{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.715306219639966,
  "eval_steps": 500,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04762358319839985,
      "grad_norm": 0.7446010708808899,
      "learning_rate": 1.986020651310564e-05,
      "loss": 2.4001,
      "mean_token_accuracy": 0.5458807892352343,
      "num_tokens": 20311447.0,
      "step": 50
    },
    {
      "epoch": 0.0952471663967997,
      "grad_norm": 1.1947827339172363,
      "learning_rate": 1.9701350277998413e-05,
      "loss": 2.0923,
      "mean_token_accuracy": 0.5837462414205075,
      "num_tokens": 40625380.0,
      "step": 100
    },
    {
      "epoch": 0.14287074959519955,
      "grad_norm": 0.47030845284461975,
      "learning_rate": 1.9542494042891185e-05,
      "loss": 1.9974,
      "mean_token_accuracy": 0.5917157411575318,
      "num_tokens": 60933259.0,
      "step": 150
    },
    {
      "epoch": 0.1904943327935994,
      "grad_norm": 0.3429001271724701,
      "learning_rate": 1.9383637807783958e-05,
      "loss": 1.9684,
      "mean_token_accuracy": 0.5942359437346458,
      "num_tokens": 81247806.0,
      "step": 200
    },
    {
      "epoch": 0.23811791599199925,
      "grad_norm": 0.30654239654541016,
      "learning_rate": 1.922478157267673e-05,
      "loss": 1.9526,
      "mean_token_accuracy": 0.596076726436615,
      "num_tokens": 101548593.0,
      "step": 250
    },
    {
      "epoch": 0.2857414991903991,
      "grad_norm": 0.3243563771247864,
      "learning_rate": 1.9065925337569502e-05,
      "loss": 1.9427,
      "mean_token_accuracy": 0.5970322663784027,
      "num_tokens": 121859649.0,
      "step": 300
    },
    {
      "epoch": 0.3333650823887989,
      "grad_norm": 0.29715102910995483,
      "learning_rate": 1.890706910246227e-05,
      "loss": 1.9305,
      "mean_token_accuracy": 0.5985876924693584,
      "num_tokens": 142175777.0,
      "step": 350
    },
    {
      "epoch": 0.3809886655871988,
      "grad_norm": 0.274366557598114,
      "learning_rate": 1.8748212867355044e-05,
      "loss": 1.9216,
      "mean_token_accuracy": 0.5999910009205341,
      "num_tokens": 162483788.0,
      "step": 400
    },
    {
      "epoch": 0.4286122487855986,
      "grad_norm": 0.35158780217170715,
      "learning_rate": 1.8589356632247816e-05,
      "loss": 1.9113,
      "mean_token_accuracy": 0.6015890350043773,
      "num_tokens": 182787334.0,
      "step": 450
    },
    {
      "epoch": 0.4762358319839985,
      "grad_norm": 0.290058970451355,
      "learning_rate": 1.843050039714059e-05,
      "loss": 1.9046,
      "mean_token_accuracy": 0.6026307411491871,
      "num_tokens": 203103375.0,
      "step": 500
    },
    {
      "epoch": 0.5238594151823983,
      "grad_norm": 0.3113633692264557,
      "learning_rate": 1.827164416203336e-05,
      "loss": 1.8955,
      "mean_token_accuracy": 0.6038400300741196,
      "num_tokens": 223412992.0,
      "step": 550
    },
    {
      "epoch": 0.5714829983807982,
      "grad_norm": 0.3326061964035034,
      "learning_rate": 1.8112787926926133e-05,
      "loss": 1.8894,
      "mean_token_accuracy": 0.604397124171257,
      "num_tokens": 243711742.0,
      "step": 600
    },
    {
      "epoch": 0.6191065815791981,
      "grad_norm": 0.3281303942203522,
      "learning_rate": 1.7953931691818906e-05,
      "loss": 1.8864,
      "mean_token_accuracy": 0.6047344071567059,
      "num_tokens": 264020914.0,
      "step": 650
    },
    {
      "epoch": 0.6667301647775978,
      "grad_norm": 0.258095920085907,
      "learning_rate": 1.7795075456711678e-05,
      "loss": 1.8765,
      "mean_token_accuracy": 0.6064859114587307,
      "num_tokens": 284324764.0,
      "step": 700
    },
    {
      "epoch": 0.7143537479759977,
      "grad_norm": 0.2982522249221802,
      "learning_rate": 1.763621922160445e-05,
      "loss": 1.88,
      "mean_token_accuracy": 0.6056106666326523,
