.2639,
      "mean_token_accuracy": 0.7612204819917678,
      "num_tokens": 450543198.0,
      "step": 6670
    },
    {
      "entropy": 0.8620423242449761,
      "epoch": 2.0236309929561465,
      "grad_norm": 0.11411301791667938,
      "learning_rate": 9.80430091833399e-05,
      "loss": 1.2553,
      "mean_token_accuracy": 0.7586001366376877,
      "num_tokens": 451220504.0,
      "step": 6680
    },
    {
      "entropy": 0.8453009083867074,
      "epoch": 2.0266606074377034,
      "grad_norm": 0.11585341393947601,
      "learning_rate": 9.80337473452344e-05,
      "loss": 1.2442,
      "mean_token_accuracy": 0.7682624146342277,
      "num_tokens": 451922244.0,
      "step": 6690
    },
    {
      "entropy": 0.8570357114076614,
      "epoch": 2.0296902219192607,
      "grad_norm": 0.11498195677995682,
      "learning_rate": 9.802446408167472e-05,
      "loss": 1.2632,
      "mean_token_accuracy": 0.7542983755469322,
      "num_tokens": 452583893.0,
      "step": 6700
    },
    {
      "entropy": 0.8643184080719948,
      "epoch": 2.032719836400818,
      "grad_norm": 0.11596842110157013,
      "learning_rate": 9.801515939680159e-05,
      "loss": 1.2712,
      "mean_token_accuracy": 0.7544256269931793,
      "num_tokens": 453260582.0,
      "step": 6710
    },
    {
      "entropy": 0.8602597936987877,
      "epoch": 2.035749450882375,
      "grad_norm": 0.10272014141082764,
      "learning_rate": 9.800583329476542e-05,
      "loss": 1.2588,
      "mean_token_accuracy": 0.7598227143287659,
      "num_tokens": 453952031.0,
      "step": 6720
    },
    {
      "entropy": 0.8591838702559471,
      "epoch": 2.0387790653639324,
      "grad_norm": 0.10272420197725296,
      "learning_rate": 9.79964857797261e-05,
      "loss": 1.2624,
      "mean_token_accuracy": 0.7543813779950141,
      "num_tokens": 454617441.0,
      "step": 6730
    },
    {
      "entropy": 0.8641746655106545,
      "epoch": 2.0418086798454897,
      "grad_norm": 0.10870486497879028,
      "learning_rate": 9.798711685585309e-05,
      "loss": 1.2716,
      "mean_token_accuracy": 0.7550192728638649,
      "num_tokens": 455278105.0,
      "step": 6740
    },
    {
      "entropy": 0.8625334724783897,
      "epoch": 2.0448382943270467,
      "grad_norm": 0.10910259187221527,
      "learning_rate": 9.79777265273254e-05,
      "loss": 1.27,
      "mean_token_accuracy": 0.7616015195846557,
      "num_tokens": 455961244.0,
      "step": 6750
    },
    {
      "entropy": 0.8660917222499848,
      "epoch": 2.047867908808604,
      "grad_norm": 0.09981189668178558,
      "learning_rate": 9.796831479833158e-05,
      "loss": 1.2661,
      "mean_token_accuracy": 0.7565690204501152,
      "num_tokens": 456640849.0,
      "step": 6760
    },
    {
      "entropy": 0.853822048008442,
      "epoch": 2.0508975232901614,
      "grad_norm": 0.10835167020559311,
      "learning_rate": 9.795888167306973e-05,
      "loss": 1.2565,
      "mean_token_accuracy": 0.7610779702663422,
      "num_tokens": 457326593.0,
      "step": 6770
    },
    {
      "entropy": 0.8495786920189857,
      "epoch": 2.0539271377717188,
      "grad_norm": 0.10651559382677078,
      "learning_rate": 9.794942715574751e-05,
      "loss": 1.2553,
      "mean_token_accuracy": 0.7608906149864196,
      "num_tokens": 458005097.0,
      "step": 6780
    },
    {
      "entropy": 0.8613103628158569,
      "epoch": 2.0569567522532757,
      "grad_norm": 0.11444416642189026,
      "learning_rate": 9.79399512505821e-05,
      "loss": 1.2674,
      "mean_token_accuracy": 0.7556070819497108,
      "num_tokens": 458676999.0,
      "step": 6790
    },
    {
      "entropy": 0.856567345559597,
      "epoch": 2.059986366734833,
      "grad_norm": 0.10138892382383347,
      "learning_rate": 9.793045396180025e-05,
      "loss": 1.2648,
      "mean_token_accuracy": 0.7604643791913986,
      "num_tokens": 459353383.0,
      "step": 6800
    },
    {
      "entropy": 0.8569768846035004,
      "epoch": 2.0630159812163904,
      "grad_norm": 0.10967062413692474,
      "learning_rate": 9.792093529363818e-05,
      "loss": 1.265,
      "mean_token_accuracy": 0.7586782589554787,
      "num_tokens": 460026096.0,
      "step": 6810
    },
    {
      "entropy": 0.8583046585321427,
      "epoch": 2.0660455956979473,
      "grad_norm": 0.1021110787987709,
      "learning_rate": 9.791139525034172e-05,
      "loss": 1.2648,
      "mean_token_accuracy": 0.7559386551380157,
      "num_tokens": 460696436.0,
      "step": 6820
    },
    {
      "entropy": 0.8555554494261741,
      "epoch": 2.0690752101795047,
      "grad_norm": 0.10578510165214539,
      "learning_rate": 9.790183383616621e-05,
      "loss": 1.2572,
      "mean_token_accuracy": 0.7565886348485946,
      "num_tokens": 461359802.0,
      "step": 6830
    },
    {
      "entropy": 0.8597806006669998,
      "epoch": 2.072104824661062,
      "grad_norm": 0.10171770304441452,
      "learning_rate": 9.78922510553765e-05,
      "loss": 1.2627,
      "mean_token_accuracy": 0.7535345032811165,
      "num_tokens": 462023955.0,
      "step": 6840
    },
    {
      "entropy": 0.8709043994545936,
      "epoch": 2.075134439142619,
      "grad_norm": 0.11991483718156815,
      "learning_rate": 9.7882646912247e-05,
      "loss": 1.2712,
      "mean_token_accuracy": 0.7572952851653099,
      "num_tokens": 462703705.0,
      "step": 6850
    },
    {
      "entropy": 0.8502332121133804,
      "epoch": 2.0781640536241763,
      "grad_norm": 0.10345316678285599,
      "learning_rate": 9.787302141106165e-05,
      "loss": 1.2562,
      "mean_token_accuracy": 0.7572643637657166,
      "num_tokens": 463373837.0,
      "step": 6860
    },
    {
      "entropy": 0.8540147677063942,
      "epoch": 2.0811936681057337,
      "grad_norm": 0.12561824917793274,
      "learning_rate": 9.78633745561139e-05,
      "loss": 1.2474,
      "mean_token_accuracy": 0.7603194698691368,
      "num_tokens": 464054524.0,
      "step": 6870
    },
    {
      "entropy": 0.8574964791536331,
      "epoch": 2.0842232825872906,
      "grad_norm": 0.11242875456809998,
      "learning_rate": 9.785370635170671e-05,
      "loss": 1.2562,
      "mean_token_accuracy": 0.7610817566514015,
      "num_tokens": 464740491.0,
      "step": 6880
    },
    {
      "entropy": 0.8604069128632545,
      "epoch": 2.087252897068848,
      "grad_norm": 0.1068437322974205,
      "learning_rate": 9.78440168021526e-05,
      "loss": 1.2665,
      "mean_token_accuracy": 0.7602398425340653,
      "num_tokens": 465423032.0,
      "step": 6890
    },
    {
      "entropy": 0.8592790767550469,
      "epoch": 2.0902825115504053,
      "grad_norm": 0.10740931332111359,
      "learning_rate": 9.78343059117736e-05,
      "loss": 1.2618,
      "mean_token_accuracy": 0.7602035805583001,
      "num_tokens": 466112075.0,
      "step": 6900
    },
    {
      "entropy": 0.8645462259650231,
      "epoch": 2.0933121260319623,
      "grad_norm": 0.10766159743070602,
      "learning_rate": 9.782457368490124e-05,
      "loss": 1.2592,
      "mean_token_accuracy": 0.757443630695343,
      "num_tokens": 466791246.0,
      "step": 6910
    },
    {
      "entropy": 0.8605925098061562,
      "epoch": 2.0963417405135196,
      "grad_norm": 0.09991318732500076,
      "learning_rate": 9.781482012587659e-05,
      "loss": 1.2693,
      "mean_token_accuracy": 0.7562457993626595,
      "num_tokens": 467461551.0,
      "step": 6920
    },
    {
      "entropy": 0.8558261439204216,
      "epoch": 2.099371354995077,
      "grad_norm": 0.10399764776229858,
      "learning_rate": 9.780504523905021e-05,
      "loss": 1.2578,
      "mean_token_accuracy": 0.7609396636486053,
      "num_tokens": 468148466.0,
      "step": 6930
    },
    {
      "entropy": 0.8557205960154534,
      "epoch": 2.102400969476634,
      "grad_norm": 0.10423752665519714,
      "learning_rate": 9.779524902878221e-05,
      "loss": 1.25,
      "mean_token_accuracy": 0.7585823208093643,
      "num_tokens": 468825740.0,
      "step": 6940
    },
    {
      "entropy": 0.8591925516724587,
      "epoch": 2.1054305839581913,
      "grad_norm": 0.10516276210546494,
      "learning_rate": 9.778543149944216e-05,
      "loss": 1.2563,
      "mean_token_accuracy": 0.7639168784022331,
      "num_tokens": 469519971.0,
      "step": 6950
    },
    {
      "entropy": 0.8553109049797059,
      "epoch": 2.1084601984397486,
      "grad_norm": 0.11408320814371109,
      "learning_rate": 9.77755926554092e-05,
      "loss": 1.2613,
      "mean_token_accuracy": 0.7561553657054901,
      "num_tokens": 470186545.0,
      "step": 6960
    },
    {
      "entropy": 0.8511288121342659,
      "epoch": 2.1114898129213056,
      "grad_norm": 0.10330954194068909,
      "learning_rate": 9.776573250107192e-05,
      "loss": 1.2591,
      "mean_token_accuracy": 0.757400181889534,
      "num_tokens": 470858012.0,
      "step": 6970
    },
    {
      "entropy": 0.8525807246565819,
      "epoch": 2.114519427402863,
      "grad_norm": 0.11110203713178635,
      "learning_rate": 9.775585104082847e-05,
      "loss": 1.2537,
      "mean_token_accuracy": 0.7565642282366752,
      "num_tokens": 471527673.0,
      "step": 6980
    },
    {
      "entropy": 0.8628541558980942,
      "epoch": 2.1175490418844203,
      "grad_norm": 0.10720645636320114,
      "learning_rate": 9.774594827908647e-05,
      "loss": 1.2731,
      "mean_token_accuracy": 0.7586317673325539,
      "num_tokens": 472201465.0,
      "step": 6990
    },
    {
      "entropy": 0.8592759609222412,
      "epoch": 2.1205786563659776,
      "grad_norm": 0.1143016666173935,
      "learning_rate": 9.773602422026302e-05,
      "loss": 1.2693,
      "mean_token_accuracy": 0.7563334733247757,
      "num_tokens": 472871361.0,
      "step": 7000
    },
    {
      "entropy": 0.8691433608531952,
      "epoch": 2.1236082708475346,
      "grad_norm": 0.12189134955406189,
      "learning_rate": 9.772607886878477e-05,
      "loss": 1.276,
      "mean_token_accuracy": 0.7545945599675179,
      "num_tokens": 473531479.0,
      "step": 7010
    },
    {
      "entropy": 0.8580952316522599,
      "epoch": 2.126637885329092,
      "grad_norm": 0.11679427325725555,
      "learning_rate": 9.771611222908785e-05,
      "loss": 1.2603,
      "mean_token_accuracy": 0.7634949997067452,
      "num_tokens": 474222414.0,
      "step": 7020
    },
    {
      "entropy": 0.8566856682300568,
      "epoch": 2.1296674998106493,
      "grad_norm": 0.10399682819843292,
      "learning_rate": 9.770612430561786e-05,
      "loss": 1.2588,
      "mean_token_accuracy": 0.7572376638650894,
      "num_tokens": 474895832.0,
      "step": 7030
    },
    {
      "entropy": 0.8566211298108101,
      "epoch": 2.132697114292206,
      "grad_norm": 0.1056089997291565,
      "learning_rate": 9.769611510282992e-05,
      "loss": 1.2697,
      "mean_token_accuracy": 0.7561067789793015,
      "num_tokens": 475556409.0,
      "step": 7040
    },
    {
      "entropy": 0.8562668204307556,
      "epoch": 2.1357267287737636,
      "grad_norm": 0.11222401261329651,
      "learning_rate": 9.768608462518865e-05,
      "loss": 1.2683,
      "mean_token_accuracy": 0.7573768883943558,
      "num_tokens": 476225073.0,
      "step": 7050
    },
    {
      "entropy": 0.8557048231363297,
      "epoch": 2.138756343255321,
      "grad_norm": 0.10089311748743057,
      "learning_rate": 9.767603287716813e-05,
      "loss": 1.2605,
      "mean_token_accuracy": 0.7541783332824707,
      "num_tokens": 476882152.0,
      "step": 7060
    },
    {
      "entropy": 0.8503266096115112,
      "epoch": 2.141785957736878,
      "grad_norm": 0.10705769062042236,
      "learning_rate": 9.766595986325193e-05,
      "loss": 1.2676,
      "mean_token_accuracy": 0.7550541222095489,
      "num_tokens": 477542573.0,
      "step": 7070
    },
    {
      "entropy": 0.8711753636598587,
      "epoch": 2.144815572218435,
      "grad_norm": 0.1165536567568779,
      "learning_rate": 9.765586558793316e-05,
      "loss": 1.2668,
      "mean_token_accuracy": 0.7530593708157539,
      "num_tokens": 478214118.0,
      "step": 7080
    },
    {
      "entropy": 0.8580114468932152,
      "epoch": 2.1478451866999926,
      "grad_norm": 0.1194147914648056,
      "learning_rate": 9.764575005571432e-05,
      "loss": 1.2658,
      "mean_token_accuracy": 0.7554805114865303,
      "num_tokens": 478881632.0,
      "step": 7090
    },
    {
      "entropy": 0.8596231237053871,
      "epoch": 2.1508748011815495,
      "grad_norm": 0.10714981704950333,
      "learning_rate": 9.763561327110749e-05,
      "loss": 1.2647,
      "mean_token_accuracy": 0.7552571997046471,
      "num_tokens": 479553212.0,
      "step": 7100
    },
    {
      "entropy": 0.8582180052995682,
      "epoch": 2.153904415663107,
      "grad_norm": 0.13480797410011292,
      "learning_rate": 9.762545523863413e-05,
      "loss": 1.2603,
      "mean_token_accuracy": 0.7590198293328285,
      "num_tokens": 480225432.0,
      "step": 7110
    },
    {
      "entropy": 0.8705548882484436,
      "epoch": 2.1569340301446642,
      "grad_norm": 0.11295425891876221,
      "learning_rate": 9.76152759628253e-05,
      "loss": 1.2793,
      "mean_token_accuracy": 0.7544082313776016,
      "num_tokens": 480893534.0,
      "step": 7120
    },
    {
      "entropy": 0.8477849125862121,
      "epoch": 2.159963644626221,
      "grad_norm": 0.12092622369527817,
      "learning_rate": 9.760507544822141e-05,
      "loss": 1.2578,
      "mean_token_accuracy": 0.7593964442610741,
      "num_tokens": 481557885.0,
      "step": 7130
    },
    {
      "entropy": 0.8525932401418685,
      "epoch": 2.1629932591077785,
      "grad_norm": 0.10847190767526627,
      "learning_rate": 9.759485369937241e-05,
      "loss": 1.2609,
      "mean_token_accuracy": 0.7636817306280136,
      "num_tokens": 482250376.0,
      "step": 7140
    },
    {
      "entropy": 0.8548767328262329,
      "epoch": 2.166022873589336,
      "grad_norm": 0.11364589631557465,
      "learning_rate": 9.758461072083773e-05,
      "loss": 1.2648,
      "mean_token_accuracy": 0.7583749040961265,
      "num_tokens": 482921431.0,
      "step": 7150
    },
    {
      "entropy": 0.8482321932911873,
      "epoch": 2.169052488070893,
      "grad_norm": 0.12237298488616943,
      "learning_rate": 9.757434651718624e-05,
      "loss": 1.2567,
      "mean_token_accuracy": 0.7610787898302078,
      "num_tokens": 483597366.0,
      "step": 7160
    },
    {
      "entropy": 0.8592027381062508,
      "epoch": 2.17208210255245,
      "grad_norm": 0.10675161331892014,
      "learning_rate": 9.756406109299629e-05,
      "loss": 1.2578,
      "mean_token_accuracy": 0.7558085158467293,
      "num_tokens": 484263747.0,
      "step": 7170
    },
    {
      "entropy": 0.8659064993262291,
      "epoch": 2.1751117170340075,
      "grad_norm": 0.11875585466623306,
      "learning_rate": 9.755375445285569e-05,
      "loss": 1.2714,
      "mean_token_accuracy": 0.7590313360095025,
      "num_tokens": 484947069.0,
      "step": 7180
    },
    {
      "entropy": 0.8590867668390274,
      "epoch": 2.1781413315155644,
      "grad_norm": 0.11141160875558853,
      "learning_rate": 9.754342660136174e-05,
      "loss": 1.2635,
      "mean_token_accuracy": 0.7573440134525299,
      "num_tokens": 485616362.0,
      "step": 7190
    },
    {
      "entropy": 0.8552271515130997,
      "epoch": 2.181170945997122,
      "grad_norm": 0.1052631139755249,
      "learning_rate": 9.753307754312114e-05,
      "loss": 1.2553,
      "mean_token_accuracy": 0.7584361836314202,
      "num_tokens": 486295062.0,
      "step": 7200
    },
    {
      "entropy": 0.8523262500762939,
      "epoch": 2.184200560478679,
      "grad_norm": 0.10578624904155731,
      "learning_rate": 9.752270728275009e-05,
      "loss": 1.2583,
      "mean_token_accuracy": 0.7606440037488937,
      "num_tokens": 486975723.0,
      "step": 7210
    },
    {
      "entropy": 0.8580429136753083,
      "epoch": 2.187230174960236,
      "grad_norm": 0.11437220126390457,
      "learning_rate": 9.751231582487428e-05,
      "loss": 1.264,
      "mean_token_accuracy": 0.7566904693841934,
      "num_tokens": 487648697.0,
      "step": 7220
    },
    {
      "entropy": 0.855644790828228,
      "epoch": 2.1902597894417934,
      "grad_norm": 0.11084471642971039,
      "learning_rate": 9.750190317412879e-05,
      "loss": 1.2592,
      "mean_token_accuracy": 0.7634907543659211,
      "num_tokens": 488338493.0,
      "step": 7230
    },
    {
      "entropy": 0.8540291801095009,
      "epoch": 2.193289403923351,
      "grad_norm": 0.11131398379802704,
      "learning_rate": 9.74914693351582e-05,
      "loss": 1.2527,
      "mean_token_accuracy": 0.7613129258155823,
      "num_tokens": 489019594.0,
      "step": 7240
    },
    {
      "entropy": 0.851766362786293,
      "epoch": 2.196319018404908,
      "grad_norm": 0.11028112471103668,
      "learning_rate": 9.748101431261652e-05,
      "loss": 1.2543,
      "mean_token_accuracy": 0.7593399941921234,
      "num_tokens": 489691617.0,
      "step": 7250
    },
    {
      "entropy": 0.8516391098499299,
      "epoch": 2.199348632886465,
      "grad_norm": 0.1113218367099762,
      "learning_rate": 9.747053811116721e-05,
      "loss": 1.2478,
      "mean_token_accuracy": 0.7576561823487282,
      "num_tokens": 490359650.0,
      "step": 7260
    },
    {
      "entropy": 0.8581887081265449,
      "epoch": 2.2023782473680225,
      "grad_norm": 0.11369206756353378,
      "learning_rate": 9.746004073548319e-05,
      "loss": 1.265,
      "mean_token_accuracy": 0.7545206531882286,
      "num_tokens": 491031446.0,
      "step": 7270
    },
    {
      "entropy": 0.8595120161771774,
      "epoch": 2.20540786184958,
      "grad_norm": 0.10878115147352219,
      "learning_rate": 9.74495221902468e-05,
      "loss": 1.2702,
      "mean_token_accuracy": 0.7577403575181961,
      "num_tokens": 491702192.0,
      "step": 7280
    },
    {
      "entropy": 0.8505108594894409,
      "epoch": 2.2084374763311367,
      "grad_norm": 0.11624005436897278,
      "learning_rate": 9.743898248014982e-05,
      "loss": 1.2443,
      "mean_token_accuracy": 0.7682089433073997,
      "num_tokens": 492403891.0,
      "step": 7290
    },
    {
      "entropy": 0.8472009196877479,
      "epoch": 2.211467090812694,
      "grad_norm": 0.10479124635457993,
      "learning_rate": 9.742842160989353e-05,
      "loss": 1.2549,
      "mean_token_accuracy": 0.7610199108719826,
      "num_tokens": 493070921.0,
      "step": 7300
    },
    {
      "entropy": 0.8508476316928864,
      "epoch": 2.2144967052942515,
      "grad_norm": 0.12949249148368835,
      "learning_rate": 9.741783958418858e-05,
      "loss": 1.2635,
      "mean_token_accuracy": 0.7580212265253067,
      "num_tokens": 493741382.0,
      "step": 7310
    },
    {
      "entropy": 0.8497532665729522,
      "epoch": 2.2175263197758084,
      "grad_norm": 0.11473718285560608,
      "learning_rate": 9.74072364077551e-05,
      "loss": 1.2611,
      "mean_token_accuracy": 0.7563960090279579,
      "num_tokens": 494399985.0,
      "step": 7320
    },
    {
      "entropy": 0.8576474860310555,
      "epoch": 2.2205559342573657,
      "grad_norm": 0.1286933571100235,
      "learning_rate": 9.739661208532263e-05,
      "loss": 1.2627,
      "mean_token_accuracy": 0.7604630246758461,
      "num_tokens": 495073813.0,
      "step": 7330
    },
    {
      "entropy": 0.8532460451126098,
      "epoch": 2.223585548738923,
      "grad_norm": 0.11478506028652191,
      "learning_rate": 9.738596662163013e-05,
      "loss": 1.2483,
      "mean_token_accuracy": 0.7595999017357826,
      "num_tokens": 495753255.0,
      "step": 7340
    },
    {
      "entropy": 0.8521913647651672,
      "epoch": 2.22661516322048,
      "grad_norm": 0.11295423656702042,
      "learning_rate": 9.737530002142605e-05,
      "loss": 1.2542,
      "mean_token_accuracy": 0.7628868624567986,
      "num_tokens": 496445396.0,
      "step": 7350
    },
    {
      "entropy": 0.8571484595537185,
      "epoch": 2.2296447777020374,
      "grad_norm": 0.10795578360557556,
      "learning_rate": 9.736461228946818e-05,
      "loss": 1.2608,
      "mean_token_accuracy": 0.7577721431851387,
      "num_tokens": 497119111.0,
      "step": 7360
    },
    {
      "entropy": 0.8618745282292366,
      "epoch": 2.2326743921835948,
      "grad_norm": 0.10621679574251175,
      "learning_rate": 9.735390343052382e-05,
      "loss": 1.2645,
      "mean_token_accuracy": 0.7595389366149903,
      "num_tokens": 497791432.0,
      "step": 7370
    },
    {
      "entropy": 0.8541598737239837,
      "epoch": 2.2357040066651517,
      "grad_norm": 0.10267043858766556,
      "learning_rate": 9.734317344936965e-05,
      "loss": 1.2558,
      "mean_token_accuracy": 0.7601015284657479,
      "num_tokens": 498466161.0,
      "step": 7380
    },
    {
      "entropy": 0.8523509263992309,
      "epoch": 2.238733621146709,
      "grad_norm": 0.11477231234312057,
      "learning_rate": 9.733242235079175e-05,
      "loss": 1.2554,
      "mean_token_accuracy": 0.7600444808602334,
      "num_tokens": 499142070.0,
      "step": 7390
    },
    {
      "entropy": 0.8656803876161575,
      "epoch": 2.2417632356282664,
      "grad_norm": 0.10855995863676071,
      "learning_rate": 9.732165013958568e-05,
      "loss": 1.2607,
      "mean_token_accuracy": 0.7546856954693795,
      "num_tokens": 499806749.0,
      "step": 7400
    },
    {
      "entropy": 0.8515672892332077,
      "epoch": 2.2447928501098233,
      "grad_norm": 0.10969103872776031,
      "learning_rate": 9.731085682055639e-05,
      "loss": 1.2594,
      "mean_token_accuracy": 0.7607250720262527,
      "num_tokens": 500488873.0,
      "step": 7410
    },
    {
      "entropy": 0.8555934444069863,
      "epoch": 2.2478224645913807,
      "grad_norm": 0.10826828330755234,
      "learning_rate": 9.73000423985182e-05,
      "loss": 1.2603,
      "mean_token_accuracy": 0.7563047707080841,
      "num_tokens": 501158774.0,
      "step": 7420
    },
    {
      "entropy": 0.8485673606395722,
      "epoch": 2.250852079072938,
      "grad_norm": 0.11583668738603592,
      "learning_rate": 9.728920687829491e-05,
      "loss": 1.2566,
      "mean_token_accuracy": 0.7643196985125542,
      "num_tokens": 501844165.0,
      "step": 7430
    },
    {
      "entropy": 0.8485406875610352,
      "epoch": 2.2538816935544954,
      "grad_norm": 0.11024606227874756,
      "learning_rate": 9.727835026471972e-05,
      "loss": 1.2562,
      "mean_token_accuracy": 0.7618746072053909,
      "num_tokens": 502525376.0,
      "step": 7440
    },
    {
      "entropy": 0.8646971046924591,
      "epoch": 2.2569113080360523,
      "grad_norm": 0.11305391043424606,
      "learning_rate": 9.726747256263518e-05,
      "loss": 1.2677,
      "mean_token_accuracy": 0.7597474843263626,
      "num_tokens": 503209294.0,
      "step": 7450
    },
    {
      "entropy": 0.864198413491249,
      "epoch": 2.2599409225176097,
      "grad_norm": 0.10352534800767899,
      "learning_rate": 9.725657377689334e-05,
      "loss": 1.2644,
      "mean_token_accuracy": 0.7597779706120491,
      "num_tokens": 503897732.0,
      "step": 7460
    },
    {
      "entropy": 0.8614189490675926,
      "epoch": 2.262970536999167,
      "grad_norm": 0.13066239655017853,
      "learning_rate": 9.724565391235556e-05,
      "loss": 1.2657,
      "mean_token_accuracy": 0.7616986453533172,
      "num_tokens": 504586957.0,
      "step": 7470
    },
    {
      "entropy": 0.8593979150056839,
      "epoch": 2.266000151480724,
      "grad_norm": 0.11965595930814743,
      "learning_rate": 9.723471297389268e-05,
      "loss": 1.2606,
      "mean_token_accuracy": 0.7532329395413399,
      "num_tokens": 505243072.0,
      "step": 7480
    },
    {
      "entropy": 0.8494557604193688,
      "epoch": 2.2690297659622813,
      "grad_norm": 0.11761012673377991,
      "learning_rate": 9.722375096638488e-05,
      "loss": 1.2519,
      "mean_token_accuracy": 0.7593458399176598,
      "num_tokens": 505915469.0,
      "step": 7490
    },
    {
      "entropy": 0.864263667166233,
      "epoch": 2.2720593804438387,
      "grad_norm": 0.10211925953626633,
      "learning_rate": 9.721276789472176e-05,
      "loss": 1.2625,
      "mean_token_accuracy": 0.7565105214715004,
      "num_tokens": 506590296.0,
      "step": 7500
    },
    {
      "entropy": 0.8466208800673485,
      "epoch": 2.2750889949253956,
      "grad_norm": 0.11636841297149658,
      "learning_rate": 9.720176376380235e-05,
      "loss": 1.245,
      "mean_token_accuracy": 0.7617765799164772,
      "num_tokens": 507272437.0,
      "step": 7510
    },
    {
      "entropy": 0.8553040236234665,
      "epoch": 2.278118609406953,
      "grad_norm": 0.12769252061843872,
      "learning_rate": 9.719073857853502e-05,
      "loss": 1.2574,
      "mean_token_accuracy": 0.7656926468014718,
      "num_tokens": 507970785.0,
      "step": 7520
    },
    {
      "entropy": 0.852819399535656,
      "epoch": 2.2811482238885104,
      "grad_norm": 0.1222531795501709,
      "learning_rate": 9.717969234383755e-05,
      "loss": 1.2604,
      "mean_token_accuracy": 0.7582620829343796,
      "num_tokens": 508647313.0,
      "step": 7530
    },
    {
      "entropy": 0.8474319830536843,
      "epoch": 2.2841778383700673,
      "grad_norm": 0.10286824405193329,
      "learning_rate": 9.716862506463714e-05,
      "loss": 1.2563,
      "mean_token_accuracy": 0.7589914947748184,
      "num_tokens": 509319096.0,
      "step": 7540
    },
    {
      "entropy": 0.8581209525465965,
      "epoch": 2.2872074528516246,
      "grad_norm": 0.10509907454252243,
      "learning_rate": 9.715753674587032e-05,
      "loss": 1.2632,
      "mean_token_accuracy": 0.7579801797866821,
      "num_tokens": 509986682.0,
      "step": 7550
    },
    {
      "entropy": 0.8578364208340645,
      "epoch": 2.290237067333182,
      "grad_norm": 0.1088315099477768,
      "learning_rate": 9.714642739248305e-05,
      "loss": 1.2674,
      "mean_token_accuracy": 0.756266288459301,
      "num_tokens": 510645010.0,
      "step": 7560
    },
    {
      "entropy": 0.8538104221224785,
      "epoch": 2.293266681814739,
      "grad_norm": 0.1265879124403,
      "learning_rate": 9.713529700943066e-05,
      "loss": 1.2618,
      "mean_token_accuracy": 0.7609230205416679,
      "num_tokens": 511324700.0,
      "step": 7570
    },
    {
      "entropy": 0.8532759755849838,
      "epoch": 2.2962962962962963,
      "grad_norm": 0.11841035634279251,
      "learning_rate": 9.712414560167785e-05,
      "loss": 1.2629,
      "mean_token_accuracy": 0.7549974635243416,
      "num_tokens": 511991352.0,
      "step": 7580
    },
    {
      "entropy": 0.8417635351419449,
      "epoch": 2.2993259107778536,
      "grad_norm": 0.13118776679039001,
      "learning_rate": 9.711297317419871e-05,
      "loss": 1.2454,
      "mean_token_accuracy": 0.7595033630728721,
      "num_tokens": 512663445.0,
      "step": 7590
    },
    {
      "entropy": 0.8628800466656685,
      "epoch": 2.3023555252594106,
      "grad_norm": 0.11126203089952469,
      "learning_rate": 9.710177973197669e-05,
      "loss": 1.2684,
      "mean_token_accuracy": 0.7575622022151947,
      "num_tokens": 513337375.0,
      "step": 7600
    },
    {
      "entropy": 0.8636928454041481,
      "epoch": 2.305385139740968,
      "grad_norm": 0.12146361172199249,
      "learning_rate": 9.709056528000466e-05,
      "loss": 1.2604,
      "mean_token_accuracy": 0.755362045764923,
      "num_tokens": 514023000.0,
      "step": 7610
    },
    {
      "entropy": 0.849440747499466,
      "epoch": 2.3084147542225253,
      "grad_norm": 0.12590289115905762,
      "learning_rate": 9.707932982328478e-05,
      "loss": 1.2552,
      "mean_token_accuracy": 0.7635484874248505,
      "num_tokens": 514706972.0,
      "step": 7620
    },
    {
      "entropy": 0.8544532179832458,
      "epoch": 2.311444368704082,
      "grad_norm": 0.10340185463428497,
      "learning_rate": 9.706807336682866e-05,
      "loss": 1.2625,
      "mean_token_accuracy": 0.7590579405426979,
      "num_tokens": 515388358.0,
      "step": 7630
    },
    {
      "entropy": 0.8562353760004043,
      "epoch": 2.3144739831856396,
      "grad_norm": 0.12309904396533966,
      "learning_rate": 9.705679591565722e-05,
      "loss": 1.2658,
      "mean_token_accuracy": 0.7605538964271545,
      "num_tokens": 516062529.0,
      "step": 7640
    },
    {
      "entropy": 0.8605162307620049,
      "epoch": 2.317503597667197,
      "grad_norm": 0.12778401374816895,
      "learning_rate": 9.704549747480078e-05,
      "loss": 1.2597,
      "mean_token_accuracy": 0.7594474971294403,
      "num_tokens": 516734700.0,
      "step": 7650
    },
    {
      "entropy": 0.8563567087054252,
      "epoch": 2.320533212148754,
      "grad_norm": 0.12819762527942657,
      "learning_rate": 9.703417804929901e-05,
      "loss": 1.265,
      "mean_token_accuracy": 0.755785447359085,
      "num_tokens": 517401137.0,
      "step": 7660
    },
    {
      "entropy": 0.8390669286251068,
      "epoch": 2.323562826630311,
      "grad_norm": 0.12742090225219727,
      "learning_rate": 9.702283764420093e-05,
      "loss": 1.2506,
      "mean_token_accuracy": 0.7612207159399986,
      "num_tokens": 518071824.0,
      "step": 7670
    },
    {
      "entropy": 0.8464131817221642,
      "epoch": 2.3265924411118686,
      "grad_norm": 0.11039908975362778,
      "learning_rate": 9.701147626456495e-05,
      "loss": 1.2511,
      "mean_token_accuracy": 0.7588544338941574,
      "num_tokens": 518731226.0,
      "step": 7680
    },
    {
      "entropy": 0.8588887944817543,
      "epoch": 2.3296220555934255,
      "grad_norm": 0.12836633622646332,
      "learning_rate": 9.700009391545878e-05,
      "loss": 1.264,
      "mean_token_accuracy": 0.7568954274058342,
      "num_tokens": 519409623.0,
      "step": 7690
    },
    {
      "entropy": 0.8526587426662445,
      "epoch": 2.332651670074983,
      "grad_norm": 0.10793481767177582,
      "learning_rate": 9.698869060195957e-05,
      "loss": 1.2673,
      "mean_token_accuracy": 0.7589220061898232,
      "num_tokens": 520080801.0,
      "step": 7700
    },
    {
      "entropy": 0.8565697595477104,
      "epoch": 2.3356812845565402,
      "grad_norm": 0.11040239036083221,
      "learning_rate": 9.69772663291537e-05,
      "loss": 1.2647,
      "mean_token_accuracy": 0.7597812503576279,
      "num_tokens": 520758394.0,
      "step": 7710
    },
    {
      "entropy": 0.8582417130470276,
      "epoch": 2.3387108990380976,
      "grad_norm": 0.11134511977434158,
      "learning_rate": 9.696582110213701e-05,
      "loss": 1.262,
      "mean_token_accuracy": 0.7580493733286857,
      "num_tokens": 521441114.0,
      "step": 7720
    },
    {
      "entropy": 0.8498915374279022,
      "epoch": 2.3417405135196545,
      "grad_norm": 0.11319853365421295,
      "learning_rate": 9.695435492601464e-05,
      "loss": 1.2523,
      "mean_token_accuracy": 0.7624587774276733,
      "num_tokens": 522135170.0,
      "step": 7730
    },
    {
      "entropy": 0.847995363175869,
      "epoch": 2.344770128001212,
      "grad_norm": 0.12444458156824112,
      "learning_rate": 9.694286780590107e-05,
      "loss": 1.2585,
      "mean_token_accuracy": 0.7576172351837158,
      "num_tokens": 522797514.0,
      "step": 7740
    },
    {
      "entropy": 0.8536500424146652,
      "epoch": 2.3477997424827692,
      "grad_norm": 0.12499086558818817,
      "learning_rate": 9.693135974692013e-05,
      "loss": 1.2678,
      "mean_token_accuracy": 0.7565039187669754,
      "num_tokens": 523456154.0,
      "step": 7750
    },
    {
      "entropy": 0.8630647584795952,
      "epoch": 2.350829356964326,
      "grad_norm": 0.11499357968568802,
      "learning_rate": 9.691983075420501e-05,
      "loss": 1.2616,
      "mean_token_accuracy": 0.7558874368667603,
      "num_tokens": 524125446.0,
      "step": 7760
    },
    {
      "entropy": 0.8613532796502114,
      "epoch": 2.3538589714458835,
      "grad_norm": 0.123953677713871,
      "learning_rate": 9.69082808328982e-05,
      "loss": 1.2654,
      "mean_token_accuracy": 0.7564760953187942,
      "num_tokens": 524801334.0,
      "step": 7770
    },
    {
      "entropy": 0.8634515941143036,
      "epoch": 2.356888585927441,
      "grad_norm": 0.12936441600322723,
      "learning_rate": 9.689670998815152e-05,
      "loss": 1.2663,
      "mean_token_accuracy": 0.757455189526081,
      "num_tokens": 525468969.0,
      "step": 7780
    },
    {
      "entropy": 0.848297119140625,
      "epoch": 2.359918200408998,
      "grad_norm": 0.11233241111040115,
      "learning_rate": 9.68851182251262e-05,
      "loss": 1.2582,
      "mean_token_accuracy": 0.7608260795474052,
      "num_tokens": 526146603.0,
      "step": 7790
    },
    {
      "entropy": 0.8517893686890602,
      "epoch": 2.362947814890555,
      "grad_norm": 0.11174257844686508,
      "learning_rate": 9.68735055489927e-05,
      "loss": 1.2597,
      "mean_token_accuracy": 0.7575798913836479,
      "num_tokens": 526811275.0,
      "step": 7800
    },
    {
      "entropy": 0.847201581299305,
      "epoch": 2.3659774293721125,
      "grad_norm": 0.10749486833810806,
      "learning_rate": 9.686187196493088e-05,
      "loss": 1.2547,
      "mean_token_accuracy": 0.7573108091950417,
      "num_tokens": 527476519.0,
      "step": 7810
    },
    {
      "entropy": 0.8587975278496742,
      "epoch": 2.3690070438536694,
      "grad_norm": 0.13182587921619415,
      "learning_rate": 9.685021747812987e-05,
      "loss": 1.2616,
      "mean_token_accuracy": 0.7562750145792961,
      "num_tokens": 528140460.0,
      "step": 7820
    },
    {
      "entropy": 0.858908225595951,
      "epoch": 2.372036658335227,
      "grad_norm": 0.10838868468999863,
      "learning_rate": 9.68385420937882e-05,
      "loss": 1.2605,
      "mean_token_accuracy": 0.7547510161995887,
      "num_tokens": 528809327.0,
      "step": 7830
    },
    {
      "entropy": 0.8573035791516304,
      "epoch": 2.375066272816784,
      "grad_norm": 0.10809095948934555,
      "learning_rate": 9.682684581711363e-05,
      "loss": 1.2669,
      "mean_token_accuracy": 0.7581486374139785,
      "num_tokens": 529484602.0,
      "step": 7840
    },
    {
      "entropy": 0.8556201368570328,
      "epoch": 2.378095887298341,
      "grad_norm": 0.11285743117332458,
      "learning_rate": 9.681512865332332e-05,
      "loss": 1.2614,
      "mean_token_accuracy": 0.7597505211830139,
      "num_tokens": 530147914.0,
      "step": 7850
    },
    {
      "entropy": 0.8480548933148384,
      "epoch": 2.3811255017798985,
      "grad_norm": 0.12628069519996643,
      "learning_rate": 9.680339060764368e-05,
      "loss": 1.265,
      "mean_token_accuracy": 0.7584131717681885,
      "num_tokens": 530804067.0,
      "step": 7860
    },
    {
      "entropy": 0.8499169021844863,
      "epoch": 2.384155116261456,
      "grad_norm": 0.10992208868265152,
      "learning_rate": 9.679163168531049e-05,
      "loss": 1.2519,
      "mean_token_accuracy": 0.7621417880058289,
      "num_tokens": 531485661.0,
      "step": 7870
    },
    {
      "entropy": 0.8616214722394944,
      "epoch": 2.387184730743013,
      "grad_norm": 0.13375818729400635,
      "learning_rate": 9.677985189156882e-05,
      "loss": 1.2666,
      "mean_token_accuracy": 0.7609955608844757,
      "num_tokens": 532167902.0,
      "step": 7880
    },
    {
      "entropy": 0.8562293291091919,
      "epoch": 2.39021434522457,
      "grad_norm": 0.11304410547018051,
      "learning_rate": 9.676805123167302e-05,
      "loss": 1.2552,
      "mean_token_accuracy": 0.7582985505461692,
      "num_tokens": 532847125.0,
      "step": 7890
    },
    {
      "entropy": 0.8536284953355789,
      "epoch": 2.3932439597061275,
      "grad_norm": 0.10804271697998047,
      "learning_rate": 9.675622971088681e-05,
      "loss": 1.2521,
      "mean_token_accuracy": 0.7590174600481987,
      "num_tokens": 533520028.0,
      "step": 7900
    },
    {
      "entropy": 0.8469849348068237,
      "epoch": 2.396273574187685,
      "grad_norm": 0.10462608188390732,
      "learning_rate": 9.674438733448314e-05,
      "loss": 1.2552,
      "mean_token_accuracy": 0.759534765779972,
      "num_tokens": 534190434.0,
      "step": 7910
    },
    {
      "entropy": 0.8400739654898643,
      "epoch": 2.3993031886692417,
      "grad_norm": 0.11971303075551987,
      "learning_rate": 9.673252410774433e-05,
      "loss": 1.2461,
      "mean_token_accuracy": 0.7663201510906219,
      "num_tokens": 534878727.0,
      "step": 7920
    },
    {
      "entropy": 0.8583844676613808,
      "epoch": 2.402332803150799,
      "grad_norm": 0.11179837584495544,
      "learning_rate": 9.672064003596197e-05,
      "loss": 1.2682,
      "mean_token_accuracy": 0.756516107916832,
      "num_tokens": 535542509.0,
      "step": 7930
    },
    {
      "entropy": 0.8607254073023796,
      "epoch": 2.4053624176323565,
      "grad_norm": 0.11369846761226654,
      "learning_rate": 9.670873512443695e-05,
      "loss": 1.2584,
      "mean_token_accuracy": 0.7623056411743164,
      "num_tokens": 536234786.0,
      "step": 7940
    },
    {
      "entropy": 0.8532637119293213,
      "epoch": 2.4083920321139134,
      "grad_norm": 0.10826956480741501,
      "learning_rate": 9.669680937847944e-05,
      "loss": 1.2541,
      "mean_token_accuracy": 0.7596175089478493,
      "num_tokens": 536918744.0,
      "step": 7950
    },
    {
      "entropy": 0.8515122517943382,
      "epoch": 2.4114216465954708,
      "grad_norm": 0.10441789776086807,
      "learning_rate": 9.668486280340894e-05,
      "loss": 1.2622,
      "mean_token_accuracy": 0.7563559800386429,
      "num_tokens": 537579555.0,
      "step": 7960
    },
    {
      "entropy": 0.848784476518631,
      "epoch": 2.414451261077028,
      "grad_norm": 0.10566443204879761,
      "learning_rate": 9.66728954045542e-05,
      "loss": 1.2506,
      "mean_token_accuracy": 0.7639730542898178,
      "num_tokens": 538265841.0,
      "step": 7970
    },
    {
      "entropy": 0.8581619426608086,
      "epoch": 2.417480875558585,
      "grad_norm": 0.1146743893623352,
      "learning_rate": 9.66609071872533e-05,
      "loss": 1.2595,
      "mean_token_accuracy": 0.7589652687311172,
      "num_tokens": 538943989.0,
      "step": 7980
    },
    {
      "entropy": 0.8564366966485977,
      "epoch": 2.4205104900401424,
      "grad_norm": 0.11719758808612823,
      "learning_rate": 9.664889815685358e-05,
      "loss": 1.2619,
      "mean_token_accuracy": 0.7580450057983399,
      "num_tokens": 539628292.0,
      "step": 7990
    },
    {
      "entropy": 0.8529757842421531,
      "epoch": 2.4235401045216998,
      "grad_norm": 0.10858716070652008,
      "learning_rate": 9.663686831871167e-05,
      "loss": 1.2606,
      "mean_token_accuracy": 0.7648607611656189,
      "num_tokens": 540314369.0,
      "step": 8000
    },
    {
      "entropy": 0.8409318551421165,
      "epoch": 2.4265697190032567,
      "grad_norm": 0.1144886389374733,
      "learning_rate": 9.662481767819345e-05,
      "loss": 1.2526,
      "mean_token_accuracy": 0.7646861433982849,
      "num_tokens": 540996019.0,
      "step": 8010
    },
    {
      "entropy": 0.8637088030576706,
      "epoch": 2.429599333484814,
      "grad_norm": 0.1376212239265442,
      "learning_rate": 9.661274624067413e-05,
      "loss": 1.2616,
      "mean_token_accuracy": 0.7563422918319702,
      "num_tokens": 541674215.0,
      "step": 8020
    },
    {
      "entropy": 0.8513024523854256,
      "epoch": 2.4326289479663714,
      "grad_norm": 0.10584916174411774,
      "learning_rate": 9.66006540115382e-05,
      "loss": 1.2619,
      "mean_token_accuracy": 0.7597479805350303,
      "num_tokens": 542352807.0,
      "step": 8030
    },
    {
      "entropy": 0.8591561034321785,
      "epoch": 2.4356585624479283,
      "grad_norm": 0.1118103638291359,
      "learning_rate": 9.658854099617935e-05,
      "loss": 1.2614,
      "mean_token_accuracy": 0.7565526068210602,
      "num_tokens": 543024165.0,
      "step": 8040
    },
    {
      "entropy": 0.8484128370881081,
      "epoch": 2.4386881769294857,
      "grad_norm": 0.10558687150478363,
      "learning_rate": 9.657640720000064e-05,
      "loss": 1.2472,
      "mean_token_accuracy": 0.7617892831563949,
      "num_tokens": 543703859.0,
      "step": 8050
    },
    {
      "entropy": 0.8600789204239845,
      "epoch": 2.441717791411043,
      "grad_norm": 0.1135752871632576,
      "learning_rate": 9.656425262841431e-05,
      "loss": 1.2675,
      "mean_token_accuracy": 0.7570716828107834,
      "num_tokens": 544375173.0,
      "step": 8060
    },
    {
      "entropy": 0.8509885743260384,
      "epoch": 2.4447474058926,
      "grad_norm": 0.11385232955217361,
      "learning_rate": 9.655207728684194e-05,
      "loss": 1.2536,
      "mean_token_accuracy": 0.76125747859478,
      "num_tokens": 545058161.0,
      "step": 8070
    },
    {
      "entropy": 0.8475042149424553,
      "epoch": 2.4477770203741573,
      "grad_norm": 0.12125519663095474,
      "learning_rate": 9.653988118071431e-05,
      "loss": 1.2549,
      "mean_token_accuracy": 0.7572567671537399,
      "num_tokens": 545726905.0,
      "step": 8080
    },
    {
      "entropy": 0.8501904651522636,
      "epoch": 2.4508066348557147,
      "grad_norm": 0.11919412761926651,
      "learning_rate": 9.652766431547151e-05,
      "loss": 1.2545,
      "mean_token_accuracy": 0.7620494902133942,
      "num_tokens": 546411973.0,
      "step": 8090
    },
    {
      "entropy": 0.8457695588469505,
      "epoch": 2.4538362493372716,
      "grad_norm": 0.12159312516450882,
      "learning_rate": 9.651542669656288e-05,
      "loss": 1.2488,
      "mean_token_accuracy": 0.7606399342417717,
      "num_tokens": 547079112.0,
      "step": 8100
    },
    {
      "entropy": 0.8524926468729973,
      "epoch": 2.456865863818829,
      "grad_norm": 0.1109381690621376,
      "learning_rate": 9.650316832944699e-05,
      "loss": 1.2619,
      "mean_token_accuracy": 0.7586096912622452,
      "num_tokens": 547747236.0,
      "step": 8110
    },
    {
      "entropy": 0.8604070246219635,
      "epoch": 2.4598954783003864,
      "grad_norm": 0.11541227251291275,
      "learning_rate": 9.649088921959171e-05,
      "loss": 1.2602,
      "mean_token_accuracy": 0.7589732840657234,
      "num_tokens": 548421661.0,
      "step": 8120
    },
    {
      "entropy": 0.849203696846962,
      "epoch": 2.4629250927819433,
      "grad_norm": 0.12876959145069122,
      "learning_rate": 9.647858937247412e-05,
      "loss": 1.2521,
      "mean_token_accuracy": 0.759013582766056,
      "num_tokens": 549098574.0,
      "step": 8130
    },
    {
      "entropy": 0.8548745185136795,
      "epoch": 2.4659547072635006,
      "grad_norm": 0.12017077207565308,
      "learning_rate": 9.646626879358055e-05,
      "loss": 1.2571,
      "mean_token_accuracy": 0.7566325277090072,
      "num_tokens": 549776319.0,
      "step": 8140
    },
    {
      "entropy": 0.8461618572473526,
      "epoch": 2.468984321745058,
      "grad_norm": 0.11218147724866867,
      "learning_rate": 9.645392748840665e-05,
      "loss": 1.255,
      "mean_token_accuracy": 0.7596992626786232,
      "num_tokens": 550452838.0,
      "step": 8150
    },
    {
      "entropy": 0.8496768981218338,
      "epoch": 2.472013936226615,
      "grad_norm": 0.11221098154783249,
      "learning_rate": 9.644156546245719e-05,
      "loss": 1.2525,
      "mean_token_accuracy": 0.7584793582558632,
      "num_tokens": 551129488.0,
      "step": 8160
    },
    {
      "entropy": 0.8606008782982826,
      "epoch": 2.4750435507081723,
      "grad_norm": 0.10581947863101959,
      "learning_rate": 9.642918272124632e-05,
      "loss": 1.2673,
      "mean_token_accuracy": 0.7597725406289101,
      "num_tokens": 551798387.0,
      "step": 8170
    },
    {
      "entropy": 0.8509682357311249,
      "epoch": 2.4780731651897296,
      "grad_norm": 0.1086905375123024,
      "learning_rate": 9.64167792702973e-05,
      "loss": 1.2501,
      "mean_token_accuracy": 0.7609071403741836,
      "num_tokens": 552473465.0,
      "step": 8180
    },
    {
      "entropy": 0.8588610589504242,
      "epoch": 2.481102779671287,
      "grad_norm": 0.14329372346401215,
      "learning_rate": 9.640435511514274e-05,
      "loss": 1.259,
      "mean_token_accuracy": 0.7589243084192276,
      "num_tokens": 553147564.0,
      "step": 8190
    },
    {
      "entropy": 0.8484311655163765,
      "epoch": 2.484132394152844,
      "grad_norm": 0.12002729624509811,
      "learning_rate": 9.63919102613244e-05,
      "loss": 1.2608,
      "mean_token_accuracy": 0.7560410052537918,
      "num_tokens": 553800562.0,
      "step": 8200
    },
    {
      "entropy": 0.8479263916611671,
      "epoch": 2.4871620086344013,
      "grad_norm": 0.11256726831197739,
      "learning_rate": 9.637944471439333e-05,
      "loss": 1.2631,
      "mean_token_accuracy": 0.7618035539984703,
      "num_tokens": 554477016.0,
      "step": 8210
    },
    {
      "entropy": 0.8599980130791665,
      "epoch": 2.4901916231159587,
      "grad_norm": 0.1089865192770958,
      "learning_rate": 9.636695847990977e-05,
      "loss": 1.2558,
      "mean_token_accuracy": 0.7554180040955544,
      "num_tokens": 555144622.0,
      "step": 8220
    },
    {
      "entropy": 0.8463609889149666,
      "epoch": 2.4932212375975156,
      "grad_norm": 0.1190873235464096,
      "learning_rate": 9.635445156344322e-05,
      "loss": 1.2517,
      "mean_token_accuracy": 0.7621619239449501,
      "num_tokens": 555832382.0,
      "step": 8230
    },
    {
      "entropy": 0.8366684496402741,
      "epoch": 2.496250852079073,
      "grad_norm": 0.11321038007736206,
      "learning_rate": 9.634192397057238e-05,
      "loss": 1.2486,
      "mean_token_accuracy": 0.7627046957612038,
      "num_tokens": 556513252.0,
      "step": 8240
    },
    {
      "entropy": 0.8561166599392891,
      "epoch": 2.4992804665606303,
      "grad_norm": 0.1392555832862854,
      "learning_rate": 9.632937570688518e-05,
      "loss": 1.2649,
      "mean_token_accuracy": 0.7513695076107979,
      "num_tokens": 557164301.0,
      "step": 8250
    },
    {
      "entropy": 0.8640503004193306,
      "epoch": 2.502310081042187,
      "grad_norm": 0.11030290275812149,
      "learning_rate": 9.631680677797879e-05,
      "loss": 1.2621,
      "mean_token_accuracy": 0.7571258053183556,
      "num_tokens": 557841790.0,
      "step": 8260
    },
    {
      "entropy": 0.8509059086441993,
      "epoch": 2.5053396955237446,
      "grad_norm": 0.11236338317394257,
      "learning_rate": 9.630421718945956e-05,
      "loss": 1.2554,
      "mean_token_accuracy": 0.756206126511097,
      "num_tokens": 558503800.0,
      "step": 8270
    },
    {
      "entropy": 0.8505407676100731,
      "epoch": 2.508369310005302,
      "grad_norm": 0.11434158682823181,
      "learning_rate": 9.629160694694311e-05,
      "loss": 1.2553,
      "mean_token_accuracy": 0.7556064292788506,
      "num_tokens": 559161417.0,
      "step": 8280
    },
    {
      "entropy": 0.8487449988722802,
      "epoch": 2.5113989244868593,
      "grad_norm": 0.1086057499051094,
      "learning_rate": 9.627897605605418e-05,
      "loss": 1.2597,
      "mean_token_accuracy": 0.7596222028136254,
      "num_tokens": 559849641.0,
      "step": 8290
    },
    {
      "entropy": 0.846830989420414,
      "epoch": 2.5144285389684162,
      "grad_norm": 0.10776392370462418,
      "learning_rate": 9.626632452242685e-05,
      "loss": 1.262,
      "mean_token_accuracy": 0.7554687961935997,
      "num_tokens": 560503057.0,
      "step": 8300
    },
    {
      "entropy": 0.8512991040945053,
      "epoch": 2.5174581534499736,
      "grad_norm": 0.10779240727424622,
      "learning_rate": 9.625365235170429e-05,
      "loss": 1.2531,
      "mean_token_accuracy": 0.7656571835279464,
      "num_tokens": 561207112.0,
      "step": 8310
    },
    {
      "entropy": 0.8476653531193733,
      "epoch": 2.520487767931531,
      "grad_norm": 0.11635958403348923,
      "learning_rate": 9.624095954953893e-05,
      "loss": 1.2461,
      "mean_token_accuracy": 0.7580608189105987,
      "num_tokens": 561887488.0,
      "step": 8320
    },
    {
      "entropy": 0.8512125059962272,
      "epoch": 2.523517382413088,
      "grad_norm": 0.11220627278089523,
      "learning_rate": 9.62282461215924e-05,
      "loss": 1.2607,
      "mean_token_accuracy": 0.7566897615790367,
      "num_tokens": 562557842.0,
      "step": 8330
    },
    {
      "entropy": 0.8469684407114982,
      "epoch": 2.5265469968946452,
      "grad_norm": 0.11146048456430435,
      "learning_rate": 9.621551207353552e-05,
      "loss": 1.256,
      "mean_token_accuracy": 0.7601208105683327,
      "num_tokens": 563235504.0,
      "step": 8340
    },
    {
      "entropy": 0.859553873538971,
      "epoch": 2.5295766113762026,
      "grad_norm": 0.13174964487552643,
      "learning_rate": 9.620275741104833e-05,
      "loss": 1.258,
      "mean_token_accuracy": 0.75794418156147,
      "num_tokens": 563910671.0,
      "step": 8350
    },
    {
      "entropy": 0.8397681131958962,
      "epoch": 2.5326062258577595,
      "grad_norm": 0.11584554612636566,
      "learning_rate": 9.618998213982003e-05,
      "loss": 1.2471,
      "mean_token_accuracy": 0.768119253218174,
      "num_tokens": 564608284.0,
      "step": 8360
    },
    {
      "entropy": 0.844032707810402,
      "epoch": 2.535635840339317,
      "grad_norm": 0.11399803310632706,
      "learning_rate": 9.617718626554903e-05,
      "loss": 1.2563,
      "mean_token_accuracy": 0.7574190989136695,
      "num_tokens": 565270666.0,
      "step": 8370
    },
    {
      "entropy": 0.8594373747706413,
      "epoch": 2.5386654548208742,
      "grad_norm": 0.11004584282636642,
      "learning_rate": 9.616436979394294e-05,
      "loss": 1.2591,
      "mean_token_accuracy": 0.7594205111265182,
      "num_tokens": 565951343.0,
      "step": 8380
    },
    {
      "entropy": 0.8535692304372787,
      "epoch": 2.541695069302431,
      "grad_norm": 0.12403662502765656,
      "learning_rate": 9.615153273071853e-05,
      "loss": 1.2528,
      "mean_token_accuracy": 0.7585661306977272,
      "num_tokens": 566630426.0,
      "step": 8390
    },
    {
      "entropy": 0.8541465491056442,
      "epoch": 2.5447246837839885,
      "grad_norm": 0.1108674630522728,
      "learning_rate": 9.613867508160179e-05,
      "loss": 1.2597,
      "mean_token_accuracy": 0.7620561316609382,
      "num_tokens": 567303716.0,
      "step": 8400
    },
    {
      "entropy": 0.8517271533608437,
      "epoch": 2.547754298265546,
      "grad_norm": 0.10456592589616776,
      "learning_rate": 9.612579685232788e-05,
      "loss": 1.2541,
      "mean_token_accuracy": 0.7625368073582649,
      "num_tokens": 567979668.0,
      "step": 8410
    },
    {
      "entropy": 0.8534714087843895,
      "epoch": 2.550783912747103,
      "grad_norm": 0.10534250736236572,
      "learning_rate": 9.611289804864112e-05,
      "loss": 1.2575,
      "mean_token_accuracy": 0.7609692618250847,
      "num_tokens": 568660485.0,
      "step": 8420
    },
    {
      "entropy": 0.8569789603352547,
      "epoch": 2.55381352722866,
      "grad_norm": 0.10718728601932526,
      "learning_rate": 9.609997867629501e-05,
      "loss": 1.2631,
      "mean_token_accuracy": 0.7570777237415314,
      "num_tokens": 569338084.0,
      "step": 8430
    },
    {
      "entropy": 0.8605900689959526,
      "epoch": 2.5568431417102175,
      "grad_norm": 0.1172102838754654,
      "learning_rate": 9.608703874105227e-05,
      "loss": 1.2597,
      "mean_token_accuracy": 0.7540593296289444,
      "num_tokens": 570007047.0,
      "step": 8440
    },
    {
      "entropy": 0.856265552341938,
      "epoch": 2.5598727561917745,
      "grad_norm": 0.11838158965110779,
      "learning_rate": 9.607407824868473e-05,
      "loss": 1.2646,
      "mean_token_accuracy": 0.7546383023262024,
      "num_tokens": 570668524.0,
      "step": 8450
    },
    {
      "entropy": 0.8531616821885109,
      "epoch": 2.562902370673332,
      "grad_norm": 0.1120709478855133,
      "learning_rate": 9.606109720497343e-05,
      "loss": 1.2518,
      "mean_token_accuracy": 0.7596362516283989,
      "num_tokens": 571342859.0,
      "step": 8460
    },
    {
      "entropy": 0.8628086686134339,
      "epoch": 2.565931985154889,
      "grad_norm": 0.11527913808822632,
      "learning_rate": 9.604809561570856e-05,
      "loss": 1.2621,
      "mean_token_accuracy": 0.7554846331477165,
      "num_tokens": 572016207.0,
      "step": 8470
    },
    {
      "entropy": 0.853224304318428,
      "epoch": 2.568961599636446,
      "grad_norm": 0.10552668571472168,
      "learning_rate": 9.603507348668947e-05,
      "loss": 1.2648,
      "mean_token_accuracy": 0.7635086283087731,
      "num_tokens": 572701837.0,
      "step": 8480
    },
    {
      "entropy": 0.8454021051526069,
      "epoch": 2.5719912141180035,
      "grad_norm": 0.09288834780454636,
      "learning_rate": 9.602203082372471e-05,
      "loss": 1.2427,
      "mean_token_accuracy": 0.7652703821659088,
      "num_tokens": 573398981.0,
      "step": 8490
    },
    {
      "entropy": 0.8564005300402642,
      "epoch": 2.575020828599561,
      "grad_norm": 0.10734734684228897,
      "learning_rate": 9.600896763263195e-05,
      "loss": 1.2621,
      "mean_token_accuracy": 0.7566014558076859,
      "num_tokens": 574067498.0,
      "step": 8500
    },
    {
      "entropy": 0.8504653915762901,
      "epoch": 2.5780504430811177,
      "grad_norm": 0.1403270959854126,
      "learning_rate": 9.5995883919238e-05,
      "loss": 1.2524,
      "mean_token_accuracy": 0.763724073767662,
      "num_tokens": 574746104.0,
      "step": 8510
    },
    {
      "entropy": 0.8560445606708527,
      "epoch": 2.581080057562675,
      "grad_norm": 0.11350507289171219,
      "learning_rate": 9.598277968937887e-05,
      "loss": 1.2587,
      "mean_token_accuracy": 0.759540781378746,
      "num_tokens": 575420436.0,
      "step": 8520
    },
    {
      "entropy": 0.8538668856024743,
      "epoch": 2.5841096720442325,
      "grad_norm": 0.12417567521333694,
      "learning_rate": 9.596965494889971e-05,
      "loss": 1.2568,
      "mean_token_accuracy": 0.7581377297639846,
      "num_tokens": 576096733.0,
      "step": 8530
    },
    {
      "entropy": 0.8500018835067749,
      "epoch": 2.5871392865257894,
      "grad_norm": 0.11514724791049957,
      "learning_rate": 9.595650970365481e-05,
      "loss": 1.2584,
      "mean_token_accuracy": 0.7608907133340835,
      "num_tokens": 576781076.0,
      "step": 8540
    },
    {
      "entropy": 0.8607743322849274,
      "epoch": 2.5901689010073468,
      "grad_norm": 0.10145779699087143,
      "learning_rate": 9.594334395950759e-05,
      "loss": 1.2569,
      "mean_token_accuracy": 0.7564357355237007,
      "num_tokens": 577459938.0,
      "step": 8550
    },
    {
      "entropy": 0.8551357820630073,
      "epoch": 2.593198515488904,
      "grad_norm": 0.11113659292459488,
      "learning_rate": 9.593015772233065e-05,
      "loss": 1.2564,
      "mean_token_accuracy": 0.761788259446621,
      "num_tokens": 578152991.0,
      "step": 8560
    },
    {
      "entropy": 0.8534735798835754,
      "epoch": 2.596228129970461,
      "grad_norm": 0.12159695476293564,
      "learning_rate": 9.591695099800571e-05,
      "loss": 1.2603,
      "mean_token_accuracy": 0.7570795536041259,
      "num_tokens": 578827823.0,
      "step": 8570
    },
    {
      "entropy": 0.8524635910987854,
      "epoch": 2.5992577444520184,
      "grad_norm": 0.15989543497562408,
      "learning_rate": 9.59037237924236e-05,
      "loss": 1.2525,
      "mean_token_accuracy": 0.7586214616894722,
      "num_tokens": 579496735.0,
      "step": 8580
    },
    {
      "entropy": 0.8524748459458351,
      "epoch": 2.6022873589335758,
      "grad_norm": 0.10653384774923325,
      "learning_rate": 9.589047611148434e-05,
      "loss": 1.2517,
      "mean_token_accuracy": 0.7634116768836975,
      "num_tokens": 580189477.0,
      "step": 8590
    },
    {
      "entropy": 0.8582893222570419,
      "epoch": 2.6053169734151327,
      "grad_norm": 0.12020713090896606,
      "learning_rate": 9.587720796109708e-05,
      "loss": 1.2625,
      "mean_token_accuracy": 0.7602693036198616,
      "num_tokens": 580870093.0,
      "step": 8600
    },
    {
      "entropy": 0.8512812659144402,
      "epoch": 2.60834658789669,
      "grad_norm": 0.11194371432065964,
      "learning_rate": 9.586391934718002e-05,
      "loss": 1.2476,
      "mean_token_accuracy": 0.7562606275081635,
      "num_tokens": 581540651.0,
      "step": 8610
    },
    {
      "entropy": 0.8489811271429062,
      "epoch": 2.6113762023782474,
      "grad_norm": 0.11133172363042831,
      "learning_rate": 9.58506102756606e-05,
      "loss": 1.2496,
      "mean_token_accuracy": 0.759826211631298,
      "num_tokens": 582208839.0,
      "step": 8620
    },
    {
      "entropy": 0.8494002401828766,
      "epoch": 2.6144058168598043,
      "grad_norm": 0.1221701055765152,
      "learning_rate": 9.583728075247532e-05,
      "loss": 1.2603,
      "mean_token_accuracy": 0.7597423046827316,
      "num_tokens": 582891996.0,
      "step": 8630
    },
    {
      "entropy": 0.8590531438589096,
      "epoch": 2.6174354313413617,
      "grad_norm": 0.10177098214626312,
      "learning_rate": 9.582393078356979e-05,
      "loss": 1.2667,
      "mean_token_accuracy": 0.7538132026791573,
      "num_tokens": 583541518.0,
      "step": 8640
    },
    {
      "entropy": 0.8468727216124534,
      "epoch": 2.620465045822919,
      "grad_norm": 0.10644321888685226,
      "learning_rate": 9.581056037489878e-05,
      "loss": 1.2517,
      "mean_token_accuracy": 0.7635338783264161,
      "num_tokens": 584221310.0,
      "step": 8650
    },
    {
      "entropy": 0.8520080730319023,
      "epoch": 2.623494660304476,
      "grad_norm": 0.1347191035747528,
      "learning_rate": 9.579716953242616e-05,
      "loss": 1.2493,
      "mean_token_accuracy": 0.7642081990838051,
      "num_tokens": 584914672.0,
      "step": 8660
    },
    {
      "entropy": 0.8513093635439872,
      "epoch": 2.6265242747860333,
      "grad_norm": 0.14387550950050354,
      "learning_rate": 9.578375826212494e-05,
      "loss": 1.2467,
      "mean_token_accuracy": 0.7608000755310058,
      "num_tokens": 585595603.0,
      "step": 8670
    },
    {
      "entropy": 0.8506295859813691,
      "epoch": 2.6295538892675907,
      "grad_norm": 0.11430154740810394,
      "learning_rate": 9.577032656997718e-05,
      "loss": 1.2562,
      "mean_token_accuracy": 0.7552939653396606,
      "num_tokens": 586260344.0,
      "step": 8680
    },
    {
      "entropy": 0.858209428191185,
      "epoch": 2.632583503749148,
      "grad_norm": 0.13120955228805542,
      "learning_rate": 9.57568744619741e-05,
      "loss": 1.2627,
      "mean_token_accuracy": 0.7558746591210366,
      "num_tokens": 586924039.0,
      "step": 8690
    },
    {
      "entropy": 0.8451697275042533,
      "epoch": 2.635613118230705,
      "grad_norm": 0.11614711582660675,
      "learning_rate": 9.574340194411602e-05,
      "loss": 1.2489,
      "mean_token_accuracy": 0.7614456757903099,
      "num_tokens": 587604761.0,
      "step": 8700
    },
    {
      "entropy": 0.8501728236675262,
      "epoch": 2.6386427327122624,
      "grad_norm": 0.1159096509218216,
      "learning_rate": 9.572990902241236e-05,
      "loss": 1.2565,
      "mean_token_accuracy": 0.76109050065279,
      "num_tokens": 588282551.0,
      "step": 8710
    },
    {
      "entropy": 0.8441739335656167,
      "epoch": 2.6416723471938197,
      "grad_norm": 0.1077175885438919,
      "learning_rate": 9.571639570288162e-05,
      "loss": 1.2532,
      "mean_token_accuracy": 0.7615225374698639,
      "num_tokens": 588960630.0,
      "step": 8720
    },
    {
      "entropy": 0.8558618754148484,
      "epoch": 2.644701961675377,
      "grad_norm": 0.10956843942403793,
      "learning_rate": 9.570286199155143e-05,
      "loss": 1.2664,
      "mean_token_accuracy": 0.7598358571529389,
      "num_tokens": 589633892.0,
      "step": 8730
    },
    {
      "entropy": 0.8522457152605056,
      "epoch": 2.647731576156934,
      "grad_norm": 0.11533006280660629,
      "learning_rate": 9.568930789445849e-05,
      "loss": 1.2527,
      "mean_token_accuracy": 0.7591142773628234,
      "num_tokens": 590310547.0,
      "step": 8740
    },
    {
      "entropy": 0.8420512557029725,
      "epoch": 2.6507611906384914,
      "grad_norm": 0.11325296014547348,
      "learning_rate": 9.567573341764862e-05,
      "loss": 1.2412,
      "mean_token_accuracy": 0.764337295293808,
      "num_tokens": 590990603.0,
      "step": 8750
    },
    {
      "entropy": 0.8466836139559746,
      "epoch": 2.6537908051200487,
      "grad_norm": 0.10968432575464249,
      "learning_rate": 9.56621385671767e-05,
      "loss": 1.2604,
      "mean_token_accuracy": 0.7618458330631256,
      "num_tokens": 591673884.0,
      "step": 8760
    },
    {
      "entropy": 0.8422190606594085,
      "epoch": 2.6568204196016056,
      "grad_norm": 0.12760990858078003,
      "learning_rate": 9.564852334910674e-05,
      "loss": 1.2533,
      "mean_token_accuracy": 0.7592846661806106,
      "num_tokens": 592329661.0,
      "step": 8770
    },
    {
      "entropy": 0.8569976061582565,
      "epoch": 2.659850034083163,
      "grad_norm": 0.10784024745225906,
      "learning_rate": 9.563488776951178e-05,
      "loss": 1.2519,
      "mean_token_accuracy": 0.7618673861026763,
      "num_tokens": 593019514.0,
      "step": 8780
    },
    {
      "entropy": 0.8481792345643043,
      "epoch": 2.6628796485647204,
      "grad_norm": 0.11991298943758011,
      "learning_rate": 9.562123183447399e-05,
      "loss": 1.2556,
      "mean_token_accuracy": 0.7591419890522957,
      "num_tokens": 593688540.0,
      "step": 8790
    },
    {
      "entropy": 0.8476454094052315,
      "epoch": 2.6659092630462773,
      "grad_norm": 0.12877626717090607,
      "learning_rate": 9.560755555008458e-05,
      "loss": 1.2492,
      "mean_token_accuracy": 0.7576417684555053,
      "num_tokens": 594355381.0,
      "step": 8800
    },
    {
      "entropy": 0.8531828731298446,
      "epoch": 2.6689388775278347,
      "grad_norm": 0.11032012104988098,
      "learning_rate": 9.559385892244388e-05,
      "loss": 1.2505,
      "mean_token_accuracy": 0.7598303601145744,
      "num_tokens": 595027868.0,
      "step": 8810
    },
    {
      "entropy": 0.8510458245873451,
      "epoch": 2.671968492009392,
      "grad_norm": 0.10477004945278168,
      "learning_rate": 9.558014195766126e-05,
      "loss": 1.2555,
      "mean_token_accuracy": 0.757819676399231,
      "num_tokens": 595698206.0,
      "step": 8820
    },
    {
      "entropy": 0.8613530620932579,
      "epoch": 2.674998106490949,
      "grad_norm": 0.10690850019454956,
      "learning_rate": 9.556640466185517e-05,
      "loss": 1.2528,
      "mean_token_accuracy": 0.7590271100401879,
      "num_tokens": 596384294.0,
      "step": 8830
    },
    {
      "entropy": 0.8479463741183281,
      "epoch": 2.6780277209725063,
      "grad_norm": 0.10583047568798065,
      "learning_rate": 9.555264704115313e-05,
      "loss": 1.258,
      "mean_token_accuracy": 0.7629940390586853,
      "num_tokens": 597068799.0,
      "step": 8840
    },
    {
      "entropy": 0.8475633636116982,
      "epoch": 2.6810573354540637,
      "grad_norm": 0.11271261423826218,
      "learning_rate": 9.553886910169174e-05,
      "loss": 1.2439,
      "mean_token_accuracy": 0.7668149173259735,
      "num_tokens": 597757095.0,
      "step": 8850
    },
    {
      "entropy": 0.8524195119738579,
      "epoch": 2.6840869499356206,
      "grad_norm": 0.11528690904378891,
      "learning_rate": 9.552507084961664e-05,
      "loss": 1.2622,
      "mean_token_accuracy": 0.7573121473193168,
      "num_tokens": 598434167.0,
      "step": 8860
    },
    {
      "entropy": 0.8472784489393235,
      "epoch": 2.687116564417178,
      "grad_norm": 0.11534834653139114,
      "learning_rate": 9.551125229108253e-05,
      "loss": 1.2424,
      "mean_token_accuracy": 0.7616955548524856,
      "num_tokens": 599119603.0,
      "step": 8870
    },
    {
      "entropy": 0.8592986464500427,
      "epoch": 2.6901461788987353,
      "grad_norm": 0.11634981632232666,
      "learning_rate": 9.54974134322532e-05,
      "loss": 1.2594,
      "mean_token_accuracy": 0.7608154565095901,
      "num_tokens": 599802386.0,
      "step": 8880
    },
    {
      "entropy": 0.8419868484139442,
      "epoch": 2.6931757933802922,
      "grad_norm": 0.11178780347108841,
      "learning_rate": 9.548355427930147e-05,
      "loss": 1.2455,
      "mean_token_accuracy": 0.7621794462203979,
      "num_tokens": 600476266.0,
      "step": 8890
    },
    {
      "entropy": 0.8469268649816513,
      "epoch": 2.6962054078618496,
      "grad_norm": 0.11938447505235672,
      "learning_rate": 9.54696748384092e-05,
      "loss": 1.262,
      "mean_token_accuracy": 0.7612688824534416,
      "num_tokens": 601144964.0,
      "step": 8900
    },
    {
      "entropy": 0.84125135242939,
      "epoch": 2.699235022343407,
      "grad_norm": 0.11642032116651535,
      "learning_rate": 9.545577511576732e-05,
      "loss": 1.2435,
      "mean_token_accuracy": 0.7643848404288291,
      "num_tokens": 601833405.0,
      "step": 8910
    },
    {
      "entropy": 0.853831721842289,
      "epoch": 2.702264636824964,
      "grad_norm": 0.11114302277565002,
      "learning_rate": 9.544185511757581e-05,
      "loss": 1.263,
      "mean_token_accuracy": 0.7591975644230843,
      "num_tokens": 602506466.0,
      "step": 8920
    },
    {
      "entropy": 0.8439517483115196,
      "epoch": 2.7052942513065212,
      "grad_norm": 0.11877685785293579,
      "learning_rate": 9.542791485004368e-05,
      "loss": 1.2515,
      "mean_token_accuracy": 0.7627226859331131,
      "num_tokens": 603183297.0,
      "step": 8930
    },
    {
      "entropy": 0.8474942713975906,
      "epoch": 2.7083238657880786,
      "grad_norm": 0.12535443902015686,
      "learning_rate": 9.5413954319389e-05,
      "loss": 1.2551,
      "mean_token_accuracy": 0.7631326824426651,
      "num_tokens": 603858049.0,
      "step": 8940
    },
    {
      "entropy": 0.8528895542025566,
      "epoch": 2.7113534802696355,
      "grad_norm": 0.10631478577852249,
      "learning_rate": 9.539997353183883e-05,
      "loss": 1.2533,
      "mean_token_accuracy": 0.7621230915188789,
      "num_tokens": 604549132.0,
      "step": 8950
    },
    {
      "entropy": 0.8584895536303521,
      "epoch": 2.714383094751193,
      "grad_norm": 0.10547883063554764,
      "learning_rate": 9.538597249362934e-05,
      "loss": 1.2642,
      "mean_token_accuracy": 0.7588069796562195,
      "num_tokens": 605225654.0,
      "step": 8960
    },
    {
      "entropy": 0.8447905778884888,
      "epoch": 2.7174127092327502,
      "grad_norm": 0.10729092359542847,
      "learning_rate": 9.537195121100568e-05,
      "loss": 1.2521,
      "mean_token_accuracy": 0.7592152699828147,
      "num_tokens": 605897599.0,
      "step": 8970
    },
    {
      "entropy": 0.8485624492168427,
      "epoch": 2.720442323714307,
      "grad_norm": 0.10752914845943451,
      "learning_rate": 9.535790969022202e-05,
      "loss": 1.2514,
      "mean_token_accuracy": 0.7590329334139824,
      "num_tokens": 606578210.0,
      "step": 8980
    },
    {
      "entropy": 0.842961086332798,
      "epoch": 2.7234719381958645,
      "grad_norm": 0.10626719146966934,
      "learning_rate": 9.534384793754163e-05,
      "loss": 1.2468,
      "mean_token_accuracy": 0.7642134800553322,
      "num_tokens": 607269313.0,
      "step": 8990
    },
    {
      "entropy": 0.8540323719382286,
      "epoch": 2.726501552677422,
      "grad_norm": 0.11208371818065643,
      "learning_rate": 9.53297659592367e-05,
      "loss": 1.2571,
      "mean_token_accuracy": 0.7591601192951203,
      "num_tokens": 607942514.0,
      "step": 9000
    },
    {
      "entropy": 0.846892024576664,
      "epoch": 2.729531167158979,
      "grad_norm": 0.11249063909053802,
      "learning_rate": 9.531566376158853e-05,
      "loss": 1.2543,
      "mean_token_accuracy": 0.7622483372688293,
      "num_tokens": 608624206.0,
      "step": 9010
    },
    {
      "entropy": 0.8537290051579476,
      "epoch": 2.732560781640536,
      "grad_norm": 0.10326066613197327,
      "learning_rate": 9.530154135088742e-05,
      "loss": 1.2407,
      "mean_token_accuracy": 0.7624550864100457,
      "num_tokens": 609324533.0,
      "step": 9020
    },
    {
      "entropy": 0.8529239282011986,
      "epoch": 2.7355903961220935,
      "grad_norm": 0.14154255390167236,
      "learning_rate": 9.528739873343264e-05,
      "loss": 1.2546,
      "mean_token_accuracy": 0.7607391595840454,
      "num_tokens": 610004801.0,
      "step": 9030
    },
    {
      "entropy": 0.8454217985272408,
      "epoch": 2.7386200106036505,
      "grad_norm": 0.11787353456020355,
      "learning_rate": 9.52732359155325e-05,
      "loss": 1.2589,
      "mean_token_accuracy": 0.7602300241589546,
      "num_tokens": 610675926.0,
      "step": 9040
    },
    {
      "entropy": 0.8582961186766624,
      "epoch": 2.741649625085208,
      "grad_norm": 0.11585194617509842,
      "learning_rate": 9.525905290350435e-05,
      "loss": 1.2578,
      "mean_token_accuracy": 0.7624749362468719,
      "num_tokens": 611368497.0,
      "step": 9050
    },
    {
      "entropy": 0.8577098563313484,
      "epoch": 2.744679239566765,
      "grad_norm": 0.10689109563827515,
      "learning_rate": 9.524484970367451e-05,
      "loss": 1.2568,
      "mean_token_accuracy": 0.757765942811966,
      "num_tokens": 612039252.0,
      "step": 9060
    },
    {
      "entropy": 0.8351210162043572,
      "epoch": 2.747708854048322,
      "grad_norm": 0.14586885273456573,
      "learning_rate": 9.523062632237834e-05,
      "loss": 1.2332,
      "mean_token_accuracy": 0.7655604347586632,
      "num_tokens": 612720216.0,
      "step": 9070
    },
    {
      "entropy": 0.8519888401031495,
      "epoch": 2.7507384685298795,
      "grad_norm": 0.11551733314990997,
      "learning_rate": 9.521638276596014e-05,
      "loss": 1.2524,
      "mean_token_accuracy": 0.7608656138181686,
      "num_tokens": 613399302.0,
      "step": 9080
    },
    {
      "entropy": 0.8438190400600434,
      "epoch": 2.753768083011437,
      "grad_norm": 0.10600708425045013,
      "learning_rate": 9.520211904077328e-05,
      "loss": 1.2488,
      "mean_token_accuracy": 0.7577302813529968,
      "num_tokens": 614054249.0,
      "step": 9090
    },
    {
      "entropy": 0.8623249784111977,
      "epoch": 2.7567976974929937,
      "grad_norm": 0.10966934263706207,
      "learning_rate": 9.518783515318007e-05,
      "loss": 1.2621,
      "mean_token_accuracy": 0.7588957846164703,
      "num_tokens": 614738179.0,
      "step": 9100
    },
    {
      "entropy": 0.8504249393939972,
      "epoch": 2.759827311974551,
      "grad_norm": 0.12102324515581131,
      "learning_rate": 9.517353110955186e-05,
      "loss": 1.2552,
      "mean_token_accuracy": 0.7596418187022209,
      "num_tokens": 615404091.0,
      "step": 9110
    },
    {
      "entropy": 0.85791544765234,
      "epoch": 2.7628569264561085,
      "grad_norm": 0.11166871339082718,
      "learning_rate": 9.515920691626896e-05,
      "loss": 1.2587,
      "mean_token_accuracy": 0.7599508956074714,
      "num_tokens": 616081723.0,
      "step": 9120
    },
    {
      "entropy": 0.8537204295396805,
      "epoch": 2.765886540937666,
      "grad_norm": 0.11071418970823288,
      "learning_rate": 9.514486257972069e-05,
      "loss": 1.2552,
      "mean_token_accuracy": 0.7543772354722023,
      "num_tokens": 616749032.0,
      "step": 9130
    },
    {
      "entropy": 0.8480722889304161,
      "epoch": 2.7689161554192228,
      "grad_norm": 0.1132015585899353,
      "learning_rate": 9.513049810630533e-05,
      "loss": 1.2527,
      "mean_token_accuracy": 0.7601557701826096,
      "num_tokens": 617429202.0,
      "step": 9140
    },
    {
      "entropy": 0.8600764691829681,
      "epoch": 2.77194576990078,
      "grad_norm": 0.11303647607564926,
      "learning_rate": 9.511611350243014e-05,
      "loss": 1.2642,
      "mean_token_accuracy": 0.7554978132247925,
      "num_tokens": 618089881.0,
      "step": 9150
    },
    {
      "entropy": 0.8386348068714142,
      "epoch": 2.7749753843823375,
      "grad_norm": 0.1146254763007164,
      "learning_rate": 9.51017087745114e-05,
      "loss": 1.2531,
      "mean_token_accuracy": 0.765274378657341,
      "num_tokens": 618768949.0,
      "step": 9160
    },
    {
      "entropy": 0.8436396852135658,
      "epoch": 2.7780049988638944,
      "grad_norm": 0.12573018670082092,
      "learning_rate": 9.508728392897432e-05,
      "loss": 1.2608,
      "mean_token_accuracy": 0.756932207942009,
      "num_tokens": 619426419.0,
      "step": 9170
    },
    {
      "entropy": 0.8480951756238937,
      "epoch": 2.7810346133454518,
      "grad_norm": 0.11268395185470581,
      "learning_rate": 9.50728389722531e-05,
      "loss": 1.2629,
      "mean_token_accuracy": 0.7584605038166046,
      "num_tokens": 620096506.0,
      "step": 9180
    },
    {
      "entropy": 0.8529573127627372,
      "epoch": 2.784064227827009,
      "grad_norm": 0.10895754396915436,
      "learning_rate": 9.505837391079093e-05,
      "loss": 1.2632,
      "mean_token_accuracy": 0.7552319064736366,
      "num_tokens": 620759536.0,
      "step": 9190
    },
    {
      "entropy": 0.8537846297025681,
      "epoch": 2.7870938423085665,
      "grad_norm": 0.11702003329992294,
      "learning_rate": 9.504388875103996e-05,
      "loss": 1.2607,
      "mean_token_accuracy": 0.7622180119156837,
      "num_tokens": 621453454.0,
      "step": 9200
    },
    {
      "entropy": 0.8559833288192749,
      "epoch": 2.7901234567901234,
      "grad_norm": 0.10382211953401566,
      "learning_rate": 9.502938349946125e-05,
      "loss": 1.2613,
      "mean_token_accuracy": 0.7598064169287682,
      "num_tokens": 622140751.0,
      "step": 9210
    },
    {
      "entropy": 0.8430117681622505,
      "epoch": 2.7931530712716808,
      "grad_norm": 0.11125016212463379,
      "learning_rate": 9.501485816252491e-05,
      "loss": 1.2472,
      "mean_token_accuracy": 0.7645782843232155,
      "num_tokens": 622829005.0,
      "step": 9220
    },
    {
      "entropy": 0.8464206829667091,
      "epoch": 2.796182685753238,
      "grad_norm": 0.12690483033657074,
      "learning_rate": 9.500031274670995e-05,
      "loss": 1.2519,
      "mean_token_accuracy": 0.7599757418036461,
      "num_tokens": 623504794.0,
      "step": 9230
    },
    {
      "entropy": 0.8461428165435791,
      "epoch": 2.799212300234795,
      "grad_norm": 0.11492403596639633,
      "learning_rate": 9.498574725850437e-05,
      "loss": 1.2616,
      "mean_token_accuracy": 0.7610984742641449,
      "num_tokens": 624174218.0,
      "step": 9240
    },
    {
      "entropy": 0.8441011562943459,
      "epoch": 2.8022419147163524,
      "grad_norm": 0.1064518615603447,
      "learning_rate": 9.497116170440507e-05,
      "loss": 1.2506,
      "mean_token_accuracy": 0.7619548350572586,
      "num_tokens": 624846524.0,
      "step": 9250
    },
    {
      "entropy": 0.8522069841623306,
      "epoch": 2.80527152919791,
      "grad_norm": 0.11196396499872208,
      "learning_rate": 9.495655609091799e-05,
      "loss": 1.245,
      "mean_token_accuracy": 0.7609677344560624,
      "num_tokens": 625534390.0,
      "step": 9260
    },
    {
      "entropy": 0.8514075130224228,
      "epoch": 2.8083011436794667,
      "grad_norm": 0.11803749948740005,
      "learning_rate": 9.494193042455791e-05,
      "loss": 1.2586,
      "mean_token_accuracy": 0.7552815482020379,
      "num_tokens": 626195246.0,
      "step": 9270
    },
    {
      "entropy": 0.8575049102306366,
      "epoch": 2.811330758161024,
      "grad_norm": 0.12270934879779816,
      "learning_rate": 9.492728471184864e-05,
      "loss": 1.2544,
      "mean_token_accuracy": 0.7574294328689575,
      "num_tokens": 626868897.0,
      "step": 9280
    },
    {
      "entropy": 0.8500569522380829,
      "epoch": 2.8143603726425814,
      "grad_norm": 0.12504854798316956,
      "learning_rate": 9.491261895932291e-05,
      "loss": 1.2463,
      "mean_token_accuracy": 0.7641521871089936,
      "num_tokens": 627559647.0,
      "step": 9290
    },
    {
      "entropy": 0.8494672313332557,
      "epoch": 2.8173899871241384,
      "grad_norm": 0.11502181738615036,
      "learning_rate": 9.489793317352234e-05,
      "loss": 1.2604,
      "mean_token_accuracy": 0.758509735763073,
      "num_tokens": 628243853.0,
      "step": 9300
    },
    {
      "entropy": 0.8530739128589631,
      "epoch": 2.8204196016056957,
      "grad_norm": 0.12078163027763367,
      "learning_rate": 9.488322736099755e-05,
      "loss": 1.2557,
      "mean_token_accuracy": 0.7597883760929107,
      "num_tokens": 628929142.0,
      "step": 9310
    },
    {
      "entropy": 0.8452306926250458,
      "epoch": 2.823449216087253,
      "grad_norm": 0.11570421606302261,
      "learning_rate": 9.486850152830809e-05,
      "loss": 1.2574,
      "mean_token_accuracy": 0.7557153657078743,
      "num_tokens": 629586684.0,
      "step": 9320
    },
    {
      "entropy": 0.8449615359306335,
      "epoch": 2.82647883056881,
      "grad_norm": 0.11672792583703995,
      "learning_rate": 9.485375568202238e-05,
      "loss": 1.2426,
      "mean_token_accuracy": 0.7659830927848816,
      "num_tokens": 630280739.0,
      "step": 9330
    },
    {
      "entropy": 0.8407104566693306,
      "epoch": 2.8295084450503674,
      "grad_norm": 0.11738342046737671,
      "learning_rate": 9.483898982871781e-05,
      "loss": 1.2384,
      "mean_token_accuracy": 0.7619338065385819,
      "num_tokens": 630964866.0,
      "step": 9340
    },
    {
      "entropy": 0.8456329941749573,
      "epoch": 2.8325380595319247,
      "grad_norm": 0.12471571564674377,
      "learning_rate": 9.482420397498069e-05,
      "loss": 1.2542,
      "mean_token_accuracy": 0.7682717412710189,
      "num_tokens": 631653706.0,
      "step": 9350
    },
    {
      "entropy": 0.8532849788665772,
      "epoch": 2.8355676740134816,
      "grad_norm": 0.12070607393980026,
      "learning_rate": 9.480939812740628e-05,
      "loss": 1.2589,
      "mean_token_accuracy": 0.7555012702941895,
      "num_tokens": 632313809.0,
      "step": 9360
    },
    {
      "entropy": 0.847382801771164,
      "epoch": 2.838597288495039,
      "grad_norm": 0.10841323435306549,
      "learning_rate": 9.47945722925987e-05,
      "loss": 1.2511,
      "mean_token_accuracy": 0.7587038695812225,
      "num_tokens": 632987731.0,
      "step": 9370
    },
    {
      "entropy": 0.8449444532394409,
      "epoch": 2.8416269029765964,
      "grad_norm": 0.12055893987417221,
      "learning_rate": 9.477972647717103e-05,
      "loss": 1.253,
      "mean_token_accuracy": 0.7607487499713897,
      "num_tokens": 633656711.0,
      "step": 9380
    },
    {
      "entropy": 0.8486421957612038,
      "epoch": 2.8446565174581533,
      "grad_norm": 0.11401059478521347,
      "learning_rate": 9.476486068774523e-05,
      "loss": 1.253,
      "mean_token_accuracy": 0.7612351104617119,
      "num_tokens": 634327363.0,
      "step": 9390
    },
    {
      "entropy": 0.8498527958989144,
      "epoch": 2.8476861319397107,
      "grad_norm": 0.11407098919153214,
      "learning_rate": 9.47499749309522e-05,
      "loss": 1.2492,
      "mean_token_accuracy": 0.7609426572918891,
      "num_tokens": 635004780.0,
      "step": 9400
    },
    {
      "entropy": 0.8565239489078522,
      "epoch": 2.850715746421268,
      "grad_norm": 0.10461205244064331,
      "learning_rate": 9.473506921343172e-05,
      "loss": 1.2484,
      "mean_token_accuracy": 0.7615654408931732,
      "num_tokens": 635700882.0,
      "step": 9410
    },
    {
      "entropy": 0.8410073146224022,
      "epoch": 2.853745360902825,
      "grad_norm": 0.11110935360193253,
      "learning_rate": 9.472014354183249e-05,
      "loss": 1.2372,
      "mean_token_accuracy": 0.7607096031308174,
      "num_tokens": 636375827.0,
      "step": 9420
    },
    {
      "entropy": 0.8488332957029343,
      "epoch": 2.8567749753843823,
      "grad_norm": 0.10545022040605545,
      "learning_rate": 9.47051979228121e-05,
      "loss": 1.2653,
      "mean_token_accuracy": 0.7549908071756363,
      "num_tokens": 637040252.0,
      "step": 9430
    },
    {
      "entropy": 0.8493894070386887,
      "epoch": 2.8598045898659397,
      "grad_norm": 0.1181030422449112,
      "learning_rate": 9.469023236303707e-05,
      "loss": 1.2487,
      "mean_token_accuracy": 0.7594626814126968,
      "num_tokens": 637705254.0,
      "step": 9440
    },
    {
      "entropy": 0.8443953573703766,
      "epoch": 2.8628342043474966,
      "grad_norm": 0.11179035156965256,
      "learning_rate": 9.467524686918277e-05,
      "loss": 1.248,
      "mean_token_accuracy": 0.7587132945656776,
      "num_tokens": 638373871.0,
      "step": 9450
    },
    {
      "entropy": 0.8514942198991775,
      "epoch": 2.865863818829054,
      "grad_norm": 0.11156836152076721,
      "learning_rate": 9.466024144793346e-05,
      "loss": 1.2563,
      "mean_token_accuracy": 0.7566749766469002,
      "num_tokens": 639043684.0,
      "step": 9460
    },
    {
      "entropy": 0.8520888537168503,
      "epoch": 2.8688934333106113,
      "grad_norm": 0.12392573058605194,
      "learning_rate": 9.464521610598234e-05,
      "loss": 1.255,
      "mean_token_accuracy": 0.7587416484951973,
      "num_tokens": 639718094.0,
      "step": 9470
    },
    {
      "entropy": 0.8556481033563614,
      "epoch": 2.8719230477921682,
      "grad_norm": 0.10400848090648651,
      "learning_rate": 9.463017085003145e-05,
      "loss": 1.2627,
      "mean_token_accuracy": 0.754160338640213,
      "num_tokens": 640383129.0,
      "step": 9480
    },
    {
      "entropy": 0.8500251278281212,
      "epoch": 2.8749526622737256,
      "grad_norm": 0.11797893047332764,
      "learning_rate": 9.461510568679173e-05,
      "loss": 1.2502,
      "mean_token_accuracy": 0.7616827592253685,
      "num_tokens": 641069429.0,
      "step": 9490
    },
    {
      "entropy": 0.8440890595316887,
      "epoch": 2.877982276755283,
      "grad_norm": 0.11008976399898529,
      "learning_rate": 9.460002062298301e-05,
      "loss": 1.2516,
      "mean_token_accuracy": 0.75860475897789,
      "num_tokens": 641729169.0,
      "step": 9500
    },
    {
      "entropy": 0.845183102786541,
      "epoch": 2.88101189123684,
      "grad_norm": 0.10795667767524719,
      "learning_rate": 9.458491566533395e-05,
      "loss": 1.252,
      "mean_token_accuracy": 0.7612995803356171,
      "num_tokens": 642404503.0,
      "step": 9510
    },
    {
      "entropy": 0.8522583171725273,
      "epoch": 2.8840415057183972,
      "grad_norm": 0.11366670578718185,
      "learning_rate": 9.456979082058214e-05,
      "loss": 1.2602,
      "mean_token_accuracy": 0.7658773332834243,
      "num_tokens": 643089662.0,
      "step": 9520
    },
    {
      "entropy": 0.8500687897205352,
      "epoch": 2.8870711201999546,
      "grad_norm": 0.1287437230348587,
      "learning_rate": 9.455464609547401e-05,
      "loss": 1.2587,
      "mean_token_accuracy": 0.7604455679655076,
      "num_tokens": 643768453.0,
      "step": 9530
    },
    {
      "entropy": 0.8454829633235932,
      "epoch": 2.8901007346815115,
      "grad_norm": 0.11401085555553436,
      "learning_rate": 9.453948149676489e-05,
      "loss": 1.2562,
      "mean_token_accuracy": 0.7634023189544678,
      "num_tokens": 644446987.0,
      "step": 9540
    },
    {
      "entropy": 0.855300472676754,
      "epoch": 2.893130349163069,
      "grad_norm": 0.11452749371528625,
      "learning_rate": 9.45242970312189e-05,
      "loss": 1.2562,
      "mean_token_accuracy": 0.7571460902690887,
      "num_tokens": 645111355.0,
      "step": 9550
    },
    {
      "entropy": 0.8505562752485275,
      "epoch": 2.8961599636446262,
      "grad_norm": 0.11944160610437393,
      "learning_rate": 9.45090927056091e-05,
      "loss": 1.2546,
