{
  "best_global_step": 14000,
  "best_metric": 0.2339961975812912,
  "best_model_checkpoint": "/output/checkpoint-14000",
  "epoch": 1.0,
  "eval_steps": 1000,
  "global_step": 14973,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.9760674737393855,
      "epoch": 0.0006678967097738335,
      "grad_norm": 1.4888298511505127,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 0.5752,
      "mean_token_accuracy": 0.8778422638773918,
      "num_tokens": 73872.0,
      "step": 10
    },
    {
      "entropy": 1.9662065237760544,
      "epoch": 0.001335793419547667,
      "grad_norm": 1.5005749464035034,
      "learning_rate": 6.333333333333333e-07,
      "loss": 0.584,
      "mean_token_accuracy": 0.8795657966285944,
      "num_tokens": 149660.0,
      "step": 20
    },
    {
      "entropy": 1.9851406954228878,
      "epoch": 0.0020036901293215004,
      "grad_norm": 1.6113650798797607,
      "learning_rate": 9.666666666666668e-07,
      "loss": 0.583,
      "mean_token_accuracy": 0.875131793692708,
      "num_tokens": 224839.0,
      "step": 30
    },
    {
      "entropy": 1.9667830072343349,
      "epoch": 0.002671586839095334,
      "grad_norm": 1.4257019758224487,
      "learning_rate": 1.3e-06,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8770494349300861,
      "num_tokens": 299436.0,
      "step": 40
    },
    {
      "entropy": 1.9775025263428687,
      "epoch": 0.0033394835488691674,
      "grad_norm": 1.5861653089523315,
      "learning_rate": 1.6333333333333333e-06,
      "loss": 0.5535,
      "mean_token_accuracy": 0.877307004109025,
      "num_tokens": 371307.0,
      "step": 50
    },
    {
      "entropy": 1.9649440452456475,
      "epoch": 0.004007380258643001,
      "grad_norm": 1.2340285778045654,
      "learning_rate": 1.9666666666666668e-06,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8859475124627352,
      "num_tokens": 445488.0,
      "step": 60
    },
    {
      "entropy": 1.9546820975840091,
      "epoch": 0.0046752769684168345,
      "grad_norm": 0.6933176517486572,
      "learning_rate": 2.3e-06,
      "loss": 0.4616,
      "mean_token_accuracy": 0.9024042788892984,
      "num_tokens": 519666.0,
      "step": 70
    },
    {
      "entropy": 1.966797484457493,
      "epoch": 0.005343173678190668,
      "grad_norm": 0.5129616260528564,
      "learning_rate": 2.6333333333333337e-06,
      "loss": 0.4088,
      "mean_token_accuracy": 0.9164085119962693,
      "num_tokens": 596205.0,
      "step": 80
    },
    {
      "entropy": 1.9509434498846532,
      "epoch": 0.006011070387964501,
      "grad_norm": 0.5326862335205078,
      "learning_rate": 2.966666666666667e-06,
      "loss": 0.3999,
      "mean_token_accuracy": 0.9183030348271132,
      "num_tokens": 669801.0,
      "step": 90
    },
    {
      "entropy": 1.9310075022280215,
      "epoch": 0.006678967097738335,
      "grad_norm": 0.44036024808883667,
      "learning_rate": 3.3e-06,
      "loss": 0.3727,
      "mean_token_accuracy": 0.9235392130911351,
      "num_tokens": 743229.0,
      "step": 100
    },
    {
      "entropy": 1.9586247511208057,
      "epoch": 0.007346863807512169,
      "grad_norm": 0.38791316747665405,
      "learning_rate": 3.633333333333334e-06,
      "loss": 0.3721,
      "mean_token_accuracy": 0.9228693630546332,
      "num_tokens": 818798.0,
      "step": 110
    },
    {
      "entropy": 1.9791630908846856,
      "epoch": 0.008014760517286001,
      "grad_norm": 0.38443174958229065,
      "learning_rate": 3.966666666666667e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.9218462239950895,
      "num_tokens": 891493.0,
      "step": 120
    },
    {
      "entropy": 1.9667378969490528,
      "epoch": 0.008682657227059834,
      "grad_norm": 0.4603234827518463,
      "learning_rate": 4.2999999999999995e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.9221717469394207,
      "num_tokens": 967318.0,
      "step": 130
    },
    {
      "entropy": 1.958837092667818,
      "epoch": 0.009350553936833669,
      "grad_norm": 0.4569478929042816,
      "learning_rate": 4.633333333333334e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.9262894965708256,
      "num_tokens": 1044796.0,
      "step": 140
    },
    {
      "entropy": 1.9429421983659267,
      "epoch": 0.010018450646607502,
      "grad_norm": 0.3992716670036316,
      "learning_rate": 4.966666666666667e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.9276390682905913,
      "num_tokens": 1120596.0,
      "step": 150
    },
    {
      "entropy": 1.9479005947709083,
      "epoch": 0.010686347356381336,
      "grad_norm": 0.5025230646133423,
      "learning_rate": 5.3e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.9234297219663858,
      "num_tokens": 1195637.0,
      "step": 160
    },
    {
      "entropy": 1.9376395657658576,
      "epoch": 0.01135424406615517,
      "grad_norm": 0.39437243342399597,
      "learning_rate": 5.633333333333333e-06,
      "loss": 0.3201,
      "mean_token_accuracy": 0.9296781037002801,
      "num_tokens": 1270664.0,
      "step": 170
    },
    {
      "entropy": 1.906531110405922,
      "epoch": 0.012022140775929002,
      "grad_norm": 0.5597942471504211,
      "learning_rate": 5.9666666666666666e-06,
      "loss": 0.3072,
      "mean_token_accuracy": 0.9306742526590824,
      "num_tokens": 1346257.0,
      "step": 180
    },
    {
      "entropy": 1.9218647584319115,
      "epoch": 0.012690037485702837,
      "grad_norm": 0.4542200267314911,
      "learning_rate": 6.300000000000001e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.9245744608342648,
      "num_tokens": 1420995.0,
      "step": 190
    },
    {
      "entropy": 1.9272994086146356,
      "epoch": 0.01335793419547667,
      "grad_norm": 0.48947060108184814,
      "learning_rate": 6.633333333333333e-06,
      "loss": 0.3015,
      "mean_token_accuracy": 0.9292313981801271,
      "num_tokens": 1494673.0,
      "step": 200
    },
    {
      "entropy": 1.907231931388378,
      "epoch": 0.014025830905250503,
      "grad_norm": 0.43454205989837646,
      "learning_rate": 6.966666666666667e-06,
      "loss": 0.2854,
      "mean_token_accuracy": 0.9343169935047626,
      "num_tokens": 1571582.0,
      "step": 210
    },
    {
      "entropy": 1.921570435166359,
      "epoch": 0.014693727615024337,
      "grad_norm": 2.662630081176758,
      "learning_rate": 7.2999999999999996e-06,
      "loss": 0.2966,
      "mean_token_accuracy": 0.9321486636996269,
      "num_tokens": 1647109.0,
      "step": 220
    },
    {
      "entropy": 1.9276616074144841,
      "epoch": 0.01536162432479817,
      "grad_norm": 0.8096469640731812,
      "learning_rate": 7.633333333333334e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.9320836074650287,
      "num_tokens": 1721663.0,
      "step": 230
    },
    {
      "entropy": 1.9333215937018395,
      "epoch": 0.016029521034572003,
      "grad_norm": 0.39333871006965637,
      "learning_rate": 7.966666666666666e-06,
      "loss": 0.2976,
      "mean_token_accuracy": 0.9332765292376279,
      "num_tokens": 1796477.0,
      "step": 240
    },
    {
      "entropy": 1.95228513777256,
      "epoch": 0.016697417744345838,
      "grad_norm": 0.42003384232521057,
      "learning_rate": 8.3e-06,
      "loss": 0.3032,
      "mean_token_accuracy": 0.9320260498672723,
      "num_tokens": 1872998.0,
      "step": 250
    },
    {
      "entropy": 1.9603711016476155,
      "epoch": 0.01736531445411967,
      "grad_norm": 0.3542785942554474,
      "learning_rate": 8.633333333333334e-06,
      "loss": 0.2961,
      "mean_token_accuracy": 0.9334841158241034,
      "num_tokens": 1948525.0,
      "step": 260
    },
    {
      "entropy": 1.9958221301436425,
      "epoch": 0.018033211163893503,
      "grad_norm": 0.5043424963951111,
      "learning_rate": 8.966666666666668e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.927190800011158,
      "num_tokens": 2024886.0,
      "step": 270
    },
    {
      "entropy": 1.9348272271454334,
      "epoch": 0.018701107873667338,
      "grad_norm": 0.4625996947288513,
      "learning_rate": 9.3e-06,
      "loss": 0.3055,
      "mean_token_accuracy": 0.9334663469344378,
      "num_tokens": 2098633.0,
      "step": 280
    },
    {
      "entropy": 1.9415634967386723,
      "epoch": 0.019369004583441173,
      "grad_norm": 0.5108323693275452,
      "learning_rate": 9.633333333333335e-06,
      "loss": 0.3031,
      "mean_token_accuracy": 0.9323569741100073,
      "num_tokens": 2174388.0,
      "step": 290
    },
    {
      "entropy": 1.995333632081747,
      "epoch": 0.020036901293215004,
      "grad_norm": 0.6340155601501465,
      "learning_rate": 9.966666666666667e-06,
      "loss": 0.3065,
      "mean_token_accuracy": 0.9304188545793295,
      "num_tokens": 2249364.0,
      "step": 300
    },
    {
      "entropy": 2.0197862409055234,
      "epoch": 0.02070479800298884,
      "grad_norm": 0.4793742299079895,
      "learning_rate": 1.03e-05,
      "loss": 0.2886,
      "mean_token_accuracy": 0.9361212570220232,
      "num_tokens": 2322460.0,
      "step": 310
    },
    {
      "entropy": 1.9583041921257973,
      "epoch": 0.021372694712762673,
      "grad_norm": 0.5382503867149353,
      "learning_rate": 1.0633333333333334e-05,
      "loss": 0.2705,
      "mean_token_accuracy": 0.9373382680118084,
      "num_tokens": 2396574.0,
      "step": 320
    },
    {
      "entropy": 1.952714692056179,
      "epoch": 0.022040591422536504,
      "grad_norm": 0.6508220434188843,
      "learning_rate": 1.0966666666666666e-05,
      "loss": 0.2799,
      "mean_token_accuracy": 0.9333360385149717,
      "num_tokens": 2470852.0,
      "step": 330
    },
    {
      "entropy": 1.9514575116336346,
      "epoch": 0.02270848813231034,
      "grad_norm": 0.6904004216194153,
      "learning_rate": 1.13e-05,
      "loss": 0.2869,
      "mean_token_accuracy": 0.9336731854826212,
      "num_tokens": 2546716.0,
      "step": 340
    },
    {
      "entropy": 1.976607009768486,
      "epoch": 0.023376384842084173,
      "grad_norm": 0.5232467651367188,
      "learning_rate": 1.1633333333333334e-05,
      "loss": 0.3064,
      "mean_token_accuracy": 0.9295540764927864,
      "num_tokens": 2620754.0,
      "step": 350
    },
    {
      "entropy": 1.9353629119694233,
      "epoch": 0.024044281551858004,
      "grad_norm": 0.5086498856544495,
      "learning_rate": 1.1966666666666668e-05,
      "loss": 0.2768,
      "mean_token_accuracy": 0.9334879882633687,
      "num_tokens": 2696812.0,
      "step": 360
    },
    {
      "entropy": 1.931234911084175,
      "epoch": 0.02471217826163184,
      "grad_norm": 0.49109113216400146,
      "learning_rate": 1.23e-05,
      "loss": 0.2681,
      "mean_token_accuracy": 0.9392643854022026,
      "num_tokens": 2773201.0,
      "step": 370
    },
    {
      "entropy": 1.9188028134405613,
      "epoch": 0.025380074971405674,
      "grad_norm": 0.5545594692230225,
      "learning_rate": 1.2633333333333333e-05,
      "loss": 0.2727,
      "mean_token_accuracy": 0.9355106376111507,
      "num_tokens": 2847967.0,
      "step": 380
    },
    {
      "entropy": 1.9422715984284877,
      "epoch": 0.026047971681179505,
      "grad_norm": 0.6035627126693726,
      "learning_rate": 1.2966666666666669e-05,
      "loss": 0.274,
      "mean_token_accuracy": 0.9372076652944088,
      "num_tokens": 2923598.0,
      "step": 390
    },
    {
      "entropy": 1.927716436982155,
      "epoch": 0.02671586839095334,
      "grad_norm": 0.5023415684700012,
      "learning_rate": 1.3300000000000001e-05,
      "loss": 0.2971,
      "mean_token_accuracy": 0.9322077054530382,
      "num_tokens": 2998451.0,
      "step": 400
    },
    {
      "entropy": 1.9238519445061684,
      "epoch": 0.027383765100727174,
      "grad_norm": 0.7115201950073242,
      "learning_rate": 1.3633333333333334e-05,
      "loss": 0.2635,
      "mean_token_accuracy": 0.937430489063263,
      "num_tokens": 3073826.0,
      "step": 410
    },
    {
      "entropy": 1.9112761169672012,
      "epoch": 0.028051661810501005,
      "grad_norm": 0.5181813836097717,
      "learning_rate": 1.3966666666666666e-05,
      "loss": 0.2711,
      "mean_token_accuracy": 0.9371557362377644,
      "num_tokens": 3147042.0,
      "step": 420
    },
    {
      "entropy": 1.9072541758418082,
      "epoch": 0.02871955852027484,
      "grad_norm": 0.5583788752555847,
      "learning_rate": 1.43e-05,
      "loss": 0.2909,
      "mean_token_accuracy": 0.9292095638811588,
      "num_tokens": 3221391.0,
      "step": 430
    },
    {
      "entropy": 1.929425187408924,
      "epoch": 0.029387455230048674,
      "grad_norm": 0.5658089518547058,
      "learning_rate": 1.4633333333333334e-05,
      "loss": 0.2773,
      "mean_token_accuracy": 0.9358743727207184,
      "num_tokens": 3297391.0,
      "step": 440
    },
    {
      "entropy": 1.9186258763074875,
      "epoch": 0.030055351939822506,
      "grad_norm": 0.5475177764892578,
      "learning_rate": 1.4966666666666668e-05,
      "loss": 0.2611,
      "mean_token_accuracy": 0.9375769957900048,
      "num_tokens": 3370044.0,
      "step": 450
    },
    {
      "entropy": 1.961367942392826,
      "epoch": 0.03072324864959634,
      "grad_norm": 0.6251649260520935,
      "learning_rate": 1.53e-05,
      "loss": 0.2749,
      "mean_token_accuracy": 0.9348032645881176,
      "num_tokens": 3443720.0,
      "step": 460
    },
    {
      "entropy": 1.9164916306734086,
      "epoch": 0.031391145359370175,
      "grad_norm": 0.6629922986030579,
      "learning_rate": 1.563333333333333e-05,
      "loss": 0.2599,
      "mean_token_accuracy": 0.9388559281826019,
      "num_tokens": 3518739.0,
      "step": 470
    },
    {
      "entropy": 1.9666702941060066,
      "epoch": 0.032059042069144006,
      "grad_norm": 0.5181350111961365,
      "learning_rate": 1.5966666666666667e-05,
      "loss": 0.2754,
      "mean_token_accuracy": 0.9346308093518019,
      "num_tokens": 3591893.0,
      "step": 480
    },
    {
      "entropy": 1.9497105732560158,
      "epoch": 0.03272693877891784,
      "grad_norm": 0.4939156770706177,
      "learning_rate": 1.63e-05,
      "loss": 0.2661,
      "mean_token_accuracy": 0.9378727618604898,
      "num_tokens": 3666092.0,
      "step": 490
    },
    {
      "entropy": 1.9019627816975118,
      "epoch": 0.033394835488691675,
      "grad_norm": 0.5023748278617859,
      "learning_rate": 1.6633333333333336e-05,
      "loss": 0.2801,
      "mean_token_accuracy": 0.937250430509448,
      "num_tokens": 3741532.0,
      "step": 500
    },
    {
      "entropy": 1.9006634928286075,
      "epoch": 0.034062732198465506,
      "grad_norm": 0.49872279167175293,
      "learning_rate": 1.6966666666666668e-05,
      "loss": 0.2745,
      "mean_token_accuracy": 0.9351544339209795,
      "num_tokens": 3818222.0,
      "step": 510
    },
    {
      "entropy": 1.953433834016323,
      "epoch": 0.03473062890823934,
      "grad_norm": 0.6907939314842224,
      "learning_rate": 1.73e-05,
      "loss": 0.2841,
      "mean_token_accuracy": 0.9363411076366901,
      "num_tokens": 3889013.0,
      "step": 520
    },
    {
      "entropy": 1.8758891873061656,
      "epoch": 0.035398525618013175,
      "grad_norm": 0.597548246383667,
      "learning_rate": 1.7633333333333336e-05,
      "loss": 0.2726,
      "mean_token_accuracy": 0.9369025427848101,
      "num_tokens": 3966179.0,
      "step": 530
    },
    {
      "entropy": 1.9165897980332374,
      "epoch": 0.03606642232778701,
      "grad_norm": 0.5589783787727356,
      "learning_rate": 1.796666666666667e-05,
      "loss": 0.2448,
      "mean_token_accuracy": 0.9418925859034062,
      "num_tokens": 4043207.0,
      "step": 540
    },
    {
      "entropy": 1.940786676108837,
      "epoch": 0.03673431903756084,
      "grad_norm": 0.5549450516700745,
      "learning_rate": 1.83e-05,
      "loss": 0.2697,
      "mean_token_accuracy": 0.937014838308096,
      "num_tokens": 4117563.0,
      "step": 550
    },
    {
      "entropy": 1.8771196544170379,
      "epoch": 0.037402215747334676,
      "grad_norm": 0.49763694405555725,
      "learning_rate": 1.8633333333333333e-05,
      "loss": 0.2544,
      "mean_token_accuracy": 0.9363558165729046,
      "num_tokens": 4192004.0,
      "step": 560
    },
    {
      "entropy": 1.8844777837395668,
      "epoch": 0.03807011245710851,
      "grad_norm": 0.43477630615234375,
      "learning_rate": 1.896666666666667e-05,
      "loss": 0.2664,
      "mean_token_accuracy": 0.9387191701680422,
      "num_tokens": 4267951.0,
      "step": 570
    },
    {
      "entropy": 1.896211576461792,
      "epoch": 0.038738009166882345,
      "grad_norm": 0.4881097376346588,
      "learning_rate": 1.93e-05,
      "loss": 0.2718,
      "mean_token_accuracy": 0.936591349542141,
      "num_tokens": 4342071.0,
      "step": 580
    },
    {
      "entropy": 1.8810176849365234,
      "epoch": 0.039405905876656176,
      "grad_norm": 0.5791126489639282,
      "learning_rate": 1.9633333333333334e-05,
      "loss": 0.2556,
      "mean_token_accuracy": 0.9391982112079859,
      "num_tokens": 4416234.0,
      "step": 590
    },
    {
      "entropy": 1.8867235787212848,
      "epoch": 0.04007380258643001,
      "grad_norm": 0.560542106628418,
      "learning_rate": 1.9966666666666666e-05,
      "loss": 0.2718,
      "mean_token_accuracy": 0.9353368133306503,
      "num_tokens": 4491695.0,
      "step": 600
    },
    {
      "entropy": 1.8580191478133201,
      "epoch": 0.040741699296203845,
      "grad_norm": 0.43044254183769226,
      "learning_rate": 2.0300000000000002e-05,
      "loss": 0.2425,
      "mean_token_accuracy": 0.9416664239019156,
      "num_tokens": 4567521.0,
      "step": 610
    },
    {
      "entropy": 1.9017898730933667,
      "epoch": 0.04140959600597768,
      "grad_norm": 0.4845420718193054,
      "learning_rate": 2.0633333333333335e-05,
      "loss": 0.2562,
      "mean_token_accuracy": 0.942466263845563,
      "num_tokens": 4641373.0,
      "step": 620
    },
    {
      "entropy": 1.8724591456353665,
      "epoch": 0.04207749271575151,
