{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.492197949175212,
  "eval_steps": 500,
  "global_step": 2520,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.017833259028087384,
      "grad_norm": 33.452999114990234,
      "learning_rate": 3.2142857142857143e-06,
      "loss": 8.9944,
      "mean_token_accuracy": 0.32701094649964946,
      "num_tokens": 2621440.0,
      "step": 10
    },
    {
      "epoch": 0.03566651805617477,
      "grad_norm": 2.596579074859619,
      "learning_rate": 6.785714285714285e-06,
      "loss": 4.9649,
      "mean_token_accuracy": 0.5848119974136352,
      "num_tokens": 5242880.0,
      "step": 20
    },
    {
      "epoch": 0.05349977708426215,
      "grad_norm": 1.8979898691177368,
      "learning_rate": 1.0357142857142859e-05,
      "loss": 4.1262,
      "mean_token_accuracy": 0.5906830102205276,
      "num_tokens": 7864320.0,
      "step": 30
    },
    {
      "epoch": 0.07133303611234953,
      "grad_norm": 1.1842058897018433,
      "learning_rate": 1.392857142857143e-05,
      "loss": 3.7327,
      "mean_token_accuracy": 0.608327355235815,
      "num_tokens": 10485760.0,
      "step": 40
    },
    {
      "epoch": 0.08916629514043692,
      "grad_norm": 0.922074556350708,
      "learning_rate": 1.75e-05,
      "loss": 3.4492,
      "mean_token_accuracy": 0.6281662203371525,
      "num_tokens": 13107200.0,
      "step": 50
    },
    {
      "epoch": 0.1069995541685243,
      "grad_norm": 0.7813541889190674,
      "learning_rate": 2.107142857142857e-05,
      "loss": 3.1416,
      "mean_token_accuracy": 0.648241326212883,
      "num_tokens": 15728640.0,
      "step": 60
    },
    {
      "epoch": 0.12483281319661169,
      "grad_norm": 1.1468125581741333,
      "learning_rate": 2.4642857142857145e-05,
      "loss": 2.8778,
      "mean_token_accuracy": 0.6583510167896748,
      "num_tokens": 18350080.0,
      "step": 70
    },
    {
      "epoch": 0.14266607222469907,
      "grad_norm": 1.0490431785583496,
      "learning_rate": 2.8214285714285714e-05,
      "loss": 2.7108,
      "mean_token_accuracy": 0.663086223602295,
      "num_tokens": 20971520.0,
      "step": 80
    },
    {
      "epoch": 0.16049933125278645,
      "grad_norm": 0.7813755869865417,
      "learning_rate": 3.1785714285714284e-05,
      "loss": 2.6039,
      "mean_token_accuracy": 0.6650082431733608,
      "num_tokens": 23592960.0,
      "step": 90
    },
    {
      "epoch": 0.17833259028087384,
      "grad_norm": 0.6500654220581055,
      "learning_rate": 3.5357142857142864e-05,
      "loss": 2.4513,
      "mean_token_accuracy": 0.674950385093689,
      "num_tokens": 26214400.0,
      "step": 100
    },
    {
      "epoch": 0.19616584930896122,
      "grad_norm": 0.81620192527771,
      "learning_rate": 3.892857142857143e-05,
      "loss": 2.3602,
      "mean_token_accuracy": 0.6777540303766727,
      "num_tokens": 28835840.0,
      "step": 110
    },
    {
      "epoch": 0.2139991083370486,
      "grad_norm": 0.6131238341331482,
      "learning_rate": 4.25e-05,
      "loss": 2.3545,
      "mean_token_accuracy": 0.6758881151676178,
      "num_tokens": 31457280.0,
      "step": 120
    },
    {
      "epoch": 0.231832367365136,
      "grad_norm": 0.9951689839363098,
      "learning_rate": 4.607142857142857e-05,
      "loss": 2.2624,
      "mean_token_accuracy": 0.6806462205946445,
      "num_tokens": 34078720.0,
      "step": 130
    },
    {
      "epoch": 0.24966562639322337,
      "grad_norm": 0.7796569466590881,
      "learning_rate": 4.964285714285715e-05,
      "loss": 2.1505,
      "mean_token_accuracy": 0.6918394304811954,
      "num_tokens": 36700160.0,
      "step": 140
    },
    {
      "epoch": 0.26749888542131073,
      "grad_norm": 0.8614610433578491,
      "learning_rate": 4.9998587698517684e-05,
      "loss": 2.1591,
      "mean_token_accuracy": 0.68697941750288,
      "num_tokens": 39321600.0,
      "step": 150
    },
    {
      "epoch": 0.28533214444939814,
      "grad_norm": 0.6178464889526367,
      "learning_rate": 4.9993705873562665e-05,
      "loss": 2.1048,
      "mean_token_accuracy": 0.6911555036902428,
      "num_tokens": 41943040.0,
      "step": 160
    },
    {
      "epoch": 0.3031654034774855,
      "grad_norm": 0.7255717515945435,
      "learning_rate": 4.998533777009954e-05,
      "loss": 2.0847,
      "mean_token_accuracy": 0.6920825429260731,
      "num_tokens": 44564480.0,
      "step": 170
    },
    {
      "epoch": 0.3209986625055729,
      "grad_norm": 0.7800336480140686,
      "learning_rate": 4.9973484555363726e-05,
      "loss": 2.049,
      "mean_token_accuracy": 0.6982546880841255,
      "num_tokens": 47185920.0,
      "step": 180
    },
    {
      "epoch": 0.33883192153366026,
      "grad_norm": 0.886745274066925,
      "learning_rate": 4.995814788271582e-05,
      "loss": 2.0301,
      "mean_token_accuracy": 0.697366189956665,
      "num_tokens": 49807360.0,
      "step": 190
    },
    {
      "epoch": 0.3566651805617477,
      "grad_norm": 0.6190801858901978,
      "learning_rate": 4.993932989141088e-05,
      "loss": 1.9408,
      "mean_token_accuracy": 0.7095360569655895,
      "num_tokens": 52428800.0,
      "step": 200
    },
    {
      "epoch": 0.37449843958983503,
      "grad_norm": 0.6333149671554565,
      "learning_rate": 4.991703320630011e-05,
      "loss": 1.9076,
      "mean_token_accuracy": 0.7127362459897995,
      "num_tokens": 55050240.0,
      "step": 210
    },
    {
      "epoch": 0.39233169861792244,
      "grad_norm": 0.9046627879142761,
      "learning_rate": 4.989126093746468e-05,
      "loss": 1.8913,
      "mean_token_accuracy": 0.7148448958992958,
      "num_tokens": 57671680.0,
      "step": 220
    },
    {
      "epoch": 0.4101649576460098,
      "grad_norm": 0.7778757214546204,
      "learning_rate": 4.986201667978193e-05,
      "loss": 1.9016,
      "mean_token_accuracy": 0.7118340134620667,
      "num_tokens": 60293120.0,
      "step": 230
    },
    {
      "epoch": 0.4279982166740972,
      "grad_norm": 0.6346383690834045,
      "learning_rate": 4.9829304512423944e-05,
      "loss": 1.898,
      "mean_token_accuracy": 0.7123855009675026,
      "num_tokens": 62914560.0,
      "step": 240
    },
    {
      "epoch": 0.44583147570218457,
      "grad_norm": 0.7308981418609619,
      "learning_rate": 4.979312899828853e-05,
      "loss": 1.8435,
      "mean_token_accuracy": 0.7184088006615639,
      "num_tokens": 65536000.0,
      "step": 250
    },
    {
      "epoch": 0.463664734730272,
      "grad_norm": 0.6174861788749695,
      "learning_rate": 4.9753495183362796e-05,
      "loss": 1.8242,
      "mean_token_accuracy": 0.7238672457635402,
      "num_tokens": 68157440.0,
      "step": 260
    },
    {
      "epoch": 0.48149799375835933,
      "grad_norm": 0.6507575511932373,
      "learning_rate": 4.971040859601927e-05,
      "loss": 1.791,
      "mean_token_accuracy": 0.7251060992479325,
      "num_tokens": 70778880.0,
      "step": 270
    },
    {
      "epoch": 0.49933125278644674,
      "grad_norm": 0.8092382550239563,
      "learning_rate": 4.966387524624482e-05,
      "loss": 1.7879,
      "mean_token_accuracy": 0.7255102708935738,
      "num_tokens": 73400320.0,
      "step": 280
    },
    {
      "epoch": 0.5171645118145342,
      "grad_norm": 0.5636172294616699,
      "learning_rate": 4.9613901624802245e-05,
      "loss": 1.7351,
      "mean_token_accuracy": 0.7330663032829762,
      "num_tokens": 76021760.0,
      "step": 290
    },
    {
      "epoch": 0.5349977708426215,
      "grad_norm": 0.655262291431427,
      "learning_rate": 4.956049470232503e-05,
      "loss": 1.7464,
      "mean_token_accuracy": 0.7302443347871304,
      "num_tokens": 78643200.0,
      "step": 300
    },
    {
      "epoch": 0.5528310298707089,
      "grad_norm": 0.9413332343101501,
      "learning_rate": 4.9503661928344944e-05,
      "loss": 1.7277,
      "mean_token_accuracy": 0.730958030372858,
      "num_tokens": 81264640.0,
      "step": 310
    },
    {
      "epoch": 0.5706642888987963,
      "grad_norm": 0.7439999580383301,
      "learning_rate": 4.9443411230252956e-05,
      "loss": 1.6947,
      "mean_token_accuracy": 0.7346593342721462,
      "num_tokens": 83886080.0,