      "mean_token_accuracy": 0.7573200240731239,
      "num_tokens": 645771986.0,
      "step": 9560
    },
    {
      "entropy": 0.8464237809181213,
      "epoch": 2.8991895781261836,
      "grad_norm": 0.12742820382118225,
      "learning_rate": 9.449386852671739e-05,
      "loss": 1.252,
      "mean_token_accuracy": 0.7640555202960968,
      "num_tokens": 646459124.0,
      "step": 9570
    },
    {
      "entropy": 0.8454401388764381,
      "epoch": 2.9022191926077405,
      "grad_norm": 0.11775665730237961,
      "learning_rate": 9.44786245013345e-05,
      "loss": 1.2455,
      "mean_token_accuracy": 0.7580542922019958,
      "num_tokens": 647138288.0,
      "step": 9580
    },
    {
      "entropy": 0.8477908670902252,
      "epoch": 2.905248807089298,
      "grad_norm": 0.11539655923843384,
      "learning_rate": 9.446336063626005e-05,
      "loss": 1.2587,
      "mean_token_accuracy": 0.760678480565548,
      "num_tokens": 647808908.0,
      "step": 9590
    },
    {
      "entropy": 0.8531346842646599,
      "epoch": 2.9082784215708553,
      "grad_norm": 0.10464304685592651,
      "learning_rate": 9.444807693830244e-05,
      "loss": 1.2516,
      "mean_token_accuracy": 0.7555674359202385,
      "num_tokens": 648486324.0,
      "step": 9600
    },
    {
      "entropy": 0.858202300965786,
      "epoch": 2.911308036052412,
      "grad_norm": 0.11530975997447968,
      "learning_rate": 9.443277341427901e-05,
      "loss": 1.2631,
      "mean_token_accuracy": 0.7608538568019867,
      "num_tokens": 649183303.0,
      "step": 9610
    },
    {
      "entropy": 0.8597147583961486,
      "epoch": 2.9143376505339695,
      "grad_norm": 0.11264371871948242,
      "learning_rate": 9.441745007101588e-05,
      "loss": 1.2619,
      "mean_token_accuracy": 0.7595389813184739,
      "num_tokens": 649870902.0,
      "step": 9620
    },
    {
      "entropy": 0.85275067538023,
      "epoch": 2.917367265015527,
      "grad_norm": 0.10551595687866211,
      "learning_rate": 9.440210691534802e-05,
      "loss": 1.2655,
      "mean_token_accuracy": 0.7552141278982163,
      "num_tokens": 650532897.0,
      "step": 9630
    },
    {
      "entropy": 0.839555224776268,
      "epoch": 2.9203968794970843,
      "grad_norm": 0.11519236117601395,
      "learning_rate": 9.438674395411926e-05,
      "loss": 1.2466,
      "mean_token_accuracy": 0.7642404139041901,
      "num_tokens": 651223401.0,
      "step": 9640
    },
    {
      "entropy": 0.8424392610788345,
      "epoch": 2.923426493978641,
      "grad_norm": 0.10668787360191345,
      "learning_rate": 9.437136119418222e-05,
      "loss": 1.2483,
      "mean_token_accuracy": 0.7609749466180802,
      "num_tokens": 651896297.0,
      "step": 9650
    },
    {
      "entropy": 0.8510879307985306,
      "epoch": 2.9264561084601985,
      "grad_norm": 0.1254030466079712,
      "learning_rate": 9.435595864239843e-05,
      "loss": 1.2548,
      "mean_token_accuracy": 0.7605242788791656,
      "num_tokens": 652575573.0,
      "step": 9660
    },
    {
      "entropy": 0.8578651532530784,
      "epoch": 2.929485722941756,
      "grad_norm": 0.10282514989376068,
      "learning_rate": 9.434053630563817e-05,
      "loss": 1.255,
      "mean_token_accuracy": 0.7591852471232414,
      "num_tokens": 653263336.0,
      "step": 9670
    },
    {
      "entropy": 0.8433485999703407,
      "epoch": 2.932515337423313,
      "grad_norm": 0.1159236878156662,
      "learning_rate": 9.432509419078057e-05,
      "loss": 1.248,
      "mean_token_accuracy": 0.7621194779872894,
      "num_tokens": 653942478.0,
      "step": 9680
    },
    {
      "entropy": 0.8420448079705238,
      "epoch": 2.93554495190487,
      "grad_norm": 0.11720075458288193,
      "learning_rate": 9.430963230471359e-05,
      "loss": 1.247,
      "mean_token_accuracy": 0.7566432520747185,
      "num_tokens": 654606479.0,
      "step": 9690
    },
    {
      "entropy": 0.8508445441722869,
      "epoch": 2.9385745663864276,
      "grad_norm": 0.11400197446346283,
      "learning_rate": 9.429415065433401e-05,
      "loss": 1.2606,
      "mean_token_accuracy": 0.751141057908535,
      "num_tokens": 655253275.0,
      "step": 9700
    },
    {
      "entropy": 0.8569239139556885,
      "epoch": 2.9416041808679845,
      "grad_norm": 0.10377223789691925,
      "learning_rate": 9.42786492465474e-05,
      "loss": 1.2569,
      "mean_token_accuracy": 0.7572843462228775,
      "num_tokens": 655934707.0,
      "step": 9710
    },
    {
      "entropy": 0.838527712225914,
      "epoch": 2.944633795349542,
      "grad_norm": 0.11711885035037994,
      "learning_rate": 9.42631280882682e-05,
      "loss": 1.2376,
      "mean_token_accuracy": 0.7575570166110992,
      "num_tokens": 656602773.0,
      "step": 9720
    },
    {
      "entropy": 0.8440342873334885,
      "epoch": 2.947663409831099,
      "grad_norm": 0.11179562658071518,
      "learning_rate": 9.424758718641958e-05,
      "loss": 1.2476,
      "mean_token_accuracy": 0.7630826935172081,
      "num_tokens": 657283167.0,
      "step": 9730
    },
    {
      "entropy": 0.8516422033309936,
      "epoch": 2.950693024312656,
      "grad_norm": 0.10201055556535721,
      "learning_rate": 9.42320265479336e-05,
      "loss": 1.254,
      "mean_token_accuracy": 0.7659321084618569,
      "num_tokens": 657968237.0,
      "step": 9740
    },
    {
      "entropy": 0.8578488811850548,
      "epoch": 2.9537226387942135,
      "grad_norm": 0.15761013329029083,
      "learning_rate": 9.421644617975104e-05,
      "loss": 1.2592,
      "mean_token_accuracy": 0.7623055756092072,
      "num_tokens": 658656196.0,
      "step": 9750
    },
    {
      "entropy": 0.8493423119187355,
      "epoch": 2.956752253275771,
      "grad_norm": 0.1358635574579239,
      "learning_rate": 9.420084608882158e-05,
      "loss": 1.2489,
      "mean_token_accuracy": 0.7584482207894325,
      "num_tokens": 659323514.0,
      "step": 9760
    },
    {
      "entropy": 0.8442327380180359,
      "epoch": 2.9597818677573278,
      "grad_norm": 0.1336761713027954,
      "learning_rate": 9.41852262821036e-05,
      "loss": 1.2525,
      "mean_token_accuracy": 0.76187454611063,
      "num_tokens": 659999460.0,
      "step": 9770
    },
    {
      "entropy": 0.8411857530474662,
      "epoch": 2.962811482238885,
      "grad_norm": 0.12662717700004578,
      "learning_rate": 9.416958676656433e-05,
      "loss": 1.2513,
      "mean_token_accuracy": 0.7660849794745446,
      "num_tokens": 660680894.0,
      "step": 9780
    },
    {
      "entropy": 0.8411580920219421,
      "epoch": 2.9658410967204425,
      "grad_norm": 0.11733353137969971,
      "learning_rate": 9.415392754917976e-05,
      "loss": 1.2424,
      "mean_token_accuracy": 0.7625678777694702,
      "num_tokens": 661357990.0,
      "step": 9790
    },
    {
      "entropy": 0.8468214213848114,
      "epoch": 2.9688707112019994,
      "grad_norm": 0.1152617484331131,
      "learning_rate": 9.41382486369347e-05,
      "loss": 1.2461,
      "mean_token_accuracy": 0.761156065762043,
      "num_tokens": 662039212.0,
      "step": 9800
    },
    {
      "entropy": 0.8458827570080757,
      "epoch": 2.9719003256835568,
      "grad_norm": 0.11981873214244843,
      "learning_rate": 9.412255003682273e-05,
      "loss": 1.2563,
      "mean_token_accuracy": 0.7581099763512611,
      "num_tokens": 662702316.0,
      "step": 9810
    },
    {
      "entropy": 0.8398707866668701,
      "epoch": 2.974929940165114,
      "grad_norm": 0.10634782165288925,
      "learning_rate": 9.41068317558462e-05,
      "loss": 1.2449,
      "mean_token_accuracy": 0.7624537348747253,
      "num_tokens": 663383844.0,
      "step": 9820
    },
    {
      "entropy": 0.8493513718247414,
      "epoch": 2.977959554646671,
      "grad_norm": 0.11004682630300522,
      "learning_rate": 9.409109380101626e-05,
      "loss": 1.2588,
      "mean_token_accuracy": 0.7634190067648887,
      "num_tokens": 664055624.0,
      "step": 9830
    },
    {
      "entropy": 0.844525645673275,
      "epoch": 2.9809891691282284,
      "grad_norm": 0.11285140365362167,
      "learning_rate": 9.407533617935281e-05,
      "loss": 1.2616,
      "mean_token_accuracy": 0.7566604182124138,
      "num_tokens": 664718688.0,
      "step": 9840
    },
    {
      "entropy": 0.8467271819710731,
      "epoch": 2.984018783609786,
      "grad_norm": 0.12681548297405243,
      "learning_rate": 9.405955889788456e-05,
      "loss": 1.2504,
      "mean_token_accuracy": 0.7627516657114028,
      "num_tokens": 665394806.0,
      "step": 9850
    },
    {
      "entropy": 0.8490105450153351,
      "epoch": 2.9870483980913427,
      "grad_norm": 0.12867045402526855,
      "learning_rate": 9.404376196364895e-05,
      "loss": 1.2515,
      "mean_token_accuracy": 0.7587300792336464,
      "num_tokens": 666065349.0,
      "step": 9860
    },
    {
      "entropy": 0.8448463037610054,
      "epoch": 2.9900780125729,
      "grad_norm": 0.13521568477153778,
      "learning_rate": 9.40279453836922e-05,
      "loss": 1.2541,
      "mean_token_accuracy": 0.7597836315631866,
      "num_tokens": 666731085.0,
      "step": 9870
    },
    {
      "entropy": 0.8395397663116455,
      "epoch": 2.9931076270544574,
      "grad_norm": 0.1719292253255844,
      "learning_rate": 9.401210916506931e-05,
      "loss": 1.2468,
      "mean_token_accuracy": 0.7587765008211136,
      "num_tokens": 667396946.0,
      "step": 9880
    },
    {
      "entropy": 0.8458210021257401,
      "epoch": 2.9961372415360144,
      "grad_norm": 0.13949954509735107,
      "learning_rate": 9.399625331484404e-05,
      "loss": 1.2556,
      "mean_token_accuracy": 0.7590783119201661,
      "num_tokens": 668060048.0,
      "step": 9890
    },
    {
      "entropy": 0.8394185081124306,
      "epoch": 2.9991668560175717,
      "grad_norm": 0.11261207610368729,
      "learning_rate": 9.398037784008884e-05,
      "loss": 1.2522,
      "mean_token_accuracy": 0.7638598397374153,
      "num_tokens": 668730199.0,
      "step": 9900
    },
    {
      "entropy": 0.8383808151269571,
      "epoch": 3.00212073013709,
      "grad_norm": 0.11749439686536789,
      "learning_rate": 9.396448274788501e-05,
      "loss": 1.2364,
      "mean_token_accuracy": 0.7646774450937907,
      "num_tokens": 669393471.0,
      "step": 9910
    },
    {
      "entropy": 0.8383400484919548,
      "epoch": 3.0051503446186474,
      "grad_norm": 0.1319848746061325,
      "learning_rate": 9.394856804532253e-05,
      "loss": 1.2414,
      "mean_token_accuracy": 0.7644588783383369,
      "num_tokens": 670088022.0,
      "step": 9920
    },
    {
      "entropy": 0.8331081718206406,
      "epoch": 3.0081799591002043,
      "grad_norm": 0.11742068082094193,
      "learning_rate": 9.393263373950019e-05,
      "loss": 1.2302,
      "mean_token_accuracy": 0.7647489890456199,
      "num_tokens": 670768593.0,
      "step": 9930
    },
    {
      "entropy": 0.8413022235035896,
      "epoch": 3.0112095735817617,
      "grad_norm": 0.1186077669262886,
      "learning_rate": 9.391667983752545e-05,
      "loss": 1.2459,
      "mean_token_accuracy": 0.7609393790364265,
      "num_tokens": 671443863.0,
      "step": 9940
    },
    {
      "entropy": 0.846348412334919,
      "epoch": 3.014239188063319,
      "grad_norm": 0.13650287687778473,
      "learning_rate": 9.390070634651458e-05,
      "loss": 1.2441,
      "mean_token_accuracy": 0.7602412641048432,
      "num_tokens": 672137443.0,
      "step": 9950
    },
    {
      "entropy": 0.8487726628780365,
      "epoch": 3.017268802544876,
      "grad_norm": 0.1162915825843811,
      "learning_rate": 9.388471327359253e-05,
      "loss": 1.2464,
      "mean_token_accuracy": 0.7582907408475876,
      "num_tokens": 672819109.0,
      "step": 9960
    },
    {
      "entropy": 0.8265037685632706,
      "epoch": 3.0202984170264333,
      "grad_norm": 0.12178266048431396,
      "learning_rate": 9.386870062589301e-05,
      "loss": 1.2318,
      "mean_token_accuracy": 0.76134063154459,
      "num_tokens": 673478301.0,
      "step": 9970
    },
    {
      "entropy": 0.8373795002698898,
      "epoch": 3.0233280315079907,
      "grad_norm": 0.11897668987512589,
      "learning_rate": 9.385266841055849e-05,
      "loss": 1.2412,
      "mean_token_accuracy": 0.7585442930459976,
      "num_tokens": 674148719.0,
      "step": 9980
    },
    {
      "entropy": 0.8317443907260895,
      "epoch": 3.026357645989548,
      "grad_norm": 0.11980799585580826,
      "learning_rate": 9.383661663474011e-05,
      "loss": 1.2352,
      "mean_token_accuracy": 0.7605805337429047,
      "num_tokens": 674825439.0,
      "step": 9990
    },
    {
      "entropy": 0.8385573863983155,
      "epoch": 3.029387260471105,
      "grad_norm": 0.12173496186733246,
      "learning_rate": 9.382054530559779e-05,
      "loss": 1.2393,
      "mean_token_accuracy": 0.7657988607883454,
      "num_tokens": 675505171.0,
      "step": 10000
    },
    {
      "entropy": 0.8310229733586312,
      "epoch": 3.0324168749526623,
      "grad_norm": 0.10542073100805283,
      "learning_rate": 9.380445443030011e-05,
      "loss": 1.2316,
      "mean_token_accuracy": 0.7675056278705596,
      "num_tokens": 676199738.0,
      "step": 10010
    },
    {
      "entropy": 0.8390552252531052,
      "epoch": 3.0354464894342197,
      "grad_norm": 0.12213842570781708,
      "learning_rate": 9.378834401602444e-05,
      "loss": 1.2343,
      "mean_token_accuracy": 0.761818116903305,
      "num_tokens": 676876944.0,
      "step": 10020
    },
    {
      "entropy": 0.835042978823185,
      "epoch": 3.0384761039157766,
      "grad_norm": 0.11653364449739456,
      "learning_rate": 9.377221406995683e-05,
      "loss": 1.2393,
      "mean_token_accuracy": 0.759946870803833,
      "num_tokens": 677544094.0,
      "step": 10030
    },
    {
      "entropy": 0.8409670516848564,
      "epoch": 3.041505718397334,
      "grad_norm": 0.12055815011262894,
      "learning_rate": 9.375606459929202e-05,
      "loss": 1.2363,
      "mean_token_accuracy": 0.7609785079956055,
      "num_tokens": 678233082.0,
      "step": 10040
    },
    {
      "entropy": 0.8427327066659928,
      "epoch": 3.0445353328788913,
      "grad_norm": 0.13013114035129547,
      "learning_rate": 9.373989561123351e-05,
      "loss": 1.2578,
      "mean_token_accuracy": 0.7567624032497406,
      "num_tokens": 678895505.0,
      "step": 10050
    },
    {
      "entropy": 0.8381952315568924,
      "epoch": 3.0475649473604483,
      "grad_norm": 0.13608799874782562,
      "learning_rate": 9.372370711299344e-05,
      "loss": 1.2403,
      "mean_token_accuracy": 0.7570142805576324,
      "num_tokens": 679566458.0,
      "step": 10060
    },
    {
      "entropy": 0.8234174102544785,
      "epoch": 3.0505945618420056,
      "grad_norm": 0.11896231770515442,
      "learning_rate": 9.370749911179272e-05,
      "loss": 1.2314,
      "mean_token_accuracy": 0.764320008456707,
      "num_tokens": 680241292.0,
      "step": 10070
    },
    {
      "entropy": 0.8354604631662369,
      "epoch": 3.053624176323563,
      "grad_norm": 0.1257018893957138,
      "learning_rate": 9.369127161486093e-05,
      "loss": 1.2323,
      "mean_token_accuracy": 0.7660845428705215,
      "num_tokens": 680927192.0,
      "step": 10080
    },
    {
      "entropy": 0.8369146749377251,
      "epoch": 3.05665379080512,
      "grad_norm": 0.1210678443312645,
      "learning_rate": 9.367502462943634e-05,
      "loss": 1.2363,
      "mean_token_accuracy": 0.7641615822911263,
      "num_tokens": 681606369.0,
      "step": 10090
    },
    {
      "entropy": 0.834869273006916,
      "epoch": 3.0596834052866773,
      "grad_norm": 0.12320520728826523,
      "learning_rate": 9.36587581627659e-05,
      "loss": 1.2454,
      "mean_token_accuracy": 0.7614041447639466,
      "num_tokens": 682271842.0,
      "step": 10100
    },
    {
      "entropy": 0.8204829946160317,
      "epoch": 3.0627130197682346,
      "grad_norm": 0.1113852709531784,
      "learning_rate": 9.364247222210529e-05,
      "loss": 1.2235,
      "mean_token_accuracy": 0.7688616901636124,
      "num_tokens": 682962488.0,
      "step": 10110
    },
    {
      "entropy": 0.8474239587783814,
      "epoch": 3.0657426342497915,
      "grad_norm": 0.13586878776550293,
      "learning_rate": 9.362616681471885e-05,
      "loss": 1.2513,
      "mean_token_accuracy": 0.7578161045908928,
      "num_tokens": 683631167.0,
      "step": 10120
    },
    {
      "entropy": 0.8334722653031349,
      "epoch": 3.068772248731349,
      "grad_norm": 0.11796333640813828,
      "learning_rate": 9.360984194787958e-05,
      "loss": 1.2295,
      "mean_token_accuracy": 0.7639172211289406,
      "num_tokens": 684308264.0,
      "step": 10130
    },
    {
      "entropy": 0.8372927248477936,
      "epoch": 3.0718018632129063,
      "grad_norm": 0.12656259536743164,
      "learning_rate": 9.359349762886925e-05,
      "loss": 1.2366,
      "mean_token_accuracy": 0.7626647621393203,
      "num_tokens": 684987150.0,
      "step": 10140
    },
    {
      "entropy": 0.8288713663816452,
      "epoch": 3.074831477694463,
      "grad_norm": 0.11318030208349228,
      "learning_rate": 9.357713386497819e-05,
      "loss": 1.2382,
      "mean_token_accuracy": 0.7620742693543434,
      "num_tokens": 685650353.0,
      "step": 10150
    },
    {
      "entropy": 0.8404918104410172,
      "epoch": 3.0778610921760206,
      "grad_norm": 0.12882491946220398,
      "learning_rate": 9.356075066350547e-05,
      "loss": 1.2387,
      "mean_token_accuracy": 0.7626294881105423,
      "num_tokens": 686337380.0,
      "step": 10160
    },
    {
      "entropy": 0.8249937057495117,
      "epoch": 3.080890706657578,
      "grad_norm": 0.11957137286663055,
      "learning_rate": 9.35443480317588e-05,
      "loss": 1.2305,
      "mean_token_accuracy": 0.758065702021122,
      "num_tokens": 686999032.0,
      "step": 10170
    },
    {
      "entropy": 0.8310855448246002,
      "epoch": 3.083920321139135,
      "grad_norm": 0.12973418831825256,
      "learning_rate": 9.352792597705462e-05,
      "loss": 1.2318,
      "mean_token_accuracy": 0.7651878580451011,
      "num_tokens": 687684037.0,
      "step": 10180
    },
    {
      "entropy": 0.8338297083973885,
      "epoch": 3.086949935620692,
      "grad_norm": 0.13006281852722168,
      "learning_rate": 9.351148450671796e-05,
      "loss": 1.2387,
      "mean_token_accuracy": 0.7593083173036576,
      "num_tokens": 688341905.0,
      "step": 10190
    },
    {
      "entropy": 0.8397548958659172,
      "epoch": 3.0899795501022496,
      "grad_norm": 0.12187941372394562,
      "learning_rate": 9.349502362808255e-05,
      "loss": 1.242,
      "mean_token_accuracy": 0.7532096341252327,
      "num_tokens": 688989343.0,
      "step": 10200
    },
    {
      "entropy": 0.8340567663311959,
      "epoch": 3.093009164583807,
      "grad_norm": 0.12024606764316559,
      "learning_rate": 9.347854334849073e-05,
      "loss": 1.2446,
      "mean_token_accuracy": 0.759178264439106,
      "num_tokens": 689644171.0,
      "step": 10210
    },
    {
      "entropy": 0.8291385605931282,
      "epoch": 3.096038779065364,
      "grad_norm": 0.11374744027853012,
      "learning_rate": 9.346204367529359e-05,
      "loss": 1.2291,
      "mean_token_accuracy": 0.7584012091159821,
      "num_tokens": 690303957.0,
      "step": 10220
    },
    {
      "entropy": 0.8290810093283654,
      "epoch": 3.099068393546921,
      "grad_norm": 0.1235131099820137,
      "learning_rate": 9.344552461585074e-05,
      "loss": 1.2315,
      "mean_token_accuracy": 0.7684821531176567,
      "num_tokens": 691005466.0,
      "step": 10230
    },
    {
      "entropy": 0.8295746877789497,
      "epoch": 3.1020980080284786,
      "grad_norm": 0.12789379060268402,
      "learning_rate": 9.342898617753056e-05,
      "loss": 1.2258,
      "mean_token_accuracy": 0.7688084349036217,
      "num_tokens": 691706441.0,
      "step": 10240
    },
    {
      "entropy": 0.8276756659150124,
      "epoch": 3.1051276225100355,
      "grad_norm": 0.11145539581775665,
      "learning_rate": 9.341242836771e-05,
      "loss": 1.2299,
      "mean_token_accuracy": 0.7633999362587929,
      "num_tokens": 692381849.0,
      "step": 10250
    },
    {
      "entropy": 0.8328027233481408,
      "epoch": 3.108157236991593,
      "grad_norm": 0.11050871014595032,
      "learning_rate": 9.339585119377468e-05,
      "loss": 1.2355,
      "mean_token_accuracy": 0.7645954817533493,
      "num_tokens": 693057008.0,
      "step": 10260
    },
    {
      "entropy": 0.8326576486229896,
      "epoch": 3.11118685147315,
      "grad_norm": 0.12428499013185501,
      "learning_rate": 9.337925466311883e-05,
      "loss": 1.2403,
      "mean_token_accuracy": 0.7647704347968102,
      "num_tokens": 693731362.0,
      "step": 10270
    },
    {
      "entropy": 0.8436383217573166,
      "epoch": 3.114216465954707,
      "grad_norm": 0.12997333705425262,
      "learning_rate": 9.336263878314536e-05,
      "loss": 1.2481,
      "mean_token_accuracy": 0.7621228620409966,
      "num_tokens": 694396353.0,
      "step": 10280
    },
    {
      "entropy": 0.8330873399972916,
      "epoch": 3.1172460804362645,
      "grad_norm": 0.11834542453289032,
      "learning_rate": 9.334600356126575e-05,
      "loss": 1.2338,
      "mean_token_accuracy": 0.7599046424031257,
      "num_tokens": 695064359.0,
      "step": 10290
    },
    {
      "entropy": 0.8297991320490837,
      "epoch": 3.120275694917822,
      "grad_norm": 0.1220044493675232,
      "learning_rate": 9.332934900490018e-05,
      "loss": 1.2328,
      "mean_token_accuracy": 0.7637186720967293,
      "num_tokens": 695740694.0,
      "step": 10300
    },
    {
      "entropy": 0.835815292596817,
      "epoch": 3.123305309399379,
      "grad_norm": 0.12407717108726501,
      "learning_rate": 9.331267512147739e-05,
      "loss": 1.2361,
      "mean_token_accuracy": 0.7623964294791221,
      "num_tokens": 696426667.0,
      "step": 10310
    },
    {
      "entropy": 0.833484211564064,
      "epoch": 3.126334923880936,
      "grad_norm": 0.14744287729263306,
      "learning_rate": 9.329598191843478e-05,
      "loss": 1.2382,
      "mean_token_accuracy": 0.7577978476881981,
      "num_tokens": 697086202.0,
      "step": 10320
    },
    {
      "entropy": 0.8317460626363754,
      "epoch": 3.1293645383624935,
      "grad_norm": 0.11609450727701187,
      "learning_rate": 9.327926940321833e-05,
      "loss": 1.2422,
      "mean_token_accuracy": 0.7573364853858948,
      "num_tokens": 697735782.0,
      "step": 10330
    },
    {
      "entropy": 0.8358004689216614,
      "epoch": 3.1323941528440504,
      "grad_norm": 0.11458564549684525,
      "learning_rate": 9.326253758328269e-05,
      "loss": 1.242,
      "mean_token_accuracy": 0.7641695529222489,
      "num_tokens": 698409023.0,
      "step": 10340
    },
    {
      "entropy": 0.8372511133551598,
      "epoch": 3.135423767325608,
      "grad_norm": 0.14164815843105316,
      "learning_rate": 9.324578646609106e-05,
      "loss": 1.2385,
      "mean_token_accuracy": 0.7637021109461785,
      "num_tokens": 699084224.0,
      "step": 10350
    },
    {
      "entropy": 0.8225207179784775,
      "epoch": 3.138453381807165,
      "grad_norm": 0.12220578640699387,
      "learning_rate": 9.32290160591153e-05,
      "loss": 1.2345,
      "mean_token_accuracy": 0.7701056480407715,
      "num_tokens": 699787268.0,
      "step": 10360
    },
    {
      "entropy": 0.8347537323832512,
      "epoch": 3.141482996288722,
      "grad_norm": 0.1229391098022461,
      "learning_rate": 9.321222636983581e-05,
      "loss": 1.2374,
      "mean_token_accuracy": 0.7667005762457848,
      "num_tokens": 700482735.0,
      "step": 10370
    },
    {
      "entropy": 0.8306485459208488,
      "epoch": 3.1445126107702794,
      "grad_norm": 0.11771932244300842,
      "learning_rate": 9.319541740574169e-05,
      "loss": 1.2281,
      "mean_token_accuracy": 0.765276812016964,
      "num_tokens": 701165632.0,
      "step": 10380
    },
    {
      "entropy": 0.8387805730104446,
      "epoch": 3.147542225251837,
      "grad_norm": 0.11932957917451859,
      "learning_rate": 9.317858917433053e-05,
      "loss": 1.2384,
      "mean_token_accuracy": 0.7616775602102279,
      "num_tokens": 701847430.0,
      "step": 10390
    },
    {
      "entropy": 0.8426155775785447,
      "epoch": 3.1505718397333937,
      "grad_norm": 0.10999251157045364,
      "learning_rate": 9.31617416831086e-05,
      "loss": 1.2543,
      "mean_token_accuracy": 0.7565399438142777,
      "num_tokens": 702502651.0,
      "step": 10400
    },
    {
      "entropy": 0.8212842464447021,
      "epoch": 3.153601454214951,
      "grad_norm": 0.1370965838432312,
      "learning_rate": 9.31448749395907e-05,
      "loss": 1.2322,
      "mean_token_accuracy": 0.7635385498404503,
      "num_tokens": 703176499.0,
      "step": 10410
    },
    {
      "entropy": 0.8400916963815689,
      "epoch": 3.1566310686965084,
      "grad_norm": 0.11962023377418518,
      "learning_rate": 9.312798895130024e-05,
      "loss": 1.2352,
      "mean_token_accuracy": 0.7685959905385971,
      "num_tokens": 703883910.0,
      "step": 10420
    },
    {
      "entropy": 0.8473451912403107,
      "epoch": 3.159660683178066,
      "grad_norm": 0.13253596425056458,
      "learning_rate": 9.311108372576923e-05,
      "loss": 1.2481,
      "mean_token_accuracy": 0.7600852161645889,
      "num_tokens": 704558968.0,
      "step": 10430
    },
    {
      "entropy": 0.8328047186136246,
      "epoch": 3.1626902976596227,
      "grad_norm": 0.12212762981653214,
      "learning_rate": 9.309415927053825e-05,
      "loss": 1.2324,
      "mean_token_accuracy": 0.7633449226617813,
      "num_tokens": 705240129.0,
      "step": 10440
    },
    {
      "entropy": 0.8332623779773712,
      "epoch": 3.16571991214118,
      "grad_norm": 0.11805460602045059,
      "learning_rate": 9.307721559315644e-05,
      "loss": 1.2407,
      "mean_token_accuracy": 0.7617170751094818,
      "num_tokens": 705910419.0,
      "step": 10450
    },
    {
      "entropy": 0.8405819192528725,
      "epoch": 3.1687495266227375,
      "grad_norm": 0.11516073346138,
      "learning_rate": 9.30602527011815e-05,
      "loss": 1.2472,
      "mean_token_accuracy": 0.7590625390410424,
      "num_tokens": 706572226.0,
      "step": 10460
    },
    {
      "entropy": 0.8321726381778717,
      "epoch": 3.1717791411042944,
      "grad_norm": 0.11908750236034393,
      "learning_rate": 9.30432706021798e-05,
      "loss": 1.229,
      "mean_token_accuracy": 0.7679396644234657,
      "num_tokens": 707273378.0,
      "step": 10470
    },
    {
      "entropy": 0.8422200918197632,
      "epoch": 3.1748087555858517,
      "grad_norm": 0.12345686554908752,
      "learning_rate": 9.302626930372613e-05,
      "loss": 1.2456,
      "mean_token_accuracy": 0.7596239730715751,
      "num_tokens": 707941328.0,
      "step": 10480
    },
    {
      "entropy": 0.8314088940620422,
      "epoch": 3.177838370067409,
      "grad_norm": 0.10889705270528793,
      "learning_rate": 9.300924881340397e-05,
      "loss": 1.2275,
      "mean_token_accuracy": 0.7671215176582337,
      "num_tokens": 708635710.0,
      "step": 10490
    },
    {
      "entropy": 0.8333190977573395,
      "epoch": 3.180867984548966,
      "grad_norm": 0.11417516320943832,
      "learning_rate": 9.29922091388053e-05,
      "loss": 1.2368,
      "mean_token_accuracy": 0.7640294909477234,
      "num_tokens": 709313698.0,
      "step": 10500
    },
    {
      "entropy": 0.8399480730295181,
      "epoch": 3.1838975990305234,
      "grad_norm": 0.12141840904951096,
      "learning_rate": 9.297515028753065e-05,
      "loss": 1.2379,
      "mean_token_accuracy": 0.7629139348864555,
      "num_tokens": 710004897.0,
      "step": 10510
    },
    {
      "entropy": 0.8384797543287277,
      "epoch": 3.1869272135120807,
      "grad_norm": 0.17831021547317505,
      "learning_rate": 9.295807226718913e-05,
      "loss": 1.2441,
      "mean_token_accuracy": 0.7584279000759124,
      "num_tokens": 710660224.0,
      "step": 10520
    },
    {
      "entropy": 0.8309674605727195,
      "epoch": 3.1899568279936377,
      "grad_norm": 0.12027305364608765,
      "learning_rate": 9.294097508539841e-05,
      "loss": 1.2309,
      "mean_token_accuracy": 0.7637091666460037,
      "num_tokens": 711350685.0,
      "step": 10530
    },
    {
      "entropy": 0.8398581445217133,
      "epoch": 3.192986442475195,
      "grad_norm": 0.12445776909589767,
      "learning_rate": 9.292385874978467e-05,
      "loss": 1.2428,
      "mean_token_accuracy": 0.7635554268956184,
      "num_tokens": 712029223.0,
      "step": 10540
    },
    {
      "entropy": 0.8296842262148857,
      "epoch": 3.1960160569567524,
      "grad_norm": 0.11100509762763977,
      "learning_rate": 9.290672326798264e-05,
      "loss": 1.2374,
      "mean_token_accuracy": 0.7651006951928139,
      "num_tokens": 712718620.0,
      "step": 10550
    },
    {
      "entropy": 0.8360298171639442,
      "epoch": 3.1990456714383093,
      "grad_norm": 0.12844832241535187,
      "learning_rate": 9.288956864763564e-05,
      "loss": 1.2378,
      "mean_token_accuracy": 0.7630566358566284,
      "num_tokens": 713401466.0,
      "step": 10560
    },
    {
      "entropy": 0.8393256574869156,
      "epoch": 3.2020752859198667,
      "grad_norm": 0.13267198204994202,
      "learning_rate": 9.287239489639545e-05,
      "loss": 1.2478,
      "mean_token_accuracy": 0.7581466615200043,
      "num_tokens": 714065572.0,
      "step": 10570
    },
    {
      "entropy": 0.8247277021408081,
      "epoch": 3.205104900401424,
      "grad_norm": 0.11070942878723145,
      "learning_rate": 9.285520202192244e-05,
      "loss": 1.2358,
      "mean_token_accuracy": 0.760840256512165,
      "num_tokens": 714725833.0,
      "step": 10580
    },
    {
      "entropy": 0.832079616189003,
      "epoch": 3.208134514882981,
      "grad_norm": 0.11494448035955429,
      "learning_rate": 9.283799003188549e-05,
      "loss": 1.2348,
      "mean_token_accuracy": 0.7642373278737068,
      "num_tokens": 715410879.0,
      "step": 10590
    },
    {
      "entropy": 0.832519993185997,
      "epoch": 3.2111641293645383,
      "grad_norm": 0.1271791309118271,
      "learning_rate": 9.2820758933962e-05,
      "loss": 1.2334,
      "mean_token_accuracy": 0.7576231256127357,
      "num_tokens": 716079706.0,
      "step": 10600
    },
    {
      "entropy": 0.8424416229128837,
      "epoch": 3.2141937438460957,
      "grad_norm": 0.1314493715763092,
      "learning_rate": 9.280350873583792e-05,
      "loss": 1.2455,
      "mean_token_accuracy": 0.7570266142487526,
      "num_tokens": 716742397.0,
      "step": 10610
    },
    {
      "entropy": 0.8290569439530373,
      "epoch": 3.2172233583276526,
      "grad_norm": 0.13425303995609283,
      "learning_rate": 9.27862394452077e-05,
      "loss": 1.2326,
      "mean_token_accuracy": 0.7638805478811264,
      "num_tokens": 717422160.0,
      "step": 10620
    },
    {
      "entropy": 0.8295670047402381,
      "epoch": 3.22025297280921,
      "grad_norm": 0.11815513670444489,
      "learning_rate": 9.276895106977428e-05,
      "loss": 1.2295,
      "mean_token_accuracy": 0.7663416728377342,
      "num_tokens": 718108091.0,
      "step": 10630
    },
    {
      "entropy": 0.8376719772815704,
      "epoch": 3.2232825872907673,
      "grad_norm": 0.13028369843959808,
      "learning_rate": 9.275164361724917e-05,
      "loss": 1.2336,
      "mean_token_accuracy": 0.7605687156319618,
      "num_tokens": 718791576.0,
      "step": 10640
    },
    {
      "entropy": 0.8381056413054466,
      "epoch": 3.2263122017723243,
      "grad_norm": 0.155443012714386,
      "learning_rate": 9.273431709535235e-05,
      "loss": 1.2366,
      "mean_token_accuracy": 0.7605232700705529,
      "num_tokens": 719472329.0,
      "step": 10650
    },
    {
      "entropy": 0.8404705911874771,
      "epoch": 3.2293418162538816,
      "grad_norm": 0.11871460825204849,
      "learning_rate": 9.271697151181231e-05,
      "loss": 1.2388,
      "mean_token_accuracy": 0.7651655808091163,
      "num_tokens": 720153666.0,
      "step": 10660
    },
    {
      "entropy": 0.8290793329477311,
      "epoch": 3.232371430735439,
      "grad_norm": 0.12072983384132385,
      "learning_rate": 9.269960687436606e-05,
      "loss": 1.234,
      "mean_token_accuracy": 0.7592705100774765,
      "num_tokens": 720811003.0,
      "step": 10670
    },
    {
      "entropy": 0.8383875161409378,
      "epoch": 3.2354010452169963,
      "grad_norm": 0.1183423399925232,
      "learning_rate": 9.268222319075906e-05,
      "loss": 1.2426,
      "mean_token_accuracy": 0.761049710214138,
      "num_tokens": 721495710.0,
      "step": 10680
    },
    {
      "entropy": 0.8362769678235054,
      "epoch": 3.2384306596985533,
      "grad_norm": 0.12930332124233246,
      "learning_rate": 9.266482046874536e-05,
      "loss": 1.239,
      "mean_token_accuracy": 0.7631333693861961,
      "num_tokens": 722178590.0,
      "step": 10690
    },
    {
      "entropy": 0.8385419920086861,
      "epoch": 3.2414602741801106,
      "grad_norm": 0.11784379929304123,
      "learning_rate": 9.26473987160874e-05,
      "loss": 1.2395,
      "mean_token_accuracy": 0.7591364905238152,
      "num_tokens": 722847005.0,
      "step": 10700
    },
    {
      "entropy": 0.832073763012886,
      "epoch": 3.244489888661668,
      "grad_norm": 0.1162286102771759,
      "learning_rate": 9.262995794055618e-05,
      "loss": 1.2252,
      "mean_token_accuracy": 0.768974381685257,
      "num_tokens": 723548894.0,
      "step": 10710
    },
    {
      "entropy": 0.8261068508028984,
      "epoch": 3.247519503143225,
      "grad_norm": 0.12522152066230774,
      "learning_rate": 9.261249814993115e-05,
      "loss": 1.2191,
      "mean_token_accuracy": 0.7679961457848549,
      "num_tokens": 724244841.0,
      "step": 10720
    },
    {
      "entropy": 0.8232755184173584,
      "epoch": 3.2505491176247823,
      "grad_norm": 0.1232825443148613,
      "learning_rate": 9.259501935200024e-05,
      "loss": 1.2253,
      "mean_token_accuracy": 0.7634338319301606,
      "num_tokens": 724914821.0,
      "step": 10730
    },
    {
      "entropy": 0.8415536910295487,
      "epoch": 3.2535787321063396,
      "grad_norm": 0.1149967685341835,
      "learning_rate": 9.25775215545599e-05,
      "loss": 1.2431,
      "mean_token_accuracy": 0.761506487429142,
      "num_tokens": 725590181.0,
      "step": 10740
    },
    {
      "entropy": 0.8311128839850426,
      "epoch": 3.2566083465878966,
      "grad_norm": 0.11484793573617935,
      "learning_rate": 9.256000476541496e-05,
      "loss": 1.2373,
      "mean_token_accuracy": 0.7669485747814179,
      "num_tokens": 726266920.0,
      "step": 10750
    },
    {
      "entropy": 0.8289522245526314,
      "epoch": 3.259637961069454,
      "grad_norm": 0.1118326410651207,
      "learning_rate": 9.254246899237885e-05,
      "loss": 1.2284,
      "mean_token_accuracy": 0.7667224913835525,
      "num_tokens": 726950506.0,
      "step": 10760
    },
    {
      "entropy": 0.8402706578373909,
      "epoch": 3.2626675755510113,
      "grad_norm": 0.12299591302871704,
      "learning_rate": 9.252491424327337e-05,
      "loss": 1.2511,
      "mean_token_accuracy": 0.7580578938126564,
      "num_tokens": 727616277.0,
      "step": 10770
    },
    {
      "entropy": 0.8413041412830353,
      "epoch": 3.265697190032568,
      "grad_norm": 0.1593431681394577,
      "learning_rate": 9.250734052592883e-05,
      "loss": 1.2417,
      "mean_token_accuracy": 0.7616387084126472,
      "num_tokens": 728301131.0,
      "step": 10780
    },
    {
      "entropy": 0.833856324851513,
      "epoch": 3.2687268045141256,
      "grad_norm": 0.11823420226573944,
      "learning_rate": 9.248974784818396e-05,
      "loss": 1.2339,
      "mean_token_accuracy": 0.7589184165000915,
      "num_tokens": 728961023.0,
      "step": 10790
    },
    {
      "entropy": 0.8375367939472198,
      "epoch": 3.271756418995683,
      "grad_norm": 0.1314859688282013,
      "learning_rate": 9.2472136217886e-05,
      "loss": 1.2383,
      "mean_token_accuracy": 0.7644320324063301,
      "num_tokens": 729645651.0,
      "step": 10800
    },
    {
      "entropy": 0.8438203558325768,
      "epoch": 3.27478603347724,
      "grad_norm": 0.11647716164588928,
      "learning_rate": 9.245450564289061e-05,
      "loss": 1.2434,
      "mean_token_accuracy": 0.7553828537464142,
      "num_tokens": 730314758.0,
      "step": 10810
    },
    {
      "entropy": 0.8331206291913986,
      "epoch": 3.277815647958797,
      "grad_norm": 0.11722481995820999,
      "learning_rate": 9.243685613106191e-05,
      "loss": 1.2336,
      "mean_token_accuracy": 0.7630361914634705,
      "num_tokens": 730998053.0,
      "step": 10820
    },
    {
      "entropy": 0.8310005441308022,
      "epoch": 3.2808452624403546,
      "grad_norm": 0.1268925815820694,
      "learning_rate": 9.241918769027247e-05,
      "loss": 1.2402,
      "mean_token_accuracy": 0.7594552874565125,
      "num_tokens": 731664781.0,
      "step": 10830
    },
    {
      "entropy": 0.8313428923487663,
      "epoch": 3.283874876921912,
      "grad_norm": 0.12375863641500473,
      "learning_rate": 9.240150032840329e-05,
      "loss": 1.2356,
      "mean_token_accuracy": 0.7621963039040566,
      "num_tokens": 732347797.0,
      "step": 10840
    },
    {
      "entropy": 0.8397682949900627,
      "epoch": 3.286904491403469,
      "grad_norm": 0.1231660544872284,
      "learning_rate": 9.238379405334382e-05,
      "loss": 1.2365,
      "mean_token_accuracy": 0.7627838492393494,
      "num_tokens": 733028509.0,
      "step": 10850
    },
    {
      "entropy": 0.8354284122586251,
      "epoch": 3.289934105885026,
      "grad_norm": 0.1508273482322693,
      "learning_rate": 9.236606887299195e-05,
      "loss": 1.2431,
      "mean_token_accuracy": 0.7645542830228805,
      "num_tokens": 733705677.0,
      "step": 10860
    },
    {
      "entropy": 0.8260786205530166,
      "epoch": 3.2929637203665836,
      "grad_norm": 0.12614572048187256,
      "learning_rate": 9.2348324795254e-05,
      "loss": 1.2337,
      "mean_token_accuracy": 0.7677811950445175,
      "num_tokens": 734388784.0,
      "step": 10870
    },
    {
      "entropy": 0.8300410374999047,
      "epoch": 3.2959933348481405,
      "grad_norm": 0.1307694911956787,
      "learning_rate": 9.233056182804469e-05,
      "loss": 1.2478,
      "mean_token_accuracy": 0.7653857231140136,
      "num_tokens": 735055958.0,
      "step": 10880
    },
    {
      "entropy": 0.8297205582261086,
      "epoch": 3.299022949329698,
      "grad_norm": 0.12933626770973206,
      "learning_rate": 9.231277997928722e-05,
      "loss": 1.2322,
      "mean_token_accuracy": 0.7599256366491318,
      "num_tokens": 735725145.0,
      "step": 10890
    },
    {
      "entropy": 0.8447857826948166,
      "epoch": 3.3020525638112552,
      "grad_norm": 0.11631935834884644,
      "learning_rate": 9.229497925691318e-05,
      "loss": 1.2409,
      "mean_token_accuracy": 0.7583971887826919,
      "num_tokens": 736402252.0,
      "step": 10900
    },
    {
      "entropy": 0.8415904492139816,
      "epoch": 3.305082178292812,
      "grad_norm": 0.12990765273571014,
      "learning_rate": 9.227715966886257e-05,
      "loss": 1.2405,
      "mean_token_accuracy": 0.7592979416251182,
      "num_tokens": 737086144.0,
      "step": 10910
    },
    {
      "entropy": 0.8309503868222237,
      "epoch": 3.3081117927743695,
      "grad_norm": 0.13030767440795898,
      "learning_rate": 9.225932122308384e-05,
      "loss": 1.2364,
      "mean_token_accuracy": 0.7657891899347306,
      "num_tokens": 737767932.0,
      "step": 10920
    },
    {
      "entropy": 0.8368337139487266,
      "epoch": 3.311141407255927,
      "grad_norm": 0.12232133001089096,
      "learning_rate": 9.224146392753381e-05,
      "loss": 1.2416,
      "mean_token_accuracy": 0.7590382680296898,
      "num_tokens": 738429360.0,
      "step": 10930
    },
    {
      "entropy": 0.8323225796222686,
      "epoch": 3.314171021737484,
      "grad_norm": 0.115901418030262,
      "learning_rate": 9.222358779017772e-05,
      "loss": 1.232,
      "mean_token_accuracy": 0.7627633005380631,
      "num_tokens": 739113087.0,
      "step": 10940
    },
    {
      "entropy": 0.8303186953067779,
      "epoch": 3.317200636219041,
      "grad_norm": 0.13112275302410126,
      "learning_rate": 9.220569281898925e-05,
      "loss": 1.2341,
      "mean_token_accuracy": 0.7631768271327019,
      "num_tokens": 739780446.0,
      "step": 10950
    },
    {
      "entropy": 0.8251298353075981,
      "epoch": 3.3202302507005985,
      "grad_norm": 0.14712879061698914,
      "learning_rate": 9.218777902195043e-05,
      "loss": 1.24,
      "mean_token_accuracy": 0.7636583000421524,
      "num_tokens": 740454648.0,
      "step": 10960
    },
    {
      "entropy": 0.8412736356258392,
      "epoch": 3.3232598651821554,
      "grad_norm": 0.11796010285615921,
      "learning_rate": 9.216984640705171e-05,
      "loss": 1.2435,
      "mean_token_accuracy": 0.7623559609055519,
      "num_tokens": 741138483.0,
      "step": 10970
    },
    {
      "entropy": 0.8381548568606376,
      "epoch": 3.326289479663713,
      "grad_norm": 0.12395946681499481,
      "learning_rate": 9.215189498229194e-05,
      "loss": 1.2378,
      "mean_token_accuracy": 0.7647039264440536,
      "num_tokens": 741815394.0,
      "step": 10980
    },
    {
      "entropy": 0.8362372472882271,
      "epoch": 3.32931909414527,
      "grad_norm": 0.1348508596420288,
      "learning_rate": 9.213392475567836e-05,
      "loss": 1.2352,
      "mean_token_accuracy": 0.7623049661517143,
      "num_tokens": 742492113.0,
      "step": 10990
    },
    {
      "entropy": 0.8433797970414162,
      "epoch": 3.332348708626827,
      "grad_norm": 0.12744256854057312,
      "learning_rate": 9.211593573522658e-05,
      "loss": 1.2502,
      "mean_token_accuracy": 0.7595667660236358,
      "num_tokens": 743163984.0,
      "step": 11000
    },
    {
      "entropy": 0.8397553741931916,
      "epoch": 3.3353783231083844,
      "grad_norm": 0.11812698841094971,
      "learning_rate": 9.20979279289606e-05,
      "loss": 1.2466,
      "mean_token_accuracy": 0.7593050569295883,
      "num_tokens": 743828395.0,
      "step": 11010
    },
    {
      "entropy": 0.8322840496897698,
      "epoch": 3.338407937589942,
      "grad_norm": 0.11905736476182938,
      "learning_rate": 9.207990134491281e-05,
      "loss": 1.2265,
      "mean_token_accuracy": 0.7691733837127686,
      "num_tokens": 744527699.0,
      "step": 11020
    },
    {
      "entropy": 0.8369256645441056,
      "epoch": 3.3414375520714987,
      "grad_norm": 0.11343682557344437,
      "learning_rate": 9.206185599112397e-05,
      "loss": 1.2382,
      "mean_token_accuracy": 0.7615094974637031,
      "num_tokens": 745205385.0,
      "step": 11030
    },
    {
      "entropy": 0.8328930526971817,
      "epoch": 3.344467166553056,
      "grad_norm": 0.12398137152194977,
      "learning_rate": 9.204379187564321e-05,
      "loss": 1.2451,
      "mean_token_accuracy": 0.7567608639597893,
      "num_tokens": 745856589.0,
      "step": 11040
    },
    {
      "entropy": 0.8329830259084702,
      "epoch": 3.3474967810346135,
      "grad_norm": 0.12614665925502777,
      "learning_rate": 9.202570900652801e-05,
      "loss": 1.234,
      "mean_token_accuracy": 0.7624053329229354,
      "num_tokens": 746527617.0,
      "step": 11050
    },
    {
      "entropy": 0.8424591690301895,
      "epoch": 3.3505263955161704,
      "grad_norm": 0.12217242270708084,
      "learning_rate": 9.200760739184426e-05,
      "loss": 1.2453,
      "mean_token_accuracy": 0.7612332418560982,
      "num_tokens": 747202230.0,
      "step": 11060
    },
    {
      "entropy": 0.8356316089630127,
      "epoch": 3.3535560099977277,
      "grad_norm": 0.13591527938842773,
      "learning_rate": 9.198948703966618e-05,
      "loss": 1.2388,
      "mean_token_accuracy": 0.7614769533276557,
      "num_tokens": 747886885.0,
      "step": 11070
    },
    {
      "entropy": 0.8369050204753876,
      "epoch": 3.356585624479285,
      "grad_norm": 0.1345531940460205,
      "learning_rate": 9.197134795807634e-05,
      "loss": 1.2425,
      "mean_token_accuracy": 0.7628663569688797,
      "num_tokens": 748562805.0,
      "step": 11080
    },
    {
      "entropy": 0.8288183137774467,
      "epoch": 3.359615238960842,
      "grad_norm": 0.1224965900182724,
      "learning_rate": 9.19531901551657e-05,
      "loss": 1.2309,
      "mean_token_accuracy": 0.7656287491321564,
      "num_tokens": 749244768.0,
      "step": 11090
    },
    {
      "entropy": 0.8335251197218895,
      "epoch": 3.3626448534423994,
      "grad_norm": 0.11962300539016724,
      "learning_rate": 9.193501363903351e-05,
      "loss": 1.24,
      "mean_token_accuracy": 0.7606292203068733,
      "num_tokens": 749918649.0,
      "step": 11100
    },
    {
      "entropy": 0.8271193683147431,
      "epoch": 3.3656744679239567,
      "grad_norm": 0.12110050022602081,
      "learning_rate": 9.191681841778746e-05,
      "loss": 1.2358,
      "mean_token_accuracy": 0.7678375959396362,
      "num_tokens": 750608272.0,
      "step": 11110
    },
    {
      "entropy": 0.8416226848959922,
      "epoch": 3.3687040824055137,
      "grad_norm": 0.13516546785831451,
      "learning_rate": 9.189860449954349e-05,
      "loss": 1.2448,
      "mean_token_accuracy": 0.7645494818687439,
      "num_tokens": 751304553.0,
      "step": 11120
    },
    {
      "entropy": 0.8248679772019386,
      "epoch": 3.371733696887071,
      "grad_norm": 0.12417922168970108,
      "learning_rate": 9.188037189242593e-05,
      "loss": 1.2361,
      "mean_token_accuracy": 0.7616899490356446,
      "num_tokens": 751967652.0,
      "step": 11130
    },
    {
      "entropy": 0.8328699260950089,
      "epoch": 3.3747633113686284,
      "grad_norm": 0.12202231585979462,
      "learning_rate": 9.186212060456743e-05,
      "loss": 1.2296,
      "mean_token_accuracy": 0.76222343146801,
      "num_tokens": 752645899.0,
      "step": 11140
    },
    {
      "entropy": 0.8371228188276291,
      "epoch": 3.3777929258501858,
      "grad_norm": 0.1218470111489296,
      "learning_rate": 9.184385064410898e-05,
      "loss": 1.2519,
      "mean_token_accuracy": 0.7579439014196396,
      "num_tokens": 753300276.0,
      "step": 11150
    },
    {
      "entropy": 0.8341500461101532,
      "epoch": 3.3808225403317427,
      "grad_norm": 0.1183963418006897,
      "learning_rate": 9.182556201919989e-05,
      "loss": 1.2268,
      "mean_token_accuracy": 0.764517468214035,
      "num_tokens": 753982832.0,
      "step": 11160
    },
    {
      "entropy": 0.8343546494841576,
      "epoch": 3.3838521548133,
      "grad_norm": 0.11850113421678543,
      "learning_rate": 9.180725473799782e-05,
      "loss": 1.2323,
      "mean_token_accuracy": 0.7607134401798248,
      "num_tokens": 754655077.0,
      "step": 11170
    },
    {
      "entropy": 0.8380969345569611,
      "epoch": 3.3868817692948574,
      "grad_norm": 0.12320259213447571,
      "learning_rate": 9.178892880866871e-05,
      "loss": 1.2452,
      "mean_token_accuracy": 0.7631853714585304,
      "num_tokens": 755330996.0,
      "step": 11180
    },
    {
      "entropy": 0.8282105028629303,
      "epoch": 3.3899113837764143,
      "grad_norm": 0.12607748806476593,
      "learning_rate": 9.177058423938687e-05,
      "loss": 1.23,
      "mean_token_accuracy": 0.7632102146744728,
      "num_tokens": 755999983.0,
      "step": 11190
    },
    {
      "entropy": 0.8313679173588753,
      "epoch": 3.3929409982579717,
      "grad_norm": 0.12297471612691879,
      "learning_rate": 9.175222103833488e-05,
      "loss": 1.2331,
      "mean_token_accuracy": 0.7653903663158417,
      "num_tokens": 756675895.0,
      "step": 11200
    },
    {
      "entropy": 0.8348143219947814,
      "epoch": 3.395970612739529,
      "grad_norm": 0.12022623419761658,
      "learning_rate": 9.173383921370363e-05,
      "loss": 1.2302,
      "mean_token_accuracy": 0.7649526372551918,
      "num_tokens": 757359109.0,
      "step": 11210
    },
    {
      "entropy": 0.8278435617685318,
      "epoch": 3.399000227221086,
      "grad_norm": 0.1261090785264969,
      "learning_rate": 9.171543877369236e-05,
      "loss": 1.2347,
      "mean_token_accuracy": 0.7597841456532478,
      "num_tokens": 758030443.0,
      "step": 11220
    },
    {
      "entropy": 0.834585751593113,
      "epoch": 3.4020298417026433,
      "grad_norm": 0.12466879189014435,
      "learning_rate": 9.169701972650857e-05,
      "loss": 1.2333,
      "mean_token_accuracy": 0.7644132405519486,
      "num_tokens": 758709443.0,
      "step": 11230
    },
    {
      "entropy": 0.8404792815446853,
      "epoch": 3.4050594561842007,
      "grad_norm": 0.1162450760602951,
      "learning_rate": 9.167858208036808e-05,
      "loss": 1.2398,
      "mean_token_accuracy": 0.7603865653276444,
      "num_tokens": 759385292.0,
      "step": 11240
    },
    {
      "entropy": 0.8283190071582794,
      "epoch": 3.4080890706657576,
      "grad_norm": 0.11930322647094727,
      "learning_rate": 9.166012584349502e-05,
      "loss": 1.2418,
      "mean_token_accuracy": 0.7585573822259903,
      "num_tokens": 760040634.0,
      "step": 11250
    },
    {
      "entropy": 0.8307083159685135,
      "epoch": 3.411118685147315,
      "grad_norm": 0.1354369968175888,
      "learning_rate": 9.164165102412175e-05,
      "loss": 1.2406,
      "mean_token_accuracy": 0.7601440161466598,
      "num_tokens": 760713312.0,
      "step": 11260
    },
    {
      "entropy": 0.8361027002334595,
      "epoch": 3.4141482996288723,
      "grad_norm": 0.1479794681072235,
      "learning_rate": 9.162315763048901e-05,
      "loss": 1.2421,
      "mean_token_accuracy": 0.7593563422560692,
      "num_tokens": 761374551.0,
      "step": 11270
    },
    {
      "entropy": 0.8356462597846985,
      "epoch": 3.4171779141104293,
      "grad_norm": 0.12720368802547455,
      "learning_rate": 9.160464567084577e-05,
      "loss": 1.2375,
      "mean_token_accuracy": 0.7608398467302322,
      "num_tokens": 762054075.0,
      "step": 11280
    },
    {
      "entropy": 0.8298819020390511,
      "epoch": 3.4202075285919866,
      "grad_norm": 0.1229054182767868,
      "learning_rate": 9.158611515344926e-05,
      "loss": 1.2251,
      "mean_token_accuracy": 0.7662895604968071,
      "num_tokens": 762742531.0,
      "step": 11290
    },
    {
      "entropy": 0.8287305623292923,
      "epoch": 3.423237143073544,
      "grad_norm": 0.12051542103290558,
      "learning_rate": 9.156756608656506e-05,
      "loss": 1.2386,
      "mean_token_accuracy": 0.7643614545464515,
      "num_tokens": 763413614.0,
      "step": 11300
    },
    {
      "entropy": 0.8374639958143234,
      "epoch": 3.4262667575551013,
      "grad_norm": 0.127227321267128,
      "learning_rate": 9.154899847846694e-05,
      "loss": 1.2393,
      "mean_token_accuracy": 0.7565767720341683,
      "num_tokens": 764071880.0,
      "step": 11310
    },
    {
      "entropy": 0.827917568385601,
      "epoch": 3.4292963720366583,
      "grad_norm": 0.1351393312215805,
      "learning_rate": 9.153041233743702e-05,
      "loss": 1.235,
      "mean_token_accuracy": 0.7639175683259964,
      "num_tokens": 764743387.0,
      "step": 11320
    },
    {
      "entropy": 0.8264069110155106,
      "epoch": 3.4323259865182156,
      "grad_norm": 0.13566623628139496,
      "learning_rate": 9.151180767176563e-05,
      "loss": 1.2372,
      "mean_token_accuracy": 0.7655463546514512,
      "num_tokens": 765419689.0,
      "step": 11330
    },
    {
      "entropy": 0.8286457657814026,
      "epoch": 3.435355600999773,
      "grad_norm": 0.12399102747440338,
      "learning_rate": 9.149318448975138e-05,
      "loss": 1.2383,
      "mean_token_accuracy": 0.7649675861001015,
      "num_tokens": 766100907.0,
      "step": 11340
    },
    {
      "entropy": 0.8278289854526519,
      "epoch": 3.43838521548133,
      "grad_norm": 0.12072256207466125,
      "learning_rate": 9.147454279970114e-05,
      "loss": 1.2217,
      "mean_token_accuracy": 0.7657185092568397,
      "num_tokens": 766783645.0,
      "step": 11350
    },
    {
      "entropy": 0.8294686049222946,
      "epoch": 3.4414148299628873,
      "grad_norm": 0.1228259727358818,
      "learning_rate": 9.145588260993006e-05,
      "loss": 1.2303,
      "mean_token_accuracy": 0.7650302588939667,
      "num_tokens": 767461938.0,
      "step": 11360
    },
    {
      "entropy": 0.8382543668150901,
      "epoch": 3.4444444444444446,
      "grad_norm": 0.1262565404176712,
      "learning_rate": 9.143720392876148e-05,
      "loss": 1.2309,
      "mean_token_accuracy": 0.7626958236098289,
      "num_tokens": 768143985.0,
      "step": 11370
    },
    {
      "entropy": 0.8315334230661392,
      "epoch": 3.4474740589260016,
      "grad_norm": 0.12471026182174683,
      "learning_rate": 9.141850676452704e-05,
      "loss": 1.2347,
      "mean_token_accuracy": 0.7572854548692703,
      "num_tokens": 768805960.0,
      "step": 11380
    },
    {
      "entropy": 0.8320801615715027,
      "epoch": 3.450503673407559,
      "grad_norm": 0.11635059863328934,
      "learning_rate": 9.139979112556662e-05,
      "loss": 1.2444,
      "mean_token_accuracy": 0.7593884602189064,
      "num_tokens": 769470007.0,
      "step": 11390
    },
    {
      "entropy": 0.8333582326769828,
      "epoch": 3.4535332878891163,
      "grad_norm": 0.11514585465192795,
      "learning_rate": 9.138105702022834e-05,
      "loss": 1.2366,
      "mean_token_accuracy": 0.7615718305110931,
      "num_tokens": 770134661.0,
      "step": 11400
    },
    {
      "entropy": 0.8287696212530136,
      "epoch": 3.456562902370673,
      "grad_norm": 0.11926062405109406,
      "learning_rate": 9.136230445686853e-05,
      "loss": 1.2302,
      "mean_token_accuracy": 0.7648595631122589,
      "num_tokens": 770811961.0,
      "step": 11410
    },
    {
      "entropy": 0.8429672375321389,
      "epoch": 3.4595925168522306,
      "grad_norm": 0.12452052533626556,
      "learning_rate": 9.134353344385176e-05,
      "loss": 1.2458,
      "mean_token_accuracy": 0.7575023502111435,
      "num_tokens": 771468667.0,
      "step": 11420
    },
    {
      "entropy": 0.844814071059227,
      "epoch": 3.462622131333788,
      "grad_norm": 0.13977554440498352,
      "learning_rate": 9.132474398955087e-05,
      "loss": 1.2402,
      "mean_token_accuracy": 0.7594343468546867,
      "num_tokens": 772150584.0,
      "step": 11430
    },
    {
      "entropy": 0.8297743394970893,
      "epoch": 3.465651745815345,
      "grad_norm": 0.12803715467453003,
      "learning_rate": 9.130593610234689e-05,
      "loss": 1.2274,
      "mean_token_accuracy": 0.7670684143900871,
      "num_tokens": 772834313.0,
      "step": 11440
    },
    {
      "entropy": 0.8327987894415856,
      "epoch": 3.468681360296902,
      "grad_norm": 0.1271650195121765,
      "learning_rate": 9.128710979062907e-05,
      "loss": 1.2362,
      "mean_token_accuracy": 0.7619610667228699,
      "num_tokens": 773511436.0,
      "step": 11450
    },
    {
      "entropy": 0.8360430836677551,
      "epoch": 3.4717109747784596,
      "grad_norm": 0.12318103760480881,
      "learning_rate": 9.126826506279487e-05,
      "loss": 1.2331,
      "mean_token_accuracy": 0.7648727923631669,
      "num_tokens": 774192858.0,
      "step": 11460
    },
    {
      "entropy": 0.8295892536640167,
      "epoch": 3.4747405892600165,
      "grad_norm": 0.11842337995767593,
      "learning_rate": 9.124940192725002e-05,
      "loss": 1.2365,
      "mean_token_accuracy": 0.7600327342748642,
      "num_tokens": 774868822.0,
      "step": 11470
    },
    {
      "entropy": 0.8435277476906776,
      "epoch": 3.477770203741574,
      "grad_norm": 0.1237192377448082,
      "learning_rate": 9.12305203924084e-05,
      "loss": 1.2435,
      "mean_token_accuracy": 0.7555145010352134,
      "num_tokens": 775527906.0,
      "step": 11480
    },
    {
      "entropy": 0.8364113673567772,
      "epoch": 3.4807998182231312,
      "grad_norm": 0.1254260092973709,
      "learning_rate": 9.121162046669212e-05,
      "loss": 1.2405,
      "mean_token_accuracy": 0.7616367593407631,
      "num_tokens": 776210142.0,
      "step": 11490
    },
    {
      "entropy": 0.8302595242857933,
      "epoch": 3.483829432704688,
      "grad_norm": 0.122586190700531,
      "learning_rate": 9.119270215853149e-05,
      "loss": 1.2376,
      "mean_token_accuracy": 0.7600247159600257,
      "num_tokens": 776873062.0,
      "step": 11500
    },
    {
      "entropy": 0.839820210635662,
      "epoch": 3.4868590471862455,
      "grad_norm": 0.13961449265480042,
      "learning_rate": 9.117376547636502e-05,
      "loss": 1.2391,
      "mean_token_accuracy": 0.7607545420527458,
      "num_tokens": 777541322.0,
      "step": 11510
    },
    {
      "entropy": 0.8306924447417259,
      "epoch": 3.489888661667803,
      "grad_norm": 0.1306026726961136,
      "learning_rate": 9.115481042863943e-05,
      "loss": 1.2321,
      "mean_token_accuracy": 0.7651106625795364,
      "num_tokens": 778221429.0,
      "step": 11520
    },
    {
      "entropy": 0.8247424677014351,
      "epoch": 3.49291827614936,
      "grad_norm": 0.1263517290353775,
      "learning_rate": 9.113583702380961e-05,
      "loss": 1.2284,
      "mean_token_accuracy": 0.7655428737401963,
      "num_tokens": 778898343.0,
      "step": 11530
    },
    {
      "entropy": 0.8408169224858284,
      "epoch": 3.495947890630917,
      "grad_norm": 0.1276242733001709,
      "learning_rate": 9.111684527033865e-05,
      "loss": 1.2423,
      "mean_token_accuracy": 0.7609571158885956,
      "num_tokens": 779576822.0,
      "step": 11540
    },
    {
      "entropy": 0.8287448465824128,
      "epoch": 3.4989775051124745,
      "grad_norm": 0.1321870982646942,
      "learning_rate": 9.109783517669783e-05,
      "loss": 1.2261,
      "mean_token_accuracy": 0.7656113281846046,
      "num_tokens": 780263497.0,
      "step": 11550
    },
    {
      "entropy": 0.828518845140934,
      "epoch": 3.5020071195940314,
      "grad_norm": 0.12340894341468811,
      "learning_rate": 9.107880675136658e-05,
      "loss": 1.2318,
      "mean_token_accuracy": 0.7637648627161979,
      "num_tokens": 780945082.0,
      "step": 11560
    },
    {
      "entropy": 0.8328303188085556,
      "epoch": 3.505036734075589,
      "grad_norm": 0.11579950153827667,
      "learning_rate": 9.105976000283256e-05,
      "loss": 1.2408,
      "mean_token_accuracy": 0.7612169280648231,
      "num_tokens": 781613662.0,
      "step": 11570
    },
    {
      "entropy": 0.8312623202800751,
      "epoch": 3.508066348557146,
      "grad_norm": 0.1133250892162323,
      "learning_rate": 9.104069493959156e-05,
      "loss": 1.2367,
      "mean_token_accuracy": 0.7617637991905213,
      "num_tokens": 782288520.0,
      "step": 11580
    },
    {
      "entropy": 0.8360737174749374,
      "epoch": 3.511095963038703,
      "grad_norm": 0.1433398276567459,
      "learning_rate": 9.102161157014753e-05,
      "loss": 1.2361,
      "mean_token_accuracy": 0.7610009625554085,
      "num_tokens": 782964782.0,
      "step": 11590
    },
    {
      "entropy": 0.8245732337236404,
      "epoch": 3.5141255775202604,
      "grad_norm": 0.12671203911304474,
      "learning_rate": 9.100250990301264e-05,
      "loss": 1.2356,
      "mean_token_accuracy": 0.7634012997150421,
      "num_tokens": 783630482.0,
      "step": 11600
    },
    {
      "entropy": 0.8308870613574981,
      "epoch": 3.517155192001818,
      "grad_norm": 0.12995724380016327,
      "learning_rate": 9.098338994670719e-05,
      "loss": 1.2299,
      "mean_token_accuracy": 0.7653521984815598,
      "num_tokens": 784312380.0,
      "step": 11610
    },
    {
      "entropy": 0.8430249497294426,
      "epoch": 3.5201848064833747,
      "grad_norm": 0.1239878460764885,
      "learning_rate": 9.09642517097596e-05,
      "loss": 1.2396,
      "mean_token_accuracy": 0.7610897243022918,
      "num_tokens": 784995872.0,
      "step": 11620
    },
    {
      "entropy": 0.8321187064051628,
      "epoch": 3.523214420964932,
      "grad_norm": 0.1331980973482132,
      "learning_rate": 9.094509520070651e-05,
      "loss": 1.2354,
      "mean_token_accuracy": 0.763390064239502,
      "num_tokens": 785668728.0,
      "step": 11630
    },
    {
      "entropy": 0.8272807270288467,
      "epoch": 3.5262440354464895,
      "grad_norm": 0.12257704138755798,
      "learning_rate": 9.092592042809267e-05,
      "loss": 1.2327,
      "mean_token_accuracy": 0.7632969573140145,
      "num_tokens": 786344132.0,
      "step": 11640
    },
    {
      "entropy": 0.8298324644565582,
      "epoch": 3.529273649928047,
      "grad_norm": 0.14654377102851868,
      "learning_rate": 9.090672740047102e-05,
      "loss": 1.2352,
      "mean_token_accuracy": 0.7644375175237655,
      "num_tokens": 787024814.0,
      "step": 11650
    },
    {
      "entropy": 0.8290773898363113,
      "epoch": 3.5323032644096037,
      "grad_norm": 0.1257285624742508,
      "learning_rate": 9.088751612640255e-05,
      "loss": 1.2287,
      "mean_token_accuracy": 0.7638806805014611,
      "num_tokens": 787705829.0,
      "step": 11660
    },
    {
      "entropy": 0.8311083659529686,
      "epoch": 3.535332878891161,
      "grad_norm": 0.1146702840924263,
      "learning_rate": 9.08682866144565e-05,
      "loss": 1.2356,
      "mean_token_accuracy": 0.759133493900299,
      "num_tokens": 788369869.0,
      "step": 11670
    },
    {
      "entropy": 0.834751708805561,
      "epoch": 3.5383624933727185,
      "grad_norm": 0.12474443763494492,
      "learning_rate": 9.084903887321016e-05,
      "loss": 1.2401,
      "mean_token_accuracy": 0.7602774262428283,
      "num_tokens": 789042854.0,
      "step": 11680
    },
    {
      "entropy": 0.8291317448019981,
      "epoch": 3.5413921078542754,
      "grad_norm": 0.12132880091667175,
      "learning_rate": 9.082977291124903e-05,
      "loss": 1.2371,
      "mean_token_accuracy": 0.7625534549355507,
      "num_tokens": 789707404.0,
      "step": 11690
    },
    {
      "entropy": 0.8243643850088119,
      "epoch": 3.5444217223358327,
      "grad_norm": 0.13487163186073303,
      "learning_rate": 9.081048873716666e-05,
      "loss": 1.2281,
      "mean_token_accuracy": 0.763888205587864,
      "num_tokens": 790373929.0,
      "step": 11700
    },
    {
      "entropy": 0.8182727739214897,
      "epoch": 3.54745133681739,
      "grad_norm": 0.11959255486726761,
      "learning_rate": 9.079118635956473e-05,
      "loss": 1.2266,
      "mean_token_accuracy": 0.7675241082906723,
      "num_tokens": 791052543.0,
      "step": 11710
    },
    {
      "entropy": 0.8281921803951263,
      "epoch": 3.5504809512989475,
      "grad_norm": 0.12217836081981659,
      "learning_rate": 9.077186578705314e-05,
      "loss": 1.2443,
      "mean_token_accuracy": 0.7592179551720619,
      "num_tokens": 791705086.0,
      "step": 11720
    },
    {
      "entropy": 0.838127113878727,
      "epoch": 3.5535105657805044,
      "grad_norm": 0.1235826388001442,
      "learning_rate": 9.075252702824976e-05,
      "loss": 1.2424,
      "mean_token_accuracy": 0.7577555432915688,
      "num_tokens": 792372054.0,
      "step": 11730
    },
    {
      "entropy": 0.8330118611454964,
      "epoch": 3.5565401802620618,
      "grad_norm": 0.11890354752540588,
      "learning_rate": 9.073317009178068e-05,
      "loss": 1.237,
      "mean_token_accuracy": 0.7636782974004745,
      "num_tokens": 793049582.0,
      "step": 11740
    },
    {
      "entropy": 0.8319677591323853,
      "epoch": 3.559569794743619,
      "grad_norm": 0.12073475867509842,
      "learning_rate": 9.071379498628008e-05,
      "loss": 1.2451,
      "mean_token_accuracy": 0.7606442883610726,
      "num_tokens": 793718001.0,
      "step": 11750
    },
    {
      "entropy": 0.8443213135004044,
      "epoch": 3.562599409225176,
      "grad_norm": 0.11481815576553345,
      "learning_rate": 9.069440172039017e-05,
      "loss": 1.2456,
      "mean_token_accuracy": 0.7587682247161865,
      "num_tokens": 794388471.0,
      "step": 11760
    },
    {
      "entropy": 0.82033039778471,
      "epoch": 3.5656290237067334,
      "grad_norm": 0.14372728765010834,
      "learning_rate": 9.067499030276136e-05,
      "loss": 1.2299,
      "mean_token_accuracy": 0.7666828289628029,
      "num_tokens": 795067679.0,
      "step": 11770
    },
    {
      "entropy": 0.8215655073523521,
      "epoch": 3.5686586381882908,
      "grad_norm": 0.12320514023303986,
      "learning_rate": 9.065556074205212e-05,
      "loss": 1.2276,
      "mean_token_accuracy": 0.7630557641386986,
      "num_tokens": 795746543.0,
      "step": 11780
    },
    {
      "entropy": 0.8349901750683785,
      "epoch": 3.5716882526698477,
      "grad_norm": 0.13253360986709595,
      "learning_rate": 9.063611304692896e-05,
      "loss": 1.2411,
      "mean_token_accuracy": 0.7601036667823792,
      "num_tokens": 796412165.0,
      "step": 11790
    },
    {
      "entropy": 0.8316451072692871,
      "epoch": 3.574717867151405,
      "grad_norm": 0.11468390375375748,
      "learning_rate": 9.061664722606657e-05,
      "loss": 1.2405,
      "mean_token_accuracy": 0.7620562642812729,
      "num_tokens": 797077751.0,
      "step": 11800
    },
    {
      "entropy": 0.824966624379158,
      "epoch": 3.5777474816329624,
      "grad_norm": 0.12100493907928467,
      "learning_rate": 9.059716328814765e-05,
      "loss": 1.2261,
      "mean_token_accuracy": 0.7641262948513031,
      "num_tokens": 797751465.0,
      "step": 11810
    },
    {
      "entropy": 0.8425392359495163,
      "epoch": 3.5807770961145193,
      "grad_norm": 0.12134084850549698,
      "learning_rate": 9.057766124186303e-05,
      "loss": 1.2399,
      "mean_token_accuracy": 0.7614474728703499,
      "num_tokens": 798435722.0,
      "step": 11820
    },
    {
      "entropy": 0.8275352597236634,
      "epoch": 3.5838067105960767,
      "grad_norm": 0.11163262277841568,
      "learning_rate": 9.055814109591159e-05,
      "loss": 1.2239,
      "mean_token_accuracy": 0.7658421337604523,
      "num_tokens": 799115112.0,
      "step": 11830
    },
    {
      "entropy": 0.8259432733058929,
      "epoch": 3.586836325077634,
      "grad_norm": 0.1207873523235321,
      "learning_rate": 9.05386028590003e-05,
      "loss": 1.2251,
      "mean_token_accuracy": 0.7633491337299347,
      "num_tokens": 799790970.0,
      "step": 11840
    },
    {
      "entropy": 0.8228763297200203,
      "epoch": 3.589865939559191,
      "grad_norm": 0.11874350905418396,
      "learning_rate": 9.051904653984418e-05,
      "loss": 1.2213,
      "mean_token_accuracy": 0.7662205755710602,
      "num_tokens": 800475946.0,
      "step": 11850
    },
    {
      "entropy": 0.8243829488754273,
      "epoch": 3.5928955540407483,
      "grad_norm": 0.12391141802072525,
      "learning_rate": 9.049947214716634e-05,
      "loss": 1.2273,
      "mean_token_accuracy": 0.7658755227923393,
      "num_tokens": 801149538.0,
      "step": 11860
    },
    {
      "entropy": 0.8380412116646767,
      "epoch": 3.5959251685223057,
      "grad_norm": 0.14538021385669708,
      "learning_rate": 9.047987968969792e-05,
      "loss": 1.2373,
      "mean_token_accuracy": 0.7669436022639274,
      "num_tokens": 801842586.0,
      "step": 11870
    },
    {
      "entropy": 0.8286140769720077,
      "epoch": 3.5989547830038626,
      "grad_norm": 0.1327095776796341,
      "learning_rate": 9.046026917617813e-05,
      "loss": 1.2296,
      "mean_token_accuracy": 0.7633321568369865,
      "num_tokens": 802512963.0,
      "step": 11880
    },
    {
      "entropy": 0.8307457745075226,
      "epoch": 3.60198439748542,
      "grad_norm": 0.12284592539072037,
      "learning_rate": 9.044064061535428e-05,
      "loss": 1.2318,
      "mean_token_accuracy": 0.7620765492320061,
      "num_tokens": 803191402.0,
      "step": 11890
    },
    {
      "entropy": 0.8326386421918869,
      "epoch": 3.6050140119669773,
      "grad_norm": 0.1168675646185875,
      "learning_rate": 9.042099401598166e-05,
      "loss": 1.2353,
      "mean_token_accuracy": 0.7611799791455269,
      "num_tokens": 803859566.0,
      "step": 11900
    },
    {
      "entropy": 0.8209079593420029,
      "epoch": 3.6080436264485343,
      "grad_norm": 0.12039514631032944,
      "learning_rate": 9.040132938682365e-05,
      "loss": 1.2339,
      "mean_token_accuracy": 0.7661359086632729,
      "num_tokens": 804531546.0,
      "step": 11910
    },
    {
      "entropy": 0.8414869412779808,
      "epoch": 3.6110732409300916,
      "grad_norm": 0.12295954674482346,
      "learning_rate": 9.038164673665162e-05,
      "loss": 1.2396,
      "mean_token_accuracy": 0.7620622083544731,
      "num_tokens": 805216610.0,
      "step": 11920
    },
    {
      "entropy": 0.8272681817412376,
      "epoch": 3.614102855411649,
      "grad_norm": 0.12621666491031647,
      "learning_rate": 9.036194607424508e-05,
      "loss": 1.2334,
      "mean_token_accuracy": 0.7656963005661964,
      "num_tokens": 805900069.0,
      "step": 11930
    },
    {
      "entropy": 0.8419021785259246,
      "epoch": 3.617132469893206,
      "grad_norm": 0.1228092685341835,
      "learning_rate": 9.034222740839147e-05,
      "loss": 1.2459,
      "mean_token_accuracy": 0.760851913690567,
      "num_tokens": 806568925.0,
      "step": 11940
    },
    {
      "entropy": 0.8313126280903816,
      "epoch": 3.6201620843747633,
      "grad_norm": 0.13267569243907928,
      "learning_rate": 9.03224907478863e-05,
      "loss": 1.2332,
      "mean_token_accuracy": 0.7593785226345062,
      "num_tokens": 807240671.0,
      "step": 11950
    },
    {
      "entropy": 0.8313692197203636,
      "epoch": 3.6231916988563206,
      "grad_norm": 0.12366776168346405,
      "learning_rate": 9.030273610153312e-05,
      "loss": 1.2361,
      "mean_token_accuracy": 0.7617954403162003,
      "num_tokens": 807917683.0,
      "step": 11960
    },
    {
      "entropy": 0.8314715504646302,
      "epoch": 3.6262213133378776,
      "grad_norm": 0.11919554322957993,
      "learning_rate": 9.028296347814349e-05,
      "loss": 1.2322,
      "mean_token_accuracy": 0.7616859465837479,
      "num_tokens": 808602683.0,
      "step": 11970
    },
    {
      "entropy": 0.8367446169257164,
      "epoch": 3.629250927819435,
      "grad_norm": 0.13895639777183533,
      "learning_rate": 9.026317288653698e-05,
      "loss": 1.2404,
      "mean_token_accuracy": 0.7621679157018661,
      "num_tokens": 809277786.0,
      "step": 11980
    },
    {
      "entropy": 0.8247834354639053,
      "epoch": 3.6322805423009923,
      "grad_norm": 0.12393562495708466,
      "learning_rate": 9.02433643355412e-05,
      "loss": 1.2367,
      "mean_token_accuracy": 0.7608552008867264,
      "num_tokens": 809943651.0,
      "step": 11990
    },
    {
      "entropy": 0.8331888973712921,
      "epoch": 3.635310156782549,
      "grad_norm": 0.12172278016805649,
      "learning_rate": 9.022353783399175e-05,
      "loss": 1.2348,
      "mean_token_accuracy": 0.7604128733277321,
      "num_tokens": 810615890.0,
      "step": 12000
    },
    {
      "entropy": 0.8270653799176216,
      "epoch": 3.6383397712641066,
      "grad_norm": 0.1276981234550476,
      "learning_rate": 9.020369339073222e-05,
      "loss": 1.232,
      "mean_token_accuracy": 0.7668903693556786,
      "num_tokens": 811299502.0,
      "step": 12010
    },
    {
      "entropy": 0.8299973547458649,
      "epoch": 3.641369385745664,
      "grad_norm": 0.12148939073085785,
      "learning_rate": 9.018383101461425e-05,
      "loss": 1.2367,
      "mean_token_accuracy": 0.7597500324249268,
      "num_tokens": 811963216.0,
      "step": 12020
    },
    {
      "entropy": 0.8284254223108292,
      "epoch": 3.644399000227221,
      "grad_norm": 0.1537405252456665,
      "learning_rate": 9.016395071449745e-05,
      "loss": 1.221,
      "mean_token_accuracy": 0.771775534749031,
      "num_tokens": 812666893.0,
      "step": 12030
    },
    {
      "entropy": 0.8277289435267449,
      "epoch": 3.647428614708778,
      "grad_norm": 0.1311744749546051,
      "learning_rate": 9.014405249924942e-05,
      "loss": 1.242,
      "mean_token_accuracy": 0.7604265868663788,
      "num_tokens": 813324365.0,
      "step": 12040
    },
    {
      "entropy": 0.8355601668357849,
      "epoch": 3.6504582291903356,
      "grad_norm": 0.12726181745529175,
      "learning_rate": 9.012413637774576e-05,
      "loss": 1.2402,
      "mean_token_accuracy": 0.7586885496973992,
      "num_tokens": 813995063.0,
      "step": 12050
    },
    {
      "entropy": 0.8477053657174111,
      "epoch": 3.6534878436718925,
      "grad_norm": 0.12102725356817245,
      "learning_rate": 9.010420235887007e-05,
      "loss": 1.2542,
      "mean_token_accuracy": 0.7584789529442787,
      "num_tokens": 814657398.0,
      "step": 12060
    },
    {
      "entropy": 0.8242482304573059,
      "epoch": 3.65651745815345,
      "grad_norm": 0.1336752474308014,
      "learning_rate": 9.00842504515139e-05,
      "loss": 1.2279,
      "mean_token_accuracy": 0.7599899381399154,
      "num_tokens": 815318538.0,
      "step": 12070
    },
    {
      "entropy": 0.8387343779206275,
      "epoch": 3.6595470726350072,
      "grad_norm": 0.13991276919841766,
      "learning_rate": 9.00642806645768e-05,
      "loss": 1.2416,
      "mean_token_accuracy": 0.7603589996695519,
      "num_tokens": 815999407.0,
      "step": 12080
    },
    {
      "entropy": 0.8195283800363541,
      "epoch": 3.662576687116564,
      "grad_norm": 0.13147521018981934,
      "learning_rate": 9.004429300696631e-05,
      "loss": 1.2229,
      "mean_token_accuracy": 0.7663108944892884,
      "num_tokens": 816682090.0,
      "step": 12090
    },
    {
      "entropy": 0.8434464514255524,
      "epoch": 3.6656063015981215,
      "grad_norm": 0.11861646920442581,
      "learning_rate": 9.002428748759791e-05,
      "loss": 1.2406,
      "mean_token_accuracy": 0.7598227173089981,
      "num_tokens": 817354308.0,
      "step": 12100
    },
    {
      "entropy": 0.8313905954360962,
      "epoch": 3.668635916079679,
      "grad_norm": 0.12398044764995575,
      "learning_rate": 9.000426411539507e-05,
      "loss": 1.2373,
      "mean_token_accuracy": 0.7607077971100807,
      "num_tokens": 818022898.0,
      "step": 12110
    },
    {
      "entropy": 0.8373366162180901,
      "epoch": 3.6716655305612362,
      "grad_norm": 0.11641517281532288,
      "learning_rate": 8.998422289928923e-05,
      "loss": 1.2387,
      "mean_token_accuracy": 0.7564888164401055,
      "num_tokens": 818682512.0,
      "step": 12120
    },
    {
      "entropy": 0.8366742789745331,
      "epoch": 3.674695145042793,
      "grad_norm": 0.12284574657678604,
      "learning_rate": 8.996416384821974e-05,
      "loss": 1.2304,
      "mean_token_accuracy": 0.7645779192447663,
      "num_tokens": 819367189.0,
      "step": 12130
    },
    {
      "entropy": 0.8274043366312981,
      "epoch": 3.6777247595243505,
      "grad_norm": 0.11731072515249252,
      "learning_rate": 8.994408697113397e-05,
      "loss": 1.2267,
      "mean_token_accuracy": 0.7654698178172111,
      "num_tokens": 820052012.0,
      "step": 12140
    },
    {
      "entropy": 0.8292063072323799,
      "epoch": 3.680754374005908,
      "grad_norm": 0.12219222635030746,
      "learning_rate": 8.992399227698721e-05,
      "loss": 1.2339,
      "mean_token_accuracy": 0.7619503021240235,
      "num_tokens": 820720978.0,
      "step": 12150
    },
    {
      "entropy": 0.8339642301201821,
      "epoch": 3.6837839884874652,
      "grad_norm": 0.12715142965316772,
      "learning_rate": 8.990387977474266e-05,
      "loss": 1.2385,
      "mean_token_accuracy": 0.7595003217458725,
      "num_tokens": 821383669.0,
      "step": 12160
    },
    {
      "entropy": 0.8279847532510758,
      "epoch": 3.686813602969022,
      "grad_norm": 0.13311992585659027,
      "learning_rate": 8.988374947337155e-05,
      "loss": 1.2384,
      "mean_token_accuracy": 0.7625073760747909,
      "num_tokens": 822064780.0,
      "step": 12170
    },
    {
      "entropy": 0.8269133359193802,
      "epoch": 3.6898432174505795,
      "grad_norm": 0.12409292161464691,
      "learning_rate": 8.9863601381853e-05,
      "loss": 1.2282,
      "mean_token_accuracy": 0.7661284163594246,
      "num_tokens": 822751650.0,
      "step": 12180
    },
    {
      "entropy": 0.8387887254357338,
      "epoch": 3.692872831932137,
      "grad_norm": 0.12265481799840927,
      "learning_rate": 8.984343550917403e-05,
      "loss": 1.2363,
      "mean_token_accuracy": 0.7628649190068245,
      "num_tokens": 823433050.0,
      "step": 12190
    },
    {
      "entropy": 0.8265530884265899,
      "epoch": 3.695902446413694,
      "grad_norm": 0.12023020535707474,
      "learning_rate": 8.982325186432966e-05,
      "loss": 1.2224,
      "mean_token_accuracy": 0.7683528974652291,
      "num_tokens": 824119564.0,
      "step": 12200
    },
    {
      "entropy": 0.8196702033281327,
      "epoch": 3.698932060895251,
      "grad_norm": 0.11864407360553741,
      "learning_rate": 8.980305045632278e-05,
      "loss": 1.2316,
      "mean_token_accuracy": 0.7656280145049095,
      "num_tokens": 824794803.0,
      "step": 12210
    },
    {
      "entropy": 0.8307268977165222,
      "epoch": 3.7019616753768085,
      "grad_norm": 0.11842304468154907,
      "learning_rate": 8.978283129416424e-05,
      "loss": 1.2327,
      "mean_token_accuracy": 0.7590238392353058,
      "num_tokens": 825466266.0,
      "step": 12220
    },
    {
      "entropy": 0.8390016943216324,
      "epoch": 3.7049912898583655,
      "grad_norm": 0.12841185927391052,
      "learning_rate": 8.976259438687282e-05,
      "loss": 1.229,
      "mean_token_accuracy": 0.7644271805882454,
      "num_tokens": 826153789.0,
      "step": 12230
    },
    {
      "entropy": 0.8306706994771957,
      "epoch": 3.708020904339923,
      "grad_norm": 0.14335404336452484,
      "learning_rate": 8.974233974347517e-05,
      "loss": 1.2374,
      "mean_token_accuracy": 0.7622088611125946,
      "num_tokens": 826818162.0,
      "step": 12240
    },
    {
      "entropy": 0.826574458181858,
      "epoch": 3.71105051882148,
      "grad_norm": 0.1161520928144455,
      "learning_rate": 8.972206737300586e-05,
      "loss": 1.2267,
      "mean_token_accuracy": 0.7656951203942299,
      "num_tokens": 827503722.0,
      "step": 12250
    },
    {
      "entropy": 0.8206147626042366,
      "epoch": 3.714080133303037,
      "grad_norm": 0.13360336422920227,
      "learning_rate": 8.970177728450742e-05,
      "loss": 1.2275,
      "mean_token_accuracy": 0.7604197725653649,
      "num_tokens": 828166082.0,
      "step": 12260
    },
    {
      "entropy": 0.8373993262648582,
      "epoch": 3.7171097477845945,
      "grad_norm": 0.13123397529125214,
      "learning_rate": 8.96814694870302e-05,
      "loss": 1.2429,
      "mean_token_accuracy": 0.7625438809394837,
      "num_tokens": 828844434.0,
      "step": 12270
    },
    {
      "entropy": 0.8399862021207809,
      "epoch": 3.720139362266152,
      "grad_norm": 0.1167200431227684,
      "learning_rate": 8.966114398963252e-05,
      "loss": 1.2449,
      "mean_token_accuracy": 0.7576163247227669,
      "num_tokens": 829509473.0,
      "step": 12280
    },
    {
      "entropy": 0.8254438489675522,
      "epoch": 3.7231689767477087,
      "grad_norm": 0.16813643276691437,
      "learning_rate": 8.96408008013806e-05,
      "loss": 1.226,
      "mean_token_accuracy": 0.7664303198456764,
      "num_tokens": 830186083.0,
      "step": 12290
    },
    {
      "entropy": 0.8378492087125778,
      "epoch": 3.726198591229266,
      "grad_norm": 0.12639133632183075,
      "learning_rate": 8.962043993134846e-05,
      "loss": 1.2413,
      "mean_token_accuracy": 0.7567730829119682,
      "num_tokens": 830853542.0,
      "step": 12300
    },
    {
      "entropy": 0.8167424246668815,
      "epoch": 3.7292282057108235,
      "grad_norm": 0.11713540554046631,
      "learning_rate": 8.960006138861808e-05,
      "loss": 1.2216,
      "mean_token_accuracy": 0.7664415761828423,
      "num_tokens": 831533067.0,
      "step": 12310
    },
    {
      "entropy": 0.8392841696739197,
      "epoch": 3.7322578201923804,
      "grad_norm": 0.12706278264522552,
      "learning_rate": 8.957966518227934e-05,
      "loss": 1.244,
      "mean_token_accuracy": 0.759884363412857,
      "num_tokens": 832205180.0,
      "step": 12320
    },
    {
      "entropy": 0.8391022115945816,
      "epoch": 3.7352874346739378,
      "grad_norm": 0.12109837681055069,
      "learning_rate": 8.955925132142992e-05,
      "loss": 1.2275,
      "mean_token_accuracy": 0.7620567560195923,
      "num_tokens": 832885884.0,
      "step": 12330
    },
    {
      "entropy": 0.8387953415513039,
      "epoch": 3.738317049155495,
      "grad_norm": 0.12044090032577515,
      "learning_rate": 8.953881981517549e-05,
      "loss": 1.2328,
      "mean_token_accuracy": 0.765028627216816,
      "num_tokens": 833571123.0,
      "step": 12340
    },
    {
      "entropy": 0.8334729716181755,
      "epoch": 3.741346663637052,
      "grad_norm": 0.126737579703331,
      "learning_rate": 8.951837067262946e-05,
      "loss": 1.2366,
      "mean_token_accuracy": 0.7619820713996888,
      "num_tokens": 834242306.0,
      "step": 12350
    },
    {
      "entropy": 0.8303900763392449,
      "epoch": 3.7443762781186094,
      "grad_norm": 0.12160580605268478,
      "learning_rate": 8.949790390291321e-05,
      "loss": 1.2314,
      "mean_token_accuracy": 0.7624058470129966,
      "num_tokens": 834922541.0,
      "step": 12360
    },
    {
      "entropy": 0.8399101763963699,
      "epoch": 3.7474058926001668,
      "grad_norm": 0.12068485468626022,
      "learning_rate": 8.947741951515594e-05,
      "loss": 1.2405,
      "mean_token_accuracy": 0.7593312174081802,
      "num_tokens": 835591469.0,
      "step": 12370
    },
    {
      "entropy": 0.8358460009098053,
      "epoch": 3.7504355070817237,
      "grad_norm": 0.12638629972934723,
      "learning_rate": 8.945691751849469e-05,
      "loss": 1.2305,
      "mean_token_accuracy": 0.7640237674117089,
      "num_tokens": 836271264.0,
      "step": 12380
    },
    {
      "entropy": 0.835583382844925,
      "epoch": 3.753465121563281,
      "grad_norm": 0.1262970119714737,
      "learning_rate": 8.943639792207438e-05,
      "loss": 1.2406,
      "mean_token_accuracy": 0.7630352824926376,
      "num_tokens": 836951359.0,
      "step": 12390
    },
    {
      "entropy": 0.8302667006850243,
      "epoch": 3.7564947360448384,
      "grad_norm": 0.13091841340065002,
      "learning_rate": 8.94158607350478e-05,
      "loss": 1.2264,
      "mean_token_accuracy": 0.7653634011745453,
      "num_tokens": 837639885.0,
      "step": 12400
    },
    {
      "entropy": 0.8293650209903717,
      "epoch": 3.7595243505263953,
      "grad_norm": 0.11191177368164062,
      "learning_rate": 8.939530596657553e-05,
      "loss": 1.2303,
      "mean_token_accuracy": 0.7612865820527077,
      "num_tokens": 838305937.0,
      "step": 12410
    },
    {
      "entropy": 0.8327704191207885,
      "epoch": 3.7625539650079527,
      "grad_norm": 0.11614000797271729,
      "learning_rate": 8.937473362582605e-05,
      "loss": 1.2357,
      "mean_token_accuracy": 0.7675615847110748,
      "num_tokens": 839009703.0,
      "step": 12420
    },
    {
      "entropy": 0.825287601351738,
      "epoch": 3.76558357948951,
      "grad_norm": 0.12107538431882858,
      "learning_rate": 8.935414372197566e-05,
      "loss": 1.2268,
      "mean_token_accuracy": 0.7634308502078057,
      "num_tokens": 839688040.0,
      "step": 12430
    },
    {
      "entropy": 0.8294574156403541,
      "epoch": 3.768613193971067,
      "grad_norm": 0.13719813525676727,
      "learning_rate": 8.933353626420845e-05,
      "loss": 1.2403,
      "mean_token_accuracy": 0.759545375406742,
      "num_tokens": 840359150.0,