      "grad_norm": 0.5080016255378723,
      "learning_rate": 2.0966666666666667e-05,
      "loss": 0.2822,
      "mean_token_accuracy": 0.9362268142402173,
      "num_tokens": 4716043.0,
      "step": 630
    },
    {
      "entropy": 1.8777315124869347,
      "epoch": 0.042745389425525346,
      "grad_norm": 0.46452513337135315,
      "learning_rate": 2.13e-05,
      "loss": 0.2653,
      "mean_token_accuracy": 0.9359228540211916,
      "num_tokens": 4792781.0,
      "step": 640
    },
    {
      "entropy": 1.8617530040442944,
      "epoch": 0.04341328613529918,
      "grad_norm": 0.6030691862106323,
      "learning_rate": 2.1633333333333332e-05,
      "loss": 0.2568,
      "mean_token_accuracy": 0.9388835299760103,
      "num_tokens": 4868270.0,
      "step": 650
    },
    {
      "entropy": 1.8857033200562001,
      "epoch": 0.04408118284507301,
      "grad_norm": 0.4643787443637848,
      "learning_rate": 2.1966666666666668e-05,
      "loss": 0.2634,
      "mean_token_accuracy": 0.9371110614389181,
      "num_tokens": 4942605.0,
      "step": 660
    },
    {
      "entropy": 1.840136867761612,
      "epoch": 0.044749079554846846,
      "grad_norm": 0.45097482204437256,
      "learning_rate": 2.23e-05,
      "loss": 0.2742,
      "mean_token_accuracy": 0.9363797027617693,
      "num_tokens": 5018019.0,
      "step": 670
    },
    {
      "entropy": 1.834625767916441,
      "epoch": 0.04541697626462068,
      "grad_norm": 0.47868287563323975,
      "learning_rate": 2.2633333333333336e-05,
      "loss": 0.272,
      "mean_token_accuracy": 0.938065306097269,
      "num_tokens": 5092429.0,
      "step": 680
    },
    {
      "entropy": 1.84593363404274,
      "epoch": 0.04608487297439451,
      "grad_norm": 0.4770377278327942,
      "learning_rate": 2.2966666666666668e-05,
      "loss": 0.2561,
      "mean_token_accuracy": 0.9369400054216385,
      "num_tokens": 5167616.0,
      "step": 690
    },
    {
      "entropy": 1.8952500484883785,
      "epoch": 0.04675276968416835,
      "grad_norm": 0.47353577613830566,
      "learning_rate": 2.3300000000000004e-05,
      "loss": 0.2912,
      "mean_token_accuracy": 0.9355636093765497,
      "num_tokens": 5240137.0,
      "step": 700
    },
    {
      "entropy": 1.8826350688934326,
      "epoch": 0.04742066639394218,
      "grad_norm": 0.4633476138114929,
      "learning_rate": 2.3633333333333336e-05,
      "loss": 0.2736,
      "mean_token_accuracy": 0.935032133758068,
      "num_tokens": 5314092.0,
      "step": 710
    },
    {
      "entropy": 1.904780062288046,
      "epoch": 0.04808856310371601,
      "grad_norm": 0.49474969506263733,
      "learning_rate": 2.396666666666667e-05,
      "loss": 0.2557,
      "mean_token_accuracy": 0.9400186624377966,
      "num_tokens": 5390094.0,
      "step": 720
    },
    {
      "entropy": 1.8624416716396808,
      "epoch": 0.04875645981348985,
      "grad_norm": 0.5992990136146545,
      "learning_rate": 2.43e-05,
      "loss": 0.2588,
      "mean_token_accuracy": 0.939788530766964,
      "num_tokens": 5463761.0,
      "step": 730
    },
    {
      "entropy": 1.8350231386721134,
      "epoch": 0.04942435652326368,
      "grad_norm": 0.43471571803092957,
      "learning_rate": 2.4633333333333334e-05,
      "loss": 0.2559,
      "mean_token_accuracy": 0.9391793996095658,
      "num_tokens": 5539989.0,
      "step": 740
    },
    {
      "entropy": 1.895322470366955,
      "epoch": 0.05009225323303751,
      "grad_norm": 0.4460161030292511,
      "learning_rate": 2.496666666666667e-05,
      "loss": 0.2667,
      "mean_token_accuracy": 0.9358015276491642,
      "num_tokens": 5612816.0,
      "step": 750
    },
    {
      "entropy": 1.8648789323866368,
      "epoch": 0.05076014994281135,
      "grad_norm": 0.367523193359375,
      "learning_rate": 2.5300000000000002e-05,
      "loss": 0.2606,
      "mean_token_accuracy": 0.9388170868158341,
      "num_tokens": 5686591.0,
      "step": 760
    },
    {
      "entropy": 1.8876647353172302,
      "epoch": 0.05142804665258518,
      "grad_norm": 0.5680462718009949,
      "learning_rate": 2.5633333333333338e-05,
      "loss": 0.2845,
      "mean_token_accuracy": 0.9361521992832422,
      "num_tokens": 5757603.0,
      "step": 770
    },
    {
      "entropy": 1.8878997050225734,
      "epoch": 0.05209594336235901,
      "grad_norm": 0.4175680875778198,
      "learning_rate": 2.5966666666666667e-05,
      "loss": 0.2531,
      "mean_token_accuracy": 0.9397067219018936,
      "num_tokens": 5834420.0,
      "step": 780
    },
    {
      "entropy": 1.8271722830832005,
      "epoch": 0.05276384007213285,
      "grad_norm": 0.5116245746612549,
      "learning_rate": 2.6300000000000002e-05,
      "loss": 0.251,
      "mean_token_accuracy": 0.9420883610844613,
      "num_tokens": 5909083.0,
      "step": 790
    },
    {
      "entropy": 1.8641077525913716,
      "epoch": 0.05343173678190668,
      "grad_norm": 0.4581059515476227,
      "learning_rate": 2.663333333333333e-05,
      "loss": 0.2723,
      "mean_token_accuracy": 0.9368813514709473,
      "num_tokens": 5983992.0,
      "step": 800
    },
    {
      "entropy": 1.857724067568779,
      "epoch": 0.05409963349168051,
      "grad_norm": 0.43746429681777954,
      "learning_rate": 2.6966666666666667e-05,
      "loss": 0.2347,
      "mean_token_accuracy": 0.9450010515749454,
      "num_tokens": 6062463.0,
      "step": 810
    },
    {
      "entropy": 1.8728003606200219,
      "epoch": 0.05476753020145435,
      "grad_norm": 0.38662680983543396,
      "learning_rate": 2.7300000000000003e-05,
      "loss": 0.247,
      "mean_token_accuracy": 0.9403760474175215,
      "num_tokens": 6136824.0,
      "step": 820
    },
    {
      "entropy": 1.898129551112652,
      "epoch": 0.05543542691122818,
      "grad_norm": 0.556283175945282,
      "learning_rate": 2.7633333333333332e-05,
      "loss": 0.2743,
      "mean_token_accuracy": 0.9356398023664951,
      "num_tokens": 6210464.0,
      "step": 830
    },
    {
      "entropy": 1.880861686170101,
      "epoch": 0.05610332362100201,
      "grad_norm": 0.4746190309524536,
      "learning_rate": 2.7966666666666668e-05,
      "loss": 0.265,
      "mean_token_accuracy": 0.936080627143383,
      "num_tokens": 6283597.0,
      "step": 840
    },
    {
      "entropy": 1.854306884109974,
      "epoch": 0.05677122033077585,
      "grad_norm": 0.4764083921909332,
      "learning_rate": 2.83e-05,
      "loss": 0.2639,
      "mean_token_accuracy": 0.9392316583544016,
      "num_tokens": 6357784.0,
      "step": 850
    },
    {
      "entropy": 1.8325285643339158,
      "epoch": 0.05743911704054968,
      "grad_norm": 0.4434099793434143,
      "learning_rate": 2.8633333333333336e-05,
      "loss": 0.2573,
      "mean_token_accuracy": 0.9396443590521812,
      "num_tokens": 6431639.0,
      "step": 860
    },
    {
      "entropy": 1.8051417395472527,
      "epoch": 0.05810701375032351,
      "grad_norm": 0.4234469532966614,
      "learning_rate": 2.8966666666666668e-05,
      "loss": 0.2565,
      "mean_token_accuracy": 0.941559237241745,
      "num_tokens": 6508365.0,
      "step": 870
    },
    {
      "entropy": 1.8686822913587093,
      "epoch": 0.05877491046009735,
      "grad_norm": 0.5077882409095764,
      "learning_rate": 2.93e-05,
      "loss": 0.2659,
      "mean_token_accuracy": 0.9374727483838796,
      "num_tokens": 6582801.0,
      "step": 880
    },
    {
      "entropy": 1.8479201644659042,
      "epoch": 0.05944280716987118,
      "grad_norm": 0.47870588302612305,
      "learning_rate": 2.9633333333333336e-05,
      "loss": 0.2788,
      "mean_token_accuracy": 0.9344479534775019,
      "num_tokens": 6656807.0,
      "step": 890
    },
    {
      "entropy": 1.7955858707427979,
      "epoch": 0.06011070387964501,
      "grad_norm": 0.4300808012485504,
      "learning_rate": 2.9966666666666672e-05,
      "loss": 0.252,
      "mean_token_accuracy": 0.941497677192092,
      "num_tokens": 6732485.0,
      "step": 900
    },
    {
      "entropy": 1.7755248837172986,
      "epoch": 0.06077860058941885,
      "grad_norm": 0.4369029402732849,
      "learning_rate": 3.03e-05,
      "loss": 0.2473,
      "mean_token_accuracy": 0.939782527461648,
      "num_tokens": 6809791.0,
      "step": 910
    },
    {
      "entropy": 1.7887912310659886,
      "epoch": 0.06144649729919268,
      "grad_norm": 0.4319849908351898,
      "learning_rate": 3.063333333333334e-05,
      "loss": 0.2496,
      "mean_token_accuracy": 0.939714727923274,
      "num_tokens": 6886872.0,
      "step": 920
    },
    {
      "entropy": 1.7954976260662079,
      "epoch": 0.06211439400896651,
      "grad_norm": 0.4967508018016815,
      "learning_rate": 3.096666666666666e-05,
      "loss": 0.2511,
      "mean_token_accuracy": 0.9389847174286843,
      "num_tokens": 6960501.0,
      "step": 930
    },
    {
      "entropy": 1.8278648495674132,
      "epoch": 0.06278229071874035,
      "grad_norm": 0.48559603095054626,
      "learning_rate": 3.13e-05,
      "loss": 0.2664,
      "mean_token_accuracy": 0.9423621110618114,
      "num_tokens": 7034210.0,
      "step": 940
    },
    {
      "entropy": 1.9057058945298195,
      "epoch": 0.06345018742851417,
      "grad_norm": 0.4104534685611725,
      "learning_rate": 3.1633333333333334e-05,
      "loss": 0.2665,
      "mean_token_accuracy": 0.9355318166315556,
      "num_tokens": 7107489.0,
      "step": 950
    },
    {
      "entropy": 1.8848116263747214,
      "epoch": 0.06411808413828801,
      "grad_norm": 0.41475149989128113,
      "learning_rate": 3.196666666666667e-05,
      "loss": 0.2457,
      "mean_token_accuracy": 0.9418462350964546,
      "num_tokens": 7182362.0,
      "step": 960
    },
    {
      "entropy": 1.8569753460586071,
      "epoch": 0.06478598084806185,
      "grad_norm": 0.4193580448627472,
      "learning_rate": 3.2300000000000006e-05,
      "loss": 0.2525,
      "mean_token_accuracy": 0.940886041894555,
      "num_tokens": 7259091.0,
      "step": 970
    },
    {
      "entropy": 1.846326508373022,
      "epoch": 0.06545387755783567,
      "grad_norm": 0.4658212661743164,
      "learning_rate": 3.263333333333333e-05,
      "loss": 0.2583,
      "mean_token_accuracy": 0.9409648180007935,
      "num_tokens": 7334380.0,
      "step": 980
    },
    {
      "entropy": 1.851350438594818,
      "epoch": 0.06612177426760951,
      "grad_norm": 0.43780967593193054,
      "learning_rate": 3.296666666666667e-05,
      "loss": 0.2501,
      "mean_token_accuracy": 0.9397942151874303,
      "num_tokens": 7411056.0,
      "step": 990
    },
    {
      "entropy": 1.853810054063797,
      "epoch": 0.06678967097738335,
      "grad_norm": 0.42440807819366455,
      "learning_rate": 3.33e-05,
      "loss": 0.2659,
      "mean_token_accuracy": 0.9382550328969955,
      "num_tokens": 7486847.0,
      "step": 1000
    },
    {
      "epoch": 0.06678967097738335,
      "eval_entropy": 1.8648166310548782,
      "eval_loss": 0.2746169865131378,
      "eval_mean_token_accuracy": 0.9396639612078667,
      "eval_num_tokens": 7486847.0,
      "eval_runtime": 549.5366,
      "eval_samples_per_second": 36.394,
      "eval_steps_per_second": 9.099,
      "step": 1000
    },
    {
      "entropy": 1.8722092419862748,
      "epoch": 0.06745756768715717,
      "grad_norm": 0.414480060338974,
      "learning_rate": 3.3633333333333335e-05,
      "loss": 0.2446,
      "mean_token_accuracy": 0.9428178198635578,
      "num_tokens": 7560292.0,
      "step": 1010
    },
    {
      "entropy": 1.8885526061058044,
      "epoch": 0.06812546439693101,
      "grad_norm": 0.39335963129997253,
      "learning_rate": 3.396666666666667e-05,
      "loss": 0.2707,
      "mean_token_accuracy": 0.9378925491124391,
      "num_tokens": 7635193.0,
      "step": 1020
    },
    {
      "entropy": 1.903367231041193,
      "epoch": 0.06879336110670485,
      "grad_norm": 0.5621363520622253,
      "learning_rate": 3.430000000000001e-05,
      "loss": 0.2653,
      "mean_token_accuracy": 0.9378355484455824,
      "num_tokens": 7709282.0,
      "step": 1030
    },
    {
      "entropy": 1.8545162931084633,
      "epoch": 0.06946125781647867,
      "grad_norm": 0.4494113624095917,
      "learning_rate": 3.463333333333333e-05,
      "loss": 0.2433,
      "mean_token_accuracy": 0.9431285582482815,
      "num_tokens": 7787510.0,
      "step": 1040
    },
    {
      "entropy": 1.8824216805398464,
      "epoch": 0.07012915452625251,
      "grad_norm": 0.34349653124809265,
      "learning_rate": 3.496666666666667e-05,
      "loss": 0.2519,
      "mean_token_accuracy": 0.9404412806034088,
      "num_tokens": 7859430.0,
      "step": 1050
    },
    {
      "entropy": 1.8358929216861726,
      "epoch": 0.07079705123602635,
      "grad_norm": 0.4624893069267273,
      "learning_rate": 3.53e-05,
      "loss": 0.282,
      "mean_token_accuracy": 0.934838455170393,
      "num_tokens": 7933964.0,
      "step": 1060
    },
    {
      "entropy": 1.9042016871273517,
      "epoch": 0.07146494794580018,
      "grad_norm": 0.4160533547401428,
      "learning_rate": 3.563333333333334e-05,
      "loss": 0.2736,
      "mean_token_accuracy": 0.9366271026432514,
      "num_tokens": 8008810.0,
      "step": 1070
    },
    {
      "entropy": 1.8691378943622112,
      "epoch": 0.07213284465557401,
      "grad_norm": 0.5466485619544983,
      "learning_rate": 3.596666666666667e-05,
      "loss": 0.2518,
      "mean_token_accuracy": 0.9419682942330837,
      "num_tokens": 8081971.0,
      "step": 1080
    },
    {
      "entropy": 1.8676304399967194,
      "epoch": 0.07280074136534785,
      "grad_norm": 0.45920252799987793,
      "learning_rate": 3.63e-05,
      "loss": 0.2669,
      "mean_token_accuracy": 0.9384940352290869,
      "num_tokens": 8160031.0,
      "step": 1090
    },
    {
      "entropy": 1.8852659337222577,
      "epoch": 0.07346863807512168,
      "grad_norm": 0.39921435713768005,
      "learning_rate": 3.6633333333333334e-05,
      "loss": 0.2448,
      "mean_token_accuracy": 0.9403918713331223,
      "num_tokens": 8238983.0,
      "step": 1100
    },
    {
      "entropy": 1.8415560849010943,
      "epoch": 0.07413653478489551,
      "grad_norm": 0.40769678354263306,
      "learning_rate": 3.6966666666666666e-05,
      "loss": 0.245,
      "mean_token_accuracy": 0.9413307942450047,
      "num_tokens": 8313284.0,
      "step": 1110
    },
    {
      "entropy": 1.8405307993292808,
      "epoch": 0.07480443149466935,
      "grad_norm": 0.39178162813186646,
      "learning_rate": 3.73e-05,
      "loss": 0.2448,
      "mean_token_accuracy": 0.9406428840011358,
      "num_tokens": 8387088.0,
      "step": 1120
    },
    {
      "entropy": 1.8549068100750445,
      "epoch": 0.07547232820444319,
      "grad_norm": 0.4287874400615692,
      "learning_rate": 3.763333333333334e-05,
      "loss": 0.2463,
      "mean_token_accuracy": 0.9380009301006794,
      "num_tokens": 8461174.0,
      "step": 1130
    },
    {
      "entropy": 1.863625731319189,
      "epoch": 0.07614022491421701,
      "grad_norm": 0.3632742166519165,
      "learning_rate": 3.796666666666667e-05,
      "loss": 0.2375,
      "mean_token_accuracy": 0.9424182560294867,
      "num_tokens": 8537462.0,
      "step": 1140
    },
    {
      "entropy": 1.8476528838276862,
      "epoch": 0.07680812162399085,
      "grad_norm": 0.3408449590206146,
      "learning_rate": 3.83e-05,
      "loss": 0.2393,
      "mean_token_accuracy": 0.9426622208207845,
      "num_tokens": 8613005.0,
      "step": 1150
    },
    {
      "entropy": 1.8576123289763928,
      "epoch": 0.07747601833376469,
      "grad_norm": 0.3981676995754242,
      "learning_rate": 3.8633333333333335e-05,
      "loss": 0.2516,
      "mean_token_accuracy": 0.9414995286613703,
      "num_tokens": 8689743.0,
      "step": 1160
    },
    {
      "entropy": 1.8964253507554532,
      "epoch": 0.07814391504353851,
      "grad_norm": 0.40773627161979675,
      "learning_rate": 3.896666666666667e-05,
      "loss": 0.2459,
      "mean_token_accuracy": 0.9434265218675136,
      "num_tokens": 8761752.0,
      "step": 1170
    },
    {
      "entropy": 1.8615813329815865,
      "epoch": 0.07881181175331235,
      "grad_norm": 0.34091347455978394,
      "learning_rate": 3.9300000000000007e-05,
      "loss": 0.2639,
      "mean_token_accuracy": 0.9368136961013078,
      "num_tokens": 8835914.0,
      "step": 1180
    },
    {
      "entropy": 1.8253806605935097,
      "epoch": 0.07947970846308619,
      "grad_norm": 0.32698726654052734,
      "learning_rate": 3.963333333333333e-05,
      "loss": 0.2554,
      "mean_token_accuracy": 0.9404144994914532,
      "num_tokens": 8910566.0,
      "step": 1190
    },
    {
      "entropy": 1.8871965758502482,
      "epoch": 0.08014760517286001,
      "grad_norm": 0.32207944989204407,
      "learning_rate": 3.996666666666667e-05,
      "loss": 0.2593,
      "mean_token_accuracy": 0.9369674678891897,
      "num_tokens": 8987453.0,
      "step": 1200
    },
    {
      "entropy": 1.8854286082088947,
      "epoch": 0.08081550188263385,
      "grad_norm": 0.40935617685317993,
      "learning_rate": 4.0300000000000004e-05,
      "loss": 0.2691,
      "mean_token_accuracy": 0.9358607489615679,
      "num_tokens": 9061161.0,
      "step": 1210
    },
    {
      "entropy": 1.8666034035384655,
      "epoch": 0.08148339859240769,
      "grad_norm": 0.40217721462249756,
      "learning_rate": 4.0633333333333336e-05,
      "loss": 0.2474,
      "mean_token_accuracy": 0.9405681319534779,
      "num_tokens": 9135309.0,
      "step": 1220
    },
    {
      "entropy": 1.8529215894639492,
      "epoch": 0.08215129530218152,
      "grad_norm": 0.3745631277561188,
      "learning_rate": 4.096666666666667e-05,
      "loss": 0.2383,
      "mean_token_accuracy": 0.9417877610772848,
      "num_tokens": 9206268.0,
      "step": 1230
    },
    {
      "entropy": 1.8744257062673568,
      "epoch": 0.08281919201195535,
      "grad_norm": 0.41536879539489746,
      "learning_rate": 4.13e-05,
      "loss": 0.2558,
      "mean_token_accuracy": 0.9366620749235153,