      "num_tokens": 304637938.0,
      "step": 750
    },
    {
      "epoch": 0.7619773311743976,
      "grad_norm": 0.2931162416934967,
      "learning_rate": 1.7477362986497223e-05,
      "loss": 1.8727,
      "mean_token_accuracy": 0.6068724053800106,
      "num_tokens": 324946960.0,
      "step": 800
    },
    {
      "epoch": 0.8096009143727975,
      "grad_norm": 0.2975282371044159,
      "learning_rate": 1.7318506751389995e-05,
      "loss": 1.8714,
      "mean_token_accuracy": 0.6067778829336167,
      "num_tokens": 345256190.0,
      "step": 850
    },
    {
      "epoch": 0.8572244975711972,
      "grad_norm": 0.2823193073272705,
      "learning_rate": 1.7159650516282767e-05,
      "loss": 1.8683,
      "mean_token_accuracy": 0.6071519528925419,
      "num_tokens": 365565286.0,
      "step": 900
    },
    {
      "epoch": 0.9048480807695971,
      "grad_norm": 0.30169978737831116,
      "learning_rate": 1.7000794281175536e-05,
      "loss": 1.8564,
      "mean_token_accuracy": 0.6089305272996426,
      "num_tokens": 385873629.0,
      "step": 950
    },
    {
      "epoch": 0.952471663967997,
      "grad_norm": 0.25320759415626526,
      "learning_rate": 1.684193804606831e-05,
      "loss": 1.8566,
      "mean_token_accuracy": 0.6090085391402245,
      "num_tokens": 406182430.0,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.27295562624931335,
      "learning_rate": 1.668308181096108e-05,
      "loss": 1.8501,
      "mean_token_accuracy": 0.6099938274266008,
      "num_tokens": 426442647.0,
      "step": 1050
    },
    {
      "epoch": 1.0476235831983998,
      "grad_norm": 0.30006057024002075,
      "learning_rate": 1.6524225575853854e-05,
      "loss": 1.8486,
      "mean_token_accuracy": 0.6099475947320461,
      "num_tokens": 446749239.0,
      "step": 1100
    },
    {
      "epoch": 1.0952471663967998,
      "grad_norm": 0.2858481705188751,
      "learning_rate": 1.6365369340746626e-05,
      "loss": 1.8458,
      "mean_token_accuracy": 0.61038392162323,
      "num_tokens": 467052014.0,
      "step": 1150
    },
    {
      "epoch": 1.1428707495951995,
      "grad_norm": 0.27023717761039734,
      "learning_rate": 1.6206513105639398e-05,
      "loss": 1.8438,
      "mean_token_accuracy": 0.6108892421722412,
      "num_tokens": 487361380.0,
      "step": 1200
    },
    {
      "epoch": 1.1904943327935995,
      "grad_norm": 0.2528296113014221,
      "learning_rate": 1.6047656870532167e-05,
      "loss": 1.8439,
      "mean_token_accuracy": 0.610313931196928,
      "num_tokens": 507679299.0,
      "step": 1250
    },
    {
      "epoch": 1.2381179159919993,
      "grad_norm": 0.2592320740222931,
      "learning_rate": 1.5888800635424943e-05,
      "loss": 1.841,
      "mean_token_accuracy": 0.6109654635787011,
      "num_tokens": 527989121.0,
      "step": 1300
    },
    {
      "epoch": 1.285741499190399,
      "grad_norm": 0.2790074348449707,
      "learning_rate": 1.5729944400317715e-05,
      "loss": 1.8382,
      "mean_token_accuracy": 0.6114664600789547,
      "num_tokens": 548300795.0,
      "step": 1350
    },
    {
      "epoch": 1.333365082388799,
      "grad_norm": 0.3099632263183594,
      "learning_rate": 1.5571088165210488e-05,
      "loss": 1.832,
      "mean_token_accuracy": 0.6125494522750378,
      "num_tokens": 568607989.0,
      "step": 1400
    },
    {
      "epoch": 1.3809886655871988,
      "grad_norm": 0.23622950911521912,
      "learning_rate": 1.541223193010326e-05,
      "loss": 1.8343,
      "mean_token_accuracy": 0.6117785506546497,
      "num_tokens": 588914448.0,
      "step": 1450
    },
    {
      "epoch": 1.4286122487855986,
      "grad_norm": 0.2750089168548584,
      "learning_rate": 1.5253375694996029e-05,
      "loss": 1.8294,
      "mean_token_accuracy": 0.6126738254725933,
      "num_tokens": 609223320.0,
      "step": 1500
    },
    {