      "step": 12440
    },
    {
      "entropy": 0.8285234406590462,
      "epoch": 3.7716428084526243,
      "grad_norm": 0.1382272094488144,
      "learning_rate": 8.931291126171641e-05,
      "loss": 1.234,
      "mean_token_accuracy": 0.7611021846532822,
      "num_tokens": 841021983.0,
      "step": 12450
    },
    {
      "entropy": 0.836207677423954,
      "epoch": 3.7746724229341817,
      "grad_norm": 0.12078831344842911,
      "learning_rate": 8.929226872369931e-05,
      "loss": 1.2335,
      "mean_token_accuracy": 0.7623356714844703,
      "num_tokens": 841706968.0,
      "step": 12460
    },
    {
      "entropy": 0.8264741778373719,
      "epoch": 3.7777020374157386,
      "grad_norm": 0.1228814572095871,
      "learning_rate": 8.927160865936476e-05,
      "loss": 1.2339,
      "mean_token_accuracy": 0.7623338222503662,
      "num_tokens": 842372400.0,
      "step": 12470
    },
    {
      "entropy": 0.827699501812458,
      "epoch": 3.780731651897296,
      "grad_norm": 0.12411075830459595,
      "learning_rate": 8.925093107792818e-05,
      "loss": 1.2319,
      "mean_token_accuracy": 0.7617390379309654,
      "num_tokens": 843040839.0,
      "step": 12480
    },
    {
      "entropy": 0.8308931544423104,
      "epoch": 3.7837612663788533,
      "grad_norm": 0.12857046723365784,
      "learning_rate": 8.92302359886128e-05,
      "loss": 1.2352,
      "mean_token_accuracy": 0.7628135606646538,
      "num_tokens": 843709430.0,
      "step": 12490
    },
    {
      "entropy": 0.8338014483451843,
      "epoch": 3.7867908808604103,
      "grad_norm": 0.12557992339134216,
      "learning_rate": 8.920952340064968e-05,
      "loss": 1.2323,
      "mean_token_accuracy": 0.7621446952223778,
      "num_tokens": 844385709.0,
      "step": 12500
    },
    {
      "entropy": 0.8305648371577263,
      "epoch": 3.7898204953419676,
      "grad_norm": 0.14182014763355255,
      "learning_rate": 8.918879332327765e-05,
      "loss": 1.2302,
      "mean_token_accuracy": 0.763672611117363,
      "num_tokens": 845077281.0,
      "step": 12510
    },
    {
      "entropy": 0.8348019689321518,
      "epoch": 3.792850109823525,
      "grad_norm": 0.11439071595668793,
      "learning_rate": 8.916804576574337e-05,
      "loss": 1.2339,
      "mean_token_accuracy": 0.7650744035840035,
      "num_tokens": 845760517.0,
      "step": 12520
    },
    {
      "entropy": 0.8238743111491204,
      "epoch": 3.795879724305082,
      "grad_norm": 0.12323829531669617,
      "learning_rate": 8.91472807373013e-05,
      "loss": 1.2279,
      "mean_token_accuracy": 0.7617947995662689,
      "num_tokens": 846422146.0,
      "step": 12530
    },
    {
      "entropy": 0.8199110209941864,
      "epoch": 3.7989093387866393,
      "grad_norm": 0.11645542085170746,
      "learning_rate": 8.912649824721367e-05,
      "loss": 1.2203,
      "mean_token_accuracy": 0.7677612766623497,
      "num_tokens": 847107414.0,
      "step": 12540
    },
    {
      "entropy": 0.8264547348022461,
      "epoch": 3.8019389532681966,
      "grad_norm": 0.12416919320821762,
      "learning_rate": 8.910569830475051e-05,
      "loss": 1.2431,
      "mean_token_accuracy": 0.7657420426607132,
      "num_tokens": 847782719.0,
      "step": 12550
    },
    {
      "entropy": 0.8251286134123802,
      "epoch": 3.804968567749754,
      "grad_norm": 0.13622988760471344,
      "learning_rate": 8.908488091918963e-05,
      "loss": 1.2271,
      "mean_token_accuracy": 0.7634443789720535,
      "num_tokens": 848454904.0,
      "step": 12560
    },
    {
      "entropy": 0.8287327572703361,
      "epoch": 3.807998182231311,
      "grad_norm": 0.1126769632101059,
      "learning_rate": 8.906404609981664e-05,
      "loss": 1.2322,
      "mean_token_accuracy": 0.763195650279522,
      "num_tokens": 849125354.0,
      "step": 12570
    },
    {
      "entropy": 0.8304981291294098,
      "epoch": 3.8110277967128683,
      "grad_norm": 0.13094691932201385,
      "learning_rate": 8.90431938559249e-05,
      "loss": 1.2344,
      "mean_token_accuracy": 0.7624080702662468,
      "num_tokens": 849790788.0,
      "step": 12580
    },
    {
      "entropy": 0.8332023620605469,
      "epoch": 3.8140574111944257,
      "grad_norm": 0.1280962973833084,
      "learning_rate": 8.902232419681557e-05,
      "loss": 1.2293,
      "mean_token_accuracy": 0.7648158460855484,
      "num_tokens": 850474638.0,
      "step": 12590
    },
    {
      "entropy": 0.8346199035644531,
      "epoch": 3.8170870256759826,
      "grad_norm": 0.1164981797337532,
      "learning_rate": 8.900143713179754e-05,
      "loss": 1.2371,
      "mean_token_accuracy": 0.7623358771204949,
      "num_tokens": 851145955.0,
      "step": 12600
    },
    {
      "entropy": 0.8156718030571938,
      "epoch": 3.82011664015754,
      "grad_norm": 0.11042694747447968,
      "learning_rate": 8.898053267018749e-05,
      "loss": 1.2215,
      "mean_token_accuracy": 0.7646094113588333,
      "num_tokens": 851816436.0,
      "step": 12610
    },
    {
      "entropy": 0.8390977829694748,
      "epoch": 3.8231462546390973,
      "grad_norm": 0.12022904306650162,
      "learning_rate": 8.895961082130989e-05,
      "loss": 1.2421,
      "mean_token_accuracy": 0.7592789888381958,
      "num_tokens": 852484169.0,
      "step": 12620
    },
    {
      "entropy": 0.830880680680275,
      "epoch": 3.8261758691206547,
      "grad_norm": 0.12470162659883499,
      "learning_rate": 8.893867159449687e-05,
      "loss": 1.2287,
      "mean_token_accuracy": 0.7626701295375824,
      "num_tokens": 853162486.0,
      "step": 12630
    },
    {
      "entropy": 0.8348073527216912,
      "epoch": 3.8292054836022116,
      "grad_norm": 0.12988479435443878,
      "learning_rate": 8.891771499908841e-05,
      "loss": 1.2395,
      "mean_token_accuracy": 0.762461656332016,
      "num_tokens": 853836077.0,
      "step": 12640
    },
    {
      "entropy": 0.8428922668099403,
      "epoch": 3.832235098083769,
      "grad_norm": 0.13711707293987274,
      "learning_rate": 8.889674104443221e-05,
      "loss": 1.2467,
      "mean_token_accuracy": 0.761611770093441,
      "num_tokens": 854518670.0,
      "step": 12650
    },
    {
      "entropy": 0.8193681225180626,
      "epoch": 3.8352647125653263,
      "grad_norm": 0.12601418793201447,
      "learning_rate": 8.887574973988368e-05,
      "loss": 1.2221,
      "mean_token_accuracy": 0.763744181394577,
      "num_tokens": 855196632.0,
      "step": 12660
    },
    {
      "entropy": 0.8222943872213364,
      "epoch": 3.8382943270468832,
      "grad_norm": 0.12922759354114532,
      "learning_rate": 8.8854741094806e-05,
      "loss": 1.2326,
      "mean_token_accuracy": 0.7638288870453834,
      "num_tokens": 855871139.0,
      "step": 12670
    },
    {
      "entropy": 0.8280269578099251,
      "epoch": 3.8413239415284406,
      "grad_norm": 0.1370943933725357,
      "learning_rate": 8.883371511857008e-05,
      "loss": 1.2299,
      "mean_token_accuracy": 0.7631358623504638,
      "num_tokens": 856534323.0,
      "step": 12680
    },
    {
      "entropy": 0.8340017095208168,
      "epoch": 3.844353556009998,
      "grad_norm": 0.1301402598619461,
      "learning_rate": 8.881267182055455e-05,
      "loss": 1.2324,
      "mean_token_accuracy": 0.7608328521251678,
      "num_tokens": 857207021.0,
      "step": 12690
    },
    {
      "entropy": 0.8315331548452377,
      "epoch": 3.847383170491555,
      "grad_norm": 0.11876879632472992,
      "learning_rate": 8.879161121014578e-05,
      "loss": 1.232,
      "mean_token_accuracy": 0.7626194566488266,
      "num_tokens": 857894691.0,
      "step": 12700
    },
    {
      "entropy": 0.8218623787164688,
      "epoch": 3.8504127849731122,
      "grad_norm": 0.12570799887180328,
      "learning_rate": 8.877053329673786e-05,
      "loss": 1.2282,
      "mean_token_accuracy": 0.7642956376075745,
      "num_tokens": 858566188.0,
      "step": 12710
    },
    {
      "entropy": 0.8263124644756317,
      "epoch": 3.8534423994546696,
      "grad_norm": 0.12733717262744904,
      "learning_rate": 8.874943808973257e-05,
      "loss": 1.2305,
      "mean_token_accuracy": 0.7641133055090904,
      "num_tokens": 859242879.0,
      "step": 12720
    },
    {
      "entropy": 0.837696586549282,
      "epoch": 3.8564720139362265,
      "grad_norm": 0.12091173976659775,
      "learning_rate": 8.872832559853946e-05,
      "loss": 1.2409,
      "mean_token_accuracy": 0.7607804715633393,
      "num_tokens": 859915787.0,
      "step": 12730
    },
    {
      "entropy": 0.8271465063095093,
      "epoch": 3.859501628417784,
      "grad_norm": 0.14471746981143951,
      "learning_rate": 8.870719583257574e-05,
      "loss": 1.2297,
      "mean_token_accuracy": 0.7651708543300628,
      "num_tokens": 860594248.0,
      "step": 12740
    },
    {
      "entropy": 0.812814648449421,
      "epoch": 3.8625312428993412,
      "grad_norm": 0.12557920813560486,
      "learning_rate": 8.868604880126634e-05,
      "loss": 1.2209,
      "mean_token_accuracy": 0.7714944124221802,
      "num_tokens": 861289719.0,
      "step": 12750
    },
    {
      "entropy": 0.8259367436170578,
      "epoch": 3.865560857380898,
      "grad_norm": 0.12413983792066574,
      "learning_rate": 8.86648845140439e-05,
      "loss": 1.2268,
      "mean_token_accuracy": 0.7617395550012589,
      "num_tokens": 861959403.0,
      "step": 12760
    },
    {
      "entropy": 0.8290757730603218,
      "epoch": 3.8685904718624555,
      "grad_norm": 0.117474764585495,
      "learning_rate": 8.864370298034875e-05,
      "loss": 1.2258,
      "mean_token_accuracy": 0.7611317902803421,
      "num_tokens": 862633647.0,
      "step": 12770
    },
    {
      "entropy": 0.8299058496952056,
      "epoch": 3.871620086344013,
      "grad_norm": 0.1349320411682129,
      "learning_rate": 8.862250420962892e-05,
      "loss": 1.2396,
      "mean_token_accuracy": 0.7624454960227013,
      "num_tokens": 863312912.0,
      "step": 12780
    },
    {
      "entropy": 0.8241622790694236,
      "epoch": 3.87464970082557,
      "grad_norm": 0.11762060225009918,
      "learning_rate": 8.86012882113401e-05,
      "loss": 1.2339,
      "mean_token_accuracy": 0.7630821943283081,
      "num_tokens": 863985023.0,
      "step": 12790
    },
    {
      "entropy": 0.8294262751936913,
      "epoch": 3.877679315307127,
      "grad_norm": 0.13306935131549835,
      "learning_rate": 8.858005499494573e-05,
      "loss": 1.2319,
      "mean_token_accuracy": 0.7615495726466179,
      "num_tokens": 864658888.0,
      "step": 12800
    },
    {
      "entropy": 0.8245569050312043,
      "epoch": 3.8807089297886845,
      "grad_norm": 0.1198149248957634,
      "learning_rate": 8.855880456991684e-05,
      "loss": 1.2298,
      "mean_token_accuracy": 0.7609682634472847,
      "num_tokens": 865318098.0,
      "step": 12810
    },
    {
      "entropy": 0.8243506833910942,
      "epoch": 3.8837385442702415,
      "grad_norm": 0.13753031194210052,
      "learning_rate": 8.853753694573219e-05,
      "loss": 1.223,
      "mean_token_accuracy": 0.7650005280971527,
      "num_tokens": 865995887.0,
      "step": 12820
    },
    {
      "entropy": 0.8278594672679901,
      "epoch": 3.886768158751799,
      "grad_norm": 0.11619915813207626,
      "learning_rate": 8.851625213187823e-05,
      "loss": 1.233,
      "mean_token_accuracy": 0.7637771308422089,
      "num_tokens": 866660519.0,
      "step": 12830
    },
    {
      "entropy": 0.8225967958569527,
      "epoch": 3.889797773233356,
      "grad_norm": 0.1255115568637848,
      "learning_rate": 8.849495013784904e-05,
      "loss": 1.2309,
      "mean_token_accuracy": 0.7614174097776413,
      "num_tokens": 867321427.0,
      "step": 12840
    },
    {
      "entropy": 0.8249985828995705,
      "epoch": 3.892827387714913,
      "grad_norm": 0.12730218470096588,
      "learning_rate": 8.847363097314636e-05,
      "loss": 1.2214,
      "mean_token_accuracy": 0.766840037703514,
      "num_tokens": 868019621.0,
      "step": 12850
    },
    {
      "entropy": 0.8264155417680741,
      "epoch": 3.8958570021964705,
      "grad_norm": 0.1336270421743393,
      "learning_rate": 8.845229464727962e-05,
      "loss": 1.2239,
      "mean_token_accuracy": 0.7623827084898949,
      "num_tokens": 868687725.0,
      "step": 12860
    },
    {
      "entropy": 0.8309557393193245,
      "epoch": 3.898886616678028,
      "grad_norm": 0.11706342548131943,
      "learning_rate": 8.84309411697659e-05,
      "loss": 1.233,
      "mean_token_accuracy": 0.7596857607364654,
      "num_tokens": 869360519.0,
      "step": 12870
    },
    {
      "entropy": 0.8345236212015152,
      "epoch": 3.9019162311595847,
      "grad_norm": 0.12713146209716797,
      "learning_rate": 8.840957055012989e-05,
      "loss": 1.2398,
      "mean_token_accuracy": 0.7650113448500633,
      "num_tokens": 870041084.0,
      "step": 12880
    },
    {
      "entropy": 0.8312649458646775,
      "epoch": 3.904945845641142,
      "grad_norm": 0.13826613128185272,
      "learning_rate": 8.838818279790396e-05,
      "loss": 1.2412,
      "mean_token_accuracy": 0.765237282216549,
      "num_tokens": 870714772.0,
      "step": 12890
    },
    {
      "entropy": 0.8348280221223832,
      "epoch": 3.9079754601226995,
      "grad_norm": 0.12400850653648376,
      "learning_rate": 8.836677792262814e-05,
      "loss": 1.2468,
      "mean_token_accuracy": 0.7586354508996009,
      "num_tokens": 871377777.0,
      "step": 12900
    },
    {
      "entropy": 0.8311003044247627,
      "epoch": 3.9110050746042564,
      "grad_norm": 0.12937553226947784,
      "learning_rate": 8.834535593385007e-05,
      "loss": 1.2447,
      "mean_token_accuracy": 0.7613857537508011,
      "num_tokens": 872042117.0,
      "step": 12910
    },
    {
      "entropy": 0.8208565428853035,
      "epoch": 3.9140346890858138,
      "grad_norm": 0.11717119812965393,
      "learning_rate": 8.8323916841125e-05,
      "loss": 1.2245,
      "mean_token_accuracy": 0.7649449229240417,
      "num_tokens": 872711042.0,
      "step": 12920
    },
    {
      "entropy": 0.8317764848470688,
      "epoch": 3.917064303567371,
      "grad_norm": 0.11867055296897888,
      "learning_rate": 8.830246065401588e-05,
      "loss": 1.2318,
      "mean_token_accuracy": 0.762212210893631,
      "num_tokens": 873383179.0,
      "step": 12930
    },
    {
      "entropy": 0.832485431432724,
      "epoch": 3.920093918048928,
      "grad_norm": 0.12630002200603485,
      "learning_rate": 8.82809873820932e-05,
      "loss": 1.2376,
      "mean_token_accuracy": 0.7598984807729721,
      "num_tokens": 874040069.0,
      "step": 12940
    },
    {
      "entropy": 0.839852300286293,
      "epoch": 3.9231235325304854,
      "grad_norm": 0.15134549140930176,
      "learning_rate": 8.825949703493515e-05,
      "loss": 1.2456,
      "mean_token_accuracy": 0.757933346927166,
      "num_tokens": 874695045.0,
      "step": 12950
    },
    {
      "entropy": 0.8406209647655487,
      "epoch": 3.9261531470120428,
      "grad_norm": 0.12075337767601013,
      "learning_rate": 8.823798962212748e-05,
      "loss": 1.2412,
      "mean_token_accuracy": 0.7612475737929344,
      "num_tokens": 875381984.0,
      "step": 12960
    },
    {
      "entropy": 0.8319113612174988,
      "epoch": 3.9291827614935997,
      "grad_norm": 0.1266404390335083,
      "learning_rate": 8.821646515326359e-05,
      "loss": 1.2391,
      "mean_token_accuracy": 0.7626997500658035,
      "num_tokens": 876044980.0,
      "step": 12970
    },
    {
      "entropy": 0.8321159169077873,
      "epoch": 3.932212375975157,
      "grad_norm": 0.12632417678833008,
      "learning_rate": 8.819492363794444e-05,
      "loss": 1.2379,
      "mean_token_accuracy": 0.7624675497412682,
      "num_tokens": 876720301.0,
      "step": 12980
    },
    {
      "entropy": 0.8342352926731109,
      "epoch": 3.9352419904567144,
      "grad_norm": 0.1273949295282364,
      "learning_rate": 8.817336508577865e-05,
      "loss": 1.2323,
      "mean_token_accuracy": 0.764403061568737,
      "num_tokens": 877405746.0,
      "step": 12990
    },
    {
      "entropy": 0.8261652573943138,
      "epoch": 3.9382716049382713,
      "grad_norm": 0.11959680169820786,
      "learning_rate": 8.815178950638239e-05,
      "loss": 1.2306,
      "mean_token_accuracy": 0.7622801020741463,
      "num_tokens": 878069456.0,
      "step": 13000
    },
    {
      "entropy": 0.837689746916294,
      "epoch": 3.9413012194198287,
      "grad_norm": 0.1293964833021164,
      "learning_rate": 8.813019690937947e-05,
      "loss": 1.2296,
      "mean_token_accuracy": 0.7628279864788056,
      "num_tokens": 878761245.0,
      "step": 13010
    },
    {
      "entropy": 0.8373796567320824,
      "epoch": 3.944330833901386,
      "grad_norm": 0.14157019555568695,
      "learning_rate": 8.810858730440126e-05,
      "loss": 1.2406,
      "mean_token_accuracy": 0.759574468433857,
      "num_tokens": 879432362.0,
      "step": 13020
    },
    {
      "entropy": 0.8271543443202972,
      "epoch": 3.9473604483829434,
      "grad_norm": 0.12580905854701996,
      "learning_rate": 8.808696070108671e-05,
      "loss": 1.243,
      "mean_token_accuracy": 0.7620757684111595,
      "num_tokens": 880097240.0,
      "step": 13030
    },
    {
      "entropy": 0.826493576169014,
      "epoch": 3.9503900628645003,
      "grad_norm": 0.1348680853843689,
      "learning_rate": 8.806531710908239e-05,
      "loss": 1.2325,
      "mean_token_accuracy": 0.7682598948478698,
      "num_tokens": 880783720.0,
      "step": 13040
    },
    {
      "entropy": 0.821631669998169,
      "epoch": 3.9534196773460577,
      "grad_norm": 0.13794854283332825,
      "learning_rate": 8.804365653804243e-05,
      "loss": 1.2168,
      "mean_token_accuracy": 0.7667650580406189,
      "num_tokens": 881478582.0,
      "step": 13050
    },
    {
      "entropy": 0.828932423889637,
      "epoch": 3.956449291827615,
      "grad_norm": 0.11824832856655121,
      "learning_rate": 8.802197899762849e-05,
      "loss": 1.2323,
      "mean_token_accuracy": 0.7659588649868965,
      "num_tokens": 882159416.0,
      "step": 13060
    },
    {
      "entropy": 0.8260334685444832,
      "epoch": 3.9594789063091724,
      "grad_norm": 0.12015210092067719,
      "learning_rate": 8.800028449750987e-05,
      "loss": 1.2287,
      "mean_token_accuracy": 0.7666187763214112,
      "num_tokens": 882839711.0,
      "step": 13070
    },
    {
      "entropy": 0.830867451429367,
      "epoch": 3.9625085207907293,
      "grad_norm": 0.11991837620735168,
      "learning_rate": 8.797857304736342e-05,
      "loss": 1.239,
      "mean_token_accuracy": 0.7593985632061958,
      "num_tokens": 883510673.0,
      "step": 13080
    },
    {
      "entropy": 0.8303544789552688,
      "epoch": 3.9655381352722867,
      "grad_norm": 0.11978612095117569,
      "learning_rate": 8.795684465687347e-05,
      "loss": 1.2344,
      "mean_token_accuracy": 0.7616124227643013,
      "num_tokens": 884194771.0,
      "step": 13090
    },
    {
      "entropy": 0.8189367249608039,
      "epoch": 3.968567749753844,
      "grad_norm": 0.11828684061765671,
      "learning_rate": 8.793509933573201e-05,
      "loss": 1.2246,
      "mean_token_accuracy": 0.7643954247236252,
      "num_tokens": 884865730.0,
      "step": 13100
    },
    {
      "entropy": 0.8249358102679253,
      "epoch": 3.971597364235401,
      "grad_norm": 0.12646810710430145,
      "learning_rate": 8.791333709363856e-05,
      "loss": 1.2308,
      "mean_token_accuracy": 0.7633047968149185,
      "num_tokens": 885537265.0,
      "step": 13110
    },
    {
      "entropy": 0.8147301137447357,
      "epoch": 3.9746269787169584,
      "grad_norm": 0.11637123674154282,
      "learning_rate": 8.789155794030012e-05,
      "loss": 1.2162,
      "mean_token_accuracy": 0.7662988126277923,
      "num_tokens": 886220738.0,
      "step": 13120
    },
    {
      "entropy": 0.8361765071749687,
      "epoch": 3.9776565931985157,
      "grad_norm": 0.13231906294822693,
      "learning_rate": 8.786976188543133e-05,
      "loss": 1.2292,
      "mean_token_accuracy": 0.7626230180263519,
      "num_tokens": 886920302.0,
      "step": 13130
    },
    {
      "entropy": 0.8213987976312638,
      "epoch": 3.9806862076800726,
      "grad_norm": 0.12882764637470245,
      "learning_rate": 8.784794893875428e-05,
      "loss": 1.223,
      "mean_token_accuracy": 0.7662776485085487,
      "num_tokens": 887604571.0,
      "step": 13140
    },
    {
      "entropy": 0.8245305463671684,
      "epoch": 3.98371582216163,
      "grad_norm": 0.12455525249242783,
      "learning_rate": 8.782611910999865e-05,
      "loss": 1.2324,
      "mean_token_accuracy": 0.7616021037101746,
      "num_tokens": 888262413.0,
      "step": 13150
    },
    {
      "entropy": 0.8323712050914764,
      "epoch": 3.9867454366431874,
      "grad_norm": 0.1199771836400032,
      "learning_rate": 8.780427240890166e-05,
      "loss": 1.2349,
      "mean_token_accuracy": 0.7597421824932098,
      "num_tokens": 888933393.0,
      "step": 13160
    },
    {
      "entropy": 0.8318541258573532,
      "epoch": 3.9897750511247443,
      "grad_norm": 0.11837221682071686,
      "learning_rate": 8.778240884520798e-05,
      "loss": 1.2325,
      "mean_token_accuracy": 0.7667552843689919,
      "num_tokens": 889621517.0,
      "step": 13170
    },
    {
      "entropy": 0.8278790339827538,
      "epoch": 3.9928046656063017,
      "grad_norm": 0.13204719126224518,
      "learning_rate": 8.776052842866992e-05,
      "loss": 1.227,
      "mean_token_accuracy": 0.7613449260592461,
      "num_tokens": 890300645.0,
      "step": 13180
    },
    {
      "entropy": 0.8214327931404114,
      "epoch": 3.995834280087859,
      "grad_norm": 0.1523841768503189,
      "learning_rate": 8.773863116904718e-05,
      "loss": 1.2325,
      "mean_token_accuracy": 0.7664753451943398,
      "num_tokens": 890970218.0,
      "step": 13190
    },
    {
      "entropy": 0.8335215345025062,
      "epoch": 3.998863894569416,
      "grad_norm": 0.13725320994853973,
      "learning_rate": 8.771671707610707e-05,
      "loss": 1.2467,
      "mean_token_accuracy": 0.7620050013065338,
      "num_tokens": 891640496.0,
      "step": 13200
    },
    {
      "entropy": 0.8312990482036884,
      "epoch": 4.001817768688935,
      "grad_norm": 0.1220497339963913,
      "learning_rate": 8.769478615962436e-05,
      "loss": 1.2284,
      "mean_token_accuracy": 0.7663144346995231,
      "num_tokens": 892314225.0,
      "step": 13210
    },
    {
      "entropy": 0.811907809972763,
      "epoch": 4.004847383170492,
      "grad_norm": 0.1275583654642105,
      "learning_rate": 8.767283842938133e-05,
      "loss": 1.2171,
      "mean_token_accuracy": 0.7655479878187179,
      "num_tokens": 892990894.0,
      "step": 13220
    },
    {
      "entropy": 0.8137328565120697,
      "epoch": 4.0078769976520485,
      "grad_norm": 0.1291702687740326,
      "learning_rate": 8.765087389516775e-05,
      "loss": 1.2217,
      "mean_token_accuracy": 0.765879887342453,
      "num_tokens": 893667329.0,
      "step": 13230
    },
    {
      "entropy": 0.8127585962414742,
      "epoch": 4.010906612133606,
      "grad_norm": 0.1276501566171646,
      "learning_rate": 8.762889256678095e-05,
      "loss": 1.2071,
      "mean_token_accuracy": 0.7639842689037323,
      "num_tokens": 894343030.0,
      "step": 13240
    },
    {
      "entropy": 0.8192038983106613,
      "epoch": 4.013936226615163,
      "grad_norm": 0.13279663026332855,
      "learning_rate": 8.760689445402567e-05,
      "loss": 1.2175,
      "mean_token_accuracy": 0.7616088435053825,
      "num_tokens": 895016670.0,
      "step": 13250
    },
    {
      "entropy": 0.8052625715732574,
      "epoch": 4.01696584109672,
      "grad_norm": 0.14576607942581177,
      "learning_rate": 8.758487956671415e-05,
      "loss": 1.2088,
      "mean_token_accuracy": 0.770484383404255,
      "num_tokens": 895710879.0,
      "step": 13260
    },
    {
      "entropy": 0.8075692936778068,
      "epoch": 4.019995455578278,
      "grad_norm": 0.13980166614055634,
      "learning_rate": 8.756284791466617e-05,
      "loss": 1.2171,
      "mean_token_accuracy": 0.7611551076173783,
      "num_tokens": 896362662.0,
      "step": 13270
    },
    {
      "entropy": 0.7997980833053588,
      "epoch": 4.023025070059835,
      "grad_norm": 0.13313572108745575,
      "learning_rate": 8.754079950770894e-05,
      "loss": 1.2079,
      "mean_token_accuracy": 0.7700257301330566,
      "num_tokens": 897039094.0,
      "step": 13280
    },
    {
      "entropy": 0.8125384911894799,
      "epoch": 4.026054684541392,
      "grad_norm": 0.1359756737947464,
      "learning_rate": 8.751873435567715e-05,
      "loss": 1.2113,
      "mean_token_accuracy": 0.7646631121635437,
      "num_tokens": 897717124.0,
      "step": 13290
    },
    {
      "entropy": 0.8125770598649978,
      "epoch": 4.02908429902295,
      "grad_norm": 0.14367371797561646,
      "learning_rate": 8.749665246841293e-05,
      "loss": 1.2127,
      "mean_token_accuracy": 0.7631704553961753,
      "num_tokens": 898378571.0,
      "step": 13300
    },
    {
      "entropy": 0.8129245296120644,
      "epoch": 4.0321139135045065,
      "grad_norm": 0.14231349527835846,
      "learning_rate": 8.747455385576595e-05,
      "loss": 1.2146,
      "mean_token_accuracy": 0.7658406496047974,
      "num_tokens": 899052172.0,
      "step": 13310
    },
    {
      "entropy": 0.8133532971143722,
      "epoch": 4.0351435279860635,
      "grad_norm": 0.15407222509384155,
      "learning_rate": 8.745243852759328e-05,
      "loss": 1.2186,
      "mean_token_accuracy": 0.7606793612241745,
      "num_tokens": 899704453.0,
      "step": 13320
    },
    {
      "entropy": 0.8136651709675788,
      "epoch": 4.038173142467621,
      "grad_norm": 0.1306125670671463,
      "learning_rate": 8.743030649375944e-05,
      "loss": 1.2049,
      "mean_token_accuracy": 0.7665054619312286,
      "num_tokens": 900388361.0,
      "step": 13330
    },
    {
      "entropy": 0.8115026831626893,
      "epoch": 4.041202756949178,
      "grad_norm": 0.12938164174556732,
      "learning_rate": 8.740815776413649e-05,
      "loss": 1.2175,
      "mean_token_accuracy": 0.7635725408792495,
      "num_tokens": 901057525.0,
      "step": 13340
    },
    {
      "entropy": 0.8118620291352272,
      "epoch": 4.044232371430735,
      "grad_norm": 0.1395217478275299,
      "learning_rate": 8.73859923486038e-05,
      "loss": 1.2151,
      "mean_token_accuracy": 0.7668686181306839,
      "num_tokens": 901734076.0,
      "step": 13350
    },
    {
      "entropy": 0.813816773891449,
      "epoch": 4.047261985912293,
      "grad_norm": 0.14438989758491516,
      "learning_rate": 8.736381025704827e-05,
      "loss": 1.2091,
      "mean_token_accuracy": 0.7655443996191025,
      "num_tokens": 902418339.0,
      "step": 13360
    },
    {
      "entropy": 0.8176364481449128,
      "epoch": 4.05029160039385,
      "grad_norm": 0.13676360249519348,
      "learning_rate": 8.734161149936426e-05,
      "loss": 1.2147,
      "mean_token_accuracy": 0.7684245362877846,
      "num_tokens": 903101313.0,
      "step": 13370
    },
    {
      "entropy": 0.8075049206614494,
      "epoch": 4.053321214875407,
      "grad_norm": 0.13936209678649902,
      "learning_rate": 8.73193960854535e-05,
      "loss": 1.2114,
      "mean_token_accuracy": 0.7677946135401725,
      "num_tokens": 903786652.0,
      "step": 13380
    },
    {
      "entropy": 0.815261285007,
      "epoch": 4.056350829356965,
      "grad_norm": 0.12821030616760254,
      "learning_rate": 8.729716402522517e-05,
      "loss": 1.215,
      "mean_token_accuracy": 0.7635524079203606,
      "num_tokens": 904449904.0,
      "step": 13390
    },
    {
      "entropy": 0.8139819651842117,
      "epoch": 4.0593804438385215,
      "grad_norm": 0.12646350264549255,
      "learning_rate": 8.727491532859587e-05,
      "loss": 1.2166,
      "mean_token_accuracy": 0.7698590472340584,
      "num_tokens": 905133903.0,
      "step": 13400
    },
    {
      "entropy": 0.8102973580360413,
      "epoch": 4.062410058320078,
      "grad_norm": 0.13327458500862122,
      "learning_rate": 8.725265000548967e-05,
      "loss": 1.2118,
      "mean_token_accuracy": 0.7633210942149162,
      "num_tokens": 905805761.0,
      "step": 13410
    },
    {
      "entropy": 0.8091414451599122,
      "epoch": 4.065439672801636,
      "grad_norm": 0.1316269487142563,
      "learning_rate": 8.7230368065838e-05,
      "loss": 1.2106,
      "mean_token_accuracy": 0.7654315799474716,
      "num_tokens": 906479368.0,
      "step": 13420
    },
    {
      "entropy": 0.8207312270998954,
      "epoch": 4.068469287283193,
      "grad_norm": 0.1382802426815033,
      "learning_rate": 8.72080695195797e-05,
      "loss": 1.2193,
      "mean_token_accuracy": 0.7628284618258476,
      "num_tokens": 907148763.0,
      "step": 13430
    },
    {
      "entropy": 0.8118870124220848,
      "epoch": 4.07149890176475,
      "grad_norm": 0.13254445791244507,
      "learning_rate": 8.718575437666107e-05,
      "loss": 1.2173,
      "mean_token_accuracy": 0.7619264394044876,
      "num_tokens": 907810001.0,
      "step": 13440
    },
    {
      "entropy": 0.8125631526112557,
      "epoch": 4.074528516246308,
      "grad_norm": 0.1385975182056427,
      "learning_rate": 8.716342264703577e-05,
      "loss": 1.2106,
      "mean_token_accuracy": 0.7686882153153419,
      "num_tokens": 908492593.0,
      "step": 13450
    },
    {
      "entropy": 0.8135557323694229,
      "epoch": 4.077558130727865,
      "grad_norm": 0.13102348148822784,
      "learning_rate": 8.714107434066485e-05,
      "loss": 1.2093,
      "mean_token_accuracy": 0.77032929956913,
      "num_tokens": 909188260.0,
      "step": 13460
    },
    {
      "entropy": 0.8188177153468132,
      "epoch": 4.080587745209422,
      "grad_norm": 0.1288897544145584,
      "learning_rate": 8.711870946751677e-05,
      "loss": 1.2183,
      "mean_token_accuracy": 0.7633365780115128,
      "num_tokens": 909867008.0,
      "step": 13470
    },
    {
      "entropy": 0.8076186761260032,
      "epoch": 4.0836173596909795,
      "grad_norm": 0.13218490779399872,
      "learning_rate": 8.709632803756743e-05,
      "loss": 1.213,
      "mean_token_accuracy": 0.7662195295095444,
      "num_tokens": 910534417.0,
      "step": 13480
    },
    {
      "entropy": 0.8244075238704681,
      "epoch": 4.086646974172536,
      "grad_norm": 0.13766668736934662,
      "learning_rate": 8.70739300608e-05,
      "loss": 1.2176,
      "mean_token_accuracy": 0.7656984314322471,
      "num_tokens": 911198736.0,
      "step": 13490
    },
    {
      "entropy": 0.8187069833278656,
      "epoch": 4.089676588654093,
      "grad_norm": 0.13429325819015503,
      "learning_rate": 8.705151554720516e-05,
      "loss": 1.2158,
      "mean_token_accuracy": 0.7615954145789147,
      "num_tokens": 911867620.0,
      "step": 13500
    },
    {
      "entropy": 0.8119468748569488,
      "epoch": 4.092706203135651,
      "grad_norm": 0.12528271973133087,
      "learning_rate": 8.702908450678088e-05,
      "loss": 1.2082,
      "mean_token_accuracy": 0.7680399090051651,
      "num_tokens": 912557718.0,
      "step": 13510
    },
    {
      "entropy": 0.8165273770689965,
      "epoch": 4.095735817617208,
      "grad_norm": 0.13239304721355438,
      "learning_rate": 8.700663694953253e-05,
      "loss": 1.2087,
      "mean_token_accuracy": 0.7682377830147743,
      "num_tokens": 913259735.0,
      "step": 13520
    },
    {
      "entropy": 0.8079084277153015,
      "epoch": 4.098765432098766,
      "grad_norm": 0.1349460929632187,
      "learning_rate": 8.698417288547282e-05,
      "loss": 1.2039,
      "mean_token_accuracy": 0.7641630545258522,
      "num_tokens": 913931114.0,
      "step": 13530
    },
    {
      "entropy": 0.8251795873045922,
      "epoch": 4.101795046580323,
      "grad_norm": 0.13713224232196808,
      "learning_rate": 8.696169232462186e-05,
      "loss": 1.2312,
      "mean_token_accuracy": 0.7643203541636467,
      "num_tokens": 914605485.0,
      "step": 13540
    },
    {
      "entropy": 0.8222374960780143,
      "epoch": 4.10482466106188,
      "grad_norm": 0.1352066695690155,
      "learning_rate": 8.693919527700712e-05,
      "loss": 1.228,
      "mean_token_accuracy": 0.7634329333901405,
      "num_tokens": 915283506.0,
      "step": 13550
    },
    {
      "entropy": 0.8097446709871292,
      "epoch": 4.1078542755434375,
      "grad_norm": 0.13599705696105957,
      "learning_rate": 8.691668175266339e-05,
      "loss": 1.2089,
      "mean_token_accuracy": 0.7686758816242218,
      "num_tokens": 915964558.0,
      "step": 13560
    },
    {
      "entropy": 0.8166528180241585,
      "epoch": 4.110883890024994,
      "grad_norm": 0.1307644248008728,
      "learning_rate": 8.689415176163284e-05,
      "loss": 1.2236,
      "mean_token_accuracy": 0.7618932396173477,
      "num_tokens": 916628512.0,
      "step": 13570
    },
    {
      "entropy": 0.8150522410869598,
      "epoch": 4.113913504506551,
      "grad_norm": 0.14148427546024323,
      "learning_rate": 8.687160531396497e-05,
      "loss": 1.2116,
      "mean_token_accuracy": 0.766238908469677,
      "num_tokens": 917296021.0,
      "step": 13580
    },
    {
      "entropy": 0.8116999998688698,
      "epoch": 4.116943118988109,
      "grad_norm": 0.13924165070056915,
      "learning_rate": 8.68490424197166e-05,
      "loss": 1.2225,
      "mean_token_accuracy": 0.760797917842865,
      "num_tokens": 917948662.0,
      "step": 13590
    },
    {
      "entropy": 0.8144552662968636,
      "epoch": 4.119972733469666,
      "grad_norm": 0.13732901215553284,
      "learning_rate": 8.682646308895198e-05,
      "loss": 1.225,
      "mean_token_accuracy": 0.7634315997362137,
      "num_tokens": 918612165.0,
      "step": 13600
    },
    {
      "entropy": 0.8196040391921997,
      "epoch": 4.123002347951223,
      "grad_norm": 0.1463479995727539,
      "learning_rate": 8.680386733174252e-05,
      "loss": 1.2255,
      "mean_token_accuracy": 0.7612081736326217,
      "num_tokens": 919268989.0,
      "step": 13610
    },
    {
      "entropy": 0.8091532364487648,
      "epoch": 4.126031962432781,
      "grad_norm": 0.1418394297361374,
      "learning_rate": 8.678125515816715e-05,
      "loss": 1.2138,
      "mean_token_accuracy": 0.7636659547686577,
      "num_tokens": 919930574.0,
      "step": 13620
    },
    {
      "entropy": 0.8077966541051864,
      "epoch": 4.129061576914338,
      "grad_norm": 0.13525235652923584,
      "learning_rate": 8.675862657831198e-05,
      "loss": 1.2105,
      "mean_token_accuracy": 0.7673310711979866,
      "num_tokens": 920608705.0,
      "step": 13630
    },
    {
      "entropy": 0.8101565122604371,
      "epoch": 4.132091191395895,
      "grad_norm": 0.14173512160778046,
      "learning_rate": 8.673598160227048e-05,
      "loss": 1.2029,
      "mean_token_accuracy": 0.7678299769759178,
      "num_tokens": 921299297.0,
      "step": 13640
    },
    {
      "entropy": 0.81236642152071,
      "epoch": 4.1351208058774525,
      "grad_norm": 0.13491219282150269,
      "learning_rate": 8.671332024014347e-05,
      "loss": 1.2141,
      "mean_token_accuracy": 0.7643201470375061,
      "num_tokens": 921969151.0,
      "step": 13650
    },
    {
      "entropy": 0.814178629219532,
      "epoch": 4.138150420359009,
      "grad_norm": 0.13452528417110443,
      "learning_rate": 8.669064250203905e-05,
      "loss": 1.209,
      "mean_token_accuracy": 0.7639429673552514,
      "num_tokens": 922644973.0,
      "step": 13660
    },
    {
      "entropy": 0.8204580396413803,
      "epoch": 4.141180034840566,
      "grad_norm": 0.13058091700077057,
      "learning_rate": 8.666794839807262e-05,
      "loss": 1.2231,
      "mean_token_accuracy": 0.7653258740901947,
      "num_tokens": 923327982.0,
      "step": 13670
    },
    {
      "entropy": 0.8129162982106208,
      "epoch": 4.144209649322124,
      "grad_norm": 0.13659673929214478,
      "learning_rate": 8.664523793836688e-05,
      "loss": 1.2108,
      "mean_token_accuracy": 0.7651533886790276,
      "num_tokens": 924003653.0,
      "step": 13680
    },
    {
      "entropy": 0.8184856981039047,
      "epoch": 4.147239263803681,
      "grad_norm": 0.13943171501159668,
      "learning_rate": 8.662251113305184e-05,
      "loss": 1.2201,
      "mean_token_accuracy": 0.7643186420202255,
      "num_tokens": 924675200.0,
      "step": 13690
    },
    {
      "entropy": 0.810869137942791,
      "epoch": 4.150268878285238,
      "grad_norm": 0.13399697840213776,
      "learning_rate": 8.65997679922648e-05,
      "loss": 1.2063,
      "mean_token_accuracy": 0.7663172572851181,
      "num_tokens": 925345355.0,
      "step": 13700
    },
    {
      "entropy": 0.8122163027524948,
      "epoch": 4.153298492766796,
      "grad_norm": 0.17282022535800934,
      "learning_rate": 8.65770085261503e-05,
      "loss": 1.2168,
      "mean_token_accuracy": 0.7658040776848793,
      "num_tokens": 926015775.0,
      "step": 13710
    },
    {
      "entropy": 0.8136539295315742,
      "epoch": 4.156328107248353,
      "grad_norm": 0.1364956945180893,
      "learning_rate": 8.655423274486023e-05,
      "loss": 1.2206,
      "mean_token_accuracy": 0.7697982728481293,
      "num_tokens": 926704437.0,
      "step": 13720
    },
    {
      "entropy": 0.8150509908795357,
      "epoch": 4.15935772172991,
      "grad_norm": 0.13956472277641296,
      "learning_rate": 8.653144065855374e-05,
      "loss": 1.219,
      "mean_token_accuracy": 0.7586035177111625,
      "num_tokens": 927351346.0,
      "step": 13730
    },
    {
      "entropy": 0.8135612204670906,
      "epoch": 4.162387336211467,
      "grad_norm": 0.1430983990430832,
      "learning_rate": 8.650863227739722e-05,
      "loss": 1.2204,
      "mean_token_accuracy": 0.766981391608715,
      "num_tokens": 928026910.0,
      "step": 13740
    },
    {
      "entropy": 0.795094721019268,
      "epoch": 4.165416950693024,
      "grad_norm": 0.13425539433956146,
      "learning_rate": 8.648580761156434e-05,
      "loss": 1.206,
      "mean_token_accuracy": 0.7714309111237526,
      "num_tokens": 928702950.0,
      "step": 13750
    },
    {
      "entropy": 0.8066852271556855,
      "epoch": 4.168446565174581,
      "grad_norm": 0.13715395331382751,
      "learning_rate": 8.646296667123607e-05,
      "loss": 1.2129,
      "mean_token_accuracy": 0.7653296887874603,
      "num_tokens": 929371855.0,
      "step": 13760
    },
    {
      "entropy": 0.8244137406349182,
      "epoch": 4.171476179656139,
      "grad_norm": 0.13420824706554413,
      "learning_rate": 8.64401094666006e-05,
      "loss": 1.2228,
      "mean_token_accuracy": 0.7606052219867706,
      "num_tokens": 930047999.0,
      "step": 13770
    },
    {
      "entropy": 0.8052446603775024,
      "epoch": 4.174505794137696,
      "grad_norm": 0.13088949024677277,
      "learning_rate": 8.641723600785338e-05,
      "loss": 1.208,
      "mean_token_accuracy": 0.7691586732864379,
      "num_tokens": 930730923.0,
      "step": 13780
    },
    {
      "entropy": 0.80958491563797,
      "epoch": 4.177535408619253,
      "grad_norm": 0.1393127143383026,
      "learning_rate": 8.639434630519712e-05,
      "loss": 1.2196,
      "mean_token_accuracy": 0.7635791033506394,
      "num_tokens": 931400971.0,
      "step": 13790
    },
    {
      "entropy": 0.8069665089249611,
      "epoch": 4.180565023100811,
      "grad_norm": 0.16349659860134125,
      "learning_rate": 8.637144036884178e-05,
      "loss": 1.2022,
      "mean_token_accuracy": 0.7647842794656754,
      "num_tokens": 932073176.0,
      "step": 13800
    },
    {
      "entropy": 0.8113209322094918,
      "epoch": 4.183594637582368,
      "grad_norm": 0.1333882361650467,
      "learning_rate": 8.634851820900455e-05,
      "loss": 1.224,
      "mean_token_accuracy": 0.7642734766006469,
      "num_tokens": 932735046.0,
      "step": 13810
    },
    {
      "entropy": 0.8123429358005524,
      "epoch": 4.1866242520639245,
      "grad_norm": 0.13888338208198547,
      "learning_rate": 8.632557983590985e-05,
      "loss": 1.2188,
      "mean_token_accuracy": 0.7666823834180831,
      "num_tokens": 933411711.0,
      "step": 13820
    },
    {
      "entropy": 0.8132064625620842,
      "epoch": 4.189653866545482,
      "grad_norm": 0.14748670160770416,
      "learning_rate": 8.630262525978937e-05,
      "loss": 1.2112,
      "mean_token_accuracy": 0.7644979268312454,
      "num_tokens": 934080196.0,
      "step": 13830
    },
    {
      "entropy": 0.8083377107977867,
      "epoch": 4.192683481027039,
      "grad_norm": 0.13556091487407684,
      "learning_rate": 8.627965449088196e-05,
      "loss": 1.2118,
      "mean_token_accuracy": 0.7660816133022308,
      "num_tokens": 934758164.0,
      "step": 13840
    },
    {
      "entropy": 0.8108568266034126,
      "epoch": 4.195713095508596,
      "grad_norm": 0.1261628419160843,
      "learning_rate": 8.625666753943375e-05,
      "loss": 1.2123,
      "mean_token_accuracy": 0.7633062645792961,
      "num_tokens": 935419773.0,
      "step": 13850
    },
    {
      "entropy": 0.8159422129392624,
      "epoch": 4.198742709990154,
      "grad_norm": 0.13253046572208405,
      "learning_rate": 8.623366441569806e-05,
      "loss": 1.2145,
      "mean_token_accuracy": 0.7646999746561051,
      "num_tokens": 936101595.0,
      "step": 13860
    },
    {
      "entropy": 0.8183444306254387,
      "epoch": 4.201772324471711,
      "grad_norm": 0.14163215458393097,
      "learning_rate": 8.621064512993548e-05,
      "loss": 1.2193,
      "mean_token_accuracy": 0.7625939324498177,
      "num_tokens": 936769221.0,
      "step": 13870
    },
    {
      "entropy": 0.811541984975338,
      "epoch": 4.204801938953268,
      "grad_norm": 0.13868501782417297,
      "learning_rate": 8.618760969241371e-05,
      "loss": 1.2109,
      "mean_token_accuracy": 0.7672978028655052,
      "num_tokens": 937458407.0,
      "step": 13880
    },
    {
      "entropy": 0.8121904909610749,
      "epoch": 4.207831553434826,
      "grad_norm": 0.13706368207931519,
      "learning_rate": 8.616455811340774e-05,
      "loss": 1.2108,
      "mean_token_accuracy": 0.7664934456348419,
      "num_tokens": 938139899.0,
      "step": 13890
    },
    {
      "entropy": 0.8236121952533721,
      "epoch": 4.2108611679163825,
      "grad_norm": 0.13133464753627777,
      "learning_rate": 8.614149040319968e-05,
      "loss": 1.2219,
      "mean_token_accuracy": 0.7693839401006699,
      "num_tokens": 938837457.0,
      "step": 13900
    },
    {
      "entropy": 0.8124981418251991,
      "epoch": 4.2138907823979395,
      "grad_norm": 0.136515811085701,
      "learning_rate": 8.611840657207895e-05,
      "loss": 1.2257,
      "mean_token_accuracy": 0.7673719182610512,
      "num_tokens": 939510878.0,
      "step": 13910
    },
    {
      "entropy": 0.8288216248154641,
      "epoch": 4.216920396879497,
      "grad_norm": 0.1298895627260208,
      "learning_rate": 8.609530663034204e-05,
      "loss": 1.2249,
      "mean_token_accuracy": 0.7594271808862686,
      "num_tokens": 940182253.0,
      "step": 13920
    },
    {
      "entropy": 0.8282486319541931,
      "epoch": 4.219950011361054,
      "grad_norm": 0.1499754935503006,
      "learning_rate": 8.60721905882927e-05,
      "loss": 1.2236,
      "mean_token_accuracy": 0.7609874993562699,
      "num_tokens": 940851472.0,
      "step": 13930
    },
    {
      "entropy": 0.8192692086100578,
      "epoch": 4.222979625842611,
      "grad_norm": 0.13645020127296448,
      "learning_rate": 8.604905845624184e-05,
      "loss": 1.2211,
      "mean_token_accuracy": 0.7661446630954742,
      "num_tokens": 941531104.0,
      "step": 13940
    },
    {
      "entropy": 0.8111198768019676,
      "epoch": 4.226009240324169,
      "grad_norm": 0.1310085952281952,
      "learning_rate": 8.602591024450757e-05,
      "loss": 1.2083,
      "mean_token_accuracy": 0.7674522623419762,
      "num_tokens": 942213304.0,
      "step": 13950
    },
    {
      "entropy": 0.8215175598859787,
      "epoch": 4.229038854805726,
      "grad_norm": 0.15496326982975006,
      "learning_rate": 8.600274596341511e-05,
      "loss": 1.2154,
      "mean_token_accuracy": 0.7666410595178604,
      "num_tokens": 942905829.0,
      "step": 13960
    },
    {
      "entropy": 0.8067178651690483,
      "epoch": 4.232068469287283,
      "grad_norm": 0.13284319639205933,
      "learning_rate": 8.597956562329691e-05,
      "loss": 1.207,
      "mean_token_accuracy": 0.7688114762306213,
      "num_tokens": 943582951.0,
      "step": 13970
    },
    {
      "entropy": 0.8132088467478752,
      "epoch": 4.235098083768841,
      "grad_norm": 0.1390438824892044,
      "learning_rate": 8.595636923449256e-05,
      "loss": 1.2129,
      "mean_token_accuracy": 0.7628922238945961,
      "num_tokens": 944258139.0,
      "step": 13980
    },
    {
      "entropy": 0.8245077699422836,
      "epoch": 4.2381276982503975,
      "grad_norm": 0.1278022974729538,
      "learning_rate": 8.593315680734881e-05,
      "loss": 1.2203,
      "mean_token_accuracy": 0.7631716653704643,
      "num_tokens": 944938300.0,
      "step": 13990
    },
    {
      "entropy": 0.8103908881545067,
      "epoch": 4.241157312731955,
      "grad_norm": 0.15149471163749695,
      "learning_rate": 8.590992835221955e-05,
      "loss": 1.2129,
      "mean_token_accuracy": 0.7649084851145744,
      "num_tokens": 945609140.0,
      "step": 14000
    },
    {
      "entropy": 0.8094377368688583,
      "epoch": 4.244186927213512,
      "grad_norm": 0.1317652463912964,
      "learning_rate": 8.588668387946586e-05,
      "loss": 1.2131,
      "mean_token_accuracy": 0.7669806465506553,
      "num_tokens": 946282228.0,
      "step": 14010
    },
    {
      "entropy": 0.8111464530229568,
      "epoch": 4.247216541695069,
      "grad_norm": 0.1401730626821518,
      "learning_rate": 8.58634233994559e-05,
      "loss": 1.2114,
      "mean_token_accuracy": 0.7699093639850616,
      "num_tokens": 946973660.0,
      "step": 14020
    },
    {
      "entropy": 0.8090970531105995,
      "epoch": 4.250246156176627,
      "grad_norm": 0.13968342542648315,
      "learning_rate": 8.584014692256506e-05,
      "loss": 1.2121,
      "mean_token_accuracy": 0.7649696469306946,
      "num_tokens": 947645356.0,
      "step": 14030
    },
    {
      "entropy": 0.8118878945708274,
      "epoch": 4.253275770658184,
      "grad_norm": 0.14246892929077148,
      "learning_rate": 8.581685445917578e-05,
      "loss": 1.2202,
      "mean_token_accuracy": 0.7633164018392563,
      "num_tokens": 948311021.0,
      "step": 14040
    },
    {
      "entropy": 0.8142295688390732,
      "epoch": 4.256305385139741,
      "grad_norm": 0.12931780517101288,
      "learning_rate": 8.579354601967766e-05,
      "loss": 1.212,
      "mean_token_accuracy": 0.7635192066431046,
      "num_tokens": 948980277.0,
      "step": 14050
    },
    {
      "entropy": 0.8129866749048233,
      "epoch": 4.259334999621299,
      "grad_norm": 0.13530075550079346,
      "learning_rate": 8.577022161446742e-05,
      "loss": 1.2063,
      "mean_token_accuracy": 0.7671345919370651,
      "num_tokens": 949665068.0,
      "step": 14060
    },
    {
      "entropy": 0.811836202442646,
      "epoch": 4.2623646141028555,
      "grad_norm": 0.1332283765077591,
      "learning_rate": 8.574688125394894e-05,
      "loss": 1.213,
      "mean_token_accuracy": 0.7662887141108513,
      "num_tokens": 950344526.0,
      "step": 14070
    },
    {
      "entropy": 0.817075201869011,
      "epoch": 4.265394228584412,
      "grad_norm": 0.13864247500896454,
      "learning_rate": 8.572352494853319e-05,
      "loss": 1.2227,
      "mean_token_accuracy": 0.7603160515427589,
      "num_tokens": 950998511.0,
      "step": 14080
    },
    {
      "entropy": 0.8112716108560563,
      "epoch": 4.26842384306597,
      "grad_norm": 0.14149050414562225,
      "learning_rate": 8.570015270863822e-05,
      "loss": 1.2079,
      "mean_token_accuracy": 0.7637853011488914,
      "num_tokens": 951673719.0,
      "step": 14090
    },
    {
      "entropy": 0.8229402303695679,
      "epoch": 4.271453457547527,
      "grad_norm": 0.1343902349472046,
      "learning_rate": 8.567676454468925e-05,
      "loss": 1.226,
      "mean_token_accuracy": 0.7641852721571922,
      "num_tokens": 952351935.0,
      "step": 14100
    },
    {
      "entropy": 0.8148226022720337,
      "epoch": 4.274483072029084,
      "grad_norm": 0.1378401517868042,
      "learning_rate": 8.565336046711854e-05,
      "loss": 1.2222,
      "mean_token_accuracy": 0.7640312537550926,
      "num_tokens": 953032820.0,
      "step": 14110
    },
    {
      "entropy": 0.8163686230778694,
      "epoch": 4.277512686510642,
      "grad_norm": 0.13915479183197021,
      "learning_rate": 8.56299404863655e-05,
      "loss": 1.2077,
      "mean_token_accuracy": 0.7669281914830208,
      "num_tokens": 953725056.0,
      "step": 14120
    },
    {
      "entropy": 0.8072247564792633,
      "epoch": 4.280542300992199,
      "grad_norm": 0.15108776092529297,
      "learning_rate": 8.560650461287662e-05,
      "loss": 1.2135,
      "mean_token_accuracy": 0.7661883860826493,
      "num_tokens": 954392636.0,
      "step": 14130
    },
    {
      "entropy": 0.8094939455389977,
      "epoch": 4.283571915473756,
      "grad_norm": 0.13692191243171692,
      "learning_rate": 8.558305285710546e-05,
      "loss": 1.2155,
      "mean_token_accuracy": 0.7612668544054031,
      "num_tokens": 955055043.0,
      "step": 14140
    },
    {
      "entropy": 0.816562020778656,
      "epoch": 4.2866015299553135,
      "grad_norm": 0.1471427083015442,
      "learning_rate": 8.555958522951269e-05,
      "loss": 1.2206,
      "mean_token_accuracy": 0.7659133851528168,
      "num_tokens": 955731236.0,
      "step": 14150
    },
    {
      "entropy": 0.81419358253479,
      "epoch": 4.28963114443687,
      "grad_norm": 0.1354820430278778,
      "learning_rate": 8.553610174056601e-05,
      "loss": 1.2191,
      "mean_token_accuracy": 0.7625333324074746,
      "num_tokens": 956391286.0,
      "step": 14160
    },
    {
      "entropy": 0.8200589090585708,
      "epoch": 4.292660758918427,
      "grad_norm": 0.1293133646249771,
      "learning_rate": 8.551260240074026e-05,
      "loss": 1.2203,
      "mean_token_accuracy": 0.7590763792395592,
      "num_tokens": 957043185.0,
      "step": 14170
    },
    {
      "entropy": 0.8099162101745605,
      "epoch": 4.295690373399985,
      "grad_norm": 0.14451120793819427,
      "learning_rate": 8.548908722051733e-05,
      "loss": 1.215,
      "mean_token_accuracy": 0.7641231268644333,
      "num_tokens": 957713860.0,
      "step": 14180
    },
    {
      "entropy": 0.8195161297917366,
      "epoch": 4.298719987881542,
      "grad_norm": 0.12840956449508667,
      "learning_rate": 8.546555621038613e-05,
      "loss": 1.2212,
      "mean_token_accuracy": 0.762025710940361,
      "num_tokens": 958374608.0,
      "step": 14190
    },
    {
      "entropy": 0.8093711003661156,
      "epoch": 4.301749602363099,
      "grad_norm": 0.14335192739963531,
      "learning_rate": 8.54420093808427e-05,
      "loss": 1.2084,
      "mean_token_accuracy": 0.7724227145314216,
      "num_tokens": 959069213.0,
      "step": 14200
    },
    {
      "entropy": 0.8138993307948112,
      "epoch": 4.304779216844657,
      "grad_norm": 0.14377017319202423,
      "learning_rate": 8.54184467423901e-05,
      "loss": 1.2102,
      "mean_token_accuracy": 0.7692980572581292,
      "num_tokens": 959757398.0,
      "step": 14210
    },
    {
      "entropy": 0.8112579181790351,
      "epoch": 4.307808831326214,
      "grad_norm": 0.1374746412038803,
      "learning_rate": 8.539486830553845e-05,
      "loss": 1.2086,
      "mean_token_accuracy": 0.7667512550950051,
      "num_tokens": 960439284.0,
      "step": 14220
    },
    {
      "entropy": 0.8109935387969017,
      "epoch": 4.310838445807771,
      "grad_norm": 0.1405801922082901,
      "learning_rate": 8.537127408080488e-05,
      "loss": 1.2115,
      "mean_token_accuracy": 0.7662670150399208,
      "num_tokens": 961120781.0,
      "step": 14230
    },
    {
      "entropy": 0.8141834795475006,
      "epoch": 4.3138680602893285,
      "grad_norm": 0.13471923768520355,
      "learning_rate": 8.534766407871362e-05,
      "loss": 1.2173,
      "mean_token_accuracy": 0.7663844376802444,
      "num_tokens": 961788829.0,
      "step": 14240
    },
    {
      "entropy": 0.8285132646560669,
      "epoch": 4.316897674770885,
      "grad_norm": 0.1346365064382553,
      "learning_rate": 8.532403830979591e-05,
      "loss": 1.2243,
      "mean_token_accuracy": 0.766338886320591,
      "num_tokens": 962482284.0,
      "step": 14250
    },
    {
      "entropy": 0.8150933369994163,
      "epoch": 4.319927289252442,
      "grad_norm": 0.1380923092365265,
      "learning_rate": 8.530039678459001e-05,
      "loss": 1.2102,
      "mean_token_accuracy": 0.7657512426376343,
      "num_tokens": 963166506.0,
      "step": 14260
    },
    {
      "entropy": 0.812044833600521,
      "epoch": 4.322956903734,
      "grad_norm": 0.12815997004508972,
      "learning_rate": 8.527673951364126e-05,
      "loss": 1.2138,
      "mean_token_accuracy": 0.7688316002488136,
      "num_tokens": 963851087.0,
      "step": 14270
    },
    {
      "entropy": 0.8042856082320213,
      "epoch": 4.325986518215557,
      "grad_norm": 0.13710647821426392,
      "learning_rate": 8.525306650750196e-05,
      "loss": 1.2128,
      "mean_token_accuracy": 0.7665573492646217,
      "num_tokens": 964516982.0,
      "step": 14280
    },
    {
      "entropy": 0.8157733410596848,
      "epoch": 4.329016132697114,
      "grad_norm": 0.13905055820941925,
      "learning_rate": 8.522937777673144e-05,
      "loss": 1.2233,
      "mean_token_accuracy": 0.7639014184474945,
      "num_tokens": 965184940.0,
      "step": 14290
    },
    {
      "entropy": 0.8091638401150704,
      "epoch": 4.332045747178672,
      "grad_norm": 0.1665315479040146,
      "learning_rate": 8.520567333189608e-05,
      "loss": 1.2096,
      "mean_token_accuracy": 0.7671777755022049,
      "num_tokens": 965856379.0,
      "step": 14300
    },
    {
      "entropy": 0.827835176885128,
      "epoch": 4.335075361660229,
      "grad_norm": 0.13923880457878113,
      "learning_rate": 8.518195318356925e-05,
      "loss": 1.2216,
      "mean_token_accuracy": 0.7618249371647835,
      "num_tokens": 966539254.0,
      "step": 14310
    },
    {
      "entropy": 0.8134009554982186,
      "epoch": 4.338104976141786,
      "grad_norm": 0.137948676943779,
      "learning_rate": 8.515821734233132e-05,
      "loss": 1.2145,
      "mean_token_accuracy": 0.7653644904494286,
      "num_tokens": 967221257.0,
      "step": 14320
    },
    {
      "entropy": 0.8160528019070625,
      "epoch": 4.341134590623343,
      "grad_norm": 0.14116792380809784,
      "learning_rate": 8.513446581876965e-05,
      "loss": 1.2152,
      "mean_token_accuracy": 0.7628973737359047,
      "num_tokens": 967891935.0,
      "step": 14330
    },
    {
      "entropy": 0.8142945170402527,
      "epoch": 4.3441642051049,
      "grad_norm": 0.1404132843017578,
      "learning_rate": 8.511069862347862e-05,
      "loss": 1.2119,
      "mean_token_accuracy": 0.7677595824003219,
      "num_tokens": 968574266.0,
      "step": 14340
    },
    {
      "entropy": 0.822813107073307,
      "epoch": 4.347193819586457,
      "grad_norm": 0.1303926408290863,
      "learning_rate": 8.508691576705959e-05,
      "loss": 1.2249,
      "mean_token_accuracy": 0.7640576988458634,
      "num_tokens": 969257136.0,
      "step": 14350
    },
    {
      "entropy": 0.819758141040802,
      "epoch": 4.350223434068015,
      "grad_norm": 0.13424734771251678,
      "learning_rate": 8.50631172601209e-05,
      "loss": 1.2139,
      "mean_token_accuracy": 0.7648958772420883,
      "num_tokens": 969940681.0,
      "step": 14360
    },
    {
      "entropy": 0.8146367132663727,
      "epoch": 4.353253048549572,
      "grad_norm": 0.15804411470890045,
      "learning_rate": 8.503930311327788e-05,
      "loss": 1.2141,
      "mean_token_accuracy": 0.7659523740410805,
      "num_tokens": 970617776.0,
      "step": 14370
    },
    {
      "entropy": 0.8044301435351372,
      "epoch": 4.356282663031129,
      "grad_norm": 0.1360851377248764,
      "learning_rate": 8.501547333715281e-05,
      "loss": 1.2156,
      "mean_token_accuracy": 0.7675289556384086,
      "num_tokens": 971284578.0,
      "step": 14380
    },
    {
      "entropy": 0.81045912951231,
      "epoch": 4.359312277512687,
      "grad_norm": 0.1404189020395279,
      "learning_rate": 8.499162794237499e-05,
      "loss": 1.2129,
      "mean_token_accuracy": 0.7676175624132157,
      "num_tokens": 971969518.0,
      "step": 14390
    },
    {
      "entropy": 0.8229447811841964,
      "epoch": 4.362341891994244,
      "grad_norm": 0.14740420877933502,
      "learning_rate": 8.496776693958062e-05,
      "loss": 1.2234,
      "mean_token_accuracy": 0.7622767046093941,
      "num_tokens": 972637831.0,
      "step": 14400
    },
    {
      "entropy": 0.8168444901704788,
      "epoch": 4.365371506475801,
      "grad_norm": 0.1315646916627884,
      "learning_rate": 8.494389033941296e-05,
      "loss": 1.2167,
      "mean_token_accuracy": 0.7641868308186531,
      "num_tokens": 973320160.0,
      "step": 14410
    },
    {
      "entropy": 0.810432693362236,
      "epoch": 4.368401120957358,
      "grad_norm": 0.1290990114212036,
      "learning_rate": 8.491999815252212e-05,
      "loss": 1.2153,
      "mean_token_accuracy": 0.7621484085917473,
      "num_tokens": 973987722.0,
      "step": 14420
    },
    {
      "entropy": 0.8143949195742607,
      "epoch": 4.371430735438915,
      "grad_norm": 0.1370246559381485,
      "learning_rate": 8.489609038956523e-05,
      "loss": 1.2159,
      "mean_token_accuracy": 0.7636998429894447,
      "num_tokens": 974650043.0,
      "step": 14430
    },
    {
      "entropy": 0.8091020226478577,
      "epoch": 4.374460349920472,
      "grad_norm": 0.14625504612922668,
      "learning_rate": 8.487216706120634e-05,
      "loss": 1.2203,
      "mean_token_accuracy": 0.7658385097980499,
      "num_tokens": 975320079.0,
      "step": 14440
    },
    {
      "entropy": 0.8171315059065819,
      "epoch": 4.37748996440203,
      "grad_norm": 0.1365855187177658,
      "learning_rate": 8.484822817811644e-05,
      "loss": 1.2143,
      "mean_token_accuracy": 0.7674818530678749,
      "num_tokens": 976001958.0,
      "step": 14450
    },
    {
      "entropy": 0.8118990361690521,
      "epoch": 4.380519578883587,
      "grad_norm": 0.1415693610906601,
      "learning_rate": 8.482427375097351e-05,
      "loss": 1.2136,
      "mean_token_accuracy": 0.7660479754209518,
      "num_tokens": 976689225.0,
      "step": 14460
    },
    {
      "entropy": 0.814107233285904,
      "epoch": 4.383549193365145,
      "grad_norm": 0.14036540687084198,
      "learning_rate": 8.480030379046238e-05,
      "loss": 1.211,
      "mean_token_accuracy": 0.7664266467094422,
      "num_tokens": 977366619.0,
      "step": 14470
    },
    {
      "entropy": 0.8128906980156898,
      "epoch": 4.386578807846702,
      "grad_norm": 0.14060281217098236,
      "learning_rate": 8.477631830727486e-05,
      "loss": 1.2106,
      "mean_token_accuracy": 0.762708805501461,
      "num_tokens": 978036800.0,
      "step": 14480
    },
    {
      "entropy": 0.8185534238815307,
      "epoch": 4.3896084223282585,
      "grad_norm": 0.142329141497612,
      "learning_rate": 8.475231731210971e-05,
      "loss": 1.2251,
      "mean_token_accuracy": 0.7630772069096565,
      "num_tokens": 978705526.0,
      "step": 14490
    },
    {
      "entropy": 0.813590244948864,
      "epoch": 4.392638036809816,
      "grad_norm": 0.13582834601402283,
      "learning_rate": 8.472830081567253e-05,
      "loss": 1.2102,
      "mean_token_accuracy": 0.7681605830788613,
      "num_tokens": 979384960.0,
      "step": 14500
    },
    {
      "entropy": 0.8131711512804032,
      "epoch": 4.395667651291373,
      "grad_norm": 0.1515778750181198,
      "learning_rate": 8.470426882867591e-05,
      "loss": 1.2193,
      "mean_token_accuracy": 0.7632424727082252,
      "num_tokens": 980053631.0,
      "step": 14510
    },
    {
      "entropy": 0.8127683132886887,
      "epoch": 4.39869726577293,
      "grad_norm": 0.14411474764347076,
      "learning_rate": 8.46802213618393e-05,
      "loss": 1.209,
      "mean_token_accuracy": 0.7665164932608605,
      "num_tokens": 980731299.0,
      "step": 14520
    },
    {
      "entropy": 0.8086910650134087,
      "epoch": 4.401726880254488,
      "grad_norm": 0.13578078150749207,
      "learning_rate": 8.465615842588908e-05,
      "loss": 1.2092,
      "mean_token_accuracy": 0.7675674170255661,
      "num_tokens": 981411298.0,
      "step": 14530
    },
    {
      "entropy": 0.8203605756163597,
      "epoch": 4.404756494736045,
      "grad_norm": 0.1365104466676712,
      "learning_rate": 8.463208003155851e-05,
      "loss": 1.2169,
      "mean_token_accuracy": 0.7646425098180771,
      "num_tokens": 982087097.0,
      "step": 14540
    },
    {
      "entropy": 0.8082945555448532,
      "epoch": 4.407786109217602,
      "grad_norm": 0.13865475356578827,
      "learning_rate": 8.460798618958778e-05,
      "loss": 1.2056,
      "mean_token_accuracy": 0.7731250032782555,
      "num_tokens": 982779112.0,
      "step": 14550
    },
    {
      "entropy": 0.8139575839042663,
      "epoch": 4.41081572369916,
      "grad_norm": 0.14405213296413422,
      "learning_rate": 8.458387691072395e-05,
      "loss": 1.2133,
      "mean_token_accuracy": 0.7653507739305496,
      "num_tokens": 983446998.0,
      "step": 14560
    },
    {
      "entropy": 0.8129200100898742,
      "epoch": 4.413845338180717,
      "grad_norm": 0.1305195540189743,
      "learning_rate": 8.455975220572094e-05,
      "loss": 1.2234,
      "mean_token_accuracy": 0.7651848569512367,
      "num_tokens": 984119489.0,
      "step": 14570
    },
    {
      "entropy": 0.8138880968093872,
      "epoch": 4.4168749526622735,
      "grad_norm": 0.13817481696605682,
      "learning_rate": 8.453561208533958e-05,
      "loss": 1.2174,
      "mean_token_accuracy": 0.7658104941248893,
      "num_tokens": 984799393.0,
      "step": 14580
    },
    {
      "entropy": 0.811996553838253,
      "epoch": 4.419904567143831,
      "grad_norm": 0.14861634373664856,
      "learning_rate": 8.45114565603476e-05,
      "loss": 1.2149,
      "mean_token_accuracy": 0.7660542353987694,
      "num_tokens": 985483715.0,
      "step": 14590
    },
    {
      "entropy": 0.8117525488138199,
      "epoch": 4.422934181625388,
      "grad_norm": 0.13194222748279572,
      "learning_rate": 8.448728564151954e-05,
      "loss": 1.2192,
      "mean_token_accuracy": 0.7686347112059593,
      "num_tokens": 986165074.0,
      "step": 14600
    },
    {
      "entropy": 0.8027610033750534,
      "epoch": 4.425963796106945,
      "grad_norm": 0.14185863733291626,
      "learning_rate": 8.446309933963687e-05,
      "loss": 1.2086,
      "mean_token_accuracy": 0.7731486231088638,
      "num_tokens": 986851050.0,
      "step": 14610
    },
    {
      "entropy": 0.8131787851452827,
      "epoch": 4.428993410588503,
      "grad_norm": 0.1457170844078064,
      "learning_rate": 8.443889766548785e-05,
      "loss": 1.2172,
      "mean_token_accuracy": 0.7673168465495109,
      "num_tokens": 987532410.0,
      "step": 14620
    },
    {
      "entropy": 0.8106110647320748,
      "epoch": 4.43202302507006,
      "grad_norm": 0.1411815732717514,
      "learning_rate": 8.441468062986767e-05,
      "loss": 1.207,
      "mean_token_accuracy": 0.7676076322793961,
      "num_tokens": 988204759.0,
      "step": 14630
    },
    {
      "entropy": 0.8080695107579231,
      "epoch": 4.435052639551617,
      "grad_norm": 0.14832152426242828,
      "learning_rate": 8.439044824357833e-05,
      "loss": 1.2152,
      "mean_token_accuracy": 0.7709108591079712,
      "num_tokens": 988895306.0,
      "step": 14640
    },
    {
      "entropy": 0.8206983223557472,
      "epoch": 4.438082254033175,
      "grad_norm": 0.13751220703125,
      "learning_rate": 8.436620051742867e-05,
      "loss": 1.2135,
      "mean_token_accuracy": 0.7643748372793198,
      "num_tokens": 989588875.0,
      "step": 14650
    },
    {
      "entropy": 0.8055402368307114,
      "epoch": 4.4411118685147315,
      "grad_norm": 0.13688570261001587,
      "learning_rate": 8.434193746223443e-05,
      "loss": 1.2069,
      "mean_token_accuracy": 0.7714090630412102,
      "num_tokens": 990272859.0,
      "step": 14660
    },
    {
      "entropy": 0.8120644554495812,
      "epoch": 4.444141482996288,
      "grad_norm": 0.1294250637292862,
      "learning_rate": 8.431765908881811e-05,
      "loss": 1.2107,
      "mean_token_accuracy": 0.7651125490665436,
      "num_tokens": 990949476.0,
      "step": 14670
    },
    {
      "entropy": 0.8165026217699051,
      "epoch": 4.447171097477846,
      "grad_norm": 0.14319740235805511,
      "learning_rate": 8.429336540800908e-05,
      "loss": 1.2131,
      "mean_token_accuracy": 0.763539296388626,
      "num_tokens": 991613763.0,
      "step": 14680
    },
    {
      "entropy": 0.7998925268650054,
      "epoch": 4.450200711959403,
      "grad_norm": 0.1286431849002838,
      "learning_rate": 8.426905643064355e-05,
      "loss": 1.2086,
      "mean_token_accuracy": 0.7659187391400337,
      "num_tokens": 992281176.0,
      "step": 14690
    },
    {
      "entropy": 0.8212928548455238,
      "epoch": 4.45323032644096,
      "grad_norm": 0.15050755441188812,
      "learning_rate": 8.424473216756456e-05,
      "loss": 1.2217,
      "mean_token_accuracy": 0.7637400344014168,
      "num_tokens": 992957794.0,
      "step": 14700
    },
    {
      "entropy": 0.8027153119444848,
      "epoch": 4.456259940922518,
      "grad_norm": 0.13645991683006287,
      "learning_rate": 8.42203926296219e-05,
      "loss": 1.2167,
      "mean_token_accuracy": 0.768967580795288,
      "num_tokens": 993632473.0,
      "step": 14710
    },
    {
      "entropy": 0.8125939905643463,
      "epoch": 4.459289555404075,
      "grad_norm": 0.13304193317890167,
      "learning_rate": 8.419603782767225e-05,
      "loss": 1.2154,
      "mean_token_accuracy": 0.7635219246149063,
      "num_tokens": 994303273.0,
      "step": 14720
    },
    {
      "entropy": 0.8155429273843765,
      "epoch": 4.462319169885632,
      "grad_norm": 0.1383727341890335,
      "learning_rate": 8.417166777257907e-05,
      "loss": 1.2102,
      "mean_token_accuracy": 0.7683589532971382,
      "num_tokens": 994983943.0,
      "step": 14730
    },
    {
      "entropy": 0.8064889863133431,
      "epoch": 4.4653487843671895,
      "grad_norm": 0.16005361080169678,
      "learning_rate": 8.414728247521262e-05,
      "loss": 1.2068,
      "mean_token_accuracy": 0.7632562965154648,
      "num_tokens": 995649623.0,
      "step": 14740
    },
    {
      "entropy": 0.8158274561166763,
      "epoch": 4.468378398848746,
      "grad_norm": 0.14108243584632874,
      "learning_rate": 8.412288194644997e-05,
      "loss": 1.2134,
      "mean_token_accuracy": 0.7660295411944389,
      "num_tokens": 996323907.0,
      "step": 14750
    },
    {
      "entropy": 0.8111421972513199,
      "epoch": 4.471408013330303,
      "grad_norm": 0.13207165896892548,
      "learning_rate": 8.409846619717496e-05,
      "loss": 1.2112,
      "mean_token_accuracy": 0.7684741467237473,
      "num_tokens": 997013381.0,
      "step": 14760
    },
    {
      "entropy": 0.8017989575862885,
      "epoch": 4.474437627811861,
      "grad_norm": 0.1371523141860962,
      "learning_rate": 8.407403523827823e-05,
      "loss": 1.216,
      "mean_token_accuracy": 0.7698475867509842,
      "num_tokens": 997686186.0,
      "step": 14770
    },
    {
      "entropy": 0.8003453806042671,
      "epoch": 4.477467242293418,
      "grad_norm": 0.13464106619358063,
      "learning_rate": 8.404958908065725e-05,
      "loss": 1.1972,
      "mean_token_accuracy": 0.7718183264136315,
      "num_tokens": 998382987.0,
      "step": 14780
    },
    {
      "entropy": 0.8182407349348069,
      "epoch": 4.480496856774975,
      "grad_norm": 0.13492155075073242,
      "learning_rate": 8.40251277352162e-05,
      "loss": 1.22,
      "mean_token_accuracy": 0.7610447928309441,
      "num_tokens": 999061110.0,
      "step": 14790
    },
    {
      "entropy": 0.8050194546580315,
      "epoch": 4.483526471256533,
      "grad_norm": 0.13588498532772064,
      "learning_rate": 8.400065121286607e-05,
      "loss": 1.2069,
      "mean_token_accuracy": 0.7692304775118828,
      "num_tokens": 999740484.0,
      "step": 14800
    },
    {
      "entropy": 0.8066660225391388,
      "epoch": 4.48655608573809,
      "grad_norm": 0.15822049975395203,
      "learning_rate": 8.39761595245246e-05,
      "loss": 1.2098,
      "mean_token_accuracy": 0.7670860573649406,
      "num_tokens": 1000415961.0,
      "step": 14810
    },
    {
      "entropy": 0.810329258441925,
      "epoch": 4.489585700219647,
      "grad_norm": 0.14247414469718933,
      "learning_rate": 8.395165268111632e-05,
      "loss": 1.2161,
      "mean_token_accuracy": 0.7617546677589416,
      "num_tokens": 1001081174.0,
      "step": 14820
    },
    {
      "entropy": 0.8143522471189499,
      "epoch": 4.4926153147012045,
      "grad_norm": 0.13341271877288818,
      "learning_rate": 8.39271306935725e-05,
      "loss": 1.2172,
      "mean_token_accuracy": 0.766852393746376,
      "num_tokens": 1001764471.0,
      "step": 14830
    },
    {
      "entropy": 0.8150116130709648,
      "epoch": 4.495644929182761,
      "grad_norm": 0.140299990773201,
      "learning_rate": 8.390259357283118e-05,
      "loss": 1.2196,
      "mean_token_accuracy": 0.7648457616567612,
      "num_tokens": 1002436289.0,
      "step": 14840
    },
    {
      "entropy": 0.8112609446048736,
      "epoch": 4.498674543664318,
      "grad_norm": 0.13958598673343658,
      "learning_rate": 8.387804132983716e-05,
      "loss": 1.2127,
      "mean_token_accuracy": 0.7658768385648728,
      "num_tokens": 1003110419.0,
      "step": 14850
    },
    {
      "entropy": 0.8080875903367997,
      "epoch": 4.501704158145876,
      "grad_norm": 0.13154947757720947,
      "learning_rate": 8.385347397554195e-05,
      "loss": 1.2195,
      "mean_token_accuracy": 0.7668353110551834,
      "num_tokens": 1003781756.0,
      "step": 14860
    },
    {
      "entropy": 0.8050442636013031,
      "epoch": 4.504733772627433,
      "grad_norm": 0.1507265269756317,
      "learning_rate": 8.382889152090382e-05,
      "loss": 1.2001,
      "mean_token_accuracy": 0.7674769267439843,
      "num_tokens": 1004456980.0,
      "step": 14870
    },
    {
      "entropy": 0.802332204580307,
      "epoch": 4.507763387108991,
      "grad_norm": 0.13116209208965302,
      "learning_rate": 8.380429397688777e-05,
      "loss": 1.2075,
      "mean_token_accuracy": 0.7701631054282189,
      "num_tokens": 1005139514.0,
      "step": 14880
    },
    {
      "entropy": 0.8175762817263603,
      "epoch": 4.510793001590548,
      "grad_norm": 0.14305265247821808,
      "learning_rate": 8.377968135446555e-05,
      "loss": 1.2113,
      "mean_token_accuracy": 0.7696104750037194,
      "num_tokens": 1005837509.0,
      "step": 14890
    },
    {
      "entropy": 0.8158002674579621,
      "epoch": 4.513822616072105,
      "grad_norm": 0.14322514832019806,
      "learning_rate": 8.37550536646156e-05,
      "loss": 1.2165,
      "mean_token_accuracy": 0.7621502950787544,
      "num_tokens": 1006504325.0,
      "step": 14900
    },
    {
      "entropy": 0.8165830463171005,
      "epoch": 4.516852230553662,
      "grad_norm": 0.13803119957447052,
      "learning_rate": 8.373041091832312e-05,
      "loss": 1.2082,
      "mean_token_accuracy": 0.7644147738814354,
      "num_tokens": 1007183989.0,
      "step": 14910
    },
    {
      "entropy": 0.8164145991206169,
      "epoch": 4.519881845035219,
      "grad_norm": 0.13508456945419312,
      "learning_rate": 8.370575312658e-05,
      "loss": 1.2197,
      "mean_token_accuracy": 0.7663424417376519,
      "num_tokens": 1007865080.0,
      "step": 14920
    },
    {
      "entropy": 0.817217543721199,
      "epoch": 4.522911459516776,
      "grad_norm": 0.13762493431568146,
      "learning_rate": 8.368108030038485e-05,
      "loss": 1.2156,
      "mean_token_accuracy": 0.7620110109448432,
      "num_tokens": 1008535404.0,
      "step": 14930
    },
    {
      "entropy": 0.8187590450048446,
      "epoch": 4.525941073998334,
      "grad_norm": 0.1280880570411682,
      "learning_rate": 8.365639245074298e-05,
      "loss": 1.2058,
      "mean_token_accuracy": 0.7672760635614395,
      "num_tokens": 1009224753.0,
      "step": 14940
    },
    {
      "entropy": 0.8178510829806328,
      "epoch": 4.528970688479891,
      "grad_norm": 0.1395222544670105,
      "learning_rate": 8.363168958866642e-05,
      "loss": 1.2156,
      "mean_token_accuracy": 0.7642698660492897,
      "num_tokens": 1009898856.0,
      "step": 14950
    },
    {
      "entropy": 0.8200342208147049,
      "epoch": 4.532000302961448,
      "grad_norm": 0.14817102253437042,
      "learning_rate": 8.360697172517386e-05,
      "loss": 1.2233,
      "mean_token_accuracy": 0.7670670941472053,
      "num_tokens": 1010580410.0,
      "step": 14960
    },
    {
      "entropy": 0.8099397957324982,
      "epoch": 4.535029917443006,
      "grad_norm": 0.13535571098327637,
      "learning_rate": 8.358223887129074e-05,
      "loss": 1.2117,
      "mean_token_accuracy": 0.7699789434671402,
      "num_tokens": 1011272061.0,
      "step": 14970
    },
    {
      "entropy": 0.8149744793772697,
      "epoch": 4.538059531924563,
      "grad_norm": 0.13294681906700134,
      "learning_rate": 8.35574910380491e-05,
      "loss": 1.2198,
      "mean_token_accuracy": 0.7701617524027824,
      "num_tokens": 1011972676.0,
      "step": 14980
    },
    {
      "entropy": 0.8180561020970345,
      "epoch": 4.54108914640612,
      "grad_norm": 0.1370512843132019,
      "learning_rate": 8.353272823648774e-05,
      "loss": 1.2169,
      "mean_token_accuracy": 0.7560781911015511,
      "num_tokens": 1012625462.0,
      "step": 14990
    },
    {
      "entropy": 0.8148312240839004,
      "epoch": 4.544118760887677,
      "grad_norm": 0.13839320838451385,
      "learning_rate": 8.350795047765212e-05,
      "loss": 1.2109,
      "mean_token_accuracy": 0.7686563178896904,
      "num_tokens": 1013318928.0,
      "step": 15000
    },
    {
      "entropy": 0.8100945860147476,
      "epoch": 4.547148375369234,
      "grad_norm": 0.14533695578575134,
      "learning_rate": 8.348315777259433e-05,
      "loss": 1.2244,
      "mean_token_accuracy": 0.762176875770092,
      "num_tokens": 1013973728.0,
      "step": 15010
    },
    {
      "entropy": 0.8118842095136642,
      "epoch": 4.550177989850791,
      "grad_norm": 0.138160839676857,
      "learning_rate": 8.345835013237318e-05,
      "loss": 1.214,
      "mean_token_accuracy": 0.7652218982577323,
      "num_tokens": 1014652356.0,
      "step": 15020
    },
    {
      "entropy": 0.8077614337205887,
      "epoch": 4.553207604332349,
      "grad_norm": 0.13375826179981232,
      "learning_rate": 8.343352756805413e-05,
      "loss": 1.2085,
      "mean_token_accuracy": 0.7711880415678024,
      "num_tokens": 1015334978.0,
      "step": 15030
    },
    {
      "entropy": 0.8138082325458527,
      "epoch": 4.556237218813906,
      "grad_norm": 0.14000007510185242,
      "learning_rate": 8.340869009070924e-05,
      "loss": 1.224,
      "mean_token_accuracy": 0.7669972136616707,
      "num_tokens": 1016006651.0,
      "step": 15040
    },
    {
      "entropy": 0.8117908909916878,
      "epoch": 4.559266833295463,
      "grad_norm": 0.13255485892295837,
      "learning_rate": 8.338383771141731e-05,
      "loss": 1.2188,
      "mean_token_accuracy": 0.762118911743164,
      "num_tokens": 1016668648.0,
      "step": 15050
    },
    {
      "entropy": 0.8168422803282738,
      "epoch": 4.562296447777021,
      "grad_norm": 0.13548730313777924,
      "learning_rate": 8.335897044126372e-05,
      "loss": 1.2194,
      "mean_token_accuracy": 0.7666918858885765,
      "num_tokens": 1017347166.0,
      "step": 15060
    },
    {
      "entropy": 0.814100868999958,
      "epoch": 4.565326062258578,
      "grad_norm": 0.14705607295036316,
      "learning_rate": 8.333408829134055e-05,
      "loss": 1.2156,
      "mean_token_accuracy": 0.7637822508811951,
      "num_tokens": 1018016924.0,
      "step": 15070
    },
    {
      "entropy": 0.8101621180772781,
      "epoch": 4.5683556767401345,
      "grad_norm": 0.139455646276474,
      "learning_rate": 8.330919127274643e-05,
      "loss": 1.2143,
      "mean_token_accuracy": 0.7600699231028557,
      "num_tokens": 1018678644.0,
      "step": 15080
    },
    {
      "entropy": 0.8048439994454384,
      "epoch": 4.571385291221692,
      "grad_norm": 0.14077451825141907,
      "learning_rate": 8.328427939658672e-05,
      "loss": 1.2094,
      "mean_token_accuracy": 0.7688601195812226,
      "num_tokens": 1019365627.0,
      "step": 15090
    },
    {
      "entropy": 0.8014249309897423,
      "epoch": 4.574414905703249,
      "grad_norm": 0.13245782256126404,
      "learning_rate": 8.325935267397333e-05,
      "loss": 1.1945,
      "mean_token_accuracy": 0.7756657838821411,
      "num_tokens": 1020066102.0,
      "step": 15100
    },
    {
      "entropy": 0.8136430740356445,
      "epoch": 4.577444520184806,
      "grad_norm": 0.14787311851978302,
      "learning_rate": 8.323441111602486e-05,
      "loss": 1.2117,
      "mean_token_accuracy": 0.7652172610163689,
      "num_tokens": 1020741192.0,
      "step": 15110
    },
    {
      "entropy": 0.8075628504157066,
      "epoch": 4.580474134666364,
      "grad_norm": 0.13490545749664307,
      "learning_rate": 8.320945473386647e-05,
      "loss": 1.2085,
      "mean_token_accuracy": 0.7699533253908157,
      "num_tokens": 1021434348.0,
      "step": 15120
    },
    {
      "entropy": 0.8200042188167572,
      "epoch": 4.583503749147921,
      "grad_norm": 0.1354069709777832,
      "learning_rate": 8.318448353862995e-05,
      "loss": 1.2175,
      "mean_token_accuracy": 0.7627965018153191,
      "num_tokens": 1022095966.0,
      "step": 15130
    },
    {
      "entropy": 0.8047011941671371,
      "epoch": 4.586533363629478,
      "grad_norm": 0.13225625455379486,
      "learning_rate": 8.315949754145372e-05,
      "loss": 1.2077,
      "mean_token_accuracy": 0.7679173469543457,
      "num_tokens": 1022772442.0,
      "step": 15140
    },
    {
      "entropy": 0.8104792013764381,
      "epoch": 4.589562978111036,
      "grad_norm": 0.13480982184410095,
      "learning_rate": 8.313449675348278e-05,
      "loss": 1.2048,
      "mean_token_accuracy": 0.76620062738657,
      "num_tokens": 1023443677.0,
      "step": 15150
    },
    {
      "entropy": 0.8011610880494118,
      "epoch": 4.592592592592593,
      "grad_norm": 0.13761554658412933,
      "learning_rate": 8.310948118586872e-05,
      "loss": 1.2054,
      "mean_token_accuracy": 0.7683380633592606,
      "num_tokens": 1024113965.0,
      "step": 15160
    },
    {
      "entropy": 0.8223711282014847,
      "epoch": 4.5956222070741495,
      "grad_norm": 0.1307128667831421,
      "learning_rate": 8.308445084976977e-05,
      "loss": 1.2176,
      "mean_token_accuracy": 0.7671125277876853,
      "num_tokens": 1024789279.0,
      "step": 15170
    },
    {
      "entropy": 0.7965401828289032,
      "epoch": 4.598651821555707,
      "grad_norm": 0.13591811060905457,
      "learning_rate": 8.305940575635064e-05,
      "loss": 1.2049,
      "mean_token_accuracy": 0.7685336083173752,
      "num_tokens": 1025462406.0,
      "step": 15180
    },
    {
      "entropy": 0.8169444009661675,
      "epoch": 4.601681436037264,
      "grad_norm": 0.13895291090011597,
      "learning_rate": 8.303434591678279e-05,
      "loss": 1.2171,
      "mean_token_accuracy": 0.7644126906991004,
      "num_tokens": 1026134747.0,
      "step": 15190
    },
    {
      "entropy": 0.8088120698928833,
      "epoch": 4.604711050518821,
      "grad_norm": 0.13839735090732574,
      "learning_rate": 8.300927134224409e-05,
      "loss": 1.2119,
      "mean_token_accuracy": 0.7721304342150688,
      "num_tokens": 1026824488.0,
      "step": 15200
    },
    {
      "entropy": 0.8069801807403565,
      "epoch": 4.607740665000379,
      "grad_norm": 0.14297841489315033,
      "learning_rate": 8.298418204391907e-05,
      "loss": 1.2093,
      "mean_token_accuracy": 0.7674991518259049,
      "num_tokens": 1027506067.0,
      "step": 15210
    },
    {
      "entropy": 0.8225396156311036,
      "epoch": 4.610770279481936,
      "grad_norm": 0.1426960974931717,
      "learning_rate": 8.295907803299883e-05,
      "loss": 1.2168,
      "mean_token_accuracy": 0.7645534157752991,
      "num_tokens": 1028192598.0,
      "step": 15220
    },
    {
      "entropy": 0.8128435805439949,
      "epoch": 4.613799893963493,
      "grad_norm": 0.1445625275373459,
      "learning_rate": 8.293395932068101e-05,
      "loss": 1.2113,
      "mean_token_accuracy": 0.7692129746079445,
      "num_tokens": 1028883590.0,
      "step": 15230
    },
    {
      "entropy": 0.8124251589179039,
      "epoch": 4.616829508445051,
      "grad_norm": 0.14045129716396332,
      "learning_rate": 8.290882591816979e-05,
      "loss": 1.2158,
      "mean_token_accuracy": 0.7653728276491165,
      "num_tokens": 1029554945.0,
      "step": 15240
    },
    {
      "entropy": 0.8150558784604073,
      "epoch": 4.6198591229266075,
      "grad_norm": 0.15508978068828583,
      "learning_rate": 8.288367783667596e-05,
      "loss": 1.2166,
      "mean_token_accuracy": 0.7606960788369179,
      "num_tokens": 1030212555.0,
      "step": 15250
    },
    {
      "entropy": 0.8126764863729476,
      "epoch": 4.622888737408164,
      "grad_norm": 0.14180020987987518,
      "learning_rate": 8.285851508741678e-05,
      "loss": 1.2126,
      "mean_token_accuracy": 0.7690615206956863,
      "num_tokens": 1030891731.0,
      "step": 15260
    },
    {
      "entropy": 0.8112331315875053,
      "epoch": 4.625918351889722,
      "grad_norm": 0.13965746760368347,
      "learning_rate": 8.283333768161615e-05,
      "loss": 1.2261,
      "mean_token_accuracy": 0.762207692861557,
      "num_tokens": 1031547425.0,
      "step": 15270
    },
    {
      "entropy": 0.8134624063968658,
      "epoch": 4.628947966371279,
      "grad_norm": 0.13768364489078522,
      "learning_rate": 8.280814563050438e-05,
      "loss": 1.2249,
      "mean_token_accuracy": 0.7601321175694465,
      "num_tokens": 1032206076.0,
      "step": 15280
    },
    {
      "entropy": 0.8125778660178185,
      "epoch": 4.631977580852837,
      "grad_norm": 0.14565123617649078,
      "learning_rate": 8.278293894531845e-05,
      "loss": 1.2119,
      "mean_token_accuracy": 0.7681752011179924,
      "num_tokens": 1032888534.0,
      "step": 15290
    },
    {
      "entropy": 0.8162658482789993,
      "epoch": 4.635007195334394,
      "grad_norm": 0.14811156690120697,
      "learning_rate": 8.275771763730178e-05,
      "loss": 1.2153,
      "mean_token_accuracy": 0.7628666043281556,
      "num_tokens": 1033563510.0,
      "step": 15300
    },
    {
      "entropy": 0.8310571193695069,
      "epoch": 4.638036809815951,
      "grad_norm": 0.14319728314876556,
      "learning_rate": 8.273248171770433e-05,
      "loss": 1.2266,
      "mean_token_accuracy": 0.7628302797675133,
      "num_tokens": 1034248909.0,
      "step": 15310
    },
    {
      "entropy": 0.8047871500253677,
      "epoch": 4.641066424297508,
      "grad_norm": 0.13362020254135132,
      "learning_rate": 8.270723119778259e-05,
      "loss": 1.2134,
      "mean_token_accuracy": 0.7651651829481125,
      "num_tokens": 1034914348.0,
      "step": 15320
    },
    {
      "entropy": 0.8066971078515053,
      "epoch": 4.6440960387790655,
      "grad_norm": 0.1409444361925125,