      "step": 320
    },
    {
      "epoch": 0.5884975479268837,
      "grad_norm": 0.7665299773216248,
      "learning_rate": 4.93797510121935e-05,
      "loss": 1.7136,
      "mean_token_accuracy": 0.7324949607253075,
      "num_tokens": 86507520.0,
      "step": 330
    },
    {
      "epoch": 0.606330806954971,
      "grad_norm": 0.7366372346878052,
      "learning_rate": 4.931269015389217e-05,
      "loss": 1.676,
      "mean_token_accuracy": 0.7371297933161258,
      "num_tokens": 89128960.0,
      "step": 340
    },
    {
      "epoch": 0.6241640659830584,
      "grad_norm": 0.6571912169456482,
      "learning_rate": 4.9242238009417175e-05,
      "loss": 1.6252,
      "mean_token_accuracy": 0.7434912383556366,
      "num_tokens": 91750400.0,
      "step": 350
    },
    {
      "epoch": 0.6419973250111458,
      "grad_norm": 0.7496657967567444,
      "learning_rate": 4.9168404405874505e-05,
      "loss": 1.657,
      "mean_token_accuracy": 0.737925547361374,
      "num_tokens": 94371840.0,
      "step": 360
    },
    {
      "epoch": 0.6598305840392331,
      "grad_norm": 0.5275835394859314,
      "learning_rate": 4.9091199642037244e-05,
      "loss": 1.6723,
      "mean_token_accuracy": 0.7358783438801766,
      "num_tokens": 96993280.0,
      "step": 370
    },
    {
      "epoch": 0.6776638430673205,
      "grad_norm": 0.6143308281898499,
      "learning_rate": 4.901063448690901e-05,
      "loss": 1.6292,
      "mean_token_accuracy": 0.7412570238113403,
      "num_tokens": 99614720.0,
      "step": 380
    },
    {
      "epoch": 0.6954971020954079,
      "grad_norm": 0.860007643699646,
      "learning_rate": 4.892672017822183e-05,
      "loss": 1.6281,
      "mean_token_accuracy": 0.7422710858285427,
      "num_tokens": 102236160.0,
      "step": 390
    },
    {
      "epoch": 0.7133303611234953,
      "grad_norm": 0.5295814275741577,
      "learning_rate": 4.8839468420868606e-05,
      "loss": 1.6112,
      "mean_token_accuracy": 0.7446457460522652,
      "num_tokens": 104857600.0,
      "step": 400
    },
    {
      "epoch": 0.7311636201515826,
      "grad_norm": 0.7366165518760681,
      "learning_rate": 4.8748891385270495e-05,
      "loss": 1.6197,
      "mean_token_accuracy": 0.7417318046092987,
      "num_tokens": 107479040.0,
      "step": 410
    },
    {
      "epoch": 0.7489968791796701,
      "grad_norm": 0.7033573985099792,
      "learning_rate": 4.865500170567926e-05,
      "loss": 1.5763,
      "mean_token_accuracy": 0.7474127545952797,
      "num_tokens": 110100480.0,
      "step": 420
    },
    {
      "epoch": 0.7668301382077575,
      "grad_norm": 0.6527330875396729,
      "learning_rate": 4.855781247841498e-05,
      "loss": 1.5714,
      "mean_token_accuracy": 0.7477886863052845,
      "num_tokens": 112721920.0,
      "step": 430
    },
    {
      "epoch": 0.7846633972358449,
      "grad_norm": 0.7344551682472229,
      "learning_rate": 4.845733726003928e-05,
      "loss": 1.5682,
      "mean_token_accuracy": 0.7475516766309738,
      "num_tokens": 115343360.0,
      "step": 440
    },
    {
      "epoch": 0.8024966562639322,
      "grad_norm": 0.6840972304344177,
      "learning_rate": 4.835359006546443e-05,
      "loss": 1.5955,
      "mean_token_accuracy": 0.7427977688610554,
      "num_tokens": 117964800.0,
      "step": 450
    },
    {
      "epoch": 0.8203299152920196,
      "grad_norm": 0.635522723197937,
      "learning_rate": 4.824658536599836e-05,
      "loss": 1.5459,
      "mean_token_accuracy": 0.7526196874678135,
      "num_tokens": 120586240.0,
      "step": 460
    },
    {
      "epoch": 0.838163174320107,
      "grad_norm": 0.6874827146530151,
      "learning_rate": 4.8136338087326216e-05,
      "loss": 1.6108,
      "mean_token_accuracy": 0.7413325920701027,
      "num_tokens": 123207680.0,
      "step": 470
    },
    {
      "epoch": 0.8559964333481944,
      "grad_norm": 0.6571274995803833,
      "learning_rate": 4.8022863607428356e-05,
      "loss": 1.539,
      "mean_token_accuracy": 0.7521994858980179,
      "num_tokens": 125829120.0,
      "step": 480
    },
    {
      "epoch": 0.8738296923762817,
      "grad_norm": 0.8841336369514465,
      "learning_rate": 4.790617775443537e-05,
      "loss": 1.5486,
      "mean_token_accuracy": 0.7492702730000019,
      "num_tokens": 128450560.0,
      "step": 490
    },
    {
      "epoch": 0.8916629514043691,
      "grad_norm": 0.690990149974823,
      "learning_rate": 4.7786296804420274e-05,
      "loss": 1.5646,
      "mean_token_accuracy": 0.7455468386411667,
      "num_tokens": 131072000.0,
      "step": 500
    },
    {
      "epoch": 0.9094962104324565,
      "grad_norm": 0.7215222716331482,
      "learning_rate": 4.766323747912818e-05,
      "loss": 1.5085,
      "mean_token_accuracy": 0.754824897646904,
      "num_tokens": 133693440.0,
      "step": 510
    },
    {
      "epoch": 0.927329469460544,
      "grad_norm": 0.6024391651153564,
      "learning_rate": 4.753701694364392e-05,
      "loss": 1.4856,
      "mean_token_accuracy": 0.7570274345576763,
      "num_tokens": 136314880.0,
      "step": 520
    },
    {
      "epoch": 0.9451627284886313,
      "grad_norm": 0.6131455302238464,
      "learning_rate": 4.740765280399767e-05,
      "loss": 1.5433,
      "mean_token_accuracy": 0.7495988786220551,
      "num_tokens": 138936320.0,
      "step": 530
    },
    {
      "epoch": 0.9629959875167187,
      "grad_norm": 0.7875815629959106,
      "learning_rate": 4.72751631047092e-05,
      "loss": 1.4917,
      "mean_token_accuracy": 0.7559954360127449,
      "num_tokens": 141557760.0,
      "step": 540
    },
    {
      "epoch": 0.9808292465448061,
      "grad_norm": 0.7112794518470764,
      "learning_rate": 4.71395663262709e-05,
      "loss": 1.4784,
      "mean_token_accuracy": 0.7572461232542992,
      "num_tokens": 144179200.0,
      "step": 550
    },
    {
      "epoch": 0.9986625055728935,
      "grad_norm": 0.6807007789611816,
      "learning_rate": 4.7000881382570014e-05,
      "loss": 1.5159,
      "mean_token_accuracy": 0.7523536741733551,
      "num_tokens": 146800640.0,
      "step": 560
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.494895339012146,
      "eval_mean_token_accuracy": 0.7405715703964233,
      "eval_num_tokens": 146980864.0,
      "eval_runtime": 2.6056,
      "eval_samples_per_second": 27.633,
      "eval_steps_per_second": 1.919,
      "step": 561
    },
    {
      "epoch": 1.0160499331252786,
      "grad_norm": 0.5480894446372986,
      "learning_rate": 4.685912761825038e-05,
      "loss": 1.4428,
      "mean_token_accuracy": 0.7618689170250525,
      "num_tokens": 149340160.0,
      "step": 570
    },
    {
      "epoch": 1.033883192153366,
      "grad_norm": 0.6192435622215271,
      "learning_rate": 4.671432480601416e-05,
      "loss": 1.4711,
      "mean_token_accuracy": 0.7556988887488842,
      "num_tokens": 151961600.0,
      "step": 580
    },
    {
      "epoch": 1.0517164511814534,
      "grad_norm": 0.5988848209381104,
      "learning_rate": 4.656649314386378e-05,
      "loss": 1.433,
      "mean_token_accuracy": 0.7611474089324475,
      "num_tokens": 154583040.0,
      "step": 590
    },
    {
      "epoch": 1.0695497102095408,
      "grad_norm": 0.7051531672477722,
      "learning_rate": 4.641565325228464e-05,
      "loss": 1.415,
      "mean_token_accuracy": 0.7643376767635346,
      "num_tokens": 157204480.0,
      "step": 600
    },
    {
      "epoch": 1.0873829692376282,
      "grad_norm": 0.6170579791069031,
      "learning_rate": 4.6261826171368774e-05,
      "loss": 1.4674,
      "mean_token_accuracy": 0.7571297205984593,
      "num_tokens": 159825920.0,
      "step": 610
    },
    {
      "epoch": 1.1052162282657156,
      "grad_norm": 0.5691844820976257,
      "learning_rate": 4.6105033357880104e-05,
      "loss": 1.407,
      "mean_token_accuracy": 0.7646078862249851,
      "num_tokens": 162447360.0,
      "step": 620
    },
    {
      "epoch": 1.123049487293803,
      "grad_norm": 0.5824171900749207,
      "learning_rate": 4.594529668226146e-05,
      "loss": 1.4185,
      "mean_token_accuracy": 0.7614389970898628,
      "num_tokens": 165068800.0,
      "step": 630
    },
    {
      "epoch": 1.1408827463218902,