      "epoch": 1.4762358319839985,
      "grad_norm": 0.2539812922477722,
      "learning_rate": 1.5094519459888801e-05,
      "loss": 1.8274,
      "mean_token_accuracy": 0.6131484903097153,
      "num_tokens": 20316041.0,
      "step": 1550
    },
    {
      "epoch": 1.5238594151823983,
      "grad_norm": 0.26810041069984436,
      "learning_rate": 1.4935663224781574e-05,
      "loss": 1.823,
      "mean_token_accuracy": 0.6137671604454518,
      "num_tokens": 40625658.0,
      "step": 1600
    },
    {
      "epoch": 1.571482998380798,
      "grad_norm": 0.28397223353385925,
      "learning_rate": 1.4776806989674346e-05,
      "loss": 1.8224,
      "mean_token_accuracy": 0.6135454473793507,
      "num_tokens": 60924408.0,
      "step": 1650
    },
    {
      "epoch": 1.619106581579198,
      "grad_norm": 0.2595246434211731,
      "learning_rate": 1.4617950754567119e-05,
      "loss": 1.8226,
      "mean_token_accuracy": 0.613398359388113,
      "num_tokens": 81233580.0,
      "step": 1700
    },
    {
      "epoch": 1.6667301647775978,
      "grad_norm": 0.2445673793554306,
      "learning_rate": 1.4459094519459891e-05,
      "loss": 1.8158,
      "mean_token_accuracy": 0.6146850402355194,
      "num_tokens": 101537430.0,
      "step": 1750
    },
    {
      "epoch": 1.7143537479759976,
      "grad_norm": 0.2608117461204529,
      "learning_rate": 1.4300238284352662e-05,
      "loss": 1.8219,
      "mean_token_accuracy": 0.6135120939016342,
      "num_tokens": 121850604.0,
      "step": 1800
    },
    {
      "epoch": 1.7619773311743976,
      "grad_norm": 0.26758599281311035,
      "learning_rate": 1.4141382049245434e-05,
      "loss": 1.8171,
      "mean_token_accuracy": 0.6143426752090454,
      "num_tokens": 142159626.0,
      "step": 1850
    },
    {
      "epoch": 1.8096009143727976,
      "grad_norm": 0.25524309277534485,
      "learning_rate": 1.3982525814138206e-05,
      "loss": 1.8178,
      "mean_token_accuracy": 0.6139899947345256,
      "num_tokens": 162468856.0,
      "step": 1900
    },
    {
      "epoch": 1.8572244975711971,
      "grad_norm": 0.26049578189849854,
      "learning_rate": 1.3823669579030979e-05,
      "loss": 1.8168,
      "mean_token_accuracy": 0.6140655097961426,
      "num_tokens": 182777952.0,
      "step": 1950
    },
    {
      "epoch": 1.904848080769597,
      "grad_norm": 0.2717042863368988,
      "learning_rate": 1.3664813343923751e-05,
      "loss": 1.8075,
      "mean_token_accuracy": 0.6155479912161828,
      "num_tokens": 203086295.0,
      "step": 2000
    },
    {
      "epoch": 1.952471663967997,
      "grad_norm": 0.24529357254505157,
      "learning_rate": 1.3505957108816523e-05,
      "loss": 1.8089,
      "mean_token_accuracy": 0.6154809602499008,
      "num_tokens": 223395096.0,
      "step": 2050
    },
    {
      "epoch": 2.000952471663968,
      "grad_norm": 0.35854172706604004,
      "learning_rate": 1.3347100873709294e-05,
      "loss": 1.8407,
      "mean_token_accuracy": 0.6161111466484126,
      "num_tokens": 244064913.0,
      "step": 2100
    },
    {
      "epoch": 2.0485760548623677,
      "grad_norm": 0.23567642271518707,
      "learning_rate": 1.3188244638602066e-05,
      "loss": 1.8052,
      "mean_token_accuracy": 0.6162393952012062,
      "num_tokens": 264375896.0,
      "step": 2150
    },
    {
      "epoch": 2.0961996380607677,
      "grad_norm": 0.2530265152454376,
      "learning_rate": 1.3029388403494839e-05,
      "loss": 1.8034,
      "mean_token_accuracy": 0.6161407787799835,
      "num_tokens": 284687901.0,
      "step": 2200
    },
    {
      "epoch": 2.1438232212591677,
      "grad_norm": 0.2370193600654602,
      "learning_rate": 1.2870532168387611e-05,
      "loss": 1.8033,
      "mean_token_accuracy": 0.6161760824322701,
      "num_tokens": 304988460.0,
      "step": 2250
    },
    {
      "epoch": 2.1914468044575672,