      "learning_rate": 8.268196608879954e-05,
      "loss": 1.1975,
      "mean_token_accuracy": 0.7695615276694298,
      "num_tokens": 1035599061.0,
      "step": 15330
    },
    {
      "entropy": 0.8144468575716018,
      "epoch": 4.647125653260622,
      "grad_norm": 0.13992279767990112,
      "learning_rate": 8.265668640202471e-05,
      "loss": 1.2152,
      "mean_token_accuracy": 0.7671654686331749,
      "num_tokens": 1036283892.0,
      "step": 15340
    },
    {
      "entropy": 0.8210369318723678,
      "epoch": 4.65015526774218,
      "grad_norm": 0.1452576220035553,
      "learning_rate": 8.263139214873406e-05,
      "loss": 1.2228,
      "mean_token_accuracy": 0.7624813586473465,
      "num_tokens": 1036952450.0,
      "step": 15350
    },
    {
      "entropy": 0.8076359361410141,
      "epoch": 4.653184882223737,
      "grad_norm": 0.15487810969352722,
      "learning_rate": 8.260608334021016e-05,
      "loss": 1.2064,
      "mean_token_accuracy": 0.768352934718132,
      "num_tokens": 1037632946.0,
      "step": 15360
    },
    {
      "entropy": 0.8118151858448982,
      "epoch": 4.656214496705294,
      "grad_norm": 0.13978509604930878,
      "learning_rate": 8.258075998774193e-05,
      "loss": 1.2127,
      "mean_token_accuracy": 0.7646442472934722,
      "num_tokens": 1038295444.0,
      "step": 15370
    },
    {
      "entropy": 0.8052182763814926,
      "epoch": 4.659244111186851,
      "grad_norm": 0.14350928366184235,
      "learning_rate": 8.25554221026249e-05,
      "loss": 1.2111,
      "mean_token_accuracy": 0.7669170662760735,
      "num_tokens": 1038968930.0,
      "step": 15380
    },
    {
      "entropy": 0.8079338103532792,
      "epoch": 4.662273725668409,
      "grad_norm": 0.14178936183452606,
      "learning_rate": 8.253006969616104e-05,
      "loss": 1.2074,
      "mean_token_accuracy": 0.7662130638957023,
      "num_tokens": 1039646283.0,
      "step": 15390
    },
    {
      "entropy": 0.8065927296876907,
      "epoch": 4.665303340149966,
      "grad_norm": 0.14469890296459198,
      "learning_rate": 8.250470277965876e-05,
      "loss": 1.2163,
      "mean_token_accuracy": 0.7637876242399215,
      "num_tokens": 1040309806.0,
      "step": 15400
    },
    {
      "entropy": 0.8208244100213051,
      "epoch": 4.6683329546315235,
      "grad_norm": 0.1422407031059265,
      "learning_rate": 8.247932136443298e-05,
      "loss": 1.2144,
      "mean_token_accuracy": 0.763920234143734,
      "num_tokens": 1040980401.0,
      "step": 15410
    },
    {
      "entropy": 0.8125808432698249,
      "epoch": 4.6713625691130805,
      "grad_norm": 0.14303629100322723,
      "learning_rate": 8.24539254618051e-05,
      "loss": 1.2171,
      "mean_token_accuracy": 0.7659008145332337,
      "num_tokens": 1041651725.0,
      "step": 15420
    },
    {
      "entropy": 0.8139560624957085,
      "epoch": 4.674392183594637,
      "grad_norm": 0.13927346467971802,
      "learning_rate": 8.242851508310294e-05,
      "loss": 1.2167,
      "mean_token_accuracy": 0.7657284542918206,
      "num_tokens": 1042328003.0,
      "step": 15430
    },
    {
      "entropy": 0.8144581004977226,
      "epoch": 4.677421798076195,
      "grad_norm": 0.14626960456371307,
      "learning_rate": 8.240309023966081e-05,
      "loss": 1.2151,
      "mean_token_accuracy": 0.7684321507811547,
      "num_tokens": 1043023342.0,
      "step": 15440
    },
    {
      "entropy": 0.8181347772479057,
      "epoch": 4.680451412557752,
      "grad_norm": 0.13941973447799683,
      "learning_rate": 8.237765094281948e-05,
      "loss": 1.2173,
      "mean_token_accuracy": 0.7630537569522857,
      "num_tokens": 1043689394.0,
      "step": 15450
    },
    {
      "entropy": 0.809430742263794,
      "epoch": 4.683481027039309,
      "grad_norm": 0.14009128510951996,
      "learning_rate": 8.235219720392611e-05,
      "loss": 1.22,
      "mean_token_accuracy": 0.7659080073237419,
      "num_tokens": 1044355961.0,
      "step": 15460
    },
    {
      "entropy": 0.8158665135502815,
      "epoch": 4.686510641520867,
      "grad_norm": 0.1422731578350067,
      "learning_rate": 8.232672903433438e-05,
      "loss": 1.2192,
      "mean_token_accuracy": 0.7604943737387657,
      "num_tokens": 1045016913.0,
      "step": 15470
    },
    {
      "entropy": 0.816311040520668,
      "epoch": 4.689540256002424,
      "grad_norm": 0.12189248949289322,
      "learning_rate": 8.230124644540434e-05,
      "loss": 1.2149,
      "mean_token_accuracy": 0.7640485346317292,
      "num_tokens": 1045695457.0,
      "step": 15480
    },
    {
      "entropy": 0.8113370686769485,
      "epoch": 4.692569870483981,
      "grad_norm": 0.13859044015407562,
      "learning_rate": 8.227574944850252e-05,
      "loss": 1.2117,
      "mean_token_accuracy": 0.7680921629071236,
      "num_tokens": 1046381557.0,
      "step": 15490
    },
    {
      "entropy": 0.8088567674160003,
      "epoch": 4.6955994849655385,
      "grad_norm": 0.14636880159378052,
      "learning_rate": 8.225023805500184e-05,
      "loss": 1.2158,
      "mean_token_accuracy": 0.765949147939682,
      "num_tokens": 1047053551.0,
      "step": 15500
    },
    {
      "entropy": 0.8066520035266876,
      "epoch": 4.698629099447095,
      "grad_norm": 0.13433587551116943,
      "learning_rate": 8.222471227628166e-05,
      "loss": 1.2054,
      "mean_token_accuracy": 0.767923741042614,
      "num_tokens": 1047729798.0,
      "step": 15510
    },
    {
      "entropy": 0.8008198201656341,
      "epoch": 4.701658713928652,
      "grad_norm": 0.13879390060901642,
      "learning_rate": 8.219917212372776e-05,
      "loss": 1.2056,
      "mean_token_accuracy": 0.7654800340533257,
      "num_tokens": 1048386762.0,
      "step": 15520
    },
    {
      "entropy": 0.8036904603242874,
      "epoch": 4.70468832841021,
      "grad_norm": 0.14773288369178772,
      "learning_rate": 8.217361760873233e-05,
      "loss": 1.2092,
      "mean_token_accuracy": 0.7680824026465416,
      "num_tokens": 1049058279.0,
      "step": 15530
    },
    {
      "entropy": 0.8194151535630226,
      "epoch": 4.707717942891767,
      "grad_norm": 0.14728638529777527,
      "learning_rate": 8.214804874269395e-05,
      "loss": 1.2201,
      "mean_token_accuracy": 0.7612240836024284,
      "num_tokens": 1049721734.0,
      "step": 15540
    },
    {
      "entropy": 0.8121907562017441,
      "epoch": 4.710747557373324,
      "grad_norm": 0.14110548794269562,
      "learning_rate": 8.212246553701764e-05,
      "loss": 1.2169,
      "mean_token_accuracy": 0.7651133015751839,
      "num_tokens": 1050385605.0,
      "step": 15550
    },
    {
      "entropy": 0.814926028251648,
      "epoch": 4.713777171854882,
      "grad_norm": 0.13291309773921967,
      "learning_rate": 8.209686800311473e-05,
      "loss": 1.2149,
      "mean_token_accuracy": 0.763114994764328,
      "num_tokens": 1051072736.0,
      "step": 15560
    },
    {
      "entropy": 0.8138626217842102,
      "epoch": 4.716806786336439,
      "grad_norm": 0.12934871017932892,
      "learning_rate": 8.207125615240307e-05,
      "loss": 1.2142,
      "mean_token_accuracy": 0.769535356760025,
      "num_tokens": 1051764345.0,
      "step": 15570
    },
    {
      "entropy": 0.8107445687055588,
      "epoch": 4.719836400817996,
      "grad_norm": 0.13547451794147491,
      "learning_rate": 8.20456299963068e-05,
      "loss": 1.2107,
      "mean_token_accuracy": 0.764206412434578,
      "num_tokens": 1052431484.0,
      "step": 15580
    },
    {
      "entropy": 0.8193677708506584,
      "epoch": 4.722866015299553,
      "grad_norm": 0.1325022578239441,
      "learning_rate": 8.201998954625645e-05,
      "loss": 1.2149,
      "mean_token_accuracy": 0.7634567454457283,
      "num_tokens": 1053100447.0,
      "step": 15590
    },
    {
      "entropy": 0.8121628552675247,
      "epoch": 4.72589562978111,
      "grad_norm": 0.1375740021467209,
      "learning_rate": 8.199433481368895e-05,
      "loss": 1.2052,
      "mean_token_accuracy": 0.769801427423954,
      "num_tokens": 1053790415.0,
      "step": 15600
    },
    {
      "entropy": 0.8133607849478721,
      "epoch": 4.728925244262667,
      "grad_norm": 0.14360255002975464,
      "learning_rate": 8.196866581004761e-05,
      "loss": 1.2173,
      "mean_token_accuracy": 0.7652547299861908,
      "num_tokens": 1054454432.0,
      "step": 15610
    },
    {
      "entropy": 0.8189873456954956,
      "epoch": 4.731954858744225,
      "grad_norm": 0.1521763801574707,
      "learning_rate": 8.194298254678208e-05,
      "loss": 1.2289,
      "mean_token_accuracy": 0.761813010275364,
      "num_tokens": 1055127197.0,
      "step": 15620
    },
    {
      "entropy": 0.8223283037543296,
      "epoch": 4.734984473225782,
      "grad_norm": 0.13551150262355804,
      "learning_rate": 8.191728503534839e-05,
      "loss": 1.2177,
      "mean_token_accuracy": 0.7663288027048111,
      "num_tokens": 1055815660.0,
      "step": 15630
    },
    {
      "entropy": 0.8093197181820869,
      "epoch": 4.738014087707339,
      "grad_norm": 0.13915053009986877,
      "learning_rate": 8.189157328720889e-05,
      "loss": 1.2076,
      "mean_token_accuracy": 0.7658779978752136,
      "num_tokens": 1056486953.0,
      "step": 15640
    },
    {
      "entropy": 0.8062524825334549,
      "epoch": 4.741043702188897,
      "grad_norm": 0.13573576509952545,
      "learning_rate": 8.186584731383233e-05,
      "loss": 1.208,
      "mean_token_accuracy": 0.770762488245964,
      "num_tokens": 1057173002.0,
      "step": 15650
    },
    {
      "entropy": 0.8034259334206582,
      "epoch": 4.744073316670454,
      "grad_norm": 0.14011842012405396,
      "learning_rate": 8.184010712669378e-05,
      "loss": 1.2152,
      "mean_token_accuracy": 0.7674999088048935,
      "num_tokens": 1057840865.0,
      "step": 15660
    },
    {
      "entropy": 0.8159258872270584,
      "epoch": 4.7471029311520105,
      "grad_norm": 0.15081876516342163,
      "learning_rate": 8.181435273727464e-05,
      "loss": 1.2176,
      "mean_token_accuracy": 0.7630300164222718,
      "num_tokens": 1058499702.0,
      "step": 15670
    },
    {
      "entropy": 0.8117739945650101,
      "epoch": 4.750132545633568,
      "grad_norm": 0.1396917849779129,
      "learning_rate": 8.178858415706266e-05,
      "loss": 1.2143,
      "mean_token_accuracy": 0.7646714493632316,
      "num_tokens": 1059172595.0,
      "step": 15680
    },
    {
      "entropy": 0.8117744654417038,
      "epoch": 4.753162160115125,
      "grad_norm": 0.13941623270511627,
      "learning_rate": 8.176280139755191e-05,
      "loss": 1.2087,
      "mean_token_accuracy": 0.764184258878231,
      "num_tokens": 1059845063.0,
      "step": 15690
    },
    {
      "entropy": 0.8059625402092934,
      "epoch": 4.756191774596682,
      "grad_norm": 0.140048086643219,
      "learning_rate": 8.17370044702428e-05,
      "loss": 1.2122,
      "mean_token_accuracy": 0.7721580326557159,
      "num_tokens": 1060530286.0,
      "step": 15700
    },
    {
      "entropy": 0.8040213257074356,
      "epoch": 4.75922138907824,
      "grad_norm": 0.1433405876159668,
      "learning_rate": 8.171119338664206e-05,
      "loss": 1.2035,
      "mean_token_accuracy": 0.7708293154835701,
      "num_tokens": 1061220258.0,
      "step": 15710
    },
    {
      "entropy": 0.8147042289376258,
      "epoch": 4.762251003559797,
      "grad_norm": 0.1303349882364273,
      "learning_rate": 8.168536815826271e-05,
      "loss": 1.2179,
      "mean_token_accuracy": 0.7646678447723388,
      "num_tokens": 1061898600.0,
      "step": 15720
    },
    {
      "entropy": 0.8121464133262635,
      "epoch": 4.765280618041354,
      "grad_norm": 0.13437813520431519,
      "learning_rate": 8.165952879662408e-05,
      "loss": 1.2082,
      "mean_token_accuracy": 0.7666034877300263,
      "num_tokens": 1062586608.0,
      "step": 15730
    },
    {
      "entropy": 0.8070258751511574,
      "epoch": 4.768310232522912,
      "grad_norm": 0.13249161839485168,
      "learning_rate": 8.163367531325186e-05,
      "loss": 1.2061,
      "mean_token_accuracy": 0.7662536963820458,
      "num_tokens": 1063278022.0,
      "step": 15740
    },
    {
      "entropy": 0.8094275057315826,
      "epoch": 4.771339847004469,
      "grad_norm": 0.1256585270166397,
      "learning_rate": 8.160780771967796e-05,
      "loss": 1.2113,
      "mean_token_accuracy": 0.7595454663038254,
      "num_tokens": 1063935349.0,
      "step": 15750
    },
    {
      "entropy": 0.8062768653035164,
      "epoch": 4.774369461486026,
      "grad_norm": 0.13668043911457062,
      "learning_rate": 8.158192602744062e-05,
      "loss": 1.2145,
      "mean_token_accuracy": 0.7655957818031311,
      "num_tokens": 1064602766.0,
      "step": 15760
    },
    {
      "entropy": 0.8141536340117455,
      "epoch": 4.777399075967583,
      "grad_norm": 0.1468982696533203,
      "learning_rate": 8.155603024808438e-05,
      "loss": 1.2121,
      "mean_token_accuracy": 0.7617022007703781,
      "num_tokens": 1065259235.0,
      "step": 15770
    },
    {
      "entropy": 0.8113908216357231,
      "epoch": 4.78042869044914,
      "grad_norm": 0.13862980902194977,
      "learning_rate": 8.153012039316005e-05,
      "loss": 1.2126,
      "mean_token_accuracy": 0.7633839502930642,
      "num_tokens": 1065922337.0,
      "step": 15780
    },
    {
      "entropy": 0.8185302555561066,
      "epoch": 4.783458304930697,
      "grad_norm": 0.14436276257038116,
      "learning_rate": 8.150419647422473e-05,
      "loss": 1.2222,
      "mean_token_accuracy": 0.7613570883870124,
      "num_tokens": 1066581990.0,
      "step": 15790
    },
    {
      "entropy": 0.8092867791652679,
      "epoch": 4.786487919412255,
      "grad_norm": 0.13496606051921844,
      "learning_rate": 8.14782585028418e-05,
      "loss": 1.2158,
      "mean_token_accuracy": 0.7661177188158035,
      "num_tokens": 1067256000.0,
      "step": 15800
    },
    {
      "entropy": 0.7992815613746643,
      "epoch": 4.789517533893812,
      "grad_norm": 0.14180412888526917,
      "learning_rate": 8.145230649058085e-05,
      "loss": 1.2011,
      "mean_token_accuracy": 0.7731818363070488,
      "num_tokens": 1067944433.0,
      "step": 15810
    },
    {
      "entropy": 0.8109717398881913,
      "epoch": 4.79254714837537,
      "grad_norm": 0.13364098966121674,
      "learning_rate": 8.14263404490178e-05,
      "loss": 1.2052,
      "mean_token_accuracy": 0.7660579293966293,
      "num_tokens": 1068615734.0,
      "step": 15820
    },
    {
      "entropy": 0.8127626776695251,
      "epoch": 4.795576762856927,
      "grad_norm": 0.14590229094028473,
      "learning_rate": 8.140036038973482e-05,
      "loss": 1.2074,
      "mean_token_accuracy": 0.7680396258831024,
      "num_tokens": 1069300896.0,
      "step": 15830
    },
    {
      "entropy": 0.8128359496593476,
      "epoch": 4.7986063773384835,
      "grad_norm": 0.15021905303001404,
      "learning_rate": 8.137436632432027e-05,
      "loss": 1.2202,
      "mean_token_accuracy": 0.7599052727222443,
      "num_tokens": 1069962948.0,
      "step": 15840
    },
    {
      "entropy": 0.8056706175208092,
      "epoch": 4.80163599182004,
      "grad_norm": 0.1618979126214981,
      "learning_rate": 8.134835826436886e-05,
      "loss": 1.2154,
      "mean_token_accuracy": 0.7665325850248337,
      "num_tokens": 1070628681.0,
      "step": 15850
    },
    {
      "entropy": 0.8111644566059113,
      "epoch": 4.804665606301598,
      "grad_norm": 0.13437557220458984,
      "learning_rate": 8.132233622148145e-05,
      "loss": 1.2176,
      "mean_token_accuracy": 0.7641709834337235,
      "num_tokens": 1071291345.0,
      "step": 15860
    },
    {
      "entropy": 0.8157501295208931,
      "epoch": 4.807695220783155,
      "grad_norm": 0.13300754129886627,
      "learning_rate": 8.129630020726516e-05,
      "loss": 1.2182,
      "mean_token_accuracy": 0.7630707338452339,
      "num_tokens": 1071959301.0,
      "step": 15870
    },
    {
      "entropy": 0.8204719394445419,
      "epoch": 4.810724835264713,
      "grad_norm": 0.1367742419242859,
      "learning_rate": 8.127025023333339e-05,
      "loss": 1.2218,
      "mean_token_accuracy": 0.7607513844966889,
      "num_tokens": 1072630676.0,
      "step": 15880
    },
    {
      "entropy": 0.8045290946960449,
      "epoch": 4.81375444974627,
      "grad_norm": 0.1460738480091095,
      "learning_rate": 8.124418631130572e-05,
      "loss": 1.202,
      "mean_token_accuracy": 0.7666138365864754,
      "num_tokens": 1073309078.0,
      "step": 15890
    },
    {
      "entropy": 0.8093014433979988,
      "epoch": 4.816784064227827,
      "grad_norm": 0.14502641558647156,
      "learning_rate": 8.121810845280795e-05,
      "loss": 1.2075,
      "mean_token_accuracy": 0.7641166388988495,
      "num_tokens": 1073976360.0,
      "step": 15900
    },
    {
      "entropy": 0.8139350295066834,
      "epoch": 4.819813678709385,
      "grad_norm": 0.13717105984687805,
      "learning_rate": 8.119201666947212e-05,
      "loss": 1.2153,
      "mean_token_accuracy": 0.7647068202495575,
      "num_tokens": 1074654333.0,
      "step": 15910
    },
    {
      "entropy": 0.800288000702858,
      "epoch": 4.8228432931909415,
      "grad_norm": 0.13237711787223816,
      "learning_rate": 8.116591097293649e-05,
      "loss": 1.2013,
      "mean_token_accuracy": 0.7709699377417565,
      "num_tokens": 1075345241.0,
      "step": 15920
    },
    {
      "entropy": 0.8054830610752106,
      "epoch": 4.825872907672498,
      "grad_norm": 0.1496029943227768,
      "learning_rate": 8.113979137484547e-05,
      "loss": 1.2128,
      "mean_token_accuracy": 0.7653148591518402,
      "num_tokens": 1076010481.0,
      "step": 15930
    },
    {
      "entropy": 0.8229808688163758,
      "epoch": 4.828902522154056,
      "grad_norm": 0.13943567872047424,
      "learning_rate": 8.111365788684973e-05,
      "loss": 1.2253,
      "mean_token_accuracy": 0.7621565163135529,
      "num_tokens": 1076684694.0,
      "step": 15940
    },
    {
      "entropy": 0.8060909911990166,
      "epoch": 4.831932136635613,
      "grad_norm": 0.14224740862846375,
      "learning_rate": 8.108751052060609e-05,
      "loss": 1.1991,
      "mean_token_accuracy": 0.7679936498403549,
      "num_tokens": 1077364793.0,
      "step": 15950
    },
    {
      "entropy": 0.8114812031388283,
      "epoch": 4.83496175111717,
      "grad_norm": 0.14297319948673248,
      "learning_rate": 8.106134928777762e-05,
      "loss": 1.2151,
      "mean_token_accuracy": 0.7653910368680954,
      "num_tokens": 1078046079.0,
      "step": 15960
    },
    {
      "entropy": 0.8102113500237464,
      "epoch": 4.837991365598728,
      "grad_norm": 0.13165238499641418,
      "learning_rate": 8.103517420003351e-05,
      "loss": 1.2126,
      "mean_token_accuracy": 0.7717357322573661,
      "num_tokens": 1078737750.0,
      "step": 15970
    },
    {
      "entropy": 0.8052035465836525,
      "epoch": 4.841020980080285,
      "grad_norm": 0.14232152700424194,
      "learning_rate": 8.100898526904916e-05,
      "loss": 1.1977,
      "mean_token_accuracy": 0.7688471049070358,
      "num_tokens": 1079427959.0,
      "step": 15980
    },
    {
      "entropy": 0.8029644057154656,
      "epoch": 4.844050594561842,
      "grad_norm": 0.13276511430740356,
      "learning_rate": 8.098278250650616e-05,
      "loss": 1.2066,
      "mean_token_accuracy": 0.7717874437570572,
      "num_tokens": 1080122748.0,
      "step": 15990
    },
    {
      "entropy": 0.8129354000091553,
      "epoch": 4.8470802090433995,
      "grad_norm": 0.13995599746704102,
      "learning_rate": 8.095656592409224e-05,
      "loss": 1.2177,
      "mean_token_accuracy": 0.7612034097313881,
      "num_tokens": 1080781107.0,
      "step": 16000
    },
    {
      "entropy": 0.8003242433071136,
      "epoch": 4.8501098235249565,
      "grad_norm": 0.13097521662712097,
      "learning_rate": 8.09303355335013e-05,
      "loss": 1.2095,
      "mean_token_accuracy": 0.7696071028709411,
      "num_tokens": 1081464992.0,
      "step": 16010
    },
    {
      "entropy": 0.8115729466080666,
      "epoch": 4.853139438006513,
      "grad_norm": 0.14724335074424744,
      "learning_rate": 8.090409134643344e-05,
      "loss": 1.2132,
      "mean_token_accuracy": 0.7653060346841812,
      "num_tokens": 1082140319.0,
      "step": 16020
    },
    {
      "entropy": 0.8205862417817116,
      "epoch": 4.856169052488071,
      "grad_norm": 0.15980929136276245,
      "learning_rate": 8.087783337459483e-05,
      "loss": 1.2209,
      "mean_token_accuracy": 0.763490405678749,
      "num_tokens": 1082820861.0,
      "step": 16030
    },
    {
      "entropy": 0.8085527703166008,
      "epoch": 4.859198666969628,
      "grad_norm": 0.13519975543022156,
      "learning_rate": 8.08515616296979e-05,
      "loss": 1.2044,
      "mean_token_accuracy": 0.7669231325387955,
      "num_tokens": 1083495114.0,
      "step": 16040
    },
    {
      "entropy": 0.8118745133280754,
      "epoch": 4.862228281451185,
      "grad_norm": 0.1396489143371582,
      "learning_rate": 8.082527612346109e-05,
      "loss": 1.2027,
      "mean_token_accuracy": 0.7672496333718299,
      "num_tokens": 1084181821.0,
      "step": 16050
    },
    {
      "entropy": 0.8091390267014503,
      "epoch": 4.865257895932743,
      "grad_norm": 0.14365805685520172,
      "learning_rate": 8.079897686760911e-05,
      "loss": 1.2156,
      "mean_token_accuracy": 0.7642460703849793,
      "num_tokens": 1084848903.0,
      "step": 16060
    },
    {
      "entropy": 0.8168342247605324,
      "epoch": 4.8682875104143,
      "grad_norm": 0.1340799331665039,
      "learning_rate": 8.077266387387274e-05,
      "loss": 1.2185,
      "mean_token_accuracy": 0.7625942230224609,
      "num_tokens": 1085517417.0,
      "step": 16070
    },
    {
      "entropy": 0.8122792810201644,
      "epoch": 4.871317124895857,
      "grad_norm": 0.13644464313983917,
      "learning_rate": 8.074633715398885e-05,
      "loss": 1.208,
      "mean_token_accuracy": 0.7683076694607734,
      "num_tokens": 1086199795.0,
      "step": 16080
    },
    {
      "entropy": 0.8055481418967247,
      "epoch": 4.8743467393774145,
      "grad_norm": 0.14579367637634277,
      "learning_rate": 8.07199967197005e-05,
      "loss": 1.2053,
      "mean_token_accuracy": 0.7688647583127022,
      "num_tokens": 1086882422.0,
      "step": 16090
    },
    {
      "entropy": 0.8040731713175774,
      "epoch": 4.877376353858971,
      "grad_norm": 0.14209434390068054,
      "learning_rate": 8.069364258275686e-05,
      "loss": 1.2103,
      "mean_token_accuracy": 0.7682266667485237,
      "num_tokens": 1087554619.0,
      "step": 16100
    },
    {
      "entropy": 0.813388267159462,
      "epoch": 4.880405968340528,
      "grad_norm": 0.14394550025463104,
      "learning_rate": 8.066727475491316e-05,
      "loss": 1.2162,
      "mean_token_accuracy": 0.7615847617387772,
      "num_tokens": 1088225692.0,
      "step": 16110
    },
    {
      "entropy": 0.8020743995904922,
      "epoch": 4.883435582822086,
      "grad_norm": 0.13503175973892212,
      "learning_rate": 8.064089324793078e-05,
      "loss": 1.209,
      "mean_token_accuracy": 0.7675877928733825,
      "num_tokens": 1088893457.0,
      "step": 16120
    },
    {
      "entropy": 0.8262182638049126,
      "epoch": 4.886465197303643,
      "grad_norm": 0.13238561153411865,
      "learning_rate": 8.06144980735772e-05,
      "loss": 1.2207,
      "mean_token_accuracy": 0.7621331796050071,
      "num_tokens": 1089560857.0,
      "step": 16130
    },
    {
      "entropy": 0.8153786823153496,
      "epoch": 4.8894948117852,
      "grad_norm": 0.14539356529712677,
      "learning_rate": 8.058808924362598e-05,
      "loss": 1.2161,
      "mean_token_accuracy": 0.7684778615832328,
      "num_tokens": 1090242445.0,
      "step": 16140
    },
    {
      "entropy": 0.8062364771962166,
      "epoch": 4.892524426266758,
      "grad_norm": 0.147408589720726,
      "learning_rate": 8.056166676985679e-05,
      "loss": 1.21,
      "mean_token_accuracy": 0.7645332485437393,
      "num_tokens": 1090914161.0,
      "step": 16150
    },
    {
      "entropy": 0.8069047465920448,
      "epoch": 4.895554040748315,
      "grad_norm": 0.14002974331378937,
      "learning_rate": 8.053523066405536e-05,
      "loss": 1.2112,
      "mean_token_accuracy": 0.7636183395981788,
      "num_tokens": 1091582994.0,
      "step": 16160
    },
    {
      "entropy": 0.807197442650795,
      "epoch": 4.8985836552298725,
      "grad_norm": 0.13712963461875916,
      "learning_rate": 8.050878093801352e-05,
      "loss": 1.2124,
      "mean_token_accuracy": 0.7733067601919175,
      "num_tokens": 1092281317.0,
      "step": 16170
    },
    {
      "entropy": 0.8109682485461235,
      "epoch": 4.901613269711429,
      "grad_norm": 0.1413911134004593,
      "learning_rate": 8.048231760352919e-05,
      "loss": 1.2185,
      "mean_token_accuracy": 0.7638120874762535,
      "num_tokens": 1092950999.0,
      "step": 16180
    },
    {
      "entropy": 0.8190348640084266,
      "epoch": 4.904642884192986,
      "grad_norm": 0.13088011741638184,
      "learning_rate": 8.045584067240632e-05,
      "loss": 1.2139,
      "mean_token_accuracy": 0.7646927267313004,
      "num_tokens": 1093622838.0,
      "step": 16190
    },
    {
      "entropy": 0.8030320763587951,
      "epoch": 4.907672498674543,
      "grad_norm": 0.1281590461730957,
      "learning_rate": 8.042935015645494e-05,
      "loss": 1.2027,
      "mean_token_accuracy": 0.7652181640267373,
      "num_tokens": 1094284012.0,
      "step": 16200
    },
    {
      "entropy": 0.8119194015860558,
      "epoch": 4.910702113156101,
      "grad_norm": 0.14856310188770294,
      "learning_rate": 8.040284606749118e-05,
      "loss": 1.2199,
      "mean_token_accuracy": 0.76348657310009,
      "num_tokens": 1094955929.0,
      "step": 16210
    },
    {
      "entropy": 0.8177849322557449,
      "epoch": 4.913731727637658,
      "grad_norm": 0.14055438339710236,
      "learning_rate": 8.037632841733715e-05,
      "loss": 1.2181,
      "mean_token_accuracy": 0.76286730915308,
      "num_tokens": 1095619214.0,
      "step": 16220
    },
    {
      "entropy": 0.8127135917544365,
      "epoch": 4.916761342119216,
      "grad_norm": 0.13536129891872406,
      "learning_rate": 8.034979721782108e-05,
      "loss": 1.2073,
      "mean_token_accuracy": 0.765118145942688,
      "num_tokens": 1096299356.0,
      "step": 16230
    },
    {
      "entropy": 0.8101259261369705,
      "epoch": 4.919790956600773,
      "grad_norm": 0.13673269748687744,
      "learning_rate": 8.032325248077718e-05,
      "loss": 1.2087,
      "mean_token_accuracy": 0.7702684447169303,
      "num_tokens": 1096992769.0,
      "step": 16240
    },
    {
      "entropy": 0.8072752684354783,
      "epoch": 4.92282057108233,
      "grad_norm": 0.14365045726299286,
      "learning_rate": 8.029669421804575e-05,
      "loss": 1.2088,
      "mean_token_accuracy": 0.7633398428559304,
      "num_tokens": 1097654101.0,
      "step": 16250
    },
    {
      "entropy": 0.809220640361309,
      "epoch": 4.9258501855638865,
      "grad_norm": 0.14049658179283142,
      "learning_rate": 8.02701224414731e-05,
      "loss": 1.2244,
      "mean_token_accuracy": 0.7636054441332817,
      "num_tokens": 1098319432.0,
      "step": 16260
    },
    {
      "entropy": 0.8204178407788276,
      "epoch": 4.928879800045444,
      "grad_norm": 0.1522514373064041,
      "learning_rate": 8.02435371629116e-05,
      "loss": 1.221,
      "mean_token_accuracy": 0.7643356367945671,
      "num_tokens": 1098999276.0,
      "step": 16270
    },
    {
      "entropy": 0.8060835391283036,
      "epoch": 4.931909414527001,
      "grad_norm": 0.14118434488773346,
      "learning_rate": 8.021693839421955e-05,
      "loss": 1.2148,
      "mean_token_accuracy": 0.7658647254109383,
      "num_tokens": 1099665803.0,
      "step": 16280
    },
    {
      "entropy": 0.813490042090416,
      "epoch": 4.934939029008559,
      "grad_norm": 0.1457880437374115,
      "learning_rate": 8.019032614726138e-05,
      "loss": 1.2166,
      "mean_token_accuracy": 0.7657451406121254,
      "num_tokens": 1100336693.0,
      "step": 16290
    },
    {
      "entropy": 0.8225024595856667,
      "epoch": 4.937968643490116,
      "grad_norm": 0.14237776398658752,
      "learning_rate": 8.016370043390749e-05,
      "loss": 1.2242,
      "mean_token_accuracy": 0.7616039767861367,
      "num_tokens": 1101004222.0,
      "step": 16300
    },
    {
      "entropy": 0.8141745552420616,
      "epoch": 4.940998257971673,
      "grad_norm": 0.17437851428985596,
      "learning_rate": 8.013706126603425e-05,
      "loss": 1.2038,
      "mean_token_accuracy": 0.7661175921559333,
      "num_tokens": 1101689786.0,
      "step": 16310
    },
    {
      "entropy": 0.8207089826464653,
      "epoch": 4.94402787245323,
      "grad_norm": 0.14756397902965546,
      "learning_rate": 8.011040865552407e-05,
      "loss": 1.2189,
      "mean_token_accuracy": 0.7631851613521576,
      "num_tokens": 1102375768.0,
      "step": 16320
    },
    {
      "entropy": 0.8151626020669938,
      "epoch": 4.947057486934788,
      "grad_norm": 0.1360650360584259,
      "learning_rate": 8.008374261426537e-05,
      "loss": 1.2185,
      "mean_token_accuracy": 0.7666550606489182,
      "num_tokens": 1103050669.0,
      "step": 16330
    },
    {
      "entropy": 0.809863132238388,
      "epoch": 4.950087101416345,
      "grad_norm": 0.14102232456207275,
      "learning_rate": 8.005706315415249e-05,
      "loss": 1.216,
      "mean_token_accuracy": 0.7645842403173446,
      "num_tokens": 1103713172.0,
      "step": 16340
    },
    {
      "entropy": 0.8130029693245888,
      "epoch": 4.953116715897902,
      "grad_norm": 0.14143683016300201,
      "learning_rate": 8.003037028708586e-05,
      "loss": 1.2159,
      "mean_token_accuracy": 0.7645550161600113,
      "num_tokens": 1104386892.0,
      "step": 16350
    },
    {
      "entropy": 0.819441930949688,
      "epoch": 4.956146330379459,
      "grad_norm": 0.13079939782619476,
      "learning_rate": 8.00036640249718e-05,
      "loss": 1.219,
      "mean_token_accuracy": 0.760621365904808,
      "num_tokens": 1105058085.0,
      "step": 16360
    },
    {
      "entropy": 0.8078571721911431,
      "epoch": 4.959175944861016,
      "grad_norm": 0.13994356989860535,
      "learning_rate": 7.997694437972262e-05,
      "loss": 1.203,
      "mean_token_accuracy": 0.7705216273665428,
      "num_tokens": 1105744822.0,
      "step": 16370
    },
    {
      "entropy": 0.8017561569809913,
      "epoch": 4.962205559342574,
      "grad_norm": 0.13538730144500732,
      "learning_rate": 7.995021136325668e-05,
      "loss": 1.2093,
      "mean_token_accuracy": 0.7655615821480751,
      "num_tokens": 1106418696.0,
      "step": 16380
    },
    {
      "entropy": 0.8173870503902435,
      "epoch": 4.965235173824131,
      "grad_norm": 0.12956133484840393,
      "learning_rate": 7.992346498749819e-05,
      "loss": 1.2046,
      "mean_token_accuracy": 0.7658839270472526,
      "num_tokens": 1107102625.0,
      "step": 16390
    },
    {
      "entropy": 0.8095331564545631,
      "epoch": 4.968264788305688,
      "grad_norm": 0.15149347484111786,
      "learning_rate": 7.98967052643774e-05,
      "loss": 1.2192,
      "mean_token_accuracy": 0.7668271765112877,
      "num_tokens": 1107769722.0,
      "step": 16400
    },
    {
      "entropy": 0.7979603424668312,
      "epoch": 4.971294402787246,
      "grad_norm": 0.1289897859096527,
      "learning_rate": 7.986993220583048e-05,
      "loss": 1.1936,
      "mean_token_accuracy": 0.7726369246840477,
      "num_tokens": 1108454674.0,
      "step": 16410
    },
    {
      "entropy": 0.816437041759491,
      "epoch": 4.974324017268803,
      "grad_norm": 0.14153149724006653,
      "learning_rate": 7.984314582379953e-05,
      "loss": 1.2168,
      "mean_token_accuracy": 0.7692676842212677,
      "num_tokens": 1109140993.0,
      "step": 16420
    },
    {
      "entropy": 0.8106490090489388,
      "epoch": 4.9773536317503595,
      "grad_norm": 0.13549388945102692,
      "learning_rate": 7.981634613023264e-05,
      "loss": 1.2113,
      "mean_token_accuracy": 0.7671712547540664,
      "num_tokens": 1109824322.0,
      "step": 16430
    },
    {
      "entropy": 0.7994436040520668,
      "epoch": 4.980383246231917,
      "grad_norm": 0.12522515654563904,
      "learning_rate": 7.978953313708381e-05,
      "loss": 1.2035,
      "mean_token_accuracy": 0.772608882188797,
      "num_tokens": 1110507272.0,
      "step": 16440
    },
    {
      "entropy": 0.8076819390058517,
      "epoch": 4.983412860713474,
      "grad_norm": 0.14474385976791382,
      "learning_rate": 7.976270685631298e-05,
      "loss": 1.2131,
      "mean_token_accuracy": 0.7663069173693657,
      "num_tokens": 1111179622.0,
      "step": 16450
    },
    {
      "entropy": 0.8016024693846703,
      "epoch": 4.986442475195031,
      "grad_norm": 0.1426592469215393,
      "learning_rate": 7.9735867299886e-05,
      "loss": 1.2004,
      "mean_token_accuracy": 0.7635600075125695,
      "num_tokens": 1111846400.0,
      "step": 16460
    },
    {
      "entropy": 0.8050459831953048,
      "epoch": 4.989472089676589,
      "grad_norm": 0.13626575469970703,
      "learning_rate": 7.970901447977468e-05,
      "loss": 1.2052,
      "mean_token_accuracy": 0.7665823057293892,
      "num_tokens": 1112520449.0,
      "step": 16470
    },
    {
      "entropy": 0.7991900339722633,
      "epoch": 4.992501704158146,
      "grad_norm": 0.1464741975069046,
      "learning_rate": 7.968214840795669e-05,
      "loss": 1.1984,
      "mean_token_accuracy": 0.7744982793927193,
      "num_tokens": 1113220416.0,
      "step": 16480
    },
    {
      "entropy": 0.8126252934336662,
      "epoch": 4.995531318639703,
      "grad_norm": 0.1428397297859192,
      "learning_rate": 7.965526909641565e-05,
      "loss": 1.2223,
      "mean_token_accuracy": 0.7577746540307999,
      "num_tokens": 1113869987.0,
      "step": 16490
    },
    {
      "entropy": 0.8123869523406029,
      "epoch": 4.998560933121261,
      "grad_norm": 0.14128269255161285,
      "learning_rate": 7.962837655714108e-05,
      "loss": 1.2153,
      "mean_token_accuracy": 0.7669962048530579,
      "num_tokens": 1114536654.0,
      "step": 16500
    },
    {
      "entropy": 0.8037411188467954,
      "epoch": 5.0015148072407785,
      "grad_norm": 0.1507730633020401,
      "learning_rate": 7.96014708021284e-05,
      "loss": 1.198,
      "mean_token_accuracy": 0.7682625345694714,
      "num_tokens": 1115195349.0,
      "step": 16510
    },
    {
      "entropy": 0.780905881524086,
      "epoch": 5.004544421722336,
      "grad_norm": 0.14938247203826904,
      "learning_rate": 7.957455184337894e-05,
      "loss": 1.1795,
      "mean_token_accuracy": 0.7759376361966133,
      "num_tokens": 1115881351.0,
      "step": 16520
    },
    {
      "entropy": 0.7980219900608063,
      "epoch": 5.007574036203893,
      "grad_norm": 0.1593325436115265,
      "learning_rate": 7.954761969289988e-05,
      "loss": 1.1897,
      "mean_token_accuracy": 0.7703300774097442,
      "num_tokens": 1116566528.0,
      "step": 16530
    },
    {
      "entropy": 0.7913754537701607,
      "epoch": 5.01060365068545,
      "grad_norm": 0.1513516902923584,
      "learning_rate": 7.952067436270432e-05,
      "loss": 1.194,
      "mean_token_accuracy": 0.7686186075210572,
      "num_tokens": 1117232615.0,
      "step": 16540
    },
    {
      "entropy": 0.7906406939029693,
      "epoch": 5.013633265167008,
      "grad_norm": 0.15083470940589905,
      "learning_rate": 7.94937158648112e-05,
      "loss": 1.1853,
      "mean_token_accuracy": 0.7708072856068611,
      "num_tokens": 1117914393.0,
      "step": 16550
    },
    {
      "entropy": 0.7962068796157837,
      "epoch": 5.016662879648565,
      "grad_norm": 0.15922969579696655,
      "learning_rate": 7.946674421124538e-05,
      "loss": 1.1984,
      "mean_token_accuracy": 0.7660602360963822,
      "num_tokens": 1118575694.0,
      "step": 16560
    },
    {
      "entropy": 0.7868657410144806,
      "epoch": 5.019692494130122,
      "grad_norm": 0.1561596542596817,
      "learning_rate": 7.943975941403758e-05,
      "loss": 1.1842,
      "mean_token_accuracy": 0.7732081234455108,
      "num_tokens": 1119264774.0,
      "step": 16570
    },
    {
      "entropy": 0.7847541496157646,
      "epoch": 5.02272210861168,
      "grad_norm": 0.16248828172683716,
      "learning_rate": 7.941276148522434e-05,
      "loss": 1.187,
      "mean_token_accuracy": 0.762403430044651,
      "num_tokens": 1119909359.0,
      "step": 16580
    },
    {
      "entropy": 0.7921791404485703,
      "epoch": 5.0257517230932365,
      "grad_norm": 0.15669487416744232,
      "learning_rate": 7.93857504368481e-05,
      "loss": 1.1931,
      "mean_token_accuracy": 0.7709335729479789,
      "num_tokens": 1120583297.0,
      "step": 16590
    },
    {
      "entropy": 0.796550589799881,
      "epoch": 5.028781337574793,
      "grad_norm": 0.15571054816246033,
      "learning_rate": 7.935872628095714e-05,
      "loss": 1.1912,
      "mean_token_accuracy": 0.7698184594511985,
      "num_tokens": 1121261775.0,
      "step": 16600
    },
    {
      "entropy": 0.7889431893825531,
      "epoch": 5.031810952056351,
      "grad_norm": 0.1492854803800583,
      "learning_rate": 7.933168902960558e-05,
      "loss": 1.1963,
      "mean_token_accuracy": 0.7704354137182235,
      "num_tokens": 1121934145.0,
      "step": 16610
    },
    {
      "entropy": 0.7852740064263344,
      "epoch": 5.034840566537908,
      "grad_norm": 0.1636449098587036,
      "learning_rate": 7.93046386948534e-05,
      "loss": 1.1812,
      "mean_token_accuracy": 0.7721859395503998,
      "num_tokens": 1122616894.0,
      "step": 16620
    },
    {
      "entropy": 0.7963797673583031,
      "epoch": 5.037870181019465,
      "grad_norm": 0.1561834067106247,
      "learning_rate": 7.927757528876638e-05,
      "loss": 1.1985,
      "mean_token_accuracy": 0.7642228111624718,
      "num_tokens": 1123279115.0,
      "step": 16630
    },
    {
      "entropy": 0.7851715862751008,
      "epoch": 5.040899795501023,
      "grad_norm": 0.15005038678646088,
      "learning_rate": 7.925049882341617e-05,
      "loss": 1.1736,
      "mean_token_accuracy": 0.7784592941403389,
      "num_tokens": 1123975413.0,
      "step": 16640
    },
    {
      "entropy": 0.7886789575219154,
      "epoch": 5.04392940998258,
      "grad_norm": 0.14715303480625153,
      "learning_rate": 7.922340931088023e-05,
      "loss": 1.1867,
      "mean_token_accuracy": 0.7671677350997925,
      "num_tokens": 1124645664.0,
      "step": 16650
    },
    {
      "entropy": 0.7879483878612519,
      "epoch": 5.046959024464137,
      "grad_norm": 0.15014539659023285,
      "learning_rate": 7.919630676324182e-05,
      "loss": 1.1893,
      "mean_token_accuracy": 0.77218779027462,
      "num_tokens": 1125325049.0,
      "step": 16660
    },
    {
      "entropy": 0.8027291014790535,
      "epoch": 5.0499886389456945,
      "grad_norm": 0.16073153913021088,
      "learning_rate": 7.916919119259005e-05,
      "loss": 1.1953,
      "mean_token_accuracy": 0.7679773375391961,
      "num_tokens": 1126003373.0,
      "step": 16670
    },
    {
      "entropy": 0.7961043208837509,
      "epoch": 5.053018253427251,
      "grad_norm": 0.1571631133556366,
      "learning_rate": 7.914206261101981e-05,
      "loss": 1.1946,
      "mean_token_accuracy": 0.7720628321170807,
      "num_tokens": 1126693716.0,
      "step": 16680
    },
    {
      "entropy": 0.7894376143813133,
      "epoch": 5.056047867908808,
      "grad_norm": 0.15614300966262817,
      "learning_rate": 7.911492103063182e-05,
      "loss": 1.1897,
      "mean_token_accuracy": 0.7688526049256325,
      "num_tokens": 1127378061.0,
      "step": 16690
    },
    {
      "entropy": 0.7944885432720185,
      "epoch": 5.059077482390366,
      "grad_norm": 0.1512015461921692,
      "learning_rate": 7.908776646353256e-05,
      "loss": 1.1956,
      "mean_token_accuracy": 0.7652768597006798,
      "num_tokens": 1128032509.0,
      "step": 16700
    },
    {
      "entropy": 0.8022823825478553,
      "epoch": 5.062107096871923,
      "grad_norm": 0.1586894690990448,
      "learning_rate": 7.906059892183435e-05,
      "loss": 1.2,
      "mean_token_accuracy": 0.7627685695886612,
      "num_tokens": 1128694417.0,
      "step": 16710
    },
    {
      "entropy": 0.7928654015064239,
      "epoch": 5.06513671135348,
      "grad_norm": 0.16155175864696503,
      "learning_rate": 7.903341841765525e-05,
      "loss": 1.1953,
      "mean_token_accuracy": 0.7680094838142395,
      "num_tokens": 1129362984.0,
      "step": 16720
    },
    {
      "entropy": 0.7847378894686698,
      "epoch": 5.068166325835038,
      "grad_norm": 0.15931692719459534,
      "learning_rate": 7.900622496311913e-05,
      "loss": 1.1856,
      "mean_token_accuracy": 0.7690101683139801,
      "num_tokens": 1130024492.0,
      "step": 16730
    },
    {
      "entropy": 0.7901332467794419,
      "epoch": 5.071195940316595,
      "grad_norm": 0.15880520641803741,
      "learning_rate": 7.897901857035564e-05,
      "loss": 1.1917,
      "mean_token_accuracy": 0.7675976231694221,
      "num_tokens": 1130685618.0,
      "step": 16740
    },
    {
      "entropy": 0.786475557088852,
      "epoch": 5.074225554798152,
      "grad_norm": 0.15323227643966675,
      "learning_rate": 7.895179925150018e-05,
      "loss": 1.1852,
      "mean_token_accuracy": 0.7693305298686027,
      "num_tokens": 1131353060.0,
      "step": 16750
    },
    {
      "entropy": 0.7913914695382118,
      "epoch": 5.077255169279709,
      "grad_norm": 0.16168981790542603,
      "learning_rate": 7.892456701869395e-05,
      "loss": 1.1898,
      "mean_token_accuracy": 0.7700254335999489,
      "num_tokens": 1132024439.0,
      "step": 16760
    },
    {
      "entropy": 0.7973674595355987,
      "epoch": 5.080284783761266,
      "grad_norm": 0.16982251405715942,
      "learning_rate": 7.889732188408388e-05,
      "loss": 1.1911,
      "mean_token_accuracy": 0.7697751298546791,
      "num_tokens": 1132709850.0,
      "step": 16770
    },
    {
      "entropy": 0.7975167855620384,
      "epoch": 5.083314398242823,
      "grad_norm": 0.15392808616161346,
      "learning_rate": 7.887006385982263e-05,
      "loss": 1.196,
      "mean_token_accuracy": 0.7692336022853852,
      "num_tokens": 1133385764.0,
      "step": 16780
    },
    {
      "entropy": 0.7845405787229538,
      "epoch": 5.086344012724381,
      "grad_norm": 0.15405374765396118,
      "learning_rate": 7.884279295806869e-05,
      "loss": 1.1895,
      "mean_token_accuracy": 0.7696777954697609,
      "num_tokens": 1134061937.0,
      "step": 16790
    },
    {
      "entropy": 0.7827245593070984,
      "epoch": 5.089373627205938,
      "grad_norm": 0.15893754363059998,
      "learning_rate": 7.881550919098622e-05,
      "loss": 1.187,
      "mean_token_accuracy": 0.7711850360035897,
      "num_tokens": 1134733794.0,
      "step": 16800
    },
    {
      "entropy": 0.7979673892259598,
      "epoch": 5.092403241687495,
      "grad_norm": 0.15813829004764557,
      "learning_rate": 7.878821257074515e-05,
      "loss": 1.1941,
      "mean_token_accuracy": 0.7615123555064202,
      "num_tokens": 1135390702.0,
      "step": 16810
    },
    {
      "entropy": 0.7882887244224548,
      "epoch": 5.095432856169053,
      "grad_norm": 0.1594170778989792,
      "learning_rate": 7.876090310952114e-05,
      "loss": 1.1883,
      "mean_token_accuracy": 0.7671463191509247,
      "num_tokens": 1136048360.0,
      "step": 16820
    },
    {
      "entropy": 0.7984920248389245,
      "epoch": 5.09846247065061,
      "grad_norm": 0.16381298005580902,
      "learning_rate": 7.873358081949556e-05,
      "loss": 1.1943,
      "mean_token_accuracy": 0.765522375702858,
      "num_tokens": 1136729881.0,
      "step": 16830
    },
    {
      "entropy": 0.7995185390114784,
      "epoch": 5.101492085132167,
      "grad_norm": 0.1667272001504898,
      "learning_rate": 7.870624571285554e-05,
      "loss": 1.2002,
      "mean_token_accuracy": 0.7661367550492286,
      "num_tokens": 1137397887.0,
      "step": 16840
    },
    {
      "entropy": 0.7874711185693741,
      "epoch": 5.104521699613724,
      "grad_norm": 0.1503990888595581,
      "learning_rate": 7.867889780179391e-05,
      "loss": 1.1856,
      "mean_token_accuracy": 0.7681826263666153,
      "num_tokens": 1138069343.0,
      "step": 16850
    },
    {
      "entropy": 0.8023771598935128,
      "epoch": 5.107551314095281,
      "grad_norm": 0.1579052209854126,
      "learning_rate": 7.865153709850918e-05,
      "loss": 1.202,
      "mean_token_accuracy": 0.7645575031638145,
      "num_tokens": 1138735463.0,
      "step": 16860
    },
    {
      "entropy": 0.7982661604881287,
      "epoch": 5.110580928576838,
      "grad_norm": 0.16254378855228424,
      "learning_rate": 7.862416361520561e-05,
      "loss": 1.1893,
      "mean_token_accuracy": 0.7708167925477027,
      "num_tokens": 1139414205.0,
      "step": 16870
    },
    {
      "entropy": 0.7918319016695022,
      "epoch": 5.113610543058396,
      "grad_norm": 0.15832503139972687,
      "learning_rate": 7.859677736409314e-05,
      "loss": 1.1854,
      "mean_token_accuracy": 0.7706602737307549,
      "num_tokens": 1140087064.0,
      "step": 16880
    },
    {
      "entropy": 0.7866656050086022,
      "epoch": 5.116640157539953,
      "grad_norm": 0.1559450775384903,
      "learning_rate": 7.85693783573874e-05,
      "loss": 1.1823,
      "mean_token_accuracy": 0.7663189440965652,
      "num_tokens": 1140758625.0,
      "step": 16890
    },
    {
      "entropy": 0.786269998550415,
      "epoch": 5.11966977202151,
      "grad_norm": 0.16720271110534668,
      "learning_rate": 7.854196660730973e-05,
      "loss": 1.1854,
      "mean_token_accuracy": 0.7719661965966225,
      "num_tokens": 1141450650.0,
      "step": 16900
    },
    {
      "entropy": 0.793235857784748,
      "epoch": 5.122699386503068,
      "grad_norm": 0.16150441765785217,
      "learning_rate": 7.851454212608715e-05,
      "loss": 1.2012,
      "mean_token_accuracy": 0.7687240347266198,
      "num_tokens": 1142130031.0,
      "step": 16910
    },
    {
      "entropy": 0.7887409865856171,
      "epoch": 5.125729000984625,
      "grad_norm": 0.15608227252960205,
      "learning_rate": 7.848710492595232e-05,
      "loss": 1.1968,
      "mean_token_accuracy": 0.7706323504447937,
      "num_tokens": 1142801822.0,
      "step": 16920
    },
    {
      "entropy": 0.791064091026783,
      "epoch": 5.1287586154661815,
      "grad_norm": 0.16037173569202423,
      "learning_rate": 7.845965501914362e-05,
      "loss": 1.1873,
      "mean_token_accuracy": 0.7689804404973983,
      "num_tokens": 1143480608.0,
      "step": 16930
    },
    {
      "entropy": 0.7976777598261833,
      "epoch": 5.131788229947739,
      "grad_norm": 0.15852181613445282,
      "learning_rate": 7.843219241790509e-05,
      "loss": 1.1988,
      "mean_token_accuracy": 0.7673050552606583,
      "num_tokens": 1144153187.0,
      "step": 16940
    },
    {
      "entropy": 0.7921642765402794,
      "epoch": 5.134817844429296,
      "grad_norm": 0.1522085815668106,
      "learning_rate": 7.84047171344864e-05,
      "loss": 1.193,
      "mean_token_accuracy": 0.7688619747757912,
      "num_tokens": 1144831260.0,
      "step": 16950
    },
    {
      "entropy": 0.7973412096500396,
      "epoch": 5.137847458910853,
      "grad_norm": 0.15903477370738983,
      "learning_rate": 7.837722918114295e-05,
      "loss": 1.1951,
      "mean_token_accuracy": 0.7696291863918304,
      "num_tokens": 1145512417.0,
      "step": 16960
    },
    {
      "entropy": 0.791239495575428,
      "epoch": 5.140877073392411,
      "grad_norm": 0.1490985006093979,
      "learning_rate": 7.834972857013569e-05,
      "loss": 1.1892,
      "mean_token_accuracy": 0.7675310984253884,
      "num_tokens": 1146196717.0,
      "step": 16970
    },
    {
      "entropy": 0.8050056949257851,
      "epoch": 5.143906687873968,
      "grad_norm": 0.16099320352077484,
      "learning_rate": 7.83222153137313e-05,
      "loss": 1.1949,
      "mean_token_accuracy": 0.7670167252421379,
      "num_tokens": 1146879833.0,
      "step": 16980
    },
    {
      "entropy": 0.8021310731768608,
      "epoch": 5.146936302355526,
      "grad_norm": 0.16402247548103333,
      "learning_rate": 7.829468942420203e-05,
      "loss": 1.2004,
      "mean_token_accuracy": 0.7693236470222473,
      "num_tokens": 1147557609.0,
      "step": 16990
    },
    {
      "entropy": 0.7945751905441284,
      "epoch": 5.149965916837083,
      "grad_norm": 0.16784335672855377,
      "learning_rate": 7.826715091382586e-05,
      "loss": 1.1976,
      "mean_token_accuracy": 0.7710944831371307,
      "num_tokens": 1148231438.0,
      "step": 17000
    },
    {
      "entropy": 0.7846817806363106,
      "epoch": 5.1529955313186395,
      "grad_norm": 0.16117709875106812,
      "learning_rate": 7.823959979488629e-05,
      "loss": 1.181,
      "mean_token_accuracy": 0.7752747610211372,
      "num_tokens": 1148930952.0,
      "step": 17010
    },
    {
      "entropy": 0.7957236215472221,
      "epoch": 5.156025145800197,
      "grad_norm": 0.15562129020690918,
      "learning_rate": 7.821203607967254e-05,
      "loss": 1.1963,
      "mean_token_accuracy": 0.7648106575012207,
      "num_tokens": 1149603103.0,
      "step": 17020
    },
    {
      "entropy": 0.7977365210652352,
      "epoch": 5.159054760281754,
      "grad_norm": 0.1677064448595047,
      "learning_rate": 7.818445978047936e-05,
      "loss": 1.1947,
      "mean_token_accuracy": 0.7639513969421386,
      "num_tokens": 1150263365.0,
      "step": 17030
    },
    {
      "entropy": 0.7925516247749329,
      "epoch": 5.162084374763311,
      "grad_norm": 0.15624701976776123,
      "learning_rate": 7.815687090960721e-05,
      "loss": 1.1923,
      "mean_token_accuracy": 0.7755320727825165,
      "num_tokens": 1150953297.0,
      "step": 17040
    },
    {
      "entropy": 0.7901217445731163,
      "epoch": 5.165113989244869,
      "grad_norm": 0.15500527620315552,
      "learning_rate": 7.812926947936208e-05,
      "loss": 1.1905,
      "mean_token_accuracy": 0.7713487282395363,
      "num_tokens": 1151631635.0,
      "step": 17050
    },
    {
      "entropy": 0.7947961136698722,
      "epoch": 5.168143603726426,
      "grad_norm": 0.1531098335981369,
      "learning_rate": 7.810165550205559e-05,
      "loss": 1.1933,
      "mean_token_accuracy": 0.7699369430541992,
      "num_tokens": 1152309119.0,
      "step": 17060
    },
    {
      "entropy": 0.7894000634551048,
      "epoch": 5.171173218207983,
      "grad_norm": 0.15637709200382233,
      "learning_rate": 7.807402899000496e-05,
      "loss": 1.1934,
      "mean_token_accuracy": 0.7707509502768517,
      "num_tokens": 1152979896.0,
      "step": 17070
    },
    {
      "entropy": 0.7942910522222519,
      "epoch": 5.174202832689541,
      "grad_norm": 0.14667530357837677,
      "learning_rate": 7.804638995553297e-05,
      "loss": 1.1959,
      "mean_token_accuracy": 0.7722572177648545,
      "num_tokens": 1153663084.0,
      "step": 17080
    },
    {
      "entropy": 0.7933478116989136,
      "epoch": 5.1772324471710975,
      "grad_norm": 0.15234503149986267,
      "learning_rate": 7.801873841096806e-05,
      "loss": 1.1929,
      "mean_token_accuracy": 0.7653992846608162,
      "num_tokens": 1154331840.0,
      "step": 17090
    },
    {
      "entropy": 0.789708323776722,
      "epoch": 5.1802620616526545,
      "grad_norm": 0.1696467250585556,
      "learning_rate": 7.799107436864417e-05,
      "loss": 1.1968,
      "mean_token_accuracy": 0.7668629556894302,
      "num_tokens": 1154986076.0,
      "step": 17100
    },
    {
      "entropy": 0.7896785274147987,
      "epoch": 5.183291676134212,
      "grad_norm": 0.16039060056209564,
      "learning_rate": 7.796339784090085e-05,
      "loss": 1.1843,
      "mean_token_accuracy": 0.7684125617146492,
      "num_tokens": 1155666259.0,
      "step": 17110
    },
    {
      "entropy": 0.7867568179965019,
      "epoch": 5.186321290615769,
      "grad_norm": 0.15491805970668793,
      "learning_rate": 7.79357088400832e-05,
      "loss": 1.1882,
      "mean_token_accuracy": 0.7716071099042893,
      "num_tokens": 1156355994.0,
      "step": 17120
    },
    {
      "entropy": 0.7851361095905304,
      "epoch": 5.189350905097326,
      "grad_norm": 0.1565706729888916,
      "learning_rate": 7.790800737854195e-05,
      "loss": 1.1745,
      "mean_token_accuracy": 0.771054470539093,
      "num_tokens": 1157043126.0,
      "step": 17130
    },
    {
      "entropy": 0.7874387934803962,
      "epoch": 5.192380519578884,
      "grad_norm": 0.1554880291223526,
      "learning_rate": 7.78802934686333e-05,
      "loss": 1.1883,
      "mean_token_accuracy": 0.7711294487118721,
      "num_tokens": 1157721720.0,
      "step": 17140
    },
    {
      "entropy": 0.7959549397230148,
      "epoch": 5.195410134060441,
      "grad_norm": 0.15542958676815033,
      "learning_rate": 7.785256712271904e-05,
      "loss": 1.1973,
      "mean_token_accuracy": 0.767873203754425,
      "num_tokens": 1158404116.0,
      "step": 17150
    },
    {
      "entropy": 0.796953085064888,
      "epoch": 5.198439748541998,
      "grad_norm": 0.15236327052116394,
      "learning_rate": 7.782482835316652e-05,
      "loss": 1.195,
      "mean_token_accuracy": 0.7699533119797707,
      "num_tokens": 1159095642.0,
      "step": 17160
    },
    {
      "entropy": 0.8015800729393959,
      "epoch": 5.201469363023556,
      "grad_norm": 0.15539121627807617,
      "learning_rate": 7.77970771723486e-05,
      "loss": 1.2046,
      "mean_token_accuracy": 0.7635770738124847,
      "num_tokens": 1159750670.0,
      "step": 17170
    },
    {
      "entropy": 0.7865219786763191,
      "epoch": 5.2044989775051125,
      "grad_norm": 0.1567796766757965,
      "learning_rate": 7.776931359264372e-05,
      "loss": 1.1858,
      "mean_token_accuracy": 0.7690562218427658,
      "num_tokens": 1160420566.0,
      "step": 17180
    },
    {
      "entropy": 0.7902747079730034,
      "epoch": 5.207528591986669,
      "grad_norm": 0.15823183953762054,
      "learning_rate": 7.774153762643582e-05,
      "loss": 1.1898,
      "mean_token_accuracy": 0.7678332209587098,
      "num_tokens": 1161094600.0,
      "step": 17190
    },
    {
      "entropy": 0.7934893280267715,
      "epoch": 5.210558206468227,
      "grad_norm": 0.15117739140987396,
      "learning_rate": 7.771374928611435e-05,
      "loss": 1.1961,
      "mean_token_accuracy": 0.7713102355599404,
      "num_tokens": 1161775732.0,
      "step": 17200
    },
    {
      "entropy": 0.7978516027331353,
      "epoch": 5.213587820949784,
      "grad_norm": 0.15321889519691467,
      "learning_rate": 7.768594858407431e-05,
      "loss": 1.1883,
      "mean_token_accuracy": 0.7674748927354813,
      "num_tokens": 1162474209.0,
      "step": 17210
    },
    {
      "entropy": 0.7908621087670327,
      "epoch": 5.216617435431341,
      "grad_norm": 0.15823061764240265,
      "learning_rate": 7.765813553271622e-05,
      "loss": 1.1942,
      "mean_token_accuracy": 0.7713371187448501,
      "num_tokens": 1163156017.0,
      "step": 17220
    },
    {
      "entropy": 0.7892576485872269,
      "epoch": 5.219647049912899,
      "grad_norm": 0.15625379979610443,
      "learning_rate": 7.763031014444605e-05,
      "loss": 1.1863,
      "mean_token_accuracy": 0.7694304808974266,
      "num_tokens": 1163828041.0,
      "step": 17230
    },
    {
      "entropy": 0.7902972042560578,
      "epoch": 5.222676664394456,
      "grad_norm": 0.1579941213130951,
      "learning_rate": 7.760247243167537e-05,
      "loss": 1.1918,
      "mean_token_accuracy": 0.7715951129794121,
      "num_tokens": 1164522773.0,
      "step": 17240
    },
    {
      "entropy": 0.7882058292627334,
      "epoch": 5.225706278876013,
      "grad_norm": 0.1585836261510849,
      "learning_rate": 7.757462240682119e-05,
      "loss": 1.2006,
      "mean_token_accuracy": 0.769963701069355,
      "num_tokens": 1165188335.0,
      "step": 17250
    },
    {
      "entropy": 0.7864675641059875,
      "epoch": 5.2287358933575705,
      "grad_norm": 0.15874925255775452,
      "learning_rate": 7.754676008230595e-05,
      "loss": 1.1889,
      "mean_token_accuracy": 0.7740319341421127,
      "num_tokens": 1165870965.0,
      "step": 17260
    },
    {
      "entropy": 0.7968963533639908,
      "epoch": 5.231765507839127,
      "grad_norm": 0.15225261449813843,
      "learning_rate": 7.75188854705577e-05,
      "loss": 1.2002,
      "mean_token_accuracy": 0.769596466422081,
      "num_tokens": 1166544318.0,
      "step": 17270
    },
    {
      "entropy": 0.7892345637083054,
      "epoch": 5.234795122320684,
      "grad_norm": 0.1502452939748764,
      "learning_rate": 7.749099858400989e-05,
      "loss": 1.1929,
      "mean_token_accuracy": 0.7687133431434632,
      "num_tokens": 1167211314.0,
      "step": 17280
    },
    {
      "entropy": 0.7954817578196526,
      "epoch": 5.237824736802242,
      "grad_norm": 0.15724745392799377,
      "learning_rate": 7.746309943510148e-05,
      "loss": 1.193,
      "mean_token_accuracy": 0.7699337929487229,
      "num_tokens": 1167905279.0,
      "step": 17290
    },
    {
      "entropy": 0.7950238794088363,
      "epoch": 5.240854351283799,
      "grad_norm": 0.15903352200984955,
      "learning_rate": 7.743518803627688e-05,
      "loss": 1.1932,
      "mean_token_accuracy": 0.76488067060709,
      "num_tokens": 1168568163.0,
      "step": 17300
    },
    {
      "entropy": 0.792098093032837,
      "epoch": 5.243883965765356,
      "grad_norm": 0.15061922371387482,
      "learning_rate": 7.740726439998594e-05,
      "loss": 1.196,
      "mean_token_accuracy": 0.7639679074287414,
      "num_tokens": 1169228184.0,
      "step": 17310
    },
    {
      "entropy": 0.7873492568731308,
      "epoch": 5.246913580246914,
      "grad_norm": 0.15553319454193115,
      "learning_rate": 7.737932853868404e-05,
      "loss": 1.1919,
      "mean_token_accuracy": 0.7729442462325096,
      "num_tokens": 1169911211.0,
      "step": 17320
    },
    {
      "entropy": 0.7864971786737442,
      "epoch": 5.249943194728471,
      "grad_norm": 0.17369091510772705,
      "learning_rate": 7.735138046483196e-05,
      "loss": 1.1872,
      "mean_token_accuracy": 0.7715881854295731,
      "num_tokens": 1170586535.0,
      "step": 17330
    },
    {
      "entropy": 0.795674116909504,
      "epoch": 5.252972809210028,
      "grad_norm": 0.1560736447572708,
      "learning_rate": 7.732342019089593e-05,
      "loss": 1.1922,
      "mean_token_accuracy": 0.7709483101963996,
      "num_tokens": 1171271147.0,
      "step": 17340
    },
    {
      "entropy": 0.7958288848400116,
      "epoch": 5.256002423691585,
      "grad_norm": 0.15873171389102936,
      "learning_rate": 7.729544772934764e-05,
      "loss": 1.1962,
      "mean_token_accuracy": 0.7691018924117088,
      "num_tokens": 1171944249.0,
      "step": 17350
    },
    {
      "entropy": 0.7819776698946953,
      "epoch": 5.259032038173142,
      "grad_norm": 0.15958859026432037,
      "learning_rate": 7.72674630926642e-05,
      "loss": 1.187,
      "mean_token_accuracy": 0.7761292487382889,
      "num_tokens": 1172625656.0,
      "step": 17360
    },
    {
      "entropy": 0.7825765237212181,
      "epoch": 5.262061652654699,
      "grad_norm": 0.15552274882793427,
      "learning_rate": 7.723946629332817e-05,
      "loss": 1.1863,
      "mean_token_accuracy": 0.7742067262530327,
      "num_tokens": 1173310261.0,
      "step": 17370
    },
    {
      "entropy": 0.7916419252753257,
      "epoch": 5.265091267136257,
      "grad_norm": 0.15146741271018982,
      "learning_rate": 7.721145734382749e-05,
      "loss": 1.1892,
      "mean_token_accuracy": 0.7693522796034813,
      "num_tokens": 1173977891.0,
      "step": 17380
    },
    {
      "entropy": 0.7943963199853897,
      "epoch": 5.268120881617814,
      "grad_norm": 0.1591690331697464,
      "learning_rate": 7.718343625665559e-05,
      "loss": 1.1894,
      "mean_token_accuracy": 0.7696996122598648,
      "num_tokens": 1174663181.0,
      "step": 17390
    },
    {
      "entropy": 0.7926700994372368,
      "epoch": 5.271150496099372,
      "grad_norm": 0.160016730427742,
      "learning_rate": 7.715540304431128e-05,
      "loss": 1.1914,
      "mean_token_accuracy": 0.7728733107447624,
      "num_tokens": 1175359104.0,
      "step": 17400
    },
    {
      "entropy": 0.8013308346271515,
      "epoch": 5.274180110580929,
      "grad_norm": 0.1549455225467682,
      "learning_rate": 7.712735771929873e-05,
      "loss": 1.2046,
      "mean_token_accuracy": 0.7645180448889732,
      "num_tokens": 1176030111.0,
      "step": 17410
    },
    {
      "entropy": 0.8020660355687141,
      "epoch": 5.277209725062486,
      "grad_norm": 0.15448881685733795,
      "learning_rate": 7.709930029412762e-05,
      "loss": 1.1964,
      "mean_token_accuracy": 0.7696937382221222,
      "num_tokens": 1176709324.0,
      "step": 17420
    },
    {
      "entropy": 0.7843293473124504,
      "epoch": 5.280239339544043,
      "grad_norm": 0.15545623004436493,
      "learning_rate": 7.70712307813129e-05,
      "loss": 1.1921,
      "mean_token_accuracy": 0.7684095501899719,
      "num_tokens": 1177375159.0,
      "step": 17430
    },
    {
      "entropy": 0.8017087817192078,
      "epoch": 5.2832689540256,
      "grad_norm": 0.15782956779003143,
      "learning_rate": 7.704314919337504e-05,
      "loss": 1.1971,
      "mean_token_accuracy": 0.7661067008972168,
      "num_tokens": 1178046557.0,
      "step": 17440
    },
    {
      "entropy": 0.7860934764146805,
      "epoch": 5.286298568507157,
      "grad_norm": 0.15507152676582336,
      "learning_rate": 7.701505554283977e-05,
      "loss": 1.1891,
      "mean_token_accuracy": 0.7704472541809082,
      "num_tokens": 1178720516.0,
      "step": 17450
    },
    {
      "entropy": 0.7910739570856095,
      "epoch": 5.289328182988715,
      "grad_norm": 0.1577184647321701,
      "learning_rate": 7.698694984223831e-05,
      "loss": 1.1902,
      "mean_token_accuracy": 0.7696097001433373,
      "num_tokens": 1179391162.0,
      "step": 17460
    },
    {
      "entropy": 0.78514154702425,
      "epoch": 5.292357797470272,
      "grad_norm": 0.15412352979183197,
      "learning_rate": 7.69588321041072e-05,
      "loss": 1.1819,
      "mean_token_accuracy": 0.7715278282761574,
      "num_tokens": 1180072891.0,
      "step": 17470
    },
    {
      "entropy": 0.7782529965043068,
      "epoch": 5.295387411951829,
      "grad_norm": 0.1585690826177597,
      "learning_rate": 7.693070234098834e-05,
      "loss": 1.1831,
      "mean_token_accuracy": 0.7694088965654373,
      "num_tokens": 1180742973.0,
      "step": 17480
    },
    {
      "entropy": 0.8020209014415741,
      "epoch": 5.298417026433387,
      "grad_norm": 0.15189771354198456,
      "learning_rate": 7.690256056542903e-05,
      "loss": 1.1988,
      "mean_token_accuracy": 0.7645947188138962,
      "num_tokens": 1181415326.0,
      "step": 17490
    },
    {
      "entropy": 0.8022048667073249,
      "epoch": 5.301446640914944,
      "grad_norm": 0.15041834115982056,
      "learning_rate": 7.68744067899819e-05,
      "loss": 1.1947,
      "mean_token_accuracy": 0.7714553833007812,
      "num_tokens": 1182104379.0,
      "step": 17500
    },
    {
      "entropy": 0.8034254133701324,
      "epoch": 5.304476255396501,
      "grad_norm": 0.15402550995349884,
      "learning_rate": 7.684624102720495e-05,
      "loss": 1.1918,
      "mean_token_accuracy": 0.7663976415991783,
      "num_tokens": 1182791512.0,
      "step": 17510
    },
    {
      "entropy": 0.7869031742215157,
      "epoch": 5.307505869878058,
      "grad_norm": 0.15747013688087463,
      "learning_rate": 7.68180632896615e-05,
      "loss": 1.1935,
      "mean_token_accuracy": 0.7666245847940445,
      "num_tokens": 1183458053.0,
      "step": 17520
    },
    {
      "entropy": 0.7984980285167694,
      "epoch": 5.310535484359615,
      "grad_norm": 0.1575598120689392,
      "learning_rate": 7.678987358992026e-05,
      "loss": 1.1934,
      "mean_token_accuracy": 0.7688501477241516,
      "num_tokens": 1184122101.0,
      "step": 17530
    },
    {
      "entropy": 0.7883581951260566,
      "epoch": 5.313565098841172,
      "grad_norm": 0.17190022766590118,
      "learning_rate": 7.676167194055523e-05,
      "loss": 1.1914,
      "mean_token_accuracy": 0.7717085897922515,
      "num_tokens": 1184800988.0,
      "step": 17540
    },
    {
      "entropy": 0.7959327608346939,
      "epoch": 5.31659471332273,
      "grad_norm": 0.15310914814472198,
      "learning_rate": 7.673345835414575e-05,
      "loss": 1.1903,
      "mean_token_accuracy": 0.7683573260903358,
      "num_tokens": 1185475110.0,
      "step": 17550
    },
    {
      "entropy": 0.7981788322329522,
      "epoch": 5.319624327804287,
      "grad_norm": 0.1599741131067276,
      "learning_rate": 7.67052328432765e-05,
      "loss": 1.2044,
      "mean_token_accuracy": 0.7693395286798477,
      "num_tokens": 1186148233.0,
      "step": 17560
    },
    {
      "entropy": 0.7871917247772217,
      "epoch": 5.322653942285844,
      "grad_norm": 0.16612312197685242,
      "learning_rate": 7.667699542053748e-05,
      "loss": 1.1906,
      "mean_token_accuracy": 0.7699298709630966,
      "num_tokens": 1186822751.0,
      "step": 17570
    },
    {
      "entropy": 0.7859062150120735,
      "epoch": 5.325683556767402,
      "grad_norm": 0.17001472413539886,
      "learning_rate": 7.664874609852397e-05,
      "loss": 1.1978,
      "mean_token_accuracy": 0.7713186040520668,
      "num_tokens": 1187493996.0,
      "step": 17580
    },
    {
      "entropy": 0.7932150304317475,
      "epoch": 5.328713171248959,
      "grad_norm": 0.1645856350660324,
      "learning_rate": 7.662048488983658e-05,
      "loss": 1.1869,
      "mean_token_accuracy": 0.7701080903410912,
      "num_tokens": 1188171986.0,
      "step": 17590
    },
    {
      "entropy": 0.7961829259991646,
      "epoch": 5.3317427857305155,
      "grad_norm": 0.14714856445789337,
      "learning_rate": 7.659221180708126e-05,
      "loss": 1.1883,
      "mean_token_accuracy": 0.7689448133111,
      "num_tokens": 1188859031.0,
      "step": 17600
    },
    {
      "entropy": 0.7916832432150841,
      "epoch": 5.334772400212073,
      "grad_norm": 0.16263554990291595,
      "learning_rate": 7.656392686286918e-05,
      "loss": 1.1831,
      "mean_token_accuracy": 0.7754878297448158,
      "num_tokens": 1189551929.0,
      "step": 17610
    },
    {
      "entropy": 0.7899981632828712,
      "epoch": 5.33780201469363,
      "grad_norm": 0.1638210117816925,
      "learning_rate": 7.653563006981685e-05,
      "loss": 1.1937,
      "mean_token_accuracy": 0.7634357333183288,
      "num_tokens": 1190204767.0,
      "step": 17620
    },
    {
      "entropy": 0.7990321606397629,
      "epoch": 5.340831629175187,
      "grad_norm": 0.1562741994857788,
      "learning_rate": 7.650732144054607e-05,
      "loss": 1.2078,
      "mean_token_accuracy": 0.7641814783215523,
      "num_tokens": 1190859135.0,
      "step": 17630
    },
    {
      "entropy": 0.795019020140171,
      "epoch": 5.343861243656745,
      "grad_norm": 0.14475341141223907,
      "learning_rate": 7.64790009876839e-05,
      "loss": 1.1927,
      "mean_token_accuracy": 0.7689029216766358,
      "num_tokens": 1191530293.0,
      "step": 17640
    },
    {
      "entropy": 0.7977340087294579,
      "epoch": 5.346890858138302,
      "grad_norm": 0.1600002646446228,
      "learning_rate": 7.645066872386266e-05,
      "loss": 1.2017,
      "mean_token_accuracy": 0.7684462234377861,
      "num_tokens": 1192217113.0,
      "step": 17650
    },
    {
      "entropy": 0.7936551615595817,
      "epoch": 5.349920472619859,
      "grad_norm": 0.1655941754579544,
      "learning_rate": 7.642232466171997e-05,
      "loss": 1.1848,
      "mean_token_accuracy": 0.7676477208733559,
      "num_tokens": 1192884086.0,
      "step": 17660
    },
    {
      "entropy": 0.7909827798604965,
      "epoch": 5.352950087101417,
      "grad_norm": 0.14532136917114258,
      "learning_rate": 7.639396881389872e-05,
      "loss": 1.1917,
      "mean_token_accuracy": 0.7680937454104424,
      "num_tokens": 1193553087.0,
      "step": 17670
    },
    {
      "entropy": 0.7936312824487686,
      "epoch": 5.3559797015829735,
      "grad_norm": 0.18110795319080353,
      "learning_rate": 7.6365601193047e-05,
      "loss": 1.1932,
      "mean_token_accuracy": 0.7639016881585121,
      "num_tokens": 1194221589.0,
      "step": 17680
    },
    {
      "entropy": 0.7801448628306389,
      "epoch": 5.3590093160645305,
      "grad_norm": 0.1572643667459488,
      "learning_rate": 7.633722181181819e-05,
      "loss": 1.1811,
      "mean_token_accuracy": 0.7772280246019363,
      "num_tokens": 1194915590.0,
      "step": 17690
    },
    {
      "entropy": 0.79637009203434,
      "epoch": 5.362038930546088,
      "grad_norm": 0.15378724038600922,
      "learning_rate": 7.630883068287096e-05,
      "loss": 1.1969,
      "mean_token_accuracy": 0.7703924238681793,
      "num_tokens": 1195587875.0,
      "step": 17700
    },
    {
      "entropy": 0.7883513554930687,
      "epoch": 5.365068545027645,
      "grad_norm": 0.15346989035606384,
      "learning_rate": 7.628042781886911e-05,
      "loss": 1.1852,
      "mean_token_accuracy": 0.7732666581869125,
      "num_tokens": 1196274595.0,
      "step": 17710
    },
    {
      "entropy": 0.7764868423342705,
      "epoch": 5.368098159509202,
      "grad_norm": 0.14909599721431732,
      "learning_rate": 7.625201323248179e-05,
      "loss": 1.1815,
      "mean_token_accuracy": 0.7732973173260689,
      "num_tokens": 1196948431.0,
      "step": 17720
    },
    {
      "entropy": 0.7900894731283188,
      "epoch": 5.37112777399076,
      "grad_norm": 0.1467720866203308,
      "learning_rate": 7.622358693638329e-05,
      "loss": 1.1905,
      "mean_token_accuracy": 0.769812461733818,
      "num_tokens": 1197620161.0,
      "step": 17730
    },
    {
      "entropy": 0.7899725899100304,
      "epoch": 5.374157388472317,
      "grad_norm": 0.15409912168979645,
      "learning_rate": 7.61951489432532e-05,
      "loss": 1.1898,
      "mean_token_accuracy": 0.7687654465436935,
      "num_tokens": 1198284040.0,
      "step": 17740
    },
    {
      "entropy": 0.7853469118475914,
      "epoch": 5.377187002953874,
      "grad_norm": 0.15078231692314148,
      "learning_rate": 7.616669926577625e-05,
      "loss": 1.1803,
      "mean_token_accuracy": 0.7739398464560509,
      "num_tokens": 1198967676.0,
      "step": 17750
    },
    {
      "entropy": 0.7953891828656197,
      "epoch": 5.380216617435432,
      "grad_norm": 0.15821775794029236,
      "learning_rate": 7.613823791664244e-05,
      "loss": 1.1971,
      "mean_token_accuracy": 0.7657200857996941,
      "num_tokens": 1199635094.0,
      "step": 17760
    },
    {
      "entropy": 0.8010059207677841,
      "epoch": 5.3832462319169885,
      "grad_norm": 0.16155438125133514,
      "learning_rate": 7.610976490854696e-05,
      "loss": 1.1946,
      "mean_token_accuracy": 0.7681468084454537,
      "num_tokens": 1200322029.0,
      "step": 17770
    },
    {
      "entropy": 0.7836634263396263,
      "epoch": 5.386275846398545,
      "grad_norm": 0.15151795744895935,
      "learning_rate": 7.608128025419017e-05,
      "loss": 1.1793,
      "mean_token_accuracy": 0.7685976520180702,
      "num_tokens": 1201000866.0,
      "step": 17780
    },
    {
      "entropy": 0.7943032741546631,
      "epoch": 5.389305460880103,
      "grad_norm": 0.16315065324306488,
      "learning_rate": 7.605278396627766e-05,
      "loss": 1.192,
      "mean_token_accuracy": 0.7686750799417496,
      "num_tokens": 1201685014.0,
      "step": 17790
    },
    {
      "entropy": 0.7829157680273056,
      "epoch": 5.39233507536166,
      "grad_norm": 0.1587747037410736,
      "learning_rate": 7.602427605752023e-05,
      "loss": 1.1857,
      "mean_token_accuracy": 0.7704502210021019,
      "num_tokens": 1202350608.0,
      "step": 17800
    },
    {
      "entropy": 0.8006536051630974,
      "epoch": 5.395364689843217,
      "grad_norm": 0.1580352634191513,
      "learning_rate": 7.599575654063378e-05,
      "loss": 1.2079,
      "mean_token_accuracy": 0.7639207169413567,
      "num_tokens": 1203010524.0,
      "step": 17810
    },
    {
      "entropy": 0.7890890315175056,
      "epoch": 5.398394304324775,
      "grad_norm": 0.15163546800613403,
      "learning_rate": 7.596722542833949e-05,
      "loss": 1.1899,
      "mean_token_accuracy": 0.76937026232481,
      "num_tokens": 1203682275.0,
      "step": 17820
    },
    {
      "entropy": 0.8010433107614517,
      "epoch": 5.401423918806332,
      "grad_norm": 0.15345145761966705,
      "learning_rate": 7.593868273336364e-05,
      "loss": 1.1963,
      "mean_token_accuracy": 0.7656665444374084,
      "num_tokens": 1204349962.0,
      "step": 17830
    },
    {
      "entropy": 0.7970649287104606,
      "epoch": 5.404453533287889,
      "grad_norm": 0.16714106500148773,
      "learning_rate": 7.591012846843771e-05,
      "loss": 1.1999,
      "mean_token_accuracy": 0.7701074466109276,
      "num_tokens": 1205021424.0,
      "step": 17840
    },
    {
      "entropy": 0.7867384389042854,
      "epoch": 5.4074831477694465,
      "grad_norm": 0.14899560809135437,
      "learning_rate": 7.588156264629831e-05,
      "loss": 1.183,
      "mean_token_accuracy": 0.7700198113918304,
      "num_tokens": 1205698314.0,
      "step": 17850
    },
    {
      "entropy": 0.8040710180997849,
      "epoch": 5.410512762251003,
      "grad_norm": 0.15596075356006622,
      "learning_rate": 7.585298527968726e-05,
      "loss": 1.1995,
      "mean_token_accuracy": 0.7669637724757195,
      "num_tokens": 1206372716.0,
      "step": 17860
    },
    {
      "entropy": 0.7969091936945916,
      "epoch": 5.413542376732561,
      "grad_norm": 0.15836764872074127,
      "learning_rate": 7.582439638135146e-05,
      "loss": 1.1986,
      "mean_token_accuracy": 0.770765969157219,
      "num_tokens": 1207050369.0,
      "step": 17870
    },
    {
      "entropy": 0.7875916823744774,
      "epoch": 5.416571991214118,
      "grad_norm": 0.14866401255130768,
      "learning_rate": 7.579579596404301e-05,
      "loss": 1.1873,
      "mean_token_accuracy": 0.7716606304049491,
      "num_tokens": 1207729643.0,
      "step": 17880
    },
    {
      "entropy": 0.8014676854014396,
      "epoch": 5.419601605695675,
      "grad_norm": 0.15639302134513855,
      "learning_rate": 7.576718404051912e-05,
      "loss": 1.1979,
      "mean_token_accuracy": 0.7659138143062592,
      "num_tokens": 1208406975.0,
      "step": 17890
    },
    {
      "entropy": 0.7919231161475182,
      "epoch": 5.422631220177232,
      "grad_norm": 0.15693379938602448,
      "learning_rate": 7.573856062354214e-05,
      "loss": 1.1901,
      "mean_token_accuracy": 0.7700849264860153,
      "num_tokens": 1209086200.0,
      "step": 17900
    },
    {
      "entropy": 0.7928600683808327,
      "epoch": 5.42566083465879,
      "grad_norm": 0.15709640085697174,
      "learning_rate": 7.570992572587955e-05,
      "loss": 1.1882,
      "mean_token_accuracy": 0.7692574381828308,
      "num_tokens": 1209781026.0,
      "step": 17910
    },
    {
      "entropy": 0.7872502312064171,
      "epoch": 5.428690449140347,
      "grad_norm": 0.16345106065273285,
      "learning_rate": 7.568127936030394e-05,
      "loss": 1.1912,
      "mean_token_accuracy": 0.7697390720248223,
      "num_tokens": 1210464753.0,
      "step": 17920
    },
    {
      "entropy": 0.797100330889225,
      "epoch": 5.4317200636219045,
      "grad_norm": 0.16022591292858124,
      "learning_rate": 7.565262153959301e-05,
      "loss": 1.194,
      "mean_token_accuracy": 0.7657869562506676,
      "num_tokens": 1211130813.0,
      "step": 17930
    },
    {
      "entropy": 0.7876862704753875,
      "epoch": 5.434749678103461,
      "grad_norm": 0.14714138209819794,
      "learning_rate": 7.562395227652962e-05,
      "loss": 1.1903,
      "mean_token_accuracy": 0.7710764214396477,
      "num_tokens": 1211815871.0,
      "step": 17940
    },
    {
      "entropy": 0.7902646377682686,
      "epoch": 5.437779292585018,
      "grad_norm": 0.15407568216323853,
      "learning_rate": 7.559527158390166e-05,
      "loss": 1.1894,
      "mean_token_accuracy": 0.7710242703557014,
      "num_tokens": 1212497763.0,
      "step": 17950
    },
    {
      "entropy": 0.7920958489179611,
      "epoch": 5.440808907066576,
      "grad_norm": 0.15582390129566193,
      "learning_rate": 7.556657947450217e-05,
      "loss": 1.1944,
      "mean_token_accuracy": 0.7680147767066956,
      "num_tokens": 1213167488.0,
      "step": 17960
    },
    {
      "entropy": 0.7920024797320366,
      "epoch": 5.443838521548133,
      "grad_norm": 0.16489140689373016,
      "learning_rate": 7.553787596112927e-05,
      "loss": 1.1871,
      "mean_token_accuracy": 0.7655975177884102,
      "num_tokens": 1213829908.0,
      "step": 17970
    },
    {
      "entropy": 0.7848840445280075,
      "epoch": 5.44686813602969,
      "grad_norm": 0.16326844692230225,
      "learning_rate": 7.550916105658614e-05,
      "loss": 1.1861,
      "mean_token_accuracy": 0.7712792336940766,
      "num_tokens": 1214503562.0,
      "step": 17980
    },
    {
      "entropy": 0.7959401324391365,
      "epoch": 5.449897750511248,
      "grad_norm": 0.15023329854011536,
      "learning_rate": 7.548043477368108e-05,
      "loss": 1.1896,
      "mean_token_accuracy": 0.7691247791051865,
      "num_tokens": 1215183259.0,
      "step": 17990
    },
    {
      "entropy": 0.7946044862270355,
      "epoch": 5.452927364992805,
      "grad_norm": 0.1672666370868683,
      "learning_rate": 7.545169712522745e-05,
      "loss": 1.1974,
      "mean_token_accuracy": 0.7653155639767647,
      "num_tokens": 1215843813.0,
      "step": 18000
    },
    {
      "entropy": 0.7864162981510162,
      "epoch": 5.455956979474362,
      "grad_norm": 0.15071184933185577,
      "learning_rate": 7.54229481240437e-05,
      "loss": 1.1825,
      "mean_token_accuracy": 0.7732902497053147,
      "num_tokens": 1216541863.0,
      "step": 18010
    },
    {
      "entropy": 0.797180961072445,
      "epoch": 5.4589865939559195,
      "grad_norm": 0.16377383470535278,
      "learning_rate": 7.539418778295328e-05,
      "loss": 1.1943,
      "mean_token_accuracy": 0.7646159008145332,
      "num_tokens": 1217206095.0,
      "step": 18020
    },
    {
      "entropy": 0.7854379042983055,
      "epoch": 5.462016208437476,
      "grad_norm": 0.15638822317123413,
      "learning_rate": 7.536541611478475e-05,
      "loss": 1.1871,
      "mean_token_accuracy": 0.7728175222873688,
      "num_tokens": 1217885808.0,
      "step": 18030
    },
    {
      "entropy": 0.7983357831835747,
      "epoch": 5.465045822919033,
      "grad_norm": 0.15034106373786926,
      "learning_rate": 7.533663313237174e-05,
      "loss": 1.197,
      "mean_token_accuracy": 0.7691913187503815,
      "num_tokens": 1218549867.0,
      "step": 18040
    },
    {
      "entropy": 0.7936286747455596,
      "epoch": 5.468075437400591,
      "grad_norm": 0.1538112610578537,
      "learning_rate": 7.530783884855288e-05,
      "loss": 1.1996,
      "mean_token_accuracy": 0.7685976907610893,
      "num_tokens": 1219220339.0,
      "step": 18050
    },
    {
      "entropy": 0.7856300622224808,
      "epoch": 5.471105051882148,
      "grad_norm": 0.1578034907579422,
      "learning_rate": 7.527903327617185e-05,
      "loss": 1.1933,
      "mean_token_accuracy": 0.766740171611309,
      "num_tokens": 1219885319.0,
      "step": 18060
    },
    {
      "entropy": 0.7912115275859832,
      "epoch": 5.474134666363705,
      "grad_norm": 0.15818874537944794,
      "learning_rate": 7.52502164280774e-05,
      "loss": 1.1858,
      "mean_token_accuracy": 0.7702404156327247,
      "num_tokens": 1220555918.0,
      "step": 18070
    },
    {
      "entropy": 0.7877009674906731,
      "epoch": 5.477164280845263,
      "grad_norm": 0.15725462138652802,
      "learning_rate": 7.522138831712329e-05,
      "loss": 1.1929,
      "mean_token_accuracy": 0.7679716780781746,
      "num_tokens": 1221224817.0,
      "step": 18080
    },
    {
      "entropy": 0.7976912975311279,
      "epoch": 5.48019389532682,
      "grad_norm": 0.15438312292099,
      "learning_rate": 7.519254895616828e-05,
      "loss": 1.1928,
      "mean_token_accuracy": 0.7668454617261886,
      "num_tokens": 1221898011.0,
      "step": 18090
    },
    {
      "entropy": 0.7954000815749168,
      "epoch": 5.483223509808377,
      "grad_norm": 0.15981487929821014,
      "learning_rate": 7.516369835807615e-05,
      "loss": 1.1948,
      "mean_token_accuracy": 0.7648048520088195,
      "num_tokens": 1222570179.0,
      "step": 18100
    },
    {
      "entropy": 0.7921616017818451,
      "epoch": 5.486253124289934,
      "grad_norm": 0.16003654897212982,
      "learning_rate": 7.513483653571577e-05,
      "loss": 1.1958,
      "mean_token_accuracy": 0.7714850351214408,
      "num_tokens": 1223238713.0,
      "step": 18110
    },
    {
      "entropy": 0.7942561611533165,
      "epoch": 5.489282738771491,
      "grad_norm": 0.15539222955703735,
      "learning_rate": 7.510596350196091e-05,
      "loss": 1.2015,
      "mean_token_accuracy": 0.7683041110634804,
      "num_tokens": 1223908044.0,
      "step": 18120
    },
    {
      "entropy": 0.7851544454693794,
      "epoch": 5.492312353253048,
      "grad_norm": 0.1554105132818222,
      "learning_rate": 7.50770792696904e-05,
      "loss": 1.1918,
      "mean_token_accuracy": 0.7712665885686875,
      "num_tokens": 1224574438.0,
      "step": 18130
    },
    {
      "entropy": 0.7835942476987838,
      "epoch": 5.495341967734606,
      "grad_norm": 0.16624827682971954,
      "learning_rate": 7.504818385178803e-05,
      "loss": 1.1864,
      "mean_token_accuracy": 0.7709278538823128,
      "num_tokens": 1225244698.0,
      "step": 18140
    },
    {
      "entropy": 0.797513335943222,
      "epoch": 5.498371582216163,
      "grad_norm": 0.16395531594753265,
      "learning_rate": 7.501927726114264e-05,
      "loss": 1.2023,
      "mean_token_accuracy": 0.7699491426348686,
      "num_tokens": 1225935554.0,
      "step": 18150
    },
    {
      "entropy": 0.7859024435281754,
      "epoch": 5.50140119669772,
      "grad_norm": 0.15835483372211456,
      "learning_rate": 7.499035951064799e-05,
      "loss": 1.1841,
      "mean_token_accuracy": 0.7728272244334221,
      "num_tokens": 1226609336.0,
      "step": 18160
    },
    {
      "entropy": 0.7876161232590675,
      "epoch": 5.504430811179278,
      "grad_norm": 0.1516963094472885,
      "learning_rate": 7.496143061320284e-05,
      "loss": 1.1965,
      "mean_token_accuracy": 0.7686561167240142,
      "num_tokens": 1227276724.0,
      "step": 18170
    },
    {
      "entropy": 0.7917749524116516,
      "epoch": 5.507460425660835,
      "grad_norm": 0.15597166121006012,
      "learning_rate": 7.493249058171096e-05,
      "loss": 1.1836,
      "mean_token_accuracy": 0.7711313754320145,
      "num_tokens": 1227967930.0,
      "step": 18180
    },
    {
      "entropy": 0.7929075747728348,
      "epoch": 5.5104900401423915,
      "grad_norm": 0.14477355778217316,
      "learning_rate": 7.490353942908103e-05,
      "loss": 1.1947,
      "mean_token_accuracy": 0.7674643844366074,
      "num_tokens": 1228634174.0,
      "step": 18190
    },
    {
      "entropy": 0.796661950647831,
      "epoch": 5.513519654623949,
      "grad_norm": 0.1461552232503891,
      "learning_rate": 7.48745771682267e-05,
      "loss": 1.189,
      "mean_token_accuracy": 0.7670310020446778,
      "num_tokens": 1229307940.0,
      "step": 18200
    },
    {
      "entropy": 0.7936533480882645,