      "num_tokens": 9280551.0,
      "step": 1240
    },
    {
      "entropy": 1.860572198778391,
      "epoch": 0.08348708872172919,
      "grad_norm": 0.38959744572639465,
      "learning_rate": 4.1633333333333333e-05,
      "loss": 0.2545,
      "mean_token_accuracy": 0.9395768508315087,
      "num_tokens": 9355255.0,
      "step": 1250
    },
    {
      "entropy": 1.8331671096384525,
      "epoch": 0.08415498543150302,
      "grad_norm": 0.38928812742233276,
      "learning_rate": 4.196666666666667e-05,
      "loss": 0.2547,
      "mean_token_accuracy": 0.9387797378003597,
      "num_tokens": 9430504.0,
      "step": 1260
    },
    {
      "entropy": 1.8536636359989642,
      "epoch": 0.08482288214127685,
      "grad_norm": 0.33527395129203796,
      "learning_rate": 4.23e-05,
      "loss": 0.2446,
      "mean_token_accuracy": 0.9408517919480801,
      "num_tokens": 9506407.0,
      "step": 1270
    },
    {
      "entropy": 1.8080134324729442,
      "epoch": 0.08549077885105069,
      "grad_norm": 0.420559287071228,
      "learning_rate": 4.263333333333334e-05,
      "loss": 0.2545,
      "mean_token_accuracy": 0.9404942344874143,
      "num_tokens": 9582031.0,
      "step": 1280
    },
    {
      "entropy": 1.8781551785767079,
      "epoch": 0.08615867556082452,
      "grad_norm": 0.487653911113739,
      "learning_rate": 4.296666666666666e-05,
      "loss": 0.2578,
      "mean_token_accuracy": 0.935891279578209,
      "num_tokens": 9656802.0,
      "step": 1290
    },
    {
      "entropy": 1.852930338680744,
      "epoch": 0.08682657227059835,
      "grad_norm": 0.33225956559181213,
      "learning_rate": 4.33e-05,
      "loss": 0.253,
      "mean_token_accuracy": 0.9419618722051382,
      "num_tokens": 9730623.0,
      "step": 1300
    },
    {
      "entropy": 1.8633646860718727,
      "epoch": 0.08749446898037219,
      "grad_norm": 0.39087024331092834,
      "learning_rate": 4.3633333333333335e-05,
      "loss": 0.2464,
      "mean_token_accuracy": 0.9423338647931814,
      "num_tokens": 9803462.0,
      "step": 1310
    },
    {
      "entropy": 1.808484634757042,
      "epoch": 0.08816236569014602,
      "grad_norm": 0.3942349851131439,
      "learning_rate": 4.396666666666667e-05,
      "loss": 0.2413,
      "mean_token_accuracy": 0.9396913766860961,
      "num_tokens": 9879104.0,
      "step": 1320
    },
    {
      "entropy": 1.8243934623897076,
      "epoch": 0.08883026239991985,
      "grad_norm": 0.4019410014152527,
      "learning_rate": 4.43e-05,
      "loss": 0.2474,
      "mean_token_accuracy": 0.9448523435741663,
      "num_tokens": 9954180.0,
      "step": 1330
    },
    {
      "entropy": 1.8570265740156173,
      "epoch": 0.08949815910969369,
      "grad_norm": 0.357666015625,
      "learning_rate": 4.463333333333334e-05,
      "loss": 0.2558,
      "mean_token_accuracy": 0.9409103903919458,
      "num_tokens": 10028028.0,
      "step": 1340
    },
    {
      "entropy": 1.8285759411752225,
      "epoch": 0.09016605581946752,
      "grad_norm": 0.4638359248638153,
      "learning_rate": 4.496666666666667e-05,
      "loss": 0.2386,
      "mean_token_accuracy": 0.9455917440354824,
      "num_tokens": 10103949.0,
      "step": 1350
    },
    {
      "entropy": 1.8921574287116527,
      "epoch": 0.09083395252924135,
      "grad_norm": 0.3664829730987549,
      "learning_rate": 4.53e-05,
      "loss": 0.2514,
      "mean_token_accuracy": 0.9409088790416718,
      "num_tokens": 10179965.0,
      "step": 1360
    },
    {
      "entropy": 1.909767533838749,
      "epoch": 0.09150184923901519,
      "grad_norm": 0.3954828679561615,
      "learning_rate": 4.5633333333333336e-05,
      "loss": 0.2641,
      "mean_token_accuracy": 0.9392028257250786,
      "num_tokens": 10253031.0,
      "step": 1370
    },
    {
      "entropy": 1.897636280953884,
      "epoch": 0.09216974594878902,
      "grad_norm": 0.3309478461742401,
      "learning_rate": 4.596666666666667e-05,
      "loss": 0.2536,
      "mean_token_accuracy": 0.9406967829912901,
      "num_tokens": 10327483.0,
      "step": 1380
    },
    {
      "entropy": 1.8735629595816135,
      "epoch": 0.09283764265856286,
      "grad_norm": 0.3709751069545746,
      "learning_rate": 4.630000000000001e-05,
      "loss": 0.2559,
      "mean_token_accuracy": 0.9381070993840694,
      "num_tokens": 10400671.0,
      "step": 1390
    },
    {
      "entropy": 1.8461422972381114,
      "epoch": 0.0935055393683367,
      "grad_norm": 0.34269604086875916,
      "learning_rate": 4.663333333333333e-05,
      "loss": 0.2486,
      "mean_token_accuracy": 0.9399528317153454,
      "num_tokens": 10475441.0,
      "step": 1400
    },
    {
      "entropy": 1.8211704172194003,
      "epoch": 0.09417343607811052,
      "grad_norm": 0.401239275932312,
      "learning_rate": 4.696666666666667e-05,
      "loss": 0.248,
      "mean_token_accuracy": 0.9393475376069546,
      "num_tokens": 10550394.0,
      "step": 1410
    },
    {
      "entropy": 1.892047306150198,
      "epoch": 0.09484133278788436,
      "grad_norm": 0.39942097663879395,
      "learning_rate": 4.73e-05,
      "loss": 0.2635,
      "mean_token_accuracy": 0.9364817954599858,
      "num_tokens": 10628054.0,
      "step": 1420
    },
    {
      "entropy": 1.8521707750856877,
      "epoch": 0.0955092294976582,
      "grad_norm": 0.4278346598148346,
      "learning_rate": 4.763333333333334e-05,
      "loss": 0.2323,
      "mean_token_accuracy": 0.9443262364715338,
      "num_tokens": 10701626.0,
      "step": 1430
    },
    {
      "entropy": 1.8943075209856033,
      "epoch": 0.09617712620743202,
      "grad_norm": 0.3699016869068146,
      "learning_rate": 4.796666666666667e-05,
      "loss": 0.2737,
      "mean_token_accuracy": 0.9367077428847551,
      "num_tokens": 10774072.0,
      "step": 1440
    },
    {
      "entropy": 1.8664367534220219,
      "epoch": 0.09684502291720586,
      "grad_norm": 0.3882374167442322,
      "learning_rate": 4.83e-05,
      "loss": 0.2576,
      "mean_token_accuracy": 0.9367181625217199,
      "num_tokens": 10847357.0,
      "step": 1450
    },
    {
      "entropy": 1.7965730026364326,
      "epoch": 0.0975129196269797,
      "grad_norm": 0.32984229922294617,
      "learning_rate": 4.8633333333333334e-05,
      "loss": 0.2264,
      "mean_token_accuracy": 0.9448713529855013,
      "num_tokens": 10924525.0,
      "step": 1460
    },
    {
      "entropy": 1.8047875829041005,
      "epoch": 0.09818081633675352,
      "grad_norm": 0.4185302257537842,
      "learning_rate": 4.8966666666666667e-05,
      "loss": 0.2462,
      "mean_token_accuracy": 0.9445567015558481,
      "num_tokens": 10997448.0,
      "step": 1470
    },
    {
      "entropy": 1.8770555928349495,
      "epoch": 0.09884871304652736,
      "grad_norm": 0.34383824467658997,
      "learning_rate": 4.93e-05,
      "loss": 0.2565,
      "mean_token_accuracy": 0.9404840290546417,
      "num_tokens": 11073389.0,
      "step": 1480
    },
    {
      "entropy": 1.843934503942728,
      "epoch": 0.0995166097563012,
      "grad_norm": 0.3678169846534729,
      "learning_rate": 4.963333333333334e-05,
      "loss": 0.2416,
      "mean_token_accuracy": 0.9416554532945156,
      "num_tokens": 11147812.0,
      "step": 1490
    },
    {
      "entropy": 1.837416634708643,
      "epoch": 0.10018450646607502,
      "grad_norm": 0.372890830039978,
      "learning_rate": 4.996666666666667e-05,
      "loss": 0.2605,
      "mean_token_accuracy": 0.9366992872208357,
      "num_tokens": 11222157.0,
      "step": 1500
    },
    {
      "entropy": 1.804855602234602,
      "epoch": 0.10085240317584886,
      "grad_norm": 0.40893515944480896,
      "learning_rate": 4.9999945059002565e-05,
      "loss": 0.2549,
      "mean_token_accuracy": 0.940057422220707,
      "num_tokens": 11295930.0,
      "step": 1510
    },
    {
      "entropy": 1.7543303154408931,
      "epoch": 0.1015202998856227,
      "grad_norm": 0.3919351398944855,
      "learning_rate": 4.999975513981592e-05,
      "loss": 0.2164,
      "mean_token_accuracy": 0.9464322041720152,
      "num_tokens": 11370899.0,
      "step": 1520
    },
    {
      "entropy": 1.8050099305808545,
      "epoch": 0.10218819659539652,
      "grad_norm": 0.3314148187637329,
      "learning_rate": 4.99994295651885e-05,
      "loss": 0.2334,
      "mean_token_accuracy": 0.9432413805276155,
      "num_tokens": 11447828.0,
      "step": 1530
    },
    {
      "entropy": 1.8214650124311447,
      "epoch": 0.10285609330517036,
      "grad_norm": 0.316817045211792,
      "learning_rate": 4.999896833689051e-05,
      "loss": 0.2443,
      "mean_token_accuracy": 0.9421243309974671,
      "num_tokens": 11520817.0,
      "step": 1540
    },
    {
      "entropy": 1.8117822676897049,
      "epoch": 0.1035239900149442,
      "grad_norm": 0.33268994092941284,
      "learning_rate": 4.999837145742971e-05,
      "loss": 0.2618,
      "mean_token_accuracy": 0.9390293166041375,
      "num_tokens": 11594522.0,
      "step": 1550
    },
    {
      "entropy": 1.8389830827713012,
      "epoch": 0.10419188672471802,
      "grad_norm": 0.4531651735305786,
      "learning_rate": 4.9997638930051434e-05,
      "loss": 0.2413,
      "mean_token_accuracy": 0.9419956561177969,
      "num_tokens": 11668587.0,
      "step": 1560
    },
    {
      "entropy": 1.7972249872982502,
      "epoch": 0.10485978343449186,
      "grad_norm": 0.35895809531211853,
      "learning_rate": 4.999677075873853e-05,
      "loss": 0.2458,
      "mean_token_accuracy": 0.9400338169187308,
      "num_tokens": 11745785.0,
      "step": 1570
    },
    {
      "entropy": 1.8368936344981193,
      "epoch": 0.1055276801442657,
      "grad_norm": 0.37567827105522156,
      "learning_rate": 4.999576694821137e-05,
      "loss": 0.2593,
      "mean_token_accuracy": 0.9367735404521227,
      "num_tokens": 11819941.0,
      "step": 1580
    },
    {
      "entropy": 1.7924836911261082,
      "epoch": 0.10619557685403952,
      "grad_norm": 0.3485127091407776,
      "learning_rate": 4.999462750392782e-05,
      "loss": 0.252,
      "mean_token_accuracy": 0.9390415012836456,
      "num_tokens": 11894898.0,
      "step": 1590
    },
    {
      "entropy": 1.7923610121011735,
      "epoch": 0.10686347356381336,
      "grad_norm": 0.4060329496860504,
      "learning_rate": 4.9993352432083206e-05,
      "loss": 0.2386,
      "mean_token_accuracy": 0.9422769341617823,
      "num_tokens": 11971167.0,
      "step": 1600
    },
    {
      "entropy": 1.8334701210260391,
      "epoch": 0.1075313702735872,
      "grad_norm": 0.36048150062561035,
      "learning_rate": 4.999194173961028e-05,
      "loss": 0.2367,
      "mean_token_accuracy": 0.9419048838317394,
      "num_tokens": 12045417.0,
      "step": 1610
    },
    {
      "entropy": 1.7842422433197498,
      "epoch": 0.10819926698336102,
      "grad_norm": 0.36341556906700134,
      "learning_rate": 4.9990395434179176e-05,
      "loss": 0.2528,
      "mean_token_accuracy": 0.9400053527206182,
      "num_tokens": 12120969.0,
      "step": 1620
    },
    {
      "entropy": 1.8372097842395305,
      "epoch": 0.10886716369313486,
      "grad_norm": 0.35848885774612427,
      "learning_rate": 4.998871352419739e-05,
      "loss": 0.2706,
      "mean_token_accuracy": 0.9359296333044768,
      "num_tokens": 12194830.0,
      "step": 1630
    },
    {
      "entropy": 1.8942107573151588,
      "epoch": 0.1095350604029087,
      "grad_norm": 0.37302595376968384,
      "learning_rate": 4.998689601880969e-05,
      "loss": 0.252,
      "mean_token_accuracy": 0.9393172737210989,
      "num_tokens": 12268478.0,
      "step": 1640
    },
    {
      "entropy": 1.846186748892069,
      "epoch": 0.11020295711268252,
      "grad_norm": 0.36369460821151733,
      "learning_rate": 4.998494292789814e-05,
      "loss": 0.2571,
      "mean_token_accuracy": 0.9378128115087747,
      "num_tokens": 12341656.0,
      "step": 1650
    },
    {
      "entropy": 1.8226500421762466,
      "epoch": 0.11087085382245636,
      "grad_norm": 0.35995057225227356,
      "learning_rate": 4.998285426208197e-05,
      "loss": 0.2593,
      "mean_token_accuracy": 0.9392500806599855,
      "num_tokens": 12417077.0,
      "step": 1660
    },
    {
      "entropy": 1.7849808871746062,
      "epoch": 0.1115387505322302,
      "grad_norm": 0.31248095631599426,
      "learning_rate": 4.998063003271756e-05,
      "loss": 0.2382,
      "mean_token_accuracy": 0.9415343657135964,
      "num_tokens": 12494685.0,
      "step": 1670
    },
    {
      "entropy": 1.8088674053549767,
      "epoch": 0.11220664724200402,
      "grad_norm": 0.3842123746871948,
      "learning_rate": 4.997827025189837e-05,
      "loss": 0.261,
      "mean_token_accuracy": 0.9392439745366573,
      "num_tokens": 12568525.0,
      "step": 1680
    },
    {
      "entropy": 1.7959242567420006,
      "epoch": 0.11287454395177786,
      "grad_norm": 0.40647968649864197,
      "learning_rate": 4.997577493245486e-05,
      "loss": 0.2695,
      "mean_token_accuracy": 0.9369698017835617,
      "num_tokens": 12642491.0,
      "step": 1690
    },
    {
      "entropy": 1.829491986334324,
      "epoch": 0.1135424406615517,
      "grad_norm": 0.32783016562461853,
      "learning_rate": 4.997314408795446e-05,
      "loss": 0.2409,
      "mean_token_accuracy": 0.9426718652248383,
      "num_tokens": 12715799.0,
      "step": 1700
    },
    {
      "entropy": 1.7980324625968933,
      "epoch": 0.11421033737132552,
      "grad_norm": 0.38863658905029297,
      "learning_rate": 4.997037773270143e-05,
      "loss": 0.2246,
      "mean_token_accuracy": 0.9470432173460722,
      "num_tokens": 12791044.0,
      "step": 1710
    },
    {
      "entropy": 1.780588860809803,
      "epoch": 0.11487823408109936,
      "grad_norm": 0.36860790848731995,
      "learning_rate": 4.996747588173688e-05,
      "loss": 0.2704,
      "mean_token_accuracy": 0.9387739099562168,
      "num_tokens": 12864717.0,
      "step": 1720
    },
    {
      "entropy": 1.8167456299066544,
      "epoch": 0.1155461307908732,
      "grad_norm": 0.40437668561935425,
      "learning_rate": 4.9964438550838566e-05,
      "loss": 0.2482,
      "mean_token_accuracy": 0.9406302180141211,
      "num_tokens": 12936931.0,
      "step": 1730
    },
    {
      "entropy": 1.7739999622106553,
      "epoch": 0.11621402750064702,
      "grad_norm": 0.42025813460350037,
      "learning_rate": 4.9961265756520914e-05,
      "loss": 0.2621,
      "mean_token_accuracy": 0.9397819988429547,
      "num_tokens": 13009783.0,
      "step": 1740
    },
    {
      "entropy": 1.8471427731215955,
      "epoch": 0.11688192421042086,
      "grad_norm": 0.24152179062366486,
      "learning_rate": 4.9957957516034854e-05,
      "loss": 0.2583,
      "mean_token_accuracy": 0.9385230287909507,
      "num_tokens": 13087594.0,
      "step": 1750
    },
    {
      "entropy": 1.829470805078745,
      "epoch": 0.1175498209201947,
      "grad_norm": 0.3349377512931824,
      "learning_rate": 4.995451384736778e-05,
      "loss": 0.2385,
      "mean_token_accuracy": 0.942843297868967,
      "num_tokens": 13162846.0,
      "step": 1760
    },
    {
      "entropy": 1.8083517365157604,
      "epoch": 0.11821771762996852,
      "grad_norm": 0.33394306898117065,
      "learning_rate": 4.995093476924341e-05,
      "loss": 0.268,
      "mean_token_accuracy": 0.9358123634010553,
      "num_tokens": 13235219.0,
      "step": 1770
    },
    {
      "entropy": 1.8390862189233304,
      "epoch": 0.11888561433974236,
      "grad_norm": 0.2808772623538971,
      "learning_rate": 4.9947220301121714e-05,
      "loss": 0.2561,
      "mean_token_accuracy": 0.9398859009146691,
      "num_tokens": 13309921.0,
      "step": 1780
    },
    {
      "entropy": 1.8172144912183286,
      "epoch": 0.1195535110495162,
      "grad_norm": 0.29831817746162415,
      "learning_rate": 4.994337046319881e-05,
      "loss": 0.2499,
      "mean_token_accuracy": 0.9418265763670206,
      "num_tokens": 13383573.0,
      "step": 1790
    },
    {
      "entropy": 1.7965184792876243,
      "epoch": 0.12022140775929002,
      "grad_norm": 0.34291186928749084,
      "learning_rate": 4.9939385276406804e-05,
      "loss": 0.2532,
      "mean_token_accuracy": 0.9384090960025787,
      "num_tokens": 13459808.0,
      "step": 1800
    },
    {
      "entropy": 1.7756263740360736,
      "epoch": 0.12088930446906386,
      "grad_norm": 0.3427332639694214,
      "learning_rate": 4.9935264762413744e-05,
      "loss": 0.2329,
      "mean_token_accuracy": 0.9460334207862615,
      "num_tokens": 13534986.0,
      "step": 1810
    },
    {
      "entropy": 1.7850582242012023,
      "epoch": 0.1215572011788377,
      "grad_norm": 0.4083283841609955,
      "learning_rate": 4.993100894362345e-05,
      "loss": 0.23,
      "mean_token_accuracy": 0.9445747293531894,
      "num_tokens": 13611445.0,
      "step": 1820
    },
    {
      "entropy": 1.810771866887808,
      "epoch": 0.12222509788861152,
      "grad_norm": 0.3793506324291229,
      "learning_rate": 4.992661784317545e-05,
      "loss": 0.2515,
      "mean_token_accuracy": 0.9381474364548922,
      "num_tokens": 13686160.0,
      "step": 1830
    },
    {
      "entropy": 1.772794821858406,
      "epoch": 0.12289299459838536,
      "grad_norm": 0.41430842876434326,
      "learning_rate": 4.9922091484944776e-05,
      "loss": 0.2567,
      "mean_token_accuracy": 0.9369965393096209,
      "num_tokens": 13760446.0,
      "step": 1840
    },
    {
      "entropy": 1.8180281892418861,
      "epoch": 0.1235608913081592,
      "grad_norm": 0.4711560904979706,
      "learning_rate": 4.991742989354192e-05,
      "loss": 0.2548,
      "mean_token_accuracy": 0.9408668961375952,
      "num_tokens": 13834551.0,
      "step": 1850
    },
    {
      "entropy": 1.7984913356602192,
      "epoch": 0.12422878801793302,
      "grad_norm": 0.33564695715904236,
      "learning_rate": 4.991263309431259e-05,
      "loss": 0.2328,
      "mean_token_accuracy": 0.9441933192312717,
      "num_tokens": 13910991.0,
      "step": 1860