      "grad_norm": 0.5765758156776428,
      "learning_rate": 4.578263842558402e-05,
      "loss": 1.42,
      "mean_token_accuracy": 0.7633033894002438,
      "num_tokens": 167690240.0,
      "step": 640
    },
    {
      "epoch": 1.1587160053499777,
      "grad_norm": 0.5261256694793701,
      "learning_rate": 4.5617081276439355e-05,
      "loss": 1.4384,
      "mean_token_accuracy": 0.7604428693652153,
      "num_tokens": 170311680.0,
      "step": 650
    },
    {
      "epoch": 1.176549264378065,
      "grad_norm": 0.913208544254303,
      "learning_rate": 4.544864832777469e-05,
      "loss": 1.4775,
      "mean_token_accuracy": 0.754877183586359,
      "num_tokens": 172933120.0,
      "step": 660
    },
    {
      "epoch": 1.1943825234061525,
      "grad_norm": 0.5799976587295532,
      "learning_rate": 4.527736307367179e-05,
      "loss": 1.4387,
      "mean_token_accuracy": 0.7589173935353756,
      "num_tokens": 175554560.0,
      "step": 670
    },
    {
      "epoch": 1.2122157824342399,
      "grad_norm": 0.635047435760498,
      "learning_rate": 4.510324940606979e-05,
      "loss": 1.4439,
      "mean_token_accuracy": 0.7597284145653248,
      "num_tokens": 178176000.0,
      "step": 680
    },
    {
      "epoch": 1.2300490414623273,
      "grad_norm": 0.5884873270988464,
      "learning_rate": 4.4926331611432687e-05,
      "loss": 1.3835,
      "mean_token_accuracy": 0.7659787967801094,
      "num_tokens": 180797440.0,
      "step": 690
    },
    {
      "epoch": 1.2478823004904147,
      "grad_norm": 0.7232898473739624,
      "learning_rate": 4.4746634367361634e-05,
      "loss": 1.3629,
      "mean_token_accuracy": 0.7691854700446129,
      "num_tokens": 183418880.0,
      "step": 700
    },
    {
      "epoch": 1.265715559518502,
      "grad_norm": 0.6569238305091858,
      "learning_rate": 4.4564182739152805e-05,
      "loss": 1.397,
      "mean_token_accuracy": 0.7651830442249775,
      "num_tokens": 186040320.0,
      "step": 710
    },
    {
      "epoch": 1.2835488185465893,
      "grad_norm": 0.8630965352058411,
      "learning_rate": 4.437900217630109e-05,
      "loss": 1.4256,
      "mean_token_accuracy": 0.7605882868170738,
      "num_tokens": 188661760.0,
      "step": 720
    },
    {
      "epoch": 1.3013820775746767,
      "grad_norm": 0.7600955367088318,
      "learning_rate": 4.419111850895028e-05,
      "loss": 1.3975,
      "mean_token_accuracy": 0.7634797111153603,
      "num_tokens": 191283200.0,
      "step": 730
    },
    {
      "epoch": 1.3192153366027641,
      "grad_norm": 0.6687177419662476,
      "learning_rate": 4.4000557944290086e-05,
      "loss": 1.3616,
      "mean_token_accuracy": 0.7703544825315476,
      "num_tokens": 193904640.0,
      "step": 740
    },
    {
      "epoch": 1.3370485956308515,
      "grad_norm": 0.719222903251648,
      "learning_rate": 4.3807347062900624e-05,
      "loss": 1.3788,
      "mean_token_accuracy": 0.7648544363677502,
      "num_tokens": 196526080.0,
      "step": 750
    },
    {
      "epoch": 1.354881854658939,
      "grad_norm": 0.5986538529396057,
      "learning_rate": 4.361151281504474e-05,
      "loss": 1.364,
      "mean_token_accuracy": 0.770479665696621,
      "num_tokens": 199147520.0,
      "step": 760
    },
    {
      "epoch": 1.3727151136870264,
      "grad_norm": 0.6472992897033691,
      "learning_rate": 4.3413082516908893e-05,
      "loss": 1.4045,
      "mean_token_accuracy": 0.7644098080694676,
      "num_tokens": 201768960.0,
      "step": 770
    },
    {
      "epoch": 1.3905483727151138,
      "grad_norm": 0.44522207975387573,
      "learning_rate": 4.321208384679285e-05,
      "loss": 1.3314,
      "mean_token_accuracy": 0.773588253557682,
      "num_tokens": 204390400.0,
      "step": 780
    },
    {
      "epoch": 1.4083816317432012,
      "grad_norm": 0.5019171833992004,
      "learning_rate": 4.3008544841248986e-05,
      "loss": 1.3591,
      "mean_token_accuracy": 0.7695087313652038,
      "num_tokens": 207011840.0,
      "step": 790
    },
    {
      "epoch": 1.4262148907712884,
      "grad_norm": 0.7287567257881165,
      "learning_rate": 4.280249389117157e-05,
      "loss": 1.3773,
      "mean_token_accuracy": 0.7675351880490779,
      "num_tokens": 209633280.0,
      "step": 800
    },
    {
      "epoch": 1.4440481497993758,
      "grad_norm": 1.0294618606567383,
      "learning_rate": 4.2593959737836604e-05,
      "loss": 1.3719,
      "mean_token_accuracy": 0.7668005004525185,
      "num_tokens": 212254720.0,
      "step": 810
    },
    {
      "epoch": 1.4618814088274632,
      "grad_norm": 0.6181490421295166,
      "learning_rate": 4.2382971468892806e-05,
      "loss": 1.3168,
      "mean_token_accuracy": 0.7758259050548076,
      "num_tokens": 214876160.0,
      "step": 820
    },
    {
      "epoch": 1.4797146678555506,
      "grad_norm": 0.5945810079574585,
      "learning_rate": 4.216955851430432e-05,
      "loss": 1.3536,
      "mean_token_accuracy": 0.7710937492549419,
      "num_tokens": 217497600.0,
      "step": 830
    },
    {
      "epoch": 1.497547926883638,
      "grad_norm": 0.6803934574127197,
      "learning_rate": 4.195375064224562e-05,
      "loss": 1.324,
      "mean_token_accuracy": 0.7753530308604241,
      "num_tokens": 220119040.0,
      "step": 840
    },
    {
      "epoch": 1.5153811859117252,
      "grad_norm": 0.5748502612113953,
      "learning_rate": 4.1735577954949225e-05,
      "loss": 1.3057,
      "mean_token_accuracy": 0.7755480580031872,
      "num_tokens": 222740480.0,
      "step": 850
    },
    {
      "epoch": 1.5332144449398126,
      "grad_norm": 0.5675086379051208,
      "learning_rate": 4.151507088450692e-05,
      "loss": 1.302,
      "mean_token_accuracy": 0.7765964694321156,
      "num_tokens": 225361920.0,
      "step": 860
    },
    {
      "epoch": 1.5510477039679,
      "grad_norm": 0.6266924142837524,
      "learning_rate": 4.129226018862484e-05,
      "loss": 1.353,
      "mean_token_accuracy": 0.770804837346077,
      "num_tokens": 227983360.0,
      "step": 870
    },
    {
      "epoch": 1.5688809629959874,
      "grad_norm": 0.551527202129364,
      "learning_rate": 4.106717694633321e-05,
      "loss": 1.3477,
      "mean_token_accuracy": 0.77206010222435,
      "num_tokens": 230604800.0,
      "step": 880
    },
    {
      "epoch": 1.5867142220240749,
      "grad_norm": 0.6628434658050537,
      "learning_rate": 4.0839852553651265e-05,
      "loss": 1.3244,
      "mean_token_accuracy": 0.773486353456974,
      "num_tokens": 233226240.0,
      "step": 890
    },
    {
      "epoch": 1.6045474810521623,
      "grad_norm": 0.5956223011016846,
      "learning_rate": 4.0610318719207906e-05,
      "loss": 1.3298,
      "mean_token_accuracy": 0.7733359761536122,
      "num_tokens": 235847680.0,
      "step": 900
    },
    {
      "epoch": 1.6223807400802497,
      "grad_norm": 0.6573621034622192,
      "learning_rate": 4.037860745981881e-05,
      "loss": 1.283,
      "mean_token_accuracy": 0.7787142746150494,
      "num_tokens": 238469120.0,
      "step": 910
    },
    {
      "epoch": 1.640213999108337,
      "grad_norm": 0.607742965221405,
      "learning_rate": 4.01447510960205e-05,
      "loss": 1.2955,
      "mean_token_accuracy": 0.7780937045812607,
      "num_tokens": 241090560.0,
      "step": 920
    },
    {
      "epoch": 1.6580472581364245,
      "grad_norm": 0.6278334856033325,
      "learning_rate": 3.99087822475621e-05,
      "loss": 1.3215,
      "mean_token_accuracy": 0.7741382211446762,
      "num_tokens": 243712000.0,
      "step": 930
    },
    {
      "epoch": 1.675880517164512,
      "grad_norm": 0.5402811169624329,
      "learning_rate": 3.96707338288553e-05,
      "loss": 1.3113,
      "mean_token_accuracy": 0.7751454092562199,
      "num_tokens": 246333440.0,
      "step": 940
    },
    {
      "epoch": 1.6937137761925993,
      "grad_norm": 0.6165571212768555,
      "learning_rate": 3.9430639044383274e-05,
      "loss": 1.3093,
      "mean_token_accuracy": 0.7768628627061844,
      "num_tokens": 248954880.0,
      "step": 950
    },
    {
      "epoch": 1.7115470352206867,
      "grad_norm": 0.6086046695709229,
      "learning_rate": 3.9188531384069096e-05,