      "epoch": 5.516549269105506,
      "grad_norm": 0.16306835412979126,
      "learning_rate": 7.48456038120666e-05,
      "loss": 1.1884,
      "mean_token_accuracy": 0.7639216050505638,
      "num_tokens": 1229976902.0,
      "step": 18210
    },
    {
      "entropy": 0.7884961009025574,
      "epoch": 5.519578883587063,
      "grad_norm": 0.1640898585319519,
      "learning_rate": 7.481661937352431e-05,
      "loss": 1.1738,
      "mean_token_accuracy": 0.7729783669114113,
      "num_tokens": 1230657835.0,
      "step": 18220
    },
    {
      "entropy": 0.7897786647081375,
      "epoch": 5.522608498068621,
      "grad_norm": 0.15668991208076477,
      "learning_rate": 7.47876238655283e-05,
      "loss": 1.1922,
      "mean_token_accuracy": 0.770969046652317,
      "num_tokens": 1231340515.0,
      "step": 18230
    },
    {
      "entropy": 0.8003821015357971,
      "epoch": 5.525638112550178,
      "grad_norm": 0.16316929459571838,
      "learning_rate": 7.475861730101204e-05,
      "loss": 1.196,
      "mean_token_accuracy": 0.7662308365106583,
      "num_tokens": 1232007819.0,
      "step": 18240
    },
    {
      "entropy": 0.7810446426272393,
      "epoch": 5.528667727031735,
      "grad_norm": 0.1480761468410492,
      "learning_rate": 7.472959969291393e-05,
      "loss": 1.1869,
      "mean_token_accuracy": 0.7710848569869995,
      "num_tokens": 1232678244.0,
      "step": 18250
    },
    {
      "entropy": 0.7922813922166825,
      "epoch": 5.531697341513293,
      "grad_norm": 0.15463969111442566,
      "learning_rate": 7.470057105417725e-05,
      "loss": 1.1856,
      "mean_token_accuracy": 0.7781815335154534,
      "num_tokens": 1233385639.0,
      "step": 18260
    },
    {
      "entropy": 0.793801961839199,
      "epoch": 5.5347269559948495,
      "grad_norm": 0.1465470790863037,
      "learning_rate": 7.467153139775022e-05,
      "loss": 1.188,
      "mean_token_accuracy": 0.7720064729452133,
      "num_tokens": 1234076432.0,
      "step": 18270
    },
    {
      "entropy": 0.7815561160445214,
      "epoch": 5.537756570476407,
      "grad_norm": 0.15440933406352997,
      "learning_rate": 7.464248073658599e-05,
      "loss": 1.1862,
      "mean_token_accuracy": 0.7724433094263077,
      "num_tokens": 1234754653.0,
      "step": 18280
    },
    {
      "entropy": 0.7791904523968697,
      "epoch": 5.540786184957964,
      "grad_norm": 0.16058407723903656,
      "learning_rate": 7.461341908364261e-05,
      "loss": 1.1848,
      "mean_token_accuracy": 0.7732041746377945,
      "num_tokens": 1235437436.0,
      "step": 18290
    },
    {
      "entropy": 0.7996877983212471,
      "epoch": 5.543815799439521,
      "grad_norm": 0.15439097583293915,
      "learning_rate": 7.458434645188304e-05,
      "loss": 1.1997,
      "mean_token_accuracy": 0.7694524973630905,
      "num_tokens": 1236110412.0,
      "step": 18300
    },
    {
      "entropy": 0.7916013330221177,
      "epoch": 5.546845413921078,
      "grad_norm": 0.15042059123516083,
      "learning_rate": 7.45552628542751e-05,
      "loss": 1.183,
      "mean_token_accuracy": 0.7714708939194679,
      "num_tokens": 1236801923.0,
      "step": 18310
    },
    {
      "entropy": 0.7963166251778603,
      "epoch": 5.549875028402636,
      "grad_norm": 0.14816337823867798,
      "learning_rate": 7.452616830379156e-05,
      "loss": 1.1963,
      "mean_token_accuracy": 0.7706051111221314,
      "num_tokens": 1237476287.0,
      "step": 18320
    },
    {
      "entropy": 0.7852500066161155,
      "epoch": 5.552904642884193,
      "grad_norm": 0.14444084465503693,
      "learning_rate": 7.449706281341e-05,
      "loss": 1.1888,
      "mean_token_accuracy": 0.7696308091282844,
      "num_tokens": 1238147258.0,
      "step": 18330
    },
    {
      "entropy": 0.796066838502884,
      "epoch": 5.555934257365751,
      "grad_norm": 0.1868371218442917,
      "learning_rate": 7.446794639611298e-05,
      "loss": 1.1928,
      "mean_token_accuracy": 0.7696374446153641,
      "num_tokens": 1238820762.0,
      "step": 18340
    },
    {
      "entropy": 0.7891002267599105,
      "epoch": 5.558963871847308,
      "grad_norm": 0.1523510366678238,
      "learning_rate": 7.443881906488786e-05,
      "loss": 1.1798,
      "mean_token_accuracy": 0.7739343196153641,
      "num_tokens": 1239507757.0,
      "step": 18350
    },
    {
      "entropy": 0.8047911122441291,
      "epoch": 5.5619934863288645,
      "grad_norm": 0.15410935878753662,
      "learning_rate": 7.440968083272688e-05,
      "loss": 1.1976,
      "mean_token_accuracy": 0.7676858559250832,
      "num_tokens": 1240189486.0,
      "step": 18360
    },
    {
      "entropy": 0.7843504250049591,
      "epoch": 5.565023100810421,
      "grad_norm": 0.15792490541934967,
      "learning_rate": 7.438053171262715e-05,
      "loss": 1.1952,
      "mean_token_accuracy": 0.7682094275951385,
      "num_tokens": 1240846048.0,
      "step": 18370
    },
    {
      "entropy": 0.7906922951340676,
      "epoch": 5.568052715291979,
      "grad_norm": 0.1568026840686798,
      "learning_rate": 7.435137171759062e-05,
      "loss": 1.1846,
      "mean_token_accuracy": 0.7710306391119957,
      "num_tokens": 1241524774.0,
      "step": 18380
    },
    {
      "entropy": 0.7865373685956001,
      "epoch": 5.571082329773536,
      "grad_norm": 0.15185314416885376,
      "learning_rate": 7.432220086062414e-05,
      "loss": 1.1851,
      "mean_token_accuracy": 0.7710539147257804,
      "num_tokens": 1242203296.0,
      "step": 18390
    },
    {
      "entropy": 0.7973109602928161,
      "epoch": 5.574111944255094,
      "grad_norm": 0.14071327447891235,
      "learning_rate": 7.429301915473935e-05,
      "loss": 1.1965,
      "mean_token_accuracy": 0.7697199746966362,
      "num_tokens": 1242890670.0,
      "step": 18400
    },
    {
      "entropy": 0.7858362942934036,
      "epoch": 5.577141558736651,
      "grad_norm": 0.16116997599601746,
      "learning_rate": 7.426382661295275e-05,
      "loss": 1.1804,
      "mean_token_accuracy": 0.7727782040834427,
      "num_tokens": 1243573883.0,
      "step": 18410
    },
    {
      "entropy": 0.7910760179162025,
      "epoch": 5.580171173218208,
      "grad_norm": 0.16359494626522064,
      "learning_rate": 7.423462324828567e-05,
      "loss": 1.1888,
      "mean_token_accuracy": 0.77058724462986,
      "num_tokens": 1244257292.0,
      "step": 18420
    },
    {
      "entropy": 0.7906682461500167,
      "epoch": 5.583200787699766,
      "grad_norm": 0.15936478972434998,
      "learning_rate": 7.42054090737643e-05,
      "loss": 1.1946,
      "mean_token_accuracy": 0.7710719019174576,
      "num_tokens": 1244940279.0,
      "step": 18430
    },
    {
      "entropy": 0.782247719168663,
      "epoch": 5.5862304021813225,
      "grad_norm": 0.16135825216770172,
      "learning_rate": 7.417618410241959e-05,
      "loss": 1.1838,
      "mean_token_accuracy": 0.7702401608228684,
      "num_tokens": 1245610404.0,
      "step": 18440
    },
    {
      "entropy": 0.7918521538376808,
      "epoch": 5.589260016662879,
      "grad_norm": 0.1581607162952423,
      "learning_rate": 7.414694834728736e-05,
      "loss": 1.1946,
      "mean_token_accuracy": 0.7683800473809242,
      "num_tokens": 1246285274.0,
      "step": 18450
    },
    {
      "entropy": 0.7975309431552887,
      "epoch": 5.592289631144437,
      "grad_norm": 0.17319001257419586,
      "learning_rate": 7.411770182140821e-05,
      "loss": 1.1892,
      "mean_token_accuracy": 0.7679103955626487,
      "num_tokens": 1246967242.0,
      "step": 18460
    },
    {
      "entropy": 0.7915191993117332,
      "epoch": 5.595319245625994,
      "grad_norm": 0.15570372343063354,
      "learning_rate": 7.408844453782755e-05,
      "loss": 1.192,
      "mean_token_accuracy": 0.7694740876555443,
      "num_tokens": 1247643658.0,
      "step": 18470
    },
    {
      "entropy": 0.7896543279290199,
      "epoch": 5.598348860107551,
      "grad_norm": 0.1625194251537323,
      "learning_rate": 7.405917650959561e-05,
      "loss": 1.1951,
      "mean_token_accuracy": 0.769255644083023,
      "num_tokens": 1248312462.0,
      "step": 18480
    },
    {
      "entropy": 0.7988270029425621,
      "epoch": 5.601378474589109,
      "grad_norm": 0.16404829919338226,
      "learning_rate": 7.402989774976737e-05,
      "loss": 1.1969,
      "mean_token_accuracy": 0.761144632101059,
      "num_tokens": 1248965124.0,
      "step": 18490
    },
    {
      "entropy": 0.7940766930580139,
      "epoch": 5.604408089070666,
      "grad_norm": 0.15311947464942932,
      "learning_rate": 7.400060827140263e-05,
      "loss": 1.1903,
      "mean_token_accuracy": 0.7688000202178955,
      "num_tokens": 1249647917.0,
      "step": 18500
    },
    {
      "entropy": 0.7972461000084877,
      "epoch": 5.607437703552223,
      "grad_norm": 0.16149654984474182,
      "learning_rate": 7.397130808756597e-05,
      "loss": 1.1975,
      "mean_token_accuracy": 0.7670831084251404,
      "num_tokens": 1250314808.0,
      "step": 18510
    },
    {
      "entropy": 0.7870167657732964,
      "epoch": 5.6104673180337805,
      "grad_norm": 0.1804661601781845,
      "learning_rate": 7.394199721132675e-05,
      "loss": 1.1917,
      "mean_token_accuracy": 0.773274627327919,
      "num_tokens": 1251001863.0,
      "step": 18520
    },
    {
      "entropy": 0.8022281333804131,
      "epoch": 5.613496932515337,
      "grad_norm": 0.153724804520607,
      "learning_rate": 7.391267565575905e-05,
      "loss": 1.2029,
      "mean_token_accuracy": 0.7655744567513466,
      "num_tokens": 1251667408.0,
      "step": 18530
    },
    {
      "entropy": 0.7879312083125114,
      "epoch": 5.616526546996894,
      "grad_norm": 0.16561438143253326,
      "learning_rate": 7.388334343394179e-05,
      "loss": 1.193,
      "mean_token_accuracy": 0.7718574240803718,
      "num_tokens": 1252351615.0,
      "step": 18540
    },
    {
      "entropy": 0.7981021985411644,
      "epoch": 5.619556161478452,
      "grad_norm": 0.16246850788593292,
      "learning_rate": 7.385400055895857e-05,
      "loss": 1.1871,
      "mean_token_accuracy": 0.7688254058361054,
      "num_tokens": 1253036617.0,
      "step": 18550
    },
    {
      "entropy": 0.7941867128014565,
      "epoch": 5.622585775960009,
      "grad_norm": 0.15055938065052032,
      "learning_rate": 7.382464704389783e-05,
      "loss": 1.1978,
      "mean_token_accuracy": 0.7644386947154999,
      "num_tokens": 1253693899.0,
      "step": 18560
    },
    {
      "entropy": 0.8000770881772041,
      "epoch": 5.625615390441566,
      "grad_norm": 0.161276713013649,
      "learning_rate": 7.379528290185265e-05,
      "loss": 1.1939,
      "mean_token_accuracy": 0.7682380676269531,
      "num_tokens": 1254383135.0,
      "step": 18570
    },
    {
      "entropy": 0.7930928155779838,
      "epoch": 5.628645004923124,
      "grad_norm": 0.16263698041439056,
      "learning_rate": 7.376590814592095e-05,
      "loss": 1.1872,
      "mean_token_accuracy": 0.7758932799100876,
      "num_tokens": 1255091820.0,
      "step": 18580
    },
    {
      "entropy": 0.792005829513073,
      "epoch": 5.631674619404681,
      "grad_norm": 0.15446345508098602,
      "learning_rate": 7.373652278920532e-05,
      "loss": 1.1912,
      "mean_token_accuracy": 0.7698688164353371,
      "num_tokens": 1255764662.0,
      "step": 18590
    },
    {
      "entropy": 0.7881263554096222,
      "epoch": 5.634704233886238,
      "grad_norm": 0.1543988734483719,
      "learning_rate": 7.37071268448131e-05,
      "loss": 1.1978,
      "mean_token_accuracy": 0.7717945784330368,
      "num_tokens": 1256437468.0,
      "step": 18600
    },
    {
      "entropy": 0.7846297726035119,
      "epoch": 5.6377338483677955,
      "grad_norm": 0.15434779226779938,
      "learning_rate": 7.367772032585634e-05,
      "loss": 1.1856,
      "mean_token_accuracy": 0.7687615990638733,
      "num_tokens": 1257111812.0,
      "step": 18610
    },
    {
      "entropy": 0.7753724783658982,
      "epoch": 5.640763462849352,
      "grad_norm": 0.15832781791687012,
      "learning_rate": 7.364830324545187e-05,
      "loss": 1.1737,
      "mean_token_accuracy": 0.7739689543843269,
      "num_tokens": 1257785256.0,
      "step": 18620
    },
    {
      "entropy": 0.7910309210419655,
      "epoch": 5.643793077330909,
      "grad_norm": 0.15043175220489502,
      "learning_rate": 7.361887561672111e-05,
      "loss": 1.1846,
      "mean_token_accuracy": 0.7723511561751366,
      "num_tokens": 1258477035.0,
      "step": 18630
    },
    {
      "entropy": 0.7858791396021843,
      "epoch": 5.646822691812467,
      "grad_norm": 0.15828905999660492,
      "learning_rate": 7.358943745279029e-05,
      "loss": 1.1948,
      "mean_token_accuracy": 0.7697947084903717,
      "num_tokens": 1259147551.0,
      "step": 18640
    },
    {
      "entropy": 0.8024312183260918,
      "epoch": 5.649852306294024,
      "grad_norm": 0.1596386730670929,
      "learning_rate": 7.35599887667903e-05,
      "loss": 1.1964,
      "mean_token_accuracy": 0.7663578435778617,
      "num_tokens": 1259809721.0,
      "step": 18650
    },
    {
      "entropy": 0.7919725030660629,
      "epoch": 5.652881920775581,
      "grad_norm": 0.16339971125125885,
      "learning_rate": 7.353052957185673e-05,
      "loss": 1.1914,
      "mean_token_accuracy": 0.7671084478497505,
      "num_tokens": 1260475216.0,
      "step": 18660
    },
    {
      "entropy": 0.7944097578525543,
      "epoch": 5.655911535257139,
      "grad_norm": 0.15540051460266113,
      "learning_rate": 7.350105988112987e-05,
      "loss": 1.1922,
      "mean_token_accuracy": 0.7699928566813469,
      "num_tokens": 1261157376.0,
      "step": 18670
    },
    {
      "entropy": 0.7971886783838272,
      "epoch": 5.658941149738696,
      "grad_norm": 0.1517374962568283,
      "learning_rate": 7.347157970775463e-05,
      "loss": 1.1972,
      "mean_token_accuracy": 0.7699651926755905,
      "num_tokens": 1261834200.0,
      "step": 18680
    },
    {
      "entropy": 0.80126001983881,
      "epoch": 5.6619707642202535,
      "grad_norm": 0.1589142233133316,
      "learning_rate": 7.34420890648807e-05,
      "loss": 1.193,
      "mean_token_accuracy": 0.7652816787362099,
      "num_tokens": 1262514719.0,
      "step": 18690
    },
    {
      "entropy": 0.7946578770875931,
      "epoch": 5.66500037870181,
      "grad_norm": 0.15374290943145752,
      "learning_rate": 7.341258796566233e-05,
      "loss": 1.1853,
      "mean_token_accuracy": 0.7693043813109398,
      "num_tokens": 1263195002.0,
      "step": 18700
    },
    {
      "entropy": 0.7874152943491936,
      "epoch": 5.668029993183367,
      "grad_norm": 0.15570077300071716,
      "learning_rate": 7.338307642325852e-05,
      "loss": 1.189,
      "mean_token_accuracy": 0.7674408495426178,
      "num_tokens": 1263863468.0,
      "step": 18710
    },
    {
      "entropy": 0.7889522552490235,
      "epoch": 5.671059607664924,
      "grad_norm": 0.15601657330989838,
      "learning_rate": 7.335355445083288e-05,
      "loss": 1.1973,
      "mean_token_accuracy": 0.7666130721569061,
      "num_tokens": 1264526777.0,
      "step": 18720
    },
    {
      "entropy": 0.7890931233763695,
      "epoch": 5.674089222146482,
      "grad_norm": 0.16730496287345886,
      "learning_rate": 7.332402206155369e-05,
      "loss": 1.1919,
      "mean_token_accuracy": 0.7724406734108925,
      "num_tokens": 1265198857.0,
      "step": 18730
    },
    {
      "entropy": 0.7928927630186081,
      "epoch": 5.677118836628039,
      "grad_norm": 0.16443376243114471,
      "learning_rate": 7.329447926859388e-05,
      "loss": 1.188,
      "mean_token_accuracy": 0.7705048725008965,
      "num_tokens": 1265886577.0,
      "step": 18740
    },
    {
      "entropy": 0.7997891709208489,
      "epoch": 5.680148451109597,
      "grad_norm": 0.16134659945964813,
      "learning_rate": 7.326492608513098e-05,
      "loss": 1.1975,
      "mean_token_accuracy": 0.7630728602409362,
      "num_tokens": 1266559041.0,
      "step": 18750
    },
    {
      "entropy": 0.79414232224226,
      "epoch": 5.683178065591154,
      "grad_norm": 0.16661733388900757,
      "learning_rate": 7.323536252434724e-05,
      "loss": 1.1932,
      "mean_token_accuracy": 0.7700477302074432,
      "num_tokens": 1267238094.0,
      "step": 18760
    },
    {
      "entropy": 0.7935743287205697,
      "epoch": 5.686207680072711,
      "grad_norm": 0.15668271481990814,
      "learning_rate": 7.320578859942945e-05,
      "loss": 1.1891,
      "mean_token_accuracy": 0.7664597243070602,
      "num_tokens": 1267909228.0,
      "step": 18770
    },
    {
      "entropy": 0.7928174495697021,
      "epoch": 5.6892372945542675,
      "grad_norm": 0.16351042687892914,
      "learning_rate": 7.317620432356907e-05,
      "loss": 1.1862,
      "mean_token_accuracy": 0.7701626896858216,
      "num_tokens": 1268588648.0,
      "step": 18780
    },
    {
      "entropy": 0.8047268077731132,
      "epoch": 5.692266909035825,
      "grad_norm": 0.15795977413654327,
      "learning_rate": 7.314660970996217e-05,
      "loss": 1.2056,
      "mean_token_accuracy": 0.7643048569560051,
      "num_tokens": 1269254322.0,
      "step": 18790
    },
    {
      "entropy": 0.8028568893671035,
      "epoch": 5.695296523517382,
      "grad_norm": 0.15192359685897827,
      "learning_rate": 7.311700477180944e-05,
      "loss": 1.2016,
      "mean_token_accuracy": 0.7660802662372589,
      "num_tokens": 1269931439.0,
      "step": 18800
    },
    {
      "entropy": 0.7914088234305382,
      "epoch": 5.69832613799894,
      "grad_norm": 0.1599586009979248,
      "learning_rate": 7.308738952231614e-05,
      "loss": 1.1873,
      "mean_token_accuracy": 0.7692271962761879,
      "num_tokens": 1270601240.0,
      "step": 18810
    },
    {
      "entropy": 0.7939092203974724,
      "epoch": 5.701355752480497,
      "grad_norm": 0.14729230105876923,
      "learning_rate": 7.305776397469214e-05,
      "loss": 1.1926,
      "mean_token_accuracy": 0.7719800055027009,
      "num_tokens": 1271291215.0,
      "step": 18820
    },
    {
      "entropy": 0.7896327704191208,
      "epoch": 5.704385366962054,
      "grad_norm": 0.1626734584569931,
      "learning_rate": 7.302812814215198e-05,
      "loss": 1.1893,
      "mean_token_accuracy": 0.767254875600338,
      "num_tokens": 1271959921.0,
      "step": 18830
    },
    {
      "entropy": 0.7918659880757332,
      "epoch": 5.707414981443611,
      "grad_norm": 0.14825868606567383,
      "learning_rate": 7.299848203791467e-05,
      "loss": 1.1853,
      "mean_token_accuracy": 0.7726846173405647,
      "num_tokens": 1272651250.0,
      "step": 18840
    },
    {
      "entropy": 0.789652743935585,
      "epoch": 5.710444595925169,
      "grad_norm": 0.16363964974880219,
      "learning_rate": 7.296882567520385e-05,
      "loss": 1.1801,
      "mean_token_accuracy": 0.7706850856542588,
      "num_tokens": 1273330270.0,
      "step": 18850
    },
    {
      "entropy": 0.7887715801596642,
      "epoch": 5.7134742104067255,
      "grad_norm": 0.15937940776348114,
      "learning_rate": 7.293915906724778e-05,
      "loss": 1.1944,
      "mean_token_accuracy": 0.7677953779697418,
      "num_tokens": 1273998084.0,
      "step": 18860
    },
    {
      "entropy": 0.7939240783452988,
      "epoch": 5.716503824888283,
      "grad_norm": 0.15611006319522858,
      "learning_rate": 7.29094822272792e-05,
      "loss": 1.1955,
      "mean_token_accuracy": 0.7675396710634231,
      "num_tokens": 1274670896.0,
      "step": 18870
    },
    {
      "entropy": 0.781504075229168,
      "epoch": 5.71953343936984,
      "grad_norm": 0.17100565135478973,
      "learning_rate": 7.287979516853553e-05,
      "loss": 1.1761,
      "mean_token_accuracy": 0.7743938907980918,
      "num_tokens": 1275342835.0,
      "step": 18880
    },
    {
      "entropy": 0.7951151594519615,
      "epoch": 5.722563053851397,
      "grad_norm": 0.16005516052246094,
      "learning_rate": 7.285009790425863e-05,
      "loss": 1.1931,
      "mean_token_accuracy": 0.7707011327147484,
      "num_tokens": 1276026509.0,
      "step": 18890
    },
    {
      "entropy": 0.7936810344457627,
      "epoch": 5.725592668332955,
      "grad_norm": 0.166903555393219,
      "learning_rate": 7.282039044769499e-05,
      "loss": 1.1979,
      "mean_token_accuracy": 0.769514313340187,
      "num_tokens": 1276706121.0,
      "step": 18900
    },
    {
      "entropy": 0.7938599124550819,
      "epoch": 5.728622282814512,
      "grad_norm": 0.1707628220319748,
      "learning_rate": 7.279067281209563e-05,
      "loss": 1.1941,
      "mean_token_accuracy": 0.7680787861347198,
      "num_tokens": 1277378551.0,
      "step": 18910
    },
    {
      "entropy": 0.7916633740067482,
      "epoch": 5.731651897296069,
      "grad_norm": 0.15222841501235962,
      "learning_rate": 7.276094501071607e-05,
      "loss": 1.1871,
      "mean_token_accuracy": 0.772497744858265,
      "num_tokens": 1278072838.0,
      "step": 18920
    },
    {
      "entropy": 0.7980836063623429,
      "epoch": 5.734681511777627,
      "grad_norm": 0.16312545537948608,
      "learning_rate": 7.273120705681643e-05,
      "loss": 1.1988,
      "mean_token_accuracy": 0.7672946929931641,
      "num_tokens": 1278736216.0,
      "step": 18930
    },
    {
      "entropy": 0.7914466798305512,
      "epoch": 5.737711126259184,
      "grad_norm": 0.15629099309444427,
      "learning_rate": 7.27014589636613e-05,
      "loss": 1.1896,
      "mean_token_accuracy": 0.7754410624504089,
      "num_tokens": 1279423182.0,
      "step": 18940
    },
    {
      "entropy": 0.7761142745614051,
      "epoch": 5.7407407407407405,
      "grad_norm": 0.16202078759670258,
      "learning_rate": 7.267170074451983e-05,
      "loss": 1.1829,
      "mean_token_accuracy": 0.7714239031076431,
      "num_tokens": 1280103864.0,
      "step": 18950
    },
    {
      "entropy": 0.8018604129552841,
      "epoch": 5.743770355222298,
      "grad_norm": 0.15807020664215088,
      "learning_rate": 7.264193241266568e-05,
      "loss": 1.1949,
      "mean_token_accuracy": 0.7640351712703705,
      "num_tokens": 1280776689.0,
      "step": 18960
    },
    {
      "entropy": 0.7881122648715972,
      "epoch": 5.746799969703855,
      "grad_norm": 0.15949994325637817,
      "learning_rate": 7.2612153981377e-05,
      "loss": 1.185,
      "mean_token_accuracy": 0.7737415760755539,
      "num_tokens": 1281465727.0,
      "step": 18970
    },
    {
      "entropy": 0.797203715145588,
      "epoch": 5.749829584185412,
      "grad_norm": 0.15690580010414124,
      "learning_rate": 7.258236546393645e-05,
      "loss": 1.1887,
      "mean_token_accuracy": 0.7698037967085838,
      "num_tokens": 1282148924.0,
      "step": 18980
    },
    {
      "entropy": 0.8052161574363709,
      "epoch": 5.75285919866697,
      "grad_norm": 0.16531206667423248,
      "learning_rate": 7.255256687363124e-05,
      "loss": 1.1977,
      "mean_token_accuracy": 0.7619235455989838,
      "num_tokens": 1282816960.0,
      "step": 18990
    },
    {
      "entropy": 0.7997000426054001,
      "epoch": 5.755888813148527,
      "grad_norm": 0.152301624417305,
      "learning_rate": 7.252275822375301e-05,
      "loss": 1.2022,
      "mean_token_accuracy": 0.7650304600596428,
      "num_tokens": 1283487423.0,
      "step": 19000
    },
    {
      "entropy": 0.8115334004163742,
      "epoch": 5.758918427630084,
      "grad_norm": 0.16885791718959808,
      "learning_rate": 7.249293952759789e-05,
      "loss": 1.2024,
      "mean_token_accuracy": 0.7626567095518112,
      "num_tokens": 1284162524.0,
      "step": 19010
    },
    {
      "entropy": 0.8028330817818642,
      "epoch": 5.761948042111642,
      "grad_norm": 0.15730434656143188,
      "learning_rate": 7.246311079846654e-05,
      "loss": 1.2063,
      "mean_token_accuracy": 0.7625677794218063,
      "num_tokens": 1284825914.0,
      "step": 19020
    },
    {
      "entropy": 0.7898641049861908,
      "epoch": 5.7649776565931985,
      "grad_norm": 0.16985639929771423,
      "learning_rate": 7.243327204966406e-05,
      "loss": 1.1799,
      "mean_token_accuracy": 0.7730851233005523,
      "num_tokens": 1285507408.0,
      "step": 19030
    },
    {
      "entropy": 0.7899454146623611,
      "epoch": 5.768007271074755,
      "grad_norm": 0.16257216036319733,
      "learning_rate": 7.24034232945e-05,
      "loss": 1.1978,
      "mean_token_accuracy": 0.7655003413558006,
      "num_tokens": 1286160339.0,
      "step": 19040
    },
    {
      "entropy": 0.7982629612088203,
      "epoch": 5.771036885556313,
      "grad_norm": 0.15875552594661713,
      "learning_rate": 7.237356454628841e-05,
      "loss": 1.2007,
      "mean_token_accuracy": 0.7625041589140892,
      "num_tokens": 1286820905.0,
      "step": 19050
    },
    {
      "entropy": 0.8031689330935479,
      "epoch": 5.77406650003787,
      "grad_norm": 0.1723564863204956,
      "learning_rate": 7.234369581834783e-05,
      "loss": 1.199,
      "mean_token_accuracy": 0.7648579478263855,
      "num_tokens": 1287494716.0,
      "step": 19060
    },
    {
      "entropy": 0.7959000512957572,
      "epoch": 5.777096114519427,
      "grad_norm": 0.15248917043209076,
      "learning_rate": 7.231381712400115e-05,
      "loss": 1.1898,
      "mean_token_accuracy": 0.7684626832604409,
      "num_tokens": 1288172111.0,
      "step": 19070
    },
    {
      "entropy": 0.796762265264988,
      "epoch": 5.780125729000985,
      "grad_norm": 0.1766333132982254,
      "learning_rate": 7.228392847657577e-05,
      "loss": 1.1982,
      "mean_token_accuracy": 0.7703196331858635,
      "num_tokens": 1288845492.0,
      "step": 19080
    },
    {
      "entropy": 0.7944855257868767,
      "epoch": 5.783155343482542,
      "grad_norm": 0.1556379795074463,
      "learning_rate": 7.225402988940352e-05,
      "loss": 1.1894,
      "mean_token_accuracy": 0.7687153145670891,
      "num_tokens": 1289525200.0,
      "step": 19090
    },
    {
      "entropy": 0.7888134881854058,
      "epoch": 5.786184957964099,
      "grad_norm": 0.15383945405483246,
      "learning_rate": 7.222412137582071e-05,
      "loss": 1.187,
      "mean_token_accuracy": 0.7703951612114907,
      "num_tokens": 1290195608.0,
      "step": 19100
    },
    {
      "entropy": 0.7984809562563896,
      "epoch": 5.7892145724456565,
      "grad_norm": 0.2070225477218628,
      "learning_rate": 7.219420294916798e-05,
      "loss": 1.1874,
      "mean_token_accuracy": 0.7712823674082756,
      "num_tokens": 1290892584.0,
      "step": 19110
    },
    {
      "entropy": 0.784035250544548,
      "epoch": 5.792244186927213,
      "grad_norm": 0.16227585077285767,
      "learning_rate": 7.216427462279047e-05,
      "loss": 1.1972,
      "mean_token_accuracy": 0.7716500088572502,
      "num_tokens": 1291561531.0,
      "step": 19120
    },
    {
      "entropy": 0.7916143849492073,
      "epoch": 5.79527380140877,
      "grad_norm": 0.15463599562644958,
      "learning_rate": 7.213433641003771e-05,
      "loss": 1.1966,
      "mean_token_accuracy": 0.7703131452202797,
      "num_tokens": 1292238265.0,
      "step": 19130
    },
    {
      "entropy": 0.795188945531845,
      "epoch": 5.798303415890328,
      "grad_norm": 0.15650375187397003,
      "learning_rate": 7.210438832426363e-05,
      "loss": 1.194,
      "mean_token_accuracy": 0.76949143409729,
      "num_tokens": 1292912695.0,
      "step": 19140
    },
    {
      "entropy": 0.7866457670927047,
      "epoch": 5.801333030371885,
      "grad_norm": 0.15094313025474548,
      "learning_rate": 7.207443037882659e-05,
      "loss": 1.1753,
      "mean_token_accuracy": 0.7699485033750534,
      "num_tokens": 1293607617.0,
      "step": 19150
    },
    {
      "entropy": 0.7934285506606102,
      "epoch": 5.804362644853443,
      "grad_norm": 0.159365713596344,
      "learning_rate": 7.204446258708931e-05,
      "loss": 1.1979,
      "mean_token_accuracy": 0.7682175755500793,
      "num_tokens": 1294279292.0,
      "step": 19160
    },
    {
      "entropy": 0.7908381313085556,
      "epoch": 5.807392259335,
      "grad_norm": 0.14878462255001068,
      "learning_rate": 7.201448496241896e-05,
      "loss": 1.1952,
      "mean_token_accuracy": 0.767137984931469,
      "num_tokens": 1294950994.0,
      "step": 19170
    },
    {
      "entropy": 0.7872956618666649,
      "epoch": 5.810421873816557,
      "grad_norm": 0.1709751933813095,
      "learning_rate": 7.198449751818702e-05,
      "loss": 1.1906,
      "mean_token_accuracy": 0.7710516393184662,
      "num_tokens": 1295627328.0,
      "step": 19180
    },
    {
      "entropy": 0.7985885605216027,
      "epoch": 5.813451488298114,
      "grad_norm": 0.15704365074634552,
      "learning_rate": 7.195450026776943e-05,
      "loss": 1.1973,
      "mean_token_accuracy": 0.7730544298887253,
      "num_tokens": 1296327150.0,
      "step": 19190
    },
    {
      "entropy": 0.795363686978817,
      "epoch": 5.8164811027796715,
      "grad_norm": 0.15271684527397156,
      "learning_rate": 7.192449322454642e-05,
      "loss": 1.1979,
      "mean_token_accuracy": 0.7606057360768318,
      "num_tokens": 1296982357.0,
      "step": 19200
    },
    {
      "entropy": 0.78994672447443,
      "epoch": 5.819510717261228,
      "grad_norm": 0.16777105629444122,
      "learning_rate": 7.189447640190267e-05,
      "loss": 1.1892,
      "mean_token_accuracy": 0.7716968283057213,
      "num_tokens": 1297670726.0,
      "step": 19210
    },
    {
      "entropy": 0.7922337487339973,
      "epoch": 5.822540331742786,
      "grad_norm": 0.15544579923152924,
      "learning_rate": 7.186444981322715e-05,
      "loss": 1.1848,
      "mean_token_accuracy": 0.7683053568005562,
      "num_tokens": 1298347347.0,
      "step": 19220
    },
    {
      "entropy": 0.791400583088398,
      "epoch": 5.825569946224343,
      "grad_norm": 0.1687878668308258,
      "learning_rate": 7.183441347191326e-05,
      "loss": 1.1846,
      "mean_token_accuracy": 0.7777029946446419,
      "num_tokens": 1299056544.0,
      "step": 19230
    },
    {
      "entropy": 0.7915820553898811,
      "epoch": 5.8285995607059,
      "grad_norm": 0.14781229197978973,
      "learning_rate": 7.18043673913587e-05,
      "loss": 1.1927,
      "mean_token_accuracy": 0.7701451942324639,
      "num_tokens": 1299743640.0,
      "step": 19240
    },
    {
      "entropy": 0.7932293877005577,
      "epoch": 5.831629175187457,
      "grad_norm": 0.1546233743429184,
      "learning_rate": 7.17743115849655e-05,
      "loss": 1.1925,
      "mean_token_accuracy": 0.7677518740296364,
      "num_tokens": 1300418189.0,
      "step": 19250
    },
    {
      "entropy": 0.7910405933856964,
      "epoch": 5.834658789669015,
      "grad_norm": 0.16984455287456512,
      "learning_rate": 7.174424606614008e-05,
      "loss": 1.1863,
      "mean_token_accuracy": 0.7770651251077652,
      "num_tokens": 1301116441.0,
      "step": 19260
    },
    {
      "entropy": 0.8009702622890472,
      "epoch": 5.837688404150572,
      "grad_norm": 0.15600861608982086,
      "learning_rate": 7.171417084829315e-05,
      "loss": 1.1978,
      "mean_token_accuracy": 0.7650102362036705,
      "num_tokens": 1301784912.0,
      "step": 19270
    },
    {
      "entropy": 0.7929332584142685,
      "epoch": 5.8407180186321295,
      "grad_norm": 0.15729230642318726,
      "learning_rate": 7.168408594483975e-05,
      "loss": 1.1938,
      "mean_token_accuracy": 0.7671513259410858,
      "num_tokens": 1302460911.0,
      "step": 19280
    },
    {
      "entropy": 0.7850655108690262,
      "epoch": 5.843747633113686,
      "grad_norm": 0.15478479862213135,
      "learning_rate": 7.16539913691993e-05,
      "loss": 1.1817,
      "mean_token_accuracy": 0.7702582538127899,
      "num_tokens": 1303139014.0,
      "step": 19290
    },
    {
      "entropy": 0.7898176908493042,
      "epoch": 5.846777247595243,
      "grad_norm": 0.15592904388904572,
      "learning_rate": 7.162388713479544e-05,
      "loss": 1.1866,
      "mean_token_accuracy": 0.7677637532353401,
      "num_tokens": 1303808926.0,
      "step": 19300
    },
    {
      "entropy": 0.790964238345623,
      "epoch": 5.8498068620768,
      "grad_norm": 0.1592186987400055,
      "learning_rate": 7.159377325505617e-05,
      "loss": 1.1929,
      "mean_token_accuracy": 0.7658483535051346,
      "num_tokens": 1304468962.0,
      "step": 19310
    },
    {
      "entropy": 0.7826980590820313,
      "epoch": 5.852836476558358,
      "grad_norm": 0.15509718656539917,
      "learning_rate": 7.15636497434138e-05,
      "loss": 1.1879,
      "mean_token_accuracy": 0.7703955799341202,
      "num_tokens": 1305147093.0,
      "step": 19320
    },
    {
      "entropy": 0.7993157789111137,
      "epoch": 5.855866091039915,
      "grad_norm": 0.1570509821176529,
      "learning_rate": 7.153351661330492e-05,
      "loss": 1.2041,
      "mean_token_accuracy": 0.7632969990372658,
      "num_tokens": 1305810051.0,
      "step": 19330
    },
    {
      "entropy": 0.7967530325055122,
      "epoch": 5.858895705521473,
      "grad_norm": 0.16697965562343597,
      "learning_rate": 7.150337387817042e-05,
      "loss": 1.1946,
      "mean_token_accuracy": 0.7688148751854896,
      "num_tokens": 1306484715.0,
      "step": 19340
    },
    {
      "entropy": 0.7971322774887085,
      "epoch": 5.86192532000303,
      "grad_norm": 0.16007335484027863,
      "learning_rate": 7.147322155145544e-05,
      "loss": 1.1977,
      "mean_token_accuracy": 0.7678506687283516,
      "num_tokens": 1307157150.0,
      "step": 19350
    },
    {
      "entropy": 0.7920721471309662,
      "epoch": 5.864954934484587,
      "grad_norm": 0.15300531685352325,
      "learning_rate": 7.144305964660947e-05,
      "loss": 1.1928,
      "mean_token_accuracy": 0.7665317371487618,
      "num_tokens": 1307823563.0,
      "step": 19360
    },
    {
      "entropy": 0.7872507840394973,
      "epoch": 5.867984548966144,
      "grad_norm": 0.15557889640331268,
      "learning_rate": 7.14128881770862e-05,
      "loss": 1.1988,
      "mean_token_accuracy": 0.7663601219654084,
      "num_tokens": 1308485053.0,
      "step": 19370
    },
    {
      "entropy": 0.8012245669960976,
      "epoch": 5.871014163447701,
      "grad_norm": 0.15355554223060608,
      "learning_rate": 7.138270715634362e-05,
      "loss": 1.1988,
      "mean_token_accuracy": 0.7701532348990441,
      "num_tokens": 1309169126.0,
      "step": 19380
    },
    {
      "entropy": 0.7840667784214019,
      "epoch": 5.874043777929258,
      "grad_norm": 0.15598563849925995,
      "learning_rate": 7.135251659784396e-05,
      "loss": 1.1861,
      "mean_token_accuracy": 0.769740016758442,
      "num_tokens": 1309849883.0,
      "step": 19390
    },
    {
      "entropy": 0.7877208113670349,
      "epoch": 5.877073392410816,
      "grad_norm": 0.15426141023635864,
      "learning_rate": 7.132231651505377e-05,
      "loss": 1.2014,
      "mean_token_accuracy": 0.7657011389732361,
      "num_tokens": 1310511116.0,
      "step": 19400
    },
    {
      "entropy": 0.7917147055268288,
      "epoch": 5.880103006892373,
      "grad_norm": 0.15134041011333466,
      "learning_rate": 7.129210692144374e-05,
      "loss": 1.1986,
      "mean_token_accuracy": 0.7654175668954849,
      "num_tokens": 1311168820.0,
      "step": 19410
    },
    {
      "entropy": 0.7980439901351929,
      "epoch": 5.88313262137393,
      "grad_norm": 0.1536293923854828,
      "learning_rate": 7.126188783048888e-05,
      "loss": 1.2001,
      "mean_token_accuracy": 0.7672934398055077,
      "num_tokens": 1311843261.0,
      "step": 19420
    },
    {
      "entropy": 0.7802516609430313,
      "epoch": 5.886162235855488,
      "grad_norm": 0.1546633094549179,
      "learning_rate": 7.123165925566842e-05,
      "loss": 1.1839,
      "mean_token_accuracy": 0.7746330320835113,
      "num_tokens": 1312533249.0,
      "step": 19430
    },
    {
      "entropy": 0.7927233129739761,
      "epoch": 5.889191850337045,
      "grad_norm": 0.15313683450222015,
      "learning_rate": 7.120142121046581e-05,
      "loss": 1.1906,
      "mean_token_accuracy": 0.7654723390936852,
      "num_tokens": 1313197685.0,
      "step": 19440
    },
    {
      "entropy": 0.7918976411223412,
      "epoch": 5.8922214648186015,
      "grad_norm": 0.1636413335800171,
      "learning_rate": 7.117117370836873e-05,
      "loss": 1.1846,
      "mean_token_accuracy": 0.7683885022997856,
      "num_tokens": 1313872546.0,
      "step": 19450
    },
    {
      "entropy": 0.7887762054800987,
      "epoch": 5.895251079300159,
      "grad_norm": 0.163103386759758,
      "learning_rate": 7.11409167628691e-05,
      "loss": 1.195,
      "mean_token_accuracy": 0.7676061347126961,
      "num_tokens": 1314542011.0,
      "step": 19460
    },
    {
      "entropy": 0.7964710831642151,
      "epoch": 5.898280693781716,
      "grad_norm": 0.1580525040626526,
      "learning_rate": 7.1110650387463e-05,
      "loss": 1.19,
      "mean_token_accuracy": 0.7661490544676781,
      "num_tokens": 1315215987.0,
      "step": 19470
    },
    {
      "entropy": 0.7858775854110718,
      "epoch": 5.901310308263273,
      "grad_norm": 0.16609440743923187,
      "learning_rate": 7.108037459565078e-05,
      "loss": 1.1866,
      "mean_token_accuracy": 0.7727140128612519,
      "num_tokens": 1315903511.0,
      "step": 19480
    },
    {
      "entropy": 0.8043234646320343,
      "epoch": 5.904339922744831,
      "grad_norm": 0.16437289118766785,
      "learning_rate": 7.105008940093694e-05,
      "loss": 1.2015,
      "mean_token_accuracy": 0.7644197762012481,
      "num_tokens": 1316581139.0,
      "step": 19490
    },
    {
      "entropy": 0.7888402491807938,
      "epoch": 5.907369537226388,
      "grad_norm": 0.15607315301895142,
      "learning_rate": 7.101979481683018e-05,
      "loss": 1.1898,
      "mean_token_accuracy": 0.7767038270831108,
      "num_tokens": 1317282784.0,
      "step": 19500
    },
    {
      "entropy": 0.7878912091255188,
      "epoch": 5.910399151707945,
      "grad_norm": 0.16913576424121857,
      "learning_rate": 7.098949085684342e-05,
      "loss": 1.19,
      "mean_token_accuracy": 0.7655957534909248,
      "num_tokens": 1317939963.0,
      "step": 19510
    },
    {
      "entropy": 0.785400713980198,
      "epoch": 5.913428766189503,
      "grad_norm": 0.1648530811071396,
      "learning_rate": 7.095917753449373e-05,
      "loss": 1.1815,
      "mean_token_accuracy": 0.7728199392557145,
      "num_tokens": 1318624430.0,
      "step": 19520
    },
    {
      "entropy": 0.7962307214736939,
      "epoch": 5.91645838067106,
      "grad_norm": 0.16219139099121094,
      "learning_rate": 7.092885486330239e-05,
      "loss": 1.1963,
      "mean_token_accuracy": 0.7642073258757591,
      "num_tokens": 1319285362.0,
      "step": 19530
    },
    {
      "entropy": 0.7846979156136513,
      "epoch": 5.9194879951526165,
      "grad_norm": 0.1602168083190918,
      "learning_rate": 7.089852285679479e-05,
      "loss": 1.181,
      "mean_token_accuracy": 0.766859495639801,
      "num_tokens": 1319945502.0,
      "step": 19540
    },
    {
      "entropy": 0.786933034658432,
      "epoch": 5.922517609634174,
      "grad_norm": 0.16247202455997467,
      "learning_rate": 7.086818152850055e-05,
      "loss": 1.1898,
      "mean_token_accuracy": 0.770433597266674,
      "num_tokens": 1320615059.0,
      "step": 19550
    },
    {
      "entropy": 0.794632551074028,
      "epoch": 5.925547224115731,
      "grad_norm": 0.16537994146347046,
      "learning_rate": 7.083783089195341e-05,
      "loss": 1.1899,
      "mean_token_accuracy": 0.767238911986351,
      "num_tokens": 1321296288.0,
      "step": 19560
    },
    {
      "entropy": 0.7873178437352181,
      "epoch": 5.928576838597288,
      "grad_norm": 0.15839438140392303,
      "learning_rate": 7.08074709606913e-05,
      "loss": 1.1878,
      "mean_token_accuracy": 0.7728889375925064,
      "num_tokens": 1321973030.0,
      "step": 19570
    },
    {
      "entropy": 0.7870238736271858,
      "epoch": 5.931606453078846,
      "grad_norm": 0.16366255283355713,
      "learning_rate": 7.077710174825623e-05,
      "loss": 1.19,
      "mean_token_accuracy": 0.7699470698833466,
      "num_tokens": 1322645930.0,
      "step": 19580
    },
    {
      "entropy": 0.7962046086788177,
      "epoch": 5.934636067560403,
      "grad_norm": 0.17779672145843506,
      "learning_rate": 7.07467232681944e-05,
      "loss": 1.2058,
      "mean_token_accuracy": 0.7600207984447479,
      "num_tokens": 1323287817.0,
      "step": 19590
    },
    {
      "entropy": 0.7926531061530113,
      "epoch": 5.93766568204196,
      "grad_norm": 0.15745460987091064,
      "learning_rate": 7.071633553405614e-05,
      "loss": 1.1926,
      "mean_token_accuracy": 0.7657125771045685,
      "num_tokens": 1323957898.0,
      "step": 19600
    },
    {
      "entropy": 0.7850239261984825,
      "epoch": 5.940695296523518,
      "grad_norm": 0.15340839326381683,
      "learning_rate": 7.06859385593959e-05,
      "loss": 1.1834,
      "mean_token_accuracy": 0.771136499941349,
      "num_tokens": 1324638511.0,
      "step": 19610
    },
    {
      "entropy": 0.7995823636651039,
      "epoch": 5.9437249110050745,
      "grad_norm": 0.14871494472026825,
      "learning_rate": 7.065553235777222e-05,
      "loss": 1.2005,
      "mean_token_accuracy": 0.7643187955021858,
      "num_tokens": 1325312190.0,
      "step": 19620
    },
    {
      "entropy": 0.7911097630858421,
      "epoch": 5.946754525486632,
      "grad_norm": 0.15272296965122223,
      "learning_rate": 7.062511694274783e-05,
      "loss": 1.1918,
      "mean_token_accuracy": 0.7669488146901131,
      "num_tokens": 1325992657.0,
      "step": 19630
    },
    {
      "entropy": 0.7895109668374062,
      "epoch": 5.949784139968189,
      "grad_norm": 0.16099727153778076,
      "learning_rate": 7.05946923278895e-05,
      "loss": 1.1893,
      "mean_token_accuracy": 0.7747374445199966,
      "num_tokens": 1326677656.0,
      "step": 19640
    },
    {
      "entropy": 0.7924715891480446,
      "epoch": 5.952813754449746,
      "grad_norm": 0.16094346344470978,
      "learning_rate": 7.056425852676815e-05,
      "loss": 1.1934,
      "mean_token_accuracy": 0.7675824925303459,
      "num_tokens": 1327354617.0,
      "step": 19650
    },
    {
      "entropy": 0.7775761127471924,
      "epoch": 5.955843368931303,
      "grad_norm": 0.15954801440238953,
      "learning_rate": 7.053381555295875e-05,
      "loss": 1.1722,
      "mean_token_accuracy": 0.777365431189537,
      "num_tokens": 1328049564.0,
      "step": 19660
    },
    {
      "entropy": 0.7921777665615082,
      "epoch": 5.958872983412861,
      "grad_norm": 0.1612180471420288,
      "learning_rate": 7.050336342004041e-05,
      "loss": 1.1975,
      "mean_token_accuracy": 0.7655911773443222,
      "num_tokens": 1328712963.0,
      "step": 19670
    },
    {
      "entropy": 0.8030157566070557,
      "epoch": 5.961902597894418,
      "grad_norm": 0.1663007289171219,
      "learning_rate": 7.047290214159632e-05,
      "loss": 1.2032,
      "mean_token_accuracy": 0.7662871927022934,
      "num_tokens": 1329379221.0,
      "step": 19680
    },
    {
      "entropy": 0.784817336499691,
      "epoch": 5.964932212375976,
      "grad_norm": 0.15503285825252533,
      "learning_rate": 7.04424317312137e-05,
      "loss": 1.1892,
      "mean_token_accuracy": 0.7714136093854904,
      "num_tokens": 1330057020.0,
      "step": 19690
    },
    {
      "entropy": 0.7827223688364029,
      "epoch": 5.9679618268575325,
      "grad_norm": 0.15545709431171417,
      "learning_rate": 7.04119522024839e-05,
      "loss": 1.1907,
      "mean_token_accuracy": 0.7702628016471863,
      "num_tokens": 1330722574.0,
      "step": 19700
    },
    {
      "entropy": 0.7823386698961258,
      "epoch": 5.970991441339089,
      "grad_norm": 0.154511496424675,
      "learning_rate": 7.038146356900229e-05,
      "loss": 1.1786,
      "mean_token_accuracy": 0.7735481590032578,
      "num_tokens": 1331397459.0,
      "step": 19710
    },
    {
      "entropy": 0.7853929430246354,
      "epoch": 5.974021055820646,
      "grad_norm": 0.16265028715133667,
      "learning_rate": 7.035096584436835e-05,
      "loss": 1.1951,
      "mean_token_accuracy": 0.7678526639938354,
      "num_tokens": 1332060462.0,
      "step": 19720
    },
    {
      "entropy": 0.7929315254092216,
      "epoch": 5.977050670302204,
      "grad_norm": 0.15164761245250702,
      "learning_rate": 7.032045904218559e-05,
      "loss": 1.1865,
      "mean_token_accuracy": 0.7695974588394165,
      "num_tokens": 1332737535.0,
      "step": 19730
    },
    {
      "entropy": 0.7914792761206627,
      "epoch": 5.980080284783761,
      "grad_norm": 0.1563020944595337,
      "learning_rate": 7.028994317606156e-05,
      "loss": 1.195,
      "mean_token_accuracy": 0.7703168600797653,
      "num_tokens": 1333424842.0,
      "step": 19740
    },
    {
      "entropy": 0.7891993552446366,
      "epoch": 5.983109899265319,
      "grad_norm": 0.15779948234558105,
      "learning_rate": 7.025941825960786e-05,
      "loss": 1.1873,
      "mean_token_accuracy": 0.7709450766444206,
      "num_tokens": 1334100599.0,
      "step": 19750
    },
    {
      "entropy": 0.7896355286240577,
      "epoch": 5.986139513746876,
      "grad_norm": 0.1670207530260086,
      "learning_rate": 7.022888430644015e-05,
      "loss": 1.1927,
      "mean_token_accuracy": 0.765450245141983,
      "num_tokens": 1334753309.0,
      "step": 19760
    },
    {
      "entropy": 0.7893132537603378,
      "epoch": 5.989169128228433,
      "grad_norm": 0.14903050661087036,
      "learning_rate": 7.019834133017806e-05,
      "loss": 1.1859,
      "mean_token_accuracy": 0.7656871825456619,
      "num_tokens": 1335410874.0,
      "step": 19770
    },
    {
      "entropy": 0.7974526420235634,
      "epoch": 5.9921987427099905,
      "grad_norm": 0.1609318107366562,
      "learning_rate": 7.016778934444534e-05,
      "loss": 1.1953,
      "mean_token_accuracy": 0.7665823966264724,
      "num_tokens": 1336087207.0,
      "step": 19780
    },
    {
      "entropy": 0.781538225710392,
      "epoch": 5.9952283571915475,
      "grad_norm": 0.1576438695192337,
      "learning_rate": 7.013722836286964e-05,
      "loss": 1.18,
      "mean_token_accuracy": 0.7733356311917305,
      "num_tokens": 1336780452.0,
      "step": 19790
    },
    {
      "entropy": 0.7825202971696854,
      "epoch": 5.998257971673104,
      "grad_norm": 0.16125264763832092,
      "learning_rate": 7.010665839908275e-05,
      "loss": 1.1815,
      "mean_token_accuracy": 0.771214810013771,
      "num_tokens": 1337455601.0,
      "step": 19800
    },
    {
      "entropy": 0.7791619728773068,
      "epoch": 6.001211845792623,
      "grad_norm": 0.15400418639183044,
      "learning_rate": 7.007607946672037e-05,
      "loss": 1.1656,
      "mean_token_accuracy": 0.7763047447571387,
      "num_tokens": 1338128078.0,
      "step": 19810
    },
    {
      "entropy": 0.7685077264904976,
      "epoch": 6.00424146027418,
      "grad_norm": 0.1843770295381546,
      "learning_rate": 7.004549157942224e-05,
      "loss": 1.1758,
      "mean_token_accuracy": 0.7730411633849144,
      "num_tokens": 1338802408.0,
      "step": 19820
    },
    {
      "entropy": 0.764088423550129,
      "epoch": 6.007271074755737,
      "grad_norm": 0.1830776333808899,
      "learning_rate": 7.001489475083208e-05,
      "loss": 1.1551,
      "mean_token_accuracy": 0.7786645635962486,
      "num_tokens": 1339493060.0,
      "step": 19830
    },
    {
      "entropy": 0.7704333201050758,
      "epoch": 6.010300689237295,
      "grad_norm": 0.1900692731142044,
      "learning_rate": 6.998428899459761e-05,
      "loss": 1.1623,
      "mean_token_accuracy": 0.771507452428341,
      "num_tokens": 1340164718.0,
      "step": 19840
    },
    {
      "entropy": 0.7672426775097847,
      "epoch": 6.013330303718852,
      "grad_norm": 0.17119291424751282,
      "learning_rate": 6.995367432437054e-05,
      "loss": 1.1562,
      "mean_token_accuracy": 0.7723920196294785,
      "num_tokens": 1340836623.0,
      "step": 19850
    },
    {
      "entropy": 0.7724898129701614,
      "epoch": 6.016359918200409,
      "grad_norm": 0.1718260645866394,
      "learning_rate": 6.992305075380654e-05,
      "loss": 1.1679,
      "mean_token_accuracy": 0.7711360037326813,
      "num_tokens": 1341513959.0,
      "step": 19860
    },
    {
      "entropy": 0.7720463991165161,
      "epoch": 6.019389532681966,
      "grad_norm": 0.1606324315071106,
      "learning_rate": 6.989241829656523e-05,
      "loss": 1.1679,
      "mean_token_accuracy": 0.7754394978284835,
      "num_tokens": 1342201281.0,
      "step": 19870
    },
    {
      "entropy": 0.7681100949645042,
      "epoch": 6.022419147163523,
      "grad_norm": 0.1752772033214569,
      "learning_rate": 6.986177696631026e-05,
      "loss": 1.1645,
      "mean_token_accuracy": 0.7750775516033173,
      "num_tokens": 1342876140.0,
      "step": 19880
    },
    {
      "entropy": 0.7687909170985222,
      "epoch": 6.02544876164508,
      "grad_norm": 0.19374622404575348,
      "learning_rate": 6.983112677670917e-05,
      "loss": 1.1653,
      "mean_token_accuracy": 0.7729068502783776,
      "num_tokens": 1343553676.0,
      "step": 19890
    },
    {
      "entropy": 0.7666820168495179,
      "epoch": 6.028478376126638,
      "grad_norm": 0.17111173272132874,
      "learning_rate": 6.980046774143348e-05,
      "loss": 1.1635,
      "mean_token_accuracy": 0.7752456784248352,
      "num_tokens": 1344228850.0,
      "step": 19900
    },
    {
      "entropy": 0.767454381287098,
      "epoch": 6.031507990608195,
      "grad_norm": 0.17312178015708923,
      "learning_rate": 6.976979987415865e-05,
      "loss": 1.1711,
      "mean_token_accuracy": 0.7684289693832398,
      "num_tokens": 1344890022.0,
      "step": 19910
    },
    {
      "entropy": 0.768279604613781,
      "epoch": 6.034537605089752,
      "grad_norm": 0.18300095200538635,
      "learning_rate": 6.973912318856409e-05,
      "loss": 1.1626,
      "mean_token_accuracy": 0.7727115780115128,
      "num_tokens": 1345562713.0,
      "step": 19920
    },
    {
      "entropy": 0.7703651890158654,
      "epoch": 6.03756721957131,
      "grad_norm": 0.181358203291893,
      "learning_rate": 6.970843769833313e-05,
      "loss": 1.1622,
      "mean_token_accuracy": 0.7731382727622986,
      "num_tokens": 1346241555.0,
      "step": 19930
    },
    {
      "entropy": 0.7753687739372254,
      "epoch": 6.040596834052867,
      "grad_norm": 0.17933306097984314,
      "learning_rate": 6.967774341715305e-05,
      "loss": 1.1737,
      "mean_token_accuracy": 0.7695334941148758,
      "num_tokens": 1346918603.0,
      "step": 19940
    },
    {
      "entropy": 0.763907541334629,
      "epoch": 6.043626448534424,
      "grad_norm": 0.17744694650173187,
      "learning_rate": 6.964704035871503e-05,
      "loss": 1.1644,
      "mean_token_accuracy": 0.7717218473553658,
      "num_tokens": 1347587724.0,
      "step": 19950
    },
    {
      "entropy": 0.7658626720309257,
      "epoch": 6.046656063015981,
      "grad_norm": 0.18007683753967285,
      "learning_rate": 6.961632853671414e-05,
      "loss": 1.1621,
      "mean_token_accuracy": 0.7754075735807419,
      "num_tokens": 1348267404.0,
      "step": 19960
    },
    {
      "entropy": 0.7742058008909225,
      "epoch": 6.049685677497538,
      "grad_norm": 0.1880507916212082,
      "learning_rate": 6.958560796484944e-05,
      "loss": 1.1697,
      "mean_token_accuracy": 0.7779480457305908,
      "num_tokens": 1348959197.0,
      "step": 19970
    },
    {
      "entropy": 0.778060233592987,
      "epoch": 6.052715291979096,
      "grad_norm": 0.18712382018566132,
      "learning_rate": 6.955487865682379e-05,
      "loss": 1.1759,
      "mean_token_accuracy": 0.7741430401802063,
      "num_tokens": 1349639599.0,
      "step": 19980
    },
    {
      "entropy": 0.7732942149043083,
      "epoch": 6.055744906460653,
      "grad_norm": 0.18797504901885986,
      "learning_rate": 6.952414062634403e-05,
      "loss": 1.1719,
      "mean_token_accuracy": 0.7757289350032807,
      "num_tokens": 1350327928.0,
      "step": 19990
    },
    {
      "entropy": 0.7683482483029366,
      "epoch": 6.05877452094221,
      "grad_norm": 0.18267159163951874,
      "learning_rate": 6.949339388712086e-05,
      "loss": 1.1685,
      "mean_token_accuracy": 0.7765683248639107,
      "num_tokens": 1351015808.0,
      "step": 20000
    },
    {
      "entropy": 0.7739752158522606,
      "epoch": 6.061804135423768,
      "grad_norm": 0.18127068877220154,
      "learning_rate": 6.946263845286884e-05,
      "loss": 1.1667,
      "mean_token_accuracy": 0.774583350121975,
      "num_tokens": 1351703087.0,
      "step": 20010
    },
    {
      "entropy": 0.7762703031301499,
      "epoch": 6.064833749905325,
      "grad_norm": 0.2021704465150833,
      "learning_rate": 6.943187433730647e-05,
      "loss": 1.1623,
      "mean_token_accuracy": 0.7790561065077781,
      "num_tokens": 1352405058.0,
      "step": 20020
    },
    {
      "entropy": 0.7643448635935783,
      "epoch": 6.067863364386882,
      "grad_norm": 0.1878141462802887,
      "learning_rate": 6.940110155415606e-05,
      "loss": 1.1676,
      "mean_token_accuracy": 0.7734850108623504,
      "num_tokens": 1353074092.0,
      "step": 20030
    },
    {
      "entropy": 0.7669717848300934,
      "epoch": 6.070892978868439,
      "grad_norm": 0.1895095407962799,
      "learning_rate": 6.937032011714379e-05,
      "loss": 1.166,
      "mean_token_accuracy": 0.7723661869764328,
      "num_tokens": 1353752868.0,
      "step": 20040
    },
    {
      "entropy": 0.756782203912735,
      "epoch": 6.073922593349996,
      "grad_norm": 0.1780415028333664,
      "learning_rate": 6.933953003999977e-05,
      "loss": 1.1609,
      "mean_token_accuracy": 0.7720997422933579,
      "num_tokens": 1354412022.0,
      "step": 20050
    },
    {
      "entropy": 0.7747316166758538,
      "epoch": 6.076952207831553,
      "grad_norm": 0.17755159735679626,
      "learning_rate": 6.930873133645787e-05,
      "loss": 1.1664,
      "mean_token_accuracy": 0.7734985128045082,
      "num_tokens": 1355089984.0,
      "step": 20060
    },
    {
      "entropy": 0.7698041707277298,
      "epoch": 6.079981822313111,
      "grad_norm": 0.18307293951511383,
      "learning_rate": 6.927792402025588e-05,
      "loss": 1.1701,
      "mean_token_accuracy": 0.768767488002777,
      "num_tokens": 1355750467.0,
      "step": 20070
    },
    {
      "entropy": 0.7640743359923363,
      "epoch": 6.083011436794668,
      "grad_norm": 0.18154790997505188,
      "learning_rate": 6.924710810513542e-05,
      "loss": 1.1586,
      "mean_token_accuracy": 0.7760217905044555,
      "num_tokens": 1356431932.0,
      "step": 20080
    },
    {
      "entropy": 0.7717065319418908,
      "epoch": 6.086041051276225,
      "grad_norm": 0.16322270035743713,
      "learning_rate": 6.921628360484186e-05,
      "loss": 1.1709,
      "mean_token_accuracy": 0.769880224764347,
      "num_tokens": 1357108067.0,
      "step": 20090
    },
    {
      "entropy": 0.7686224833130837,
      "epoch": 6.089070665757783,
      "grad_norm": 0.18724380433559418,
      "learning_rate": 6.918545053312454e-05,
      "loss": 1.1674,
      "mean_token_accuracy": 0.7708944946527481,
      "num_tokens": 1357783406.0,
      "step": 20100
    },
    {
      "entropy": 0.7616812288761139,
      "epoch": 6.09210028023934,
      "grad_norm": 0.18448813259601593,
      "learning_rate": 6.915460890373652e-05,
      "loss": 1.168,
      "mean_token_accuracy": 0.7714104011654854,
      "num_tokens": 1358442411.0,
      "step": 20110
    },
    {
      "entropy": 0.7682848289608956,
      "epoch": 6.0951298947208965,
      "grad_norm": 0.1796640306711197,
      "learning_rate": 6.912375873043474e-05,
      "loss": 1.1689,
      "mean_token_accuracy": 0.7742606103420258,
      "num_tokens": 1359125270.0,
      "step": 20120
    },
    {
      "entropy": 0.7673468515276909,
      "epoch": 6.098159509202454,
      "grad_norm": 0.17251326143741608,
      "learning_rate": 6.909290002697987e-05,
      "loss": 1.1717,
      "mean_token_accuracy": 0.7708658576011658,
      "num_tokens": 1359796282.0,
      "step": 20130
    },
    {
      "entropy": 0.7758640006184578,
      "epoch": 6.101189123684011,
      "grad_norm": 0.18424230813980103,
      "learning_rate": 6.906203280713647e-05,
      "loss": 1.1686,
      "mean_token_accuracy": 0.7709204509854317,
      "num_tokens": 1360462142.0,
      "step": 20140
    },
    {
      "entropy": 0.7815222010016442,
      "epoch": 6.104218738165568,
      "grad_norm": 0.17566242814064026,
      "learning_rate": 6.903115708467285e-05,
      "loss": 1.1779,
      "mean_token_accuracy": 0.7686811447143554,
      "num_tokens": 1361144650.0,
      "step": 20150
    },
    {
      "entropy": 0.7736230954527855,
      "epoch": 6.107248352647126,
      "grad_norm": 0.18372832238674164,
      "learning_rate": 6.900027287336115e-05,
      "loss": 1.1684,
      "mean_token_accuracy": 0.7732756614685059,
      "num_tokens": 1361815906.0,
      "step": 20160
    },
    {
      "entropy": 0.7611715704202652,
      "epoch": 6.110277967128683,
      "grad_norm": 0.18187449872493744,
      "learning_rate": 6.896938018697723e-05,
      "loss": 1.1686,
      "mean_token_accuracy": 0.7724242389202118,
      "num_tokens": 1362480459.0,
      "step": 20170
    },
    {
      "entropy": 0.7746902257204056,
      "epoch": 6.11330758161024,
      "grad_norm": 0.19529913365840912,
      "learning_rate": 6.893847903930082e-05,
      "loss": 1.1659,
      "mean_token_accuracy": 0.7734479159116745,
      "num_tokens": 1363166426.0,
      "step": 20180
    },
    {
      "entropy": 0.7706331163644791,
      "epoch": 6.116337196091798,
      "grad_norm": 0.18860577046871185,
      "learning_rate": 6.890756944411537e-05,
      "loss": 1.1766,
      "mean_token_accuracy": 0.7709536388516426,
      "num_tokens": 1363825369.0,
      "step": 20190
    },
    {
      "entropy": 0.7658568829298019,
      "epoch": 6.1193668105733545,
      "grad_norm": 0.178918719291687,
      "learning_rate": 6.887665141520807e-05,
      "loss": 1.169,
      "mean_token_accuracy": 0.7709693178534508,
      "num_tokens": 1364492287.0,
      "step": 20200
    },
    {
      "entropy": 0.7668952539563179,
      "epoch": 6.122396425054911,
      "grad_norm": 0.18230096995830536,
      "learning_rate": 6.884572496636993e-05,
      "loss": 1.1635,
      "mean_token_accuracy": 0.772887147963047,
      "num_tokens": 1365157281.0,
      "step": 20210
    },
    {
      "entropy": 0.7698351189494133,
      "epoch": 6.125426039536469,
      "grad_norm": 0.18724210560321808,
      "learning_rate": 6.881479011139571e-05,
      "loss": 1.1717,
      "mean_token_accuracy": 0.7695181295275688,
      "num_tokens": 1365808415.0,
      "step": 20220
    },
    {
      "entropy": 0.7700842469930649,
      "epoch": 6.128455654018026,
      "grad_norm": 0.17615756392478943,
      "learning_rate": 6.878384686408387e-05,
      "loss": 1.1652,
      "mean_token_accuracy": 0.7756535485386848,
      "num_tokens": 1366489041.0,
      "step": 20230
    },
    {
      "entropy": 0.7667704582214355,
      "epoch": 6.131485268499583,
      "grad_norm": 0.18842682242393494,
      "learning_rate": 6.875289523823669e-05,
      "loss": 1.1744,
      "mean_token_accuracy": 0.7754338160157204,
      "num_tokens": 1367160173.0,
      "step": 20240
    },
    {
      "entropy": 0.7679364934563637,
      "epoch": 6.134514882981141,
      "grad_norm": 0.1914799064397812,
      "learning_rate": 6.87219352476601e-05,
      "loss": 1.1636,
      "mean_token_accuracy": 0.7715009048581123,
      "num_tokens": 1367823763.0,
      "step": 20250
    },
    {
      "entropy": 0.7675938904285431,
      "epoch": 6.137544497462698,
      "grad_norm": 0.18526522815227509,
      "learning_rate": 6.869096690616383e-05,
      "loss": 1.1677,
      "mean_token_accuracy": 0.770045630633831,
      "num_tokens": 1368483957.0,
      "step": 20260
    },
    {
      "entropy": 0.7649978309869766,
      "epoch": 6.140574111944255,
      "grad_norm": 0.17641954123973846,
      "learning_rate": 6.865999022756128e-05,
      "loss": 1.1583,
      "mean_token_accuracy": 0.7767908945679665,
      "num_tokens": 1369165404.0,
      "step": 20270
    },
    {
      "entropy": 0.7735571846365928,
      "epoch": 6.1436037264258125,
      "grad_norm": 0.18176063895225525,
      "learning_rate": 6.862900522566962e-05,
      "loss": 1.1677,
      "mean_token_accuracy": 0.7714340478181839,
      "num_tokens": 1369831429.0,
      "step": 20280
    },
    {
      "entropy": 0.7654300466179847,
      "epoch": 6.1466333409073695,
      "grad_norm": 0.17600718140602112,
      "learning_rate": 6.859801191430976e-05,
      "loss": 1.167,
      "mean_token_accuracy": 0.7764873564243316,
      "num_tokens": 1370511767.0,
      "step": 20290
    },
    {
      "entropy": 0.7687219202518463,
      "epoch": 6.149662955388926,
      "grad_norm": 0.19362908601760864,
      "learning_rate": 6.856701030730617e-05,
      "loss": 1.1697,
      "mean_token_accuracy": 0.7721666499972344,
      "num_tokens": 1371180860.0,
      "step": 20300
    },
    {
      "entropy": 0.7682798579335213,
      "epoch": 6.152692569870484,
      "grad_norm": 0.18298585712909698,
      "learning_rate": 6.853600041848718e-05,
      "loss": 1.1635,
      "mean_token_accuracy": 0.7707045659422874,
      "num_tokens": 1371841453.0,
      "step": 20310
    },
    {
      "entropy": 0.7731047704815864,
      "epoch": 6.155722184352041,
      "grad_norm": 0.17189756035804749,
      "learning_rate": 6.850498226168473e-05,
      "loss": 1.1631,
      "mean_token_accuracy": 0.7727273076772689,
      "num_tokens": 1372517132.0,
      "step": 20320
    },
    {
      "entropy": 0.7688058093190193,
      "epoch": 6.158751798833598,
      "grad_norm": 0.17937079071998596,
      "learning_rate": 6.847395585073448e-05,
      "loss": 1.1647,
      "mean_token_accuracy": 0.7755904942750931,
      "num_tokens": 1373200818.0,
      "step": 20330
    },
    {
      "entropy": 0.766862353682518,
      "epoch": 6.161781413315156,
      "grad_norm": 0.18698036670684814,
      "learning_rate": 6.844292119947576e-05,
      "loss": 1.1649,
      "mean_token_accuracy": 0.7730498358607292,
      "num_tokens": 1373870605.0,
      "step": 20340
    },
    {
      "entropy": 0.7802311062812806,
      "epoch": 6.164811027796713,
      "grad_norm": 0.1781839281320572,
      "learning_rate": 6.841187832175157e-05,
      "loss": 1.1644,
      "mean_token_accuracy": 0.7746206551790238,
      "num_tokens": 1374569387.0,
      "step": 20350
    },
    {
      "entropy": 0.7692409440875053,
      "epoch": 6.16784064227827,
      "grad_norm": 0.19574400782585144,
      "learning_rate": 6.83808272314086e-05,
      "loss": 1.1728,
      "mean_token_accuracy": 0.7712500303983688,
      "num_tokens": 1375233750.0,
      "step": 20360
    },
    {
      "entropy": 0.7716340467333793,
      "epoch": 6.1708702567598275,
      "grad_norm": 0.1821366846561432,
      "learning_rate": 6.834976794229717e-05,
      "loss": 1.1673,
      "mean_token_accuracy": 0.7683703422546386,
      "num_tokens": 1375887021.0,
      "step": 20370
    },
    {
      "entropy": 0.7655805602669716,
      "epoch": 6.173899871241384,
      "grad_norm": 0.1861383318901062,
      "learning_rate": 6.83187004682713e-05,
      "loss": 1.157,
      "mean_token_accuracy": 0.773718175292015,
      "num_tokens": 1376564944.0,
      "step": 20380
    },
    {
      "entropy": 0.7723828539252281,
      "epoch": 6.176929485722942,
      "grad_norm": 0.18086256086826324,
      "learning_rate": 6.828762482318861e-05,
      "loss": 1.1684,
      "mean_token_accuracy": 0.7714751392602921,
      "num_tokens": 1377250732.0,
      "step": 20390
    },
    {
      "entropy": 0.777621379494667,
      "epoch": 6.179959100204499,
      "grad_norm": 0.19169875979423523,
      "learning_rate": 6.825654102091041e-05,
      "loss": 1.1687,
      "mean_token_accuracy": 0.769960243999958,
      "num_tokens": 1377927823.0,
      "step": 20400
    },
    {
      "entropy": 0.7638231918215752,
      "epoch": 6.182988714686056,
      "grad_norm": 0.18605810403823853,
      "learning_rate": 6.822544907530164e-05,
      "loss": 1.1667,
      "mean_token_accuracy": 0.7757900342345238,
      "num_tokens": 1378605440.0,
      "step": 20410
    },
    {
      "entropy": 0.7629158541560173,
      "epoch": 6.186018329167614,
      "grad_norm": 0.19657094776630402,
      "learning_rate": 6.819434900023083e-05,
      "loss": 1.1624,
      "mean_token_accuracy": 0.776650607585907,
      "num_tokens": 1379286782.0,
      "step": 20420
    },
    {
      "entropy": 0.7685953602194786,
      "epoch": 6.189047943649171,
      "grad_norm": 0.18294546008110046,
      "learning_rate": 6.816324080957022e-05,
      "loss": 1.1691,
      "mean_token_accuracy": 0.7680969446897506,
      "num_tokens": 1379940653.0,
      "step": 20430
    },
    {
      "entropy": 0.7718799218535424,
      "epoch": 6.192077558130728,
      "grad_norm": 0.17661738395690918,
      "learning_rate": 6.813212451719558e-05,
      "loss": 1.1652,
      "mean_token_accuracy": 0.7731961011886597,
      "num_tokens": 1380615271.0,
      "step": 20440
    },
    {
      "entropy": 0.7688711792230606,
      "epoch": 6.1951071726122855,
      "grad_norm": 0.17443495988845825,
      "learning_rate": 6.810100013698632e-05,
      "loss": 1.1673,
      "mean_token_accuracy": 0.7736601695418358,
      "num_tokens": 1381291690.0,
      "step": 20450
    },
    {
      "entropy": 0.7773081868886947,
      "epoch": 6.198136787093842,
      "grad_norm": 0.18575076758861542,
      "learning_rate": 6.806986768282553e-05,
      "loss": 1.1596,
      "mean_token_accuracy": 0.7757630035281181,
      "num_tokens": 1381987252.0,
      "step": 20460
    },
    {
      "entropy": 0.7711373716592789,
      "epoch": 6.201166401575399,
      "grad_norm": 0.18245543539524078,
      "learning_rate": 6.803872716859979e-05,
      "loss": 1.1621,
      "mean_token_accuracy": 0.7761328637599945,
      "num_tokens": 1382683321.0,
      "step": 20470
    },
    {
      "entropy": 0.7689271911978721,
      "epoch": 6.204196016056957,
      "grad_norm": 0.18027953803539276,
      "learning_rate": 6.800757860819933e-05,
      "loss": 1.1639,
      "mean_token_accuracy": 0.7739423215389252,
      "num_tokens": 1383357143.0,
      "step": 20480
    },
    {
      "entropy": 0.7688937187194824,
      "epoch": 6.207225630538514,
      "grad_norm": 0.17201583087444305,
      "learning_rate": 6.797642201551799e-05,
      "loss": 1.171,
      "mean_token_accuracy": 0.7725284457206726,
      "num_tokens": 1384029840.0,
      "step": 20490
    },
    {
      "entropy": 0.7655343785881996,
      "epoch": 6.210255245020071,
      "grad_norm": 0.17956753075122833,
      "learning_rate": 6.794525740445317e-05,
      "loss": 1.1594,
      "mean_token_accuracy": 0.7772940382361412,
      "num_tokens": 1384709893.0,
      "step": 20500
    },
    {
      "entropy": 0.7810137525200844,
      "epoch": 6.213284859501629,
      "grad_norm": 0.18092070519924164,
      "learning_rate": 6.791408478890581e-05,
      "loss": 1.1818,
      "mean_token_accuracy": 0.7734193980693818,
      "num_tokens": 1385390270.0,
      "step": 20510
    },
    {
      "entropy": 0.7619176179170608,
      "epoch": 6.216314473983186,
      "grad_norm": 0.18442845344543457,
      "learning_rate": 6.788290418278049e-05,
      "loss": 1.156,
      "mean_token_accuracy": 0.7772478967905044,
      "num_tokens": 1386072852.0,
      "step": 20520
    },
    {
      "entropy": 0.7646384924650192,
      "epoch": 6.219344088464743,
      "grad_norm": 0.1901431381702423,
      "learning_rate": 6.785171559998531e-05,
      "loss": 1.1661,
      "mean_token_accuracy": 0.77126996666193,
      "num_tokens": 1386738809.0,
      "step": 20530
    },
    {
      "entropy": 0.7642429247498512,
      "epoch": 6.2223737029463,
      "grad_norm": 0.1791820526123047,
      "learning_rate": 6.782051905443193e-05,
      "loss": 1.1675,
      "mean_token_accuracy": 0.7767746433615684,
      "num_tokens": 1387419720.0,
      "step": 20540
    },
    {
      "entropy": 0.766360518336296,
      "epoch": 6.225403317427857,
      "grad_norm": 0.19383500516414642,
      "learning_rate": 6.778931456003558e-05,
      "loss": 1.1683,
      "mean_token_accuracy": 0.7729840204119682,
      "num_tokens": 1388085492.0,
      "step": 20550
    },
    {
      "entropy": 0.7726920738816261,
      "epoch": 6.228432931909414,
      "grad_norm": 0.18173034489154816,
      "learning_rate": 6.775810213071503e-05,
      "loss": 1.1706,
      "mean_token_accuracy": 0.7684736415743828,
      "num_tokens": 1388753867.0,
      "step": 20560
    },
    {
      "entropy": 0.7729493036866188,
      "epoch": 6.231462546390972,
      "grad_norm": 0.18822112679481506,
      "learning_rate": 6.772688178039257e-05,
      "loss": 1.1659,
      "mean_token_accuracy": 0.7745878711342812,
      "num_tokens": 1389437894.0,
      "step": 20570
    },
    {
      "entropy": 0.7687354937195778,
      "epoch": 6.234492160872529,
      "grad_norm": 0.1805078536272049,
      "learning_rate": 6.769565352299404e-05,
      "loss": 1.1648,
      "mean_token_accuracy": 0.7734712541103363,
      "num_tokens": 1390119855.0,
      "step": 20580
    },
    {
      "entropy": 0.7803248658776283,
      "epoch": 6.237521775354086,
      "grad_norm": 0.1833615005016327,
      "learning_rate": 6.766441737244881e-05,
      "loss": 1.1665,
      "mean_token_accuracy": 0.7708469524979591,
      "num_tokens": 1390794339.0,
      "step": 20590
    },
    {
      "entropy": 0.7634388208389282,
      "epoch": 6.240551389835644,
      "grad_norm": 0.18130077421665192,
      "learning_rate": 6.763317334268979e-05,
      "loss": 1.164,
      "mean_token_accuracy": 0.7763359144330024,
      "num_tokens": 1391467689.0,
      "step": 20600
    },
    {
      "entropy": 0.7681302428245544,
      "epoch": 6.243581004317201,
      "grad_norm": 0.1837979257106781,
      "learning_rate": 6.760192144765333e-05,
      "loss": 1.1684,
      "mean_token_accuracy": 0.7731573298573494,
      "num_tokens": 1392138378.0,
      "step": 20610
    },
    {
      "entropy": 0.7732110306620598,
      "epoch": 6.246610618798758,
      "grad_norm": 0.18030866980552673,
      "learning_rate": 6.757066170127939e-05,
      "loss": 1.1698,
      "mean_token_accuracy": 0.7718074098229408,
      "num_tokens": 1392807259.0,
      "step": 20620
    },
    {
      "entropy": 0.7699881076812745,
      "epoch": 6.249640233280315,
      "grad_norm": 0.1618461310863495,
      "learning_rate": 6.753939411751135e-05,
      "loss": 1.1691,
      "mean_token_accuracy": 0.772416403889656,
      "num_tokens": 1393470765.0,
      "step": 20630
    },
    {
      "entropy": 0.7750715002417564,
      "epoch": 6.252669847761872,
      "grad_norm": 0.18438762426376343,
      "learning_rate": 6.750811871029611e-05,
      "loss": 1.1698,
      "mean_token_accuracy": 0.7706983745098114,
      "num_tokens": 1394148049.0,
      "step": 20640
    },
    {
      "entropy": 0.7704644456505776,
      "epoch": 6.255699462243429,
      "grad_norm": 0.1812216341495514,
      "learning_rate": 6.74768354935841e-05,
      "loss": 1.1674,
      "mean_token_accuracy": 0.7730281010270119,
      "num_tokens": 1394821669.0,
      "step": 20650
    },
    {
      "entropy": 0.7773768618702889,
      "epoch": 6.258729076724987,
      "grad_norm": 0.1881628930568695,
      "learning_rate": 6.744554448132915e-05,
      "loss": 1.1767,
      "mean_token_accuracy": 0.7692205518484115,
      "num_tokens": 1395489466.0,
      "step": 20660
    },
    {
      "entropy": 0.7700340136885643,
      "epoch": 6.261758691206544,
      "grad_norm": 0.18863986432552338,
      "learning_rate": 6.741424568748867e-05,
      "loss": 1.1694,
      "mean_token_accuracy": 0.7716132119297981,
      "num_tokens": 1396150850.0,
      "step": 20670
    },
    {
      "entropy": 0.7674183547496796,
      "epoch": 6.264788305688101,
      "grad_norm": 0.17054618895053864,
      "learning_rate": 6.738293912602343e-05,
      "loss": 1.1559,
      "mean_token_accuracy": 0.7718262419104576,
      "num_tokens": 1396833321.0,
      "step": 20680
    },
    {
      "entropy": 0.7717158362269402,
      "epoch": 6.267817920169659,
      "grad_norm": 0.17838239669799805,
      "learning_rate": 6.735162481089779e-05,
      "loss": 1.1586,
      "mean_token_accuracy": 0.7769655585289001,
      "num_tokens": 1397526053.0,
      "step": 20690
    },
    {
      "entropy": 0.7677412182092667,
      "epoch": 6.270847534651216,
      "grad_norm": 0.1865473985671997,
      "learning_rate": 6.732030275607946e-05,
      "loss": 1.1654,
      "mean_token_accuracy": 0.774252912402153,
      "num_tokens": 1398206362.0,
      "step": 20700
    },
    {
      "entropy": 0.7620617344975471,
      "epoch": 6.2738771491327725,
      "grad_norm": 0.1808382123708725,
      "learning_rate": 6.728897297553964e-05,
      "loss": 1.1644,
      "mean_token_accuracy": 0.7775876253843308,
      "num_tokens": 1398882277.0,
      "step": 20710
    },
    {
      "entropy": 0.7642869144678116,
      "epoch": 6.27690676361433,
      "grad_norm": 0.17975032329559326,
      "learning_rate": 6.725763548325298e-05,
      "loss": 1.1677,
      "mean_token_accuracy": 0.7721994742751122,
      "num_tokens": 1399549887.0,
      "step": 20720
    },
    {
      "entropy": 0.7756190747022629,
      "epoch": 6.279936378095887,
      "grad_norm": 0.192529559135437,
      "learning_rate": 6.72262902931976e-05,
      "loss": 1.1738,
      "mean_token_accuracy": 0.7731177315115929,
      "num_tokens": 1400227892.0,
      "step": 20730
    },
    {
      "entropy": 0.7728633806109428,
      "epoch": 6.282965992577444,
      "grad_norm": 0.18479959666728973,
      "learning_rate": 6.719493741935498e-05,
      "loss": 1.1683,
      "mean_token_accuracy": 0.7704318210482597,
      "num_tokens": 1400897290.0,
      "step": 20740
    },
    {
      "entropy": 0.7752651602029801,
      "epoch": 6.285995607059002,
      "grad_norm": 0.18089070916175842,
      "learning_rate": 6.71635768757101e-05,
      "loss": 1.1775,
      "mean_token_accuracy": 0.7714816778898239,
      "num_tokens": 1401565432.0,
      "step": 20750
    },
    {
      "entropy": 0.7818360552191734,
      "epoch": 6.289025221540559,
      "grad_norm": 0.18499885499477386,
      "learning_rate": 6.713220867625133e-05,
      "loss": 1.1665,
      "mean_token_accuracy": 0.7688517034053802,
      "num_tokens": 1402235614.0,
      "step": 20760
    },
    {
      "entropy": 0.7780620649456977,
      "epoch": 6.292054836022116,
      "grad_norm": 0.17752474546432495,
      "learning_rate": 6.710083283497045e-05,
      "loss": 1.1742,
      "mean_token_accuracy": 0.7699872747063636,
      "num_tokens": 1402905570.0,
      "step": 20770
    },
    {
      "entropy": 0.7712685331702233,
      "epoch": 6.295084450503674,
      "grad_norm": 0.17574584484100342,
      "learning_rate": 6.706944936586264e-05,
      "loss": 1.1677,
      "mean_token_accuracy": 0.7757701098918914,
      "num_tokens": 1403590577.0,
      "step": 20780
    },
    {
      "entropy": 0.7641827404499054,
      "epoch": 6.2981140649852305,
      "grad_norm": 0.18991263210773468,
      "learning_rate": 6.703805828292652e-05,
      "loss": 1.1651,
      "mean_token_accuracy": 0.7717865869402886,
      "num_tokens": 1404257085.0,
      "step": 20790
    },
    {
      "entropy": 0.7810153186321258,
      "epoch": 6.301143679466787,
      "grad_norm": 0.19716088473796844,
      "learning_rate": 6.700665960016408e-05,
      "loss": 1.1733,
      "mean_token_accuracy": 0.7687037140130997,
      "num_tokens": 1404934702.0,
      "step": 20800
    },
    {
      "entropy": 0.7753670707345008,
      "epoch": 6.304173293948345,
      "grad_norm": 0.1863342970609665,
      "learning_rate": 6.69752533315807e-05,
      "loss": 1.1625,
      "mean_token_accuracy": 0.7753713056445122,
      "num_tokens": 1405622781.0,
      "step": 20810
    },
    {
      "entropy": 0.7637102693319321,
      "epoch": 6.307202908429902,
      "grad_norm": 0.1791115254163742,
      "learning_rate": 6.694383949118512e-05,
      "loss": 1.1606,
      "mean_token_accuracy": 0.7726149007678031,
      "num_tokens": 1406300935.0,
      "step": 20820
    },
    {
      "entropy": 0.7656391814351082,
      "epoch": 6.310232522911459,
      "grad_norm": 0.18493247032165527,
      "learning_rate": 6.691241809298952e-05,
      "loss": 1.1603,
      "mean_token_accuracy": 0.7783811420202256,
      "num_tokens": 1406980777.0,
      "step": 20830
    },
    {
      "entropy": 0.7651975587010383,
      "epoch": 6.313262137393017,
      "grad_norm": 0.20166723430156708,
      "learning_rate": 6.688098915100942e-05,
      "loss": 1.164,
      "mean_token_accuracy": 0.7746549114584923,
      "num_tokens": 1407651515.0,
      "step": 20840
    },
    {
      "entropy": 0.7801913648843766,
      "epoch": 6.316291751874574,
      "grad_norm": 0.17842784523963928,
      "learning_rate": 6.684955267926367e-05,
      "loss": 1.1705,
      "mean_token_accuracy": 0.7728280574083328,
      "num_tokens": 1408333492.0,
      "step": 20850
    },
    {
      "entropy": 0.7690182015299797,
      "epoch": 6.319321366356132,
      "grad_norm": 0.18181131780147552,
      "learning_rate": 6.681810869177452e-05,
      "loss": 1.17,
      "mean_token_accuracy": 0.7699664264917374,
      "num_tokens": 1408993748.0,
      "step": 20860
    },
    {
      "entropy": 0.775431276857853,
      "epoch": 6.3223509808376885,
      "grad_norm": 0.18599024415016174,
      "learning_rate": 6.678665720256759e-05,
      "loss": 1.172,
      "mean_token_accuracy": 0.7708151072263718,
      "num_tokens": 1409669977.0,
      "step": 20870
    },
    {
      "entropy": 0.7725912168622017,
      "epoch": 6.3253805953192455,
      "grad_norm": 0.17442113161087036,
      "learning_rate": 6.675519822567175e-05,
      "loss": 1.1683,
      "mean_token_accuracy": 0.7700779899954796,
      "num_tokens": 1410334412.0,
      "step": 20880
    },
    {
      "entropy": 0.7692646980285645,
      "epoch": 6.328410209800802,
      "grad_norm": 0.19657650589942932,
      "learning_rate": 6.672373177511935e-05,
      "loss": 1.1602,
      "mean_token_accuracy": 0.7722544282674789,
      "num_tokens": 1411019522.0,
      "step": 20890
    },
    {
      "entropy": 0.7827627912163735,
      "epoch": 6.33143982428236,
      "grad_norm": 0.18095193803310394,
      "learning_rate": 6.669225786494597e-05,
      "loss": 1.1814,
      "mean_token_accuracy": 0.7697557792067528,
      "num_tokens": 1411691157.0,
      "step": 20900
    },
    {
      "entropy": 0.7719816759228706,
      "epoch": 6.334469438763917,
      "grad_norm": 0.18132972717285156,
      "learning_rate": 6.666077650919054e-05,
      "loss": 1.1691,
      "mean_token_accuracy": 0.770168337225914,
      "num_tokens": 1412368112.0,
      "step": 20910
    },
    {
      "entropy": 0.7707783684134484,
      "epoch": 6.337499053245475,
      "grad_norm": 0.19714146852493286,
      "learning_rate": 6.662928772189532e-05,
      "loss": 1.1656,
      "mean_token_accuracy": 0.7698728039860725,
      "num_tokens": 1413049734.0,
      "step": 20920
    },
    {
      "entropy": 0.7757603093981743,
      "epoch": 6.340528667727032,
      "grad_norm": 0.17612393200397491,
      "learning_rate": 6.65977915171059e-05,
      "loss": 1.1727,
      "mean_token_accuracy": 0.7731116905808448,
      "num_tokens": 1413729512.0,
      "step": 20930
    },
    {
      "entropy": 0.7686492502689362,
      "epoch": 6.343558282208589,
      "grad_norm": 0.17846208810806274,
      "learning_rate": 6.656628790887117e-05,
      "loss": 1.1647,
      "mean_token_accuracy": 0.7792172431945801,
      "num_tokens": 1414427392.0,
      "step": 20940
    },
    {
      "entropy": 0.7680420055985451,
      "epoch": 6.346587896690147,
      "grad_norm": 0.18803076446056366,
      "learning_rate": 6.653477691124329e-05,
      "loss": 1.1667,
      "mean_token_accuracy": 0.7723503828048706,
      "num_tokens": 1415093152.0,
      "step": 20950
    },
    {
      "entropy": 0.7752744823694229,
      "epoch": 6.3496175111717035,
      "grad_norm": 0.1949007511138916,
      "learning_rate": 6.650325853827775e-05,
      "loss": 1.1696,
      "mean_token_accuracy": 0.7694061130285264,
      "num_tokens": 1415769889.0,
      "step": 20960
    },
    {
      "entropy": 0.7765705659985542,
      "epoch": 6.35264712565326,
      "grad_norm": 0.18722619116306305,
      "learning_rate": 6.647173280403334e-05,
      "loss": 1.1713,
      "mean_token_accuracy": 0.7727076575160027,
      "num_tokens": 1416450068.0,
      "step": 20970
    },
    {
      "entropy": 0.7655983179807663,
      "epoch": 6.355676740134818,
      "grad_norm": 0.17996463179588318,
      "learning_rate": 6.644019972257207e-05,
      "loss": 1.1561,
      "mean_token_accuracy": 0.7806936472654342,
      "num_tokens": 1417142308.0,
      "step": 20980
    },
    {
      "entropy": 0.7769755586981774,
      "epoch": 6.358706354616375,
      "grad_norm": 0.17622792720794678,
      "learning_rate": 6.640865930795933e-05,
      "loss": 1.1705,
      "mean_token_accuracy": 0.7706779524683952,
      "num_tokens": 1417815837.0,
      "step": 20990
    },
    {
      "entropy": 0.7751562669873238,
      "epoch": 6.361735969097932,
      "grad_norm": 0.18360324203968048,
      "learning_rate": 6.637711157426367e-05,
      "loss": 1.1658,
      "mean_token_accuracy": 0.7699324071407319,
      "num_tokens": 1418493989.0,
      "step": 21000
    },
    {
      "entropy": 0.7727232158184052,
      "epoch": 6.36476558357949,
      "grad_norm": 0.18577717244625092,
      "learning_rate": 6.634555653555701e-05,
      "loss": 1.1738,
      "mean_token_accuracy": 0.7678659647703171,
      "num_tokens": 1419152661.0,
      "step": 21010
    },
    {
      "entropy": 0.7622781082987785,
      "epoch": 6.367795198061047,
      "grad_norm": 0.18314194679260254,
      "learning_rate": 6.631399420591443e-05,
      "loss": 1.1643,
      "mean_token_accuracy": 0.7772304236888885,
      "num_tokens": 1419836556.0,
      "step": 21020
    },
    {
      "entropy": 0.7705505311489105,
      "epoch": 6.370824812542604,
      "grad_norm": 0.16682995855808258,
      "learning_rate": 6.628242459941429e-05,
      "loss": 1.1711,
      "mean_token_accuracy": 0.7730928897857666,
      "num_tokens": 1420505280.0,
      "step": 21030
    },
    {
      "entropy": 0.7660258039832115,
      "epoch": 6.3738544270241615,
      "grad_norm": 0.17763105034828186,
      "learning_rate": 6.62508477301383e-05,
      "loss": 1.1752,
      "mean_token_accuracy": 0.7703719526529312,
      "num_tokens": 1421168671.0,
      "step": 21040
    },
    {
      "entropy": 0.7788975641131402,
      "epoch": 6.376884041505718,
      "grad_norm": 0.18896135687828064,
      "learning_rate": 6.621926361217123e-05,
      "loss": 1.1765,
      "mean_token_accuracy": 0.7684742927551269,
      "num_tokens": 1421833344.0,
      "step": 21050
    },
    {
      "entropy": 0.7638883426785469,
      "epoch": 6.379913655987275,
      "grad_norm": 0.18676747381687164,
      "learning_rate": 6.618767225960124e-05,
      "loss": 1.1669,
      "mean_token_accuracy": 0.7727542161941529,
      "num_tokens": 1422507130.0,
      "step": 21060
    },
    {
      "entropy": 0.761621555685997,
      "epoch": 6.382943270468833,
      "grad_norm": 0.19346609711647034,
      "learning_rate": 6.615607368651964e-05,
      "loss": 1.1664,
      "mean_token_accuracy": 0.7779986679553985,
      "num_tokens": 1423180466.0,
      "step": 21070
    },
    {
      "entropy": 0.7653658419847489,
      "epoch": 6.38597288495039,
      "grad_norm": 0.1793552190065384,
      "learning_rate": 6.612446790702094e-05,
      "loss": 1.1616,
      "mean_token_accuracy": 0.7750341057777405,
      "num_tokens": 1423855718.0,