      "grad_norm": 0.2497035413980484,
      "learning_rate": 1.2711675933280384e-05,
      "loss": 1.8068,
      "mean_token_accuracy": 0.6154370903074742,
      "num_tokens": 325293299.0,
      "step": 2300
    },
    {
      "epoch": 2.239070387655967,
      "grad_norm": 0.23987485468387604,
      "learning_rate": 1.2552819698173156e-05,
      "loss": 1.8033,
      "mean_token_accuracy": 0.6162745767831802,
      "num_tokens": 345604102.0,
      "step": 2350
    },
    {
      "epoch": 2.286693970854367,
      "grad_norm": 0.23024563491344452,
      "learning_rate": 1.2393963463065925e-05,
      "loss": 1.7997,
      "mean_token_accuracy": 0.6168436284661293,
      "num_tokens": 365917355.0,
      "step": 2400
    },
    {
      "epoch": 2.3343175540527668,
      "grad_norm": 0.24297727644443512,
      "learning_rate": 1.2235107227958697e-05,
      "loss": 1.8001,
      "mean_token_accuracy": 0.6166748906970024,
      "num_tokens": 386222500.0,
      "step": 2450
    },
    {
      "epoch": 2.3819411372511667,
      "grad_norm": 0.23780129849910736,
      "learning_rate": 1.207625099285147e-05,
      "loss": 1.7973,
      "mean_token_accuracy": 0.6169657715559006,
      "num_tokens": 406530949.0,
      "step": 2500
    },
    {
      "epoch": 2.4286122487855986,
      "grad_norm": 0.2644389867782593,
      "learning_rate": 1.1917394757744242e-05,
      "loss": 1.7985,
      "mean_token_accuracy": 0.6168324590623379,
      "num_tokens": 20303546.0,
      "step": 2550
    },
    {
      "epoch": 2.4762358319839985,
      "grad_norm": 0.23990584909915924,
      "learning_rate": 1.1758538522637014e-05,
      "loss": 1.7939,
      "mean_token_accuracy": 0.6177124035656452,
      "num_tokens": 40619587.0,
      "step": 2600
    },
    {
      "epoch": 2.5238594151823985,
      "grad_norm": 0.25453951954841614,
      "learning_rate": 1.1599682287529785e-05,
      "loss": 1.7904,
      "mean_token_accuracy": 0.6182607997953892,
      "num_tokens": 60929204.0,
      "step": 2650
    },
    {
      "epoch": 2.571482998380798,
      "grad_norm": 0.2683112919330597,
      "learning_rate": 1.1440826052422557e-05,
      "loss": 1.7911,
      "mean_token_accuracy": 0.6178435170054436,
      "num_tokens": 81227954.0,
      "step": 2700
    },
    {
      "epoch": 2.619106581579198,
      "grad_norm": 0.24393855035305023,
      "learning_rate": 1.128196981731533e-05,
      "loss": 1.7921,
      "mean_token_accuracy": 0.6176374254822731,
      "num_tokens": 101537126.0,
      "step": 2750
    },
    {
      "epoch": 2.666730164777598,
      "grad_norm": 0.2376239001750946,
      "learning_rate": 1.1123113582208102e-05,
      "loss": 1.7862,
      "mean_token_accuracy": 0.6187356041073799,
      "num_tokens": 121840976.0,
      "step": 2800
    },
    {
      "epoch": 2.7143537479759976,
      "grad_norm": 0.24735721945762634,
      "learning_rate": 1.0964257347100875e-05,
      "loss": 1.7929,
      "mean_token_accuracy": 0.617545694053173,
      "num_tokens": 142154150.0,
      "step": 2850
    },
    {
      "epoch": 2.7619773311743976,
      "grad_norm": 0.2544368803501129,
      "learning_rate": 1.0805401111993647e-05,
      "loss": 1.789,
      "mean_token_accuracy": 0.6183735141754151,
      "num_tokens": 162463172.0,
      "step": 2900
    },
    {
      "epoch": 2.8096009143727976,
      "grad_norm": 0.23869812488555908,
      "learning_rate": 1.0646544876886418e-05,
      "loss": 1.7901,
      "mean_token_accuracy": 0.6179508809447288,
      "num_tokens": 182772402.0,
      "step": 2950
    },
    {
      "epoch": 2.857224497571197,
      "grad_norm": 0.2459404468536377,
      "learning_rate": 1.048768864177919e-05,
      "loss": 1.7898,
      "mean_token_accuracy": 0.6179825856387615,
      "num_tokens": 203081498.0,
      "step": 3000
    },
    {
      "epoch": 2.904848080769597,
      "grad_norm": 0.25019368529319763,