    },
    {
      "entropy": 1.7778552062809467,
      "epoch": 0.12489668472770686,
      "grad_norm": 0.32167258858680725,
      "learning_rate": 4.990770111333772e-05,
      "loss": 0.2361,
      "mean_token_accuracy": 0.9451872866600752,
      "num_tokens": 13985904.0,
      "step": 1870
    },
    {
      "entropy": 1.77905383259058,
      "epoch": 0.1255645814374807,
      "grad_norm": 0.3849715292453766,
      "learning_rate": 4.990263397743318e-05,
      "loss": 0.2402,
      "mean_token_accuracy": 0.9438797861337662,
      "num_tokens": 14061330.0,
      "step": 1880
    },
    {
      "entropy": 1.85131838619709,
      "epoch": 0.12623247814725452,
      "grad_norm": 0.3501381576061249,
      "learning_rate": 4.989743171414973e-05,
      "loss": 0.2627,
      "mean_token_accuracy": 0.9393126614391804,
      "num_tokens": 14133348.0,
      "step": 1890
    },
    {
      "entropy": 1.7933654822409153,
      "epoch": 0.12690037485702835,
      "grad_norm": 0.3642660677433014,
      "learning_rate": 4.9892094351772844e-05,
      "loss": 0.2389,
      "mean_token_accuracy": 0.9438054759055376,
      "num_tokens": 14207435.0,
      "step": 1900
    },
    {
      "entropy": 1.8201318435370921,
      "epoch": 0.1275682715668022,
      "grad_norm": 0.35850223898887634,
      "learning_rate": 4.9886621919322515e-05,
      "loss": 0.2629,
      "mean_token_accuracy": 0.9394240003079176,
      "num_tokens": 14282314.0,
      "step": 1910
    },
    {
      "entropy": 1.8730139046907426,
      "epoch": 0.12823616827657602,
      "grad_norm": 0.3666210174560547,
      "learning_rate": 4.9881014446553165e-05,
      "loss": 0.243,
      "mean_token_accuracy": 0.9434024095535278,
      "num_tokens": 14356840.0,
      "step": 1920
    },
    {
      "entropy": 1.8122118309140205,
      "epoch": 0.12890406498634985,
      "grad_norm": 0.2802065312862396,
      "learning_rate": 4.9875271963953446e-05,
      "loss": 0.2275,
      "mean_token_accuracy": 0.9431390129029751,
      "num_tokens": 14431012.0,
      "step": 1930
    },
    {
      "entropy": 1.806647739559412,
      "epoch": 0.1295719616961237,
      "grad_norm": 0.3309171795845032,
      "learning_rate": 4.9869394502746056e-05,
      "loss": 0.2593,
      "mean_token_accuracy": 0.9391654759645462,
      "num_tokens": 14505496.0,
      "step": 1940
    },
    {
      "entropy": 1.836927954852581,
      "epoch": 0.13023985840589752,
      "grad_norm": 0.26602479815483093,
      "learning_rate": 4.9863382094887604e-05,
      "loss": 0.2564,
      "mean_token_accuracy": 0.9381305709481239,
      "num_tokens": 14580129.0,
      "step": 1950
    },
    {
      "entropy": 1.8698457337915897,
      "epoch": 0.13090775511567135,
      "grad_norm": 0.2941421866416931,
      "learning_rate": 4.9857234773068425e-05,
      "loss": 0.2503,
      "mean_token_accuracy": 0.939929100498557,
      "num_tokens": 14656073.0,
      "step": 1960
    },
    {
      "entropy": 1.8299203373491764,
      "epoch": 0.1315756518254452,
      "grad_norm": 0.34756720066070557,
      "learning_rate": 4.985095257071241e-05,
      "loss": 0.2369,
      "mean_token_accuracy": 0.9416007433086634,
      "num_tokens": 14729935.0,
      "step": 1970
    },
    {
      "entropy": 1.8041098691523074,
      "epoch": 0.13224354853521902,
      "grad_norm": 0.3420882523059845,
      "learning_rate": 4.98445355219768e-05,
      "loss": 0.2276,
      "mean_token_accuracy": 0.9446675062179566,
      "num_tokens": 14804005.0,
      "step": 1980
    },
    {
      "entropy": 1.7574528865516186,
      "epoch": 0.13291144524499285,
      "grad_norm": 0.31510859727859497,
      "learning_rate": 4.983798366175202e-05,
      "loss": 0.2476,
      "mean_token_accuracy": 0.9438613794744015,
      "num_tokens": 14879452.0,
      "step": 1990
    },
    {
      "entropy": 1.8129021555185318,
      "epoch": 0.1335793419547667,
      "grad_norm": 0.31414422392845154,
      "learning_rate": 4.9831297025661483e-05,
      "loss": 0.2398,
      "mean_token_accuracy": 0.9419260278344155,
      "num_tokens": 14956825.0,
      "step": 2000
    },
    {
      "epoch": 0.1335793419547667,
      "eval_entropy": 1.7998176928043366,
      "eval_loss": 0.2604806125164032,
      "eval_mean_token_accuracy": 0.9421368992447853,
      "eval_num_tokens": 14956825.0,
      "eval_runtime": 553.5898,
      "eval_samples_per_second": 36.128,
      "eval_steps_per_second": 9.032,
      "step": 2000
    },
    {
      "entropy": 1.7944769166409968,
      "epoch": 0.13424723866454052,
      "grad_norm": 0.4020504355430603,
      "learning_rate": 4.98244756500614e-05,
      "loss": 0.2509,
      "mean_token_accuracy": 0.9413325522094965,
      "num_tokens": 15031758.0,
      "step": 2010
    },
    {
      "entropy": 1.7625232838094234,
      "epoch": 0.13491513537431435,
      "grad_norm": 0.35861051082611084,
      "learning_rate": 4.9817519572040584e-05,
      "loss": 0.2421,
      "mean_token_accuracy": 0.9444488797336816,
      "num_tokens": 15106692.0,
      "step": 2020
    },
    {
      "entropy": 1.8512812927365303,
      "epoch": 0.1355830320840882,
      "grad_norm": 0.2726729214191437,
      "learning_rate": 4.981042882942024e-05,
      "loss": 0.2524,
      "mean_token_accuracy": 0.9382127381861209,
      "num_tokens": 15179306.0,
      "step": 2030
    },
    {
      "entropy": 1.80864247828722,
      "epoch": 0.13625092879386203,
      "grad_norm": 0.3192000985145569,
      "learning_rate": 4.980320346075376e-05,
      "loss": 0.2513,
      "mean_token_accuracy": 0.9387617770582437,
      "num_tokens": 15254467.0,
      "step": 2040
    },
    {
      "entropy": 1.77226110547781,
      "epoch": 0.13691882550363585,
      "grad_norm": 0.39089086651802063,
      "learning_rate": 4.9795843505326534e-05,
      "loss": 0.2376,
      "mean_token_accuracy": 0.9425041008740663,
      "num_tokens": 15325693.0,
      "step": 2050
    },
    {
      "entropy": 1.744492419809103,
      "epoch": 0.1375867222134097,
      "grad_norm": 0.29420822858810425,
      "learning_rate": 4.97883490031557e-05,
      "loss": 0.2298,
      "mean_token_accuracy": 0.9479223694652319,
      "num_tokens": 15400502.0,
      "step": 2060
    },
    {
      "entropy": 1.7660885341465473,
      "epoch": 0.13825461892318353,
      "grad_norm": 0.36348381638526917,
      "learning_rate": 4.9780719994989946e-05,
      "loss": 0.2366,
      "mean_token_accuracy": 0.9435297749936581,
      "num_tokens": 15473805.0,
      "step": 2070
    },
    {
      "entropy": 1.780941829830408,
      "epoch": 0.13892251563295735,
      "grad_norm": 0.3287426233291626,
      "learning_rate": 4.977295652230931e-05,
      "loss": 0.2607,
      "mean_token_accuracy": 0.9359222516417504,
      "num_tokens": 15548685.0,
      "step": 2080
    },
    {
      "entropy": 1.8226717375218868,
      "epoch": 0.1395904123427312,
      "grad_norm": 0.4171643555164337,
      "learning_rate": 4.976505862732493e-05,
      "loss": 0.2607,
      "mean_token_accuracy": 0.9392563026398421,
      "num_tokens": 15621589.0,
      "step": 2090
    },
    {
      "entropy": 1.7729314967989922,
      "epoch": 0.14025830905250503,
      "grad_norm": 0.359145849943161,
      "learning_rate": 4.9757026352978784e-05,
      "loss": 0.2248,
      "mean_token_accuracy": 0.9455544374883175,
      "num_tokens": 15695417.0,
      "step": 2100
    },
    {
      "entropy": 1.7612593069672584,
      "epoch": 0.14092620576227885,
      "grad_norm": 0.3510279655456543,
      "learning_rate": 4.974885974294352e-05,
      "loss": 0.244,
      "mean_token_accuracy": 0.9400189932435751,
      "num_tokens": 15769288.0,
      "step": 2110
    },
    {
      "entropy": 1.737210039794445,
      "epoch": 0.1415941024720527,
      "grad_norm": 0.2984858751296997,
      "learning_rate": 4.97405588416222e-05,
      "loss": 0.2305,
      "mean_token_accuracy": 0.9429280549287796,
      "num_tokens": 15845511.0,
      "step": 2120
    },
    {
      "entropy": 1.737841621041298,
      "epoch": 0.14226199918182653,
      "grad_norm": 0.3221711814403534,
      "learning_rate": 4.9732123694147994e-05,
      "loss": 0.2396,
      "mean_token_accuracy": 0.9418632112443447,
      "num_tokens": 15920447.0,
      "step": 2130
    },
    {
      "entropy": 1.770331757515669,
      "epoch": 0.14292989589160035,
      "grad_norm": 0.5877280235290527,
      "learning_rate": 4.972355434638406e-05,
      "loss": 0.2596,
      "mean_token_accuracy": 0.938718581944704,
      "num_tokens": 15994725.0,
      "step": 2140
    },
    {
      "entropy": 1.8453688018023968,
      "epoch": 0.1435977926013742,
      "grad_norm": 0.34415510296821594,
      "learning_rate": 4.971485084492317e-05,
      "loss": 0.2594,
      "mean_token_accuracy": 0.9379322212189436,
      "num_tokens": 16069197.0,
      "step": 2150
    },
    {
      "entropy": 1.7769900791347026,
      "epoch": 0.14426568931114803,
      "grad_norm": 0.3126566708087921,
      "learning_rate": 4.970601323708752e-05,
      "loss": 0.2408,
      "mean_token_accuracy": 0.9445823833346367,
      "num_tokens": 16146214.0,
      "step": 2160
    },
    {
      "entropy": 1.738958951085806,
      "epoch": 0.14493358602092185,
      "grad_norm": 0.30899330973625183,
      "learning_rate": 4.969704157092848e-05,
      "loss": 0.2358,
      "mean_token_accuracy": 0.9447852328419686,
      "num_tokens": 16221287.0,
      "step": 2170
    },
    {
      "entropy": 1.746282958239317,
      "epoch": 0.1456014827306957,
      "grad_norm": 0.32895487546920776,
      "learning_rate": 4.968793589522628e-05,
      "loss": 0.2484,
      "mean_token_accuracy": 0.9414300579577685,
      "num_tokens": 16295465.0,
      "step": 2180
    },
    {
      "entropy": 1.7455839931964874,
      "epoch": 0.14626937944046953,
      "grad_norm": 0.3615387976169586,
      "learning_rate": 4.967869625948983e-05,
      "loss": 0.2491,
      "mean_token_accuracy": 0.9379068542271852,
      "num_tokens": 16368543.0,
      "step": 2190
    },
    {
      "entropy": 1.7793588049709796,
      "epoch": 0.14693727615024335,
      "grad_norm": 0.33342647552490234,
      "learning_rate": 4.966932271395635e-05,
      "loss": 0.234,
      "mean_token_accuracy": 0.94397107437253,
      "num_tokens": 16442099.0,
      "step": 2200
    },
    {
      "entropy": 1.7409492023289204,
      "epoch": 0.1476051728600172,
      "grad_norm": 0.3483268916606903,
      "learning_rate": 4.965981530959116e-05,
      "loss": 0.242,
      "mean_token_accuracy": 0.942544998973608,
      "num_tokens": 16516266.0,
      "step": 2210
    },
    {
      "entropy": 1.7093798875808717,
      "epoch": 0.14827306956979103,
      "grad_norm": 0.3046635389328003,
      "learning_rate": 4.9650174098087435e-05,
      "loss": 0.25,
      "mean_token_accuracy": 0.9399724617600441,
      "num_tokens": 16591269.0,
      "step": 2220
    },
    {
      "entropy": 1.7450269810855388,
      "epoch": 0.14894096627956485,
      "grad_norm": 0.2996240556240082,
      "learning_rate": 4.96403991318658e-05,
      "loss": 0.2525,
      "mean_token_accuracy": 0.9402870047837496,
      "num_tokens": 16664341.0,
      "step": 2230
    },
    {
      "entropy": 1.7802352659404277,
      "epoch": 0.1496088629893387,
      "grad_norm": 0.3075702488422394,
      "learning_rate": 4.963049046407419e-05,
      "loss": 0.2301,
      "mean_token_accuracy": 0.9439422283321619,
      "num_tokens": 16739026.0,
      "step": 2240
    },
    {
      "entropy": 1.8050305910408497,
      "epoch": 0.15027675969911253,
      "grad_norm": 0.27009710669517517,
      "learning_rate": 4.962044814858746e-05,
      "loss": 0.2537,
      "mean_token_accuracy": 0.9410571195185184,
      "num_tokens": 16813065.0,
      "step": 2250
    },
    {
      "entropy": 1.843242458254099,
      "epoch": 0.15094465640888638,
      "grad_norm": 0.34481319785118103,
      "learning_rate": 4.961027224000715e-05,
      "loss": 0.2425,
      "mean_token_accuracy": 0.9395177617669106,
      "num_tokens": 16885812.0,
      "step": 2260
    },
    {
      "entropy": 1.7868775509297847,
      "epoch": 0.1516125531186602,
      "grad_norm": 0.37550628185272217,
      "learning_rate": 4.959996279366114e-05,
      "loss": 0.2468,
      "mean_token_accuracy": 0.9417683627456427,
      "num_tokens": 16958783.0,
      "step": 2270
    },
    {
      "entropy": 1.799260675907135,
      "epoch": 0.15228044982843403,
      "grad_norm": 0.31568869948387146,
      "learning_rate": 4.958951986560338e-05,
      "loss": 0.2358,
      "mean_token_accuracy": 0.9450964849442244,
      "num_tokens": 17032435.0,
      "step": 2280
    },
    {
      "entropy": 1.7812663353979588,
      "epoch": 0.15294834653820788,
      "grad_norm": 0.3289982080459595,
      "learning_rate": 4.95789435126136e-05,
      "loss": 0.2498,
      "mean_token_accuracy": 0.9409911584109067,
      "num_tokens": 17107302.0,
      "step": 2290
    },
    {
      "entropy": 1.782754474878311,
      "epoch": 0.1536162432479817,
      "grad_norm": 0.364711731672287,
      "learning_rate": 4.9568233792196964e-05,
      "loss": 0.2442,
      "mean_token_accuracy": 0.9397149797528982,
      "num_tokens": 17183282.0,
      "step": 2300
    },
    {
      "entropy": 1.8161667115986346,
      "epoch": 0.15428413995775553,
      "grad_norm": 0.30088546872138977,
      "learning_rate": 4.955739076258377e-05,
      "loss": 0.2367,
      "mean_token_accuracy": 0.9431589763611555,
      "num_tokens": 17257554.0,
      "step": 2310
    },
    {
      "entropy": 1.7942778818309306,
      "epoch": 0.15495203666752938,
      "grad_norm": 0.3264505863189697,
      "learning_rate": 4.9546414482729154e-05,
      "loss": 0.2393,
      "mean_token_accuracy": 0.9417916513979435,
      "num_tokens": 17332071.0,
      "step": 2320
    },
    {
      "entropy": 1.783587858080864,
      "epoch": 0.1556199333773032,
      "grad_norm": 0.25699731707572937,
      "learning_rate": 4.9535305012312735e-05,
      "loss": 0.2223,
      "mean_token_accuracy": 0.9469919823110103,
      "num_tokens": 17407322.0,
      "step": 2330
    },
    {
      "entropy": 1.7524106569588185,
      "epoch": 0.15628783008707703,
      "grad_norm": 0.28227999806404114,
      "learning_rate": 4.952406241173832e-05,
      "loss": 0.2307,
      "mean_token_accuracy": 0.9452204752713442,
      "num_tokens": 17483224.0,
      "step": 2340
    },
    {
      "entropy": 1.8086401492357254,
      "epoch": 0.15695572679685088,
      "grad_norm": 0.3575395345687866,
      "learning_rate": 4.9512686742133554e-05,
      "loss": 0.2276,
      "mean_token_accuracy": 0.9439147464931011,
      "num_tokens": 17557422.0,
      "step": 2350
    },
    {
      "entropy": 1.7260880969464778,
      "epoch": 0.1576236235066247,
      "grad_norm": 0.3322930335998535,
      "learning_rate": 4.950117806534962e-05,
      "loss": 0.2319,
      "mean_token_accuracy": 0.9437771786004305,
      "num_tokens": 17632490.0,
      "step": 2360
    },
    {
      "entropy": 1.7858803570270538,
      "epoch": 0.15829152021639853,
      "grad_norm": 0.31739628314971924,
      "learning_rate": 4.948953644396085e-05,
      "loss": 0.2424,
      "mean_token_accuracy": 0.941823310405016,
      "num_tokens": 17703649.0,
      "step": 2370
    },
    {
      "entropy": 1.7709483467042446,
      "epoch": 0.15895941692617238,
      "grad_norm": 0.3278816044330597,
      "learning_rate": 4.947776194126443e-05,
      "loss": 0.2461,
      "mean_token_accuracy": 0.9422798965126276,
      "num_tokens": 17776440.0,
      "step": 2380
    },
    {
      "entropy": 1.7706992015242577,
      "epoch": 0.1596273136359462,
      "grad_norm": 0.28149178624153137,
      "learning_rate": 4.946585462128004e-05,
      "loss": 0.2488,
      "mean_token_accuracy": 0.9397258721292019,
      "num_tokens": 17853046.0,
      "step": 2390
    },
    {
      "entropy": 1.8093292579054832,
      "epoch": 0.16029521034572003,
      "grad_norm": 0.29292032122612,
      "learning_rate": 4.9453814548749525e-05,
      "loss": 0.2402,
      "mean_token_accuracy": 0.9416569489985704,
      "num_tokens": 17927675.0,
      "step": 2400
    },
    {
      "entropy": 1.8050937592983245,
      "epoch": 0.16096310705549388,
      "grad_norm": 0.3780844211578369,
      "learning_rate": 4.944164178913649e-05,
      "loss": 0.2579,
      "mean_token_accuracy": 0.9373075187206268,
      "num_tokens": 18000388.0,
      "step": 2410
    },
    {
      "entropy": 1.8016890078783034,
      "epoch": 0.1616310037652677,
      "grad_norm": 0.34870943427085876,
      "learning_rate": 4.942933640862599e-05,
      "loss": 0.2276,
      "mean_token_accuracy": 0.9450472913682461,
      "num_tokens": 18074770.0,
      "step": 2420
    },
    {
      "entropy": 1.7888476587831974,
      "epoch": 0.16229890047504153,
      "grad_norm": 0.29706668853759766,
      "learning_rate": 4.941689847412419e-05,
      "loss": 0.2437,
      "mean_token_accuracy": 0.9415043368935585,
      "num_tokens": 18149899.0,
      "step": 2430
    },
    {
      "entropy": 1.8147377699613572,
      "epoch": 0.16296679718481538,
      "grad_norm": 0.28822728991508484,
      "learning_rate": 4.940432805325792e-05,
      "loss": 0.2642,
      "mean_token_accuracy": 0.9362334042787552,
      "num_tokens": 18224872.0,
      "step": 2440
    },
    {
      "entropy": 1.8263470873236656,
      "epoch": 0.1636346938945892,
      "grad_norm": 0.32590338587760925,
      "learning_rate": 4.9391625214374395e-05,
      "loss": 0.2554,
      "mean_token_accuracy": 0.9387320384383202,
      "num_tokens": 18297649.0,
      "step": 2450
    },
    {
      "entropy": 1.800306037068367,
      "epoch": 0.16430259060436303,
      "grad_norm": 0.3951384425163269,
      "learning_rate": 4.93787900265408e-05,
      "loss": 0.2528,
      "mean_token_accuracy": 0.9386140391230583,
      "num_tokens": 18369828.0,
      "step": 2460
    },
    {
      "entropy": 1.804408533871174,
      "epoch": 0.16497048731413688,
      "grad_norm": 0.2890225946903229,
      "learning_rate": 4.93658225595439e-05,
      "loss": 0.2514,