      "step": 21080
    },
    {
      "entropy": 0.7796183124184608,
      "epoch": 6.389002499431947,
      "grad_norm": 0.18394005298614502,
      "learning_rate": 6.609285493520294e-05,
      "loss": 1.1765,
      "mean_token_accuracy": 0.774328587949276,
      "num_tokens": 1424547336.0,
      "step": 21090
    },
    {
      "entropy": 0.7839424893260002,
      "epoch": 6.392032113913505,
      "grad_norm": 0.17285147309303284,
      "learning_rate": 6.606123478516659e-05,
      "loss": 1.177,
      "mean_token_accuracy": 0.7707555308938027,
      "num_tokens": 1425231517.0,
      "step": 21100
    },
    {
      "entropy": 0.7662839666008949,
      "epoch": 6.395061728395062,
      "grad_norm": 0.17895035445690155,
      "learning_rate": 6.602960747101605e-05,
      "loss": 1.1701,
      "mean_token_accuracy": 0.7697966247797012,
      "num_tokens": 1425893988.0,
      "step": 21110
    },
    {
      "entropy": 0.7741492956876754,
      "epoch": 6.398091342876619,
      "grad_norm": 0.1894502490758896,
      "learning_rate": 6.59979730068587e-05,
      "loss": 1.1728,
      "mean_token_accuracy": 0.7667291045188904,
      "num_tokens": 1426561243.0,
      "step": 21120
    },
    {
      "entropy": 0.765935517847538,
      "epoch": 6.401120957358176,
      "grad_norm": 0.19292519986629486,
      "learning_rate": 6.596633140680509e-05,
      "loss": 1.1682,
      "mean_token_accuracy": 0.7751199170947075,
      "num_tokens": 1427241194.0,
      "step": 21130
    },
    {
      "entropy": 0.776461872458458,
      "epoch": 6.404150571839733,
      "grad_norm": 0.191630020737648,
      "learning_rate": 6.593468268496896e-05,
      "loss": 1.1734,
      "mean_token_accuracy": 0.7709326341748237,
      "num_tokens": 1427908234.0,
      "step": 21140
    },
    {
      "entropy": 0.7621702805161477,
      "epoch": 6.40718018632129,
      "grad_norm": 0.18412964046001434,
      "learning_rate": 6.590302685546721e-05,
      "loss": 1.1583,
      "mean_token_accuracy": 0.7765338391065597,
      "num_tokens": 1428588885.0,
      "step": 21150
    },
    {
      "entropy": 0.782624889910221,
      "epoch": 6.410209800802848,
      "grad_norm": 0.16957661509513855,
      "learning_rate": 6.587136393241993e-05,
      "loss": 1.1744,
      "mean_token_accuracy": 0.7702025204896927,
      "num_tokens": 1429271107.0,
      "step": 21160
    },
    {
      "entropy": 0.7680342867970467,
      "epoch": 6.413239415284405,
      "grad_norm": 0.18677100539207458,
      "learning_rate": 6.583969392995037e-05,
      "loss": 1.1658,
      "mean_token_accuracy": 0.7732661232352257,
      "num_tokens": 1429942334.0,
      "step": 21170
    },
    {
      "entropy": 0.7686312094330787,
      "epoch": 6.416269029765962,
      "grad_norm": 0.17361260950565338,
      "learning_rate": 6.580801686218495e-05,
      "loss": 1.1585,
      "mean_token_accuracy": 0.7795374110341072,
      "num_tokens": 1430641178.0,
      "step": 21180
    },
    {
      "entropy": 0.7684334322810173,
      "epoch": 6.41929864424752,
      "grad_norm": 0.1868567019701004,
      "learning_rate": 6.577633274325319e-05,
      "loss": 1.1649,
      "mean_token_accuracy": 0.7755624249577522,
      "num_tokens": 1431329956.0,
      "step": 21190
    },
    {
      "entropy": 0.7772552326321602,
      "epoch": 6.422328258729077,
      "grad_norm": 0.18956871330738068,
      "learning_rate": 6.574464158728781e-05,
      "loss": 1.1731,
      "mean_token_accuracy": 0.7735658958554268,
      "num_tokens": 1432010736.0,
      "step": 21200
    },
    {
      "entropy": 0.7731815323233604,
      "epoch": 6.425357873210634,
      "grad_norm": 0.18785738945007324,
      "learning_rate": 6.571294340842467e-05,
      "loss": 1.17,
      "mean_token_accuracy": 0.772030559182167,
      "num_tokens": 1432696627.0,
      "step": 21210
    },
    {
      "entropy": 0.7707874432206154,
      "epoch": 6.428387487692191,
      "grad_norm": 0.19727292656898499,
      "learning_rate": 6.568123822080273e-05,
      "loss": 1.1673,
      "mean_token_accuracy": 0.7771949291229248,
      "num_tokens": 1433401192.0,
      "step": 21220
    },
    {
      "entropy": 0.7744519487023354,
      "epoch": 6.431417102173748,
      "grad_norm": 0.18756203353405,
      "learning_rate": 6.564952603856408e-05,
      "loss": 1.1636,
      "mean_token_accuracy": 0.7736545264720917,
      "num_tokens": 1434080468.0,
      "step": 21230
    },
    {
      "entropy": 0.7654246866703034,
      "epoch": 6.434446716655305,
      "grad_norm": 0.18440191447734833,
      "learning_rate": 6.561780687585392e-05,
      "loss": 1.162,
      "mean_token_accuracy": 0.7743609964847564,
      "num_tokens": 1434756409.0,
      "step": 21240
    },
    {
      "entropy": 0.7709704726934433,
      "epoch": 6.437476331136863,
      "grad_norm": 0.17374394834041595,
      "learning_rate": 6.558608074682064e-05,
      "loss": 1.1689,
      "mean_token_accuracy": 0.7725660175085067,
      "num_tokens": 1435434873.0,
      "step": 21250
    },
    {
      "entropy": 0.7626835972070694,
      "epoch": 6.44050594561842,
      "grad_norm": 0.18436864018440247,
      "learning_rate": 6.555434766561565e-05,
      "loss": 1.1665,
      "mean_token_accuracy": 0.7708254888653755,
      "num_tokens": 1436098544.0,
      "step": 21260
    },
    {
      "entropy": 0.7821949139237404,
      "epoch": 6.443535560099978,
      "grad_norm": 0.19370079040527344,
      "learning_rate": 6.552260764639348e-05,
      "loss": 1.1765,
      "mean_token_accuracy": 0.7679042026400567,
      "num_tokens": 1436766331.0,
      "step": 21270
    },
    {
      "entropy": 0.7772647470235825,
      "epoch": 6.446565174581535,
      "grad_norm": 0.1867731362581253,
      "learning_rate": 6.549086070331179e-05,
      "loss": 1.1747,
      "mean_token_accuracy": 0.7668262258172035,
      "num_tokens": 1437434949.0,
      "step": 21280
    },
    {
      "entropy": 0.765082560479641,
      "epoch": 6.449594789063092,
      "grad_norm": 0.18199868500232697,
      "learning_rate": 6.545910685053128e-05,
      "loss": 1.1681,
      "mean_token_accuracy": 0.7766094893217087,
      "num_tokens": 1438118585.0,
      "step": 21290
    },
    {
      "entropy": 0.7694566279649735,
      "epoch": 6.4526244035446485,
      "grad_norm": 0.18930061161518097,
      "learning_rate": 6.542734610221578e-05,
      "loss": 1.1651,
      "mean_token_accuracy": 0.7690699368715286,
      "num_tokens": 1438778345.0,
      "step": 21300
    },
    {
      "entropy": 0.7629023566842079,
      "epoch": 6.455654018026206,
      "grad_norm": 0.19168712198734283,
      "learning_rate": 6.539557847253217e-05,
      "loss": 1.1659,
      "mean_token_accuracy": 0.7722378998994828,
      "num_tokens": 1439443575.0,
      "step": 21310
    },
    {
      "entropy": 0.7745316356420517,
      "epoch": 6.458683632507763,
      "grad_norm": 0.18028612434864044,
      "learning_rate": 6.536380397565038e-05,
      "loss": 1.1763,
      "mean_token_accuracy": 0.7670167997479439,
      "num_tokens": 1440106438.0,
      "step": 21320
    },
    {
      "entropy": 0.7686283111572265,
      "epoch": 6.461713246989321,
      "grad_norm": 0.18088003993034363,
      "learning_rate": 6.533202262574343e-05,
      "loss": 1.1654,
      "mean_token_accuracy": 0.7731160148978233,
      "num_tokens": 1440788951.0,
      "step": 21330
    },
    {
      "entropy": 0.7695881143212319,
      "epoch": 6.464742861470878,
      "grad_norm": 0.1931450366973877,
      "learning_rate": 6.530023443698744e-05,
      "loss": 1.1624,
      "mean_token_accuracy": 0.7759535938501358,
      "num_tokens": 1441472780.0,
      "step": 21340
    },
    {
      "entropy": 0.7654684498906136,
      "epoch": 6.467772475952435,
      "grad_norm": 0.1829850971698761,
      "learning_rate": 6.526843942356147e-05,
      "loss": 1.1685,
      "mean_token_accuracy": 0.7688734218478203,
      "num_tokens": 1442131107.0,
      "step": 21350
    },
    {
      "entropy": 0.7659239575266839,
      "epoch": 6.470802090433993,
      "grad_norm": 0.19256149232387543,
      "learning_rate": 6.523663759964772e-05,
      "loss": 1.1669,
      "mean_token_accuracy": 0.7726069211959838,
      "num_tokens": 1442814003.0,
      "step": 21360
    },
    {
      "entropy": 0.774470879137516,
      "epoch": 6.47383170491555,
      "grad_norm": 0.18819394707679749,
      "learning_rate": 6.520482897943137e-05,
      "loss": 1.1654,
      "mean_token_accuracy": 0.7749659284949303,
      "num_tokens": 1443499409.0,
      "step": 21370
    },
    {
      "entropy": 0.7676126733422279,
      "epoch": 6.4768613193971065,
      "grad_norm": 0.18283072113990784,
      "learning_rate": 6.517301357710069e-05,
      "loss": 1.1585,
      "mean_token_accuracy": 0.7723554998636246,
      "num_tokens": 1444177570.0,
      "step": 21380
    },
    {
      "entropy": 0.7656297534704208,
      "epoch": 6.479890933878664,
      "grad_norm": 0.18021374940872192,
      "learning_rate": 6.514119140684691e-05,
      "loss": 1.1671,
      "mean_token_accuracy": 0.7718273043632508,
      "num_tokens": 1444847150.0,
      "step": 21390
    },
    {
      "entropy": 0.7661767914891243,
      "epoch": 6.482920548360221,
      "grad_norm": 0.1733720302581787,
      "learning_rate": 6.510936248286432e-05,
      "loss": 1.1592,
      "mean_token_accuracy": 0.7731059432029724,
      "num_tokens": 1445527955.0,
      "step": 21400
    },
    {
      "entropy": 0.763685031235218,
      "epoch": 6.485950162841778,
      "grad_norm": 0.18219871819019318,
      "learning_rate": 6.507752681935022e-05,
      "loss": 1.1588,
      "mean_token_accuracy": 0.7770874068140984,
      "num_tokens": 1446200835.0,
      "step": 21410
    },
    {
      "entropy": 0.771403020620346,
      "epoch": 6.488979777323336,
      "grad_norm": 0.18652617931365967,
      "learning_rate": 6.50456844305049e-05,
      "loss": 1.1791,
      "mean_token_accuracy": 0.7701421588659286,
      "num_tokens": 1446862912.0,
      "step": 21420
    },
    {
      "entropy": 0.7709251582622528,
      "epoch": 6.492009391804893,
      "grad_norm": 0.1999097764492035,
      "learning_rate": 6.501383533053166e-05,
      "loss": 1.1639,
      "mean_token_accuracy": 0.7720854237675667,
      "num_tokens": 1447540393.0,
      "step": 21430
    },
    {
      "entropy": 0.7736376330256463,
      "epoch": 6.49503900628645,
      "grad_norm": 0.18020248413085938,
      "learning_rate": 6.49819795336368e-05,
      "loss": 1.1608,
      "mean_token_accuracy": 0.770463714003563,
      "num_tokens": 1448214715.0,
      "step": 21440
    },
    {
      "entropy": 0.7666349977254867,
      "epoch": 6.498068620768008,
      "grad_norm": 0.19957131147384644,
      "learning_rate": 6.495011705402958e-05,
      "loss": 1.167,
      "mean_token_accuracy": 0.7720817238092422,
      "num_tokens": 1448887278.0,
      "step": 21450
    },
    {
      "entropy": 0.757916721701622,
      "epoch": 6.5010982352495645,
      "grad_norm": 0.23013772070407867,
      "learning_rate": 6.491824790592228e-05,
      "loss": 1.1546,
      "mean_token_accuracy": 0.7753512367606163,
      "num_tokens": 1449568819.0,
      "step": 21460
    },
    {
      "entropy": 0.7611304610967636,
      "epoch": 6.5041278497311215,
      "grad_norm": 0.1955873966217041,
      "learning_rate": 6.488637210353012e-05,
      "loss": 1.1546,
      "mean_token_accuracy": 0.7735495448112488,
      "num_tokens": 1450237286.0,
      "step": 21470
    },
    {
      "entropy": 0.7752103328704834,
      "epoch": 6.507157464212679,
      "grad_norm": 0.17470267415046692,
      "learning_rate": 6.485448966107133e-05,
      "loss": 1.1663,
      "mean_token_accuracy": 0.7741127356886863,
      "num_tokens": 1450927403.0,
      "step": 21480
    },
    {
      "entropy": 0.7749665439128876,
      "epoch": 6.510187078694236,
      "grad_norm": 0.18244533240795135,
      "learning_rate": 6.482260059276703e-05,
      "loss": 1.1717,
      "mean_token_accuracy": 0.7713807061314583,
      "num_tokens": 1451597746.0,
      "step": 21490
    },
    {
      "entropy": 0.7803113788366318,
      "epoch": 6.513216693175793,
      "grad_norm": 0.1868499368429184,
      "learning_rate": 6.479070491284136e-05,
      "loss": 1.1799,
      "mean_token_accuracy": 0.7682018563151359,
      "num_tokens": 1452264988.0,
      "step": 21500
    },
    {
      "entropy": 0.7720252394676208,
      "epoch": 6.516246307657351,
      "grad_norm": 0.17205385863780975,
      "learning_rate": 6.47588026355214e-05,
      "loss": 1.162,
      "mean_token_accuracy": 0.7716523960232735,
      "num_tokens": 1452941641.0,
      "step": 21510
    },
    {
      "entropy": 0.7706853568553924,
      "epoch": 6.519275922138908,
      "grad_norm": 0.1778043806552887,
      "learning_rate": 6.472689377503718e-05,
      "loss": 1.1717,
      "mean_token_accuracy": 0.7744766980409622,
      "num_tokens": 1453617733.0,
      "step": 21520
    },
    {
      "entropy": 0.7690259456634522,
      "epoch": 6.522305536620465,
      "grad_norm": 0.18089909851551056,
      "learning_rate": 6.469497834562159e-05,
      "loss": 1.1657,
      "mean_token_accuracy": 0.7723587095737457,
      "num_tokens": 1454285958.0,
      "step": 21530
    },
    {
      "entropy": 0.7676387712359428,
      "epoch": 6.525335151102023,
      "grad_norm": 0.18544796109199524,
      "learning_rate": 6.466305636151054e-05,
      "loss": 1.1669,
      "mean_token_accuracy": 0.771427059173584,
      "num_tokens": 1454949315.0,
      "step": 21540
    },
    {
      "entropy": 0.7739736899733544,
      "epoch": 6.5283647655835795,
      "grad_norm": 0.17433473467826843,
      "learning_rate": 6.463112783694284e-05,
      "loss": 1.1695,
      "mean_token_accuracy": 0.7705166786909103,
      "num_tokens": 1455617190.0,
      "step": 21550
    },
    {
      "entropy": 0.7755619525909424,
      "epoch": 6.531394380065136,
      "grad_norm": 0.16753840446472168,
      "learning_rate": 6.459919278616019e-05,
      "loss": 1.1691,
      "mean_token_accuracy": 0.7722525209188461,
      "num_tokens": 1456295247.0,
      "step": 21560
    },
    {
      "entropy": 0.7638898327946663,
      "epoch": 6.534423994546694,
      "grad_norm": 0.1864500343799591,
      "learning_rate": 6.456725122340724e-05,
      "loss": 1.1574,
      "mean_token_accuracy": 0.7794457793235778,
      "num_tokens": 1456982654.0,
      "step": 21570
    },
    {
      "entropy": 0.7689558789134026,
      "epoch": 6.537453609028251,
      "grad_norm": 0.18878714740276337,
      "learning_rate": 6.45353031629315e-05,
      "loss": 1.1719,
      "mean_token_accuracy": 0.7743179693818092,
      "num_tokens": 1457656508.0,
      "step": 21580
    },
    {
      "entropy": 0.7655672878026962,
      "epoch": 6.540483223509808,
      "grad_norm": 0.17761436104774475,
      "learning_rate": 6.450334861898344e-05,
      "loss": 1.1623,
      "mean_token_accuracy": 0.7732078984379769,
      "num_tokens": 1458336182.0,
      "step": 21590
    },
    {
      "entropy": 0.7692855253815651,
      "epoch": 6.543512837991366,
      "grad_norm": 0.1718793660402298,
      "learning_rate": 6.447138760581635e-05,
      "loss": 1.1692,
      "mean_token_accuracy": 0.7737704396247864,
      "num_tokens": 1459007287.0,
      "step": 21600
    },
    {
      "entropy": 0.7747060254216194,
      "epoch": 6.546542452472923,
      "grad_norm": 0.1755637675523758,
      "learning_rate": 6.443942013768645e-05,
      "loss": 1.171,
      "mean_token_accuracy": 0.7702451661229134,
      "num_tokens": 1459679061.0,
      "step": 21610
    },
    {
      "entropy": 0.7761596962809563,
      "epoch": 6.54957206695448,
      "grad_norm": 0.19001683592796326,
      "learning_rate": 6.440744622885284e-05,
      "loss": 1.1686,
      "mean_token_accuracy": 0.7771530330181122,
      "num_tokens": 1460374429.0,
      "step": 21620
    },
    {
      "entropy": 0.764761796593666,
      "epoch": 6.5526016814360375,
      "grad_norm": 0.1785161793231964,
      "learning_rate": 6.437546589357748e-05,
      "loss": 1.164,
      "mean_token_accuracy": 0.7729075953364373,
      "num_tokens": 1461049779.0,
      "step": 21630
    },
    {
      "entropy": 0.7704018160700798,
      "epoch": 6.555631295917594,
      "grad_norm": 0.19004888832569122,
      "learning_rate": 6.434347914612519e-05,
      "loss": 1.1635,
      "mean_token_accuracy": 0.7719914078712463,
      "num_tokens": 1461727645.0,
      "step": 21640
    },
    {
      "entropy": 0.7683401018381119,
      "epoch": 6.558660910399151,
      "grad_norm": 0.171677827835083,
      "learning_rate": 6.431148600076368e-05,
      "loss": 1.1649,
      "mean_token_accuracy": 0.7729918599128723,
      "num_tokens": 1462404520.0,
      "step": 21650
    },
    {
      "entropy": 0.7704039484262466,
      "epoch": 6.561690524880709,
      "grad_norm": 0.1909465491771698,
      "learning_rate": 6.427948647176347e-05,
      "loss": 1.1679,
      "mean_token_accuracy": 0.7740996435284615,
      "num_tokens": 1463080948.0,
      "step": 21660
    },
    {
      "entropy": 0.7697620555758476,
      "epoch": 6.564720139362266,
      "grad_norm": 0.18859635293483734,
      "learning_rate": 6.424748057339795e-05,
      "loss": 1.1644,
      "mean_token_accuracy": 0.7722744092345237,
      "num_tokens": 1463750087.0,
      "step": 21670
    },
    {
      "entropy": 0.7678390488028526,
      "epoch": 6.567749753843824,
      "grad_norm": 0.18070949614048004,
      "learning_rate": 6.421546831994338e-05,
      "loss": 1.1629,
      "mean_token_accuracy": 0.7758280858397484,
      "num_tokens": 1464428117.0,
      "step": 21680
    },
    {
      "entropy": 0.7706129759550094,
      "epoch": 6.570779368325381,
      "grad_norm": 0.18724402785301208,
      "learning_rate": 6.418344972567881e-05,
      "loss": 1.1759,
      "mean_token_accuracy": 0.7726528882980347,
      "num_tokens": 1465103570.0,
      "step": 21690
    },
    {
      "entropy": 0.7660466030240058,
      "epoch": 6.573808982806938,
      "grad_norm": 0.19266277551651,
      "learning_rate": 6.415142480488612e-05,
      "loss": 1.1709,
      "mean_token_accuracy": 0.7720398455858231,
      "num_tokens": 1465771173.0,
      "step": 21700
    },
    {
      "entropy": 0.7690909415483475,
      "epoch": 6.576838597288495,
      "grad_norm": 0.18735866248607635,
      "learning_rate": 6.411939357185004e-05,
      "loss": 1.1704,
      "mean_token_accuracy": 0.7696846678853035,
      "num_tokens": 1466440331.0,
      "step": 21710
    },
    {
      "entropy": 0.7741300716996193,
      "epoch": 6.579868211770052,
      "grad_norm": 0.18521477282047272,
      "learning_rate": 6.40873560408581e-05,
      "loss": 1.17,
      "mean_token_accuracy": 0.7761170193552971,
      "num_tokens": 1467129411.0,
      "step": 21720
    },
    {
      "entropy": 0.7730172485113144,
      "epoch": 6.582897826251609,
      "grad_norm": 0.18900567293167114,
      "learning_rate": 6.405531222620065e-05,
      "loss": 1.1762,
      "mean_token_accuracy": 0.7701623916625977,
      "num_tokens": 1467795570.0,
      "step": 21730
    },
    {
      "entropy": 0.7690958172082901,
      "epoch": 6.585927440733167,
      "grad_norm": 0.1802007108926773,
      "learning_rate": 6.402326214217082e-05,
      "loss": 1.161,
      "mean_token_accuracy": 0.7755455136299133,
      "num_tokens": 1468470270.0,
      "step": 21740
    },
    {
      "entropy": 0.7712338462471962,
      "epoch": 6.588957055214724,
      "grad_norm": 0.1835826337337494,
      "learning_rate": 6.399120580306456e-05,
      "loss": 1.1656,
      "mean_token_accuracy": 0.7762703418731689,
      "num_tokens": 1469154893.0,
      "step": 21750
    },
    {
      "entropy": 0.7653421729803085,
      "epoch": 6.591986669696281,
      "grad_norm": 0.18978463113307953,
      "learning_rate": 6.395914322318062e-05,
      "loss": 1.1663,
      "mean_token_accuracy": 0.7751046478748321,
      "num_tokens": 1469823552.0,
      "step": 21760
    },
    {
      "entropy": 0.7720669090747834,
      "epoch": 6.595016284177838,
      "grad_norm": 0.18262676894664764,
      "learning_rate": 6.392707441682047e-05,
      "loss": 1.1645,
      "mean_token_accuracy": 0.7753222897648812,
      "num_tokens": 1470519341.0,
      "step": 21770
    },
    {
      "entropy": 0.7722769737243652,
      "epoch": 6.598045898659396,
      "grad_norm": 0.1838468313217163,
      "learning_rate": 6.389499939828842e-05,
      "loss": 1.1625,
      "mean_token_accuracy": 0.7714739501476288,
      "num_tokens": 1471192707.0,
      "step": 21780
    },
    {
      "entropy": 0.7652555599808692,
      "epoch": 6.601075513140953,
      "grad_norm": 0.183171346783638,
      "learning_rate": 6.386291818189157e-05,
      "loss": 1.169,
      "mean_token_accuracy": 0.7731975227594375,
      "num_tokens": 1471864045.0,
      "step": 21790
    },
    {
      "entropy": 0.7602854669094086,
      "epoch": 6.6041051276225105,
      "grad_norm": 0.18502184748649597,
      "learning_rate": 6.383083078193969e-05,
      "loss": 1.1561,
      "mean_token_accuracy": 0.774936942756176,
      "num_tokens": 1472549992.0,
      "step": 21800
    },
    {
      "entropy": 0.7682442143559456,
      "epoch": 6.607134742104067,
      "grad_norm": 0.18116509914398193,
      "learning_rate": 6.37987372127454e-05,
      "loss": 1.1615,
      "mean_token_accuracy": 0.770703761279583,
      "num_tokens": 1473221478.0,
      "step": 21810
    },
    {
      "entropy": 0.7739316254854203,
      "epoch": 6.610164356585624,
      "grad_norm": 0.18188564479351044,
      "learning_rate": 6.376663748862404e-05,
      "loss": 1.1745,
      "mean_token_accuracy": 0.7728255540132523,
      "num_tokens": 1473897881.0,
      "step": 21820
    },
    {
      "entropy": 0.7783873111009598,
      "epoch": 6.613193971067181,
      "grad_norm": 0.1905154585838318,
      "learning_rate": 6.373453162389365e-05,
      "loss": 1.1727,
      "mean_token_accuracy": 0.7709947720170021,
      "num_tokens": 1474562263.0,
      "step": 21830
    },
    {
      "entropy": 0.7556312188506127,
      "epoch": 6.616223585548739,
      "grad_norm": 0.19350853562355042,
      "learning_rate": 6.370241963287511e-05,
      "loss": 1.1541,
      "mean_token_accuracy": 0.774929678440094,
      "num_tokens": 1475234089.0,
      "step": 21840
    },
    {
      "entropy": 0.7703248292207718,
      "epoch": 6.619253200030296,
      "grad_norm": 0.18423984944820404,
      "learning_rate": 6.367030152989195e-05,
      "loss": 1.1741,
      "mean_token_accuracy": 0.7735764056444168,
      "num_tokens": 1475911688.0,
      "step": 21850
    },
    {
      "entropy": 0.7765189647674561,
      "epoch": 6.622282814511854,
      "grad_norm": 0.18914060294628143,
      "learning_rate": 6.363817732927044e-05,
      "loss": 1.1727,
      "mean_token_accuracy": 0.7717511236667634,
      "num_tokens": 1476587303.0,
      "step": 21860
    },
    {
      "entropy": 0.768643818795681,
      "epoch": 6.625312428993411,
      "grad_norm": 0.18642932176589966,
      "learning_rate": 6.360604704533959e-05,
      "loss": 1.1701,
      "mean_token_accuracy": 0.7708021387457847,
      "num_tokens": 1477255880.0,
      "step": 21870
    },
    {
      "entropy": 0.772607932984829,
      "epoch": 6.628342043474968,
      "grad_norm": 0.18266133964061737,
      "learning_rate": 6.357391069243112e-05,
      "loss": 1.1767,
      "mean_token_accuracy": 0.7690922453999519,
      "num_tokens": 1477933555.0,
      "step": 21880
    },
    {
      "entropy": 0.7673264116048812,
      "epoch": 6.631371657956525,
      "grad_norm": 0.18531456589698792,
      "learning_rate": 6.354176828487943e-05,
      "loss": 1.1597,
      "mean_token_accuracy": 0.7732823312282562,
      "num_tokens": 1478602707.0,
      "step": 21890
    },
    {
      "entropy": 0.7652122467756272,
      "epoch": 6.634401272438082,
      "grad_norm": 0.17667856812477112,
      "learning_rate": 6.350961983702167e-05,
      "loss": 1.1634,
      "mean_token_accuracy": 0.7752939403057099,
      "num_tokens": 1479286804.0,
      "step": 21900
    },
    {
      "entropy": 0.7755970239639283,
      "epoch": 6.637430886919639,
      "grad_norm": 0.19186754524707794,
      "learning_rate": 6.347746536319763e-05,
      "loss": 1.1706,
      "mean_token_accuracy": 0.7728856489062309,
      "num_tokens": 1479958048.0,
      "step": 21910
    },
    {
      "entropy": 0.7646700069308281,
      "epoch": 6.640460501401197,
      "grad_norm": 0.1853509545326233,
      "learning_rate": 6.344530487774982e-05,
      "loss": 1.1619,
      "mean_token_accuracy": 0.7750240325927734,
      "num_tokens": 1480637890.0,
      "step": 21920
    },
    {
      "entropy": 0.7710616961121559,
      "epoch": 6.643490115882754,
      "grad_norm": 0.1920197606086731,
      "learning_rate": 6.341313839502345e-05,
      "loss": 1.1705,
      "mean_token_accuracy": 0.7739571809768677,
      "num_tokens": 1481311429.0,
      "step": 21930
    },
    {
      "entropy": 0.7739026889204978,
      "epoch": 6.646519730364311,
      "grad_norm": 0.1981394737958908,
      "learning_rate": 6.338096592936635e-05,
      "loss": 1.174,
      "mean_token_accuracy": 0.7715424120426178,
      "num_tokens": 1481981945.0,
      "step": 21940
    },
    {
      "entropy": 0.7698251903057098,
      "epoch": 6.649549344845869,
      "grad_norm": 0.18606361746788025,
      "learning_rate": 6.334878749512906e-05,
      "loss": 1.1716,
      "mean_token_accuracy": 0.7694026336073876,
      "num_tokens": 1482652948.0,
      "step": 21950
    },
    {
      "entropy": 0.7713384702801704,
      "epoch": 6.652578959327426,
      "grad_norm": 0.19235661625862122,
      "learning_rate": 6.331660310666479e-05,
      "loss": 1.1645,
      "mean_token_accuracy": 0.7726158857345581,
      "num_tokens": 1483326600.0,
      "step": 21960
    },
    {
      "entropy": 0.7762814402580261,
      "epoch": 6.6556085738089825,
      "grad_norm": 0.1893935203552246,
      "learning_rate": 6.328441277832934e-05,
      "loss": 1.1677,
      "mean_token_accuracy": 0.7695004731416702,
      "num_tokens": 1484003122.0,
      "step": 21970
    },
    {
      "entropy": 0.7783913135528564,
      "epoch": 6.65863818829054,
      "grad_norm": 0.18274828791618347,
      "learning_rate": 6.325221652448124e-05,
      "loss": 1.1739,
      "mean_token_accuracy": 0.7683811485767365,
      "num_tokens": 1484668823.0,
      "step": 21980
    },
    {
      "entropy": 0.7588885337114334,
      "epoch": 6.661667802772097,
      "grad_norm": 0.18802721798419952,
      "learning_rate": 6.322001435948166e-05,
      "loss": 1.1531,
      "mean_token_accuracy": 0.7772417977452278,
      "num_tokens": 1485344835.0,
      "step": 21990
    },
    {
      "entropy": 0.7634912580251694,
      "epoch": 6.664697417253654,
      "grad_norm": 0.18112128973007202,
      "learning_rate": 6.31878062976943e-05,
      "loss": 1.1654,
      "mean_token_accuracy": 0.773204405605793,
      "num_tokens": 1486008505.0,
      "step": 22000
    },
    {
      "entropy": 0.7671458825469017,
      "epoch": 6.667727031735212,
      "grad_norm": 0.19336499273777008,
      "learning_rate": 6.315559235348563e-05,
      "loss": 1.1686,
      "mean_token_accuracy": 0.7707120925188065,
      "num_tokens": 1486676616.0,
      "step": 22010
    },
    {
      "entropy": 0.7553473100066185,
      "epoch": 6.670756646216769,
      "grad_norm": 0.1731991022825241,
      "learning_rate": 6.312337254122465e-05,
      "loss": 1.1478,
      "mean_token_accuracy": 0.7773711651563644,
      "num_tokens": 1487356232.0,
      "step": 22020
    },
    {
      "entropy": 0.7643198490142822,
      "epoch": 6.673786260698326,
      "grad_norm": 0.18662288784980774,
      "learning_rate": 6.309114687528305e-05,
      "loss": 1.1653,
      "mean_token_accuracy": 0.7735546737909317,
      "num_tokens": 1488033523.0,
      "step": 22030
    },
    {
      "entropy": 0.7643371403217316,
      "epoch": 6.676815875179884,
      "grad_norm": 0.18357449769973755,
      "learning_rate": 6.305891537003502e-05,
      "loss": 1.1726,
      "mean_token_accuracy": 0.7715471386909485,
      "num_tokens": 1488705079.0,
      "step": 22040
    },
    {
      "entropy": 0.77202457934618,
      "epoch": 6.6798454896614405,
      "grad_norm": 0.18090148270130157,
      "learning_rate": 6.302667803985747e-05,
      "loss": 1.1682,
      "mean_token_accuracy": 0.7699681699275971,
      "num_tokens": 1489367689.0,
      "step": 22050
    },
    {
      "entropy": 0.774673941731453,
      "epoch": 6.6828751041429975,
      "grad_norm": 0.1801070272922516,
      "learning_rate": 6.299443489912985e-05,
      "loss": 1.1738,
      "mean_token_accuracy": 0.7740351021289825,
      "num_tokens": 1490052150.0,
      "step": 22060
    },
    {
      "entropy": 0.7639446377754211,
      "epoch": 6.685904718624555,
      "grad_norm": 0.18547028303146362,
      "learning_rate": 6.296218596223422e-05,
      "loss": 1.1599,
      "mean_token_accuracy": 0.7719153627753258,
      "num_tokens": 1490717649.0,
      "step": 22070
    },
    {
      "entropy": 0.7652400135993958,
      "epoch": 6.688934333106112,
      "grad_norm": 0.18799661099910736,
      "learning_rate": 6.292993124355522e-05,
      "loss": 1.1591,
      "mean_token_accuracy": 0.7754050731658936,
      "num_tokens": 1491397320.0,
      "step": 22080
    },
    {
      "entropy": 0.770381124317646,
      "epoch": 6.691963947587669,
      "grad_norm": 0.17602533102035522,
      "learning_rate": 6.289767075748005e-05,
      "loss": 1.1703,
      "mean_token_accuracy": 0.7748016655445099,
      "num_tokens": 1492081050.0,
      "step": 22090
    },
    {
      "entropy": 0.7705921873450279,
      "epoch": 6.694993562069227,
      "grad_norm": 0.1815701276063919,
      "learning_rate": 6.286540451839853e-05,
      "loss": 1.165,
      "mean_token_accuracy": 0.7728049546480179,
      "num_tokens": 1492769410.0,
      "step": 22100
    },
    {
      "entropy": 0.7681502133607865,
      "epoch": 6.698023176550784,
      "grad_norm": 0.18458382785320282,
      "learning_rate": 6.2833132540703e-05,
      "loss": 1.1657,
      "mean_token_accuracy": 0.768888546526432,
      "num_tokens": 1493441838.0,
      "step": 22110
    },
    {
      "entropy": 0.774550287425518,
      "epoch": 6.701052791032341,
      "grad_norm": 0.19067157804965973,
      "learning_rate": 6.280085483878837e-05,
      "loss": 1.1717,
      "mean_token_accuracy": 0.772263289988041,
      "num_tokens": 1494120881.0,
      "step": 22120
    },
    {
      "entropy": 0.7727632358670234,
      "epoch": 6.704082405513899,
      "grad_norm": 0.1965942680835724,
      "learning_rate": 6.276857142705213e-05,
      "loss": 1.1788,
      "mean_token_accuracy": 0.7671290412545204,
      "num_tokens": 1494778778.0,
      "step": 22130
    },
    {
      "entropy": 0.7777154386043549,
      "epoch": 6.7071120199954555,
      "grad_norm": 0.19251009821891785,
      "learning_rate": 6.273628231989426e-05,
      "loss": 1.1733,
      "mean_token_accuracy": 0.7677260801196099,
      "num_tokens": 1495445881.0,
      "step": 22140
    },
    {
      "entropy": 0.7793846324086189,
      "epoch": 6.710141634477013,
      "grad_norm": 0.1896645426750183,
      "learning_rate": 6.270398753171734e-05,
      "loss": 1.1696,
      "mean_token_accuracy": 0.7731392189860344,
      "num_tokens": 1496131151.0,
      "step": 22150
    },
    {
      "entropy": 0.7714336290955544,
      "epoch": 6.71317124895857,
      "grad_norm": 0.18072839081287384,
      "learning_rate": 6.267168707692649e-05,
      "loss": 1.1653,
      "mean_token_accuracy": 0.771985012292862,
      "num_tokens": 1496805690.0,
      "step": 22160
    },
    {
      "entropy": 0.7614537060260773,
      "epoch": 6.716200863440127,
      "grad_norm": 0.2017190158367157,
      "learning_rate": 6.263938096992929e-05,
      "loss": 1.1659,
      "mean_token_accuracy": 0.7736597269773483,
      "num_tokens": 1497480078.0,
      "step": 22170
    },
    {
      "entropy": 0.7740855038166046,
      "epoch": 6.719230477921684,
      "grad_norm": 0.1913452297449112,
      "learning_rate": 6.260706922513587e-05,
      "loss": 1.1662,
      "mean_token_accuracy": 0.7732095837593078,
      "num_tokens": 1498160528.0,
      "step": 22180
    },
    {
      "entropy": 0.7760282531380653,
      "epoch": 6.722260092403242,
      "grad_norm": 0.18119335174560547,
      "learning_rate": 6.257475185695891e-05,
      "loss": 1.1722,
      "mean_token_accuracy": 0.7678147166967392,
      "num_tokens": 1498825082.0,
      "step": 22190
    },
    {
      "entropy": 0.7587352946400643,
      "epoch": 6.725289706884799,
      "grad_norm": 0.18721026182174683,
      "learning_rate": 6.254242887981359e-05,
      "loss": 1.1595,
      "mean_token_accuracy": 0.7760735020041466,
      "num_tokens": 1499499710.0,
      "step": 22200
    },
    {
      "entropy": 0.77638940513134,
      "epoch": 6.728319321366357,
      "grad_norm": 0.19338718056678772,
      "learning_rate": 6.251010030811755e-05,
      "loss": 1.1744,
      "mean_token_accuracy": 0.7720439538359642,
      "num_tokens": 1500183912.0,
      "step": 22210
    },
    {
      "entropy": 0.7665472030639648,
      "epoch": 6.7313489358479135,
      "grad_norm": 0.18072675168514252,
      "learning_rate": 6.247776615629094e-05,
      "loss": 1.1615,
      "mean_token_accuracy": 0.7728681981563568,
      "num_tokens": 1500854050.0,
      "step": 22220
    },
    {
      "entropy": 0.7659976661205292,
      "epoch": 6.73437855032947,
      "grad_norm": 0.1857847422361374,
      "learning_rate": 6.244542643875642e-05,
      "loss": 1.1614,
      "mean_token_accuracy": 0.7746493652462959,
      "num_tokens": 1501533385.0,
      "step": 22230
    },
    {
      "entropy": 0.7666662365198136,
      "epoch": 6.737408164811027,
      "grad_norm": 0.19038133323192596,
      "learning_rate": 6.241308116993914e-05,
      "loss": 1.1625,
      "mean_token_accuracy": 0.7739652186632157,
      "num_tokens": 1502208304.0,
      "step": 22240
    },
    {
      "entropy": 0.7609266832470893,
      "epoch": 6.740437779292585,
      "grad_norm": 0.17761822044849396,
      "learning_rate": 6.238073036426669e-05,
      "loss": 1.1628,
      "mean_token_accuracy": 0.7767347857356072,
      "num_tokens": 1502885874.0,
      "step": 22250
    },
    {
      "entropy": 0.77239308655262,
      "epoch": 6.743467393774142,
      "grad_norm": 0.18619075417518616,
      "learning_rate": 6.234837403616912e-05,
      "loss": 1.1704,
      "mean_token_accuracy": 0.7716382130980491,
      "num_tokens": 1503555544.0,
      "step": 22260
    },
    {
      "entropy": 0.7584697559475899,
      "epoch": 6.7464970082557,
      "grad_norm": 0.18827976286411285,
      "learning_rate": 6.231601220007903e-05,
      "loss": 1.1565,
      "mean_token_accuracy": 0.7730544328689575,
      "num_tokens": 1504216227.0,
      "step": 22270
    },
    {
      "entropy": 0.7744659379124641,
      "epoch": 6.749526622737257,
      "grad_norm": 0.1921151727437973,
      "learning_rate": 6.228364487043137e-05,
      "loss": 1.1652,
      "mean_token_accuracy": 0.769395387172699,
      "num_tokens": 1504883190.0,
      "step": 22280
    },
    {
      "entropy": 0.7673099309206008,
      "epoch": 6.752556237218814,
      "grad_norm": 0.2096191793680191,
      "learning_rate": 6.225127206166363e-05,
      "loss": 1.1722,
      "mean_token_accuracy": 0.772365789115429,
      "num_tokens": 1505552306.0,
      "step": 22290
    },
    {
      "entropy": 0.7724889099597931,
      "epoch": 6.7555858517003715,
      "grad_norm": 0.17974375188350677,
      "learning_rate": 6.221889378821566e-05,
      "loss": 1.1698,
      "mean_token_accuracy": 0.7738417014479637,
      "num_tokens": 1506235893.0,
      "step": 22300
    },
    {
      "entropy": 0.7580396920442581,
      "epoch": 6.758615466181928,
      "grad_norm": 0.18559345602989197,
      "learning_rate": 6.218651006452981e-05,
      "loss": 1.1575,
      "mean_token_accuracy": 0.774227486550808,
      "num_tokens": 1506920520.0,
      "step": 22310
    },
    {
      "entropy": 0.7742979377508163,
      "epoch": 6.761645080663485,
      "grad_norm": 0.18375980854034424,
      "learning_rate": 6.215412090505084e-05,
      "loss": 1.1659,
      "mean_token_accuracy": 0.7725578680634498,
      "num_tokens": 1507599440.0,
      "step": 22320
    },
    {
      "entropy": 0.7670708894729614,
      "epoch": 6.764674695145043,
      "grad_norm": 0.18639779090881348,
      "learning_rate": 6.212172632422594e-05,
      "loss": 1.1623,
      "mean_token_accuracy": 0.7718268305063247,
      "num_tokens": 1508286070.0,
      "step": 22330
    },
    {
      "entropy": 0.7699150651693344,
      "epoch": 6.7677043096266,
      "grad_norm": 0.1727694272994995,
      "learning_rate": 6.20893263365047e-05,
      "loss": 1.1667,
      "mean_token_accuracy": 0.7740846604108811,
      "num_tokens": 1508967305.0,
      "step": 22340
    },
    {
      "entropy": 0.7649823412299156,
      "epoch": 6.770733924108157,
      "grad_norm": 0.1860731542110443,
      "learning_rate": 6.205692095633915e-05,
      "loss": 1.1636,
      "mean_token_accuracy": 0.7753878831863403,
      "num_tokens": 1509649444.0,
      "step": 22350
    },
    {
      "entropy": 0.7677910998463631,
      "epoch": 6.773763538589715,
      "grad_norm": 0.1788383275270462,
      "learning_rate": 6.20245101981837e-05,
      "loss": 1.1583,
      "mean_token_accuracy": 0.778513939678669,
      "num_tokens": 1510344184.0,
      "step": 22360
    },
    {
      "entropy": 0.7659306794404983,
      "epoch": 6.776793153071272,
      "grad_norm": 0.18011821806430817,
      "learning_rate": 6.199209407649518e-05,
      "loss": 1.1654,
      "mean_token_accuracy": 0.7731028303503991,
      "num_tokens": 1511023896.0,
      "step": 22370
    },
    {
      "entropy": 0.7609267979860306,
      "epoch": 6.779822767552829,
      "grad_norm": 0.1777167022228241,
      "learning_rate": 6.195967260573278e-05,
      "loss": 1.1689,
      "mean_token_accuracy": 0.7709998160600662,
      "num_tokens": 1511678266.0,
      "step": 22380
    },
    {
      "entropy": 0.7727775529026986,
      "epoch": 6.7828523820343865,
      "grad_norm": 0.18532463908195496,
      "learning_rate": 6.192724580035814e-05,
      "loss": 1.1707,
      "mean_token_accuracy": 0.7709300830960274,
      "num_tokens": 1512354459.0,
      "step": 22390
    },
    {
      "entropy": 0.7810803890228272,
      "epoch": 6.785881996515943,
      "grad_norm": 0.18865905702114105,
      "learning_rate": 6.189481367483518e-05,
      "loss": 1.1774,
      "mean_token_accuracy": 0.7638928830623627,
      "num_tokens": 1513012403.0,
      "step": 22400
    },
    {
      "entropy": 0.7617676466703415,
      "epoch": 6.7889116109975,
      "grad_norm": 0.18379583954811096,
      "learning_rate": 6.18623762436303e-05,
      "loss": 1.1525,
      "mean_token_accuracy": 0.7848607137799263,
      "num_tokens": 1513730338.0,
      "step": 22410
    },
    {
      "entropy": 0.7729958280920982,
      "epoch": 6.791941225479058,
      "grad_norm": 0.18827588856220245,
      "learning_rate": 6.182993352121219e-05,
      "loss": 1.1645,
      "mean_token_accuracy": 0.7752539649605751,
      "num_tokens": 1514425220.0,
      "step": 22420
    },
    {
      "entropy": 0.7729710429906845,
      "epoch": 6.794970839960615,
      "grad_norm": 0.18430310487747192,
      "learning_rate": 6.179748552205193e-05,
      "loss": 1.168,
      "mean_token_accuracy": 0.772411173582077,
      "num_tokens": 1515098424.0,
      "step": 22430
    },
    {
      "entropy": 0.7681291326880455,
      "epoch": 6.798000454442172,
      "grad_norm": 0.17213250696659088,
      "learning_rate": 6.176503226062298e-05,
      "loss": 1.1613,
      "mean_token_accuracy": 0.7747751772403717,
      "num_tokens": 1515772411.0,
      "step": 22440
    },
    {
      "entropy": 0.775424613058567,
      "epoch": 6.80103006892373,
      "grad_norm": 0.19226394593715668,
      "learning_rate": 6.173257375140107e-05,
      "loss": 1.1722,
      "mean_token_accuracy": 0.7726172283291817,
      "num_tokens": 1516450283.0,
      "step": 22450
    },
    {
      "entropy": 0.7723731324076653,
      "epoch": 6.804059683405287,
      "grad_norm": 0.18939033150672913,
      "learning_rate": 6.170011000886436e-05,
      "loss": 1.1627,
      "mean_token_accuracy": 0.7734440341591835,
      "num_tokens": 1517135844.0,
      "step": 22460
    },
    {
      "entropy": 0.7599155440926552,
      "epoch": 6.807089297886844,
      "grad_norm": 0.1787664294242859,
      "learning_rate": 6.166764104749328e-05,
      "loss": 1.1578,
      "mean_token_accuracy": 0.7775285467505455,
      "num_tokens": 1517815583.0,
      "step": 22470
    },
    {
      "entropy": 0.7809656620025635,
      "epoch": 6.810118912368401,
      "grad_norm": 0.1927752047777176,
      "learning_rate": 6.163516688177061e-05,
      "loss": 1.1796,
      "mean_token_accuracy": 0.7675548031926155,
      "num_tokens": 1518483753.0,
      "step": 22480
    },
    {
      "entropy": 0.7554050832986832,
      "epoch": 6.813148526849958,
      "grad_norm": 0.18380290269851685,
      "learning_rate": 6.160268752618145e-05,
      "loss": 1.1625,
      "mean_token_accuracy": 0.7777298867702485,
      "num_tokens": 1519169091.0,
      "step": 22490
    },
    {
      "entropy": 0.7745677649974823,
      "epoch": 6.816178141331515,
      "grad_norm": 0.17235508561134338,
      "learning_rate": 6.157020299521326e-05,
      "loss": 1.1675,
      "mean_token_accuracy": 0.7761607438325882,
      "num_tokens": 1519864001.0,
      "step": 22500
    },
    {
      "entropy": 0.7699503168463707,
      "epoch": 6.819207755813073,
      "grad_norm": 0.18010473251342773,
      "learning_rate": 6.15377133033557e-05,
      "loss": 1.1685,
      "mean_token_accuracy": 0.7711123511195183,
      "num_tokens": 1520532427.0,
      "step": 22510
    },
    {
      "entropy": 0.7693112701177597,
      "epoch": 6.82223737029463,
      "grad_norm": 0.1972823143005371,
      "learning_rate": 6.150521846510084e-05,
      "loss": 1.1651,
      "mean_token_accuracy": 0.7719914734363555,
      "num_tokens": 1521204222.0,
      "step": 22520
    },
    {
      "entropy": 0.766348859667778,
      "epoch": 6.825266984776187,
      "grad_norm": 0.18678975105285645,
      "learning_rate": 6.147271849494301e-05,
      "loss": 1.1738,
      "mean_token_accuracy": 0.7723152801394463,
      "num_tokens": 1521879719.0,
      "step": 22530
    },
    {
      "entropy": 0.7674392148852348,
      "epoch": 6.828296599257745,
      "grad_norm": 0.17786964774131775,
      "learning_rate": 6.14402134073788e-05,
      "loss": 1.1724,
      "mean_token_accuracy": 0.776619590818882,
      "num_tokens": 1522563588.0,
      "step": 22540
    },
    {
      "entropy": 0.7676349014043808,
      "epoch": 6.831326213739302,
      "grad_norm": 0.18887832760810852,
      "learning_rate": 6.14077032169071e-05,
      "loss": 1.1651,
      "mean_token_accuracy": 0.7697415858507156,
      "num_tokens": 1523221949.0,
      "step": 22550
    },
    {
      "entropy": 0.7809972450137138,
      "epoch": 6.8343558282208585,
      "grad_norm": 0.1912042498588562,
      "learning_rate": 6.137518793802911e-05,
      "loss": 1.1802,
      "mean_token_accuracy": 0.7686642140150071,
      "num_tokens": 1523895779.0,
      "step": 22560
    },
    {
      "entropy": 0.7668987795710563,
      "epoch": 6.837385442702416,
      "grad_norm": 0.17426718771457672,
      "learning_rate": 6.134266758524824e-05,
      "loss": 1.1566,
      "mean_token_accuracy": 0.772175969183445,
      "num_tokens": 1524587967.0,
      "step": 22570
    },
    {
      "entropy": 0.768090882897377,
      "epoch": 6.840415057183973,
      "grad_norm": 0.2009420245885849,
      "learning_rate": 6.131014217307021e-05,
      "loss": 1.165,
      "mean_token_accuracy": 0.7689484387636185,
      "num_tokens": 1525255486.0,
      "step": 22580
    },
    {
      "entropy": 0.7724968254566192,
      "epoch": 6.84344467166553,
      "grad_norm": 0.18546654284000397,
      "learning_rate": 6.127761171600298e-05,
      "loss": 1.1728,
      "mean_token_accuracy": 0.7720789045095444,
      "num_tokens": 1525928126.0,
      "step": 22590
    },
    {
      "entropy": 0.7581162869930267,
      "epoch": 6.846474286147088,
      "grad_norm": 0.18833152949810028,
      "learning_rate": 6.124507622855677e-05,
      "loss": 1.1618,
      "mean_token_accuracy": 0.7740279600024224,
      "num_tokens": 1526603842.0,
      "step": 22600
    },
    {
      "entropy": 0.766860842704773,
      "epoch": 6.849503900628645,
      "grad_norm": 0.1974152773618698,
      "learning_rate": 6.121253572524402e-05,
      "loss": 1.1679,
      "mean_token_accuracy": 0.7746059313416481,
      "num_tokens": 1527282804.0,
      "step": 22610
    },
    {
      "entropy": 0.7733130246400833,
      "epoch": 6.852533515110203,
      "grad_norm": 0.18934597074985504,
      "learning_rate": 6.117999022057942e-05,
      "loss": 1.1663,
      "mean_token_accuracy": 0.7716334849596024,
      "num_tokens": 1527951591.0,
      "step": 22620
    },
    {
      "entropy": 0.774915198981762,
      "epoch": 6.85556312959176,
      "grad_norm": 0.18406635522842407,
      "learning_rate": 6.11474397290799e-05,
      "loss": 1.1696,
      "mean_token_accuracy": 0.7750470638275146,
      "num_tokens": 1528639116.0,
      "step": 22630
    },
    {
      "entropy": 0.7649851545691491,
      "epoch": 6.8585927440733165,
      "grad_norm": 0.1802431344985962,
      "learning_rate": 6.111488426526462e-05,
      "loss": 1.1686,
      "mean_token_accuracy": 0.769770884513855,
      "num_tokens": 1529308239.0,
      "step": 22640
    },
    {
      "entropy": 0.7749467492103577,
      "epoch": 6.8616223585548735,
      "grad_norm": 0.16982093453407288,
      "learning_rate": 6.108232384365492e-05,
      "loss": 1.1702,
      "mean_token_accuracy": 0.7691055357456207,
      "num_tokens": 1529975463.0,
      "step": 22650
    },
    {
      "entropy": 0.7667987331748009,
      "epoch": 6.864651973036431,
      "grad_norm": 0.18780744075775146,
      "learning_rate": 6.10497584787744e-05,
      "loss": 1.1663,
      "mean_token_accuracy": 0.7754819333553314,
      "num_tokens": 1530655183.0,
      "step": 22660
    },
    {
      "entropy": 0.7683945238590241,
      "epoch": 6.867681587517988,
      "grad_norm": 0.18448594212532043,
      "learning_rate": 6.1017188185148835e-05,
      "loss": 1.1593,
      "mean_token_accuracy": 0.78026642203331,
      "num_tokens": 1531356516.0,
      "step": 22670
    },
    {
      "entropy": 0.7647533610463142,
      "epoch": 6.870711201999546,
      "grad_norm": 0.17418862879276276,
      "learning_rate": 6.09846129773062e-05,
      "loss": 1.1617,
      "mean_token_accuracy": 0.7716727629303932,
      "num_tokens": 1532025501.0,
      "step": 22680
    },
    {
      "entropy": 0.771768918633461,
      "epoch": 6.873740816481103,
      "grad_norm": 0.21458394825458527,
      "learning_rate": 6.0952032869776666e-05,
      "loss": 1.1736,
      "mean_token_accuracy": 0.7672025233507156,
      "num_tokens": 1532679284.0,
      "step": 22690
    },
    {
      "entropy": 0.7655633240938187,
      "epoch": 6.87677043096266,
      "grad_norm": 0.18572165071964264,
      "learning_rate": 6.091944787709258e-05,
      "loss": 1.161,
      "mean_token_accuracy": 0.776415741443634,
      "num_tokens": 1533366501.0,
      "step": 22700
    },
    {
      "entropy": 0.7642249599099159,
      "epoch": 6.879800045444217,
      "grad_norm": 0.1808936893939972,
      "learning_rate": 6.088685801378853e-05,
      "loss": 1.1658,
      "mean_token_accuracy": 0.7749256417155266,
      "num_tokens": 1534042272.0,
      "step": 22710
    },
    {
      "entropy": 0.7605438247323036,
      "epoch": 6.882829659925775,
      "grad_norm": 0.17921391129493713,
      "learning_rate": 6.085426329440116e-05,
      "loss": 1.1635,
      "mean_token_accuracy": 0.7730612322688103,
      "num_tokens": 1534703535.0,
      "step": 22720
    },
    {
      "entropy": 0.7646952569484711,
      "epoch": 6.8858592744073315,
      "grad_norm": 0.20235541462898254,
      "learning_rate": 6.082166373346939e-05,
      "loss": 1.1583,
      "mean_token_accuracy": 0.7779818847775459,
      "num_tokens": 1535396020.0,
      "step": 22730
    },
    {
      "entropy": 0.7687417015433311,
      "epoch": 6.888888888888889,
      "grad_norm": 0.18874038755893707,
      "learning_rate": 6.078905934553424e-05,
      "loss": 1.1653,
      "mean_token_accuracy": 0.7701189145445824,
      "num_tokens": 1536063943.0,
      "step": 22740
    },
    {
      "entropy": 0.7666957467794419,
      "epoch": 6.891918503370446,
      "grad_norm": 0.18202944099903107,
      "learning_rate": 6.075645014513889e-05,
      "loss": 1.1683,
      "mean_token_accuracy": 0.768748952448368,
      "num_tokens": 1536722118.0,
      "step": 22750
    },
    {
      "entropy": 0.7854179322719574,
      "epoch": 6.894948117852003,
      "grad_norm": 0.18317753076553345,
      "learning_rate": 6.0723836146828684e-05,
      "loss": 1.1766,
      "mean_token_accuracy": 0.7681457713246346,
      "num_tokens": 1537395544.0,
      "step": 22760
    },
    {
      "entropy": 0.764841052889824,
      "epoch": 6.897977732333561,
      "grad_norm": 0.19204257428646088,
      "learning_rate": 6.0691217365151085e-05,
      "loss": 1.1629,
      "mean_token_accuracy": 0.7735884115099907,
      "num_tokens": 1538066040.0,
      "step": 22770
    },
    {
      "entropy": 0.7716961309313775,
      "epoch": 6.901007346815118,
      "grad_norm": 0.18792079389095306,
      "learning_rate": 6.0658593814655726e-05,
      "loss": 1.1708,
      "mean_token_accuracy": 0.7709859684109688,
      "num_tokens": 1538731061.0,
      "step": 22780
    },
    {
      "entropy": 0.7698103070259095,
      "epoch": 6.904036961296675,
      "grad_norm": 0.16728675365447998,
      "learning_rate": 6.062596550989432e-05,
      "loss": 1.1686,
      "mean_token_accuracy": 0.7722794011235237,
      "num_tokens": 1539407709.0,
      "step": 22790
    },
    {
      "entropy": 0.7710490822792053,
      "epoch": 6.907066575778233,
      "grad_norm": 0.17876802384853363,
      "learning_rate": 6.059333246542074e-05,
      "loss": 1.1706,
      "mean_token_accuracy": 0.770564828813076,
      "num_tokens": 1540080998.0,
      "step": 22800
    },
    {
      "entropy": 0.7533868879079819,
      "epoch": 6.9100961902597895,
      "grad_norm": 0.1808808594942093,
      "learning_rate": 6.056069469579094e-05,
      "loss": 1.1546,
      "mean_token_accuracy": 0.77374277561903,
      "num_tokens": 1540747136.0,
      "step": 22810
    },
    {
      "entropy": 0.7582154646515846,
      "epoch": 6.913125804741346,
      "grad_norm": 0.19275827705860138,
      "learning_rate": 6.0528052215563e-05,
      "loss": 1.1647,
      "mean_token_accuracy": 0.7769809141755104,
      "num_tokens": 1541421610.0,
      "step": 22820
    },
    {
      "entropy": 0.7601933673024177,
      "epoch": 6.916155419222904,
      "grad_norm": 0.1873077005147934,
      "learning_rate": 6.04954050392971e-05,
      "loss": 1.1526,
      "mean_token_accuracy": 0.7743327051401139,
      "num_tokens": 1542100777.0,
      "step": 22830
    },
    {
      "entropy": 0.7621347069740295,
      "epoch": 6.919185033704461,
      "grad_norm": 0.19351033866405487,
      "learning_rate": 6.0462753181555534e-05,
      "loss": 1.1683,
      "mean_token_accuracy": 0.7717160657048225,
      "num_tokens": 1542765304.0,
      "step": 22840
    },
    {
      "entropy": 0.7661462962627411,
      "epoch": 6.922214648186018,
      "grad_norm": 0.185185506939888,
      "learning_rate": 6.0430096656902645e-05,
      "loss": 1.1634,
      "mean_token_accuracy": 0.7741788566112519,
      "num_tokens": 1543439607.0,
      "step": 22850
    },
    {
      "entropy": 0.7748598620295525,
      "epoch": 6.925244262667576,
      "grad_norm": 0.18341249227523804,
      "learning_rate": 6.0397435479904874e-05,
      "loss": 1.165,
      "mean_token_accuracy": 0.7725229263305664,
      "num_tokens": 1544125563.0,
      "step": 22860
    },
    {
      "entropy": 0.7673856854438782,
      "epoch": 6.928273877149133,
      "grad_norm": 0.181553915143013,
      "learning_rate": 6.0364769665130735e-05,
      "loss": 1.1672,
      "mean_token_accuracy": 0.7745458260178566,
      "num_tokens": 1544806082.0,
      "step": 22870
    },
    {
      "entropy": 0.7706713765859604,
      "epoch": 6.93130349163069,
      "grad_norm": 0.18595345318317413,
      "learning_rate": 6.0332099227150853e-05,
      "loss": 1.1586,
      "mean_token_accuracy": 0.7737122029066086,
      "num_tokens": 1545493338.0,
      "step": 22880
    },
    {
      "entropy": 0.766918708384037,
      "epoch": 6.9343331061122475,
      "grad_norm": 0.1871090978384018,
      "learning_rate": 6.0299424180537835e-05,
      "loss": 1.1558,
      "mean_token_accuracy": 0.776808249950409,
      "num_tokens": 1546180215.0,
      "step": 22890
    },
    {
      "entropy": 0.7650902613997459,
      "epoch": 6.937362720593804,
      "grad_norm": 0.17750787734985352,
      "learning_rate": 6.026674453986641e-05,
      "loss": 1.1636,
      "mean_token_accuracy": 0.7759310081601143,
      "num_tokens": 1546861315.0,
      "step": 22900
    },
    {
      "entropy": 0.7809863060712814,
      "epoch": 6.940392335075361,
      "grad_norm": 0.17797423899173737,
      "learning_rate": 6.0234060319713315e-05,
      "loss": 1.1713,
      "mean_token_accuracy": 0.7697636038064957,
      "num_tokens": 1547530157.0,
      "step": 22910
    },
    {
      "entropy": 0.7729805439710618,
      "epoch": 6.943421949556919,
      "grad_norm": 0.17471666634082794,
      "learning_rate": 6.020137153465737e-05,
      "loss": 1.1718,
      "mean_token_accuracy": 0.7714817345142364,
      "num_tokens": 1548212298.0,
      "step": 22920
    },
    {
      "entropy": 0.7698503375053406,
      "epoch": 6.946451564038476,
      "grad_norm": 0.17625947296619415,
      "learning_rate": 6.016867819927938e-05,
      "loss": 1.1691,
      "mean_token_accuracy": 0.7711612269282341,
      "num_tokens": 1548881160.0,
      "step": 22930
    },
    {
      "entropy": 0.7707154870033264,
      "epoch": 6.949481178520033,
      "grad_norm": 0.19445252418518066,
      "learning_rate": 6.013598032816222e-05,
      "loss": 1.1689,
      "mean_token_accuracy": 0.773430486023426,
      "num_tokens": 1549560622.0,
      "step": 22940
    },
    {
      "entropy": 0.7669893369078636,
      "epoch": 6.952510793001591,
      "grad_norm": 0.18977509438991547,
      "learning_rate": 6.01032779358908e-05,
      "loss": 1.1634,
      "mean_token_accuracy": 0.7739262640476227,
      "num_tokens": 1550239370.0,
      "step": 22950
    },
    {
      "entropy": 0.7711462885141372,
      "epoch": 6.955540407483148,
      "grad_norm": 0.1888669729232788,
      "learning_rate": 6.0070571037051994e-05,
      "loss": 1.1709,
      "mean_token_accuracy": 0.7699010357260704,
      "num_tokens": 1550903445.0,
      "step": 22960
    },
    {
      "entropy": 0.7590415224432945,
      "epoch": 6.958570021964705,
      "grad_norm": 0.1832433044910431,
      "learning_rate": 6.003785964623472e-05,
      "loss": 1.1532,
      "mean_token_accuracy": 0.7786305457353592,
      "num_tokens": 1551588450.0,
      "step": 22970
    },
    {
      "entropy": 0.7689649447798729,
      "epoch": 6.9615996364462625,
      "grad_norm": 0.1829010248184204,
      "learning_rate": 6.0005143778029906e-05,
      "loss": 1.1672,
      "mean_token_accuracy": 0.7758778482675552,
      "num_tokens": 1552276431.0,
      "step": 22980
    },
    {
      "entropy": 0.7748028218746186,
      "epoch": 6.964629250927819,
      "grad_norm": 0.1904628574848175,
      "learning_rate": 5.997242344703043e-05,
      "loss": 1.1729,
      "mean_token_accuracy": 0.7732190832495689,
      "num_tokens": 1552958727.0,
      "step": 22990
    },
    {
      "entropy": 0.7566230967640877,
      "epoch": 6.967658865409376,
      "grad_norm": 0.19712401926517487,
      "learning_rate": 5.993969866783125e-05,
      "loss": 1.1655,
      "mean_token_accuracy": 0.7763057157397271,
      "num_tokens": 1553626742.0,
      "step": 23000
    },
    {
      "entropy": 0.7784351959824563,
      "epoch": 6.970688479890934,
      "grad_norm": 0.1927061527967453,
      "learning_rate": 5.9906969455029226e-05,
      "loss": 1.1751,
      "mean_token_accuracy": 0.7713754266500473,
      "num_tokens": 1554302878.0,
      "step": 23010
    },
    {
      "entropy": 0.7713057413697243,
      "epoch": 6.973718094372491,
      "grad_norm": 0.19877652823925018,
      "learning_rate": 5.987423582322322e-05,
      "loss": 1.1678,
      "mean_token_accuracy": 0.7730776354670524,
      "num_tokens": 1554970858.0,
      "step": 23020
    },
    {
      "entropy": 0.7634402811527252,
      "epoch": 6.976747708854048,
      "grad_norm": 0.1868930608034134,
      "learning_rate": 5.984149778701409e-05,
      "loss": 1.1733,
      "mean_token_accuracy": 0.770755261182785,
      "num_tokens": 1555625363.0,
      "step": 23030
    },
    {
      "entropy": 0.772444897890091,
      "epoch": 6.979777323335606,
      "grad_norm": 0.1826222687959671,
      "learning_rate": 5.980875536100461e-05,
      "loss": 1.172,
      "mean_token_accuracy": 0.7747440591454506,
      "num_tokens": 1556310882.0,
      "step": 23040
    },
    {
      "entropy": 0.7708359241485596,
      "epoch": 6.982806937817163,
      "grad_norm": 0.18035846948623657,
      "learning_rate": 5.977600855979957e-05,
      "loss": 1.1605,
      "mean_token_accuracy": 0.7712452054023743,
      "num_tokens": 1556987180.0,
      "step": 23050
    },
    {
      "entropy": 0.7600368797779083,
      "epoch": 6.98583655229872,
      "grad_norm": 0.19632886350154877,
      "learning_rate": 5.974325739800565e-05,
      "loss": 1.1673,
      "mean_token_accuracy": 0.772319607436657,
      "num_tokens": 1557651327.0,
      "step": 23060
    },
    {
      "entropy": 0.7694717109203338,
      "epoch": 6.988866166780277,
      "grad_norm": 0.18624110519886017,
      "learning_rate": 5.971050189023153e-05,
      "loss": 1.1644,
      "mean_token_accuracy": 0.76870878636837,
      "num_tokens": 1558312900.0,
      "step": 23070
    },
    {
      "entropy": 0.7642167627811431,
      "epoch": 6.991895781261834,
      "grad_norm": 0.1855657994747162,
      "learning_rate": 5.9677742051087804e-05,
      "loss": 1.1592,
      "mean_token_accuracy": 0.7784095257520676,
      "num_tokens": 1559002299.0,
      "step": 23080
    },
    {
      "entropy": 0.7630846992135047,
      "epoch": 6.994925395743392,
      "grad_norm": 0.17961439490318298,
      "learning_rate": 5.964497789518699e-05,
      "loss": 1.1668,
      "mean_token_accuracy": 0.7715464368462562,
      "num_tokens": 1559668049.0,
      "step": 23090
    },
    {
      "entropy": 0.7648632317781449,
      "epoch": 6.997955010224949,
      "grad_norm": 0.19289350509643555,
      "learning_rate": 5.961220943714354e-05,
      "loss": 1.1593,
      "mean_token_accuracy": 0.7767577812075614,
      "num_tokens": 1560354052.0,
      "step": 23100
    },
    {
      "entropy": 0.769464543232551,
      "epoch": 7.000908884344467,
      "grad_norm": 0.17458492517471313,
      "learning_rate": 5.957943669157382e-05,
      "loss": 1.1615,
      "mean_token_accuracy": 0.7763572549208616,
      "num_tokens": 1561022967.0,
      "step": 23110
    },
    {
      "entropy": 0.7415925338864326,
      "epoch": 7.003938498826025,
      "grad_norm": 0.20691928267478943,
      "learning_rate": 5.954665967309614e-05,
      "loss": 1.1336,
      "mean_token_accuracy": 0.783345764875412,
      "num_tokens": 1561709761.0,
      "step": 23120
    },
    {
      "entropy": 0.7527311459183693,
      "epoch": 7.006968113307582,
      "grad_norm": 0.2043900191783905,
      "learning_rate": 5.951387839633066e-05,
      "loss": 1.1407,
      "mean_token_accuracy": 0.7745445773005486,
      "num_tokens": 1562383663.0,
      "step": 23130
    },
    {
      "entropy": 0.7451196193695069,
      "epoch": 7.0099977277891385,
      "grad_norm": 0.2298509031534195,
      "learning_rate": 5.948109287589947e-05,
      "loss": 1.1344,
      "mean_token_accuracy": 0.7788399383425713,
      "num_tokens": 1563061168.0,
      "step": 23140
    },
    {
      "entropy": 0.7414814531803131,
      "epoch": 7.013027342270696,
      "grad_norm": 0.20013704895973206,
      "learning_rate": 5.94483031264266e-05,
      "loss": 1.1284,
      "mean_token_accuracy": 0.7796229213476181,
      "num_tokens": 1563742282.0,
      "step": 23150
    },
    {
      "entropy": 0.7403618216514587,
      "epoch": 7.016056956752253,
      "grad_norm": 0.21095553040504456,
      "learning_rate": 5.9415509162537855e-05,
      "loss": 1.1289,
      "mean_token_accuracy": 0.7779444932937623,
      "num_tokens": 1564417247.0,
      "step": 23160
    },
    {
      "entropy": 0.7454135611653327,
      "epoch": 7.01908657123381,
      "grad_norm": 0.22342520952224731,
      "learning_rate": 5.9382710998861005e-05,
      "loss": 1.139,
      "mean_token_accuracy": 0.7780433669686317,
      "num_tokens": 1565093792.0,
      "step": 23170
    },
    {
      "entropy": 0.7413820624351501,
      "epoch": 7.022116185715368,
      "grad_norm": 0.20998717844486237,
      "learning_rate": 5.934990865002569e-05,
      "loss": 1.1314,
      "mean_token_accuracy": 0.780006904900074,
      "num_tokens": 1565772423.0,
      "step": 23180
    },
    {
      "entropy": 0.7410301849246025,
      "epoch": 7.025145800196925,
      "grad_norm": 0.20335505902767181,
      "learning_rate": 5.931710213066339e-05,
      "loss": 1.1336,
      "mean_token_accuracy": 0.778225240111351,
      "num_tokens": 1566443073.0,
      "step": 23190
    },
    {
      "entropy": 0.73619065284729,
      "epoch": 7.028175414678482,
      "grad_norm": 0.2237924337387085,
      "learning_rate": 5.928429145540745e-05,
      "loss": 1.1259,
      "mean_token_accuracy": 0.7830476120114327,
      "num_tokens": 1567123051.0,
      "step": 23200
    },
    {
      "entropy": 0.7490300342440606,
      "epoch": 7.03120502916004,
      "grad_norm": 0.21476173400878906,
      "learning_rate": 5.9251476638893053e-05,
      "loss": 1.1298,
      "mean_token_accuracy": 0.7808086097240448,
      "num_tokens": 1567811235.0,
      "step": 23210
    },
    {
      "entropy": 0.7409378379583359,
      "epoch": 7.034234643641597,
      "grad_norm": 0.21300707757472992,
      "learning_rate": 5.921865769575729e-05,
      "loss": 1.1393,
      "mean_token_accuracy": 0.7780482247471809,
      "num_tokens": 1568488704.0,
      "step": 23220
    },
    {
      "entropy": 0.7460506349802017,
      "epoch": 7.0372642581231535,
      "grad_norm": 0.21356716752052307,
      "learning_rate": 5.918583464063903e-05,
      "loss": 1.1453,
      "mean_token_accuracy": 0.7756352841854095,
      "num_tokens": 1569155307.0,
      "step": 23230
    },
    {
      "entropy": 0.7379998996853828,
      "epoch": 7.040293872604711,
      "grad_norm": 0.200457364320755,
      "learning_rate": 5.915300748817899e-05,
      "loss": 1.1286,
      "mean_token_accuracy": 0.7798128291964531,
      "num_tokens": 1569831441.0,
      "step": 23240
    },
    {
      "entropy": 0.7388947442173958,
      "epoch": 7.043323487086268,
      "grad_norm": 0.21624140441417694,
      "learning_rate": 5.912017625301973e-05,
      "loss": 1.1412,
      "mean_token_accuracy": 0.7798016265034675,
      "num_tokens": 1570507044.0,
      "step": 23250
    },
    {
      "entropy": 0.7562513992190361,
      "epoch": 7.046353101567825,
      "grad_norm": 0.20515398681163788,
      "learning_rate": 5.908734094980564e-05,
      "loss": 1.1519,
      "mean_token_accuracy": 0.7743081346154213,
      "num_tokens": 1571175446.0,
      "step": 23260
    },
    {
      "entropy": 0.7395011693239212,
      "epoch": 7.049382716049383,
      "grad_norm": 0.21537697315216064,
      "learning_rate": 5.905450159318286e-05,
      "loss": 1.1381,
      "mean_token_accuracy": 0.7777394846081733,
      "num_tokens": 1571847808.0,
      "step": 23270
    },
    {
      "entropy": 0.7445904403924942,
      "epoch": 7.05241233053094,
      "grad_norm": 0.23160496354103088,
      "learning_rate": 5.9021658197799435e-05,
      "loss": 1.139,
      "mean_token_accuracy": 0.7750459596514702,
      "num_tokens": 1572508289.0,
      "step": 23280
    },
    {
      "entropy": 0.7437122985720634,
      "epoch": 7.055441945012497,
      "grad_norm": 0.2150079309940338,
      "learning_rate": 5.898881077830515e-05,
      "loss": 1.1339,
      "mean_token_accuracy": 0.7771567583084107,
      "num_tokens": 1573176400.0,
      "step": 23290
    },
    {
      "entropy": 0.742127226293087,
      "epoch": 7.058471559494055,
      "grad_norm": 0.21064236760139465,
      "learning_rate": 5.8955959349351574e-05,
      "loss": 1.1383,
      "mean_token_accuracy": 0.7798569083213807,
      "num_tokens": 1573854994.0,
      "step": 23300
    },
    {
      "entropy": 0.7435739010572433,
      "epoch": 7.0615011739756115,
      "grad_norm": 0.2325943112373352,
      "learning_rate": 5.8923103925592116e-05,
      "loss": 1.1404,
      "mean_token_accuracy": 0.7728467658162117,
      "num_tokens": 1574513069.0,
      "step": 23310
    },
    {
      "entropy": 0.7404571071267128,
      "epoch": 7.064530788457168,
      "grad_norm": 0.20979952812194824,
      "learning_rate": 5.889024452168193e-05,
      "loss": 1.1353,
      "mean_token_accuracy": 0.7756794095039368,
      "num_tokens": 1575178513.0,
      "step": 23320
    },
    {
      "entropy": 0.7502257347106933,
      "epoch": 7.067560402938726,
      "grad_norm": 0.23219376802444458,
      "learning_rate": 5.885738115227795e-05,
      "loss": 1.1325,
      "mean_token_accuracy": 0.7795778349041939,
      "num_tokens": 1575872216.0,
      "step": 23330
    },
    {
      "entropy": 0.7455775111913681,
      "epoch": 7.070590017420283,
      "grad_norm": 0.2104475498199463,
      "learning_rate": 5.8824513832038866e-05,
      "loss": 1.1447,
      "mean_token_accuracy": 0.7723892733454705,
      "num_tokens": 1576528943.0,
      "step": 23340
    },
    {
      "entropy": 0.7544861778616905,
      "epoch": 7.07361963190184,
      "grad_norm": 0.21834543347358704,
      "learning_rate": 5.879164257562517e-05,
      "loss": 1.1477,
      "mean_token_accuracy": 0.7774828374385834,
      "num_tokens": 1577202197.0,
      "step": 23350
    },
    {
      "entropy": 0.7377363726496696,
      "epoch": 7.076649246383398,
      "grad_norm": 0.21699722111225128,
      "learning_rate": 5.87587673976991e-05,
      "loss": 1.1412,
      "mean_token_accuracy": 0.7775470077991485,
      "num_tokens": 1577869101.0,
      "step": 23360
    },
    {
      "entropy": 0.7471830129623414,
      "epoch": 7.079678860864955,
      "grad_norm": 0.21366539597511292,
      "learning_rate": 5.8725888312924606e-05,
      "loss": 1.1414,
      "mean_token_accuracy": 0.7751694902777672,
      "num_tokens": 1578534353.0,
      "step": 23370
    },
    {
      "entropy": 0.7416027471423149,
      "epoch": 7.082708475346513,
      "grad_norm": 0.23158206045627594,
      "learning_rate": 5.86930053359674e-05,
      "loss": 1.133,
      "mean_token_accuracy": 0.7785997346043587,
      "num_tokens": 1579208524.0,
      "step": 23380
    },
    {
      "entropy": 0.7411579832434654,
      "epoch": 7.0857380898280695,
      "grad_norm": 0.20406202971935272,
      "learning_rate": 5.866011848149495e-05,
      "loss": 1.1386,
      "mean_token_accuracy": 0.780168367922306,
      "num_tokens": 1579885128.0,
      "step": 23390
    },
    {
      "entropy": 0.7510242238640785,
      "epoch": 7.088767704309626,
      "grad_norm": 0.25734636187553406,
      "learning_rate": 5.862722776417644e-05,
      "loss": 1.1439,
      "mean_token_accuracy": 0.771857887506485,
      "num_tokens": 1580550798.0,
      "step": 23400
    },
    {
      "entropy": 0.7522947192192078,
      "epoch": 7.091797318791184,
      "grad_norm": 0.20757640898227692,
      "learning_rate": 5.859433319868276e-05,
      "loss": 1.1448,
      "mean_token_accuracy": 0.7728193268179894,
      "num_tokens": 1581218703.0,
      "step": 23410
    },
    {
      "entropy": 0.7494997382164001,
      "epoch": 7.094826933272741,
      "grad_norm": 0.22695372998714447,
      "learning_rate": 5.8561434799686554e-05,
      "loss": 1.1475,
      "mean_token_accuracy": 0.7778515160083771,
      "num_tokens": 1581903891.0,
      "step": 23420
    },
    {
      "entropy": 0.739686281979084,
      "epoch": 7.097856547754298,
      "grad_norm": 0.21500404179096222,
      "learning_rate": 5.852853258186213e-05,
      "loss": 1.132,
      "mean_token_accuracy": 0.7795198231935501,
      "num_tokens": 1582582728.0,
      "step": 23430
    },
    {
      "entropy": 0.7484030231833458,
      "epoch": 7.100886162235856,
      "grad_norm": 0.21093358099460602,
      "learning_rate": 5.849562655988553e-05,
      "loss": 1.1403,
      "mean_token_accuracy": 0.7773430526256562,
      "num_tokens": 1583265598.0,
      "step": 23440
    },
    {
      "entropy": 0.7469753220677375,
      "epoch": 7.103915776717413,
      "grad_norm": 0.22998684644699097,
      "learning_rate": 5.8462716748434486e-05,
      "loss": 1.1405,
      "mean_token_accuracy": 0.7785072311758995,
      "num_tokens": 1583937787.0,
      "step": 23450
    },
    {
      "entropy": 0.7508120566606522,
      "epoch": 7.10694539119897,
      "grad_norm": 0.20284292101860046,
      "learning_rate": 5.842980316218845e-05,
      "loss": 1.1398,
      "mean_token_accuracy": 0.7788093894720077,
      "num_tokens": 1584629630.0,
      "step": 23460
    },
    {
      "entropy": 0.7422115087509156,
      "epoch": 7.1099750056805275,
      "grad_norm": 0.21174338459968567,
      "learning_rate": 5.839688581582847e-05,
      "loss": 1.1455,
      "mean_token_accuracy": 0.7801809579133987,
      "num_tokens": 1585313405.0,
      "step": 23470
    },
    {
      "entropy": 0.7515713751316071,
      "epoch": 7.1130046201620845,
      "grad_norm": 0.21566711366176605,
      "learning_rate": 5.836396472403737e-05,
      "loss": 1.1402,
      "mean_token_accuracy": 0.7787953034043312,
      "num_tokens": 1586005561.0,
      "step": 23480
    },
    {
      "entropy": 0.7425371959805489,
      "epoch": 7.116034234643641,
      "grad_norm": 0.2266668826341629,
      "learning_rate": 5.8331039901499606e-05,
      "loss": 1.1306,
      "mean_token_accuracy": 0.7758061677217484,
      "num_tokens": 1586664727.0,
      "step": 23490
    },
    {
      "entropy": 0.7493927583098412,
      "epoch": 7.119063849125199,
      "grad_norm": 0.2102845460176468,
      "learning_rate": 5.829811136290128e-05,
      "loss": 1.1434,
      "mean_token_accuracy": 0.777026678621769,
      "num_tokens": 1587340222.0,
      "step": 23500
    },
    {
      "entropy": 0.7541937798261642,
      "epoch": 7.122093463606756,
      "grad_norm": 0.20236948132514954,
      "learning_rate": 5.826517912293016e-05,
      "loss": 1.1355,
      "mean_token_accuracy": 0.7758743047714234,
      "num_tokens": 1588020769.0,
      "step": 23510
    },
    {
      "entropy": 0.7430880099534989,
      "epoch": 7.125123078088313,
      "grad_norm": 0.20082427561283112,
      "learning_rate": 5.823224319627568e-05,
      "loss": 1.1441,
      "mean_token_accuracy": 0.7754659965634346,
      "num_tokens": 1588677061.0,
      "step": 23520
    },
    {
      "entropy": 0.7427562147378921,
      "epoch": 7.128152692569871,
      "grad_norm": 0.2160859853029251,
      "learning_rate": 5.8199303597628926e-05,
      "loss": 1.1391,
      "mean_token_accuracy": 0.7824327185750007,
      "num_tokens": 1589364715.0,
      "step": 23530
    },
    {
      "entropy": 0.7464813143014908,
      "epoch": 7.131182307051428,
      "grad_norm": 0.21505455672740936,
      "learning_rate": 5.8166360341682605e-05,
      "loss": 1.1356,
      "mean_token_accuracy": 0.7731814235448837,
      "num_tokens": 1590038292.0,
      "step": 23540
    },
    {
      "entropy": 0.7509728401899338,
      "epoch": 7.134211921532985,
      "grad_norm": 0.21698719263076782,
      "learning_rate": 5.8133413443131036e-05,
      "loss": 1.1448,
      "mean_token_accuracy": 0.7785868659615517,
      "num_tokens": 1590728880.0,
      "step": 23550
    },
    {
      "entropy": 0.7430469363927841,
      "epoch": 7.1372415360145425,
      "grad_norm": 0.22430694103240967,
      "learning_rate": 5.8100462916670196e-05,
      "loss": 1.1321,
      "mean_token_accuracy": 0.7757880195975304,
      "num_tokens": 1591393754.0,
      "step": 23560
    },
    {
      "entropy": 0.7454712346196175,
      "epoch": 7.140271150496099,
      "grad_norm": 0.20653754472732544,
      "learning_rate": 5.806750877699767e-05,
      "loss": 1.1348,
      "mean_token_accuracy": 0.7810803964734078,
      "num_tokens": 1592073212.0,
      "step": 23570
    },
    {
      "entropy": 0.7382087767124176,
      "epoch": 7.143300764977656,
      "grad_norm": 0.2208268642425537,
      "learning_rate": 5.8034551038812657e-05,
      "loss": 1.1319,
      "mean_token_accuracy": 0.782666739821434,
      "num_tokens": 1592767591.0,
      "step": 23580
    },
    {
      "entropy": 0.7456942811608315,
      "epoch": 7.146330379459214,
      "grad_norm": 0.21309810876846313,
      "learning_rate": 5.800158971681595e-05,
      "loss": 1.1378,
      "mean_token_accuracy": 0.7783115655183792,
      "num_tokens": 1593459153.0,
      "step": 23590
    },
    {
      "entropy": 0.7472904816269874,
      "epoch": 7.149359993940771,
      "grad_norm": 0.21896205842494965,
      "learning_rate": 5.796862482570994e-05,
      "loss": 1.1422,
      "mean_token_accuracy": 0.7750360250473023,
      "num_tokens": 1594132093.0,
      "step": 23600
    },
    {
      "entropy": 0.7441201984882355,
      "epoch": 7.152389608422328,
      "grad_norm": 0.2231408655643463,
      "learning_rate": 5.7935656380198634e-05,
      "loss": 1.1366,
      "mean_token_accuracy": 0.7772513866424561,
      "num_tokens": 1594815093.0,
      "step": 23610
    },
    {
      "entropy": 0.7458626598119735,
      "epoch": 7.155419222903886,
      "grad_norm": 0.2119058519601822,
      "learning_rate": 5.790268439498759e-05,
      "loss": 1.149,
      "mean_token_accuracy": 0.7717437729239464,
      "num_tokens": 1595480749.0,
      "step": 23620
    },
    {
      "entropy": 0.7433493629097938,
      "epoch": 7.158448837385443,
      "grad_norm": 0.21936173737049103,
      "learning_rate": 5.7869708884783994e-05,
      "loss": 1.1385,
      "mean_token_accuracy": 0.7798545464873314,
      "num_tokens": 1596162626.0,
      "step": 23630
    },
    {
      "entropy": 0.7416732639074326,
      "epoch": 7.161478451867,
      "grad_norm": 0.21237511932849884,
      "learning_rate": 5.7836729864296537e-05,
      "loss": 1.1306,
      "mean_token_accuracy": 0.7766799658536911,
      "num_tokens": 1596833948.0,
      "step": 23640
    },
    {
      "entropy": 0.7414098903536797,
      "epoch": 7.164508066348557,
      "grad_norm": 0.21430625021457672,
      "learning_rate": 5.780374734823552e-05,
      "loss": 1.14,
      "mean_token_accuracy": 0.776218494772911,
      "num_tokens": 1597498836.0,
      "step": 23650
    },
    {
      "entropy": 0.7381272122263909,
      "epoch": 7.167537680830114,
      "grad_norm": 0.1998218148946762,
      "learning_rate": 5.777076135131281e-05,
      "loss": 1.1324,
      "mean_token_accuracy": 0.7800790205597877,
      "num_tokens": 1598178573.0,
      "step": 23660
    },
    {
      "entropy": 0.7380612477660179,
      "epoch": 7.170567295311671,
      "grad_norm": 0.22392015159130096,
      "learning_rate": 5.773777188824181e-05,
      "loss": 1.1368,
      "mean_token_accuracy": 0.7783250853419303,
      "num_tokens": 1598846556.0,
      "step": 23670
    },
    {
      "entropy": 0.7345643967390061,
      "epoch": 7.173596909793229,
      "grad_norm": 0.20985360443592072,
      "learning_rate": 5.770477897373745e-05,
      "loss": 1.1308,
      "mean_token_accuracy": 0.7828897207975387,
      "num_tokens": 1599542618.0,
      "step": 23680
    },
    {
      "entropy": 0.7467793002724648,
      "epoch": 7.176626524274786,
      "grad_norm": 0.2099127322435379,
      "learning_rate": 5.7671782622516236e-05,
      "loss": 1.142,
      "mean_token_accuracy": 0.7768888935446739,
      "num_tokens": 1600217753.0,
      "step": 23690
    },
    {
      "entropy": 0.7502148687839508,
      "epoch": 7.179656138756343,
      "grad_norm": 0.21838758885860443,
      "learning_rate": 5.7638782849296204e-05,
      "loss": 1.1347,
      "mean_token_accuracy": 0.7802770152688027,
      "num_tokens": 1600911236.0,
      "step": 23700
    },
    {
      "entropy": 0.7430243715643883,
      "epoch": 7.182685753237901,
      "grad_norm": 0.21046076714992523,
      "learning_rate": 5.760577966879688e-05,
      "loss": 1.1327,
      "mean_token_accuracy": 0.7815440043807029,
      "num_tokens": 1601595935.0,
      "step": 23710
    },
    {
      "entropy": 0.7310260012745857,
      "epoch": 7.185715367719458,
      "grad_norm": 0.2190067321062088,
      "learning_rate": 5.757277309573934e-05,
      "loss": 1.1319,
      "mean_token_accuracy": 0.7807996392250061,
      "num_tokens": 1602264612.0,
      "step": 23720
    },
    {
      "entropy": 0.7429593563079834,
      "epoch": 7.1887449822010145,
      "grad_norm": 0.21921633183956146,
      "learning_rate": 5.7539763144846184e-05,
      "loss": 1.1383,
      "mean_token_accuracy": 0.7799703419208527,
      "num_tokens": 1602944973.0,
      "step": 23730
    },
    {
      "entropy": 0.7431642904877662,
      "epoch": 7.191774596682572,
      "grad_norm": 0.2143329381942749,
      "learning_rate": 5.7506749830841476e-05,
      "loss": 1.1395,
      "mean_token_accuracy": 0.7765309974551201,
      "num_tokens": 1603607959.0,
      "step": 23740
    },
    {
      "entropy": 0.7459607213735581,
      "epoch": 7.194804211164129,
      "grad_norm": 0.22598759829998016,
      "learning_rate": 5.747373316845081e-05,
      "loss": 1.1363,
      "mean_token_accuracy": 0.7783181667327881,
      "num_tokens": 1604291864.0,
      "step": 23750
    },
    {
      "entropy": 0.7523634403944015,
      "epoch": 7.197833825645686,
      "grad_norm": 0.22408194839954376,
      "learning_rate": 5.7440713172401294e-05,
      "loss": 1.1523,
      "mean_token_accuracy": 0.7770254418253899,
      "num_tokens": 1604973582.0,
      "step": 23760
    },
    {
      "entropy": 0.7411210224032402,
      "epoch": 7.200863440127244,
      "grad_norm": 0.2086329311132431,
      "learning_rate": 5.7407689857421473e-05,
      "loss": 1.1343,
      "mean_token_accuracy": 0.7793959960341453,
      "num_tokens": 1605646230.0,
      "step": 23770
    },
    {
      "entropy": 0.7558016300201416,
      "epoch": 7.203893054608801,
      "grad_norm": 0.21734054386615753,
      "learning_rate": 5.737466323824141e-05,
      "loss": 1.1492,
      "mean_token_accuracy": 0.7727850556373597,
      "num_tokens": 1606312586.0,
      "step": 23780
    },
    {
      "entropy": 0.7510229885578156,
      "epoch": 7.206922669090358,
      "grad_norm": 0.21224647760391235,
      "learning_rate": 5.734163332959261e-05,
      "loss": 1.1469,
      "mean_token_accuracy": 0.7727331802248955,
      "num_tokens": 1606971412.0,
      "step": 23790
    },
    {
      "entropy": 0.733087569475174,
      "epoch": 7.209952283571916,
      "grad_norm": 0.22325651347637177,
      "learning_rate": 5.73086001462081e-05,
      "loss": 1.1305,
      "mean_token_accuracy": 0.7817760735750199,
      "num_tokens": 1607644861.0,
      "step": 23800
    },
    {
      "entropy": 0.7554741337895393,
      "epoch": 7.212981898053473,
      "grad_norm": 0.20936958491802216,
      "learning_rate": 5.727556370282229e-05,
      "loss": 1.1434,
      "mean_token_accuracy": 0.7714540019631386,
      "num_tokens": 1608311747.0,
      "step": 23810
    },
    {
      "entropy": 0.7569979503750801,
      "epoch": 7.2160115125350295,
      "grad_norm": 0.20916040241718292,
      "learning_rate": 5.724252401417112e-05,
      "loss": 1.1484,
      "mean_token_accuracy": 0.774019081890583,
      "num_tokens": 1608988099.0,
      "step": 23820
    },
    {
      "entropy": 0.7400409236550332,
      "epoch": 7.219041127016587,
      "grad_norm": 0.21031858026981354,
      "learning_rate": 5.7209481094991916e-05,
      "loss": 1.1435,
      "mean_token_accuracy": 0.7749984413385391,
      "num_tokens": 1609651043.0,
      "step": 23830
    },
    {
      "entropy": 0.7454067185521126,
      "epoch": 7.222070741498144,
      "grad_norm": 0.21775247156620026,
      "learning_rate": 5.717643496002351e-05,
      "loss": 1.1393,
      "mean_token_accuracy": 0.7788102000951767,
      "num_tokens": 1610331166.0,
      "step": 23840
    },
    {
      "entropy": 0.7444580152630806,
      "epoch": 7.225100355979702,
      "grad_norm": 0.21025405824184418,
      "learning_rate": 5.714338562400609e-05,
      "loss": 1.1372,
      "mean_token_accuracy": 0.7797574743628501,
      "num_tokens": 1611013078.0,
      "step": 23850
    },
    {
      "entropy": 0.7428898483514785,
      "epoch": 7.228129970461259,
      "grad_norm": 0.2147832065820694,
      "learning_rate": 5.7110333101681336e-05,
      "loss": 1.1413,
      "mean_token_accuracy": 0.7777466505765915,
      "num_tokens": 1611688453.0,
      "step": 23860
    },
    {
      "entropy": 0.7444768413901329,
      "epoch": 7.231159584942816,
      "grad_norm": 0.22270500659942627,
      "learning_rate": 5.707727740779234e-05,
      "loss": 1.1405,
      "mean_token_accuracy": 0.7768704071640968,
      "num_tokens": 1612364201.0,
      "step": 23870
    },
    {
      "entropy": 0.7423681631684304,
      "epoch": 7.234189199424374,
      "grad_norm": 0.21367275714874268,
      "learning_rate": 5.704421855708355e-05,
      "loss": 1.1326,
      "mean_token_accuracy": 0.7826312854886055,
      "num_tokens": 1613059538.0,
      "step": 23880
    },
    {
      "entropy": 0.7551495105028152,
      "epoch": 7.237218813905931,
      "grad_norm": 0.20957709848880768,
      "learning_rate": 5.701115656430092e-05,
      "loss": 1.1541,
      "mean_token_accuracy": 0.769644808769226,
      "num_tokens": 1613718151.0,
      "step": 23890
    },
    {
      "entropy": 0.7407628953456878,
      "epoch": 7.2402484283874875,
      "grad_norm": 0.2054181545972824,
      "learning_rate": 5.697809144419172e-05,
      "loss": 1.1366,
      "mean_token_accuracy": 0.7815528899431229,
      "num_tokens": 1614402887.0,
      "step": 23900
    },
    {
      "entropy": 0.7427027866244316,
      "epoch": 7.243278042869045,
      "grad_norm": 0.21740637719631195,
      "learning_rate": 5.6945023211504656e-05,
      "loss": 1.1404,
      "mean_token_accuracy": 0.7759135112166404,
      "num_tokens": 1615066380.0,
      "step": 23910
    },
    {
      "entropy": 0.7589648768305779,
      "epoch": 7.246307657350602,
      "grad_norm": 0.22193023562431335,
      "learning_rate": 5.69119518809898e-05,
      "loss": 1.154,
      "mean_token_accuracy": 0.7683237046003342,
      "num_tokens": 1615723150.0,
      "step": 23920
    },
    {
      "entropy": 0.7384623020887375,
      "epoch": 7.249337271832159,
      "grad_norm": 0.21717776358127594,
      "learning_rate": 5.687887746739866e-05,
      "loss": 1.1329,
      "mean_token_accuracy": 0.7788059458136558,
      "num_tokens": 1616397325.0,
      "step": 23930
    },
    {
      "entropy": 0.7552555456757546,
      "epoch": 7.252366886313717,
      "grad_norm": 0.2296213060617447,
      "learning_rate": 5.684579998548403e-05,
      "loss": 1.1396,
      "mean_token_accuracy": 0.772842176258564,
      "num_tokens": 1617072088.0,
      "step": 23940
    },
    {
      "entropy": 0.7496678099036217,
      "epoch": 7.255396500795274,
      "grad_norm": 0.2228216975927353,
      "learning_rate": 5.681271945000015e-05,
      "loss": 1.1468,
      "mean_token_accuracy": 0.7734100058674812,
      "num_tokens": 1617732992.0,
      "step": 23950
    },
    {
      "entropy": 0.7452170476317406,
      "epoch": 7.258426115276831,
      "grad_norm": 0.22176329791545868,
      "learning_rate": 5.67796358757026e-05,