      "learning_rate": 1.0328832406671962e-05,
      "loss": 1.7817,
      "mean_token_accuracy": 0.6192810923457146,
      "num_tokens": 223389841.0,
      "step": 3050
    },
    {
      "epoch": 2.952471663967997,
      "grad_norm": 0.23745620250701904,
      "learning_rate": 1.0169976171564735e-05,
      "loss": 1.7833,
      "mean_token_accuracy": 0.6191711721122265,
      "num_tokens": 243698642.0,
      "step": 3100
    },
    {
      "epoch": 3.000952471663968,
      "grad_norm": 0.3401544392108917,
      "learning_rate": 1.0011119936457507e-05,
      "loss": 1.8158,
      "mean_token_accuracy": 0.6196168116302996,
      "num_tokens": 264368459.0,
      "step": 3150
    },
    {
      "epoch": 3.0485760548623677,
      "grad_norm": 0.23939253389835358,
      "learning_rate": 9.85226370135028e-06,
      "loss": 1.7767,
      "mean_token_accuracy": 0.6204061754941941,
      "num_tokens": 284674076.0,
      "step": 3200
    },
    {
      "epoch": 3.0961996380607677,
      "grad_norm": 0.2439105212688446,
      "learning_rate": 9.693407466243052e-06,
      "loss": 1.7815,
      "mean_token_accuracy": 0.6191436488628388,
      "num_tokens": 304985319.0,
      "step": 3250
    },
    {
      "epoch": 3.1438232212591677,
      "grad_norm": 0.24051690101623535,
      "learning_rate": 9.534551231135823e-06,
      "loss": 1.7826,
      "mean_token_accuracy": 0.6190176220238208,
      "num_tokens": 325293877.0,
      "step": 3300
    },
    {
      "epoch": 3.1914468044575672,
      "grad_norm": 0.24046750366687775,
      "learning_rate": 9.375694996028595e-06,
      "loss": 1.7807,
      "mean_token_accuracy": 0.619534037321806,
      "num_tokens": 345604610.0,
      "step": 3350
    },
    {
      "epoch": 3.239070387655967,
      "grad_norm": 0.23331508040428162,
      "learning_rate": 9.216838760921367e-06,
      "loss": 1.7815,
      "mean_token_accuracy": 0.6193765329420566,
      "num_tokens": 365908151.0,
      "step": 3400
    },
    {
      "epoch": 3.286693970854367,
      "grad_norm": 0.24685585498809814,
      "learning_rate": 9.057982525814138e-06,
      "loss": 1.7833,
      "mean_token_accuracy": 0.6190218808948994,
      "num_tokens": 386213379.0,
      "step": 3450
    },
    {
      "epoch": 3.3343175540527668,
      "grad_norm": 0.24293000996112823,
      "learning_rate": 8.89912629070691e-06,
      "loss": 1.7812,
      "mean_token_accuracy": 0.6195805520415306,
      "num_tokens": 406527411.0,
      "step": 3500
    },
    {
      "epoch": 3.3819411372511667,
      "grad_norm": 0.24087201058864594,
      "learning_rate": 8.740270055599683e-06,
      "loss": 1.7785,
      "mean_token_accuracy": 0.6196017529964447,
      "num_tokens": 426842256.0,
      "step": 3550
    },
    {
      "epoch": 3.4295647204495667,
      "grad_norm": 0.23481643199920654,
      "learning_rate": 8.581413820492455e-06,
      "loss": 1.781,
      "mean_token_accuracy": 0.6192840512394905,
      "num_tokens": 447149406.0,
      "step": 3600
    },
    {
      "epoch": 3.4771883036479663,
      "grad_norm": 0.24832718074321747,
      "learning_rate": 8.422557585385227e-06,
      "loss": 1.7804,
      "mean_token_accuracy": 0.6194921168982983,
      "num_tokens": 467457805.0,
      "step": 3650
    },
    {
      "epoch": 3.5248118868463663,
      "grad_norm": 0.2421002835035324,
      "learning_rate": 8.263701350278e-06,
      "loss": 1.778,
      "mean_token_accuracy": 0.6197192951738835,
      "num_tokens": 487763203.0,
      "step": 3700
    },
    {
      "epoch": 3.5724354700447662,
      "grad_norm": 0.2387075573205948,
      "learning_rate": 8.10484511517077e-06,
      "loss": 1.7767,
      "mean_token_accuracy": 0.62002942019701,
      "num_tokens": 508073422.0,
      "step": 3750
    },
    {
      "epoch": 3.620059053243166,
      "grad_norm": 0.24657268822193146,
      "learning_rate": 7.945988880063543e-06,