      "loss": 1.3125,
      "mean_token_accuracy": 0.7762621335685254,
      "num_tokens": 251576320.0,
      "step": 960
    },
    {
      "epoch": 1.729380294248774,
      "grad_norm": 0.682686984539032,
      "learning_rate": 3.8944444618604356e-05,
      "loss": 1.2821,
      "mean_token_accuracy": 0.7795726217329502,
      "num_tokens": 254197760.0,
      "step": 970
    },
    {
      "epoch": 1.7472135532768613,
      "grad_norm": 0.5615099668502808,
      "learning_rate": 3.869841279473862e-05,
      "loss": 1.3095,
      "mean_token_accuracy": 0.7766918800771236,
      "num_tokens": 256819200.0,
      "step": 980
    },
    {
      "epoch": 1.7650468123049488,
      "grad_norm": 0.5341463088989258,
      "learning_rate": 3.845047023053037e-05,
      "loss": 1.2864,
      "mean_token_accuracy": 0.779373399913311,
      "num_tokens": 259440640.0,
      "step": 990
    },
    {
      "epoch": 1.7828800713330362,
      "grad_norm": 0.5641157627105713,
      "learning_rate": 3.82006515105601e-05,
      "loss": 1.2857,
      "mean_token_accuracy": 0.7798695519566536,
      "num_tokens": 262062080.0,
      "step": 1000
    },
    {
      "epoch": 1.8007133303611234,
      "grad_norm": 0.5196830630302429,
      "learning_rate": 3.794899148110628e-05,
      "loss": 1.2949,
      "mean_token_accuracy": 0.7788909852504731,
      "num_tokens": 264683520.0,
      "step": 1010
    },
    {
      "epoch": 1.8185465893892108,
      "grad_norm": 0.4990008771419525,
      "learning_rate": 3.7695525245284716e-05,
      "loss": 1.254,
      "mean_token_accuracy": 0.7839147254824639,
      "num_tokens": 267304960.0,
      "step": 1020
    },
    {
      "epoch": 1.8363798484172982,
      "grad_norm": 0.5643882751464844,
      "learning_rate": 3.7440288158152187e-05,
      "loss": 1.2446,
      "mean_token_accuracy": 0.785726061463356,
      "num_tokens": 269926400.0,
      "step": 1030
    },
    {
      "epoch": 1.8542131074453856,
      "grad_norm": 0.8107313513755798,
      "learning_rate": 3.7183315821774914e-05,
      "loss": 1.2974,
      "mean_token_accuracy": 0.7775464087724686,
      "num_tokens": 272547840.0,
      "step": 1040
    },
    {
      "epoch": 1.872046366473473,
      "grad_norm": 0.5176706314086914,
      "learning_rate": 3.692464408026253e-05,
      "loss": 1.2916,
      "mean_token_accuracy": 0.7772151328623295,
      "num_tokens": 275169280.0,
      "step": 1050
    },
    {
      "epoch": 1.8898796255015604,
      "grad_norm": 0.5898165106773376,
      "learning_rate": 3.666430901476833e-05,
      "loss": 1.2716,
      "mean_token_accuracy": 0.7806546173989772,
      "num_tokens": 277790720.0,
      "step": 1060
    },
    {
      "epoch": 1.9077128845296478,
      "grad_norm": 0.5368467569351196,
      "learning_rate": 3.640234693845645e-05,
      "loss": 1.2491,
      "mean_token_accuracy": 0.7844746112823486,
      "num_tokens": 280412160.0,
      "step": 1070
    },
    {
      "epoch": 1.9255461435577352,
      "grad_norm": 0.5310569405555725,
      "learning_rate": 3.613879439143667e-05,
      "loss": 1.2825,
      "mean_token_accuracy": 0.7795199528336525,
      "num_tokens": 283033600.0,
      "step": 1080
    },
    {
      "epoch": 1.9433794025858226,
      "grad_norm": 0.6114910840988159,
      "learning_rate": 3.58736881356676e-05,
      "loss": 1.2595,
      "mean_token_accuracy": 0.7831220239400863,
      "num_tokens": 285655040.0,
      "step": 1090
    },
    {
      "epoch": 1.96121266161391,
      "grad_norm": 0.5786811113357544,
      "learning_rate": 3.5607065149828843e-05,
      "loss": 1.2694,
      "mean_token_accuracy": 0.7822808526456356,
      "num_tokens": 288276480.0,
      "step": 1100
    },
    {
      "epoch": 1.9790459206419975,
      "grad_norm": 0.6175279021263123,
      "learning_rate": 3.533896262416302e-05,
      "loss": 1.2712,
      "mean_token_accuracy": 0.7803981445729733,
      "num_tokens": 290897920.0,
      "step": 1110
    },
    {
      "epoch": 1.9968791796700847,
      "grad_norm": 0.568803608417511,
      "learning_rate": 3.506941795528821e-05,
      "loss": 1.2576,
      "mean_token_accuracy": 0.7821117818355561,
      "num_tokens": 293519360.0,
      "step": 1120
    },
    {
      "epoch": 2.0,
      "eval_loss": 1.2395988702774048,
      "eval_mean_token_accuracy": 0.7735649704933166,
      "eval_num_tokens": 293961728.0,
      "eval_runtime": 2.8804,
      "eval_samples_per_second": 24.997,
      "eval_steps_per_second": 1.736,
      "step": 1122
    },
    {
      "epoch": 2.0142666072224698,
      "grad_norm": 0.6652095913887024,
      "learning_rate": 3.479846874098167e-05,
      "loss": 1.2741,
      "mean_token_accuracy": 0.7794669828353784,
      "num_tokens": 296058880.0,
      "step": 1130
    },
    {
      "epoch": 2.032099866250557,
      "grad_norm": 0.7581707835197449,
      "learning_rate": 3.452615277493544e-05,
      "loss": 1.199,
      "mean_token_accuracy": 0.7900658771395683,
      "num_tokens": 298680320.0,
      "step": 1140
    },
    {
      "epoch": 2.0499331252786446,
      "grad_norm": 0.5383137464523315,
      "learning_rate": 3.425250804148464e-05,
      "loss": 1.2304,
      "mean_token_accuracy": 0.7861943542957306,
      "num_tokens": 301301760.0,
      "step": 1150
    },
    {
      "epoch": 2.067766384306732,
      "grad_norm": 0.5952672958374023,
      "learning_rate": 3.3977572710309166e-05,
      "loss": 1.2445,
      "mean_token_accuracy": 0.7836834371089936,
      "num_tokens": 303923200.0,
      "step": 1160
    },
    {
      "epoch": 2.0855996433348194,
      "grad_norm": 0.646511435508728,
      "learning_rate": 3.3701385131109616e-05,
      "loss": 1.2019,
      "mean_token_accuracy": 0.7892128698527813,
      "num_tokens": 306544640.0,
      "step": 1170
    },
    {
      "epoch": 2.103432902362907,
      "grad_norm": 0.514665961265564,
      "learning_rate": 3.342398382825791e-05,
      "loss": 1.2614,
      "mean_token_accuracy": 0.7812080189585686,
      "num_tokens": 309166080.0,
      "step": 1180
    },
    {
      "epoch": 2.121266161390994,
      "grad_norm": 0.5490918159484863,
      "learning_rate": 3.31454074954238e-05,
      "loss": 1.2108,
      "mean_token_accuracy": 0.7876022845506668,
      "num_tokens": 311787520.0,
      "step": 1190
    },
    {
      "epoch": 2.1390994204190816,
      "grad_norm": 0.6549789309501648,
      "learning_rate": 3.286569499017752e-05,
      "loss": 1.2024,
      "mean_token_accuracy": 0.7898712314665317,
      "num_tokens": 314408960.0,
      "step": 1200
    },
    {
      "epoch": 2.156932679447169,
      "grad_norm": 0.5743107199668884,
      "learning_rate": 3.2584885328569785e-05,
      "loss": 1.1835,
      "mean_token_accuracy": 0.7928275369107723,
      "num_tokens": 317030400.0,
      "step": 1210
    },
    {
      "epoch": 2.1747659384752565,
      "grad_norm": 0.6201761960983276,
      "learning_rate": 3.2303017679689525e-05,
      "loss": 1.1939,
      "mean_token_accuracy": 0.7908047571778297,
      "num_tokens": 319651840.0,
      "step": 1220
    },
    {
      "epoch": 2.192599197503344,
      "grad_norm": 0.550830602645874,
      "learning_rate": 3.2020131360200354e-05,
      "loss": 1.1864,
      "mean_token_accuracy": 0.7930233284831048,
      "num_tokens": 322273280.0,
      "step": 1230
    },
    {
      "epoch": 2.2104324565314313,
      "grad_norm": 0.5972590446472168,
      "learning_rate": 3.173626582885645e-05,
      "loss": 1.2177,
      "mean_token_accuracy": 0.7885610058903694,
      "num_tokens": 324894720.0,
      "step": 1240
    },
    {
      "epoch": 2.2282657155595187,
      "grad_norm": 0.5069220066070557,
      "learning_rate": 3.145146068099858e-05,
      "loss": 1.2279,
      "mean_token_accuracy": 0.7862870991230011,
      "num_tokens": 327516160.0,
      "step": 1250
    },
    {
      "epoch": 2.246098974587606,
      "grad_norm": 0.5100940465927124,
      "learning_rate": 3.116575564303109e-05,
      "loss": 1.1777,
      "mean_token_accuracy": 0.7929233327507973,
      "num_tokens": 330137600.0,
      "step": 1260
    },
    {
      "epoch": 2.2639322336156935,
      "grad_norm": 0.6175646781921387,
      "learning_rate": 3.087919056688067e-05,
      "loss": 1.173,