      "loss": 1.1453,
      "mean_token_accuracy": 0.7765678241848946,
      "num_tokens": 1618406533.0,
      "step": 23960
    },
    {
      "entropy": 0.7457974344491959,
      "epoch": 7.261455729758389,
      "grad_norm": 0.21788766980171204,
      "learning_rate": 5.674654927734829e-05,
      "loss": 1.1403,
      "mean_token_accuracy": 0.7738040804862976,
      "num_tokens": 1619072744.0,
      "step": 23970
    },
    {
      "entropy": 0.7414055898785591,
      "epoch": 7.2644853442399455,
      "grad_norm": 0.2042253017425537,
      "learning_rate": 5.67134596696955e-05,
      "loss": 1.1428,
      "mean_token_accuracy": 0.7790959507226944,
      "num_tokens": 1619753563.0,
      "step": 23980
    },
    {
      "entropy": 0.7385856539011002,
      "epoch": 7.267514958721502,
      "grad_norm": 0.22292554378509521,
      "learning_rate": 5.668036706750386e-05,
      "loss": 1.1343,
      "mean_token_accuracy": 0.7773075819015502,
      "num_tokens": 1620419569.0,
      "step": 23990
    },
    {
      "entropy": 0.7456323817372322,
      "epoch": 7.27054457320306,
      "grad_norm": 0.2110702246427536,
      "learning_rate": 5.664727148553431e-05,
      "loss": 1.1402,
      "mean_token_accuracy": 0.7801220253109932,
      "num_tokens": 1621111439.0,
      "step": 24000
    },
    {
      "entropy": 0.7472523480653763,
      "epoch": 7.273574187684617,
      "grad_norm": 0.2149803340435028,
      "learning_rate": 5.661417293854917e-05,
      "loss": 1.1441,
      "mean_token_accuracy": 0.7775317147374153,
      "num_tokens": 1621787136.0,
      "step": 24010
    },
    {
      "entropy": 0.7425885871052742,
      "epoch": 7.276603802166174,
      "grad_norm": 0.21277956664562225,
      "learning_rate": 5.6581071441312006e-05,
      "loss": 1.1353,
      "mean_token_accuracy": 0.7793334111571312,
      "num_tokens": 1622470050.0,
      "step": 24020
    },
    {
      "entropy": 0.7456920474767685,
      "epoch": 7.279633416647732,
      "grad_norm": 0.2155601978302002,
      "learning_rate": 5.654796700858775e-05,
      "loss": 1.1425,
      "mean_token_accuracy": 0.7784823909401893,
      "num_tokens": 1623148353.0,
      "step": 24030
    },
    {
      "entropy": 0.7436433017253876,
      "epoch": 7.282663031129289,
      "grad_norm": 0.21758146584033966,
      "learning_rate": 5.651485965514267e-05,
      "loss": 1.1361,
      "mean_token_accuracy": 0.7788098677992821,
      "num_tokens": 1623814664.0,
      "step": 24040
    },
    {
      "entropy": 0.7383101120591163,
      "epoch": 7.285692645610846,
      "grad_norm": 0.22921079397201538,
      "learning_rate": 5.648174939574427e-05,
      "loss": 1.1392,
      "mean_token_accuracy": 0.7783874228596688,
      "num_tokens": 1624492342.0,
      "step": 24050
    },
    {
      "entropy": 0.7421487629413605,
      "epoch": 7.2887222600924035,
      "grad_norm": 0.21617472171783447,
      "learning_rate": 5.644863624516138e-05,
      "loss": 1.1454,
      "mean_token_accuracy": 0.7777203261852265,
      "num_tokens": 1625167670.0,
      "step": 24060
    },
    {
      "entropy": 0.7490338414907456,
      "epoch": 7.2917518745739605,
      "grad_norm": 0.20987454056739807,
      "learning_rate": 5.641552021816414e-05,
      "loss": 1.1464,
      "mean_token_accuracy": 0.777875579893589,
      "num_tokens": 1625851661.0,
      "step": 24070
    },
    {
      "entropy": 0.7433881223201751,
      "epoch": 7.294781489055517,
      "grad_norm": 0.21918809413909912,
      "learning_rate": 5.638240132952394e-05,
      "loss": 1.1417,
      "mean_token_accuracy": 0.7791951164603234,
      "num_tokens": 1626523825.0,
      "step": 24080
    },
    {
      "entropy": 0.7431725174188614,
      "epoch": 7.297811103537075,
      "grad_norm": 0.20071259140968323,
      "learning_rate": 5.634927959401346e-05,
      "loss": 1.1412,
      "mean_token_accuracy": 0.7779849901795387,
      "num_tokens": 1627191800.0,
      "step": 24090
    },
    {
      "entropy": 0.745476596057415,
      "epoch": 7.300840718018632,
      "grad_norm": 0.20968084037303925,
      "learning_rate": 5.6316155026406684e-05,
      "loss": 1.1435,
      "mean_token_accuracy": 0.7744548842310905,
      "num_tokens": 1627861146.0,
      "step": 24100
    },
    {
      "entropy": 0.7318308338522911,
      "epoch": 7.303870332500189,
      "grad_norm": 0.20997008681297302,
      "learning_rate": 5.628302764147877e-05,
      "loss": 1.1299,
      "mean_token_accuracy": 0.7816808238625527,
      "num_tokens": 1628548768.0,
      "step": 24110
    },
    {
      "entropy": 0.7426049247384071,
      "epoch": 7.306899946981747,
      "grad_norm": 0.2112835943698883,
      "learning_rate": 5.624989745400623e-05,
      "loss": 1.1378,
      "mean_token_accuracy": 0.779009498655796,
      "num_tokens": 1629220912.0,
      "step": 24120
    },
    {
      "entropy": 0.7388949811458587,
      "epoch": 7.309929561463304,
      "grad_norm": 0.21682319045066833,
      "learning_rate": 5.621676447876677e-05,
      "loss": 1.1365,
      "mean_token_accuracy": 0.7800767734646797,
      "num_tokens": 1629893582.0,
      "step": 24130
    },
    {
      "entropy": 0.7498345792293548,
      "epoch": 7.312959175944861,
      "grad_norm": 0.2141590118408203,
      "learning_rate": 5.618362873053937e-05,
      "loss": 1.1514,
      "mean_token_accuracy": 0.7738681003451348,
      "num_tokens": 1630561143.0,
      "step": 24140
    },
    {
      "entropy": 0.7501821503043175,
      "epoch": 7.3159887904264185,
      "grad_norm": 0.2280568927526474,
      "learning_rate": 5.61504902241042e-05,
      "loss": 1.1415,
      "mean_token_accuracy": 0.7810229748487473,
      "num_tokens": 1631233693.0,
      "step": 24150
    },
    {
      "entropy": 0.7449058368802071,
      "epoch": 7.319018404907975,
      "grad_norm": 0.22089658677577972,
      "learning_rate": 5.6117348974242724e-05,
      "loss": 1.1342,
      "mean_token_accuracy": 0.7829345911741257,
      "num_tokens": 1631927017.0,
      "step": 24160
    },
    {
      "entropy": 0.735488374531269,
      "epoch": 7.322048019389532,
      "grad_norm": 0.21309734880924225,
      "learning_rate": 5.6084204995737576e-05,
      "loss": 1.1332,
      "mean_token_accuracy": 0.7801588609814644,
      "num_tokens": 1632604646.0,
      "step": 24170
    },
    {
      "entropy": 0.745568074285984,
      "epoch": 7.32507763387109,
      "grad_norm": 0.21478989720344543,
      "learning_rate": 5.605105830337265e-05,
      "loss": 1.145,
      "mean_token_accuracy": 0.7778879448771476,
      "num_tokens": 1633286932.0,
      "step": 24180
    },
    {
      "entropy": 0.7399929359555244,
      "epoch": 7.328107248352647,
      "grad_norm": 0.20924386382102966,
      "learning_rate": 5.6017908911933004e-05,
      "loss": 1.1418,
      "mean_token_accuracy": 0.7734668895602226,
      "num_tokens": 1633944821.0,
      "step": 24190
    },
    {
      "entropy": 0.7454836443066597,
      "epoch": 7.331136862834204,
      "grad_norm": 0.21767070889472961,
      "learning_rate": 5.598475683620493e-05,
      "loss": 1.14,
      "mean_token_accuracy": 0.7767185181379318,
      "num_tokens": 1634613657.0,
      "step": 24200
    },
    {
      "entropy": 0.745055228471756,
      "epoch": 7.334166477315762,
      "grad_norm": 0.2201317697763443,
      "learning_rate": 5.595160209097593e-05,
      "loss": 1.1354,
      "mean_token_accuracy": 0.7779637157917023,
      "num_tokens": 1635293860.0,
      "step": 24210
    },
    {
      "entropy": 0.7456524521112442,
      "epoch": 7.337196091797319,
      "grad_norm": 0.202079176902771,
      "learning_rate": 5.591844469103467e-05,
      "loss": 1.1395,
      "mean_token_accuracy": 0.7798618689179421,
      "num_tokens": 1635992432.0,
      "step": 24220
    },
    {
      "entropy": 0.7374287351965905,
      "epoch": 7.340225706278876,
      "grad_norm": 0.22293561697006226,
      "learning_rate": 5.5885284651170986e-05,
      "loss": 1.139,
      "mean_token_accuracy": 0.7847616389393807,
      "num_tokens": 1636675788.0,
      "step": 24230
    },
    {
      "entropy": 0.7431089982390404,
      "epoch": 7.343255320760433,
      "grad_norm": 0.20412328839302063,
      "learning_rate": 5.585212198617596e-05,
      "loss": 1.1362,
      "mean_token_accuracy": 0.7805899128317833,
      "num_tokens": 1637367574.0,
      "step": 24240
    },
    {
      "entropy": 0.7460298150777817,
      "epoch": 7.34628493524199,
      "grad_norm": 0.21624484658241272,
      "learning_rate": 5.581895671084176e-05,
      "loss": 1.1359,
      "mean_token_accuracy": 0.7783966347575187,
      "num_tokens": 1638050741.0,
      "step": 24250
    },
    {
      "entropy": 0.7557017922401428,
      "epoch": 7.349314549723548,
      "grad_norm": 0.22091956436634064,
      "learning_rate": 5.5785788839961785e-05,
      "loss": 1.1445,
      "mean_token_accuracy": 0.7714565888047218,
      "num_tokens": 1638717200.0,
      "step": 24260
    },
    {
      "entropy": 0.7368078112602234,
      "epoch": 7.352344164205105,
      "grad_norm": 0.2095811516046524,
      "learning_rate": 5.575261838833054e-05,
      "loss": 1.1378,
      "mean_token_accuracy": 0.7802141949534416,
      "num_tokens": 1639401631.0,
      "step": 24270
    },
    {
      "entropy": 0.7564533114433288,
      "epoch": 7.355373778686662,
      "grad_norm": 0.2242221087217331,
      "learning_rate": 5.571944537074373e-05,
      "loss": 1.1382,
      "mean_token_accuracy": 0.7788906678557396,
      "num_tokens": 1640106395.0,
      "step": 24280
    },
    {
      "entropy": 0.737377966940403,
      "epoch": 7.358403393168219,
      "grad_norm": 0.20979861915111542,
      "learning_rate": 5.568626980199816e-05,
      "loss": 1.1301,
      "mean_token_accuracy": 0.7788542896509171,
      "num_tokens": 1640788308.0,
      "step": 24290
    },
    {
      "entropy": 0.7406774923205376,
      "epoch": 7.361433007649777,
      "grad_norm": 0.2215997725725174,
      "learning_rate": 5.565309169689179e-05,
      "loss": 1.1349,
      "mean_token_accuracy": 0.7774257019162178,
      "num_tokens": 1641460222.0,
      "step": 24300
    },
    {
      "entropy": 0.7524708792567253,
      "epoch": 7.364462622131334,
      "grad_norm": 0.21218158304691315,
      "learning_rate": 5.5619911070223764e-05,
      "loss": 1.144,
      "mean_token_accuracy": 0.7777711153030396,
      "num_tokens": 1642143261.0,
      "step": 24310
    },
    {
      "entropy": 0.7503455430269241,
      "epoch": 7.367492236612891,
      "grad_norm": 0.2106654793024063,
      "learning_rate": 5.5586727936794236e-05,
      "loss": 1.1411,
      "mean_token_accuracy": 0.776572035253048,
      "num_tokens": 1642817220.0,
      "step": 24320
    },
    {
      "entropy": 0.7529546201229096,
      "epoch": 7.370521851094448,
      "grad_norm": 0.20978443324565887,
      "learning_rate": 5.555354231140457e-05,
      "loss": 1.139,
      "mean_token_accuracy": 0.7750160679221153,
      "num_tokens": 1643490100.0,
      "step": 24330
    },
    {
      "entropy": 0.7439987063407898,
      "epoch": 7.373551465576005,
      "grad_norm": 0.21684180200099945,
      "learning_rate": 5.5520354208857215e-05,
      "loss": 1.1334,
      "mean_token_accuracy": 0.7768491759896279,
      "num_tokens": 1644171664.0,
      "step": 24340
    },
    {
      "entropy": 0.7514504194259644,
      "epoch": 7.376581080057563,
      "grad_norm": 0.2160317301750183,
      "learning_rate": 5.548716364395574e-05,
      "loss": 1.1439,
      "mean_token_accuracy": 0.7775610223412514,
      "num_tokens": 1644847717.0,
      "step": 24350
    },
    {
      "entropy": 0.7286875173449516,
      "epoch": 7.37961069453912,
      "grad_norm": 0.22281469404697418,
      "learning_rate": 5.545397063150477e-05,
      "loss": 1.1289,
      "mean_token_accuracy": 0.7795726001262665,
      "num_tokens": 1645514078.0,
      "step": 24360
    },
    {
      "entropy": 0.7473196431994438,
      "epoch": 7.382640309020677,
      "grad_norm": 0.22722819447517395,
      "learning_rate": 5.542077518631007e-05,
      "loss": 1.1409,
      "mean_token_accuracy": 0.7789425924420357,
      "num_tokens": 1646191594.0,
      "step": 24370
    },
    {
      "entropy": 0.744561393558979,
      "epoch": 7.385669923502235,
      "grad_norm": 0.2064792662858963,
      "learning_rate": 5.5387577323178464e-05,
      "loss": 1.1439,
      "mean_token_accuracy": 0.7743083328008652,
      "num_tokens": 1646855313.0,
      "step": 24380
    },
    {
      "entropy": 0.7324327319860459,
      "epoch": 7.388699537983792,
      "grad_norm": 0.21427488327026367,
      "learning_rate": 5.535437705691785e-05,
      "loss": 1.1281,
      "mean_token_accuracy": 0.7798612684011459,
      "num_tokens": 1647528294.0,
      "step": 24390
    },
    {
      "entropy": 0.7458065509796142,
      "epoch": 7.391729152465349,
      "grad_norm": 0.21999485790729523,
      "learning_rate": 5.532117440233722e-05,
      "loss": 1.1442,
      "mean_token_accuracy": 0.7761651277542114,
      "num_tokens": 1648194479.0,
      "step": 24400
    },
    {
      "entropy": 0.7495456725358963,
      "epoch": 7.394758766946906,
      "grad_norm": 0.2247866690158844,
      "learning_rate": 5.528796937424662e-05,
      "loss": 1.1481,
      "mean_token_accuracy": 0.7760565713047981,
      "num_tokens": 1648869091.0,
      "step": 24410
    },
    {
      "entropy": 0.7413565754890442,
      "epoch": 7.397788381428463,
      "grad_norm": 0.20716004073619843,
      "learning_rate": 5.525476198745714e-05,
      "loss": 1.1358,
      "mean_token_accuracy": 0.7809042319655418,
      "num_tokens": 1649549800.0,
      "step": 24420
    },
    {
      "entropy": 0.7377701357007027,
      "epoch": 7.40081799591002,
      "grad_norm": 0.22581028938293457,
      "learning_rate": 5.522155225678093e-05,
      "loss": 1.1299,
      "mean_token_accuracy": 0.7825725868344307,
      "num_tokens": 1650228436.0,
      "step": 24430
    },
    {
      "entropy": 0.7420076668262482,
      "epoch": 7.403847610391578,
      "grad_norm": 0.2277398556470871,
      "learning_rate": 5.518834019703122e-05,
      "loss": 1.1381,
      "mean_token_accuracy": 0.7727554067969322,
      "num_tokens": 1650880381.0,
      "step": 24440
    },
    {
      "entropy": 0.7543826818466186,
      "epoch": 7.406877224873135,
      "grad_norm": 0.21878106892108917,
      "learning_rate": 5.515512582302224e-05,
      "loss": 1.1432,
      "mean_token_accuracy": 0.7753280624747276,
      "num_tokens": 1651559161.0,
      "step": 24450
    },
    {
      "entropy": 0.7435263082385063,
      "epoch": 7.409906839354692,
      "grad_norm": 0.22242344915866852,
      "learning_rate": 5.5121909149569237e-05,
      "loss": 1.1346,
      "mean_token_accuracy": 0.7759518295526504,
      "num_tokens": 1652237914.0,
      "step": 24460
    },
    {
      "entropy": 0.7520292147994041,
      "epoch": 7.41293645383625,
      "grad_norm": 0.2094714492559433,
      "learning_rate": 5.508869019148855e-05,
      "loss": 1.138,
      "mean_token_accuracy": 0.7758851468563079,
      "num_tokens": 1652908954.0,
      "step": 24470
    },
    {
      "entropy": 0.7410292357206345,
      "epoch": 7.415966068317807,
      "grad_norm": 0.2083895057439804,
      "learning_rate": 5.505546896359747e-05,
      "loss": 1.1304,
      "mean_token_accuracy": 0.786678695678711,
      "num_tokens": 1653619377.0,
      "step": 24480
    },
    {
      "entropy": 0.7511360585689545,
      "epoch": 7.4189956827993635,
      "grad_norm": 0.21662305295467377,
      "learning_rate": 5.502224548071433e-05,
      "loss": 1.1489,
      "mean_token_accuracy": 0.7749097287654877,
      "num_tokens": 1654289661.0,
      "step": 24490
    },
    {
      "entropy": 0.7482529670000077,
      "epoch": 7.422025297280921,
      "grad_norm": 0.21172094345092773,
      "learning_rate": 5.4989019757658466e-05,
      "loss": 1.1416,
      "mean_token_accuracy": 0.7741121739149094,
      "num_tokens": 1654963488.0,
      "step": 24500
    },
    {
      "entropy": 0.7441295340657235,
      "epoch": 7.425054911762478,
      "grad_norm": 0.21113701164722443,
      "learning_rate": 5.4955791809250204e-05,
      "loss": 1.1447,
      "mean_token_accuracy": 0.7777657687664032,
      "num_tokens": 1655627965.0,
      "step": 24510
    },
    {
      "entropy": 0.7464437022805214,
      "epoch": 7.428084526244035,
      "grad_norm": 0.21303771436214447,
      "learning_rate": 5.49225616503109e-05,
      "loss": 1.1352,
      "mean_token_accuracy": 0.7807260543107987,
      "num_tokens": 1656316403.0,
      "step": 24520
    },
    {
      "entropy": 0.7457913666963577,
      "epoch": 7.431114140725593,
      "grad_norm": 0.2176608294248581,
      "learning_rate": 5.488932929566283e-05,
      "loss": 1.1508,
      "mean_token_accuracy": 0.7762917503714561,
      "num_tokens": 1656984317.0,
      "step": 24530
    },
    {
      "entropy": 0.7491842567920685,
      "epoch": 7.43414375520715,
      "grad_norm": 0.21883437037467957,
      "learning_rate": 5.485609476012931e-05,
      "loss": 1.1454,
      "mean_token_accuracy": 0.7750205948948861,
      "num_tokens": 1657648557.0,
      "step": 24540
    },
    {
      "entropy": 0.7478434219956398,
      "epoch": 7.437173369688707,
      "grad_norm": 0.22971990704536438,
      "learning_rate": 5.482285805853461e-05,
      "loss": 1.1351,
      "mean_token_accuracy": 0.7778780668973923,
      "num_tokens": 1658330995.0,
      "step": 24550
    },
    {
      "entropy": 0.7507547929883003,
      "epoch": 7.440202984170265,
      "grad_norm": 0.22184528410434723,
      "learning_rate": 5.478961920570394e-05,
      "loss": 1.1445,
      "mean_token_accuracy": 0.7745205998420716,
      "num_tokens": 1658996530.0,
      "step": 24560
    },
    {
      "entropy": 0.7461107209324837,
      "epoch": 7.4432325986518215,
      "grad_norm": 0.2146303355693817,
      "learning_rate": 5.47563782164635e-05,
      "loss": 1.1437,
      "mean_token_accuracy": 0.777275325357914,
      "num_tokens": 1659671191.0,
      "step": 24570
    },
    {
      "entropy": 0.7430723607540131,
      "epoch": 7.446262213133378,
      "grad_norm": 0.20267100632190704,
      "learning_rate": 5.4723135105640445e-05,
      "loss": 1.1381,
      "mean_token_accuracy": 0.7795214995741844,
      "num_tokens": 1660352543.0,
      "step": 24580
    },
    {
      "entropy": 0.7431056842207908,
      "epoch": 7.449291827614936,
      "grad_norm": 0.21331390738487244,
      "learning_rate": 5.4689889888062853e-05,
      "loss": 1.1363,
      "mean_token_accuracy": 0.7799687758088112,
      "num_tokens": 1661034375.0,
      "step": 24590
    },
    {
      "entropy": 0.7397371128201484,
      "epoch": 7.452321442096493,
      "grad_norm": 0.20924900472164154,
      "learning_rate": 5.465664257855976e-05,
      "loss": 1.1274,
      "mean_token_accuracy": 0.7805151045322418,
      "num_tokens": 1661726134.0,
      "step": 24600
    },
    {
      "entropy": 0.7526766210794449,
      "epoch": 7.45535105657805,
      "grad_norm": 0.21683505177497864,
      "learning_rate": 5.462339319196112e-05,
      "loss": 1.1512,
      "mean_token_accuracy": 0.7739051789045334,
      "num_tokens": 1662399680.0,
      "step": 24610
    },
    {
      "entropy": 0.7580562621355057,
      "epoch": 7.458380671059608,
      "grad_norm": 0.2254616767168045,
      "learning_rate": 5.459014174309785e-05,
      "loss": 1.1496,
      "mean_token_accuracy": 0.7714251548051834,
      "num_tokens": 1663063774.0,
      "step": 24620
    },
    {
      "entropy": 0.7435199916362762,
      "epoch": 7.461410285541165,
      "grad_norm": 0.20744958519935608,
      "learning_rate": 5.455688824680173e-05,
      "loss": 1.1344,
      "mean_token_accuracy": 0.7797993645071983,
      "num_tokens": 1663755581.0,
      "step": 24630
    },
    {
      "entropy": 0.7406926706433297,
      "epoch": 7.464439900022722,
      "grad_norm": 0.21881410479545593,
      "learning_rate": 5.452363271790548e-05,
      "loss": 1.1365,
      "mean_token_accuracy": 0.7793450295925141,
      "num_tokens": 1664433891.0,
      "step": 24640
    },
    {
      "entropy": 0.7477216809988022,
      "epoch": 7.4674695145042795,
      "grad_norm": 0.21395400166511536,
      "learning_rate": 5.449037517124276e-05,
      "loss": 1.1405,
      "mean_token_accuracy": 0.7774820044636727,
      "num_tokens": 1665111657.0,
      "step": 24650
    },
    {
      "entropy": 0.747513872385025,
      "epoch": 7.4704991289858365,
      "grad_norm": 0.22176103293895721,
      "learning_rate": 5.445711562164809e-05,
      "loss": 1.1404,
      "mean_token_accuracy": 0.7792335256934166,
      "num_tokens": 1665799335.0,
      "step": 24660
    },
    {
      "entropy": 0.7383706942200661,
      "epoch": 7.473528743467393,
      "grad_norm": 0.210130512714386,
      "learning_rate": 5.442385408395686e-05,
      "loss": 1.1351,
      "mean_token_accuracy": 0.7739451065659523,
      "num_tokens": 1666463124.0,
      "step": 24670
    },
    {
      "entropy": 0.7512979134917259,
      "epoch": 7.476558357948951,
      "grad_norm": 0.201966792345047,
      "learning_rate": 5.4390590573005415e-05,
      "loss": 1.1382,
      "mean_token_accuracy": 0.7743869826197625,
      "num_tokens": 1667141456.0,
      "step": 24680
    },
    {
      "entropy": 0.7560889542102813,
      "epoch": 7.479587972430508,
      "grad_norm": 0.22573871910572052,
      "learning_rate": 5.435732510363094e-05,
      "loss": 1.1475,
      "mean_token_accuracy": 0.7724868059158325,
      "num_tokens": 1667821834.0,
      "step": 24690
    },
    {
      "entropy": 0.7346357598900795,
      "epoch": 7.482617586912065,
      "grad_norm": 0.2244092971086502,
      "learning_rate": 5.4324057690671485e-05,
      "loss": 1.14,
      "mean_token_accuracy": 0.7774963542819023,
      "num_tokens": 1668492166.0,
      "step": 24700
    },
    {
      "entropy": 0.7364307418465614,
      "epoch": 7.485647201393623,
      "grad_norm": 0.22128066420555115,
      "learning_rate": 5.429078834896598e-05,
      "loss": 1.1353,
      "mean_token_accuracy": 0.7787325322628021,
      "num_tokens": 1669155086.0,
      "step": 24710
    },
    {
      "entropy": 0.7441545367240906,
      "epoch": 7.48867681587518,
      "grad_norm": 0.2105528563261032,
      "learning_rate": 5.4257517093354246e-05,
      "loss": 1.1442,
      "mean_token_accuracy": 0.7805730924010277,
      "num_tokens": 1669837761.0,
      "step": 24720
    },
    {
      "entropy": 0.7354769557714462,
      "epoch": 7.491706430356738,
      "grad_norm": 0.21838191151618958,
      "learning_rate": 5.422424393867689e-05,
      "loss": 1.1341,
      "mean_token_accuracy": 0.7799891889095306,
      "num_tokens": 1670510937.0,
      "step": 24730
    },
    {
      "entropy": 0.746819120645523,
      "epoch": 7.4947360448382945,
      "grad_norm": 0.20795127749443054,
      "learning_rate": 5.419096889977542e-05,
      "loss": 1.1453,
      "mean_token_accuracy": 0.7764046773314476,
      "num_tokens": 1671187171.0,
      "step": 24740
    },
    {
      "entropy": 0.7433664634823799,
      "epoch": 7.497765659319851,
      "grad_norm": 0.21214525401592255,
      "learning_rate": 5.4157691991492174e-05,
      "loss": 1.1427,
      "mean_token_accuracy": 0.7786703392863273,
      "num_tokens": 1671866647.0,
      "step": 24750
    },
    {
      "entropy": 0.7340948358178139,
      "epoch": 7.500795273801408,
      "grad_norm": 0.22363007068634033,
      "learning_rate": 5.412441322867031e-05,
      "loss": 1.1377,
      "mean_token_accuracy": 0.7721888646483421,
      "num_tokens": 1672519297.0,
      "step": 24760
    },
    {
      "entropy": 0.7401010930538178,
      "epoch": 7.503824888282966,
      "grad_norm": 0.21117638051509857,
      "learning_rate": 5.409113262615381e-05,
      "loss": 1.1263,
      "mean_token_accuracy": 0.7808028668165207,
      "num_tokens": 1673211238.0,
      "step": 24770
    },
    {
      "entropy": 0.7396194770932197,
      "epoch": 7.506854502764523,
      "grad_norm": 0.22600512206554413,
      "learning_rate": 5.4057850198787506e-05,
      "loss": 1.1395,
      "mean_token_accuracy": 0.7758064150810242,
      "num_tokens": 1673869540.0,
      "step": 24780
    },
    {
      "entropy": 0.7452031940221786,
      "epoch": 7.509884117246081,
      "grad_norm": 0.21701331436634064,
      "learning_rate": 5.402456596141703e-05,
      "loss": 1.1528,
      "mean_token_accuracy": 0.7733472928404808,
      "num_tokens": 1674527406.0,
      "step": 24790
    },
    {
      "entropy": 0.734237490594387,
      "epoch": 7.512913731727638,
      "grad_norm": 0.22200846672058105,
      "learning_rate": 5.399127992888878e-05,
      "loss": 1.1346,
      "mean_token_accuracy": 0.7834853321313858,
      "num_tokens": 1675208503.0,
      "step": 24800
    },
    {
      "entropy": 0.7524815320968627,
      "epoch": 7.515943346209195,
      "grad_norm": 0.21497096121311188,
      "learning_rate": 5.395799211605004e-05,
      "loss": 1.1439,
      "mean_token_accuracy": 0.7742198213934899,
      "num_tokens": 1675893075.0,
      "step": 24810
    },
    {
      "entropy": 0.7457500502467156,
      "epoch": 7.518972960690752,
      "grad_norm": 0.2262214720249176,
      "learning_rate": 5.3924702537748815e-05,
      "loss": 1.1449,
      "mean_token_accuracy": 0.7760716915130615,
      "num_tokens": 1676569241.0,
      "step": 24820
    },
    {
      "entropy": 0.740559920668602,
      "epoch": 7.522002575172309,
      "grad_norm": 0.20782777667045593,
      "learning_rate": 5.389141120883392e-05,
      "loss": 1.1446,
      "mean_token_accuracy": 0.7764463156461716,
      "num_tokens": 1677244238.0,
      "step": 24830
    },
    {
      "entropy": 0.7468500062823296,
      "epoch": 7.525032189653866,
      "grad_norm": 0.21614263951778412,
      "learning_rate": 5.385811814415496e-05,
      "loss": 1.1408,
      "mean_token_accuracy": 0.7774775564670563,
      "num_tokens": 1677921825.0,
      "step": 24840
    },
    {
      "entropy": 0.7555354133248329,
      "epoch": 7.528061804135424,
      "grad_norm": 0.23248232901096344,
      "learning_rate": 5.382482335856229e-05,
      "loss": 1.1413,
      "mean_token_accuracy": 0.7761192858219147,
      "num_tokens": 1678606445.0,
      "step": 24850
    },
    {
      "entropy": 0.7485510841012001,
      "epoch": 7.531091418616981,
      "grad_norm": 0.22634170949459076,
      "learning_rate": 5.3791526866907084e-05,
      "loss": 1.1472,
      "mean_token_accuracy": 0.7722590655088425,
      "num_tokens": 1679271499.0,
      "step": 24860
    },
    {
      "entropy": 0.7513169005513192,
      "epoch": 7.534121033098538,
      "grad_norm": 0.21921472251415253,
      "learning_rate": 5.375822868404121e-05,
      "loss": 1.1413,
      "mean_token_accuracy": 0.7759198114275933,
      "num_tokens": 1679950085.0,
      "step": 24870
    },
    {
      "entropy": 0.7495874270796776,
      "epoch": 7.537150647580096,
      "grad_norm": 0.2265593707561493,
      "learning_rate": 5.372492882481734e-05,
      "loss": 1.1484,
      "mean_token_accuracy": 0.7736220329999923,
      "num_tokens": 1680618916.0,
      "step": 24880
    },
    {
      "entropy": 0.7537253230810166,
      "epoch": 7.540180262061653,
      "grad_norm": 0.20332907140254974,
      "learning_rate": 5.369162730408889e-05,
      "loss": 1.1435,
      "mean_token_accuracy": 0.7734320342540741,
      "num_tokens": 1681285210.0,
      "step": 24890
    },
    {
      "entropy": 0.7369566798210144,
      "epoch": 7.54320987654321,
      "grad_norm": 0.2314373254776001,
      "learning_rate": 5.365832413670997e-05,
      "loss": 1.1266,
      "mean_token_accuracy": 0.7839697256684304,
      "num_tokens": 1681979698.0,
      "step": 24900
    },
    {
      "entropy": 0.7459864601492882,
      "epoch": 7.546239491024767,
      "grad_norm": 0.22535762190818787,
      "learning_rate": 5.3625019337535475e-05,
      "loss": 1.1399,
      "mean_token_accuracy": 0.7763787895441056,
      "num_tokens": 1682648758.0,
      "step": 24910
    },
    {
      "entropy": 0.7416107371449471,
      "epoch": 7.549269105506324,
      "grad_norm": 0.21356399357318878,
      "learning_rate": 5.3591712921421024e-05,
      "loss": 1.1379,
      "mean_token_accuracy": 0.7756614923477173,
      "num_tokens": 1683314524.0,
      "step": 24920
    },
    {
      "entropy": 0.7331911876797677,
      "epoch": 7.552298719987881,
      "grad_norm": 0.2103649228811264,
      "learning_rate": 5.355840490322293e-05,
      "loss": 1.1323,
      "mean_token_accuracy": 0.7800414964556694,
      "num_tokens": 1683995587.0,
      "step": 24930
    },
    {
      "entropy": 0.7561028689146042,
      "epoch": 7.555328334469439,
      "grad_norm": 0.21475647389888763,
      "learning_rate": 5.352509529779823e-05,
      "loss": 1.1496,
      "mean_token_accuracy": 0.7750357866287232,
      "num_tokens": 1684671456.0,
      "step": 24940
    },
    {
      "entropy": 0.7422101736068726,
      "epoch": 7.558357948950996,
      "grad_norm": 0.21627213060855865,
      "learning_rate": 5.3491784120004685e-05,
      "loss": 1.1407,
      "mean_token_accuracy": 0.776326359808445,
      "num_tokens": 1685338459.0,
      "step": 24950
    },
    {
      "entropy": 0.7497922003269195,
      "epoch": 7.561387563432553,
      "grad_norm": 0.22639530897140503,
      "learning_rate": 5.345847138470077e-05,
      "loss": 1.1428,
      "mean_token_accuracy": 0.7732251092791558,
      "num_tokens": 1685995806.0,
      "step": 24960
    },
    {
      "entropy": 0.7480155780911446,
      "epoch": 7.564417177914111,
      "grad_norm": 0.22280606627464294,
      "learning_rate": 5.342515710674558e-05,
      "loss": 1.1517,
      "mean_token_accuracy": 0.7730533495545387,
      "num_tokens": 1686646088.0,
      "step": 24970
    },
    {
      "entropy": 0.7508288726210595,
      "epoch": 7.567446792395668,
      "grad_norm": 0.21140004694461823,
      "learning_rate": 5.339184130099898e-05,
      "loss": 1.1407,
      "mean_token_accuracy": 0.7772667378187179,
      "num_tokens": 1687332806.0,
      "step": 24980
    },
    {
      "entropy": 0.7477628767490387,
      "epoch": 7.570476406877225,
      "grad_norm": 0.24616877734661102,
      "learning_rate": 5.3358523982321494e-05,
      "loss": 1.1394,
      "mean_token_accuracy": 0.7770170167088508,
      "num_tokens": 1688018557.0,
      "step": 24990
    },
    {
      "entropy": 0.7574859589338303,
      "epoch": 7.573506021358782,
      "grad_norm": 0.21998058259487152,
      "learning_rate": 5.3325205165574297e-05,
      "loss": 1.153,
      "mean_token_accuracy": 0.7750434413552284,
      "num_tokens": 1688703140.0,
      "step": 25000
    },
    {
      "entropy": 0.7477376744151115,
      "epoch": 7.576535635840339,
      "grad_norm": 0.21323354542255402,
      "learning_rate": 5.3291884865619256e-05,
      "loss": 1.1402,
      "mean_token_accuracy": 0.7790637165307999,
      "num_tokens": 1689382179.0,
      "step": 25010
    },
    {
      "entropy": 0.7436079770326615,
      "epoch": 7.579565250321896,
      "grad_norm": 0.20804129540920258,
      "learning_rate": 5.325856309731888e-05,
      "loss": 1.1383,
      "mean_token_accuracy": 0.7806669667363166,
      "num_tokens": 1690076304.0,
      "step": 25020
    },
    {
      "entropy": 0.7412419065833091,
      "epoch": 7.582594864803454,
      "grad_norm": 0.22202837467193604,
      "learning_rate": 5.322523987553636e-05,
      "loss": 1.1364,
      "mean_token_accuracy": 0.7770675376057625,
      "num_tokens": 1690744702.0,
      "step": 25030
    },
    {
      "entropy": 0.746268455684185,
      "epoch": 7.585624479285011,
      "grad_norm": 0.21350513398647308,
      "learning_rate": 5.31919152151355e-05,
      "loss": 1.1254,
      "mean_token_accuracy": 0.7802624836564064,
      "num_tokens": 1691441533.0,
      "step": 25040
    },
    {
      "entropy": 0.7400405585765839,
      "epoch": 7.588654093766568,
      "grad_norm": 0.21588623523712158,
      "learning_rate": 5.3158589130980783e-05,
      "loss": 1.1372,
      "mean_token_accuracy": 0.778198529779911,
      "num_tokens": 1692124994.0,
      "step": 25050
    },
    {
      "entropy": 0.7503207892179489,
      "epoch": 7.591683708248126,
      "grad_norm": 0.21895559132099152,
      "learning_rate": 5.312526163793732e-05,
      "loss": 1.1448,
      "mean_token_accuracy": 0.7747712373733521,
      "num_tokens": 1692796826.0,
      "step": 25060
    },
    {
      "entropy": 0.7457394242286682,
      "epoch": 7.594713322729683,
      "grad_norm": 0.21419158577919006,
      "learning_rate": 5.309193275087081e-05,
      "loss": 1.1391,
      "mean_token_accuracy": 0.7781403362751007,
      "num_tokens": 1693479871.0,
      "step": 25070
    },
    {
      "entropy": 0.7462506055831909,
      "epoch": 7.5977429372112395,
      "grad_norm": 0.21640583872795105,
      "learning_rate": 5.305860248464761e-05,
      "loss": 1.1383,
      "mean_token_accuracy": 0.777347669005394,
      "num_tokens": 1694158507.0,
      "step": 25080
    },
    {
      "entropy": 0.7491141736507416,
      "epoch": 7.600772551692797,
      "grad_norm": 0.21499212086200714,
      "learning_rate": 5.30252708541347e-05,
      "loss": 1.1479,
      "mean_token_accuracy": 0.7724004551768303,
      "num_tokens": 1694816905.0,
      "step": 25090
    },
    {
      "entropy": 0.7444032818078995,
      "epoch": 7.603802166174354,
      "grad_norm": 0.23837362229824066,
      "learning_rate": 5.2991937874199657e-05,
      "loss": 1.1406,
      "mean_token_accuracy": 0.7754931151866913,
      "num_tokens": 1695488226.0,
      "step": 25100
    },
    {
      "entropy": 0.7473259896039963,
      "epoch": 7.606831780655911,
      "grad_norm": 0.21419358253479004,
      "learning_rate": 5.295860355971065e-05,
      "loss": 1.136,
      "mean_token_accuracy": 0.7758742675185204,
      "num_tokens": 1696152897.0,
      "step": 25110
    },
    {
      "entropy": 0.7429569244384766,
      "epoch": 7.609861395137469,
      "grad_norm": 0.2154773473739624,
      "learning_rate": 5.292526792553645e-05,
      "loss": 1.1444,
      "mean_token_accuracy": 0.77743329256773,
      "num_tokens": 1696824855.0,
      "step": 25120
    },
    {
      "entropy": 0.7521555483341217,
      "epoch": 7.612891009619026,
      "grad_norm": 0.21548032760620117,
      "learning_rate": 5.2891930986546426e-05,
      "loss": 1.1434,
      "mean_token_accuracy": 0.7764998629689217,
      "num_tokens": 1697505906.0,
      "step": 25130
    },
    {
      "entropy": 0.746573394536972,
      "epoch": 7.615920624100584,
      "grad_norm": 0.20502008497714996,
      "learning_rate": 5.2858592757610494e-05,
      "loss": 1.1446,
      "mean_token_accuracy": 0.773410850763321,
      "num_tokens": 1698165815.0,
      "step": 25140
    },
    {
      "entropy": 0.7367995575070381,
      "epoch": 7.618950238582141,
      "grad_norm": 0.2106497585773468,
      "learning_rate": 5.282525325359919e-05,
      "loss": 1.138,
      "mean_token_accuracy": 0.7751655489206314,
      "num_tokens": 1698825451.0,
      "step": 25150
    },
    {
      "entropy": 0.7473216980695725,
      "epoch": 7.6219798530636975,
      "grad_norm": 0.2234392911195755,
      "learning_rate": 5.2791912489383624e-05,
      "loss": 1.1377,
      "mean_token_accuracy": 0.7727016642689705,
      "num_tokens": 1699495216.0,
      "step": 25160
    },
    {
      "entropy": 0.7424630865454673,
      "epoch": 7.625009467545254,
      "grad_norm": 0.22373464703559875,
      "learning_rate": 5.27585704798354e-05,
      "loss": 1.1395,
      "mean_token_accuracy": 0.7784733980894089,
      "num_tokens": 1700167107.0,
      "step": 25170
    },
    {
      "entropy": 0.7530706122517585,
      "epoch": 7.628039082026812,
      "grad_norm": 0.22579248249530792,
      "learning_rate": 5.2725227239826756e-05,
      "loss": 1.1412,
      "mean_token_accuracy": 0.776701420545578,
      "num_tokens": 1700853133.0,
      "step": 25180
    },
    {
      "entropy": 0.7394076824188233,
      "epoch": 7.631068696508369,
      "grad_norm": 0.23015308380126953,
      "learning_rate": 5.269188278423043e-05,
      "loss": 1.1262,
      "mean_token_accuracy": 0.7838101804256439,
      "num_tokens": 1701546542.0,
      "step": 25190
    },
    {
      "entropy": 0.739519964158535,
      "epoch": 7.634098310989927,
      "grad_norm": 0.2184506356716156,
      "learning_rate": 5.265853712791973e-05,
      "loss": 1.1408,
      "mean_token_accuracy": 0.7789064064621926,
      "num_tokens": 1702224022.0,
      "step": 25200
    },
    {
      "entropy": 0.7475892812013626,
      "epoch": 7.637127925471484,
      "grad_norm": 0.2101866602897644,
      "learning_rate": 5.2625190285768464e-05,
      "loss": 1.1338,
      "mean_token_accuracy": 0.7773853868246079,
      "num_tokens": 1702904070.0,
      "step": 25210
    },
    {
      "entropy": 0.7507322892546654,
      "epoch": 7.640157539953041,
      "grad_norm": 0.20114384591579437,
      "learning_rate": 5.2591842272651003e-05,
      "loss": 1.1473,
      "mean_token_accuracy": 0.7686016082763671,
      "num_tokens": 1703561907.0,
      "step": 25220
    },
    {
      "entropy": 0.7429238423705101,
      "epoch": 7.643187154434598,
      "grad_norm": 0.22367319464683533,
      "learning_rate": 5.255849310344224e-05,
      "loss": 1.1411,
      "mean_token_accuracy": 0.7787930741906166,
      "num_tokens": 1704247823.0,
      "step": 25230
    },
    {
      "entropy": 0.7425561651587487,
      "epoch": 7.6462167689161555,
      "grad_norm": 0.2082562893629074,
      "learning_rate": 5.252514279301756e-05,
      "loss": 1.1359,
      "mean_token_accuracy": 0.7764569118618965,
      "num_tokens": 1704917401.0,
      "step": 25240
    },
    {
      "entropy": 0.7391936540603637,
      "epoch": 7.6492463833977125,
      "grad_norm": 0.21968869864940643,
      "learning_rate": 5.2491791356252876e-05,
      "loss": 1.1344,
      "mean_token_accuracy": 0.7789005756378173,
      "num_tokens": 1705586357.0,
      "step": 25250
    },
    {
      "entropy": 0.7499315455555916,
      "epoch": 7.65227599787927,
      "grad_norm": 0.2282896488904953,
      "learning_rate": 5.2458438808024576e-05,
      "loss": 1.141,
      "mean_token_accuracy": 0.7771964773535729,
      "num_tokens": 1706262520.0,
      "step": 25260
    },
    {
      "entropy": 0.7449300393462182,
      "epoch": 7.655305612360827,
      "grad_norm": 0.20787255465984344,
      "learning_rate": 5.242508516320962e-05,
      "loss": 1.1449,
      "mean_token_accuracy": 0.7738001808524132,
      "num_tokens": 1706932425.0,
      "step": 25270
    },
    {
      "entropy": 0.734488981962204,
      "epoch": 7.658335226842384,
      "grad_norm": 0.21595390141010284,
      "learning_rate": 5.239173043668534e-05,
      "loss": 1.1319,
      "mean_token_accuracy": 0.7872363120317459,
      "num_tokens": 1707629153.0,
      "step": 25280
    },
    {
      "entropy": 0.7480002358555794,
      "epoch": 7.661364841323942,
      "grad_norm": 0.22453685104846954,
      "learning_rate": 5.2358374643329633e-05,
      "loss": 1.15,
      "mean_token_accuracy": 0.7736384928226471,
      "num_tokens": 1708291373.0,
      "step": 25290
    },
    {
      "entropy": 0.7501042276620865,
      "epoch": 7.664394455805499,
      "grad_norm": 0.21583005785942078,
      "learning_rate": 5.232501779802088e-05,
      "loss": 1.1425,
      "mean_token_accuracy": 0.7768845707178116,
      "num_tokens": 1708973385.0,
      "step": 25300
    },
    {
      "entropy": 0.742583692073822,
      "epoch": 7.667424070287056,
      "grad_norm": 0.2266470193862915,
      "learning_rate": 5.2291659915637866e-05,
      "loss": 1.1414,
      "mean_token_accuracy": 0.7782695710659027,
      "num_tokens": 1709653473.0,
      "step": 25310
    },
    {
      "entropy": 0.7440670862793922,
      "epoch": 7.6704536847686136,
      "grad_norm": 0.21403197944164276,
      "learning_rate": 5.225830101105988e-05,
      "loss": 1.14,
      "mean_token_accuracy": 0.7751417070627212,
      "num_tokens": 1710319434.0,
      "step": 25320
    },
    {
      "entropy": 0.753742316365242,
      "epoch": 7.6734832992501705,
      "grad_norm": 0.2077597677707672,
      "learning_rate": 5.22249410991667e-05,
      "loss": 1.1462,
      "mean_token_accuracy": 0.7745224148035049,
      "num_tokens": 1711012533.0,
      "step": 25330
    },
    {
      "entropy": 0.743423281610012,
      "epoch": 7.676512913731727,
      "grad_norm": 0.2180161029100418,
      "learning_rate": 5.2191580194838464e-05,
      "loss": 1.1358,
      "mean_token_accuracy": 0.7788239121437073,
      "num_tokens": 1711691176.0,
      "step": 25340
    },
    {
      "entropy": 0.7364968597888947,
      "epoch": 7.679542528213285,
      "grad_norm": 0.222105473279953,
      "learning_rate": 5.2158218312955845e-05,
      "loss": 1.1376,
      "mean_token_accuracy": 0.7743668854236603,
      "num_tokens": 1712351284.0,
      "step": 25350
    },
    {
      "entropy": 0.7402329221367836,
      "epoch": 7.682572142694842,
      "grad_norm": 0.2200395166873932,
      "learning_rate": 5.212485546839987e-05,
      "loss": 1.144,
      "mean_token_accuracy": 0.7772669464349746,
      "num_tokens": 1713015717.0,
      "step": 25360
    },
    {
      "entropy": 0.736244784295559,
      "epoch": 7.685601757176399,
      "grad_norm": 0.22047896683216095,
      "learning_rate": 5.209149167605207e-05,
      "loss": 1.1338,
      "mean_token_accuracy": 0.7799585655331611,
      "num_tokens": 1713692961.0,
      "step": 25370
    },
    {
      "entropy": 0.7455885216593743,
      "epoch": 7.688631371657957,
      "grad_norm": 0.21278539299964905,
      "learning_rate": 5.205812695079436e-05,
      "loss": 1.1394,
      "mean_token_accuracy": 0.7775056421756744,
      "num_tokens": 1714366935.0,
      "step": 25380
    },
    {
      "entropy": 0.743017515540123,
      "epoch": 7.691660986139514,
      "grad_norm": 0.20847824215888977,
      "learning_rate": 5.202476130750906e-05,
      "loss": 1.1231,
      "mean_token_accuracy": 0.7809359863400459,
      "num_tokens": 1715048623.0,
      "step": 25390
    },
    {
      "entropy": 0.7346621245145798,
      "epoch": 7.694690600621071,
      "grad_norm": 0.20856250822544098,
      "learning_rate": 5.1991394761078946e-05,
      "loss": 1.1331,
      "mean_token_accuracy": 0.7845747739076614,
      "num_tokens": 1715741602.0,
      "step": 25400
    },
    {
      "entropy": 0.7453661963343621,
      "epoch": 7.6977202151026285,
      "grad_norm": 0.2149330973625183,
      "learning_rate": 5.195802732638714e-05,
      "loss": 1.1363,
      "mean_token_accuracy": 0.7745748803019523,
      "num_tokens": 1716413572.0,
      "step": 25410
    },
    {
      "entropy": 0.7401582688093186,
      "epoch": 7.700749829584185,
      "grad_norm": 0.2142152488231659,
      "learning_rate": 5.192465901831718e-05,
      "loss": 1.1314,
      "mean_token_accuracy": 0.7795200228691102,
      "num_tokens": 1717095588.0,
      "step": 25420
    },
    {
      "entropy": 0.7450745359063149,
      "epoch": 7.703779444065742,
      "grad_norm": 0.22123929858207703,
      "learning_rate": 5.1891289851753034e-05,
      "loss": 1.1394,
      "mean_token_accuracy": 0.7832079187035561,
      "num_tokens": 1717803940.0,
      "step": 25430
    },
    {
      "entropy": 0.7458594724535942,
      "epoch": 7.7068090585473,
      "grad_norm": 0.21973532438278198,
      "learning_rate": 5.1857919841579006e-05,
      "loss": 1.1419,
      "mean_token_accuracy": 0.7753015145659446,
      "num_tokens": 1718474653.0,
      "step": 25440
    },
    {
      "entropy": 0.7485863000154496,
      "epoch": 7.709838673028857,
      "grad_norm": 0.22180785238742828,
      "learning_rate": 5.182454900267977e-05,
      "loss": 1.1425,
      "mean_token_accuracy": 0.7791336089372635,
      "num_tokens": 1719157344.0,
      "step": 25450
    },
    {
      "entropy": 0.7507815167307854,
      "epoch": 7.712868287510414,
      "grad_norm": 0.2153090238571167,
      "learning_rate": 5.179117734994041e-05,
      "loss": 1.1458,
      "mean_token_accuracy": 0.7759419962763786,
      "num_tokens": 1719836087.0,
      "step": 25460
    },
    {
      "entropy": 0.7442035615444184,
      "epoch": 7.715897901991972,
      "grad_norm": 0.2220461368560791,
      "learning_rate": 5.1757804898246354e-05,
      "loss": 1.1401,
      "mean_token_accuracy": 0.7786268338561058,
      "num_tokens": 1720520273.0,
      "step": 25470
    },
    {
      "entropy": 0.7401345342397689,
      "epoch": 7.718927516473529,
      "grad_norm": 0.2081901878118515,
      "learning_rate": 5.1724431662483374e-05,
      "loss": 1.1348,
      "mean_token_accuracy": 0.7731791481375694,
      "num_tokens": 1721182612.0,
      "step": 25480
    },
    {
      "entropy": 0.7447274580597878,
      "epoch": 7.721957130955086,
      "grad_norm": 0.217488631606102,
      "learning_rate": 5.169105765753761e-05,
      "loss": 1.1467,
      "mean_token_accuracy": 0.776178726553917,
      "num_tokens": 1721857650.0,
      "step": 25490
    },
    {
      "entropy": 0.7468136623501778,
      "epoch": 7.724986745436643,
      "grad_norm": 0.2096599042415619,
      "learning_rate": 5.1657682898295546e-05,
      "loss": 1.1353,
      "mean_token_accuracy": 0.7728841468691826,
      "num_tokens": 1722529175.0,
      "step": 25500
    },
    {
      "entropy": 0.7344526037573814,
      "epoch": 7.7280163599182,
      "grad_norm": 0.2137104570865631,
      "learning_rate": 5.162430739964397e-05,
      "loss": 1.1354,
      "mean_token_accuracy": 0.7796164557337761,
      "num_tokens": 1723209381.0,
      "step": 25510
    },
    {
      "entropy": 0.7481473803520202,
      "epoch": 7.731045974399757,
      "grad_norm": 0.22086377441883087,
      "learning_rate": 5.159093117647004e-05,
      "loss": 1.1525,
      "mean_token_accuracy": 0.7749302208423614,
      "num_tokens": 1723880355.0,
      "step": 25520
    },
    {
      "entropy": 0.7484655126929283,
      "epoch": 7.734075588881315,
      "grad_norm": 0.22618655860424042,
      "learning_rate": 5.155755424366122e-05,
      "loss": 1.139,
      "mean_token_accuracy": 0.7727179780602456,
      "num_tokens": 1724542211.0,
      "step": 25530
    },
    {
      "entropy": 0.7443579941987991,
      "epoch": 7.737105203362872,
      "grad_norm": 0.21905077993869781,
      "learning_rate": 5.152417661610529e-05,
      "loss": 1.1343,
      "mean_token_accuracy": 0.7727830931544304,
      "num_tokens": 1725198559.0,
      "step": 25540
    },
    {
      "entropy": 0.7451524212956429,
      "epoch": 7.740134817844429,
      "grad_norm": 0.2269979566335678,
      "learning_rate": 5.149079830869035e-05,
      "loss": 1.1447,
      "mean_token_accuracy": 0.7765819162130356,
      "num_tokens": 1725871918.0,
      "step": 25550
    },
    {
      "entropy": 0.7532629147171974,
      "epoch": 7.743164432325987,
      "grad_norm": 0.2097410261631012,
      "learning_rate": 5.145741933630477e-05,
      "loss": 1.1443,
      "mean_token_accuracy": 0.7739479586482048,
      "num_tokens": 1726548865.0,
      "step": 25560
    },
    {
      "entropy": 0.7375313773751259,
      "epoch": 7.746194046807544,
      "grad_norm": 0.22446924448013306,
      "learning_rate": 5.142403971383727e-05,
      "loss": 1.138,
      "mean_token_accuracy": 0.7751024857163429,
      "num_tokens": 1727207165.0,
      "step": 25570
    },
    {
      "entropy": 0.7458680495619774,
      "epoch": 7.749223661289101,
      "grad_norm": 0.22164654731750488,
      "learning_rate": 5.1390659456176814e-05,
      "loss": 1.1455,
      "mean_token_accuracy": 0.7725724175572395,
      "num_tokens": 1727860610.0,
      "step": 25580
    },
    {
      "entropy": 0.7698509722948075,
      "epoch": 7.752253275770658,
      "grad_norm": 0.22158177196979523,
      "learning_rate": 5.135727857821267e-05,
      "loss": 1.1567,
      "mean_token_accuracy": 0.7673335209488868,
      "num_tokens": 1728532118.0,
      "step": 25590
    },
    {
      "entropy": 0.7348514199256897,
      "epoch": 7.755282890252215,
      "grad_norm": 0.1981198638677597,
      "learning_rate": 5.132389709483438e-05,
      "loss": 1.1375,
      "mean_token_accuracy": 0.780041366815567,
      "num_tokens": 1729205410.0,
      "step": 25600
    },
    {
      "entropy": 0.7483593672513962,
      "epoch": 7.758312504733773,
      "grad_norm": 0.21139384806156158,
      "learning_rate": 5.129051502093177e-05,
      "loss": 1.1462,
      "mean_token_accuracy": 0.7759598404169082,
      "num_tokens": 1729875927.0,
      "step": 25610
    },
    {
      "entropy": 0.7465812742710114,
      "epoch": 7.76134211921533,
      "grad_norm": 0.2231806516647339,
      "learning_rate": 5.12571323713949e-05,
      "loss": 1.1429,
      "mean_token_accuracy": 0.7714732691645623,
      "num_tokens": 1730528703.0,
      "step": 25620
    },
    {
      "entropy": 0.7381371155381202,
      "epoch": 7.764371733696887,
      "grad_norm": 0.21462588012218475,
      "learning_rate": 5.12237491611141e-05,
      "loss": 1.1367,
      "mean_token_accuracy": 0.7785015001893043,
      "num_tokens": 1731206259.0,
      "step": 25630
    },
    {
      "entropy": 0.759087823331356,
      "epoch": 7.767401348178444,
      "grad_norm": 0.21570007503032684,
      "learning_rate": 5.119036540497996e-05,
      "loss": 1.1529,
      "mean_token_accuracy": 0.7716951355338096,
      "num_tokens": 1731882044.0,
      "step": 25640
    },
    {
      "entropy": 0.7455698773264885,
      "epoch": 7.770430962660002,
      "grad_norm": 0.21832765638828278,
      "learning_rate": 5.1156981117883296e-05,
      "loss": 1.1406,
      "mean_token_accuracy": 0.7762769013643265,
      "num_tokens": 1732567092.0,
      "step": 25650
    },
    {
      "entropy": 0.7408061742782592,
      "epoch": 7.773460577141559,
      "grad_norm": 0.21853670477867126,
      "learning_rate": 5.1123596314715175e-05,
      "loss": 1.1315,
      "mean_token_accuracy": 0.7758067131042481,
      "num_tokens": 1733240134.0,
      "step": 25660
    },
    {
      "entropy": 0.7455203846096993,
      "epoch": 7.776490191623116,
      "grad_norm": 0.21512551605701447,
      "learning_rate": 5.1090211010366895e-05,
      "loss": 1.1435,
      "mean_token_accuracy": 0.7802461013197899,
      "num_tokens": 1733928793.0,
      "step": 25670
    },
    {
      "entropy": 0.7461658477783203,
      "epoch": 7.779519806104673,
      "grad_norm": 0.2122987061738968,
      "learning_rate": 5.1056825219729966e-05,
      "loss": 1.1327,
      "mean_token_accuracy": 0.7792137682437896,
      "num_tokens": 1734616105.0,
      "step": 25680
    },
    {
      "entropy": 0.7498957872390747,
      "epoch": 7.78254942058623,
      "grad_norm": 0.21380706131458282,
      "learning_rate": 5.1023438957696115e-05,
      "loss": 1.1388,
      "mean_token_accuracy": 0.7798877775669097,
      "num_tokens": 1735305057.0,
      "step": 25690
    },
    {
      "entropy": 0.7470949575304985,
      "epoch": 7.785579035067787,
      "grad_norm": 0.21742387115955353,
      "learning_rate": 5.099005223915729e-05,
      "loss": 1.1437,
      "mean_token_accuracy": 0.7777863383293152,
      "num_tokens": 1735983727.0,
      "step": 25700
    },
    {
      "entropy": 0.7559949979186058,
      "epoch": 7.788608649549345,
      "grad_norm": 0.22177299857139587,
      "learning_rate": 5.095666507900566e-05,
      "loss": 1.1466,
      "mean_token_accuracy": 0.7746515914797782,
      "num_tokens": 1736668849.0,
      "step": 25710
    },
    {
      "entropy": 0.756106062233448,
      "epoch": 7.791638264030902,
      "grad_norm": 0.21292948722839355,
      "learning_rate": 5.0923277492133526e-05,
      "loss": 1.1407,
      "mean_token_accuracy": 0.7747223898768425,
      "num_tokens": 1737340633.0,
      "step": 25720
    },
    {
      "entropy": 0.7445455402135849,
      "epoch": 7.79466787851246,
      "grad_norm": 0.22191624343395233,
      "learning_rate": 5.088988949343344e-05,
      "loss": 1.1506,
      "mean_token_accuracy": 0.7734212204813957,
      "num_tokens": 1737997480.0,
      "step": 25730
    },
    {
      "entropy": 0.7500823542475701,
      "epoch": 7.797697492994017,
      "grad_norm": 0.22217002511024475,
      "learning_rate": 5.0856501097798146e-05,
      "loss": 1.1516,
      "mean_token_accuracy": 0.7724595025181771,
      "num_tokens": 1738656996.0,
      "step": 25740
    },
    {
      "entropy": 0.742697848379612,
      "epoch": 7.8007271074755735,
      "grad_norm": 0.22848986089229584,
      "learning_rate": 5.0823112320120494e-05,
      "loss": 1.1402,
      "mean_token_accuracy": 0.7776058688759804,
      "num_tokens": 1739338729.0,
      "step": 25750
    },
    {
      "entropy": 0.7325052246451378,
      "epoch": 7.803756721957131,
      "grad_norm": 0.22159868478775024,
      "learning_rate": 5.078972317529358e-05,
      "loss": 1.1337,
      "mean_token_accuracy": 0.7774195194244384,
      "num_tokens": 1739995077.0,
      "step": 25760
    },
    {
      "entropy": 0.7375677943229675,
      "epoch": 7.806786336438688,
      "grad_norm": 0.20608189702033997,
      "learning_rate": 5.075633367821061e-05,
      "loss": 1.1387,
      "mean_token_accuracy": 0.7787287041544915,
      "num_tokens": 1740667260.0,
      "step": 25770
    },
    {
      "entropy": 0.748692835867405,
      "epoch": 7.809815950920245,
      "grad_norm": 0.220370352268219,
      "learning_rate": 5.0722943843765006e-05,
      "loss": 1.1296,
      "mean_token_accuracy": 0.7749065220355987,
      "num_tokens": 1741345071.0,
      "step": 25780
    },
    {
      "entropy": 0.7556084752082824,
      "epoch": 7.812845565401803,
      "grad_norm": 0.2178974449634552,
      "learning_rate": 5.0689553686850263e-05,
      "loss": 1.1404,
      "mean_token_accuracy": 0.7793405070900917,
      "num_tokens": 1742039572.0,
      "step": 25790
    },
    {
      "entropy": 0.7367696225643158,
      "epoch": 7.81587517988336,
      "grad_norm": 0.22308918833732605,
      "learning_rate": 5.0656163222360084e-05,
      "loss": 1.1419,
      "mean_token_accuracy": 0.7798249363899231,
      "num_tokens": 1742713423.0,
      "step": 25800
    },
    {
      "entropy": 0.742949104309082,
      "epoch": 7.818904794364917,
      "grad_norm": 0.2218913733959198,
      "learning_rate": 5.062277246518831e-05,
      "loss": 1.135,
      "mean_token_accuracy": 0.7802236065268516,
      "num_tokens": 1743400089.0,
      "step": 25810
    },
    {
      "entropy": 0.7405900165438652,
      "epoch": 7.821934408846475,
      "grad_norm": 0.23014149069786072,
      "learning_rate": 5.058938143022883e-05,
      "loss": 1.1359,
      "mean_token_accuracy": 0.7767721772193908,
      "num_tokens": 1744064439.0,
      "step": 25820
    },
    {
      "entropy": 0.7466139867901802,
      "epoch": 7.8249640233280315,
      "grad_norm": 0.22135604918003082,
      "learning_rate": 5.0555990132375753e-05,
      "loss": 1.1443,
      "mean_token_accuracy": 0.7720306783914566,
      "num_tokens": 1744728361.0,
      "step": 25830
    },
    {
      "entropy": 0.748503664135933,
      "epoch": 7.8279936378095885,
      "grad_norm": 0.21388953924179077,
      "learning_rate": 5.0522598586523286e-05,
      "loss": 1.1426,
      "mean_token_accuracy": 0.77419193983078,
      "num_tokens": 1745399338.0,
      "step": 25840
    },
    {
      "entropy": 0.749726516008377,
      "epoch": 7.831023252291146,
      "grad_norm": 0.2057487815618515,
      "learning_rate": 5.048920680756568e-05,
      "loss": 1.15,
      "mean_token_accuracy": 0.7743424639105797,
      "num_tokens": 1746070743.0,
      "step": 25850
    },
    {
      "entropy": 0.7458449557423592,
      "epoch": 7.834052866772703,
      "grad_norm": 0.2177976369857788,
      "learning_rate": 5.0455814810397374e-05,
      "loss": 1.1371,
      "mean_token_accuracy": 0.7758996948599816,
      "num_tokens": 1746741146.0,
      "step": 25860
    },
    {
      "entropy": 0.7548872157931328,
      "epoch": 7.83708248125426,
      "grad_norm": 0.2201891839504242,
      "learning_rate": 5.042242260991286e-05,
      "loss": 1.1364,
      "mean_token_accuracy": 0.7745819926261902,
      "num_tokens": 1747423056.0,
      "step": 25870
    },
    {
      "entropy": 0.7475612044334412,
      "epoch": 7.840112095735818,
      "grad_norm": 0.20265506207942963,
      "learning_rate": 5.038903022100675e-05,
      "loss": 1.1384,
      "mean_token_accuracy": 0.7801030427217484,
      "num_tokens": 1748110164.0,
      "step": 25880
    },
    {
      "entropy": 0.7405172199010849,
      "epoch": 7.843141710217375,
      "grad_norm": 0.2164795845746994,
      "learning_rate": 5.035563765857367e-05,
      "loss": 1.1318,
      "mean_token_accuracy": 0.7839020133018494,
      "num_tokens": 1748805998.0,
      "step": 25890
    },
    {
      "entropy": 0.7411181926727295,
      "epoch": 7.846171324698932,
      "grad_norm": 0.21809406578540802,
      "learning_rate": 5.032224493750841e-05,
      "loss": 1.1345,
      "mean_token_accuracy": 0.7742871686816215,
      "num_tokens": 1749459915.0,
      "step": 25900
    },
    {
      "entropy": 0.7525318443775177,
      "epoch": 7.8492009391804896,
      "grad_norm": 0.21590840816497803,
      "learning_rate": 5.028885207270579e-05,
      "loss": 1.1493,
      "mean_token_accuracy": 0.7758223533630371,
      "num_tokens": 1750138323.0,
      "step": 25910
    },
    {
      "entropy": 0.7444149792194367,
      "epoch": 7.8522305536620465,
      "grad_norm": 0.22707167267799377,
      "learning_rate": 5.025545907906071e-05,
      "loss": 1.14,
      "mean_token_accuracy": 0.7741804406046867,
      "num_tokens": 1750812267.0,
      "step": 25920
    },
    {
      "entropy": 0.7396515145897865,
      "epoch": 7.855260168143603,
      "grad_norm": 0.23188693821430206,
      "learning_rate": 5.022206597146809e-05,
      "loss": 1.1429,
      "mean_token_accuracy": 0.7752804264426232,
      "num_tokens": 1751466833.0,
      "step": 25930
    },
    {
      "entropy": 0.7539496392011642,
      "epoch": 7.858289782625161,
      "grad_norm": 0.222253680229187,
      "learning_rate": 5.018867276482293e-05,
      "loss": 1.1531,
      "mean_token_accuracy": 0.7719980478286743,
      "num_tokens": 1752123862.0,
      "step": 25940
    },
    {
      "entropy": 0.7511607229709625,
      "epoch": 7.861319397106718,
      "grad_norm": 0.21358038485050201,
      "learning_rate": 5.01552794740203e-05,
      "loss": 1.1465,
      "mean_token_accuracy": 0.7767821297049522,
      "num_tokens": 1752804083.0,
      "step": 25950
    },
    {
      "entropy": 0.7481107085943222,
      "epoch": 7.864349011588275,
      "grad_norm": 0.2152298241853714,
      "learning_rate": 5.012188611395523e-05,
      "loss": 1.1423,
      "mean_token_accuracy": 0.77596565335989,
      "num_tokens": 1753488779.0,
      "step": 25960
    },
    {
      "entropy": 0.7437504604458809,
      "epoch": 7.867378626069833,
      "grad_norm": 0.2288159728050232,
      "learning_rate": 5.008849269952287e-05,
      "loss": 1.1383,
      "mean_token_accuracy": 0.7771565794944764,
      "num_tokens": 1754166460.0,
      "step": 25970
    },
    {
      "entropy": 0.7486496567726135,
      "epoch": 7.87040824055139,
      "grad_norm": 0.22141054272651672,
      "learning_rate": 5.0055099245618343e-05,
      "loss": 1.1408,
      "mean_token_accuracy": 0.7729766875505447,
      "num_tokens": 1754839682.0,
      "step": 25980
    },
    {
      "entropy": 0.7523567467927933,
      "epoch": 7.873437855032947,
      "grad_norm": 0.2228238731622696,
      "learning_rate": 5.0021705767136784e-05,
      "loss": 1.151,
      "mean_token_accuracy": 0.7777088150382042,
      "num_tokens": 1755529541.0,
      "step": 25990
    },
    {
      "entropy": 0.7315017133951187,
      "epoch": 7.8764674695145045,
      "grad_norm": 0.2139190435409546,
      "learning_rate": 4.9988312278973374e-05,
      "loss": 1.1339,
      "mean_token_accuracy": 0.7816120103001595,
      "num_tokens": 1756207843.0,
      "step": 26000
    },
    {
      "entropy": 0.7476119130849839,
      "epoch": 7.879497083996061,
      "grad_norm": 0.21859972178936005,
      "learning_rate": 4.995491879602328e-05,
      "loss": 1.1514,
      "mean_token_accuracy": 0.7729114070534706,
      "num_tokens": 1756868766.0,
      "step": 26010
    },
    {
      "entropy": 0.7385255351662636,
      "epoch": 7.882526698477619,
      "grad_norm": 0.2217828333377838,
      "learning_rate": 4.9921525333181645e-05,
      "loss": 1.1259,
      "mean_token_accuracy": 0.7782105281949043,
      "num_tokens": 1757545544.0,
      "step": 26020
    },
    {
      "entropy": 0.7463892117142678,
      "epoch": 7.885556312959176,
      "grad_norm": 0.22420364618301392,
      "learning_rate": 4.988813190534364e-05,
      "loss": 1.1373,
      "mean_token_accuracy": 0.7758491992950439,
      "num_tokens": 1758214530.0,
      "step": 26030
    },
    {
      "entropy": 0.7418407365679741,
      "epoch": 7.888585927440733,
      "grad_norm": 0.22543227672576904,
      "learning_rate": 4.98547385274044e-05,
      "loss": 1.1343,
      "mean_token_accuracy": 0.7768672183156013,
      "num_tokens": 1758890862.0,
      "step": 26040
    },
    {
      "entropy": 0.7480473071336746,
      "epoch": 7.89161554192229,
      "grad_norm": 0.2153388112783432,
      "learning_rate": 4.982134521425906e-05,
      "loss": 1.1438,
      "mean_token_accuracy": 0.7791155099868774,
      "num_tokens": 1759585957.0,
      "step": 26050
    },
    {
      "entropy": 0.7421821311116219,
      "epoch": 7.894645156403848,
      "grad_norm": 0.2179073989391327,
      "learning_rate": 4.978795198080267e-05,
      "loss": 1.1399,
      "mean_token_accuracy": 0.773826913535595,
      "num_tokens": 1760254722.0,
      "step": 26060
    },
    {
      "entropy": 0.7383495569229126,
      "epoch": 7.897674770885405,
      "grad_norm": 0.22895166277885437,
      "learning_rate": 4.975455884193031e-05,
      "loss": 1.1325,
      "mean_token_accuracy": 0.7804876893758774,
      "num_tokens": 1760929886.0,
      "step": 26070
    },
    {
      "entropy": 0.7413220927119255,
      "epoch": 7.9007043853669625,
      "grad_norm": 0.2241257280111313,
      "learning_rate": 4.9721165812536994e-05,
      "loss": 1.1428,
      "mean_token_accuracy": 0.7776457533240319,
      "num_tokens": 1761608920.0,
      "step": 26080
    },
    {
      "entropy": 0.744784663617611,
      "epoch": 7.903733999848519,
      "grad_norm": 0.2179223597049713,
      "learning_rate": 4.9687772907517665e-05,
      "loss": 1.142,
      "mean_token_accuracy": 0.7793117061257362,
      "num_tokens": 1762298896.0,
      "step": 26090
    },
    {
      "entropy": 0.7535127177834511,
      "epoch": 7.906763614330076,
      "grad_norm": 0.21491022408008575,
      "learning_rate": 4.965438014176724e-05,
      "loss": 1.1477,
      "mean_token_accuracy": 0.7741543173789978,
      "num_tokens": 1762977870.0,
      "step": 26100
    },
    {
      "entropy": 0.748827189207077,
      "epoch": 7.909793228811633,
      "grad_norm": 0.216589093208313,
      "learning_rate": 4.962098753018057e-05,
      "loss": 1.1432,
      "mean_token_accuracy": 0.773657874763012,
      "num_tokens": 1763647714.0,
      "step": 26110
    },
    {
      "entropy": 0.7505015954375267,
      "epoch": 7.912822843293191,
      "grad_norm": 0.22123666107654572,
      "learning_rate": 4.958759508765241e-05,
      "loss": 1.1455,
      "mean_token_accuracy": 0.7739342629909516,
      "num_tokens": 1764314179.0,
      "step": 26120
    },
    {
      "entropy": 0.7430455759167671,
      "epoch": 7.915852457774748,
      "grad_norm": 0.24435074627399445,
      "learning_rate": 4.955420282907748e-05,
      "loss": 1.1447,
      "mean_token_accuracy": 0.7748620107769966,
      "num_tokens": 1764971979.0,
      "step": 26130
    },
    {
      "entropy": 0.7427011296153069,
      "epoch": 7.918882072256306,
      "grad_norm": 0.21677272021770477,
      "learning_rate": 4.95208107693504e-05,
      "loss": 1.1355,
      "mean_token_accuracy": 0.7760151118040085,
      "num_tokens": 1765651991.0,
      "step": 26140
    },
    {
      "entropy": 0.7467780768871307,
      "epoch": 7.921911686737863,
      "grad_norm": 0.21139970421791077,
      "learning_rate": 4.948741892336567e-05,
      "loss": 1.135,
      "mean_token_accuracy": 0.7797481656074524,
      "num_tokens": 1766335249.0,
      "step": 26150
    },
    {
      "entropy": 0.7491904139518738,
      "epoch": 7.92494130121942,
      "grad_norm": 0.21966759860515594,
      "learning_rate": 4.9454027306017756e-05,
      "loss": 1.148,
      "mean_token_accuracy": 0.774878802895546,
      "num_tokens": 1767002053.0,
      "step": 26160
    },
    {
      "entropy": 0.7397206977009774,
      "epoch": 7.927970915700977,
      "grad_norm": 0.2051488608121872,
      "learning_rate": 4.942063593220096e-05,
      "loss": 1.13,
      "mean_token_accuracy": 0.7776805475354195,
      "num_tokens": 1767669297.0,
      "step": 26170
    },
    {
      "entropy": 0.7347766533493996,
      "epoch": 7.931000530182534,
      "grad_norm": 0.20992016792297363,
      "learning_rate": 4.9387244816809525e-05,
      "loss": 1.1346,
      "mean_token_accuracy": 0.7761421769857406,
      "num_tokens": 1768332143.0,
      "step": 26180
    },
    {
      "entropy": 0.7496741846203804,
      "epoch": 7.934030144664091,
      "grad_norm": 0.21435385942459106,
      "learning_rate": 4.9353853974737564e-05,
      "loss": 1.1375,
      "mean_token_accuracy": 0.7734736561775207,
      "num_tokens": 1769001909.0,
      "step": 26190
    },
    {
      "entropy": 0.7377337127923965,
      "epoch": 7.937059759145649,
      "grad_norm": 0.21795615553855896,
      "learning_rate": 4.9320463420879045e-05,
      "loss": 1.1331,
      "mean_token_accuracy": 0.78009432554245,
      "num_tokens": 1769677823.0,
      "step": 26200
    },
    {
      "entropy": 0.734045286476612,
      "epoch": 7.940089373627206,
      "grad_norm": 0.2170575112104416,
      "learning_rate": 4.928707317012783e-05,
      "loss": 1.1313,
      "mean_token_accuracy": 0.7810531422495842,
      "num_tokens": 1770358913.0,
      "step": 26210
    },
    {
      "entropy": 0.7450289964675904,
      "epoch": 7.943118988108763,
      "grad_norm": 0.2235105037689209,
      "learning_rate": 4.925368323737766e-05,
      "loss": 1.1424,
      "mean_token_accuracy": 0.7792221143841743,
      "num_tokens": 1771048989.0,
      "step": 26220
    },
    {
      "entropy": 0.7492415189743042,
      "epoch": 7.946148602590321,
      "grad_norm": 0.2182435542345047,
      "learning_rate": 4.922029363752209e-05,
      "loss": 1.1375,
      "mean_token_accuracy": 0.7748518586158752,
      "num_tokens": 1771723596.0,
      "step": 26230
    },
    {
      "entropy": 0.7439008742570877,
      "epoch": 7.949178217071878,
      "grad_norm": 0.2185424119234085,
      "learning_rate": 4.9186904385454566e-05,
      "loss": 1.1383,
      "mean_token_accuracy": 0.7725673720240593,
      "num_tokens": 1772387819.0,
      "step": 26240
    },
    {
      "entropy": 0.7428147852420807,
      "epoch": 7.952207831553435,
      "grad_norm": 0.21208150684833527,
      "learning_rate": 4.915351549606835e-05,
      "loss": 1.139,
      "mean_token_accuracy": 0.7782357901334762,
      "num_tokens": 1773068984.0,
      "step": 26250
    },
    {
      "entropy": 0.7407799273729324,
      "epoch": 7.955237446034992,
      "grad_norm": 0.2177264541387558,
      "learning_rate": 4.912012698425659e-05,
      "loss": 1.1366,
      "mean_token_accuracy": 0.7793653145432472,
      "num_tokens": 1773751389.0,
      "step": 26260
    },
    {
      "entropy": 0.7367604196071624,
      "epoch": 7.958267060516549,
      "grad_norm": 0.2107536643743515,
      "learning_rate": 4.9086738864912196e-05,
      "loss": 1.1379,
      "mean_token_accuracy": 0.7817483022809029,
      "num_tokens": 1774426363.0,
      "step": 26270
    },
    {
      "entropy": 0.7470024287700653,
      "epoch": 7.961296674998106,
      "grad_norm": 0.2174767702817917,
      "learning_rate": 4.905335115292795e-05,
      "loss": 1.1386,
      "mean_token_accuracy": 0.7773818820714951,
      "num_tokens": 1775098932.0,
      "step": 26280
    },
    {
      "entropy": 0.7420658662915229,
      "epoch": 7.964326289479664,
      "grad_norm": 0.21864546835422516,
      "learning_rate": 4.9019963863196464e-05,
      "loss": 1.1315,
      "mean_token_accuracy": 0.7792339041829109,
      "num_tokens": 1775776540.0,
      "step": 26290
    },
    {
      "entropy": 0.7492674008011818,
      "epoch": 7.967355903961221,
      "grad_norm": 0.2162260264158249,
      "learning_rate": 4.8986577010610104e-05,
      "loss": 1.1404,
      "mean_token_accuracy": 0.7754534050822258,
      "num_tokens": 1776466246.0,
      "step": 26300
    },
    {
      "entropy": 0.7473296016454697,
      "epoch": 7.970385518442778,
      "grad_norm": 0.22251366078853607,
      "learning_rate": 4.89531906100611e-05,
      "loss": 1.1405,
      "mean_token_accuracy": 0.7775650814175605,
      "num_tokens": 1777150534.0,
      "step": 26310
    },
    {
      "entropy": 0.7539983540773392,
      "epoch": 7.973415132924336,
      "grad_norm": 0.21550723910331726,
      "learning_rate": 4.8919804676441463e-05,
      "loss": 1.1428,
      "mean_token_accuracy": 0.7771080404520034,
      "num_tokens": 1777829444.0,
      "step": 26320
    },
    {
      "entropy": 0.7489592969417572,
      "epoch": 7.976444747405893,
      "grad_norm": 0.23051001131534576,
      "learning_rate": 4.888641922464296e-05,
      "loss": 1.1414,
      "mean_token_accuracy": 0.7754841029644013,
      "num_tokens": 1778496537.0,
      "step": 26330
    },
    {
      "entropy": 0.7378235667943954,
      "epoch": 7.9794743618874495,
      "grad_norm": 0.22835935652256012,
      "learning_rate": 4.885303426955719e-05,
      "loss": 1.1253,
      "mean_token_accuracy": 0.7837060406804085,
      "num_tokens": 1779184947.0,
      "step": 26340
    },
    {
      "entropy": 0.7488358020782471,
      "epoch": 7.982503976369007,
      "grad_norm": 0.21689769625663757,
      "learning_rate": 4.881964982607553e-05,
      "loss": 1.1429,
      "mean_token_accuracy": 0.7760648384690285,
      "num_tokens": 1779864840.0,
      "step": 26350
    },
    {
      "entropy": 0.7419907003641129,
      "epoch": 7.985533590850564,
      "grad_norm": 0.22363406419754028,
      "learning_rate": 4.8786265909089076e-05,
      "loss": 1.1388,
      "mean_token_accuracy": 0.780245891213417,
      "num_tokens": 1780544252.0,
      "step": 26360
    },
    {
      "entropy": 0.7440352529287338,
      "epoch": 7.988563205332121,
      "grad_norm": 0.20977413654327393,
      "learning_rate": 4.8752882533488756e-05,
      "loss": 1.148,
      "mean_token_accuracy": 0.7776442632079125,
      "num_tokens": 1781215187.0,
      "step": 26370
    },
    {
      "entropy": 0.7335471183061599,
      "epoch": 7.991592819813679,
      "grad_norm": 0.22370311617851257,
      "learning_rate": 4.87194997141652e-05,
      "loss": 1.1253,
      "mean_token_accuracy": 0.7811989828944206,
      "num_tokens": 1781901541.0,
      "step": 26380
    },
    {
      "entropy": 0.7404125243425369,
      "epoch": 7.994622434295236,
      "grad_norm": 0.22028976678848267,
      "learning_rate": 4.8686117466008845e-05,
      "loss": 1.141,
      "mean_token_accuracy": 0.7835704386234283,
      "num_tokens": 1782591873.0,
      "step": 26390
    },
    {
      "entropy": 0.7476471602916718,
      "epoch": 7.997652048776793,
      "grad_norm": 0.22728370130062103,
      "learning_rate": 4.8652735803909814e-05,
      "loss": 1.1371,
      "mean_token_accuracy": 0.7757069900631904,
      "num_tokens": 1783270137.0,
      "step": 26400
    },
    {
      "entropy": 0.7424362485225384,
      "epoch": 8.00060592289631,
      "grad_norm": 0.19484201073646545,
      "learning_rate": 4.861935474275802e-05,
      "loss": 1.1397,
      "mean_token_accuracy": 0.7754941108899239,
      "num_tokens": 1783925499.0,
      "step": 26410
    },
    {
      "entropy": 0.7199281647801399,
      "epoch": 8.00363553737787,
      "grad_norm": 0.25212445855140686,
      "learning_rate": 4.8585974297443074e-05,
      "loss": 1.1066,
      "mean_token_accuracy": 0.780900226533413,
      "num_tokens": 1784596583.0,
      "step": 26420
    },
    {
      "entropy": 0.722083055973053,
      "epoch": 8.006665151859426,
      "grad_norm": 0.23797836899757385,
      "learning_rate": 4.855259448285433e-05,
      "loss": 1.1062,
      "mean_token_accuracy": 0.7874858692288399,
      "num_tokens": 1785290769.0,
      "step": 26430
    },
    {
      "entropy": 0.7245304256677627,
      "epoch": 8.009694766340983,
      "grad_norm": 0.24838151037693024,
      "learning_rate": 4.851921531388085e-05,
      "loss": 1.1132,
      "mean_token_accuracy": 0.7804965898394585,
      "num_tokens": 1785970378.0,
      "step": 26440
    },
    {
      "entropy": 0.7161515861749649,
      "epoch": 8.01272438082254,
      "grad_norm": 0.2590703070163727,
      "learning_rate": 4.848583680541143e-05,
      "loss": 1.1095,
      "mean_token_accuracy": 0.7841158762574196,
      "num_tokens": 1786648781.0,
      "step": 26450
    },
    {
      "entropy": 0.7293979570269584,
      "epoch": 8.015753995304097,
      "grad_norm": 0.24743109941482544,
      "learning_rate": 4.845245897233453e-05,
      "loss": 1.1126,
      "mean_token_accuracy": 0.7821021333336831,
      "num_tokens": 1787344983.0,
      "step": 26460
    },
    {
      "entropy": 0.7267181724309921,
      "epoch": 8.018783609785654,
      "grad_norm": 0.23520056903362274,
      "learning_rate": 4.841908182953833e-05,
      "loss": 1.1109,
      "mean_token_accuracy": 0.7799164101481437,
      "num_tokens": 1788032482.0,
      "step": 26470
    },
    {
      "entropy": 0.7187842220067978,
      "epoch": 8.021813224267213,
      "grad_norm": 0.23745307326316833,
      "learning_rate": 4.8385705391910734e-05,
      "loss": 1.0985,
      "mean_token_accuracy": 0.781644044816494,
      "num_tokens": 1788711125.0,
      "step": 26480
    },
    {
      "entropy": 0.7077472984790802,
      "epoch": 8.02484283874877,
      "grad_norm": 0.24203577637672424,
      "learning_rate": 4.835232967433931e-05,
      "loss": 1.0988,
      "mean_token_accuracy": 0.783315247297287,
      "num_tokens": 1789364497.0,
      "step": 26490
    },
    {
      "entropy": 0.7164745584130288,
      "epoch": 8.027872453230326,
      "grad_norm": 0.2467387318611145,
      "learning_rate": 4.831895469171126e-05,
      "loss": 1.0961,
      "mean_token_accuracy": 0.7835220918059349,
      "num_tokens": 1790045867.0,
      "step": 26500
    },
    {
      "entropy": 0.7072773531079293,
      "epoch": 8.030902067711883,
      "grad_norm": 0.25449636578559875,
      "learning_rate": 4.82855804589135e-05,
      "loss": 1.0966,
      "mean_token_accuracy": 0.7852317169308662,
      "num_tokens": 1790726659.0,
      "step": 26510
    },
    {
      "entropy": 0.7197290822863579,
      "epoch": 8.03393168219344,
      "grad_norm": 0.2365366667509079,
      "learning_rate": 4.825220699083263e-05,
      "loss": 1.1083,
      "mean_token_accuracy": 0.7822935700416564,
      "num_tokens": 1791407340.0,
      "step": 26520
    },
    {
      "entropy": 0.7254417568445206,
      "epoch": 8.036961296674997,
      "grad_norm": 0.26874080300331116,
      "learning_rate": 4.8218834302354875e-05,
      "loss": 1.1101,
      "mean_token_accuracy": 0.7791984334588051,
      "num_tokens": 1792081333.0,
      "step": 26530
    },
    {
      "entropy": 0.7225031182169914,
      "epoch": 8.039990911156556,
      "grad_norm": 0.24366725981235504,
      "learning_rate": 4.818546240836612e-05,
      "loss": 1.1045,
      "mean_token_accuracy": 0.7792498484253884,
      "num_tokens": 1792751027.0,
      "step": 26540
    },
    {
      "entropy": 0.7129725962877274,
      "epoch": 8.043020525638113,
      "grad_norm": 0.2428189367055893,
      "learning_rate": 4.81520913237519e-05,
      "loss": 1.1037,
      "mean_token_accuracy": 0.7835505723953247,
      "num_tokens": 1793417824.0,
      "step": 26550
    },
    {
      "entropy": 0.7184087797999382,
      "epoch": 8.04605014011967,
      "grad_norm": 0.24311953783035278,
      "learning_rate": 4.81187210633974e-05,
      "loss": 1.1099,
      "mean_token_accuracy": 0.7814700752496719,
      "num_tokens": 1794089530.0,
      "step": 26560
    },
    {
      "entropy": 0.718797305226326,
      "epoch": 8.049079754601227,
      "grad_norm": 0.2504687011241913,
      "learning_rate": 4.808535164218741e-05,
      "loss": 1.1038,
      "mean_token_accuracy": 0.7837594002485275,
      "num_tokens": 1794773932.0,
      "step": 26570
    },
    {
      "entropy": 0.7104886934161186,
      "epoch": 8.052109369082784,
      "grad_norm": 0.25315359234809875,
      "learning_rate": 4.8051983075006374e-05,
      "loss": 1.1109,
      "mean_token_accuracy": 0.78170804977417,
      "num_tokens": 1795441765.0,
      "step": 26580
    },
    {
      "entropy": 0.7183498606085778,
      "epoch": 8.05513898356434,
      "grad_norm": 0.2640795111656189,
      "learning_rate": 4.801861537673833e-05,
      "loss": 1.111,
      "mean_token_accuracy": 0.7804919451475143,
      "num_tokens": 1796113390.0,
      "step": 26590
    },
    {
      "entropy": 0.7245974063873291,
      "epoch": 8.0581685980459,
      "grad_norm": 0.24883559346199036,
      "learning_rate": 4.798524856226696e-05,
      "loss": 1.1092,
      "mean_token_accuracy": 0.7838584095239639,
      "num_tokens": 1796806113.0,
      "step": 26600
    },
    {
      "entropy": 0.7104981392621994,
      "epoch": 8.061198212527456,
      "grad_norm": 0.25070619583129883,
      "learning_rate": 4.79518826464755e-05,
      "loss": 1.1059,
      "mean_token_accuracy": 0.7818232834339142,
      "num_tokens": 1797475384.0,
      "step": 26610
    },
    {
      "entropy": 0.716577960550785,
      "epoch": 8.064227827009013,
      "grad_norm": 0.25383082032203674,
      "learning_rate": 4.7918517644246855e-05,
      "loss": 1.1033,
      "mean_token_accuracy": 0.782039375603199,
      "num_tokens": 1798149303.0,
      "step": 26620
    },
    {
      "entropy": 0.7199601382017136,
      "epoch": 8.06725744149057,
      "grad_norm": 0.24829378724098206,
      "learning_rate": 4.788515357046347e-05,
      "loss": 1.1012,
      "mean_token_accuracy": 0.7868308603763581,
      "num_tokens": 1798846822.0,
      "step": 26630
    },
    {
      "entropy": 0.7276430949568748,
      "epoch": 8.070287055972127,
      "grad_norm": 0.24197372794151306,
      "learning_rate": 4.7851790440007366e-05,
      "loss": 1.1147,
      "mean_token_accuracy": 0.7819822028279304,
      "num_tokens": 1799537476.0,
      "step": 26640
    },
    {
      "entropy": 0.7098560944199562,
      "epoch": 8.073316670453686,
      "grad_norm": 0.256600022315979,
      "learning_rate": 4.78184282677602e-05,
      "loss": 1.1021,
      "mean_token_accuracy": 0.7854956075549125,
      "num_tokens": 1800219877.0,
      "step": 26650
    },
    {
      "entropy": 0.7111981973052025,
      "epoch": 8.076346284935243,
      "grad_norm": 0.2502190172672272,
      "learning_rate": 4.778506706860317e-05,
      "loss": 1.1075,
      "mean_token_accuracy": 0.7847366347908974,
      "num_tokens": 1800897818.0,
      "step": 26660
    },
    {
      "entropy": 0.7186481073498726,
      "epoch": 8.0793758994168,
      "grad_norm": 0.2546496093273163,
      "learning_rate": 4.775170685741702e-05,
      "loss": 1.1116,
      "mean_token_accuracy": 0.7852939248085022,
      "num_tokens": 1801578318.0,
      "step": 26670
    },
    {
      "entropy": 0.7206821799278259,
      "epoch": 8.082405513898356,
      "grad_norm": 0.24004603922367096,
      "learning_rate": 4.771834764908208e-05,
      "loss": 1.1161,
      "mean_token_accuracy": 0.7796136900782585,
      "num_tokens": 1802240373.0,
      "step": 26680
    },
    {
      "entropy": 0.7167475149035454,
      "epoch": 8.085435128379913,
      "grad_norm": 0.25139474868774414,
      "learning_rate": 4.768498945847823e-05,
      "loss": 1.1089,
      "mean_token_accuracy": 0.7803749486804008,
      "num_tokens": 1802899947.0,
      "step": 26690
    },
    {
      "entropy": 0.720377242565155,
      "epoch": 8.08846474286147,
      "grad_norm": 0.25136709213256836,
      "learning_rate": 4.76516323004849e-05,
      "loss": 1.1038,
      "mean_token_accuracy": 0.7825200736522675,
      "num_tokens": 1803572374.0,
      "step": 26700
    },
    {
      "entropy": 0.7177918121218682,
      "epoch": 8.091494357343029,
      "grad_norm": 0.2568497955799103,
      "learning_rate": 4.761827618998103e-05,
      "loss": 1.1117,
      "mean_token_accuracy": 0.7785768941044807,
      "num_tokens": 1804241280.0,
      "step": 26710
    },
    {
      "entropy": 0.7282487601041794,
      "epoch": 8.094523971824586,
      "grad_norm": 0.2670879364013672,
      "learning_rate": 4.758492114184513e-05,
      "loss": 1.1279,
      "mean_token_accuracy": 0.7746492862701416,
      "num_tokens": 1804897799.0,
      "step": 26720
    },
    {
      "entropy": 0.7177425503730774,
      "epoch": 8.097553586306143,
      "grad_norm": 0.2556954622268677,
      "learning_rate": 4.755156717095523e-05,
      "loss": 1.1101,
      "mean_token_accuracy": 0.7844574332237244,
      "num_tokens": 1805583806.0,
      "step": 26730
    },
    {
      "entropy": 0.7086010038852691,
      "epoch": 8.1005832007877,
      "grad_norm": 0.2505972385406494,
      "learning_rate": 4.751821429218883e-05,
      "loss": 1.1118,
      "mean_token_accuracy": 0.7858968749642372,
      "num_tokens": 1806261822.0,
      "step": 26740
    },
    {
      "entropy": 0.7242101460695267,
      "epoch": 8.103612815269257,
      "grad_norm": 0.24488812685012817,
      "learning_rate": 4.748486252042302e-05,
      "loss": 1.1112,
      "mean_token_accuracy": 0.777028651535511,
      "num_tokens": 1806926942.0,
      "step": 26750
    },
    {
      "entropy": 0.7144681587815285,
      "epoch": 8.106642429750814,
      "grad_norm": 0.23839133977890015,
      "learning_rate": 4.7451511870534345e-05,
      "loss": 1.1051,
      "mean_token_accuracy": 0.7829489514231682,
      "num_tokens": 1807600572.0,
      "step": 26760
    },
    {
      "entropy": 0.7107941642403602,
      "epoch": 8.109672044232372,
      "grad_norm": 0.2450329214334488,
      "learning_rate": 4.7418162357398866e-05,
      "loss": 1.1042,
      "mean_token_accuracy": 0.7836286336183548,
      "num_tokens": 1808275883.0,
      "step": 26770
    },
    {
      "entropy": 0.7249545127153396,
      "epoch": 8.11270165871393,
      "grad_norm": 0.25030460953712463,
      "learning_rate": 4.7384813995892126e-05,
      "loss": 1.1171,
      "mean_token_accuracy": 0.7797255769371987,
      "num_tokens": 1808949615.0,
      "step": 26780
    },
    {
      "entropy": 0.7185983180999755,
      "epoch": 8.115731273195486,
      "grad_norm": 0.2522261142730713,
      "learning_rate": 4.7351466800889183e-05,
      "loss": 1.1041,
      "mean_token_accuracy": 0.7817625656723977,
      "num_tokens": 1809619652.0,
      "step": 26790
    },
    {
      "entropy": 0.7124221190810204,
      "epoch": 8.118760887677043,
      "grad_norm": 0.2514772415161133,
      "learning_rate": 4.7318120787264525e-05,
      "loss": 1.1108,
      "mean_token_accuracy": 0.7823670014739037,
      "num_tokens": 1810285799.0,
      "step": 26800
    },
    {
      "entropy": 0.7199279174208641,
      "epoch": 8.1217905021586,
      "grad_norm": 0.2518217861652374,
      "learning_rate": 4.728477596989216e-05,
      "loss": 1.1162,
      "mean_token_accuracy": 0.7802311480045319,
      "num_tokens": 1810962658.0,
      "step": 26810
    },
    {
      "entropy": 0.7195024460554122,
      "epoch": 8.124820116640157,
      "grad_norm": 0.23763404786586761,
      "learning_rate": 4.725143236364554e-05,
      "loss": 1.1097,
      "mean_token_accuracy": 0.7863524854183197,
      "num_tokens": 1811660178.0,
      "step": 26820
    },
    {
      "entropy": 0.7217857331037522,
      "epoch": 8.127849731121715,
      "grad_norm": 0.26382240653038025,
      "learning_rate": 4.721808998339759e-05,
      "loss": 1.1144,
      "mean_token_accuracy": 0.7774293676018715,
      "num_tokens": 1812315947.0,
      "step": 26830