      "loss": 1.7739,
      "mean_token_accuracy": 0.6202718901336193,
      "num_tokens": 528379381.0,
      "step": 3800
    },
    {
      "epoch": 3.667682636441566,
      "grad_norm": 0.24887719750404358,
      "learning_rate": 7.787132644956315e-06,
      "loss": 1.7777,
      "mean_token_accuracy": 0.6198979534208775,
      "num_tokens": 548681613.0,
      "step": 3850
    },
    {
      "epoch": 3.7153062196399658,
      "grad_norm": 0.2520405054092407,
      "learning_rate": 7.628276409849087e-06,
      "loss": 1.7734,
      "mean_token_accuracy": 0.6207626007199287,
      "num_tokens": 568984321.0,
      "step": 3900
    },
    {
      "epoch": 3.7629298028383653,
      "grad_norm": 0.25676149129867554,
      "learning_rate": 7.469420174741859e-06,
      "loss": 1.7739,
      "mean_token_accuracy": 0.6203771130740643,
      "num_tokens": 589297969.0,
      "step": 3950
    },
    {
      "epoch": 3.8105533860367653,
      "grad_norm": 0.2468143105506897,
      "learning_rate": 7.3105639396346315e-06,
      "loss": 1.7755,
      "mean_token_accuracy": 0.6202954652607441,
      "num_tokens": 609609102.0,
      "step": 4000
    },
    {
      "epoch": 3.8581769692351653,
      "grad_norm": 0.24929603934288025,
      "learning_rate": 7.151707704527403e-06,
      "loss": 1.7739,
      "mean_token_accuracy": 0.6202192025780677,
      "num_tokens": 629919095.0,
      "step": 4050
    },
    {
      "epoch": 3.905800552433565,
      "grad_norm": 0.24869875609874725,
      "learning_rate": 6.992851469420175e-06,
      "loss": 1.7773,
      "mean_token_accuracy": 0.6198899661004543,
      "num_tokens": 650226880.0,
      "step": 4100
    },
    {
      "epoch": 3.953424135631965,
      "grad_norm": 0.24287043511867523,
      "learning_rate": 6.833995234312948e-06,
      "loss": 1.7683,
      "mean_token_accuracy": 0.621512754380703,
      "num_tokens": 670539372.0,
      "step": 4150
    },
    {
      "epoch": 4.000952471663968,
      "grad_norm": 0.2520332336425781,
      "learning_rate": 6.675138999205719e-06,
      "loss": 1.7797,
      "mean_token_accuracy": 0.6192158266513763,
      "num_tokens": 690811106.0,
      "step": 4200
    },
    {
      "epoch": 4.048576054862368,
      "grad_norm": 0.2522871494293213,
      "learning_rate": 6.516282764098492e-06,
      "loss": 1.7702,
      "mean_token_accuracy": 0.6210551818013191,
      "num_tokens": 711113716.0,
      "step": 4250
    },
    {
      "epoch": 4.096199638060767,
      "grad_norm": 0.25319358706474304,
      "learning_rate": 6.357426528991263e-06,
      "loss": 1.7687,
      "mean_token_accuracy": 0.6209404880702496,
      "num_tokens": 731430179.0,
      "step": 4300
    },
    {
      "epoch": 4.143823221259168,
      "grad_norm": 0.2546616792678833,
      "learning_rate": 6.1985702938840355e-06,
      "loss": 1.7665,
      "mean_token_accuracy": 0.6217049216628074,
      "num_tokens": 751740836.0,
      "step": 4350
    },
    {
      "epoch": 4.191446804457567,
      "grad_norm": 0.24473318457603455,
      "learning_rate": 6.039714058776808e-06,
      "loss": 1.7699,
      "mean_token_accuracy": 0.6211777364015579,
      "num_tokens": 772047032.0,
      "step": 4400
    },
    {
      "epoch": 4.239070387655968,
      "grad_norm": 0.24759261310100555,
      "learning_rate": 5.8808578236695794e-06,
      "loss": 1.77,
      "mean_token_accuracy": 0.6208123913705349,
      "num_tokens": 792358278.0,
      "step": 4450
    },
    {
      "epoch": 4.286693970854367,
      "grad_norm": 0.2464452087879181,
      "learning_rate": 5.722001588562352e-06,
      "loss": 1.7684,
      "mean_token_accuracy": 0.6211452582776547,
      "num_tokens": 812665585.0,
      "step": 4500
    },
    {
      "epoch": 4.334317554052767,
      "grad_norm": 0.27063560485839844,
      "learning_rate": 5.563145353455124e-06,
      "loss": 1.7671,