      "mean_token_accuracy": 0.7950430497527122,
      "num_tokens": 332759040.0,
      "step": 1270
    },
    {
      "epoch": 2.2817654926437805,
      "grad_norm": 0.5933928489685059,
      "learning_rate": 3.059180542443746e-05,
      "loss": 1.2057,
      "mean_token_accuracy": 0.7882316343486309,
      "num_tokens": 335380480.0,
      "step": 1280
    },
    {
      "epoch": 2.299598751671868,
      "grad_norm": 0.4882568120956421,
      "learning_rate": 3.0303640301979635e-05,
      "loss": 1.1982,
      "mean_token_accuracy": 0.7903318881988526,
      "num_tokens": 338001920.0,
      "step": 1290
    },
    {
      "epoch": 2.3174320106999553,
      "grad_norm": 0.5135630369186401,
      "learning_rate": 3.0014735394581823e-05,
      "loss": 1.2045,
      "mean_token_accuracy": 0.7893250770866871,
      "num_tokens": 340623360.0,
      "step": 1300
    },
    {
      "epoch": 2.3352652697280427,
      "grad_norm": 0.5496548414230347,
      "learning_rate": 2.972513100050851e-05,
      "loss": 1.1883,
      "mean_token_accuracy": 0.791755847632885,
      "num_tokens": 343244800.0,
      "step": 1310
    },
    {
      "epoch": 2.35309852875613,
      "grad_norm": 0.5601444840431213,
      "learning_rate": 2.943486751559295e-05,
      "loss": 1.1263,
      "mean_token_accuracy": 0.7998171858489513,
      "num_tokens": 345866240.0,
      "step": 1320
    },
    {
      "epoch": 2.3709317877842175,
      "grad_norm": 0.5243608355522156,
      "learning_rate": 2.9143985427602516e-05,
      "loss": 1.2274,
      "mean_token_accuracy": 0.7866500526666641,
      "num_tokens": 348487680.0,
      "step": 1330
    },
    {
      "epoch": 2.388765046812305,
      "grad_norm": 0.573567807674408,
      "learning_rate": 2.88525253105912e-05,
      "loss": 1.1977,
      "mean_token_accuracy": 0.7901960171759128,
      "num_tokens": 351109120.0,
      "step": 1340
    },
    {
      "epoch": 2.4065983058403924,
      "grad_norm": 0.4823901951313019,
      "learning_rate": 2.856052781924012e-05,
      "loss": 1.1358,
      "mean_token_accuracy": 0.7998057343065739,
      "num_tokens": 353730560.0,
      "step": 1350
    },
    {
      "epoch": 2.4244315648684798,
      "grad_norm": 0.515941858291626,
      "learning_rate": 2.8268033683186697e-05,
      "loss": 1.1881,
      "mean_token_accuracy": 0.7909879550337792,
      "num_tokens": 356352000.0,
      "step": 1360
    },
    {
      "epoch": 2.442264823896567,
      "grad_norm": 0.5508431196212769,
      "learning_rate": 2.7975083701343497e-05,
      "loss": 1.171,
      "mean_token_accuracy": 0.7945770494639873,
      "num_tokens": 358973440.0,
      "step": 1370
    },
    {
      "epoch": 2.4600980829246546,
      "grad_norm": 0.5473787188529968,
      "learning_rate": 2.7681718736207298e-05,
      "loss": 1.1309,
      "mean_token_accuracy": 0.7995454497635365,
      "num_tokens": 361594880.0,
      "step": 1380
    },
    {
      "epoch": 2.477931341952742,
      "grad_norm": 0.6680080890655518,
      "learning_rate": 2.738797970815933e-05,
      "loss": 1.1848,
      "mean_token_accuracy": 0.7917638584971428,
      "num_tokens": 364216320.0,
      "step": 1390
    },
    {
      "epoch": 2.4957646009808294,
      "grad_norm": 0.49958300590515137,
      "learning_rate": 2.7093907589757468e-05,
      "loss": 1.1922,
      "mean_token_accuracy": 0.7899307683110237,
      "num_tokens": 366837760.0,
      "step": 1400
    },
    {
      "epoch": 2.5135978600089164,
      "grad_norm": 0.4601416289806366,
      "learning_rate": 2.6799543400021132e-05,
      "loss": 1.1448,
      "mean_token_accuracy": 0.7981104202568531,
      "num_tokens": 369459200.0,
      "step": 1410
    },
    {
      "epoch": 2.531431119037004,
      "grad_norm": 0.5619862079620361,
      "learning_rate": 2.650492819870965e-05,
      "loss": 1.1371,
      "mean_token_accuracy": 0.7994878143072128,
      "num_tokens": 372080640.0,
      "step": 1420
    },
    {
      "epoch": 2.549264378065091,
      "grad_norm": 0.6030861139297485,
      "learning_rate": 2.621010308059504e-05,
      "loss": 1.1695,
      "mean_token_accuracy": 0.7944411806762218,
      "num_tokens": 374702080.0,
      "step": 1430
    },
    {
      "epoch": 2.5670976370931786,
      "grad_norm": 0.5237954258918762,
      "learning_rate": 2.5915109169729814e-05,
      "loss": 1.1372,
      "mean_token_accuracy": 0.7982218623161316,
      "num_tokens": 377323520.0,
      "step": 1440
    },
    {
      "epoch": 2.584930896121266,
      "grad_norm": 0.5144650936126709,
      "learning_rate": 2.5619987613710756e-05,
      "loss": 1.1896,
      "mean_token_accuracy": 0.7915455557405948,
      "num_tokens": 379944960.0,
      "step": 1450
    },
    {
      "epoch": 2.6027641551493534,
      "grad_norm": 0.5404797196388245,
      "learning_rate": 2.5324779577939394e-05,
      "loss": 1.1364,
      "mean_token_accuracy": 0.7987703062593937,
      "num_tokens": 382566400.0,
      "step": 1460
    },
    {
      "epoch": 2.620597414177441,
      "grad_norm": 0.4770708382129669,
      "learning_rate": 2.5029526239880015e-05,
      "loss": 1.1541,
      "mean_token_accuracy": 0.7973127588629723,
      "num_tokens": 385187840.0,
      "step": 1470
    },
    {
      "epoch": 2.6384306732055283,
      "grad_norm": 0.5579174757003784,
      "learning_rate": 2.473426878331591e-05,
      "loss": 1.1695,
      "mean_token_accuracy": 0.7945644512772561,
      "num_tokens": 387809280.0,
      "step": 1480
    },
    {
      "epoch": 2.6562639322336157,
      "grad_norm": 0.5287127494812012,
      "learning_rate": 2.443904839260488e-05,
      "loss": 1.1615,
      "mean_token_accuracy": 0.7949182473123073,
      "num_tokens": 390430720.0,
      "step": 1490
    },
    {
      "epoch": 2.674097191261703,
      "grad_norm": 0.5106226801872253,
      "learning_rate": 2.4143906246934537e-05,
      "loss": 1.1201,
      "mean_token_accuracy": 0.8007564425468445,
      "num_tokens": 393052160.0,
      "step": 1500
    },
    {
      "epoch": 2.6919304502897905,
      "grad_norm": 0.48907479643821716,
      "learning_rate": 2.384888351457838e-05,
      "loss": 1.1908,
      "mean_token_accuracy": 0.7911803111433983,
      "num_tokens": 395673600.0,
      "step": 1510
    },
    {
      "epoch": 2.709763709317878,
      "grad_norm": 0.5045855045318604,
      "learning_rate": 2.3554021347153403e-05,
      "loss": 1.1483,
      "mean_token_accuracy": 0.7968017213046551,
      "num_tokens": 398295040.0,
      "step": 1520
    },
    {
      "epoch": 2.7275969683459653,
      "grad_norm": 0.505662739276886,
      "learning_rate": 2.3259360873879992e-05,
      "loss": 1.1749,
      "mean_token_accuracy": 0.7930756144225597,
      "num_tokens": 400916480.0,
      "step": 1530
    },
    {
      "epoch": 2.7454302273740527,
      "grad_norm": 0.5283376574516296,
      "learning_rate": 2.2964943195844966e-05,
      "loss": 1.173,
      "mean_token_accuracy": 0.7921638391911984,
      "num_tokens": 403537920.0,
      "step": 1540
    },
    {
      "epoch": 2.76326348640214,
      "grad_norm": 0.5634790062904358,
      "learning_rate": 2.2670809380268548e-05,
      "loss": 1.1779,
      "mean_token_accuracy": 0.7936561174690724,
      "num_tokens": 406159360.0,
      "step": 1550
    },
    {
      "epoch": 2.7810967454302276,
      "grad_norm": 0.48228803277015686,
      "learning_rate": 2.237700045477607e-05,
      "loss": 1.14,
      "mean_token_accuracy": 0.796902097016573,
      "num_tokens": 408780800.0,
      "step": 1560
    },
    {
      "epoch": 2.798930004458315,
      "grad_norm": 0.4753395617008209,
      "learning_rate": 2.2083557401675115e-05,
      "loss": 1.1294,
      "mean_token_accuracy": 0.7987142026424408,
      "num_tokens": 411402240.0,
      "step": 1570
    },
    {
      "epoch": 2.8167632634864024,
      "grad_norm": 0.5348941683769226,
      "learning_rate": 2.1790521152239137e-05,
      "loss": 1.1581,
      "mean_token_accuracy": 0.7944865994155407,
      "num_tokens": 414023680.0,
      "step": 1580
    },
    {
      "epoch": 2.83459652251449,
      "grad_norm": 0.5610408782958984,
      "learning_rate": 2.1497932580998053e-05,
      "loss": 1.1609,
      "mean_token_accuracy": 0.795001070201397,