    },
    {
      "entropy": 0.708995920419693,
      "epoch": 8.130879345603272,
      "grad_norm": 0.25705721974372864,
      "learning_rate": 4.7184748844020695e-05,
      "loss": 1.1026,
      "mean_token_accuracy": 0.7843075603246689,
      "num_tokens": 1812985285.0,
      "step": 26840
    },
    {
      "entropy": 0.7199554100632668,
      "epoch": 8.13390896008483,
      "grad_norm": 0.24791404604911804,
      "learning_rate": 4.715140896038662e-05,
      "loss": 1.1093,
      "mean_token_accuracy": 0.7802799060940743,
      "num_tokens": 1813646439.0,
      "step": 26850
    },
    {
      "entropy": 0.7239656805992126,
      "epoch": 8.136938574566386,
      "grad_norm": 0.2485019564628601,
      "learning_rate": 4.7118070347366657e-05,
      "loss": 1.1183,
      "mean_token_accuracy": 0.7817316979169846,
      "num_tokens": 1814315361.0,
      "step": 26860
    },
    {
      "entropy": 0.7147956863045692,
      "epoch": 8.139968189047943,
      "grad_norm": 0.24146954715251923,
      "learning_rate": 4.708473301983149e-05,
      "loss": 1.1049,
      "mean_token_accuracy": 0.7799753293395042,
      "num_tokens": 1814981719.0,
      "step": 26870
    },
    {
      "entropy": 0.7090666785836219,
      "epoch": 8.1429978035295,
      "grad_norm": 0.2604823112487793,
      "learning_rate": 4.7051396992651223e-05,
      "loss": 1.1059,
      "mean_token_accuracy": 0.7828262433409691,
      "num_tokens": 1815642105.0,
      "step": 26880
    },
    {
      "entropy": 0.7162450239062309,
      "epoch": 8.146027418011059,
      "grad_norm": 0.24658261239528656,
      "learning_rate": 4.701806228069541e-05,
      "loss": 1.1113,
      "mean_token_accuracy": 0.7840412139892579,
      "num_tokens": 1816311172.0,
      "step": 26890
    },
    {
      "entropy": 0.7098590478301048,
      "epoch": 8.149057032492616,
      "grad_norm": 0.26275911927223206,
      "learning_rate": 4.698472889883298e-05,
      "loss": 1.1027,
      "mean_token_accuracy": 0.7864038065075875,
      "num_tokens": 1816980224.0,
      "step": 26900
    },
    {
      "entropy": 0.7219566449522972,
      "epoch": 8.152086646974173,
      "grad_norm": 0.25178688764572144,
      "learning_rate": 4.695139686193232e-05,
      "loss": 1.1089,
      "mean_token_accuracy": 0.781039121747017,
      "num_tokens": 1817652354.0,
      "step": 26910
    },
    {
      "entropy": 0.7122976735234261,
      "epoch": 8.15511626145573,
      "grad_norm": 0.2587715983390808,
      "learning_rate": 4.6918066184861154e-05,
      "loss": 1.102,
      "mean_token_accuracy": 0.7840895235538483,
      "num_tokens": 1818332978.0,
      "step": 26920
    },
    {
      "entropy": 0.7142306208610535,
      "epoch": 8.158145875937286,
      "grad_norm": 0.23534339666366577,
      "learning_rate": 4.688473688248664e-05,
      "loss": 1.1136,
      "mean_token_accuracy": 0.7824066042900085,
      "num_tokens": 1819012246.0,
      "step": 26930
    },
    {
      "entropy": 0.7209535852074623,
      "epoch": 8.161175490418843,
      "grad_norm": 0.24771210551261902,
      "learning_rate": 4.6851408969675334e-05,
      "loss": 1.1182,
      "mean_token_accuracy": 0.77869703322649,
      "num_tokens": 1819664571.0,
      "step": 26940
    },
    {
      "entropy": 0.7205091133713722,
      "epoch": 8.164205104900402,
      "grad_norm": 0.24920061230659485,
      "learning_rate": 4.681808246129312e-05,
      "loss": 1.1105,
      "mean_token_accuracy": 0.7805688068270683,
      "num_tokens": 1820336391.0,
      "step": 26950
    },
    {
      "entropy": 0.7151263654232025,
      "epoch": 8.167234719381959,
      "grad_norm": 0.253296822309494,
      "learning_rate": 4.678475737220531e-05,
      "loss": 1.1101,
      "mean_token_accuracy": 0.782950097322464,
      "num_tokens": 1821014808.0,
      "step": 26960
    },
    {
      "entropy": 0.7242428854107856,
      "epoch": 8.170264333863516,
      "grad_norm": 0.25193753838539124,
      "learning_rate": 4.675143371727656e-05,
      "loss": 1.1129,
      "mean_token_accuracy": 0.7782095596194267,
      "num_tokens": 1821669158.0,
      "step": 26970
    },
    {
      "entropy": 0.719010303914547,
      "epoch": 8.173293948345073,
      "grad_norm": 0.2513599991798401,
      "learning_rate": 4.671811151137088e-05,
      "loss": 1.1105,
      "mean_token_accuracy": 0.7793019488453865,
      "num_tokens": 1822330924.0,
      "step": 26980
    },
    {
      "entropy": 0.7160741940140725,
      "epoch": 8.17632356282663,
      "grad_norm": 0.23966023325920105,
      "learning_rate": 4.6684790769351634e-05,
      "loss": 1.1098,
      "mean_token_accuracy": 0.7804221719503402,
      "num_tokens": 1823001974.0,
      "step": 26990
    },
    {
      "entropy": 0.719944280385971,
      "epoch": 8.179353177308187,
      "grad_norm": 0.2498185932636261,
      "learning_rate": 4.6651471506081565e-05,
      "loss": 1.1136,
      "mean_token_accuracy": 0.7820601284503936,
      "num_tokens": 1823675112.0,
      "step": 27000
    },
    {
      "entropy": 0.7151422023773193,
      "epoch": 8.182382791789745,
      "grad_norm": 0.2516774535179138,
      "learning_rate": 4.66181537364227e-05,
      "loss": 1.1009,
      "mean_token_accuracy": 0.7837372854351997,
      "num_tokens": 1824351926.0,
      "step": 27010
    },
    {
      "entropy": 0.7109734356403351,
      "epoch": 8.185412406271302,
      "grad_norm": 0.25796839594841003,
      "learning_rate": 4.6584837475236444e-05,
      "loss": 1.1034,
      "mean_token_accuracy": 0.7837534859776497,
      "num_tokens": 1825025318.0,
      "step": 27020
    },
    {
      "entropy": 0.7213475614786148,
      "epoch": 8.18844202075286,
      "grad_norm": 0.2408466339111328,
      "learning_rate": 4.6551522737383526e-05,
      "loss": 1.116,
      "mean_token_accuracy": 0.7806976407766342,
      "num_tokens": 1825704868.0,
      "step": 27030
    },
    {
      "entropy": 0.723088102042675,
      "epoch": 8.191471635234416,
      "grad_norm": 0.27157673239707947,
      "learning_rate": 4.651820953772398e-05,
      "loss": 1.1085,
      "mean_token_accuracy": 0.7855571120977402,
      "num_tokens": 1826396242.0,
      "step": 27040
    },
    {
      "entropy": 0.7191458314657211,
      "epoch": 8.194501249715973,
      "grad_norm": 0.2600638270378113,
      "learning_rate": 4.6484897891117154e-05,
      "loss": 1.1076,
      "mean_token_accuracy": 0.7821206986904145,
      "num_tokens": 1827074091.0,
      "step": 27050
    },
    {
      "entropy": 0.7229193404316903,
      "epoch": 8.197530864197532,
      "grad_norm": 0.26254802942276,
      "learning_rate": 4.645158781242171e-05,
      "loss": 1.118,
      "mean_token_accuracy": 0.7763147711753845,
      "num_tokens": 1827728690.0,
      "step": 27060
    },
    {
      "entropy": 0.7128867775201797,
      "epoch": 8.200560478679089,
      "grad_norm": 0.26198434829711914,
      "learning_rate": 4.641827931649562e-05,
      "loss": 1.1094,
      "mean_token_accuracy": 0.7814601376652718,
      "num_tokens": 1828391383.0,
      "step": 27070
    },
    {
      "entropy": 0.7261982277035713,
      "epoch": 8.203590093160646,
      "grad_norm": 0.24718758463859558,
      "learning_rate": 4.638497241819612e-05,
      "loss": 1.1202,
      "mean_token_accuracy": 0.7774030700325966,
      "num_tokens": 1829056240.0,
      "step": 27080
    },
    {
      "entropy": 0.720742504298687,
      "epoch": 8.206619707642202,
      "grad_norm": 0.25525447726249695,
      "learning_rate": 4.6351667132379766e-05,
      "loss": 1.1086,
      "mean_token_accuracy": 0.7852376803755761,
      "num_tokens": 1829744599.0,
      "step": 27090
    },
    {
      "entropy": 0.7264063850045204,
      "epoch": 8.20964932212376,
      "grad_norm": 0.2477932572364807,
      "learning_rate": 4.631836347390239e-05,
      "loss": 1.1134,
      "mean_token_accuracy": 0.7852577731013298,
      "num_tokens": 1830441169.0,
      "step": 27100
    },
    {
      "entropy": 0.7133047297596932,
      "epoch": 8.212678936605316,
      "grad_norm": 0.25836342573165894,
      "learning_rate": 4.628506145761907e-05,
      "loss": 1.1009,
      "mean_token_accuracy": 0.7800086751580239,
      "num_tokens": 1831106159.0,
      "step": 27110
    },
    {
      "entropy": 0.7223998859524727,
      "epoch": 8.215708551086875,
      "grad_norm": 0.2607588768005371,
      "learning_rate": 4.6251761098384174e-05,
      "loss": 1.1208,
      "mean_token_accuracy": 0.7797861859202385,
      "num_tokens": 1831777748.0,
      "step": 27120
    },
    {
      "entropy": 0.7217176571488381,
      "epoch": 8.218738165568432,
      "grad_norm": 0.24700438976287842,
      "learning_rate": 4.6218462411051355e-05,
      "loss": 1.1151,
      "mean_token_accuracy": 0.7837042525410652,
      "num_tokens": 1832456356.0,
      "step": 27130
    },
    {
      "entropy": 0.7273529902100563,
      "epoch": 8.221767780049989,
      "grad_norm": 0.2521395683288574,
      "learning_rate": 4.618516541047345e-05,
      "loss": 1.1188,
      "mean_token_accuracy": 0.7798873633146286,
      "num_tokens": 1833133661.0,
      "step": 27140
    },
    {
      "entropy": 0.7186020672321319,
      "epoch": 8.224797394531546,
      "grad_norm": 0.24985063076019287,
      "learning_rate": 4.615187011150262e-05,
      "loss": 1.1138,
      "mean_token_accuracy": 0.7808549642562866,
      "num_tokens": 1833796802.0,
      "step": 27150
    },
    {
      "entropy": 0.712014751136303,
      "epoch": 8.227827009013103,
      "grad_norm": 0.2520310878753662,
      "learning_rate": 4.611857652899022e-05,
      "loss": 1.1043,
      "mean_token_accuracy": 0.7841321885585785,
      "num_tokens": 1834466999.0,
      "step": 27160
    },
    {
      "entropy": 0.7179373130202293,
      "epoch": 8.23085662349466,
      "grad_norm": 0.25570186972618103,
      "learning_rate": 4.608528467778687e-05,
      "loss": 1.1151,
      "mean_token_accuracy": 0.7850959330797196,
      "num_tokens": 1835139694.0,
      "step": 27170
    },
    {
      "entropy": 0.7286332011222839,
      "epoch": 8.233886237976218,
      "grad_norm": 0.2549498379230499,
      "learning_rate": 4.605199457274238e-05,
      "loss": 1.1162,
      "mean_token_accuracy": 0.7740892961621284,
      "num_tokens": 1835802648.0,
      "step": 27180
    },
    {
      "entropy": 0.721484349668026,
      "epoch": 8.236915852457775,
      "grad_norm": 0.24085596203804016,
      "learning_rate": 4.6018706228705825e-05,
      "loss": 1.1192,
      "mean_token_accuracy": 0.7827752307057381,
      "num_tokens": 1836477469.0,
      "step": 27190
    },
    {
      "entropy": 0.7150719866156579,
      "epoch": 8.239945466939332,
      "grad_norm": 0.2526509463787079,
      "learning_rate": 4.5985419660525434e-05,
      "loss": 1.1049,
      "mean_token_accuracy": 0.7816643178462982,
      "num_tokens": 1837145950.0,
      "step": 27200
    },
    {
      "entropy": 0.7226071417331695,
      "epoch": 8.242975081420889,
      "grad_norm": 0.2403123676776886,
      "learning_rate": 4.595213488304872e-05,
      "loss": 1.1092,
      "mean_token_accuracy": 0.7827942579984665,
      "num_tokens": 1837827447.0,
      "step": 27210
    },
    {
      "entropy": 0.7206570655107498,
      "epoch": 8.246004695902446,
      "grad_norm": 0.26418620347976685,
      "learning_rate": 4.591885191112233e-05,
      "loss": 1.1075,
      "mean_token_accuracy": 0.7818023473024368,
      "num_tokens": 1838509638.0,
      "step": 27220
    },
    {
      "entropy": 0.723764368891716,
      "epoch": 8.249034310384003,
      "grad_norm": 0.24933916330337524,
      "learning_rate": 4.588557075959214e-05,
      "loss": 1.1025,
      "mean_token_accuracy": 0.7819537103176117,
      "num_tokens": 1839193441.0,
      "step": 27230
    },
    {
      "entropy": 0.7162736654281616,
      "epoch": 8.252063924865562,
      "grad_norm": 0.23663337528705597,
      "learning_rate": 4.585229144330323e-05,
      "loss": 1.1178,
      "mean_token_accuracy": 0.778569583594799,
      "num_tokens": 1839852616.0,
      "step": 27240
    },
    {
      "entropy": 0.7260021716356277,
      "epoch": 8.255093539347119,
      "grad_norm": 0.2518142759799957,
      "learning_rate": 4.581901397709983e-05,
      "loss": 1.114,
      "mean_token_accuracy": 0.7796929970383644,
      "num_tokens": 1840528113.0,
      "step": 27250
    },
    {
      "entropy": 0.7233222261071205,
      "epoch": 8.258123153828675,
      "grad_norm": 0.2555762827396393,
      "learning_rate": 4.578573837582534e-05,
      "loss": 1.1125,
      "mean_token_accuracy": 0.7773929163813591,
      "num_tokens": 1841183735.0,
      "step": 27260
    },
    {
      "entropy": 0.723837748169899,
      "epoch": 8.261152768310232,
      "grad_norm": 0.2641092538833618,
      "learning_rate": 4.5752464654322355e-05,
      "loss": 1.1105,
      "mean_token_accuracy": 0.7805815920233726,
      "num_tokens": 1841862928.0,
      "step": 27270
    },
    {
      "entropy": 0.715344388782978,
      "epoch": 8.26418238279179,
      "grad_norm": 0.25379928946495056,
      "learning_rate": 4.5719192827432624e-05,
      "loss": 1.1122,
      "mean_token_accuracy": 0.7848491683602333,
      "num_tokens": 1842528590.0,
      "step": 27280
    },
    {
      "entropy": 0.716085834801197,
      "epoch": 8.267211997273346,
      "grad_norm": 0.256211519241333,
      "learning_rate": 4.5685922909997036e-05,
      "loss": 1.1039,
      "mean_token_accuracy": 0.7850614383816719,
      "num_tokens": 1843213297.0,
      "step": 27290
    },
    {
      "entropy": 0.7137120842933655,
      "epoch": 8.270241611754905,
      "grad_norm": 0.24229760468006134,
      "learning_rate": 4.565265491685564e-05,
      "loss": 1.1116,
      "mean_token_accuracy": 0.7800800368189812,
      "num_tokens": 1843875109.0,
      "step": 27300
    },
    {
      "entropy": 0.7221919119358062,
      "epoch": 8.273271226236462,
      "grad_norm": 0.25407901406288147,
      "learning_rate": 4.561938886284765e-05,
      "loss": 1.1129,
      "mean_token_accuracy": 0.7793138191103935,
      "num_tokens": 1844545036.0,
      "step": 27310
    },
    {
      "entropy": 0.7136831447482109,
      "epoch": 8.276300840718019,
      "grad_norm": 0.24182777106761932,
      "learning_rate": 4.5586124762811336e-05,
      "loss": 1.1104,
      "mean_token_accuracy": 0.7838802829384803,
      "num_tokens": 1845219829.0,
      "step": 27320
    },
    {
      "entropy": 0.7143359944224358,
      "epoch": 8.279330455199576,
      "grad_norm": 0.2557224929332733,
      "learning_rate": 4.555286263158419e-05,
      "loss": 1.1084,
      "mean_token_accuracy": 0.7849928557872772,
      "num_tokens": 1845892685.0,
      "step": 27330
    },
    {
      "entropy": 0.7223512455821037,
      "epoch": 8.282360069681133,
      "grad_norm": 0.24389885365962982,
      "learning_rate": 4.55196024840028e-05,
      "loss": 1.1102,
      "mean_token_accuracy": 0.7804371014237403,
      "num_tokens": 1846564946.0,
      "step": 27340
    },
    {
      "entropy": 0.7124507904052735,
      "epoch": 8.28538968416269,
      "grad_norm": 0.247615784406662,
      "learning_rate": 4.548634433490281e-05,
      "loss": 1.1105,
      "mean_token_accuracy": 0.7814389169216156,
      "num_tokens": 1847230526.0,
      "step": 27350
    },
    {
      "entropy": 0.7159087106585502,
      "epoch": 8.288419298644248,
      "grad_norm": 0.2521499991416931,
      "learning_rate": 4.5453088199119056e-05,
      "loss": 1.1024,
      "mean_token_accuracy": 0.7813684791326523,
      "num_tokens": 1847906579.0,
      "step": 27360
    },
    {
      "entropy": 0.7156479343771934,
      "epoch": 8.291448913125805,
      "grad_norm": 0.24656377732753754,
      "learning_rate": 4.5419834091485414e-05,
      "loss": 1.1032,
      "mean_token_accuracy": 0.7859241157770157,
      "num_tokens": 1848596034.0,
      "step": 27370
    },
    {
      "entropy": 0.7168033167719841,
      "epoch": 8.294478527607362,
      "grad_norm": 0.24311594665050507,
      "learning_rate": 4.5386582026834906e-05,
      "loss": 1.1113,
      "mean_token_accuracy": 0.7866822898387908,
      "num_tokens": 1849278634.0,
      "step": 27380
    },
    {
      "entropy": 0.7253716468811036,
      "epoch": 8.297508142088919,
      "grad_norm": 0.2668185830116272,
      "learning_rate": 4.535333201999959e-05,
      "loss": 1.1239,
      "mean_token_accuracy": 0.7788155362010002,
      "num_tokens": 1849954261.0,
      "step": 27390
    },
    {
      "entropy": 0.7032744482159614,
      "epoch": 8.300537756570476,
      "grad_norm": 0.24929794669151306,
      "learning_rate": 4.532008408581064e-05,
      "loss": 1.0972,
      "mean_token_accuracy": 0.7904573962092399,
      "num_tokens": 1850653969.0,
      "step": 27400
    },
    {
      "entropy": 0.7162431120872498,
      "epoch": 8.303567371052033,
      "grad_norm": 0.2612493336200714,
      "learning_rate": 4.5286838239098325e-05,
      "loss": 1.1107,
      "mean_token_accuracy": 0.7826158821582794,
      "num_tokens": 1851324108.0,
      "step": 27410
    },
    {
      "entropy": 0.7205453425645828,
      "epoch": 8.306596985533591,
      "grad_norm": 0.25268685817718506,
      "learning_rate": 4.525359449469191e-05,
      "loss": 1.1149,
      "mean_token_accuracy": 0.7815184578299522,
      "num_tokens": 1852000998.0,
      "step": 27420
    },
    {
      "entropy": 0.7282377079129219,
      "epoch": 8.309626600015148,
      "grad_norm": 0.2584766447544098,
      "learning_rate": 4.522035286741979e-05,
      "loss": 1.1175,
      "mean_token_accuracy": 0.7803824990987778,
      "num_tokens": 1852679929.0,
      "step": 27430
    },
    {
      "entropy": 0.7233649790287018,
      "epoch": 8.312656214496705,
      "grad_norm": 0.253195196390152,
      "learning_rate": 4.518711337210941e-05,
      "loss": 1.1215,
      "mean_token_accuracy": 0.7783728897571563,
      "num_tokens": 1853351086.0,
      "step": 27440
    },
    {
      "entropy": 0.7166319891810418,
      "epoch": 8.315685828978262,
      "grad_norm": 0.25387871265411377,
      "learning_rate": 4.5153876023587214e-05,
      "loss": 1.1096,
      "mean_token_accuracy": 0.7807391300797463,
      "num_tokens": 1854019055.0,
      "step": 27450
    },
    {
      "entropy": 0.7199093952775002,
      "epoch": 8.31871544345982,
      "grad_norm": 0.2522037625312805,
      "learning_rate": 4.512064083667875e-05,
      "loss": 1.1129,
      "mean_token_accuracy": 0.7819029286503791,
      "num_tokens": 1854690897.0,
      "step": 27460
    },
    {
      "entropy": 0.7243819639086724,
      "epoch": 8.321745057941378,
      "grad_norm": 0.23504793643951416,
      "learning_rate": 4.508740782620855e-05,
      "loss": 1.1181,
      "mean_token_accuracy": 0.7792690947651864,
      "num_tokens": 1855371240.0,
      "step": 27470
    },
    {
      "entropy": 0.7095865100622177,
      "epoch": 8.324774672422935,
      "grad_norm": 0.24862954020500183,
      "learning_rate": 4.505417700700023e-05,
      "loss": 1.1072,
      "mean_token_accuracy": 0.7814011991024017,
      "num_tokens": 1856033287.0,
      "step": 27480
    },
    {
      "entropy": 0.7154641449451447,
      "epoch": 8.327804286904492,
      "grad_norm": 0.2469140589237213,
      "learning_rate": 4.502094839387637e-05,
      "loss": 1.1043,
      "mean_token_accuracy": 0.7874842047691345,
      "num_tokens": 1856721284.0,
      "step": 27490
    },
    {
      "entropy": 0.7299807772040368,
      "epoch": 8.330833901386049,
      "grad_norm": 0.2515232563018799,
      "learning_rate": 4.49877220016586e-05,
      "loss": 1.1233,
      "mean_token_accuracy": 0.7815038755536079,
      "num_tokens": 1857414157.0,
      "step": 27500
    },
    {
      "entropy": 0.7260418713092804,
      "epoch": 8.333863515867606,
      "grad_norm": 0.25561216473579407,
      "learning_rate": 4.4954497845167574e-05,
      "loss": 1.1148,
      "mean_token_accuracy": 0.7742961138486862,
      "num_tokens": 1858082085.0,
      "step": 27510
    },
    {
      "entropy": 0.7206007733941078,
      "epoch": 8.336893130349162,
      "grad_norm": 0.2394639402627945,
      "learning_rate": 4.49212759392229e-05,
      "loss": 1.1089,
      "mean_token_accuracy": 0.7829590395092965,
      "num_tokens": 1858766844.0,
      "step": 27520
    },
    {
      "entropy": 0.7164310291409492,
      "epoch": 8.33992274483072,
      "grad_norm": 0.2688215970993042,
      "learning_rate": 4.4888056298643225e-05,
      "loss": 1.1143,
      "mean_token_accuracy": 0.779485197365284,
      "num_tokens": 1859431672.0,
      "step": 27530
    },
    {
      "entropy": 0.7138465449213982,
      "epoch": 8.342952359312278,
      "grad_norm": 0.25172656774520874,
      "learning_rate": 4.485483893824618e-05,
      "loss": 1.104,
      "mean_token_accuracy": 0.7824663892388344,
      "num_tokens": 1860110045.0,
      "step": 27540
    },
    {
      "entropy": 0.7216217398643494,
      "epoch": 8.345981973793835,
      "grad_norm": 0.25880149006843567,
      "learning_rate": 4.482162387284835e-05,
      "loss": 1.1136,
      "mean_token_accuracy": 0.7828172624111176,
      "num_tokens": 1860800566.0,
      "step": 27550
    },
    {
      "entropy": 0.7184495881199837,
      "epoch": 8.349011588275392,
      "grad_norm": 0.23833420872688293,
      "learning_rate": 4.4788411117265304e-05,
      "loss": 1.1101,
      "mean_token_accuracy": 0.7817671433091163,
      "num_tokens": 1861474775.0,
      "step": 27560
    },
    {
      "entropy": 0.7274298757314682,
      "epoch": 8.352041202756949,
      "grad_norm": 0.2440783977508545,
      "learning_rate": 4.475520068631162e-05,
      "loss": 1.1198,
      "mean_token_accuracy": 0.7835265845060349,
      "num_tokens": 1862156443.0,
      "step": 27570
    },
    {
      "entropy": 0.7170529827475548,
      "epoch": 8.355070817238506,
      "grad_norm": 0.26175349950790405,
      "learning_rate": 4.4721992594800794e-05,
      "loss": 1.1095,
      "mean_token_accuracy": 0.778462964296341,
      "num_tokens": 1862828400.0,
      "step": 27580
    },
    {
      "entropy": 0.7256954088807106,
      "epoch": 8.358100431720064,
      "grad_norm": 0.2389829009771347,
      "learning_rate": 4.468878685754532e-05,
      "loss": 1.1188,
      "mean_token_accuracy": 0.782408845424652,
      "num_tokens": 1863513803.0,
      "step": 27590
    },
    {
      "entropy": 0.717227679491043,
      "epoch": 8.361130046201621,
      "grad_norm": 0.2596455216407776,
      "learning_rate": 4.465558348935659e-05,
      "loss": 1.1059,
      "mean_token_accuracy": 0.7873851031064987,
      "num_tokens": 1864210837.0,
      "step": 27600
    },
    {
      "entropy": 0.7271547868847847,
      "epoch": 8.364159660683178,
      "grad_norm": 0.24676461517810822,
      "learning_rate": 4.462238250504498e-05,
      "loss": 1.1126,
      "mean_token_accuracy": 0.7838033780455589,
      "num_tokens": 1864900256.0,
      "step": 27610
    },
    {
      "entropy": 0.7318004697561264,
      "epoch": 8.367189275164735,
      "grad_norm": 0.25214236974716187,
      "learning_rate": 4.45891839194198e-05,
      "loss": 1.119,
      "mean_token_accuracy": 0.7798203274607658,
      "num_tokens": 1865577911.0,
      "step": 27620
    },
    {
      "entropy": 0.7154213026165962,
      "epoch": 8.370218889646292,
      "grad_norm": 0.25352394580841064,
      "learning_rate": 4.455598774728925e-05,
      "loss": 1.1134,
      "mean_token_accuracy": 0.7794664621353149,
      "num_tokens": 1866233536.0,
      "step": 27630
    },
    {
      "entropy": 0.7112863630056381,
      "epoch": 8.373248504127849,
      "grad_norm": 0.25486281514167786,
      "learning_rate": 4.452279400346051e-05,
      "loss": 1.1094,
      "mean_token_accuracy": 0.7859494358301162,
      "num_tokens": 1866918882.0,
      "step": 27640
    },
    {
      "entropy": 0.7169707760214805,
      "epoch": 8.376278118609408,
      "grad_norm": 0.2506760358810425,
      "learning_rate": 4.448960270273965e-05,
      "loss": 1.1064,
      "mean_token_accuracy": 0.7848715350031853,
      "num_tokens": 1867605528.0,
      "step": 27650
    },
    {
      "entropy": 0.7194483205676079,
      "epoch": 8.379307733090965,
      "grad_norm": 0.25176456570625305,
      "learning_rate": 4.445641385993163e-05,
      "loss": 1.1156,
      "mean_token_accuracy": 0.7766473367810249,
      "num_tokens": 1868263380.0,
      "step": 27660
    },
    {
      "entropy": 0.7132114320993423,
      "epoch": 8.382337347572522,
      "grad_norm": 0.26137015223503113,
      "learning_rate": 4.442322748984036e-05,
      "loss": 1.1134,
      "mean_token_accuracy": 0.7818466663360596,
      "num_tokens": 1868927424.0,
      "step": 27670
    },
    {
      "entropy": 0.7198499768972397,
      "epoch": 8.385366962054078,
      "grad_norm": 0.2476302832365036,
      "learning_rate": 4.4390043607268594e-05,
      "loss": 1.114,
      "mean_token_accuracy": 0.7790758609771729,
      "num_tokens": 1869578969.0,
      "step": 27680
    },
    {
      "entropy": 0.7200073435902595,
      "epoch": 8.388396576535635,
      "grad_norm": 0.25056976079940796,
      "learning_rate": 4.4356862227018035e-05,
      "loss": 1.1149,
      "mean_token_accuracy": 0.782424296438694,
      "num_tokens": 1870258507.0,
      "step": 27690
    },
    {
      "entropy": 0.7121904641389847,
      "epoch": 8.391426191017192,
      "grad_norm": 0.24387040734291077,
      "learning_rate": 4.4323683363889214e-05,
      "loss": 1.1058,
      "mean_token_accuracy": 0.7839852780103683,
      "num_tokens": 1870927719.0,
      "step": 27700
    },
    {
      "entropy": 0.7183660760521888,
      "epoch": 8.394455805498751,
      "grad_norm": 0.2422250360250473,
      "learning_rate": 4.4290507032681565e-05,
      "loss": 1.1123,
      "mean_token_accuracy": 0.7803048029541969,
      "num_tokens": 1871592721.0,
      "step": 27710
    },
    {
      "entropy": 0.7177755072712898,
      "epoch": 8.397485419980308,
      "grad_norm": 0.260509192943573,
      "learning_rate": 4.425733324819341e-05,
      "loss": 1.1144,
      "mean_token_accuracy": 0.7785208135843277,
      "num_tokens": 1872258357.0,
      "step": 27720
    },
    {
      "entropy": 0.7150320529937744,
      "epoch": 8.400515034461865,
      "grad_norm": 0.25201788544654846,
      "learning_rate": 4.42241620252219e-05,
      "loss": 1.1022,
      "mean_token_accuracy": 0.7828321620821953,
      "num_tokens": 1872930446.0,
      "step": 27730
    },
    {
      "entropy": 0.7162398502230645,
      "epoch": 8.403544648943422,
      "grad_norm": 0.25627806782722473,
      "learning_rate": 4.419099337856306e-05,
      "loss": 1.0997,
      "mean_token_accuracy": 0.7854035153985024,
      "num_tokens": 1873625344.0,
      "step": 27740
    },
    {
      "entropy": 0.7124519675970078,
      "epoch": 8.406574263424979,
      "grad_norm": 0.2569003105163574,
      "learning_rate": 4.415782732301177e-05,
      "loss": 1.106,
      "mean_token_accuracy": 0.7857031732797622,
      "num_tokens": 1874312978.0,
      "step": 27750
    },
    {
      "entropy": 0.721549491584301,
      "epoch": 8.409603877906536,
      "grad_norm": 0.24334709346294403,
      "learning_rate": 4.4124663873361753e-05,
      "loss": 1.1144,
      "mean_token_accuracy": 0.7795088529586792,
      "num_tokens": 1874979206.0,
      "step": 27760
    },
    {
      "entropy": 0.7219868063926697,
      "epoch": 8.412633492388094,
      "grad_norm": 0.23988959193229675,
      "learning_rate": 4.409150304440556e-05,
      "loss": 1.1108,
      "mean_token_accuracy": 0.7811971604824066,
      "num_tokens": 1875650612.0,
      "step": 27770
    },
    {
      "entropy": 0.7256921246647835,
      "epoch": 8.415663106869651,
      "grad_norm": 0.24760156869888306,
      "learning_rate": 4.405834485093459e-05,
      "loss": 1.107,
      "mean_token_accuracy": 0.7802922621369361,
      "num_tokens": 1876327876.0,
      "step": 27780
    },
    {
      "entropy": 0.7178578466176987,
      "epoch": 8.418692721351208,
      "grad_norm": 0.26577064394950867,
      "learning_rate": 4.402518930773904e-05,
      "loss": 1.1177,
      "mean_token_accuracy": 0.7766996204853058,
      "num_tokens": 1876992197.0,
      "step": 27790
    },
    {
      "entropy": 0.7079323068261146,
      "epoch": 8.421722335832765,
      "grad_norm": 0.26500049233436584,
      "learning_rate": 4.3992036429607955e-05,
      "loss": 1.1013,
      "mean_token_accuracy": 0.787224443256855,
      "num_tokens": 1877672478.0,
      "step": 27800
    },
    {
      "entropy": 0.7205834656953811,
      "epoch": 8.424751950314322,
      "grad_norm": 0.2477947175502777,
      "learning_rate": 4.395888623132916e-05,
      "loss": 1.1102,
      "mean_token_accuracy": 0.7847822561860085,
      "num_tokens": 1878354158.0,
      "step": 27810
    },
    {
      "entropy": 0.7100614115595818,
      "epoch": 8.427781564795879,
      "grad_norm": 0.23827891051769257,
      "learning_rate": 4.392573872768933e-05,
      "loss": 1.1092,
      "mean_token_accuracy": 0.7817847982048989,
      "num_tokens": 1879020492.0,
      "step": 27820
    },
    {
      "entropy": 0.7129092320799828,
      "epoch": 8.430811179277438,
      "grad_norm": 0.26287147402763367,
      "learning_rate": 4.389259393347389e-05,
      "loss": 1.1055,
      "mean_token_accuracy": 0.7855137750506401,
      "num_tokens": 1879707040.0,
      "step": 27830
    },
    {
      "entropy": 0.7174157917499542,
      "epoch": 8.433840793758995,
      "grad_norm": 0.26285824179649353,
      "learning_rate": 4.385945186346707e-05,
      "loss": 1.1052,
      "mean_token_accuracy": 0.7869385465979576,
      "num_tokens": 1880397214.0,
      "step": 27840
    },
    {
      "entropy": 0.7163793385028839,
      "epoch": 8.436870408240551,
      "grad_norm": 0.2453935742378235,
      "learning_rate": 4.3826312532451914e-05,
      "loss": 1.1104,
      "mean_token_accuracy": 0.7838813096284867,
      "num_tokens": 1881079728.0,
      "step": 27850
    },
    {
      "entropy": 0.7231308802962303,
      "epoch": 8.439900022722108,
      "grad_norm": 0.25110098719596863,
      "learning_rate": 4.37931759552102e-05,
      "loss": 1.1121,
      "mean_token_accuracy": 0.781943641602993,
      "num_tokens": 1881758803.0,
      "step": 27860
    },
    {
      "entropy": 0.716950623691082,
      "epoch": 8.442929637203665,
      "grad_norm": 0.2400575876235962,
      "learning_rate": 4.37600421465225e-05,
      "loss": 1.1116,
      "mean_token_accuracy": 0.7835470572113991,
      "num_tokens": 1882441320.0,
      "step": 27870
    },
    {
      "entropy": 0.7130150616168975,
      "epoch": 8.445959251685222,
      "grad_norm": 0.2530190348625183,
      "learning_rate": 4.372691112116817e-05,
      "loss": 1.109,
      "mean_token_accuracy": 0.7860501155257225,
      "num_tokens": 1883125135.0,
      "step": 27880
    },
    {
      "entropy": 0.7143324956297874,
      "epoch": 8.448988866166781,
      "grad_norm": 0.2560359537601471,
      "learning_rate": 4.3693782893925304e-05,
      "loss": 1.1038,
      "mean_token_accuracy": 0.7861661210656166,
      "num_tokens": 1883817683.0,
      "step": 27890
    },
    {
      "entropy": 0.7231035277247428,
      "epoch": 8.452018480648338,
      "grad_norm": 0.2547987997531891,
      "learning_rate": 4.366065747957072e-05,
      "loss": 1.1156,
      "mean_token_accuracy": 0.781479986011982,
      "num_tokens": 1884489814.0,
      "step": 27900
    },
    {
      "entropy": 0.7147853136062622,
      "epoch": 8.455048095129895,
      "grad_norm": 0.2504449784755707,
      "learning_rate": 4.362753489288003e-05,
      "loss": 1.1115,
      "mean_token_accuracy": 0.782762411236763,
      "num_tokens": 1885172135.0,
      "step": 27910
    },
    {
      "entropy": 0.7212734669446945,
      "epoch": 8.458077709611452,
      "grad_norm": 0.2554200291633606,
      "learning_rate": 4.3594415148627555e-05,
      "loss": 1.1169,
      "mean_token_accuracy": 0.7784308612346649,
      "num_tokens": 1885831076.0,
      "step": 27920
    },
    {
      "entropy": 0.7203407719731331,
      "epoch": 8.461107324093009,
      "grad_norm": 0.2583375573158264,
      "learning_rate": 4.356129826158637e-05,
      "loss": 1.1164,
      "mean_token_accuracy": 0.778656542301178,
      "num_tokens": 1886491048.0,
      "step": 27930
    },
    {
      "entropy": 0.7222388982772827,
      "epoch": 8.464136938574566,
      "grad_norm": 0.24358335137367249,
      "learning_rate": 4.352818424652826e-05,
      "loss": 1.1128,
      "mean_token_accuracy": 0.7837407380342484,
      "num_tokens": 1887175296.0,
      "step": 27940
    },
    {
      "entropy": 0.719977830350399,
      "epoch": 8.467166553056124,
      "grad_norm": 0.2525087594985962,
      "learning_rate": 4.3495073118223725e-05,
      "loss": 1.1109,
      "mean_token_accuracy": 0.7815982520580291,
      "num_tokens": 1887854011.0,
      "step": 27950
    },
    {
      "entropy": 0.7202391773462296,
      "epoch": 8.470196167537681,
      "grad_norm": 0.2515532374382019,
      "learning_rate": 4.346196489144199e-05,
      "loss": 1.1081,
      "mean_token_accuracy": 0.7765378817915917,
      "num_tokens": 1888523269.0,
      "step": 27960
    },
    {
      "entropy": 0.7178052544593811,
      "epoch": 8.473225782019238,
      "grad_norm": 0.25526100397109985,
      "learning_rate": 4.342885958095098e-05,
      "loss": 1.1044,
      "mean_token_accuracy": 0.7809532567858696,
      "num_tokens": 1889197502.0,
      "step": 27970
    },
    {
      "entropy": 0.7176271915435791,
      "epoch": 8.476255396500795,
      "grad_norm": 0.2526986598968506,
      "learning_rate": 4.3395757201517315e-05,
      "loss": 1.1087,
      "mean_token_accuracy": 0.7788459807634354,
      "num_tokens": 1889859446.0,
      "step": 27980
    },
    {
      "entropy": 0.7217741340398789,
      "epoch": 8.479285010982352,
      "grad_norm": 0.261422336101532,
      "learning_rate": 4.336265776790633e-05,
      "loss": 1.1097,
      "mean_token_accuracy": 0.7857506960630417,
      "num_tokens": 1890554079.0,
      "step": 27990
    },
    {
      "entropy": 0.7163657307624817,
      "epoch": 8.48231462546391,
      "grad_norm": 0.25897666811943054,
      "learning_rate": 4.3329561294882e-05,
      "loss": 1.1137,
      "mean_token_accuracy": 0.7787162497639656,
      "num_tokens": 1891209107.0,
      "step": 28000
    },
    {
      "entropy": 0.7168700203299523,
      "epoch": 8.485344239945467,
      "grad_norm": 0.24963456392288208,
      "learning_rate": 4.3296467797207025e-05,
      "loss": 1.1172,
      "mean_token_accuracy": 0.7804996103048325,
      "num_tokens": 1891877021.0,
      "step": 28010
    },
    {
      "entropy": 0.7277984783053398,
      "epoch": 8.488373854427024,
      "grad_norm": 0.2574210464954376,
      "learning_rate": 4.3263377289642754e-05,
      "loss": 1.1115,
      "mean_token_accuracy": 0.7805963233113289,
      "num_tokens": 1892555296.0,
      "step": 28020
    },
    {
      "entropy": 0.7244296565651893,
      "epoch": 8.491403468908581,
      "grad_norm": 0.25558575987815857,
      "learning_rate": 4.323028978694923e-05,
      "loss": 1.1129,
      "mean_token_accuracy": 0.7798798620700836,
      "num_tokens": 1893233523.0,
      "step": 28030
    },
    {
      "entropy": 0.7282684370875359,
      "epoch": 8.494433083390138,
      "grad_norm": 0.2511744201183319,
      "learning_rate": 4.3197205303885105e-05,
      "loss": 1.109,
      "mean_token_accuracy": 0.781240402162075,
      "num_tokens": 1893928697.0,
      "step": 28040
    },
    {
      "entropy": 0.716360405087471,
      "epoch": 8.497462697871695,
      "grad_norm": 0.26084938645362854,
      "learning_rate": 4.3164123855207715e-05,
      "loss": 1.1107,
      "mean_token_accuracy": 0.7831664323806763,
      "num_tokens": 1894606541.0,
      "step": 28050
    },
    {
      "entropy": 0.7177587166428566,
      "epoch": 8.500492312353254,
      "grad_norm": 0.2502937912940979,
      "learning_rate": 4.313104545567307e-05,
      "loss": 1.1035,
      "mean_token_accuracy": 0.7813873678445816,
      "num_tokens": 1895284845.0,
      "step": 28060
    },
    {
      "entropy": 0.7103653758764267,
      "epoch": 8.50352192683481,
      "grad_norm": 0.2538270950317383,
      "learning_rate": 4.309797012003576e-05,
      "loss": 1.1075,
      "mean_token_accuracy": 0.7838470160961151,
      "num_tokens": 1895957404.0,
      "step": 28070
    },
    {
      "entropy": 0.7125907152891159,
      "epoch": 8.506551541316368,
      "grad_norm": 0.2410464733839035,
      "learning_rate": 4.306489786304906e-05,
      "loss": 1.107,
      "mean_token_accuracy": 0.7825109049677849,
      "num_tokens": 1896636351.0,
      "step": 28080
    },
    {
      "entropy": 0.7232971996068954,
      "epoch": 8.509581155797925,
      "grad_norm": 0.24507884681224823,
      "learning_rate": 4.3031828699464846e-05,
      "loss": 1.1247,
      "mean_token_accuracy": 0.778888575732708,
      "num_tokens": 1897302092.0,
      "step": 28090
    },
    {
      "entropy": 0.7203190252184868,
      "epoch": 8.512610770279482,
      "grad_norm": 0.277683824300766,
      "learning_rate": 4.29987626440336e-05,
      "loss": 1.1031,
      "mean_token_accuracy": 0.7816737964749336,
      "num_tokens": 1897983133.0,
      "step": 28100
    },
    {
      "entropy": 0.7236964136362076,
      "epoch": 8.515640384761038,
      "grad_norm": 0.2470213621854782,
      "learning_rate": 4.296569971150446e-05,
      "loss": 1.1197,
      "mean_token_accuracy": 0.7798929691314698,
      "num_tokens": 1898656412.0,
      "step": 28110
    },
    {
      "entropy": 0.7217274710536004,
      "epoch": 8.518669999242597,
      "grad_norm": 0.2536276578903198,
      "learning_rate": 4.293263991662515e-05,
      "loss": 1.1039,
      "mean_token_accuracy": 0.7806472942233086,
      "num_tokens": 1899327501.0,
      "step": 28120
    },
    {
      "entropy": 0.7342067047953605,
      "epoch": 8.521699613724154,
      "grad_norm": 0.2557133734226227,
      "learning_rate": 4.2899583274141975e-05,
      "loss": 1.1191,
      "mean_token_accuracy": 0.7798523887991905,
      "num_tokens": 1900009318.0,
      "step": 28130
    },
    {
      "entropy": 0.7135517805814743,
      "epoch": 8.524729228205711,
      "grad_norm": 0.26016944646835327,
      "learning_rate": 4.2866529798799855e-05,
      "loss": 1.1108,
      "mean_token_accuracy": 0.781361098587513,
      "num_tokens": 1900671691.0,
      "step": 28140
    },
    {
      "entropy": 0.7154786333441734,
      "epoch": 8.527758842687268,
      "grad_norm": 0.2544073164463043,
      "learning_rate": 4.28334795053423e-05,
      "loss": 1.1145,
      "mean_token_accuracy": 0.7790338054299355,
      "num_tokens": 1901329773.0,
      "step": 28150
    },
    {
      "entropy": 0.7150747537612915,
      "epoch": 8.530788457168825,
      "grad_norm": 0.2526049017906189,
      "learning_rate": 4.28004324085114e-05,
      "loss": 1.1056,
      "mean_token_accuracy": 0.7860410287976265,
      "num_tokens": 1902008559.0,
      "step": 28160
    },
    {
      "entropy": 0.7090863332152366,
      "epoch": 8.533818071650382,
      "grad_norm": 0.2356344610452652,
      "learning_rate": 4.27673885230478e-05,
      "loss": 1.1094,
      "mean_token_accuracy": 0.7842254042625427,
      "num_tokens": 1902682622.0,
      "step": 28170
    },
    {
      "entropy": 0.7112705111503601,
      "epoch": 8.53684768613194,
      "grad_norm": 0.25435927510261536,
      "learning_rate": 4.273434786369072e-05,
      "loss": 1.1155,
      "mean_token_accuracy": 0.7843185245990754,
      "num_tokens": 1903355345.0,
      "step": 28180
    },
    {
      "entropy": 0.7159225881099701,
      "epoch": 8.539877300613497,
      "grad_norm": 0.25240522623062134,
      "learning_rate": 4.270131044517798e-05,
      "loss": 1.1053,
      "mean_token_accuracy": 0.7823803201317787,
      "num_tokens": 1904041698.0,
      "step": 28190
    },
    {
      "entropy": 0.7105033457279205,
      "epoch": 8.542906915095054,
      "grad_norm": 0.2581464946269989,
      "learning_rate": 4.2668276282245886e-05,
      "loss": 1.1008,
      "mean_token_accuracy": 0.7827621147036552,
      "num_tokens": 1904714535.0,
      "step": 28200
    },
    {
      "entropy": 0.7183787137269974,
      "epoch": 8.545936529576611,
      "grad_norm": 0.26322516798973083,
      "learning_rate": 4.263524538962934e-05,
      "loss": 1.1104,
      "mean_token_accuracy": 0.7877506449818611,
      "num_tokens": 1905415985.0,
      "step": 28210
    },
    {
      "entropy": 0.7210176229476929,
      "epoch": 8.548966144058168,
      "grad_norm": 0.24380862712860107,
      "learning_rate": 4.260221778206178e-05,
      "loss": 1.1111,
      "mean_token_accuracy": 0.7773053452372551,
      "num_tokens": 1906066787.0,
      "step": 28220
    },
    {
      "entropy": 0.7159445375204087,
      "epoch": 8.551995758539725,
      "grad_norm": 0.25712108612060547,
      "learning_rate": 4.256919347427516e-05,
      "loss": 1.1096,
      "mean_token_accuracy": 0.7849787518382072,
      "num_tokens": 1906746985.0,
      "step": 28230
    },
    {
      "entropy": 0.7213902294635772,
      "epoch": 8.555025373021284,
      "grad_norm": 0.2573804259300232,
      "learning_rate": 4.2536172481e-05,
      "loss": 1.1148,
      "mean_token_accuracy": 0.7856921151280403,
      "num_tokens": 1907443686.0,
      "step": 28240
    },
    {
      "entropy": 0.7158431977033615,
      "epoch": 8.55805498750284,
      "grad_norm": 0.25303325057029724,
      "learning_rate": 4.250315481696527e-05,
      "loss": 1.1027,
      "mean_token_accuracy": 0.7845643639564515,
      "num_tokens": 1908114289.0,
      "step": 28250
    },
    {
      "entropy": 0.7229256302118301,
      "epoch": 8.561084601984398,
      "grad_norm": 0.24933502078056335,
      "learning_rate": 4.247014049689852e-05,
      "loss": 1.1147,
      "mean_token_accuracy": 0.782072140276432,
      "num_tokens": 1908798709.0,
      "step": 28260
    },
    {
      "entropy": 0.7287473857402802,
      "epoch": 8.564114216465954,
      "grad_norm": 0.27063044905662537,
      "learning_rate": 4.24371295355258e-05,
      "loss": 1.1163,
      "mean_token_accuracy": 0.7813744112849236,
      "num_tokens": 1909477472.0,
      "step": 28270
    },
    {
      "entropy": 0.7227299347519874,
      "epoch": 8.567143830947511,
      "grad_norm": 0.2610662579536438,
      "learning_rate": 4.240412194757163e-05,
      "loss": 1.1064,
      "mean_token_accuracy": 0.7828714400529861,
      "num_tokens": 1910158452.0,
      "step": 28280
    },
    {
      "entropy": 0.7248746410012246,
      "epoch": 8.570173445429068,
      "grad_norm": 0.2513570487499237,
      "learning_rate": 4.2371117747759064e-05,
      "loss": 1.12,
      "mean_token_accuracy": 0.7782956436276436,
      "num_tokens": 1910826710.0,
      "step": 28290
    },
    {
      "entropy": 0.7234019681811332,
      "epoch": 8.573203059910627,
      "grad_norm": 0.26494210958480835,
      "learning_rate": 4.2338116950809626e-05,
      "loss": 1.1143,
      "mean_token_accuracy": 0.7800129473209381,
      "num_tokens": 1911500695.0,
      "step": 28300
    },
    {
      "entropy": 0.7204536572098732,
      "epoch": 8.576232674392184,
      "grad_norm": 0.24218662083148956,
      "learning_rate": 4.2305119571443305e-05,
      "loss": 1.1164,
      "mean_token_accuracy": 0.7839529857039451,
      "num_tokens": 1912176330.0,
      "step": 28310
    },
    {
      "entropy": 0.7227112546563148,
      "epoch": 8.57926228887374,
      "grad_norm": 0.23897382616996765,
      "learning_rate": 4.227212562437859e-05,
      "loss": 1.1228,
      "mean_token_accuracy": 0.7834416672587394,
      "num_tokens": 1912856100.0,
      "step": 28320
    },
    {
      "entropy": 0.7184045344591141,
      "epoch": 8.582291903355298,
      "grad_norm": 0.2519665062427521,
      "learning_rate": 4.223913512433244e-05,
      "loss": 1.1095,
      "mean_token_accuracy": 0.780766113102436,
      "num_tokens": 1913540086.0,
      "step": 28330
    },
    {
      "entropy": 0.7270558908581733,
      "epoch": 8.585321517836855,
      "grad_norm": 0.23440684378147125,
      "learning_rate": 4.220614808602026e-05,
      "loss": 1.1126,
      "mean_token_accuracy": 0.7794644325971604,
      "num_tokens": 1914209628.0,
      "step": 28340
    },
    {
      "entropy": 0.71775231808424,
      "epoch": 8.588351132318412,
      "grad_norm": 0.24744759500026703,
      "learning_rate": 4.217316452415592e-05,
      "loss": 1.1136,
      "mean_token_accuracy": 0.7830313116312027,
      "num_tokens": 1914886075.0,
      "step": 28350
    },
    {
      "entropy": 0.7131304040551185,
      "epoch": 8.59138074679997,
      "grad_norm": 0.2522117793560028,
      "learning_rate": 4.2140184453451746e-05,
      "loss": 1.1029,
      "mean_token_accuracy": 0.7822527140378952,
      "num_tokens": 1915571029.0,
      "step": 28360
    },
    {
      "entropy": 0.7166482031345367,
      "epoch": 8.594410361281527,
      "grad_norm": 0.2604322135448456,
      "learning_rate": 4.21072078886185e-05,
      "loss": 1.1192,
      "mean_token_accuracy": 0.779690945148468,
      "num_tokens": 1916233717.0,
      "step": 28370
    },
    {
      "entropy": 0.7124838411808014,
      "epoch": 8.597439975763084,
      "grad_norm": 0.2668404281139374,
      "learning_rate": 4.207423484436537e-05,
      "loss": 1.1071,
      "mean_token_accuracy": 0.785042567551136,
      "num_tokens": 1916913507.0,
      "step": 28380
    },
    {
      "entropy": 0.7189269706606864,
      "epoch": 8.600469590244641,
      "grad_norm": 0.2488768994808197,
      "learning_rate": 4.204126533539998e-05,
      "loss": 1.1125,
      "mean_token_accuracy": 0.7832905948162079,
      "num_tokens": 1917595164.0,
      "step": 28390
    },
    {
      "entropy": 0.7153389841318131,
      "epoch": 8.603499204726198,
      "grad_norm": 0.2504398822784424,
      "learning_rate": 4.2008299376428394e-05,
      "loss": 1.1093,
      "mean_token_accuracy": 0.7843625247478485,
      "num_tokens": 1918272014.0,
      "step": 28400
    },
    {
      "entropy": 0.7196563646197319,
      "epoch": 8.606528819207757,
      "grad_norm": 0.26472902297973633,
      "learning_rate": 4.197533698215507e-05,
      "loss": 1.0974,
      "mean_token_accuracy": 0.7859700217843055,
      "num_tokens": 1918963097.0,
      "step": 28410
    },
    {
      "entropy": 0.7287015676498413,
      "epoch": 8.609558433689314,
      "grad_norm": 0.2562050521373749,
      "learning_rate": 4.194237816728288e-05,
      "loss": 1.1163,
      "mean_token_accuracy": 0.7776473611593246,
      "num_tokens": 1919639991.0,
      "step": 28420
    },
    {
      "entropy": 0.7181836724281311,
      "epoch": 8.61258804817087,
      "grad_norm": 0.24757793545722961,
      "learning_rate": 4.190942294651313e-05,
      "loss": 1.1161,
      "mean_token_accuracy": 0.7788678079843521,
      "num_tokens": 1920304695.0,
      "step": 28430
    },
    {
      "entropy": 0.7180621802806855,
      "epoch": 8.615617662652427,
      "grad_norm": 0.24904799461364746,
      "learning_rate": 4.1876471334545466e-05,
      "loss": 1.1102,
      "mean_token_accuracy": 0.7853743806481361,
      "num_tokens": 1920986144.0,
      "step": 28440
    },
    {
      "entropy": 0.716050174832344,
      "epoch": 8.618647277133984,
      "grad_norm": 0.254031777381897,
      "learning_rate": 4.184352334607796e-05,
      "loss": 1.1081,
      "mean_token_accuracy": 0.7810869053006172,
      "num_tokens": 1921662755.0,
      "step": 28450
    },
    {
      "entropy": 0.7207084417343139,
      "epoch": 8.621676891615541,
      "grad_norm": 0.24686495959758759,
      "learning_rate": 4.181057899580708e-05,
      "loss": 1.108,
      "mean_token_accuracy": 0.7840207055211067,
      "num_tokens": 1922342411.0,
      "step": 28460
    },
    {
      "entropy": 0.7214000925421715,
      "epoch": 8.624706506097098,
      "grad_norm": 0.25418463349342346,
      "learning_rate": 4.1777638298427646e-05,
      "loss": 1.1091,
      "mean_token_accuracy": 0.7834576144814491,
      "num_tokens": 1923019703.0,
      "step": 28470
    },
    {
      "entropy": 0.7185854330658913,
      "epoch": 8.627736120578657,
      "grad_norm": 0.2534089982509613,
      "learning_rate": 4.174470126863285e-05,
      "loss": 1.1003,
      "mean_token_accuracy": 0.7833646684885025,
      "num_tokens": 1923702536.0,
      "step": 28480
    },
    {
      "entropy": 0.7247257903218269,
      "epoch": 8.630765735060214,
      "grad_norm": 0.2549518346786499,
      "learning_rate": 4.1711767921114265e-05,
      "loss": 1.1112,
      "mean_token_accuracy": 0.7785336881875992,
      "num_tokens": 1924376486.0,
      "step": 28490
    },
    {
      "entropy": 0.7220595791935921,
      "epoch": 8.63379534954177,
      "grad_norm": 0.2647266983985901,
      "learning_rate": 4.1678838270561814e-05,
      "loss": 1.1148,
      "mean_token_accuracy": 0.7850570365786552,
      "num_tokens": 1925068760.0,
      "step": 28500
    },
    {
      "entropy": 0.7314556479454041,
      "epoch": 8.636824964023328,
      "grad_norm": 0.2623205780982971,
      "learning_rate": 4.164591233166375e-05,
      "loss": 1.1291,
      "mean_token_accuracy": 0.7768009513616562,
      "num_tokens": 1925747773.0,
      "step": 28510
    },
    {
      "entropy": 0.7182144120335578,
      "epoch": 8.639854578504885,
      "grad_norm": 0.27247917652130127,
      "learning_rate": 4.1612990119106706e-05,
      "loss": 1.1182,
      "mean_token_accuracy": 0.7795385822653771,
      "num_tokens": 1926409866.0,
      "step": 28520
    },
    {
      "entropy": 0.7145122721791267,
      "epoch": 8.642884192986443,
      "grad_norm": 0.2504322826862335,
      "learning_rate": 4.1580071647575634e-05,
      "loss": 1.1101,
      "mean_token_accuracy": 0.7811422646045685,
      "num_tokens": 1927074050.0,
      "step": 28530
    },
    {
      "entropy": 0.7201222479343414,
      "epoch": 8.645913807468,
      "grad_norm": 0.24874699115753174,
      "learning_rate": 4.1547156931753813e-05,
      "loss": 1.1099,
      "mean_token_accuracy": 0.7810144543647766,
      "num_tokens": 1927739565.0,
      "step": 28540
    },
    {
      "entropy": 0.7202480390667916,
      "epoch": 8.648943421949557,
      "grad_norm": 0.24992485344409943,
      "learning_rate": 4.1514245986322846e-05,
      "loss": 1.1239,
      "mean_token_accuracy": 0.7761563330888748,
      "num_tokens": 1928391157.0,
      "step": 28550
    },
    {
      "entropy": 0.7083740383386612,
      "epoch": 8.651973036431114,
      "grad_norm": 0.25214624404907227,
      "learning_rate": 4.148133882596269e-05,
      "loss": 1.0964,
      "mean_token_accuracy": 0.7914482444524765,
      "num_tokens": 1929094996.0,
      "step": 28560
    },
    {
      "entropy": 0.7194327369332314,
      "epoch": 8.655002650912671,
      "grad_norm": 0.23992250859737396,
      "learning_rate": 4.144843546535155e-05,
      "loss": 1.1143,
      "mean_token_accuracy": 0.7817048177123069,
      "num_tokens": 1929772714.0,
      "step": 28570
    },
    {
      "entropy": 0.7236621171236038,
      "epoch": 8.658032265394228,
      "grad_norm": 0.25560665130615234,
      "learning_rate": 4.141553591916598e-05,
      "loss": 1.1075,
      "mean_token_accuracy": 0.7771933913230896,
      "num_tokens": 1930445435.0,
      "step": 28580
    },
    {
      "entropy": 0.725893759727478,
      "epoch": 8.661061879875787,
      "grad_norm": 0.25523504614830017,
      "learning_rate": 4.138264020208086e-05,
      "loss": 1.124,
      "mean_token_accuracy": 0.7764753505587578,
      "num_tokens": 1931098014.0,
      "step": 28590
    },
    {
      "entropy": 0.7127537101507186,
      "epoch": 8.664091494357343,
      "grad_norm": 0.25567346811294556,
      "learning_rate": 4.1349748328769265e-05,
      "loss": 1.1195,
      "mean_token_accuracy": 0.7771034657955169,
      "num_tokens": 1931748298.0,
      "step": 28600
    },
    {
      "entropy": 0.7272965505719184,
      "epoch": 8.6671211088389,
      "grad_norm": 0.25819697976112366,
      "learning_rate": 4.1316860313902666e-05,
      "loss": 1.1166,
      "mean_token_accuracy": 0.7850852489471436,
      "num_tokens": 1932437385.0,
      "step": 28610
    },
    {
      "entropy": 0.7176499485969543,
      "epoch": 8.670150723320457,
      "grad_norm": 0.24638673663139343,
      "learning_rate": 4.1283976172150726e-05,
      "loss": 1.1035,
      "mean_token_accuracy": 0.7825480386614799,
      "num_tokens": 1933116742.0,
      "step": 28620
    },
    {
      "entropy": 0.7143126979470253,
      "epoch": 8.673180337802014,
      "grad_norm": 0.2555699646472931,
      "learning_rate": 4.125109591818144e-05,
      "loss": 1.1018,
      "mean_token_accuracy": 0.7826764941215515,
      "num_tokens": 1933789342.0,
      "step": 28630
    },
    {
      "entropy": 0.7244675323367119,
      "epoch": 8.676209952283571,
      "grad_norm": 0.249857097864151,
      "learning_rate": 4.121821956666104e-05,
      "loss": 1.1089,
      "mean_token_accuracy": 0.783166480064392,
      "num_tokens": 1934482228.0,
      "step": 28640
    },
    {
      "entropy": 0.7126829192042351,
      "epoch": 8.67923956676513,
      "grad_norm": 0.2583950161933899,
      "learning_rate": 4.118534713225402e-05,
      "loss": 1.1144,
      "mean_token_accuracy": 0.7875151291489602,
      "num_tokens": 1935168965.0,
      "step": 28650
    },
    {
      "entropy": 0.7243151098489762,
      "epoch": 8.682269181246687,
      "grad_norm": 0.2539556920528412,
      "learning_rate": 4.115247862962313e-05,
      "loss": 1.1068,
      "mean_token_accuracy": 0.782958970963955,
      "num_tokens": 1935853331.0,
      "step": 28660
    },
    {
      "entropy": 0.721580271422863,
      "epoch": 8.685298795728244,
      "grad_norm": 0.2541027367115021,
      "learning_rate": 4.111961407342937e-05,
      "loss": 1.1118,
      "mean_token_accuracy": 0.7814381510019303,
      "num_tokens": 1936529254.0,
      "step": 28670
    },
    {
      "entropy": 0.7292007640004158,
      "epoch": 8.6883284102098,
      "grad_norm": 0.25219234824180603,
      "learning_rate": 4.108675347833199e-05,
      "loss": 1.1168,
      "mean_token_accuracy": 0.7778084501624107,
      "num_tokens": 1937198608.0,
      "step": 28680
    },
    {
      "entropy": 0.7166239753365516,
      "epoch": 8.691358024691358,
      "grad_norm": 0.25991955399513245,
      "learning_rate": 4.105389685898843e-05,
      "loss": 1.1103,
      "mean_token_accuracy": 0.7806353375315667,
      "num_tokens": 1937865862.0,
      "step": 28690
    },
    {
      "entropy": 0.7189975202083587,
      "epoch": 8.694387639172914,
      "grad_norm": 0.24662035703659058,
      "learning_rate": 4.1021044230054405e-05,
      "loss": 1.1086,
      "mean_token_accuracy": 0.7831170216202736,
      "num_tokens": 1938547499.0,
      "step": 28700
    },
    {
      "entropy": 0.7202324852347374,
      "epoch": 8.697417253654473,
      "grad_norm": 0.2525968849658966,
      "learning_rate": 4.098819560618384e-05,
      "loss": 1.1186,
      "mean_token_accuracy": 0.7782834306359291,
      "num_tokens": 1939204955.0,
      "step": 28710
    },
    {
      "entropy": 0.721531181037426,
      "epoch": 8.70044686813603,
      "grad_norm": 0.25534752011299133,
      "learning_rate": 4.095535100202884e-05,
      "loss": 1.112,
      "mean_token_accuracy": 0.7807690426707268,
      "num_tokens": 1939887443.0,
      "step": 28720
    },
    {
      "entropy": 0.7301719605922699,
      "epoch": 8.703476482617587,
      "grad_norm": 0.2842459976673126,
      "learning_rate": 4.092251043223975e-05,
      "loss": 1.1146,
      "mean_token_accuracy": 0.7789567887783051,
      "num_tokens": 1940559581.0,
      "step": 28730
    },
    {
      "entropy": 0.7073962211608886,
      "epoch": 8.706506097099144,
      "grad_norm": 0.24807953834533691,
      "learning_rate": 4.088967391146512e-05,
      "loss": 1.101,
      "mean_token_accuracy": 0.7865076690912247,
      "num_tokens": 1941240652.0,
      "step": 28740
    },
    {
      "entropy": 0.7230818122625351,
      "epoch": 8.7095357115807,
      "grad_norm": 0.2505377531051636,
      "learning_rate": 4.0856841454351655e-05,
      "loss": 1.1198,
      "mean_token_accuracy": 0.7763510674238205,
      "num_tokens": 1941902577.0,
      "step": 28750
    },
    {
      "entropy": 0.7187742874026298,
      "epoch": 8.712565326062258,
      "grad_norm": 0.2498081773519516,
      "learning_rate": 4.08240130755443e-05,
      "loss": 1.1179,
      "mean_token_accuracy": 0.7771988451480866,
      "num_tokens": 1942566897.0,
      "step": 28760
    },
    {
      "entropy": 0.7182139918208122,
      "epoch": 8.715594940543816,
      "grad_norm": 0.2587888240814209,
      "learning_rate": 4.079118878968615e-05,
      "loss": 1.1091,
      "mean_token_accuracy": 0.7775295615196228,
      "num_tokens": 1943229094.0,
      "step": 28770
    },
    {
      "entropy": 0.7200123488903045,
      "epoch": 8.718624555025373,
      "grad_norm": 0.25699582695961,
      "learning_rate": 4.075836861141846e-05,
      "loss": 1.1148,
      "mean_token_accuracy": 0.783364151418209,
      "num_tokens": 1943903293.0,
      "step": 28780
    },
    {
      "entropy": 0.7022015497088432,
      "epoch": 8.72165416950693,
      "grad_norm": 0.25240638852119446,
      "learning_rate": 4.072555255538068e-05,
      "loss": 1.0967,
      "mean_token_accuracy": 0.7873583644628525,
      "num_tokens": 1944576805.0,
      "step": 28790
    },
    {
      "entropy": 0.7154752492904664,
      "epoch": 8.724683783988487,
      "grad_norm": 0.2615509331226349,
      "learning_rate": 4.069274063621043e-05,
      "loss": 1.1072,
      "mean_token_accuracy": 0.7806541219353675,
      "num_tokens": 1945241096.0,
      "step": 28800
    },
    {
      "entropy": 0.7336331263184548,
      "epoch": 8.727713398470044,
      "grad_norm": 0.2593686580657959,
      "learning_rate": 4.065993286854345e-05,
      "loss": 1.1183,
      "mean_token_accuracy": 0.776524794101715,
      "num_tokens": 1945918249.0,
      "step": 28810
    },
    {
      "entropy": 0.7094599232077599,
      "epoch": 8.730743012951603,
      "grad_norm": 0.2444300800561905,
      "learning_rate": 4.0627129267013654e-05,
      "loss": 1.106,
      "mean_token_accuracy": 0.7816817224025726,
      "num_tokens": 1946588886.0,
      "step": 28820
    },
    {
      "entropy": 0.7200517654418945,
      "epoch": 8.73377262743316,
      "grad_norm": 0.2554880380630493,
      "learning_rate": 4.059432984625307e-05,
      "loss": 1.1104,
      "mean_token_accuracy": 0.7789775893092156,
      "num_tokens": 1947255963.0,
      "step": 28830
    },
    {
      "entropy": 0.7202544391155243,
      "epoch": 8.736802241914717,
      "grad_norm": 0.2426423281431198,
      "learning_rate": 4.056153462089191e-05,
      "loss": 1.1111,
      "mean_token_accuracy": 0.7811928883194923,
      "num_tokens": 1947924276.0,
      "step": 28840
    },
    {
      "entropy": 0.7059272572398185,
      "epoch": 8.739831856396274,
      "grad_norm": 0.2617579698562622,
      "learning_rate": 4.052874360555846e-05,
      "loss": 1.1139,
      "mean_token_accuracy": 0.7811008810997009,
      "num_tokens": 1948586366.0,
      "step": 28850
    },
    {
      "entropy": 0.7253514006733894,
      "epoch": 8.74286147087783,
      "grad_norm": 0.25759056210517883,
      "learning_rate": 4.049595681487916e-05,
      "loss": 1.1206,
      "mean_token_accuracy": 0.7825950235128403,
      "num_tokens": 1949261209.0,
      "step": 28860
    },
    {
      "entropy": 0.7255978435277939,
      "epoch": 8.745891085359387,
      "grad_norm": 0.25632065534591675,
      "learning_rate": 4.0463174263478575e-05,
      "loss": 1.1028,
      "mean_token_accuracy": 0.7838640242815018,
      "num_tokens": 1949951570.0,
      "step": 28870
    },
    {
      "entropy": 0.718338181078434,
      "epoch": 8.748920699840944,
      "grad_norm": 0.25563666224479675,
      "learning_rate": 4.043039596597934e-05,
      "loss": 1.1131,
      "mean_token_accuracy": 0.7802659586071968,
      "num_tokens": 1950624759.0,
      "step": 28880
    },
    {
      "entropy": 0.7295897051692009,
      "epoch": 8.751950314322503,
      "grad_norm": 0.24798616766929626,
      "learning_rate": 4.0397621937002224e-05,
      "loss": 1.1177,
      "mean_token_accuracy": 0.7814691454172135,
      "num_tokens": 1951308501.0,
      "step": 28890
    },
    {
      "entropy": 0.7243193000555038,
      "epoch": 8.75497992880406,
      "grad_norm": 0.258666068315506,
      "learning_rate": 4.036485219116608e-05,
      "loss": 1.1161,
      "mean_token_accuracy": 0.7811876088380814,
      "num_tokens": 1951982623.0,
      "step": 28900
    },
    {
      "entropy": 0.7230897203087807,
      "epoch": 8.758009543285617,
      "grad_norm": 0.24920624494552612,
      "learning_rate": 4.033208674308788e-05,
      "loss": 1.1081,
      "mean_token_accuracy": 0.781388807296753,
      "num_tokens": 1952666389.0,
      "step": 28910
    },
    {
      "entropy": 0.7174295037984848,
      "epoch": 8.761039157767174,
      "grad_norm": 0.2559826076030731,
      "learning_rate": 4.029932560738261e-05,
      "loss": 1.117,
      "mean_token_accuracy": 0.7765072166919709,
      "num_tokens": 1953338904.0,
      "step": 28920
    },
    {
      "entropy": 0.711271233856678,
      "epoch": 8.76406877224873,
      "grad_norm": 0.2394898384809494,
      "learning_rate": 4.02665687986634e-05,
      "loss": 1.1017,
      "mean_token_accuracy": 0.7824870631098747,
      "num_tokens": 1954013318.0,
      "step": 28930
    },
    {
      "entropy": 0.7250132650136948,
      "epoch": 8.76709838673029,
      "grad_norm": 0.24374224245548248,
      "learning_rate": 4.0233816331541446e-05,
      "loss": 1.1114,
      "mean_token_accuracy": 0.7829572439193726,
      "num_tokens": 1954701010.0,
      "step": 28940
    },
    {
      "entropy": 0.7257913932204246,
      "epoch": 8.770128001211846,
      "grad_norm": 0.2666158080101013,
      "learning_rate": 4.0201068220625956e-05,
      "loss": 1.1189,
      "mean_token_accuracy": 0.7795502886176109,
      "num_tokens": 1955374388.0,
      "step": 28950
    },
    {
      "entropy": 0.7138613164424896,
      "epoch": 8.773157615693403,
      "grad_norm": 0.2544364631175995,
      "learning_rate": 4.0168324480524215e-05,
      "loss": 1.112,
      "mean_token_accuracy": 0.7827801123261452,
      "num_tokens": 1956053327.0,
      "step": 28960
    },
    {
      "entropy": 0.7104690358042717,
      "epoch": 8.77618723017496,
      "grad_norm": 0.24729299545288086,
      "learning_rate": 4.013558512584159e-05,
      "loss": 1.0972,
      "mean_token_accuracy": 0.7920410320162773,
      "num_tokens": 1956759614.0,
      "step": 28970
    },
    {
      "entropy": 0.715251824259758,
      "epoch": 8.779216844656517,
      "grad_norm": 0.2622358500957489,
      "learning_rate": 4.010285017118148e-05,
      "loss": 1.1088,
      "mean_token_accuracy": 0.780117604136467,
      "num_tokens": 1957431280.0,
      "step": 28980
    },
    {
      "entropy": 0.7235038727521896,
      "epoch": 8.782246459138074,
      "grad_norm": 0.26019060611724854,
      "learning_rate": 4.007011963114529e-05,
      "loss": 1.1152,
      "mean_token_accuracy": 0.777077080309391,
      "num_tokens": 1958090531.0,
      "step": 28990
    },
    {
      "entropy": 0.7245069786906242,
      "epoch": 8.785276073619633,
      "grad_norm": 0.25815483927726746,
      "learning_rate": 4.003739352033248e-05,
      "loss": 1.1167,
      "mean_token_accuracy": 0.7816456466913223,
      "num_tokens": 1958761785.0,
      "step": 29000
    },
    {
      "entropy": 0.7195040881633759,
      "epoch": 8.78830568810119,
      "grad_norm": 0.26050928235054016,
      "learning_rate": 4.0004671853340545e-05,
      "loss": 1.1165,
      "mean_token_accuracy": 0.7832333400845528,
      "num_tokens": 1959443739.0,
      "step": 29010
    },
    {
      "entropy": 0.7293911203742027,
      "epoch": 8.791335302582747,
      "grad_norm": 0.2726181447505951,
      "learning_rate": 3.9971954644765e-05,
      "loss": 1.1218,
      "mean_token_accuracy": 0.7771417945623398,
      "num_tokens": 1960108512.0,
      "step": 29020
    },
    {
      "entropy": 0.715778224170208,
      "epoch": 8.794364917064303,
      "grad_norm": 0.24542336165905,
      "learning_rate": 3.993924190919932e-05,
      "loss": 1.1144,
      "mean_token_accuracy": 0.7793296068906784,
      "num_tokens": 1960768263.0,
      "step": 29030
    },
    {
      "entropy": 0.7286868259310723,
      "epoch": 8.79739453154586,
      "grad_norm": 0.2558838129043579,
      "learning_rate": 3.990653366123504e-05,
      "loss": 1.1209,
      "mean_token_accuracy": 0.7779481574892998,
      "num_tokens": 1961431378.0,
      "step": 29040
    },
    {
      "entropy": 0.7266527563333511,
      "epoch": 8.800424146027417,
      "grad_norm": 0.2615821361541748,
      "learning_rate": 3.987382991546169e-05,
      "loss": 1.1234,
      "mean_token_accuracy": 0.7768598183989525,
      "num_tokens": 1962099362.0,
      "step": 29050
    },
    {
      "entropy": 0.7164703905582428,
      "epoch": 8.803453760508976,
      "grad_norm": 0.2523133158683777,
      "learning_rate": 3.984113068646674e-05,
      "loss": 1.1067,
      "mean_token_accuracy": 0.7812863737344742,
      "num_tokens": 1962763998.0,
      "step": 29060
    },
    {
      "entropy": 0.7180733740329742,
      "epoch": 8.806483374990533,
      "grad_norm": 0.26194852590560913,
      "learning_rate": 3.9808435988835714e-05,
      "loss": 1.1159,
      "mean_token_accuracy": 0.7819589123129844,
      "num_tokens": 1963446007.0,
      "step": 29070
    },
    {
      "entropy": 0.715312410891056,
      "epoch": 8.80951298947209,
      "grad_norm": 0.25656309723854065,
      "learning_rate": 3.977574583715207e-05,
      "loss": 1.1066,
      "mean_token_accuracy": 0.7843555957078934,
      "num_tokens": 1964130571.0,
      "step": 29080
    },
    {
      "entropy": 0.7154703140258789,
      "epoch": 8.812542603953647,
      "grad_norm": 0.25038787722587585,
      "learning_rate": 3.974306024599725e-05,
      "loss": 1.1073,
      "mean_token_accuracy": 0.7852437317371368,
      "num_tokens": 1964806701.0,
      "step": 29090
    },
    {
      "entropy": 0.7199769169092178,
      "epoch": 8.815572218435204,
      "grad_norm": 0.25432780385017395,
      "learning_rate": 3.971037922995067e-05,
      "loss": 1.1098,
      "mean_token_accuracy": 0.7845522180199623,
      "num_tokens": 1965487561.0,
      "step": 29100
    },
    {
      "entropy": 0.7070397034287452,
      "epoch": 8.81860183291676,
      "grad_norm": 0.2585289776325226,
      "learning_rate": 3.967770280358969e-05,
      "loss": 1.101,
      "mean_token_accuracy": 0.7847616329789162,
      "num_tokens": 1966164914.0,
      "step": 29110
    },
    {
      "entropy": 0.7252490550279618,
      "epoch": 8.82163144739832,
      "grad_norm": 0.261148065328598,
      "learning_rate": 3.964503098148964e-05,
      "loss": 1.1164,
      "mean_token_accuracy": 0.779693067073822,
      "num_tokens": 1966838245.0,
      "step": 29120
    },
    {
      "entropy": 0.719429862499237,
      "epoch": 8.824661061879876,
      "grad_norm": 0.24912382662296295,
      "learning_rate": 3.961236377822377e-05,
      "loss": 1.1062,
      "mean_token_accuracy": 0.7832588732242585,
      "num_tokens": 1967537263.0,
      "step": 29130
    },
    {
      "entropy": 0.7216413915157318,
      "epoch": 8.827690676361433,
      "grad_norm": 0.25173404812812805,
      "learning_rate": 3.957970120836331e-05,
      "loss": 1.107,
      "mean_token_accuracy": 0.779204361140728,
      "num_tokens": 1968218852.0,
      "step": 29140
    },
    {
      "entropy": 0.7214172005653381,
      "epoch": 8.83072029084299,
      "grad_norm": 0.24195538461208344,
      "learning_rate": 3.954704328647739e-05,
      "loss": 1.1186,
      "mean_token_accuracy": 0.7817006081342697,
      "num_tokens": 1968893708.0,
      "step": 29150
    },
    {
      "entropy": 0.7169260442256927,
      "epoch": 8.833749905324547,
      "grad_norm": 0.25657886266708374,
      "learning_rate": 3.951439002713306e-05,
      "loss": 1.115,
      "mean_token_accuracy": 0.7818932577967643,
      "num_tokens": 1969562790.0,
      "step": 29160
    },
    {
      "entropy": 0.7181063741445541,
      "epoch": 8.836779519806104,
      "grad_norm": 0.2619813084602356,
      "learning_rate": 3.948174144489533e-05,
      "loss": 1.1079,
      "mean_token_accuracy": 0.7846121177077293,
      "num_tokens": 1970247901.0,
      "step": 29170
    },
    {
      "entropy": 0.7212286561727523,
      "epoch": 8.839809134287663,
      "grad_norm": 0.25744473934173584,
      "learning_rate": 3.9449097554327106e-05,
      "loss": 1.115,
      "mean_token_accuracy": 0.7834999904036521,
      "num_tokens": 1970934230.0,
      "step": 29180
    },
    {
      "entropy": 0.7198623836040496,
      "epoch": 8.84283874876922,
      "grad_norm": 0.24997738003730774,
      "learning_rate": 3.9416458369989165e-05,
      "loss": 1.1109,
      "mean_token_accuracy": 0.7817680180072785,
      "num_tokens": 1971611091.0,
      "step": 29190
    },
    {
      "entropy": 0.7187659323215485,
      "epoch": 8.845868363250776,
      "grad_norm": 0.2572028338909149,
      "learning_rate": 3.938382390644024e-05,
      "loss": 1.109,
      "mean_token_accuracy": 0.7830177575349808,
      "num_tokens": 1972301989.0,
      "step": 29200
    },
    {
      "entropy": 0.7165235981345177,
      "epoch": 8.848897977732333,
      "grad_norm": 0.26468753814697266,
      "learning_rate": 3.9351194178236936e-05,
      "loss": 1.1123,
      "mean_token_accuracy": 0.779882937669754,
      "num_tokens": 1972967238.0,
      "step": 29210
    },
    {
      "entropy": 0.7153317481279373,
      "epoch": 8.85192759221389,
      "grad_norm": 0.2593344748020172,
      "learning_rate": 3.9318569199933726e-05,
      "loss": 1.1063,
      "mean_token_accuracy": 0.7848672181367874,
      "num_tokens": 1973656440.0,
      "step": 29220
    },
    {
      "entropy": 0.7206306457519531,
      "epoch": 8.854957206695449,
      "grad_norm": 0.2500818967819214,
      "learning_rate": 3.9285948986082995e-05,
      "loss": 1.1131,
      "mean_token_accuracy": 0.7811333194375039,
      "num_tokens": 1974347297.0,
      "step": 29230
    },
    {
      "entropy": 0.7156893268227578,
      "epoch": 8.857986821177006,
      "grad_norm": 0.2541957497596741,
      "learning_rate": 3.9253333551234984e-05,
      "loss": 1.1027,
      "mean_token_accuracy": 0.7849383920431137,
      "num_tokens": 1975034819.0,
      "step": 29240
    },
    {
      "entropy": 0.7273165121674537,
      "epoch": 8.861016435658563,
      "grad_norm": 0.24332493543624878,
      "learning_rate": 3.922072290993783e-05,
      "loss": 1.1038,
      "mean_token_accuracy": 0.788136437535286,
      "num_tokens": 1975749273.0,
      "step": 29250
    },
    {
      "entropy": 0.7169123306870461,
      "epoch": 8.86404605014012,
      "grad_norm": 0.25201770663261414,
      "learning_rate": 3.918811707673748e-05,
      "loss": 1.1127,
      "mean_token_accuracy": 0.7837028563022613,
      "num_tokens": 1976421424.0,
      "step": 29260
    },
    {
      "entropy": 0.7232697606086731,
      "epoch": 8.867075664621677,
      "grad_norm": 0.25215113162994385,
      "learning_rate": 3.9155516066177775e-05,
      "loss": 1.1129,
      "mean_token_accuracy": 0.7839630246162415,
      "num_tokens": 1977109932.0,
      "step": 29270
    },
    {
      "entropy": 0.7059458583593369,
      "epoch": 8.870105279103234,
      "grad_norm": 0.27492555975914,
      "learning_rate": 3.912291989280042e-05,
      "loss": 1.1135,
      "mean_token_accuracy": 0.7810848966240883,
      "num_tokens": 1977765241.0,
      "step": 29280
    },
    {
      "entropy": 0.7126860484480858,
      "epoch": 8.87313489358479,
      "grad_norm": 0.2462565302848816,
      "learning_rate": 3.909032857114493e-05,
      "loss": 1.1066,
      "mean_token_accuracy": 0.78338822722435,
      "num_tokens": 1978438667.0,
      "step": 29290
    },
    {
      "entropy": 0.7065578386187553,
      "epoch": 8.87616450806635,
      "grad_norm": 0.24767723679542542,
      "learning_rate": 3.905774211574863e-05,
      "loss": 1.1116,
      "mean_token_accuracy": 0.7801328271627426,
      "num_tokens": 1979090499.0,
      "step": 29300
    },
    {
      "entropy": 0.7230974361300468,
      "epoch": 8.879194122547906,
      "grad_norm": 0.24369339644908905,
      "learning_rate": 3.9025160541146735e-05,
      "loss": 1.11,
      "mean_token_accuracy": 0.7799566656351089,
      "num_tokens": 1979775527.0,
      "step": 29310
    },
    {
      "entropy": 0.7286474034190178,
      "epoch": 8.882223737029463,
      "grad_norm": 0.2690872550010681,
      "learning_rate": 3.899258386187226e-05,
      "loss": 1.1186,
      "mean_token_accuracy": 0.7749822065234184,
      "num_tokens": 1980438431.0,
      "step": 29320
    },
    {
      "entropy": 0.7080101639032363,
      "epoch": 8.88525335151102,
      "grad_norm": 0.24646250903606415,
      "learning_rate": 3.896001209245604e-05,
      "loss": 1.1005,
      "mean_token_accuracy": 0.7858548179268837,
      "num_tokens": 1981117410.0,
      "step": 29330
    },
    {
      "entropy": 0.7161659374833107,
      "epoch": 8.888282965992577,
      "grad_norm": 0.24269527196884155,
      "learning_rate": 3.892744524742669e-05,
      "loss": 1.1195,
      "mean_token_accuracy": 0.7821093693375587,
      "num_tokens": 1981780762.0,
      "step": 29340
    },
    {
      "entropy": 0.7193952903151513,
      "epoch": 8.891312580474136,
      "grad_norm": 0.26987215876579285,
      "learning_rate": 3.889488334131067e-05,
      "loss": 1.1111,
      "mean_token_accuracy": 0.7841506749391556,
      "num_tokens": 1982456617.0,
      "step": 29350
    },
    {
      "entropy": 0.7211821421980857,
      "epoch": 8.894342194955692,
      "grad_norm": 0.25645774602890015,
      "learning_rate": 3.886232638863223e-05,
      "loss": 1.1109,
      "mean_token_accuracy": 0.7815490037202835,
      "num_tokens": 1983133499.0,
      "step": 29360
    },
    {
      "entropy": 0.7222009599208832,
      "epoch": 8.89737180943725,
      "grad_norm": 0.24122609198093414,
      "learning_rate": 3.882977440391337e-05,
      "loss": 1.1088,
      "mean_token_accuracy": 0.7885529518127441,
      "num_tokens": 1983839999.0,
      "step": 29370
    },
    {
      "entropy": 0.7147975280880928,
      "epoch": 8.900401423918806,
      "grad_norm": 0.2429054230451584,
      "learning_rate": 3.8797227401673915e-05,
      "loss": 1.1139,
      "mean_token_accuracy": 0.7818574577569961,
      "num_tokens": 1984510703.0,
      "step": 29380
    },
    {
      "entropy": 0.713473479449749,
      "epoch": 8.903431038400363,
      "grad_norm": 0.2651282250881195,
      "learning_rate": 3.876468539643147e-05,
      "loss": 1.1028,
      "mean_token_accuracy": 0.7835205689072609,
      "num_tokens": 1985180169.0,
      "step": 29390
    },
    {
      "entropy": 0.716995793581009,
      "epoch": 8.90646065288192,
      "grad_norm": 0.2531137764453888,
      "learning_rate": 3.873214840270138e-05,
      "loss": 1.1066,
      "mean_token_accuracy": 0.7841816321015358,
      "num_tokens": 1985861492.0,
      "step": 29400
    },
    {
      "entropy": 0.718168930709362,
      "epoch": 8.909490267363479,
      "grad_norm": 0.25928670167922974,
      "learning_rate": 3.8699616434996774e-05,
      "loss": 1.1031,
      "mean_token_accuracy": 0.7842026695609092,
      "num_tokens": 1986548929.0,
      "step": 29410
    },
    {
      "entropy": 0.7155626058578491,
      "epoch": 8.912519881845036,
      "grad_norm": 0.23230260610580444,
      "learning_rate": 3.866708950782856e-05,
      "loss": 1.1102,
      "mean_token_accuracy": 0.7817296728491783,
      "num_tokens": 1987225729.0,
      "step": 29420
    },
    {
      "entropy": 0.7169981315732002,
      "epoch": 8.915549496326593,
      "grad_norm": 0.25931882858276367,
      "learning_rate": 3.863456763570533e-05,
      "loss": 1.1133,
      "mean_token_accuracy": 0.7813719809055328,
      "num_tokens": 1987893133.0,
      "step": 29430
    },
    {
      "entropy": 0.711993083357811,
      "epoch": 8.91857911080815,
      "grad_norm": 0.26571208238601685,
      "learning_rate": 3.860205083313351e-05,
      "loss": 1.1114,
      "mean_token_accuracy": 0.7813589677214623,
      "num_tokens": 1988548951.0,
      "step": 29440
    },
    {
      "entropy": 0.7180174738168716,
      "epoch": 8.921608725289706,
      "grad_norm": 0.25587186217308044,
      "learning_rate": 3.8569539114617196e-05,
      "loss": 1.1186,
      "mean_token_accuracy": 0.7839012235403061,
      "num_tokens": 1989231270.0,
      "step": 29450
    },
    {
      "entropy": 0.7281316161155701,
      "epoch": 8.924638339771263,
      "grad_norm": 0.2636795938014984,
      "learning_rate": 3.853703249465826e-05,
      "loss": 1.1129,
      "mean_token_accuracy": 0.780683021247387,
      "num_tokens": 1989916503.0,
      "step": 29460
    },
    {
      "entropy": 0.710832704603672,
      "epoch": 8.927667954252822,
      "grad_norm": 0.2544217109680176,
      "learning_rate": 3.850453098775625e-05,
      "loss": 1.1019,
      "mean_token_accuracy": 0.7840294167399406,
      "num_tokens": 1990598033.0,
      "step": 29470
    },
    {
      "entropy": 0.7293656095862389,
      "epoch": 8.930697568734379,
      "grad_norm": 0.25565722584724426,
      "learning_rate": 3.8472034608408485e-05,
      "loss": 1.1162,
      "mean_token_accuracy": 0.7786854088306427,
      "num_tokens": 1991274377.0,
      "step": 29480
    },
    {
      "entropy": 0.7208972543478012,
      "epoch": 8.933727183215936,
      "grad_norm": 0.25705480575561523,
      "learning_rate": 3.8439543371109984e-05,
      "loss": 1.1143,
      "mean_token_accuracy": 0.7823522612452507,
      "num_tokens": 1991950348.0,
      "step": 29490
    },
    {
      "entropy": 0.7095813632011414,
      "epoch": 8.936756797697493,
      "grad_norm": 0.2603360116481781,
      "learning_rate": 3.8407057290353444e-05,
      "loss": 1.1124,
      "mean_token_accuracy": 0.7848794236779213,
      "num_tokens": 1992628427.0,
      "step": 29500
    },
    {
      "entropy": 0.7260840758681297,
      "epoch": 8.93978641217905,
      "grad_norm": 0.2520385682582855,
      "learning_rate": 3.8374576380629296e-05,
      "loss": 1.115,
      "mean_token_accuracy": 0.7772787541151047,
      "num_tokens": 1993300669.0,
      "step": 29510
    },
    {
      "entropy": 0.7156735628843307,
      "epoch": 8.942816026660607,
      "grad_norm": 0.26142042875289917,
      "learning_rate": 3.834210065642565e-05,
      "loss": 1.1086,
      "mean_token_accuracy": 0.7815072506666183,
      "num_tokens": 1993973719.0,
      "step": 29520
    },
    {
      "entropy": 0.7224139705300331,
      "epoch": 8.945845641142165,
      "grad_norm": 0.24675850570201874,
      "learning_rate": 3.830963013222829e-05,
      "loss": 1.1106,
      "mean_token_accuracy": 0.7821033999323845,
      "num_tokens": 1994657764.0,
      "step": 29530
    },
    {
      "entropy": 0.7191166371107102,
      "epoch": 8.948875255623722,
      "grad_norm": 0.2539442181587219,
      "learning_rate": 3.82771648225207e-05,
      "loss": 1.1128,
      "mean_token_accuracy": 0.7840816512703895,
      "num_tokens": 1995345280.0,
      "step": 29540
    },
    {
      "entropy": 0.719819089770317,
      "epoch": 8.95190487010528,
      "grad_norm": 0.2582361102104187,
      "learning_rate": 3.824470474178406e-05,
      "loss": 1.1065,
      "mean_token_accuracy": 0.782945702970028,
      "num_tokens": 1996036948.0,
      "step": 29550
    },
    {
      "entropy": 0.7191162839531898,
      "epoch": 8.954934484586836,
      "grad_norm": 0.26400452852249146,
      "learning_rate": 3.821224990449715e-05,
      "loss": 1.114,
      "mean_token_accuracy": 0.7831235095858574,
      "num_tokens": 1996705772.0,
      "step": 29560
    },
    {
      "entropy": 0.7220586121082306,
      "epoch": 8.957964099068393,
      "grad_norm": 0.25238141417503357,
      "learning_rate": 3.817980032513646e-05,
      "loss": 1.1215,
      "mean_token_accuracy": 0.7829408764839172,
      "num_tokens": 1997386912.0,
      "step": 29570
    },
    {
      "entropy": 0.7142419680953026,
      "epoch": 8.96099371354995,
      "grad_norm": 0.2720177471637726,
      "learning_rate": 3.8147356018176134e-05,
      "loss": 1.1155,
      "mean_token_accuracy": 0.7802116602659226,
      "num_tokens": 1998049925.0,
      "step": 29580
    },
    {
      "entropy": 0.7250123754143715,
      "epoch": 8.964023328031509,
      "grad_norm": 0.24264773726463318,
      "learning_rate": 3.8114916998087965e-05,
      "loss": 1.1136,
      "mean_token_accuracy": 0.7808478653430939,
      "num_tokens": 1998732906.0,
      "step": 29590
    },
    {
      "entropy": 0.718470261991024,
      "epoch": 8.967052942513066,
      "grad_norm": 0.24063296616077423,
      "learning_rate": 3.808248327934136e-05,
      "loss": 1.1078,
      "mean_token_accuracy": 0.7862830191850663,
      "num_tokens": 1999415635.0,
      "step": 29600
    },
    {
      "entropy": 0.7226502120494842,
      "epoch": 8.970082556994623,
      "grad_norm": 0.25834742188453674,
      "learning_rate": 3.8050054876403385e-05,
      "loss": 1.1172,
      "mean_token_accuracy": 0.7758548200130463,
      "num_tokens": 2000071312.0,
      "step": 29610
    },
    {
      "entropy": 0.7235815927386284,
      "epoch": 8.97311217147618,
      "grad_norm": 0.2576269805431366,
      "learning_rate": 3.8017631803738754e-05,
      "loss": 1.1174,
      "mean_token_accuracy": 0.7818986803293229,
      "num_tokens": 2000747914.0,
      "step": 29620
    },
    {
      "entropy": 0.7264309778809548,
      "epoch": 8.976141785957736,
      "grad_norm": 0.24839350581169128,
      "learning_rate": 3.7985214075809735e-05,
      "loss": 1.1238,
      "mean_token_accuracy": 0.7805746138095856,
      "num_tokens": 2001416072.0,
      "step": 29630
    },
    {
      "entropy": 0.7113376304507255,
      "epoch": 8.979171400439293,
      "grad_norm": 0.2609247863292694,
      "learning_rate": 3.795280170707631e-05,
      "loss": 1.1137,
      "mean_token_accuracy": 0.7840653508901596,
      "num_tokens": 2002086596.0,
      "step": 29640
    },
    {
      "entropy": 0.7157400205731392,
      "epoch": 8.982201014920852,
      "grad_norm": 0.2429143339395523,
      "learning_rate": 3.7920394711995954e-05,
      "loss": 1.1089,
      "mean_token_accuracy": 0.7816368266940117,
      "num_tokens": 2002763076.0,
      "step": 29650
    },
    {
      "entropy": 0.7212949648499489,
      "epoch": 8.985230629402409,
      "grad_norm": 0.26175034046173096,
      "learning_rate": 3.788799310502385e-05,
      "loss": 1.1073,
      "mean_token_accuracy": 0.7816547930240632,
      "num_tokens": 2003442438.0,
      "step": 29660
    },
    {
      "entropy": 0.7140222698450088,
      "epoch": 8.988260243883966,
      "grad_norm": 0.25381457805633545,
      "learning_rate": 3.785559690061274e-05,
      "loss": 1.1104,
      "mean_token_accuracy": 0.7850309312343597,
      "num_tokens": 2004125910.0,
      "step": 29670
    },
    {
      "entropy": 0.7134387657046318,
      "epoch": 8.991289858365523,
      "grad_norm": 0.24920444190502167,
      "learning_rate": 3.7823206113212916e-05,
      "loss": 1.1067,
      "mean_token_accuracy": 0.7890672579407692,
      "num_tokens": 2004813988.0,
      "step": 29680
    },
    {
      "entropy": 0.7175428092479705,
      "epoch": 8.99431947284708,
      "grad_norm": 0.26466912031173706,
      "learning_rate": 3.779082075727232e-05,
      "loss": 1.1148,
      "mean_token_accuracy": 0.7806145489215851,
      "num_tokens": 2005488022.0,
      "step": 29690
    },
    {
      "entropy": 0.7156146258115769,
      "epoch": 8.997349087328637,
      "grad_norm": 0.2508610188961029,
      "learning_rate": 3.775844084723644e-05,
      "loss": 1.1094,
      "mean_token_accuracy": 0.7850867241621018,
      "num_tokens": 2006180506.0,
      "step": 29700
    }
  ],
  "logging_steps": 10,
  "max_steps": 49515,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "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": 9.86333686365883e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}