      "mean_token_accuracy": 0.6214230598509312,
      "num_tokens": 832970219.0,
      "step": 4550
    },
    {
      "epoch": 4.381941137251167,
      "grad_norm": 0.25516068935394287,
      "learning_rate": 5.404289118347895e-06,
      "loss": 1.7673,
      "mean_token_accuracy": 0.6212706764042377,
      "num_tokens": 853280670.0,
      "step": 4600
    },
    {
      "epoch": 4.429564720449567,
      "grad_norm": 0.25876569747924805,
      "learning_rate": 5.245432883240667e-06,
      "loss": 1.7694,
      "mean_token_accuracy": 0.6209256626665592,
      "num_tokens": 873582095.0,
      "step": 4650
    },
    {
      "epoch": 4.477188303647966,
      "grad_norm": 0.252450168132782,
      "learning_rate": 5.08657664813344e-06,
      "loss": 1.7672,
      "mean_token_accuracy": 0.6213706638813019,
      "num_tokens": 893886676.0,
      "step": 4700
    },
    {
      "epoch": 4.524811886846367,
      "grad_norm": 0.26232871413230896,
      "learning_rate": 4.927720413026212e-06,
      "loss": 1.7652,
      "mean_token_accuracy": 0.6216549359858036,
      "num_tokens": 914201383.0,
      "step": 4750
    },
    {
      "epoch": 4.572435470044766,
      "grad_norm": 0.25896576046943665,
      "learning_rate": 4.7688641779189835e-06,
      "loss": 1.7668,
      "mean_token_accuracy": 0.6213670527338981,
      "num_tokens": 934510978.0,
      "step": 4800
    },
    {
      "epoch": 4.620059053243166,
      "grad_norm": 0.2614849805831909,
      "learning_rate": 4.610007942811756e-06,
      "loss": 1.7693,
      "mean_token_accuracy": 0.6207684008777141,
      "num_tokens": 954821977.0,
      "step": 4850
    },
    {
      "epoch": 4.667682636441566,
      "grad_norm": 0.26850801706314087,
      "learning_rate": 4.451151707704528e-06,
      "loss": 1.7641,
      "mean_token_accuracy": 0.621698461562395,
      "num_tokens": 975134425.0,
      "step": 4900
    },
    {
      "epoch": 4.715306219639966,
      "grad_norm": 0.2754788398742676,
      "learning_rate": 4.2922954725973e-06,
      "loss": 1.763,
      "mean_token_accuracy": 0.6220386314690113,
      "num_tokens": 995444510.0,
      "step": 4950
    },
    {
      "epoch": 4.762929802838365,
      "grad_norm": 0.26721155643463135,
      "learning_rate": 4.133439237490072e-06,
      "loss": 1.7646,
      "mean_token_accuracy": 0.6219381015598774,
      "num_tokens": 1015754448.0,
      "step": 5000
    },
    {
      "epoch": 4.810553386036766,
      "grad_norm": 0.25806230306625366,
      "learning_rate": 3.9745830023828445e-06,
      "loss": 1.7643,
      "mean_token_accuracy": 0.6215842329263687,
      "num_tokens": 1036064491.0,
      "step": 5050
    },
    {
      "epoch": 4.858176969235165,
      "grad_norm": 0.27059823274612427,
      "learning_rate": 3.815726767275616e-06,
      "loss": 1.769,
      "mean_token_accuracy": 0.6209043772816658,
      "num_tokens": 1056373154.0,
      "step": 5100
    },
    {
      "epoch": 4.905800552433565,
      "grad_norm": 0.2654803395271301,
      "learning_rate": 3.656870532168388e-06,
      "loss": 1.7632,
      "mean_token_accuracy": 0.6219344958364964,
      "num_tokens": 1076683766.0,
      "step": 5150
    },
    {
      "epoch": 4.953424135631965,
      "grad_norm": 0.2545090913772583,
      "learning_rate": 3.49801429706116e-06,
      "loss": 1.7623,
      "mean_token_accuracy": 0.6223813739717007,
      "num_tokens": 1096993338.0,
      "step": 5200
    },
    {
      "epoch": 5.000952471663968,
      "grad_norm": 0.24684475362300873,
      "learning_rate": 3.3391580619539323e-06,
      "loss": 1.7634,
      "mean_token_accuracy": 0.6217962507016673,
      "num_tokens": 1117253753.0,
      "step": 5250
    },
    {
      "epoch": 5.048576054862368,
      "grad_norm": 0.25752031803131104,
      "learning_rate": 3.180301826846704e-06,
      "loss": 1.7649,
      "mean_token_accuracy": 0.6215447184741497,
      "num_tokens": 1137562518.0,