      "num_tokens": 416645120.0,
      "step": 1590
    },
    {
      "epoch": 2.8524297815425768,
      "grad_norm": 0.5431333184242249,
      "learning_rate": 2.1205832500036838e-05,
      "loss": 1.1405,
      "mean_token_accuracy": 0.7983539171516896,
      "num_tokens": 419266560.0,
      "step": 1600
    },
    {
      "epoch": 2.870263040570664,
      "grad_norm": 0.5650882720947266,
      "learning_rate": 2.0914261653302747e-05,
      "loss": 1.17,
      "mean_token_accuracy": 0.7925355695188046,
      "num_tokens": 421888000.0,
      "step": 1610
    },
    {
      "epoch": 2.8880962995987516,
      "grad_norm": 0.5435991287231445,
      "learning_rate": 2.062326071092218e-05,
      "loss": 1.1562,
      "mean_token_accuracy": 0.7965150982141495,
      "num_tokens": 424509440.0,
      "step": 1620
    },
    {
      "epoch": 2.905929558626839,
      "grad_norm": 0.4466145932674408,
      "learning_rate": 2.0332870263527675e-05,
      "loss": 1.1199,
      "mean_token_accuracy": 0.8016460791230202,
      "num_tokens": 427130880.0,
      "step": 1630
    },
    {
      "epoch": 2.9237628176549264,
      "grad_norm": 0.5083450675010681,
      "learning_rate": 2.0043130816596136e-05,
      "loss": 1.1467,
      "mean_token_accuracy": 0.7959990955889225,
      "num_tokens": 429752320.0,
      "step": 1640
    },
    {
      "epoch": 2.941596076683014,
      "grad_norm": 0.49085912108421326,
      "learning_rate": 1.9754082784798843e-05,
      "loss": 1.1565,
      "mean_token_accuracy": 0.7959273479878902,
      "num_tokens": 432373760.0,
      "step": 1650
    },
    {
      "epoch": 2.9594293357111012,
      "grad_norm": 0.48444733023643494,
      "learning_rate": 1.9465766486364143e-05,
      "loss": 1.1342,
      "mean_token_accuracy": 0.797958905249834,
      "num_tokens": 434995200.0,
      "step": 1660
    },
    {
      "epoch": 2.9772625947391886,
      "grad_norm": 0.5343450903892517,
      "learning_rate": 1.9178222137453656e-05,
      "loss": 1.141,
      "mean_token_accuracy": 0.7985142156481743,
      "num_tokens": 437616640.0,
      "step": 1670
    },
    {
      "epoch": 2.995095853767276,
      "grad_norm": 0.5583868026733398,
      "learning_rate": 1.8891489846552646e-05,
      "loss": 1.1432,
      "mean_token_accuracy": 0.7971971161663532,
      "num_tokens": 440238080.0,
      "step": 1680
    },
    {
      "epoch": 3.0,
      "eval_loss": 1.1234296560287476,
      "eval_mean_token_accuracy": 0.7894601821899414,
      "eval_num_tokens": 440942592.0,
      "eval_runtime": 2.9179,
      "eval_samples_per_second": 24.675,
      "eval_steps_per_second": 1.714,
      "step": 1683
    },
    {
      "epoch": 3.0124832813196614,
      "grad_norm": 0.5347409844398499,
      "learning_rate": 1.860560960887546e-05,
      "loss": 1.1253,
      "mean_token_accuracy": 0.8009015726737487,
      "num_tokens": 442777600.0,
      "step": 1690
    },
    {
      "epoch": 3.0303165403477483,
      "grad_norm": 0.5598998069763184,
      "learning_rate": 1.8320621300786688e-05,
      "loss": 1.106,
      "mean_token_accuracy": 0.8023651242256165,
      "num_tokens": 445399040.0,
      "step": 1700
    },
    {
      "epoch": 3.0481497993758357,
      "grad_norm": 0.5113139152526855,
      "learning_rate": 1.8036564674239043e-05,
      "loss": 1.0887,
      "mean_token_accuracy": 0.8040673099458218,
      "num_tokens": 448020480.0,
      "step": 1710
    },
    {
      "epoch": 3.065983058403923,
      "grad_norm": 0.44441869854927063,
      "learning_rate": 1.7753479351228444e-05,
      "loss": 1.1185,
      "mean_token_accuracy": 0.7999229028820991,
      "num_tokens": 450641920.0,
      "step": 1720
    },
    {
      "epoch": 3.0838163174320106,
      "grad_norm": 0.47339409589767456,
      "learning_rate": 1.7471404818267316e-05,
      "loss": 1.0925,
      "mean_token_accuracy": 0.8034070417284965,
      "num_tokens": 453263360.0,
      "step": 1730
    },
    {
      "epoch": 3.101649576460098,
      "grad_norm": 0.47856712341308594,
      "learning_rate": 1.7190380420876766e-05,
      "loss": 1.0911,
      "mean_token_accuracy": 0.8045188069343567,
      "num_tokens": 455884800.0,
      "step": 1740
    },
    {
      "epoch": 3.1194828354881854,
      "grad_norm": 0.4918275773525238,
      "learning_rate": 1.6910445358098415e-05,
      "loss": 1.1109,
      "mean_token_accuracy": 0.8015380732715129,
      "num_tokens": 458506240.0,
      "step": 1750
    },
    {
      "epoch": 3.137316094516273,
      "grad_norm": 0.5043804049491882,
      "learning_rate": 1.6631638677026674e-05,
      "loss": 1.1364,
      "mean_token_accuracy": 0.7974261090159416,
      "num_tokens": 461127680.0,
      "step": 1760
    },
    {
      "epoch": 3.15514935354436,
      "grad_norm": 0.4904772937297821,
      "learning_rate": 1.6353999267362237e-05,
      "loss": 1.0838,
      "mean_token_accuracy": 0.806386636942625,
      "num_tokens": 463749120.0,
      "step": 1770
    },
    {
      "epoch": 3.1729826125724476,
      "grad_norm": 0.48294153809547424,
      "learning_rate": 1.607756585598743e-05,
      "loss": 1.1389,
      "mean_token_accuracy": 0.7980314172804356,
      "num_tokens": 466370560.0,
      "step": 1780
    },
    {
      "epoch": 3.190815871600535,
      "grad_norm": 0.5139598250389099,
      "learning_rate": 1.580237700156443e-05,
      "loss": 1.1095,
      "mean_token_accuracy": 0.8025780878961086,
      "num_tokens": 468992000.0,
      "step": 1790
    },
    {
      "epoch": 3.2086491306286224,
      "grad_norm": 0.47298020124435425,
      "learning_rate": 1.5528471089156804e-05,
      "loss": 1.0931,
      "mean_token_accuracy": 0.8041585251688957,
      "num_tokens": 471613440.0,
      "step": 1800
    },
    {
      "epoch": 3.22648238965671,
      "grad_norm": 0.41903576254844666,
      "learning_rate": 1.5255886324875363e-05,
      "loss": 1.0954,
      "mean_token_accuracy": 0.8047371134161949,
      "num_tokens": 474234880.0,
      "step": 1810
    },
    {
      "epoch": 3.2443156486847973,
      "grad_norm": 0.4610045552253723,
      "learning_rate": 1.4984660730548856e-05,
      "loss": 1.1095,
      "mean_token_accuracy": 0.8022483348846435,
      "num_tokens": 476856320.0,
      "step": 1820
    },
    {
      "epoch": 3.2621489077128847,
      "grad_norm": 0.5105413198471069,
      "learning_rate": 1.4714832138420553e-05,
      "loss": 1.1258,
      "mean_token_accuracy": 0.7983997173607349,
      "num_tokens": 479477760.0,
      "step": 1830
    },
    {
      "epoch": 3.279982166740972,
      "grad_norm": 0.4393883943557739,
      "learning_rate": 1.4446438185871092e-05,
      "loss": 1.0566,
      "mean_token_accuracy": 0.8093036323785782,
      "num_tokens": 482099200.0,
      "step": 1840
    },
    {
      "epoch": 3.297815425769059,
      "grad_norm": 0.4618268311023712,
      "learning_rate": 1.4179516310168623e-05,
      "loss": 1.0697,
      "mean_token_accuracy": 0.8082189694046974,
      "num_tokens": 484720640.0,
      "step": 1850
    },
    {
      "epoch": 3.3156486847971465,
      "grad_norm": 0.47192856669425964,
      "learning_rate": 1.3914103743246757e-05,
      "loss": 1.1203,
      "mean_token_accuracy": 0.8002961635589599,
      "num_tokens": 487342080.0,
      "step": 1860
    },
    {
      "epoch": 3.333481943825234,
      "grad_norm": 0.4617849290370941,
      "learning_rate": 1.3650237506511331e-05,
      "loss": 1.0664,
      "mean_token_accuracy": 0.8089914366602897,
      "num_tokens": 489963520.0,
      "step": 1870
    },
    {
      "epoch": 3.3513152028533213,
      "grad_norm": 0.46263769268989563,
      "learning_rate": 1.3387954405676329e-05,
      "loss": 1.0855,
      "mean_token_accuracy": 0.8045512497425079,
      "num_tokens": 492584960.0,
      "step": 1880
    },
    {
      "epoch": 3.3691484618814087,
      "grad_norm": 0.47842440009117126,
      "learning_rate": 1.3127291025630073e-05,
      "loss": 1.0705,
      "mean_token_accuracy": 0.8085929870605468,
      "num_tokens": 495206400.0,
      "step": 1890
    },
    {
      "epoch": 3.386981720909496,
      "grad_norm": 0.43500861525535583,
      "learning_rate": 1.2868283725332028e-05,
      "loss": 1.0635,