      "mean_token_accuracy": 0.9395979948341846,
      "num_tokens": 18444016.0,
      "step": 2470
    },
    {
      "entropy": 1.791828601807356,
      "epoch": 0.1656383840239107,
      "grad_norm": 0.26778239011764526,
      "learning_rate": 4.935272288388971e-05,
      "loss": 0.2335,
      "mean_token_accuracy": 0.9447437960654497,
      "num_tokens": 18518063.0,
      "step": 2480
    },
    {
      "entropy": 1.7641214445233344,
      "epoch": 0.16630628073368453,
      "grad_norm": 0.3373701870441437,
      "learning_rate": 4.9339491070803063e-05,
      "loss": 0.2314,
      "mean_token_accuracy": 0.9459409095346928,
      "num_tokens": 18592479.0,
      "step": 2490
    },
    {
      "entropy": 1.7729347363114356,
      "epoch": 0.16697417744345838,
      "grad_norm": 0.4155683219432831,
      "learning_rate": 4.932612719222723e-05,
      "loss": 0.2398,
      "mean_token_accuracy": 0.9411568857729435,
      "num_tokens": 18665984.0,
      "step": 2500
    },
    {
      "entropy": 1.803322371840477,
      "epoch": 0.1676420741532322,
      "grad_norm": 0.3024482727050781,
      "learning_rate": 4.931263132082359e-05,
      "loss": 0.2531,
      "mean_token_accuracy": 0.9418744131922722,
      "num_tokens": 18738465.0,
      "step": 2510
    },
    {
      "entropy": 1.7958635151386262,
      "epoch": 0.16830997086300603,
      "grad_norm": 0.3386994004249573,
      "learning_rate": 4.929900352997114e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9474283073097467,
      "num_tokens": 18815718.0,
      "step": 2520
    },
    {
      "entropy": 1.8138790540397167,
      "epoch": 0.16897786757277988,
      "grad_norm": 0.2985503077507019,
      "learning_rate": 4.928524389376615e-05,
      "loss": 0.2541,
      "mean_token_accuracy": 0.9393276982009411,
      "num_tokens": 18889497.0,
      "step": 2530
    },
    {
      "entropy": 1.8433697573840617,
      "epoch": 0.1696457642825537,
      "grad_norm": 0.26578304171562195,
      "learning_rate": 4.9271352487021786e-05,
      "loss": 0.2543,
      "mean_token_accuracy": 0.9370607744902373,
      "num_tokens": 18963905.0,
      "step": 2540
    },
    {
      "entropy": 1.799303326755762,
      "epoch": 0.17031366099232753,
      "grad_norm": 0.45828530192375183,
      "learning_rate": 4.92573293852676e-05,
      "loss": 0.2287,
      "mean_token_accuracy": 0.9441456735134125,
      "num_tokens": 19040441.0,
      "step": 2550
    },
    {
      "entropy": 1.808397851884365,
      "epoch": 0.17098155770210138,
      "grad_norm": 0.3327210247516632,
      "learning_rate": 4.924317466474927e-05,
      "loss": 0.2282,
      "mean_token_accuracy": 0.9457386258989573,
      "num_tokens": 19116101.0,
      "step": 2560
    },
    {
      "entropy": 1.8653571978211403,
      "epoch": 0.1716494544118752,
      "grad_norm": 0.3209848701953888,
      "learning_rate": 4.922888840242805e-05,
      "loss": 0.2542,
      "mean_token_accuracy": 0.9404212288558483,
      "num_tokens": 19190633.0,
      "step": 2570
    },
    {
      "entropy": 1.7841139681637288,
      "epoch": 0.17231735112164903,
      "grad_norm": 0.3265194296836853,
      "learning_rate": 4.921447067598042e-05,
      "loss": 0.2346,
      "mean_token_accuracy": 0.942919509112835,
      "num_tokens": 19265494.0,
      "step": 2580
    },
    {
      "entropy": 1.7643261104822159,
      "epoch": 0.17298524783142288,
      "grad_norm": 0.29099276661872864,
      "learning_rate": 4.9199921563797666e-05,
      "loss": 0.2311,
      "mean_token_accuracy": 0.9446221332997083,
      "num_tokens": 19338724.0,
      "step": 2590
    },
    {
      "entropy": 1.7880529016256332,
      "epoch": 0.1736531445411967,
      "grad_norm": 0.3842791020870209,
      "learning_rate": 4.9185241144985407e-05,
      "loss": 0.2232,
      "mean_token_accuracy": 0.9438499320298434,
      "num_tokens": 19413496.0,
      "step": 2600
    },
    {
      "entropy": 1.7623730182647706,
      "epoch": 0.17432104125097053,
      "grad_norm": 0.2518618404865265,
      "learning_rate": 4.917042949936322e-05,
      "loss": 0.2396,
      "mean_token_accuracy": 0.9409934185445309,
      "num_tokens": 19489554.0,
      "step": 2610
    },
    {
      "entropy": 1.815621118992567,
      "epoch": 0.17498893796074438,
      "grad_norm": 0.27371156215667725,
      "learning_rate": 4.915548670746418e-05,
      "loss": 0.243,
      "mean_token_accuracy": 0.9420837715268136,
      "num_tokens": 19561110.0,
      "step": 2620
    },
    {
      "entropy": 1.799737050384283,
      "epoch": 0.1756568346705182,
      "grad_norm": 0.3550506830215454,
      "learning_rate": 4.914041285053439e-05,
      "loss": 0.2688,
      "mean_token_accuracy": 0.9365224111825228,
      "num_tokens": 19634120.0,
      "step": 2630
    },
    {
      "entropy": 1.8204796224832536,
      "epoch": 0.17632473138029203,
      "grad_norm": 0.31795257329940796,
      "learning_rate": 4.912520801053262e-05,
      "loss": 0.2457,
      "mean_token_accuracy": 0.9405323587357998,
      "num_tokens": 19707007.0,
      "step": 2640
    },
    {
      "entropy": 1.7732565388083459,
      "epoch": 0.17699262809006588,
      "grad_norm": 0.3907810151576996,
      "learning_rate": 4.910987227012978e-05,
      "loss": 0.2114,
      "mean_token_accuracy": 0.9482876140624285,
      "num_tokens": 19781374.0,
      "step": 2650
    },
    {
      "entropy": 1.7709458693861961,
      "epoch": 0.1776605247998397,
      "grad_norm": 0.29372161626815796,
      "learning_rate": 4.909440571270855e-05,
      "loss": 0.2528,
      "mean_token_accuracy": 0.9390585873275995,
      "num_tokens": 19858000.0,
      "step": 2660
    },
    {
      "entropy": 1.7674592830240727,
      "epoch": 0.17832842150961353,
      "grad_norm": 0.3351532518863678,
      "learning_rate": 4.9078808422362805e-05,
      "loss": 0.2163,
      "mean_token_accuracy": 0.9467199422419071,
      "num_tokens": 19933532.0,
      "step": 2670
    },
    {
      "entropy": 1.8004688128829003,
      "epoch": 0.17899631821938738,
      "grad_norm": 0.30614417791366577,
      "learning_rate": 4.906308048389729e-05,
      "loss": 0.244,
      "mean_token_accuracy": 0.9407716311514378,
      "num_tokens": 20007969.0,
      "step": 2680
    },
    {
      "entropy": 1.79113230407238,
      "epoch": 0.1796642149291612,
      "grad_norm": 0.2519867420196533,
      "learning_rate": 4.9047221982827105e-05,
      "loss": 0.2408,
      "mean_token_accuracy": 0.9427142448723316,
      "num_tokens": 20082666.0,
      "step": 2690
    },
    {
      "entropy": 1.82162799090147,
      "epoch": 0.18033211163893503,
      "grad_norm": 0.32323145866394043,
      "learning_rate": 4.9031233005377205e-05,
      "loss": 0.2607,
      "mean_token_accuracy": 0.9416906431317329,
      "num_tokens": 20157796.0,
      "step": 2700
    },
    {
      "entropy": 1.8283264808356763,
      "epoch": 0.18100000834870889,
      "grad_norm": 0.4093096852302551,
      "learning_rate": 4.901511363848199e-05,
      "loss": 0.2477,
      "mean_token_accuracy": 0.9399809576570988,
      "num_tokens": 20231893.0,
      "step": 2710
    },
    {
      "entropy": 1.8407348856329917,
      "epoch": 0.1816679050584827,
      "grad_norm": 0.3175579905509949,
      "learning_rate": 4.89988639697848e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9450520567595959,
      "num_tokens": 20308296.0,
      "step": 2720
    },
    {
      "entropy": 1.8361914508044719,
      "epoch": 0.18233580176825653,
      "grad_norm": 0.343191534280777,
      "learning_rate": 4.8982484087637434e-05,
      "loss": 0.2444,
      "mean_token_accuracy": 0.9414731811732053,
      "num_tokens": 20381860.0,
      "step": 2730
    },
    {
      "entropy": 1.7833547614514829,
      "epoch": 0.18300369847803039,
      "grad_norm": 0.3247445225715637,
      "learning_rate": 4.8965974081099706e-05,
      "loss": 0.2349,
      "mean_token_accuracy": 0.9435523323714733,
      "num_tokens": 20458651.0,
      "step": 2740
    },
    {
      "entropy": 1.760952640324831,
      "epoch": 0.1836715951878042,
      "grad_norm": 0.3023550808429718,
      "learning_rate": 4.894933403993889e-05,
      "loss": 0.2337,
      "mean_token_accuracy": 0.9437456987798214,
      "num_tokens": 20532995.0,
      "step": 2750
    },
    {
      "entropy": 1.8257094100117683,
      "epoch": 0.18433949189757803,
      "grad_norm": 0.3648087680339813,
      "learning_rate": 4.893256405462933e-05,
      "loss": 0.2368,
      "mean_token_accuracy": 0.9426568791270256,
      "num_tokens": 20606639.0,
      "step": 2760
    },
    {
      "entropy": 1.8224825911223888,
      "epoch": 0.18500738860735189,
      "grad_norm": 0.3090004324913025,
      "learning_rate": 4.8915664216351864e-05,
      "loss": 0.2436,
      "mean_token_accuracy": 0.9429239921271801,
      "num_tokens": 20680686.0,
      "step": 2770
    },
    {
      "entropy": 1.8282977767288684,
      "epoch": 0.1856752853171257,
      "grad_norm": 0.4072166681289673,
      "learning_rate": 4.889863461699336e-05,
      "loss": 0.2481,
      "mean_token_accuracy": 0.9407894611358643,
      "num_tokens": 20753386.0,
      "step": 2780
    },
    {
      "entropy": 1.8297925420105456,
      "epoch": 0.18634318202689953,
      "grad_norm": 0.32134750485420227,
      "learning_rate": 4.888147534914623e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9439899358898401,
      "num_tokens": 20829573.0,
      "step": 2790
    },
    {
      "entropy": 1.8206966735422612,
      "epoch": 0.1870110787366734,
      "grad_norm": 0.3000108599662781,
      "learning_rate": 4.88641865061079e-05,
      "loss": 0.2377,
      "mean_token_accuracy": 0.9433749865740537,
      "num_tokens": 20904700.0,
      "step": 2800
    },
    {
      "entropy": 1.8035025045275688,
      "epoch": 0.1876789754464472,
      "grad_norm": 0.2725549042224884,
      "learning_rate": 4.884676818188032e-05,
      "loss": 0.2292,
      "mean_token_accuracy": 0.9448900416493415,
      "num_tokens": 20979907.0,
      "step": 2810
    },
    {
      "entropy": 1.7830322928726674,
      "epoch": 0.18834687215622103,
      "grad_norm": 0.43724963068962097,
      "learning_rate": 4.882922047116945e-05,
      "loss": 0.238,
      "mean_token_accuracy": 0.9456601191312075,
      "num_tokens": 21056711.0,
      "step": 2820
    },
    {
      "entropy": 1.7994263157248498,
      "epoch": 0.1890147688659949,
      "grad_norm": 0.29276999831199646,
      "learning_rate": 4.881154346938474e-05,
      "loss": 0.2368,
      "mean_token_accuracy": 0.9421561386436224,
      "num_tokens": 21132253.0,
      "step": 2830
    },
    {
      "entropy": 1.8200394950807095,
      "epoch": 0.1896826655757687,
      "grad_norm": 0.3760021924972534,
      "learning_rate": 4.879373727263861e-05,
      "loss": 0.2379,
      "mean_token_accuracy": 0.9416802622377872,
      "num_tokens": 21205481.0,
      "step": 2840
    },
    {
      "entropy": 1.82858549952507,
      "epoch": 0.19035056228554253,
      "grad_norm": 0.325962632894516,
      "learning_rate": 4.8775801977745944e-05,
      "loss": 0.2344,
      "mean_token_accuracy": 0.940724040567875,
      "num_tokens": 21279464.0,
      "step": 2850
    },
    {
      "entropy": 1.8095352806150913,
      "epoch": 0.1910184589953164,
      "grad_norm": 0.3366350531578064,
      "learning_rate": 4.875773768222354e-05,
      "loss": 0.2216,
      "mean_token_accuracy": 0.9457078546285629,
      "num_tokens": 21355141.0,
      "step": 2860
    },
    {
      "entropy": 1.7960414096713067,
      "epoch": 0.1916863557050902,
      "grad_norm": 0.32072076201438904,
      "learning_rate": 4.873954448428959e-05,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9486239220947027,
      "num_tokens": 21428320.0,
      "step": 2870
    },
    {
      "entropy": 1.8340895384550095,
      "epoch": 0.19235425241486404,
      "grad_norm": 0.28964364528656006,
      "learning_rate": 4.872122248286316e-05,
      "loss": 0.2369,
      "mean_token_accuracy": 0.943690874427557,
      "num_tokens": 21503892.0,
      "step": 2880
    },
    {
      "entropy": 1.850197684764862,
      "epoch": 0.1930221491246379,
      "grad_norm": 0.3319382965564728,
      "learning_rate": 4.8702771777563646e-05,
      "loss": 0.2495,
      "mean_token_accuracy": 0.9389934357255697,
      "num_tokens": 21579873.0,
      "step": 2890
    },
    {
      "entropy": 1.8499460443854332,
      "epoch": 0.1936900458344117,
      "grad_norm": 0.3138546943664551,
      "learning_rate": 4.868419246871018e-05,
      "loss": 0.2293,
      "mean_token_accuracy": 0.9435185875743628,
      "num_tokens": 21653351.0,
      "step": 2900
    },
    {
      "entropy": 1.8294983595609664,
      "epoch": 0.19435794254418554,
      "grad_norm": 0.28921806812286377,
      "learning_rate": 4.866548465732118e-05,
      "loss": 0.2485,
      "mean_token_accuracy": 0.9417641393840313,
      "num_tokens": 21727321.0,
      "step": 2910
    },
    {
      "entropy": 1.850615631043911,
      "epoch": 0.1950258392539594,
      "grad_norm": 0.3138139843940735,
      "learning_rate": 4.864664844511373e-05,
      "loss": 0.2316,
      "mean_token_accuracy": 0.941710140183568,
      "num_tokens": 21801693.0,
      "step": 2920
    },
    {
      "entropy": 1.8938884772360325,
      "epoch": 0.1956937359637332,
      "grad_norm": 0.3019010126590729,
      "learning_rate": 4.862768393450304e-05,
      "loss": 0.2306,
      "mean_token_accuracy": 0.9423238437622785,
      "num_tokens": 21878352.0,
      "step": 2930
    },
    {
      "entropy": 1.834979911893606,
      "epoch": 0.19636163267350704,
      "grad_norm": 0.2476569563150406,
      "learning_rate": 4.860859122860193e-05,
      "loss": 0.2213,
      "mean_token_accuracy": 0.945951184630394,
      "num_tokens": 21955790.0,
      "step": 2940
    },
    {
      "entropy": 1.8118432722985744,
      "epoch": 0.1970295293832809,
      "grad_norm": 0.3174128234386444,
      "learning_rate": 4.85893704312202e-05,
      "loss": 0.2218,
      "mean_token_accuracy": 0.9457420982420445,
      "num_tokens": 22031599.0,
      "step": 2950
    },
    {
      "entropy": 1.8221365205943585,
      "epoch": 0.1976974260930547,
      "grad_norm": 0.3072435259819031,
      "learning_rate": 4.857002164686411e-05,
      "loss": 0.2449,
      "mean_token_accuracy": 0.9402021203190088,
      "num_tokens": 22105860.0,
      "step": 2960
    },
    {
      "entropy": 1.851933331042528,
      "epoch": 0.19836532280282854,
      "grad_norm": 0.28047919273376465,
      "learning_rate": 4.855054498073581e-05,
      "loss": 0.2308,
      "mean_token_accuracy": 0.9437161486595869,
      "num_tokens": 22178655.0,
      "step": 2970
    },
    {
      "entropy": 1.8283856086432935,
      "epoch": 0.1990332195126024,
      "grad_norm": 0.3457595705986023,
      "learning_rate": 4.853094053873278e-05,
      "loss": 0.2343,
      "mean_token_accuracy": 0.9423086777329445,
      "num_tokens": 22252851.0,
      "step": 2980
    },
    {
      "entropy": 1.8148655988276006,
      "epoch": 0.1997011162223762,
      "grad_norm": 0.27360469102859497,
      "learning_rate": 4.8511208427447194e-05,
      "loss": 0.2407,
      "mean_token_accuracy": 0.9426719162613153,
      "num_tokens": 22327247.0,
      "step": 2990
    },
    {
      "entropy": 1.844046352058649,
      "epoch": 0.20036901293215004,
      "grad_norm": 0.36052948236465454,
      "learning_rate": 4.849134875416542e-05,
      "loss": 0.2425,
      "mean_token_accuracy": 0.9421751528978348,
      "num_tokens": 22402103.0,
      "step": 3000
    },
    {
      "epoch": 0.20036901293215004,
      "eval_entropy": 1.8109371545791626,
      "eval_loss": 0.25382745265960693,
      "eval_mean_token_accuracy": 0.9431343225717544,
      "eval_num_tokens": 22402103.0,
      "eval_runtime": 554.0002,
      "eval_samples_per_second": 36.101,
      "eval_steps_per_second": 9.025,
      "step": 3000
    },
    {
      "entropy": 1.8049564428627491,
      "epoch": 0.2010369096419239,
      "grad_norm": 0.3022690415382385,
      "learning_rate": 4.8471361626867385e-05,
      "loss": 0.2247,
      "mean_token_accuracy": 0.9441684324294328,
      "num_tokens": 22477985.0,
      "step": 3010
    },
    {
      "entropy": 1.8117576189339162,
      "epoch": 0.2017048063516977,
      "grad_norm": 0.3053519129753113,
      "learning_rate": 4.8451247154225994e-05,
      "loss": 0.2536,
      "mean_token_accuracy": 0.9388801921159029,
      "num_tokens": 22553192.0,
      "step": 3020
    },
    {
      "entropy": 1.7961050122976303,
      "epoch": 0.20237270306147154,
      "grad_norm": 0.2798187732696533,
      "learning_rate": 4.843100544560655e-05,
      "loss": 0.2395,
      "mean_token_accuracy": 0.9425423514097929,
      "num_tokens": 22629440.0,
      "step": 3030
    },
    {
      "entropy": 1.8528731279075146,
      "epoch": 0.2030405997712454,
      "grad_norm": 0.32638564705848694,
      "learning_rate": 4.841063661106618e-05,
      "loss": 0.2321,
      "mean_token_accuracy": 0.9433439329266549,
      "num_tokens": 22704720.0,
      "step": 3040
    },
    {
      "entropy": 1.8354654036462308,
      "epoch": 0.2037084964810192,
      "grad_norm": 0.3027002215385437,
      "learning_rate": 4.839014076135318e-05,
      "loss": 0.2388,
      "mean_token_accuracy": 0.9438424319028854,
      "num_tokens": 22778022.0,
      "step": 3050
    },
    {
      "entropy": 1.7913148306310176,
      "epoch": 0.20437639319079304,
      "grad_norm": 0.3384620249271393,
      "learning_rate": 4.8369518007906464e-05,
      "loss": 0.2292,
      "mean_token_accuracy": 0.9437200121581555,
      "num_tokens": 22851343.0,
      "step": 3060
    },
    {
      "entropy": 1.7788014896214008,
      "epoch": 0.2050442899005669,
      "grad_norm": 0.3399059772491455,
      "learning_rate": 4.834876846285494e-05,
      "loss": 0.2378,
      "mean_token_accuracy": 0.9435754213482141,
      "num_tokens": 22925055.0,
      "step": 3070
    },
    {
      "entropy": 1.7674326702952385,
      "epoch": 0.2057121866103407,
      "grad_norm": 0.3759543299674988,
      "learning_rate": 4.832789223901689e-05,