      "step": 5300
    },
    {
      "epoch": 5.096199638060767,
      "grad_norm": 0.25249016284942627,
      "learning_rate": 3.0214455917394757e-06,
      "loss": 1.7628,
      "mean_token_accuracy": 0.6217640188038349,
      "num_tokens": 1157872138.0,
      "step": 5350
    },
    {
      "epoch": 5.143823221259168,
      "grad_norm": 0.2576809823513031,
      "learning_rate": 2.862589356632248e-06,
      "loss": 1.7609,
      "mean_token_accuracy": 0.6222141864299774,
      "num_tokens": 1178186376.0,
      "step": 5400
    },
    {
      "epoch": 5.191446804457567,
      "grad_norm": 0.2503405809402466,
      "learning_rate": 2.70373312152502e-06,
      "loss": 1.7608,
      "mean_token_accuracy": 0.6223511473238468,
      "num_tokens": 1198490885.0,
      "step": 5450
    },
    {
      "epoch": 5.239070387655968,
      "grad_norm": 0.2567017376422882,
      "learning_rate": 2.544876886417792e-06,
      "loss": 1.7633,
      "mean_token_accuracy": 0.6218047961294652,
      "num_tokens": 1218801541.0,
      "step": 5500
    },
    {
      "epoch": 5.286693970854367,
      "grad_norm": 0.2591474652290344,
      "learning_rate": 2.386020651310564e-06,
      "loss": 1.7642,
      "mean_token_accuracy": 0.6217405062913894,
      "num_tokens": 1239114262.0,
      "step": 5550
    },
    {
      "epoch": 5.334317554052767,
      "grad_norm": 0.24408093094825745,
      "learning_rate": 2.2271644162033363e-06,
      "loss": 1.7585,
      "mean_token_accuracy": 0.6226994139850139,
      "num_tokens": 1259423857.0,
      "step": 5600
    },
    {
      "epoch": 5.381941137251167,
      "grad_norm": 0.25508007407188416,
      "learning_rate": 2.0683081810961083e-06,
      "loss": 1.7608,
      "mean_token_accuracy": 0.6223227469921112,
      "num_tokens": 1279728024.0,
      "step": 5650
    },
    {
      "epoch": 5.429564720449567,
      "grad_norm": 0.2491166740655899,
      "learning_rate": 1.90945194598888e-06,
      "loss": 1.7546,
      "mean_token_accuracy": 0.6232103821635246,
      "num_tokens": 1300046077.0,
      "step": 5700
    },
    {
      "epoch": 5.477188303647966,
      "grad_norm": 0.24481193721294403,
      "learning_rate": 1.7505957108816524e-06,
      "loss": 1.7626,
      "mean_token_accuracy": 0.6218702466189862,
      "num_tokens": 1320360752.0,
      "step": 5750
    },
    {
      "epoch": 5.524811886846367,
      "grad_norm": 0.2409469485282898,
      "learning_rate": 1.5917394757744243e-06,
      "loss": 1.7642,
      "mean_token_accuracy": 0.6215654002428055,
      "num_tokens": 1340669691.0,
      "step": 5800
    },
    {
      "epoch": 5.572435470044766,
      "grad_norm": 0.2315661460161209,
      "learning_rate": 1.4328832406671963e-06,
      "loss": 1.7647,
      "mean_token_accuracy": 0.6213131138384342,
      "num_tokens": 1360981891.0,
      "step": 5850
    },
    {
      "epoch": 5.620059053243166,
      "grad_norm": 0.23747174441814423,
      "learning_rate": 1.2740270055599682e-06,
      "loss": 1.7566,
      "mean_token_accuracy": 0.6231028576791287,
      "num_tokens": 1381283444.0,
      "step": 5900
    },
    {
      "epoch": 5.667682636441566,
      "grad_norm": 0.23048564791679382,
      "learning_rate": 1.1151707704527404e-06,
      "loss": 1.7589,
      "mean_token_accuracy": 0.6227406437397003,
      "num_tokens": 1401595655.0,
      "step": 5950
    },
    {
      "epoch": 5.715306219639966,
      "grad_norm": 0.22774732112884521,
      "learning_rate": 9.563145353455123e-07,
      "loss": 1.7608,
      "mean_token_accuracy": 0.6220790773630143,
      "num_tokens": 1421898736.0,
      "step": 6000
    }
  ],
  "logging_steps": 50,
  "max_steps": 6300,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "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": 1.5552084524505498e+19,
  "train_batch_size": 10,
  "trial_name": null,
  "trial_params": null
}