      "mean_token_accuracy": 0.8091395162045956,
      "num_tokens": 497827840.0,
      "step": 1900
    },
    {
      "epoch": 3.4048149799375835,
      "grad_norm": 0.508797287940979,
      "learning_rate": 1.261096863274136e-05,
      "loss": 1.1052,
      "mean_token_accuracy": 0.8021300226449967,
      "num_tokens": 500449280.0,
      "step": 1910
    },
    {
      "epoch": 3.422648238965671,
      "grad_norm": 0.42378219962120056,
      "learning_rate": 1.2355381639777472e-05,
      "loss": 1.0936,
      "mean_token_accuracy": 0.8043497301638126,
      "num_tokens": 503070720.0,
      "step": 1920
    },
    {
      "epoch": 3.4404814979937584,
      "grad_norm": 0.4547331631183624,
      "learning_rate": 1.2101558397313634e-05,
      "loss": 1.1142,
      "mean_token_accuracy": 0.8015869230031967,
      "num_tokens": 505692160.0,
      "step": 1930
    },
    {
      "epoch": 3.4583147570218458,
      "grad_norm": 0.4319451153278351,
      "learning_rate": 1.1849534310204152e-05,
      "loss": 1.0835,
      "mean_token_accuracy": 0.8058400988578797,
      "num_tokens": 508313600.0,
      "step": 1940
    },
    {
      "epoch": 3.476148016049933,
      "grad_norm": 0.4666835069656372,
      "learning_rate": 1.1599344532345877e-05,
      "loss": 1.1072,
      "mean_token_accuracy": 0.801591893285513,
      "num_tokens": 510935040.0,
      "step": 1950
    },
    {
      "epoch": 3.4939812750780206,
      "grad_norm": 0.47252121567726135,
      "learning_rate": 1.1351023961774749e-05,
      "loss": 1.1245,
      "mean_token_accuracy": 0.7985180281102657,
      "num_tokens": 513556480.0,
      "step": 1960
    },
    {
      "epoch": 3.511814534106108,
      "grad_norm": 0.5152117013931274,
      "learning_rate": 1.110460723579796e-05,
      "loss": 1.0982,
      "mean_token_accuracy": 0.8033757455646992,
      "num_tokens": 516177920.0,
      "step": 1970
    },
    {
      "epoch": 3.5296477931341954,
      "grad_norm": 0.43357202410697937,
      "learning_rate": 1.0860128726162545e-05,
      "loss": 1.0632,
      "mean_token_accuracy": 0.8095081992447376,
      "num_tokens": 518799360.0,
      "step": 1980
    },
    {
      "epoch": 3.547481052162283,
      "grad_norm": 0.45368409156799316,
      "learning_rate": 1.0617622534260985e-05,
      "loss": 1.114,
      "mean_token_accuracy": 0.8014212884008884,
      "num_tokens": 521420800.0,
      "step": 1990
    },
    {
      "epoch": 3.5653143111903702,
      "grad_norm": 0.38939619064331055,
      "learning_rate": 1.0377122486374576e-05,
      "loss": 1.0449,
      "mean_token_accuracy": 0.8123023062944412,
      "num_tokens": 524042240.0,
      "step": 2000
    },
    {
      "epoch": 3.5831475702184576,
      "grad_norm": 0.42768579721450806,
      "learning_rate": 1.0138662128955053e-05,
      "loss": 1.104,
      "mean_token_accuracy": 0.8025334358215332,
      "num_tokens": 526663680.0,
      "step": 2010
    },
    {
      "epoch": 3.600980829246545,
      "grad_norm": 0.41337716579437256,
      "learning_rate": 9.902274723945381e-06,
      "loss": 1.0788,
      "mean_token_accuracy": 0.8076182380318642,
      "num_tokens": 529285120.0,
      "step": 2020
    },
    {
      "epoch": 3.618814088274632,
      "grad_norm": 0.4830559194087982,
      "learning_rate": 9.667993244140134e-06,
      "loss": 1.0967,
      "mean_token_accuracy": 0.8028597474098206,
      "num_tokens": 531906560.0,
      "step": 2030
    },
    {
      "epoch": 3.6366473473027194,
      "grad_norm": 0.4195765554904938,
      "learning_rate": 9.435850368586286e-06,
      "loss": 1.0773,
      "mean_token_accuracy": 0.8066488340497017,
      "num_tokens": 534528000.0,
      "step": 2040
    },
    {
      "epoch": 3.654480606330807,
      "grad_norm": 0.4251314103603363,
      "learning_rate": 9.20587847802488e-06,
      "loss": 1.0819,
      "mean_token_accuracy": 0.8052599862217903,
      "num_tokens": 537149440.0,
      "step": 2050
    },
    {
      "epoch": 3.6723138653588943,
      "grad_norm": 0.4114632308483124,
      "learning_rate": 8.978109650374397e-06,
      "loss": 1.0697,
      "mean_token_accuracy": 0.8077415138483047,
      "num_tokens": 539770880.0,
      "step": 2060
    },
    {
      "epoch": 3.6901471243869817,
      "grad_norm": 0.4304318428039551,
      "learning_rate": 8.752575656256321e-06,
      "loss": 1.1046,
      "mean_token_accuracy": 0.8016697481274605,
      "num_tokens": 542392320.0,
      "step": 2070
    },
    {
      "epoch": 3.707980383415069,
      "grad_norm": 0.47687673568725586,
      "learning_rate": 8.52930795456355e-06,
      "loss": 1.0558,
      "mean_token_accuracy": 0.8107806429266929,
      "num_tokens": 545013760.0,
      "step": 2080
    },
    {
      "epoch": 3.7258136424431565,
      "grad_norm": 0.45524051785469055,
      "learning_rate": 8.30833768807237e-06,
      "loss": 1.0816,
      "mean_token_accuracy": 0.8059305518865585,
      "num_tokens": 547635200.0,
      "step": 2090
    },
    {
      "epoch": 3.743646901471244,
      "grad_norm": 0.45478421449661255,
      "learning_rate": 8.089695679098375e-06,
      "loss": 1.0961,
      "mean_token_accuracy": 0.8029544025659561,
      "num_tokens": 550256640.0,
      "step": 2100
    },
    {
      "epoch": 3.7614801604993313,
      "grad_norm": 0.45910394191741943,
      "learning_rate": 7.87341242519728e-06,
      "loss": 1.0608,
      "mean_token_accuracy": 0.8082987323403359,
      "num_tokens": 552878080.0,
      "step": 2110
    },
    {
      "epoch": 3.7793134195274187,
      "grad_norm": 0.4360869824886322,
      "learning_rate": 7.659518094910839e-06,
      "loss": 1.0806,
      "mean_token_accuracy": 0.8052332706749439,
      "num_tokens": 555499520.0,
      "step": 2120
    },
    {
      "epoch": 3.797146678555506,
      "grad_norm": 0.46505963802337646,
      "learning_rate": 7.44804252355884e-06,
      "loss": 1.1143,
      "mean_token_accuracy": 0.8007854476571084,
      "num_tokens": 558120960.0,
      "step": 2130
    },
    {
      "epoch": 3.8149799375835935,
      "grad_norm": 0.40090060234069824,
      "learning_rate": 7.2390152090773765e-06,
      "loss": 1.079,
      "mean_token_accuracy": 0.8058068990707398,
      "num_tokens": 560742400.0,
      "step": 2140
    },
    {
      "epoch": 3.8328131966116805,
      "grad_norm": 0.41377872228622437,
      "learning_rate": 7.032465307904404e-06,
      "loss": 1.0496,
      "mean_token_accuracy": 0.8097444467246533,
      "num_tokens": 563363840.0,
      "step": 2150
    },
    {
      "epoch": 3.850646455639768,
      "grad_norm": 0.45236936211586,
      "learning_rate": 6.82842163091274e-06,
      "loss": 1.0819,
      "mean_token_accuracy": 0.8053412742912769,
      "num_tokens": 565985280.0,
      "step": 2160
    },
    {
      "epoch": 3.8684797146678553,
      "grad_norm": 0.42451992630958557,
      "learning_rate": 6.626912639391377e-06,
      "loss": 1.0649,
      "mean_token_accuracy": 0.8076308310031891,
      "num_tokens": 568606720.0,
      "step": 2170
    },
    {
      "epoch": 3.8863129736959428,
      "grad_norm": 0.43241724371910095,
      "learning_rate": 6.427966441075511e-06,
      "loss": 1.0821,
      "mean_token_accuracy": 0.805367611348629,
      "num_tokens": 571228160.0,
      "step": 2180
    },
    {
      "epoch": 3.90414623272403,
      "grad_norm": 0.41288912296295166,
      "learning_rate": 6.2316107862259026e-06,
      "loss": 1.1017,
      "mean_token_accuracy": 0.8022395603358745,
      "num_tokens": 573849600.0,
      "step": 2190
    },
    {
      "epoch": 3.9219794917521176,
      "grad_norm": 0.4336395263671875,
      "learning_rate": 6.037873063758112e-06,
      "loss": 1.0546,
      "mean_token_accuracy": 0.8097028434276581,
      "num_tokens": 576471040.0,
      "step": 2200
    },
    {
      "epoch": 3.939812750780205,
      "grad_norm": 0.4041893482208252,
      "learning_rate": 5.846780297422091e-06,
      "loss": 1.0603,
      "mean_token_accuracy": 0.8089009828865528,
      "num_tokens": 579092480.0,
      "step": 2210
    },
    {
      "epoch": 3.9576460098082924,
      "grad_norm": 0.39978188276290894,
      "learning_rate": 5.6583591420327684e-06,
      "loss": 1.0752,
      "mean_token_accuracy": 0.8069194242358207,