      "loss": 0.2321,
      "mean_token_accuracy": 0.9434260703623295,
      "num_tokens": 23001078.0,
      "step": 3080
    },
    {
      "entropy": 1.823313407599926,
      "epoch": 0.20638008332011454,
      "grad_norm": 0.2871500551700592,
      "learning_rate": 4.830688944989938e-05,
      "loss": 0.2435,
      "mean_token_accuracy": 0.940041083842516,
      "num_tokens": 23077193.0,
      "step": 3090
    },
    {
      "entropy": 1.7994055196642875,
      "epoch": 0.2070479800298884,
      "grad_norm": 0.27684858441352844,
      "learning_rate": 4.828576020969762e-05,
      "loss": 0.224,
      "mean_token_accuracy": 0.9467190150171518,
      "num_tokens": 23152633.0,
      "step": 3100
    },
    {
      "entropy": 1.7518393844366074,
      "epoch": 0.20771587673966221,
      "grad_norm": 0.4524850845336914,
      "learning_rate": 4.826450463329436e-05,
      "loss": 0.2477,
      "mean_token_accuracy": 0.9389440231025219,
      "num_tokens": 23223527.0,
      "step": 3110
    },
    {
      "entropy": 1.7778724975883962,
      "epoch": 0.20838377344943604,
      "grad_norm": 0.37888818979263306,
      "learning_rate": 4.824312283625925e-05,
      "loss": 0.2388,
      "mean_token_accuracy": 0.940614327415824,
      "num_tokens": 23293540.0,
      "step": 3120
    },
    {
      "entropy": 1.7527782574295998,
      "epoch": 0.2090516701592099,
      "grad_norm": 0.29386237263679504,
      "learning_rate": 4.822161493484822e-05,
      "loss": 0.2541,
      "mean_token_accuracy": 0.9378659635782242,
      "num_tokens": 23366494.0,
      "step": 3130
    },
    {
      "entropy": 1.7559617906808853,
      "epoch": 0.20971956686898371,
      "grad_norm": 0.2850927710533142,
      "learning_rate": 4.819998104600286e-05,
      "loss": 0.2407,
      "mean_token_accuracy": 0.9400011789053678,
      "num_tokens": 23440754.0,
      "step": 3140
    },
    {
      "entropy": 1.8102621376514434,
      "epoch": 0.21038746357875754,
      "grad_norm": 0.30684709548950195,
      "learning_rate": 4.8178221287349764e-05,
      "loss": 0.245,
      "mean_token_accuracy": 0.941274481266737,
      "num_tokens": 23513489.0,
      "step": 3150
    },
    {
      "entropy": 1.8046070531010627,
      "epoch": 0.2110553602885314,
      "grad_norm": 0.32182544469833374,
      "learning_rate": 4.8156335777199906e-05,
      "loss": 0.2416,
      "mean_token_accuracy": 0.9421315606683492,
      "num_tokens": 23585829.0,
      "step": 3160
    },
    {
      "entropy": 1.755084903538227,
      "epoch": 0.21172325699830521,
      "grad_norm": 0.3017768859863281,
      "learning_rate": 4.8134324634547966e-05,
      "loss": 0.2286,
      "mean_token_accuracy": 0.9476879339665174,
      "num_tokens": 23657471.0,
      "step": 3170
    },
    {
      "entropy": 1.7735888287425041,
      "epoch": 0.21239115370807904,
      "grad_norm": 0.30247068405151367,
      "learning_rate": 4.811218797907173e-05,
      "loss": 0.226,
      "mean_token_accuracy": 0.9445185609161854,
      "num_tokens": 23731517.0,
      "step": 3180
    },
    {
      "entropy": 1.7969969406723976,
      "epoch": 0.2130590504178529,
      "grad_norm": 0.3549629747867584,
      "learning_rate": 4.808992593113141e-05,
      "loss": 0.2358,
      "mean_token_accuracy": 0.945430938899517,
      "num_tokens": 23806169.0,
      "step": 3190
    },
    {
      "entropy": 1.8177655592560769,
      "epoch": 0.21372694712762672,
      "grad_norm": 0.3350241780281067,
      "learning_rate": 4.8067538611768995e-05,
      "loss": 0.2433,
      "mean_token_accuracy": 0.943061376735568,
      "num_tokens": 23880228.0,
      "step": 3200
    },
    {
      "entropy": 1.8541865468025207,
      "epoch": 0.21439484383740054,
      "grad_norm": 0.3743235468864441,
      "learning_rate": 4.8045026142707585e-05,
      "loss": 0.2614,
      "mean_token_accuracy": 0.935718546807766,
      "num_tokens": 23952692.0,
      "step": 3210
    },
    {
      "entropy": 1.830523307621479,
      "epoch": 0.2150627405471744,
      "grad_norm": 0.291801393032074,
      "learning_rate": 4.802238864635073e-05,
      "loss": 0.2484,
      "mean_token_accuracy": 0.9374997280538082,
      "num_tokens": 24028103.0,
      "step": 3220
    },
    {
      "entropy": 1.7853223897516728,
      "epoch": 0.21573063725694822,
      "grad_norm": 0.29808318614959717,
      "learning_rate": 4.799962624578181e-05,
      "loss": 0.2304,
      "mean_token_accuracy": 0.9421184424310922,
      "num_tokens": 24103005.0,
      "step": 3230
    },
    {
      "entropy": 1.8096862316131592,
      "epoch": 0.21639853396672204,
      "grad_norm": 0.3377639055252075,
      "learning_rate": 4.797673906476328e-05,
      "loss": 0.2327,
      "mean_token_accuracy": 0.9450104378163815,
      "num_tokens": 24174517.0,
      "step": 3240
    },
    {
      "entropy": 1.7864435084164143,
      "epoch": 0.2170664306764959,
      "grad_norm": 0.35219600796699524,
      "learning_rate": 4.795372722773608e-05,
      "loss": 0.2223,
      "mean_token_accuracy": 0.9452082689851522,
      "num_tokens": 24249275.0,
      "step": 3250
    },
    {
      "entropy": 1.842170026898384,
      "epoch": 0.21773432738626972,
      "grad_norm": 0.2833854556083679,
      "learning_rate": 4.7930590859818893e-05,
      "loss": 0.245,
      "mean_token_accuracy": 0.9390489414334298,
      "num_tokens": 24324467.0,
      "step": 3260
    },
    {
      "entropy": 1.809202066808939,
      "epoch": 0.21840222409604354,
      "grad_norm": 0.3260651230812073,
      "learning_rate": 4.790733008680753e-05,
      "loss": 0.2351,
      "mean_token_accuracy": 0.9439034074544906,
      "num_tokens": 24400522.0,
      "step": 3270
    },
    {
      "entropy": 1.8304984860122204,
      "epoch": 0.2190701208058174,
      "grad_norm": 0.2720033824443817,
      "learning_rate": 4.7883945035174175e-05,
      "loss": 0.2296,
      "mean_token_accuracy": 0.9467284128069877,
      "num_tokens": 24476561.0,
      "step": 3280
    },
    {
      "entropy": 1.824673803895712,
      "epoch": 0.21973801751559122,
      "grad_norm": 0.32116538286209106,
      "learning_rate": 4.7860435832066765e-05,
      "loss": 0.2496,
      "mean_token_accuracy": 0.9417725473642349,
      "num_tokens": 24549445.0,
      "step": 3290
    },
    {
      "entropy": 1.8653734281659127,
      "epoch": 0.22040591422536504,
      "grad_norm": 0.25364255905151367,
      "learning_rate": 4.7836802605308224e-05,
      "loss": 0.2247,
      "mean_token_accuracy": 0.9449461746960879,
      "num_tokens": 24624106.0,
      "step": 3300
    },
    {
      "entropy": 1.8543301723897456,
      "epoch": 0.2210738109351389,
      "grad_norm": 0.3310909569263458,
      "learning_rate": 4.781304548339586e-05,
      "loss": 0.2431,
      "mean_token_accuracy": 0.9421414006501436,
      "num_tokens": 24698207.0,
      "step": 3310
    },
    {
      "entropy": 1.8697098560631276,
      "epoch": 0.22174170764491272,
      "grad_norm": 0.30037450790405273,
      "learning_rate": 4.778916459550059e-05,
      "loss": 0.2311,
      "mean_token_accuracy": 0.9462303768843412,
      "num_tokens": 24771943.0,
      "step": 3320
    },
    {
      "entropy": 1.816944743692875,
      "epoch": 0.22240960435468654,
      "grad_norm": 0.2789921164512634,
      "learning_rate": 4.7765160071466255e-05,
      "loss": 0.2325,
      "mean_token_accuracy": 0.9444701805710792,
      "num_tokens": 24846612.0,
      "step": 3330
    },
    {
      "entropy": 1.8986338444054127,
      "epoch": 0.2230775010644604,
      "grad_norm": 0.2924051582813263,
      "learning_rate": 4.774103204180896e-05,
      "loss": 0.2459,
      "mean_token_accuracy": 0.9418189588934183,
      "num_tokens": 24917314.0,
      "step": 3340
    },
    {
      "entropy": 1.8245647922158241,
      "epoch": 0.22374539777423422,
      "grad_norm": 0.28584960103034973,
      "learning_rate": 4.771678063771628e-05,
      "loss": 0.2307,
      "mean_token_accuracy": 0.9448101501911879,
      "num_tokens": 24994662.0,
      "step": 3350
    },
    {
      "entropy": 1.8702694348990918,
      "epoch": 0.22441329448400804,
      "grad_norm": 0.3350934684276581,
      "learning_rate": 4.7692405991046646e-05,
      "loss": 0.2314,
      "mean_token_accuracy": 0.9442384645342827,
      "num_tokens": 25068785.0,
      "step": 3360
    },
    {
      "entropy": 1.8374993085861206,
      "epoch": 0.2250811911937819,
      "grad_norm": 0.24896343052387238,
      "learning_rate": 4.766790823432852e-05,
      "loss": 0.229,
      "mean_token_accuracy": 0.9465972561389208,
      "num_tokens": 25145338.0,
      "step": 3370
    },
    {
      "entropy": 1.867978721857071,
      "epoch": 0.22574908790355572,
      "grad_norm": 0.34039804339408875,
      "learning_rate": 4.764328750075978e-05,
      "loss": 0.2367,
      "mean_token_accuracy": 0.944473821669817,
      "num_tokens": 25220859.0,
      "step": 3380
    },
    {
      "entropy": 1.8610978439450263,
      "epoch": 0.22641698461332954,
      "grad_norm": 0.3372277319431305,
      "learning_rate": 4.7618543924206935e-05,
      "loss": 0.2565,
      "mean_token_accuracy": 0.9431333035230637,
      "num_tokens": 25294781.0,
      "step": 3390
    },
    {
      "entropy": 1.8835390880703926,
      "epoch": 0.2270848813231034,
      "grad_norm": 0.3769782483577728,
      "learning_rate": 4.759367763920439e-05,
      "loss": 0.2497,
      "mean_token_accuracy": 0.9379332199692726,
      "num_tokens": 25369069.0,
      "step": 3400
    },
    {
      "entropy": 1.8768240429461003,
      "epoch": 0.22775277803287722,
      "grad_norm": 0.32036513090133667,
      "learning_rate": 4.756868878095374e-05,
      "loss": 0.2289,
      "mean_token_accuracy": 0.9450360585004092,
      "num_tokens": 25444795.0,
      "step": 3410
    },
    {
      "entropy": 1.9017126262187958,
      "epoch": 0.22842067474265104,
      "grad_norm": 0.28228071331977844,
      "learning_rate": 4.7543577485323065e-05,
      "loss": 0.2594,
      "mean_token_accuracy": 0.9393475946038962,
      "num_tokens": 25519254.0,
      "step": 3420
    },
    {
      "entropy": 1.8318243347108365,
      "epoch": 0.2290885714524249,
      "grad_norm": 0.28551679849624634,
      "learning_rate": 4.75183438888461e-05,
      "loss": 0.2339,
      "mean_token_accuracy": 0.9436967823654413,
      "num_tokens": 25593483.0,
      "step": 3430
    },
    {
      "entropy": 1.8500353939831258,
      "epoch": 0.22975646816219872,
      "grad_norm": 0.26876312494277954,
      "learning_rate": 4.7492988128721565e-05,
      "loss": 0.2368,
      "mean_token_accuracy": 0.9435034118592739,
      "num_tokens": 25669434.0,
      "step": 3440
    },
    {
      "entropy": 1.830290822684765,
      "epoch": 0.23042436487197254,
      "grad_norm": 0.2895486354827881,
      "learning_rate": 4.7467510342812435e-05,
      "loss": 0.2247,
      "mean_token_accuracy": 0.945241654664278,
      "num_tokens": 25744924.0,
      "step": 3450
    },
    {
      "entropy": 1.8303662419319153,
      "epoch": 0.2310922615817464,
      "grad_norm": 0.2706555724143982,
      "learning_rate": 4.744191066964511e-05,
      "loss": 0.2257,
      "mean_token_accuracy": 0.9461141012609005,
      "num_tokens": 25821606.0,
      "step": 3460
    },
    {
      "entropy": 1.8629770003259183,
      "epoch": 0.23176015829152022,
      "grad_norm": 0.2680894136428833,
      "learning_rate": 4.741618924840874e-05,
      "loss": 0.2191,
      "mean_token_accuracy": 0.9460224557667971,
      "num_tokens": 25898667.0,
      "step": 3470
    },
    {
      "entropy": 1.85146396830678,
      "epoch": 0.23242805500129404,
      "grad_norm": 0.34949150681495667,
      "learning_rate": 4.739034621895443e-05,
      "loss": 0.2533,
      "mean_token_accuracy": 0.9408930767327547,
      "num_tokens": 25974131.0,
      "step": 3480
    },
    {
      "entropy": 1.8315501116216182,
      "epoch": 0.2330959517110679,
      "grad_norm": 0.3140328824520111,
      "learning_rate": 4.736438172179446e-05,
      "loss": 0.2148,
      "mean_token_accuracy": 0.947515569999814,
      "num_tokens": 26050557.0,
      "step": 3490
    },
    {
      "entropy": 1.8211340382695198,
      "epoch": 0.23376384842084172,
      "grad_norm": 0.32769250869750977,
      "learning_rate": 4.73382958981016e-05,
      "loss": 0.234,
      "mean_token_accuracy": 0.942050764709711,
      "num_tokens": 26123638.0,
      "step": 3500
    },
    {
      "entropy": 1.8524242155253887,
      "epoch": 0.23443174513061554,
      "grad_norm": 0.2564964294433594,
      "learning_rate": 4.731208888970825e-05,
      "loss": 0.2335,
      "mean_token_accuracy": 0.9442535698413849,
      "num_tokens": 26197843.0,
      "step": 3510
    },
    {
      "entropy": 1.88613108843565,
      "epoch": 0.2350996418403894,
      "grad_norm": 0.24955983459949493,
      "learning_rate": 4.7285760839105706e-05,
      "loss": 0.2294,
      "mean_token_accuracy": 0.941499026119709,
      "num_tokens": 26271783.0,
      "step": 3520
    },
    {
      "entropy": 1.8038532249629498,
      "epoch": 0.23576753855016322,
      "grad_norm": 0.30114203691482544,
      "learning_rate": 4.725931188944342e-05,
      "loss": 0.2248,
      "mean_token_accuracy": 0.9469171680510045,
      "num_tokens": 26347502.0,
      "step": 3530
    },
    {
      "entropy": 1.8166924223303795,
      "epoch": 0.23643543525993704,
      "grad_norm": 0.2860889732837677,
      "learning_rate": 4.7232742184528155e-05,
      "loss": 0.2489,
      "mean_token_accuracy": 0.941021029278636,
      "num_tokens": 26422797.0,
      "step": 3540
    },
    {
      "entropy": 1.8762867979705333,
      "epoch": 0.2371033319697109,
      "grad_norm": 0.29442188143730164,
      "learning_rate": 4.7206051868823264e-05,
      "loss": 0.2365,
      "mean_token_accuracy": 0.942730788514018,
      "num_tokens": 26496425.0,
      "step": 3550
    },
    {
      "entropy": 1.8776430949568748,
      "epoch": 0.23777122867948472,
      "grad_norm": 0.26039785146713257,
      "learning_rate": 4.717924108744787e-05,
      "loss": 0.2224,
      "mean_token_accuracy": 0.9473829843103886,
      "num_tokens": 26573687.0,
      "step": 3560
    },
    {
      "entropy": 1.8724658526480198,
      "epoch": 0.23843912538925854,
      "grad_norm": 0.30388835072517395,
      "learning_rate": 4.7152309986176077e-05,
      "loss": 0.2328,
      "mean_token_accuracy": 0.9449579548090696,
      "num_tokens": 26649187.0,
      "step": 3570
    },
    {
      "entropy": 1.8556077405810356,
      "epoch": 0.2391070220990324,
      "grad_norm": 0.28188958764076233,
      "learning_rate": 4.712525871143619e-05,
      "loss": 0.2249,
      "mean_token_accuracy": 0.944434067979455,
      "num_tokens": 26724611.0,
      "step": 3580
    },
    {
      "entropy": 1.890062303096056,
      "epoch": 0.23977491880880622,
      "grad_norm": 0.34954994916915894,
      "learning_rate": 4.709808741030992e-05,
      "loss": 0.2336,
      "mean_token_accuracy": 0.9438429970294238,
      "num_tokens": 26798982.0,
      "step": 3590
    },
    {
      "entropy": 1.877310136705637,
      "epoch": 0.24044281551858004,
      "grad_norm": 0.27162179350852966,
      "learning_rate": 4.707079623053159e-05,
      "loss": 0.2297,
      "mean_token_accuracy": 0.9456120621412992,
      "num_tokens": 26873337.0,
      "step": 3600
    },
    {
      "entropy": 1.8258368089795112,
      "epoch": 0.2411107122283539,
      "grad_norm": 0.2752111256122589,
      "learning_rate": 4.704338532048728e-05,
      "loss": 0.2201,
      "mean_token_accuracy": 0.9444967415183783,
      "num_tokens": 26947217.0,
      "step": 3610
    },
    {
      "entropy": 1.8778181873261928,
      "epoch": 0.24177860893812772,
      "grad_norm": 0.30664199590682983,
      "learning_rate": 4.7015854829214104e-05,
      "loss": 0.2286,
      "mean_token_accuracy": 0.943577378988266,
      "num_tokens": 27019170.0,
      "step": 3620
    },
    {
      "entropy": 1.8909952104091645,
      "epoch": 0.24244650564790154,
      "grad_norm": 0.298827588558197,
      "learning_rate": 4.6988204906399355e-05,
      "loss": 0.2147,
      "mean_token_accuracy": 0.9472693510353565,
      "num_tokens": 27093645.0,
      "step": 3630
    },
    {
      "entropy": 1.909240224212408,
      "epoch": 0.2431144023576754,
      "grad_norm": 0.3560390770435333,
      "learning_rate": 4.696043570237964e-05,
      "loss": 0.2495,
      "mean_token_accuracy": 0.9418341901153326,
      "num_tokens": 27168714.0,
      "step": 3640
    },
    {
      "entropy": 1.8704890042543412,
      "epoch": 0.24378229906744922,
      "grad_norm": 0.316654235124588,
      "learning_rate": 4.693254736814017e-05,
      "loss": 0.2495,
      "mean_token_accuracy": 0.9396880820393563,
      "num_tokens": 27242321.0,
      "step": 3650
    },
    {
      "entropy": 1.8840535148978232,
      "epoch": 0.24445019577722304,
      "grad_norm": 0.29791778326034546,
      "learning_rate": 4.690454005531388e-05,
      "loss": 0.2328,
      "mean_token_accuracy": 0.9429134096950292,
      "num_tokens": 27316492.0,
      "step": 3660
    },
    {
      "entropy": 1.8875044971704482,
      "epoch": 0.2451180924869969,
      "grad_norm": 0.3188786506652832,
      "learning_rate": 4.6876413916180576e-05,
      "loss": 0.2233,
      "mean_token_accuracy": 0.9476851027458906,
      "num_tokens": 27390030.0,
      "step": 3670
    },
    {
      "entropy": 1.8899781942367553,
      "epoch": 0.24578598919677072,
      "grad_norm": 0.267020583152771,
      "learning_rate": 4.684816910366618e-05,
      "loss": 0.2395,
      "mean_token_accuracy": 0.9413978833705187,
      "num_tokens": 27466229.0,
      "step": 3680
    },
    {
      "entropy": 1.879320028424263,
      "epoch": 0.24645388590654455,
      "grad_norm": 0.2677568793296814,
      "learning_rate": 4.6819805771341814e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.9485225848853588,
      "num_tokens": 27543372.0,
      "step": 3690
    },
    {
      "entropy": 1.8553839147090911,
      "epoch": 0.2471217826163184,
      "grad_norm": 0.28572285175323486,