      "num_tokens": 581713920.0,
      "step": 2220
    },
    {
      "epoch": 3.97547926883638,
      "grad_norm": 0.41306570172309875,
      "learning_rate": 5.472635879752042e-06,
      "loss": 1.0757,
      "mean_token_accuracy": 0.8059477277100087,
      "num_tokens": 584335360.0,
      "step": 2230
    },
    {
      "epoch": 3.993312527864467,
      "grad_norm": 0.4413711428642273,
      "learning_rate": 5.289636416422822e-06,
      "loss": 1.1117,
      "mean_token_accuracy": 0.8009537614881992,
      "num_tokens": 586956800.0,
      "step": 2240
    },
    {
      "epoch": 4.0,
      "eval_loss": 1.0707859992980957,
      "eval_mean_token_accuracy": 0.7980337142944336,
      "eval_num_tokens": 587923456.0,
      "eval_runtime": 3.1887,
      "eval_samples_per_second": 22.579,
      "eval_steps_per_second": 1.568,
      "step": 2244
    },
    {
      "epoch": 4.010699955416852,
      "grad_norm": 0.3837050199508667,
      "learning_rate": 5.1093862779554776e-06,
      "loss": 1.0237,
      "mean_token_accuracy": 0.814745701276339,
      "num_tokens": 589496320.0,
      "step": 2250
    },
    {
      "epoch": 4.0285332144449395,
      "grad_norm": 0.39410483837127686,
      "learning_rate": 4.93191060676734e-06,
      "loss": 1.059,
      "mean_token_accuracy": 0.8092822574079037,
      "num_tokens": 592117760.0,
      "step": 2260
    },
    {
      "epoch": 4.046366473473027,
      "grad_norm": 0.4275108873844147,
      "learning_rate": 4.757234158275681e-06,
      "loss": 1.0876,
      "mean_token_accuracy": 0.8050832770764828,
      "num_tokens": 594739200.0,
      "step": 2270
    },
    {
      "epoch": 4.064199732501114,
      "grad_norm": 0.39579182863235474,
      "learning_rate": 4.585381297444688e-06,
      "loss": 1.0313,
      "mean_token_accuracy": 0.8126221284270286,
      "num_tokens": 597360640.0,
      "step": 2280
    },
    {
      "epoch": 4.082032991529202,
      "grad_norm": 0.4174662232398987,
      "learning_rate": 4.416375995386857e-06,
      "loss": 1.0677,
      "mean_token_accuracy": 0.808067063242197,
      "num_tokens": 599982080.0,
      "step": 2290
    },
    {
      "epoch": 4.099866250557289,
      "grad_norm": 0.426666796207428,
      "learning_rate": 4.2502418260193686e-06,
      "loss": 1.0551,
      "mean_token_accuracy": 0.8088761754333973,
      "num_tokens": 602603520.0,
      "step": 2300
    },
    {
      "epoch": 4.117699509585377,
      "grad_norm": 0.3763045370578766,
      "learning_rate": 4.087001962775849e-06,
      "loss": 1.0624,
      "mean_token_accuracy": 0.8099226720631123,
      "num_tokens": 605224960.0,
      "step": 2310
    },
    {
      "epoch": 4.135532768613464,
      "grad_norm": 0.40583643317222595,
      "learning_rate": 3.926679175373985e-06,
      "loss": 1.0593,
      "mean_token_accuracy": 0.8088261790573597,
      "num_tokens": 607846400.0,
      "step": 2320
    },
    {
      "epoch": 4.153366027641551,
      "grad_norm": 0.3805159330368042,
      "learning_rate": 3.7692958266394963e-06,
      "loss": 1.0686,
      "mean_token_accuracy": 0.808470857143402,
      "num_tokens": 610467840.0,
      "step": 2330
    },
    {
      "epoch": 4.171199286669639,
      "grad_norm": 0.36603036522865295,
      "learning_rate": 3.614873869386762e-06,
      "loss": 1.0475,
      "mean_token_accuracy": 0.8111096285283566,
      "num_tokens": 613089280.0,
      "step": 2340
    },
    {
      "epoch": 4.189032545697726,
      "grad_norm": 0.4105148911476135,
      "learning_rate": 3.463434843356783e-06,
      "loss": 1.0474,
      "mean_token_accuracy": 0.8102818146348,
      "num_tokens": 615710720.0,
      "step": 2350
    },
    {
      "epoch": 4.206865804725814,
      "grad_norm": 0.41211214661598206,
      "learning_rate": 3.314999872212618e-06,
      "loss": 1.1055,
      "mean_token_accuracy": 0.8027391448616982,
      "num_tokens": 618332160.0,
      "step": 2360
    },
    {
      "epoch": 4.224699063753901,
      "grad_norm": 0.409272164106369,
      "learning_rate": 3.169589660592978e-06,
      "loss": 1.0479,
      "mean_token_accuracy": 0.8102260909974575,
      "num_tokens": 620953600.0,
      "step": 2370
    },
    {
      "epoch": 4.242532322781988,
      "grad_norm": 0.4126708209514618,
      "learning_rate": 3.0272244912241453e-06,
      "loss": 1.0142,
      "mean_token_accuracy": 0.8164093054831028,
      "num_tokens": 623575040.0,
      "step": 2380
    },
    {
      "epoch": 4.260365581810076,
      "grad_norm": 0.3964686393737793,
      "learning_rate": 2.887924222090899e-06,
      "loss": 1.0644,
      "mean_token_accuracy": 0.8078506626188755,
      "num_tokens": 626196480.0,
      "step": 2390
    },
    {
      "epoch": 4.278198840838163,
      "grad_norm": 0.40106719732284546,
      "learning_rate": 2.751708283666521e-06,
      "loss": 1.0834,
      "mean_token_accuracy": 0.8057584255933762,
      "num_tokens": 628817920.0,
      "step": 2400
    },
    {
      "epoch": 4.296032099866251,
      "grad_norm": 0.3820619285106659,
      "learning_rate": 2.6185956762025592e-06,
      "loss": 1.0744,
      "mean_token_accuracy": 0.8067240163683891,
      "num_tokens": 631439360.0,
      "step": 2410
    },
    {
      "epoch": 4.313865358894338,
      "grad_norm": 0.41071221232414246,
      "learning_rate": 2.488604967078528e-06,
      "loss": 1.0805,
      "mean_token_accuracy": 0.8062465645372867,
      "num_tokens": 634060800.0,
      "step": 2420
    },
    {
      "epoch": 4.3316986179224255,
      "grad_norm": 0.3963005542755127,
      "learning_rate": 2.361754288212031e-06,
      "loss": 1.0898,
      "mean_token_accuracy": 0.8040848657488823,
      "num_tokens": 636682240.0,
      "step": 2430
    },
    {
      "epoch": 4.349531876950513,
      "grad_norm": 0.3937000036239624,
      "learning_rate": 2.2380613335296036e-06,
      "loss": 1.0211,
      "mean_token_accuracy": 0.8141208916902543,
      "num_tokens": 639303680.0,
      "step": 2440
    },
    {
      "epoch": 4.3673651359786,
      "grad_norm": 0.401140034198761,
      "learning_rate": 2.1175433564986443e-06,
      "loss": 1.0464,
      "mean_token_accuracy": 0.8111970283091068,
      "num_tokens": 641925120.0,
      "step": 2450
    },
    {
      "epoch": 4.385198395006688,
      "grad_norm": 0.39180195331573486,
      "learning_rate": 2.0002171677208026e-06,
      "loss": 1.0422,
      "mean_token_accuracy": 0.811572577059269,
      "num_tokens": 644546560.0,
      "step": 2460
    },
    {
      "epoch": 4.403031654034775,
      "grad_norm": 0.3858877122402191,
      "learning_rate": 1.8860991325871214e-06,
      "loss": 1.0625,
      "mean_token_accuracy": 0.8080498903989792,
      "num_tokens": 647168000.0,
      "step": 2470
    },
    {
      "epoch": 4.4208649130628626,
      "grad_norm": 0.38444533944129944,
      "learning_rate": 1.7752051689953087e-06,
      "loss": 1.0375,
      "mean_token_accuracy": 0.8128984488546849,
      "num_tokens": 649789440.0,
      "step": 2480
    },
    {
      "epoch": 4.43869817209095,
      "grad_norm": 0.3802605867385864,
      "learning_rate": 1.6675507451293899e-06,
      "loss": 1.0732,
      "mean_token_accuracy": 0.8079281389713288,
      "num_tokens": 652410880.0,
      "step": 2490
    },
    {
      "epoch": 4.456531431119037,
      "grad_norm": 0.39424973726272583,
      "learning_rate": 1.5631508773021165e-06,
      "loss": 1.0759,
      "mean_token_accuracy": 0.8063347272574901,
      "num_tokens": 655032320.0,
      "step": 2500
    },
    {
      "epoch": 4.474364690147125,
      "grad_norm": 0.4043511748313904,
      "learning_rate": 1.4620201278603945e-06,
      "loss": 1.0533,
      "mean_token_accuracy": 0.8101692236959934,
      "num_tokens": 657653760.0,
      "step": 2510
    },
    {
      "epoch": 4.492197949175212,
      "grad_norm": 0.4222579002380371,
      "learning_rate": 1.3641726031540431e-06,
      "loss": 1.0429,
      "mean_token_accuracy": 0.8107840739190578,
      "num_tokens": 660275200.0,
      "step": 2520
    }
  ],
  "logging_steps": 10,
  "max_steps": 2800,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 140,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.287571416664244e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}