      "learning_rate": 4.679132407342305e-05,
      "loss": 0.2354,
      "mean_token_accuracy": 0.9409180752933025,
      "num_tokens": 27616468.0,
      "step": 3700
    },
    {
      "entropy": 1.8926606640219688,
      "epoch": 0.24778967932609222,
      "grad_norm": 0.3074871003627777,
      "learning_rate": 4.6762724164769014e-05,
      "loss": 0.2404,
      "mean_token_accuracy": 0.9434654451906681,
      "num_tokens": 27691489.0,
      "step": 3710
    },
    {
      "entropy": 1.8878293678164482,
      "epoch": 0.24845757603586605,
      "grad_norm": 0.26660841703414917,
      "learning_rate": 4.673400620088155e-05,
      "loss": 0.2258,
      "mean_token_accuracy": 0.9468064650893211,
      "num_tokens": 27767282.0,
      "step": 3720
    },
    {
      "entropy": 1.9093123912811278,
      "epoch": 0.2491254727456399,
      "grad_norm": 0.242268905043602,
      "learning_rate": 4.670517033790441e-05,
      "loss": 0.2229,
      "mean_token_accuracy": 0.944462651014328,
      "num_tokens": 27842357.0,
      "step": 3730
    },
    {
      "entropy": 1.8764431931078434,
      "epoch": 0.24979336945541372,
      "grad_norm": 0.30574408173561096,
      "learning_rate": 4.6676216732622367e-05,
      "loss": 0.2333,
      "mean_token_accuracy": 0.9444595023989677,
      "num_tokens": 27918261.0,
      "step": 3740
    },
    {
      "entropy": 1.8780780017375946,
      "epoch": 0.25046126616518755,
      "grad_norm": 0.28927671909332275,
      "learning_rate": 4.6647145542460356e-05,
      "loss": 0.2226,
      "mean_token_accuracy": 0.9472949348390103,
      "num_tokens": 27996751.0,
      "step": 3750
    },
    {
      "entropy": 1.8979285068809986,
      "epoch": 0.2511291628749614,
      "grad_norm": 0.3261164426803589,
      "learning_rate": 4.6617956925482685e-05,
      "loss": 0.2536,
      "mean_token_accuracy": 0.9386015087366104,
      "num_tokens": 28069416.0,
      "step": 3760
    },
    {
      "entropy": 1.9287927523255348,
      "epoch": 0.2517970595847352,
      "grad_norm": 0.2858129143714905,
      "learning_rate": 4.658865104039208e-05,
      "loss": 0.2466,
      "mean_token_accuracy": 0.942908226698637,
      "num_tokens": 28145828.0,
      "step": 3770
    },
    {
      "entropy": 1.8883475452661513,
      "epoch": 0.25246495629450905,
      "grad_norm": 0.30416712164878845,
      "learning_rate": 4.655922804652891e-05,
      "loss": 0.2286,
      "mean_token_accuracy": 0.9443699412047863,
      "num_tokens": 28221045.0,
      "step": 3780
    },
    {
      "entropy": 1.9357228808104991,
      "epoch": 0.2531328530042829,
      "grad_norm": 0.31053051352500916,
      "learning_rate": 4.6529688103870264e-05,
      "loss": 0.2448,
      "mean_token_accuracy": 0.9393238201737404,
      "num_tokens": 28294449.0,
      "step": 3790
    },
    {
      "entropy": 1.8701168105006218,
      "epoch": 0.2538007497140567,
      "grad_norm": 0.2875557839870453,
      "learning_rate": 4.650003137302908e-05,
      "loss": 0.2263,
      "mean_token_accuracy": 0.9480938505381346,
      "num_tokens": 28370048.0,
      "step": 3800
    },
    {
      "entropy": 1.9043088875710965,
      "epoch": 0.25446864642383055,
      "grad_norm": 0.38767239451408386,
      "learning_rate": 4.647025801525333e-05,
      "loss": 0.2409,
      "mean_token_accuracy": 0.9406221088021993,
      "num_tokens": 28445602.0,
      "step": 3810
    },
    {
      "entropy": 1.9278858952224254,
      "epoch": 0.2551365431336044,
      "grad_norm": 0.3644951581954956,
      "learning_rate": 4.644036819242506e-05,
      "loss": 0.2379,
      "mean_token_accuracy": 0.9439516618847847,
      "num_tokens": 28519403.0,
      "step": 3820
    },
    {
      "entropy": 1.9014289505779742,
      "epoch": 0.2558044398433782,
      "grad_norm": 0.30445632338523865,
      "learning_rate": 4.641036206705961e-05,
      "loss": 0.2347,
      "mean_token_accuracy": 0.943390891700983,
      "num_tokens": 28595357.0,
      "step": 3830
    },
    {
      "entropy": 1.8606593564152718,
      "epoch": 0.25647233655315205,
      "grad_norm": 0.3390524387359619,
      "learning_rate": 4.63802398023046e-05,
      "loss": 0.2349,
      "mean_token_accuracy": 0.9411648731678725,
      "num_tokens": 28672079.0,
      "step": 3840
    },
    {
      "entropy": 1.8884294882416726,
      "epoch": 0.2571402332629259,
      "grad_norm": 0.4061954617500305,
      "learning_rate": 4.635000156193918e-05,
      "loss": 0.2414,
      "mean_token_accuracy": 0.9413364317268134,
      "num_tokens": 28744711.0,
      "step": 3850
    },
    {
      "entropy": 1.874824495613575,
      "epoch": 0.2578081299726997,
      "grad_norm": 0.2919199764728546,
      "learning_rate": 4.6319647510373046e-05,
      "loss": 0.2385,
      "mean_token_accuracy": 0.9417082604020834,
      "num_tokens": 28819375.0,
      "step": 3860
    },
    {
      "entropy": 1.8597666814923286,
      "epoch": 0.25847602668247355,
      "grad_norm": 0.5549324154853821,
      "learning_rate": 4.6289177812645586e-05,
      "loss": 0.2366,
      "mean_token_accuracy": 0.9450602639466524,
      "num_tokens": 28891975.0,
      "step": 3870
    },
    {
      "entropy": 1.8977480821311474,
      "epoch": 0.2591439233922474,
      "grad_norm": 0.4319974482059479,
      "learning_rate": 4.6258592634424976e-05,
      "loss": 0.2528,
      "mean_token_accuracy": 0.9396120455116034,
      "num_tokens": 28964117.0,
      "step": 3880
    },
    {
      "entropy": 1.8904687605798245,
      "epoch": 0.2598118201020212,
      "grad_norm": 0.25685828924179077,
      "learning_rate": 4.622789214200726e-05,
      "loss": 0.2409,
      "mean_token_accuracy": 0.9404903642833233,
      "num_tokens": 29039759.0,
      "step": 3890
    },
    {
      "entropy": 1.8698497407138348,
      "epoch": 0.26047971681179505,
      "grad_norm": 0.34948089718818665,
      "learning_rate": 4.6197076502315474e-05,
      "loss": 0.2402,
      "mean_token_accuracy": 0.9440975349396468,
      "num_tokens": 29113336.0,
      "step": 3900
    },
    {
      "entropy": 1.915880484879017,
      "epoch": 0.2611476135215689,
      "grad_norm": 0.34970033168792725,
      "learning_rate": 4.616614588289872e-05,
      "loss": 0.2535,
      "mean_token_accuracy": 0.9414620477706194,
      "num_tokens": 29186923.0,
      "step": 3910
    },
    {
      "entropy": 1.8937527641654015,
      "epoch": 0.2618155102313427,
      "grad_norm": 0.3331890404224396,
      "learning_rate": 4.6135100451931276e-05,
      "loss": 0.2425,
      "mean_token_accuracy": 0.9413011107593775,
      "num_tokens": 29260423.0,
      "step": 3920
    },
    {
      "entropy": 1.9137082055211068,
      "epoch": 0.26248340694111655,
      "grad_norm": 0.381989449262619,
      "learning_rate": 4.6103940378211656e-05,
      "loss": 0.2396,
      "mean_token_accuracy": 0.9424898091703653,
      "num_tokens": 29335203.0,
      "step": 3930
    },
    {
      "entropy": 1.864582545310259,
      "epoch": 0.2631513036508904,
      "grad_norm": 0.29004234075546265,
      "learning_rate": 4.607266583116171e-05,
      "loss": 0.2261,
      "mean_token_accuracy": 0.9446154166013002,
      "num_tokens": 29408831.0,
      "step": 3940
    },
    {
      "entropy": 1.8831079982221126,
      "epoch": 0.2638192003606642,
      "grad_norm": 0.3435479998588562,
      "learning_rate": 4.604127698082567e-05,
      "loss": 0.2538,
      "mean_token_accuracy": 0.940527680888772,
      "num_tokens": 29481179.0,
      "step": 3950
    },
    {
      "entropy": 1.8447520434856415,
      "epoch": 0.26448709707043805,
      "grad_norm": 0.2671530544757843,
      "learning_rate": 4.600977399786929e-05,
      "loss": 0.2331,
      "mean_token_accuracy": 0.9435788184404373,
      "num_tokens": 29556624.0,
      "step": 3960
    },
    {
      "entropy": 1.8865703962743283,
      "epoch": 0.2651549937802119,
      "grad_norm": 0.29622238874435425,
      "learning_rate": 4.5978157053578854e-05,
      "loss": 0.2623,
      "mean_token_accuracy": 0.9372790973633528,
      "num_tokens": 29629336.0,
      "step": 3970
    },
    {
      "entropy": 1.839417278766632,
      "epoch": 0.2658228904899857,
      "grad_norm": 0.2559685707092285,
      "learning_rate": 4.5946426319860284e-05,
      "loss": 0.2318,
      "mean_token_accuracy": 0.9451282382011413,
      "num_tokens": 29705947.0,
      "step": 3980
    },
    {
      "entropy": 1.8560559064149857,
      "epoch": 0.26649078719975955,
      "grad_norm": 0.38469651341438293,
      "learning_rate": 4.591458196923818e-05,
      "loss": 0.2322,
      "mean_token_accuracy": 0.9416191060096025,
      "num_tokens": 29779966.0,
      "step": 3990
    },
    {
      "entropy": 1.9010761491954327,
      "epoch": 0.2671586839095334,
      "grad_norm": 0.38259655237197876,
      "learning_rate": 4.5882624174854894e-05,
      "loss": 0.2528,
      "mean_token_accuracy": 0.9391449630260468,
      "num_tokens": 29853036.0,
      "step": 4000
    },
    {
      "epoch": 0.2671586839095334,
      "eval_entropy": 1.901148555612564,
      "eval_loss": 0.2492472380399704,
      "eval_mean_token_accuracy": 0.9440275570511818,
      "eval_num_tokens": 29853036.0,
      "eval_runtime": 553.7327,
      "eval_samples_per_second": 36.119,
      "eval_steps_per_second": 9.03,
      "step": 4000
    },
    {
      "entropy": 1.8866844855248928,
      "epoch": 0.2678265806193072,
      "grad_norm": 0.26135289669036865,
      "learning_rate": 4.5850553110469606e-05,
      "loss": 0.2216,
      "mean_token_accuracy": 0.9460945274680853,
      "num_tokens": 29928613.0,
      "step": 4010
    },
    {
      "entropy": 1.8946465224027633,
      "epoch": 0.26849447732908105,
      "grad_norm": 0.2813650667667389,
      "learning_rate": 4.5818368950457345e-05,
      "loss": 0.2403,
      "mean_token_accuracy": 0.9423510797321797,
      "num_tokens": 30005051.0,
      "step": 4020
    },
    {
      "entropy": 1.8805329866707325,
      "epoch": 0.2691623740388549,
      "grad_norm": 0.32251104712486267,
      "learning_rate": 4.578607186980807e-05,
      "loss": 0.2138,
      "mean_token_accuracy": 0.9491134345531463,
      "num_tokens": 30079427.0,
      "step": 4030
    },
    {
      "entropy": 1.870673867315054,
      "epoch": 0.2698302707486287,
      "grad_norm": 0.32249370217323303,
      "learning_rate": 4.575366204412567e-05,
      "loss": 0.2129,
      "mean_token_accuracy": 0.9503507029265166,
      "num_tokens": 30153707.0,
      "step": 4040
    },
    {
      "entropy": 1.8925836361944675,
      "epoch": 0.27049816745840255,
      "grad_norm": 0.3650279939174652,
      "learning_rate": 4.5721139649627106e-05,
      "loss": 0.2309,
      "mean_token_accuracy": 0.9465401224792004,
      "num_tokens": 30227660.0,
      "step": 4050
    },
    {
      "entropy": 1.9027898274362087,
      "epoch": 0.2711660641681764,
      "grad_norm": 0.3059157133102417,
      "learning_rate": 4.568850486314133e-05,
      "loss": 0.2364,
      "mean_token_accuracy": 0.9449421457946301,
      "num_tokens": 30301084.0,
      "step": 4060
    },
    {
      "entropy": 1.9198800154030322,
      "epoch": 0.2718339608779502,
      "grad_norm": 0.3443954885005951,
      "learning_rate": 4.5655757862108426e-05,
      "loss": 0.2385,
      "mean_token_accuracy": 0.9415464106947183,
      "num_tokens": 30373637.0,
      "step": 4070
    },
    {
      "entropy": 1.8452917598187923,
      "epoch": 0.27250185758772405,
      "grad_norm": 0.3252948820590973,
      "learning_rate": 4.56228988245786e-05,
      "loss": 0.2366,
      "mean_token_accuracy": 0.9430207271128893,
      "num_tokens": 30449169.0,
      "step": 4080
    },
    {
      "entropy": 1.9020006403326988,
      "epoch": 0.2731697542974979,
      "grad_norm": 0.3423897325992584,
      "learning_rate": 4.55899279292112e-05,
      "loss": 0.2346,
      "mean_token_accuracy": 0.9429361525923013,
      "num_tokens": 30523259.0,
      "step": 4090
    },
    {
      "entropy": 1.8965322069823742,
      "epoch": 0.2738376510072717,
      "grad_norm": 0.29990923404693604,
      "learning_rate": 4.555684535527377e-05,
      "loss": 0.2293,
      "mean_token_accuracy": 0.9427106875926257,
      "num_tokens": 30598924.0,
      "step": 4100
    },
    {
      "entropy": 1.8720484122633934,
      "epoch": 0.27450554771704555,
      "grad_norm": 0.25237590074539185,
      "learning_rate": 4.552365128264107e-05,
      "loss": 0.225,
      "mean_token_accuracy": 0.946211227029562,
      "num_tokens": 30675727.0,
      "step": 4110
    },
    {
      "entropy": 1.847211516648531,
      "epoch": 0.2751734444268194,
      "grad_norm": 0.3517327308654785,
      "learning_rate": 4.549034589179409e-05,
      "loss": 0.2322,
      "mean_token_accuracy": 0.9446510747075081,
      "num_tokens": 30749848.0,
      "step": 4120
    },
    {
      "entropy": 1.8402021639049053,
      "epoch": 0.2758413411365932,
      "grad_norm": 0.3106963038444519,
      "learning_rate": 4.545692936381907e-05,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9477717727422714,
      "num_tokens": 30824000.0,
      "step": 4130
    },
    {
      "entropy": 1.8828895375132562,
      "epoch": 0.27650923784636705,
      "grad_norm": 0.35134467482566833,
      "learning_rate": 4.5423401880406546e-05,
      "loss": 0.2261,
      "mean_token_accuracy": 0.9444015204906464,
      "num_tokens": 30896221.0,
      "step": 4140
    },
    {
      "entropy": 1.8295325577259063,
      "epoch": 0.2771771345561409,
      "grad_norm": 0.2969716787338257,
      "learning_rate": 4.53897636238503e-05,
      "loss": 0.208,
      "mean_token_accuracy": 0.949476208910346,
      "num_tokens": 30971589.0,
      "step": 4150
    },
    {
      "entropy": 1.8696855299174786,
      "epoch": 0.2778450312659147,
      "grad_norm": 0.3581697344779968,
      "learning_rate": 4.53560147770464e-05,
      "loss": 0.2359,
      "mean_token_accuracy": 0.943967666849494,
      "num_tokens": 31044164.0,
      "step": 4160
    },
    {
      "entropy": 1.8627238169312477,
      "epoch": 0.27851292797568855,
      "grad_norm": 0.2857859432697296,
      "learning_rate": 4.532215552349224e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9427007786929608,
      "num_tokens": 31117045.0,
      "step": 4170
    },
    {
      "entropy": 1.8972482278943061,
      "epoch": 0.2791808246854624,
      "grad_norm": 0.29394519329071045,
      "learning_rate": 4.52881860472855e-05,
      "loss": 0.2394,
      "mean_token_accuracy": 0.9426060736179351,
      "num_tokens": 31191061.0,
      "step": 4180
    },
    {
      "entropy": 1.8200726546347141,
      "epoch": 0.2798487213952362,
      "grad_norm": 0.3024114966392517,
      "learning_rate": 4.5254106533123154e-05,
      "loss": 0.2272,
      "mean_token_accuracy": 0.9430060673505067,
      "num_tokens": 31267549.0,
      "step": 4190
    },
    {
      "entropy": 1.8552132204174996,
      "epoch": 0.28051661810501005,
      "grad_norm": 0.3161923289299011,
      "learning_rate": 4.521991716630046e-05,
      "loss": 0.2262,
      "mean_token_accuracy": 0.9449829947203398,
      "num_tokens": 31343976.0,
      "step": 4200
    },
    {
      "entropy": 1.8349798373878001,
      "epoch": 0.2811845148147839,
      "grad_norm": 0.3157711327075958,
      "learning_rate": 4.518561813270997e-05,
      "loss": 0.2353,
      "mean_token_accuracy": 0.9441523414105177,
      "num_tokens": 31416981.0,
      "step": 4210
    },
    {
      "entropy": 1.8618488132953643,
      "epoch": 0.2818524115245577,
      "grad_norm": 0.256456196308136,
      "learning_rate": 4.5151209618840516e-05,
      "loss": 0.2258,
      "mean_token_accuracy": 0.9452581308782101,
      "num_tokens": 31492733.0,
      "step": 4220
    },
    {
      "entropy": 1.8773194417357444,
      "epoch": 0.28252030823433155,
      "grad_norm": 0.2675033509731293,
      "learning_rate": 4.511669181177618e-05,
      "loss": 0.231,
      "mean_token_accuracy": 0.9448836881667375,
      "num_tokens": 31569492.0,
      "step": 4230
    },
    {
      "entropy": 1.906337308138609,
      "epoch": 0.2831882049441054,
      "grad_norm": 0.29556792974472046,
      "learning_rate": 4.5082064899195284e-05,
      "loss": 0.2375,
      "mean_token_accuracy": 0.9377251841127873,
      "num_tokens": 31640754.0,
      "step": 4240
    },
    {
      "entropy": 1.8609922096133231,
      "epoch": 0.2838561016538792,
      "grad_norm": 0.2518376410007477,
      "learning_rate": 4.504732906936938e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9435051172971726,
      "num_tokens": 31714627.0,
      "step": 4250
    },
    {
      "entropy": 1.8560601614415646,
      "epoch": 0.28452399836365305,
      "grad_norm": 0.32658979296684265,
      "learning_rate": 4.501248451116223e-05,
      "loss": 0.22,
      "mean_token_accuracy": 0.944538202509284,
      "num_tokens": 31788036.0,
      "step": 4260
    },
    {
      "entropy": 1.8571526892483234,
      "epoch": 0.2851918950734269,
      "grad_norm": 0.32422661781311035,
      "learning_rate": 4.497753141402875e-05,
      "loss": 0.242,
      "mean_token_accuracy": 0.941166504099965,
      "num_tokens": 31860429.0,
      "step": 4270
    },
    {
      "entropy": 1.805522048473358,
      "epoch": 0.2858597917832007,
      "grad_norm": 0.29409337043762207,
      "learning_rate": 4.4942469968013986e-05,
      "loss": 0.2165,
      "mean_token_accuracy": 0.9465219881385565,
      "num_tokens": 31938096.0,
      "step": 4280
    },
    {
      "entropy": 1.8214035376906395,
      "epoch": 0.28652768849297455,
      "grad_norm": 0.2931544780731201,
      "learning_rate": 4.4907300363752136e-05,
      "loss": 0.2236,
      "mean_token_accuracy": 0.9435908664017916,
      "num_tokens": 32011325.0,
      "step": 4290
    },
    {
      "entropy": 1.8379184059798717,
      "epoch": 0.2871955852027484,
      "grad_norm": 0.2970452606678009,
      "learning_rate": 4.487202279246543e-05,
      "loss": 0.2396,
      "mean_token_accuracy": 0.9429842695593834,
      "num_tokens": 32083739.0,
      "step": 4300
    },
    {
      "entropy": 1.8201567843556403,
      "epoch": 0.2878634819125222,
      "grad_norm": 0.26993146538734436,
      "learning_rate": 4.483663744596314e-05,
      "loss": 0.2288,
      "mean_token_accuracy": 0.9451590970158577,
      "num_tokens": 32159968.0,
      "step": 4310
    },
    {
      "entropy": 1.8193272963166236,
      "epoch": 0.28853137862229605,
      "grad_norm": 0.29660362005233765,
      "learning_rate": 4.4801144516640544e-05,
      "loss": 0.2166,
      "mean_token_accuracy": 0.9471204251050949,
      "num_tokens": 32235405.0,
      "step": 4320
    },
    {
      "entropy": 1.8298594862222672,
      "epoch": 0.2891992753320699,
      "grad_norm": 0.3144473731517792,
      "learning_rate": 4.476554419747785e-05,
      "loss": 0.2418,
      "mean_token_accuracy": 0.9419256884604692,
      "num_tokens": 32308991.0,
      "step": 4330
    },
    {
      "entropy": 1.815320910513401,
      "epoch": 0.2898671720418437,
      "grad_norm": 0.2645331919193268,
      "learning_rate": 4.472983668203915e-05,
      "loss": 0.2339,
      "mean_token_accuracy": 0.9424494832754136,
      "num_tokens": 32385652.0,
      "step": 4340
    },
    {
      "entropy": 1.8737569808959962,
      "epoch": 0.29053506875161755,
      "grad_norm": 0.3032049536705017,
      "learning_rate": 4.469402216447139e-05,
      "loss": 0.2592,
      "mean_token_accuracy": 0.9386070717126132,
      "num_tokens": 32456916.0,
      "step": 4350
    },
    {
      "entropy": 1.833969661593437,
      "epoch": 0.2912029654613914,
      "grad_norm": 0.2716382145881653,
      "learning_rate": 4.4658100839503324e-05,
      "loss": 0.2356,
      "mean_token_accuracy": 0.9422187585383653,
      "num_tokens": 32530601.0,
      "step": 4360
    },
    {
      "entropy": 1.837301241606474,
      "epoch": 0.2918708621711652,
      "grad_norm": 0.32120734453201294,
      "learning_rate": 4.462207290244438e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9441341172903777,
      "num_tokens": 32604989.0,
      "step": 4370
    },
    {
      "entropy": 1.8483294509351254,
      "epoch": 0.29253875888093905,
      "grad_norm": 0.26520368456840515,
      "learning_rate": 4.458593854918372e-05,
      "loss": 0.2323,
      "mean_token_accuracy": 0.942601065710187,
      "num_tokens": 32680902.0,
      "step": 4380
    },
    {
      "entropy": 1.8409473292529583,
      "epoch": 0.2932066555907129,
      "grad_norm": 0.32361602783203125,
      "learning_rate": 4.454969797618902e-05,
      "loss": 0.2157,
      "mean_token_accuracy": 0.9451724544167519,
      "num_tokens": 32755345.0,
      "step": 4390
    },
    {
      "entropy": 1.8125441461801528,
      "epoch": 0.2938745523004867,
      "grad_norm": 0.27777454257011414,
      "learning_rate": 4.451335138050558e-05,
      "loss": 0.2301,
      "mean_token_accuracy": 0.9425783287733793,
      "num_tokens": 32831474.0,
      "step": 4400
    },
    {
      "entropy": 1.8201068609952926,
      "epoch": 0.29454244901026055,
      "grad_norm": 0.3343043029308319,
      "learning_rate": 4.44768989597551e-05,
      "loss": 0.2313,
      "mean_token_accuracy": 0.9439606107771397,
      "num_tokens": 32906299.0,
      "step": 4410
    },
    {
      "entropy": 1.8236121252179145,
      "epoch": 0.2952103457200344,
      "grad_norm": 0.22313813865184784,
      "learning_rate": 4.444034091213468e-05,
      "loss": 0.2243,
      "mean_token_accuracy": 0.9450538717210293,
      "num_tokens": 32982509.0,
      "step": 4420
    },
    {
      "entropy": 1.8397392719984054,
      "epoch": 0.2958782424298082,
      "grad_norm": 0.28908953070640564,
      "learning_rate": 4.440367743641572e-05,
      "loss": 0.2442,
      "mean_token_accuracy": 0.940560732409358,
      "num_tokens": 33057779.0,
      "step": 4430
    },
    {
      "entropy": 1.8658656142652035,
      "epoch": 0.29654613913958205,
      "grad_norm": 0.3691408336162567,
      "learning_rate": 4.436690873194289e-05,
      "loss": 0.2287,
      "mean_token_accuracy": 0.9450929258018732,
      "num_tokens": 33132977.0,
      "step": 4440
    },
    {
      "entropy": 1.8281126484274863,
      "epoch": 0.2972140358493559,
      "grad_norm": 0.2597799003124237,
      "learning_rate": 4.433003499863296e-05,
      "loss": 0.2019,
      "mean_token_accuracy": 0.9486159149557352,
      "num_tokens": 33207351.0,
      "step": 4450
    },
    {
      "entropy": 1.8303055241703987,
      "epoch": 0.2978819325591297,
      "grad_norm": 0.29256701469421387,
      "learning_rate": 4.4293056436973765e-05,
      "loss": 0.2337,
      "mean_token_accuracy": 0.9425358675420284,
      "num_tokens": 33281504.0,
      "step": 4460
    },
    {
      "entropy": 1.8612960532307625,
      "epoch": 0.29854982926890355,
      "grad_norm": 0.33854877948760986,
      "learning_rate": 4.4255973248023116e-05,
      "loss": 0.2364,
      "mean_token_accuracy": 0.9431222815066576,
      "num_tokens": 33354876.0,
      "step": 4470
    },
    {
      "entropy": 1.8458468355238438,
      "epoch": 0.2992177259786774,
      "grad_norm": 0.33116385340690613,
      "learning_rate": 4.4218785633407704e-05,
      "loss": 0.2386,
      "mean_token_accuracy": 0.9436350584030151,
      "num_tokens": 33429475.0,
      "step": 4480
    },
    {
      "entropy": 1.865982796996832,
      "epoch": 0.2998856226884512,
      "grad_norm": 0.30375996232032776,
      "learning_rate": 4.4181493795321984e-05,
      "loss": 0.2307,
      "mean_token_accuracy": 0.9446152281016111,
      "num_tokens": 33502739.0,
      "step": 4490
    },
    {
      "entropy": 1.8897246249020099,
      "epoch": 0.30055351939822506,
      "grad_norm": 0.3325091004371643,
      "learning_rate": 4.414409793652709e-05,
      "loss": 0.2166,
      "mean_token_accuracy": 0.9460424091666937,
      "num_tokens": 33577606.0,
      "step": 4500
    },
    {
      "entropy": 1.8710330106317996,
      "epoch": 0.3012214161079989,
      "grad_norm": 0.34524285793304443,
      "learning_rate": 4.4106598260349735e-05,
      "loss": 0.2124,
      "mean_token_accuracy": 0.9496726516634226,
      "num_tokens": 33652812.0,
      "step": 4510
    },
    {
      "entropy": 1.8505713641643524,
      "epoch": 0.30188931281777276,
      "grad_norm": 0.2671905755996704,
      "learning_rate": 4.406899497068111e-05,
      "loss": 0.2238,
      "mean_token_accuracy": 0.9437539499253035,
      "num_tokens": 33729880.0,
      "step": 4520
    },
    {
      "entropy": 1.9004945039749146,
      "epoch": 0.30255720952754656,
      "grad_norm": 0.3384301960468292,
      "learning_rate": 4.4031288271975734e-05,
      "loss": 0.228,
      "mean_token_accuracy": 0.9426455538719892,
      "num_tokens": 33803589.0,
      "step": 4530
    },
    {
      "entropy": 1.8961285054683685,
      "epoch": 0.3032251062373204,
      "grad_norm": 0.30734241008758545,
      "learning_rate": 4.399347836925044e-05,
      "loss": 0.2121,
      "mean_token_accuracy": 0.9471046507358551,
      "num_tokens": 33881531.0,
      "step": 4540
    },
    {
      "entropy": 1.8047952473163604,
      "epoch": 0.30389300294709426,
      "grad_norm": 0.31810906529426575,
      "learning_rate": 4.395556546808314e-05,
      "loss": 0.2068,
      "mean_token_accuracy": 0.9483619179576636,
      "num_tokens": 33956428.0,
      "step": 4550
    },
    {
      "entropy": 1.8286520220339297,
      "epoch": 0.30456089965686806,
      "grad_norm": 0.2646723985671997,
      "learning_rate": 4.3917549774611774e-05,
      "loss": 0.2201,
      "mean_token_accuracy": 0.9456077262759208,
      "num_tokens": 34030162.0,
      "step": 4560
    },
    {
      "entropy": 1.8612429611384869,
      "epoch": 0.3052287963666419,
      "grad_norm": 0.32797884941101074,
      "learning_rate": 4.38794314955332e-05,
      "loss": 0.2367,
      "mean_token_accuracy": 0.9431695096194744,
      "num_tokens": 34105703.0,
      "step": 4570
    },
    {
      "entropy": 1.9067265942692757,
      "epoch": 0.30589669307641576,
      "grad_norm": 0.3241778314113617,
      "learning_rate": 4.384121083810202e-05,
      "loss": 0.2433,
      "mean_token_accuracy": 0.940237358957529,
      "num_tokens": 34179056.0,
      "step": 4580
    },
    {
      "entropy": 1.8829899229109288,
      "epoch": 0.30656458978618956,
      "grad_norm": 0.2544192373752594,
      "learning_rate": 4.380288801012951e-05,
      "loss": 0.2342,
      "mean_token_accuracy": 0.944569843634963,
      "num_tokens": 34252687.0,
      "step": 4590
    },
    {
      "entropy": 1.852818877995014,
      "epoch": 0.3072324864959634,
      "grad_norm": 0.28105655312538147,
      "learning_rate": 4.3764463219982455e-05,
      "loss": 0.2271,
      "mean_token_accuracy": 0.9435019325464964,
      "num_tokens": 34328899.0,
      "step": 4600
    },
    {
      "entropy": 1.8698960803449154,
      "epoch": 0.30790038320573726,
      "grad_norm": 0.3155994117259979,
      "learning_rate": 4.3725936676582e-05,
      "loss": 0.2344,
      "mean_token_accuracy": 0.9412890531122684,
      "num_tokens": 34403483.0,
      "step": 4610
    },
    {
      "entropy": 1.8788681253790855,
      "epoch": 0.30856827991551106,
      "grad_norm": 0.32375720143318176,
      "learning_rate": 4.368730858940256e-05,
      "loss": 0.2166,
      "mean_token_accuracy": 0.9469292409718036,
      "num_tokens": 34477321.0,
      "step": 4620
    },
    {
      "entropy": 1.8518653787672519,
      "epoch": 0.3092361766252849,
      "grad_norm": 0.2895698845386505,
      "learning_rate": 4.364857916847066e-05,
      "loss": 0.2076,
      "mean_token_accuracy": 0.9456298138946295,
      "num_tokens": 34552247.0,
      "step": 4630
    },
    {
      "entropy": 1.879375659674406,
      "epoch": 0.30990407333505876,
      "grad_norm": 0.2914004921913147,
      "learning_rate": 4.3609748624363776e-05,
      "loss": 0.2334,
      "mean_token_accuracy": 0.9401689626276493,
      "num_tokens": 34625508.0,
      "step": 4640
    },
    {
      "entropy": 1.845840898901224,
      "epoch": 0.31057197004483256,
      "grad_norm": 0.3413725793361664,
      "learning_rate": 4.357081716820922e-05,
      "loss": 0.2296,
      "mean_token_accuracy": 0.9441549833863974,
      "num_tokens": 34701073.0,
      "step": 4650
    },
    {
      "entropy": 1.8971908934414388,
      "epoch": 0.3112398667546064,
      "grad_norm": 0.3303360342979431,
      "learning_rate": 4.353178501168297e-05,
      "loss": 0.2335,
      "mean_token_accuracy": 0.9451367352157831,
      "num_tokens": 34775956.0,
      "step": 4660
    },
    {
      "entropy": 1.898054015636444,
      "epoch": 0.31190776346438026,
      "grad_norm": 0.27976253628730774,
      "learning_rate": 4.349265236700852e-05,
      "loss": 0.2214,
      "mean_token_accuracy": 0.9466199483722448,
      "num_tokens": 34852579.0,
      "step": 4670
    },
    {
      "entropy": 1.9111323684453965,
      "epoch": 0.31257566017415406,
      "grad_norm": 0.2828115224838257,
      "learning_rate": 4.345341944695574e-05,
      "loss": 0.2311,
      "mean_token_accuracy": 0.9418555200099945,
      "num_tokens": 34925466.0,
      "step": 4680
    },
    {
      "entropy": 1.8647110000252725,
      "epoch": 0.3132435568839279,
      "grad_norm": 0.2782430052757263,
      "learning_rate": 4.341408646483973e-05,
      "loss": 0.2289,
      "mean_token_accuracy": 0.9432813376188278,
      "num_tokens": 35000102.0,
      "step": 4690
    },
    {
      "entropy": 1.9170991756021976,
      "epoch": 0.31391145359370176,
      "grad_norm": 0.27305853366851807,
      "learning_rate": 4.3374653634519606e-05,
      "loss": 0.2365,
      "mean_token_accuracy": 0.9436933804303408,
      "num_tokens": 35075247.0,
      "step": 4700
    },
    {
      "entropy": 1.8815664917230606,
      "epoch": 0.31457935030347556,
      "grad_norm": 0.38132867217063904,
      "learning_rate": 4.3335121170397396e-05,
      "loss": 0.2273,
      "mean_token_accuracy": 0.9423738639801741,
      "num_tokens": 35148766.0,
      "step": 4710
    },
    {
      "entropy": 1.8863162621855736,
      "epoch": 0.3152472470132494,
      "grad_norm": 0.2659788727760315,
      "learning_rate": 4.329548928741685e-05,
      "loss": 0.2261,
      "mean_token_accuracy": 0.9450285650789738,
      "num_tokens": 35223172.0,
      "step": 4720
    },
    {
      "entropy": 1.8451788559556008,
      "epoch": 0.31591514372302326,
      "grad_norm": 0.3314361274242401,
      "learning_rate": 4.3255758201062256e-05,
      "loss": 0.2065,
      "mean_token_accuracy": 0.9494762852787971,
      "num_tokens": 35298405.0,
      "step": 4730
    },
    {
      "entropy": 1.8919203341007234,
      "epoch": 0.31658304043279706,
      "grad_norm": 0.32431700825691223,
      "learning_rate": 4.321592812735732e-05,
      "loss": 0.2379,
      "mean_token_accuracy": 0.9421888526529074,
      "num_tokens": 35375230.0,
      "step": 4740
    },
    {
      "entropy": 1.870032712817192,
      "epoch": 0.3172509371425709,
      "grad_norm": 0.2901422679424286,
      "learning_rate": 4.317599928286392e-05,
      "loss": 0.2163,
      "mean_token_accuracy": 0.9477691803127527,
      "num_tokens": 35451126.0,
      "step": 4750
    },
    {
      "entropy": 1.8814179971814156,
      "epoch": 0.31791883385234476,
      "grad_norm": 0.32144755125045776,
      "learning_rate": 4.313597188468099e-05,
      "loss": 0.2294,
      "mean_token_accuracy": 0.9436922177672387,
      "num_tokens": 35525614.0,
      "step": 4760
    },
    {
      "entropy": 1.9087035037577151,
      "epoch": 0.31858673056211856,
      "grad_norm": 0.3281264305114746,
      "learning_rate": 4.3095846150443284e-05,
      "loss": 0.2368,
      "mean_token_accuracy": 0.9433741334825754,
      "num_tokens": 35597957.0,
      "step": 4770
    },
    {
      "entropy": 1.8724081672728061,
      "epoch": 0.3192546272718924,
      "grad_norm": 0.2671878933906555,
      "learning_rate": 4.3055622298320276e-05,
      "loss": 0.233,
      "mean_token_accuracy": 0.940693574771285,
      "num_tokens": 35672467.0,
      "step": 4780
    },
    {
      "entropy": 1.9181607067584991,
      "epoch": 0.31992252398166626,
      "grad_norm": 0.3136788308620453,
      "learning_rate": 4.301530054701486e-05,
      "loss": 0.2216,
      "mean_token_accuracy": 0.9458787549287081,
      "num_tokens": 35750079.0,
      "step": 4790
    },
    {
      "entropy": 1.861856424808502,
      "epoch": 0.32059042069144006,
      "grad_norm": 0.2824789583683014,
      "learning_rate": 4.2974881115762274e-05,
      "loss": 0.2277,
      "mean_token_accuracy": 0.9444599103182554,
      "num_tokens": 35825831.0,
      "step": 4800
    },
    {
      "entropy": 1.912356250733137,
      "epoch": 0.3212583174012139,
      "grad_norm": 0.35055530071258545,
      "learning_rate": 4.293436422432882e-05,
      "loss": 0.2389,
      "mean_token_accuracy": 0.9436374962329864,
      "num_tokens": 35899099.0,
      "step": 4810
    },
    {
      "entropy": 1.934591693431139,
      "epoch": 0.32192621411098776,
      "grad_norm": 0.36533433198928833,
      "learning_rate": 4.289375009301072e-05,
      "loss": 0.2467,
      "mean_token_accuracy": 0.940224488452077,
      "num_tokens": 35970280.0,
      "step": 4820
    },
    {
      "entropy": 1.8885440282523631,
      "epoch": 0.32259411082076156,
      "grad_norm": 0.3091188669204712,
      "learning_rate": 4.28530389426329e-05,
      "loss": 0.2243,
      "mean_token_accuracy": 0.9452458128333092,
      "num_tokens": 36044645.0,
      "step": 4830
    },
    {
      "entropy": 1.924134112149477,
      "epoch": 0.3232620075305354,
      "grad_norm": 0.3028869032859802,
      "learning_rate": 4.2812230994547806e-05,
      "loss": 0.2257,
      "mean_token_accuracy": 0.9441969476640224,
      "num_tokens": 36119364.0,
      "step": 4840
    },
    {
      "entropy": 1.8526901558041573,
      "epoch": 0.32392990424030926,
      "grad_norm": 0.26295527815818787,
      "learning_rate": 4.2771326470634165e-05,
      "loss": 0.2123,
      "mean_token_accuracy": 0.9481824822723866,
      "num_tokens": 36195357.0,
      "step": 4850
    },
    {
      "entropy": 1.877461788803339,
      "epoch": 0.32459780095008306,
      "grad_norm": 0.2593669295310974,
      "learning_rate": 4.273032559329582e-05,
      "loss": 0.2349,
      "mean_token_accuracy": 0.9420103743672371,
      "num_tokens": 36269141.0,
      "step": 4860
    },
    {
      "entropy": 1.9075257934629917,
      "epoch": 0.3252656976598569,
      "grad_norm": 0.2766685485839844,
      "learning_rate": 4.2689228585460494e-05,
      "loss": 0.2407,
      "mean_token_accuracy": 0.9416776720434428,
      "num_tokens": 36341450.0,
      "step": 4870
    },
    {
      "entropy": 1.8789460368454456,
      "epoch": 0.32593359436963076,
      "grad_norm": 0.2707120478153229,
      "learning_rate": 4.2648035670578574e-05,
      "loss": 0.2195,
      "mean_token_accuracy": 0.9438760846853256,
      "num_tokens": 36416193.0,
      "step": 4880
    },
    {
      "entropy": 1.9080967545509337,
      "epoch": 0.32660149107940456,
      "grad_norm": 0.40785717964172363,
      "learning_rate": 4.260674707262193e-05,
      "loss": 0.2526,
      "mean_token_accuracy": 0.9366442874073982,
      "num_tokens": 36487303.0,
      "step": 4890
    },
    {
      "entropy": 1.8581715993583203,
      "epoch": 0.3272693877891784,
      "grad_norm": 0.31682661175727844,
      "learning_rate": 4.256536301608265e-05,
      "loss": 0.2124,
      "mean_token_accuracy": 0.9485861904919147,
      "num_tokens": 36562424.0,
      "step": 4900
    },
    {
      "entropy": 1.926520000398159,
      "epoch": 0.32793728449895226,
      "grad_norm": 0.30442801117897034,
      "learning_rate": 4.2523883725971876e-05,
      "loss": 0.2553,
      "mean_token_accuracy": 0.9383448876440525,
      "num_tokens": 36634810.0,
      "step": 4910
    },
    {
      "entropy": 1.9025879234075547,
      "epoch": 0.32860518120872606,
      "grad_norm": 0.3349396586418152,
      "learning_rate": 4.248230942781851e-05,
      "loss": 0.2517,
      "mean_token_accuracy": 0.9404251247644424,
      "num_tokens": 36706875.0,
      "step": 4920
    },
    {
      "entropy": 1.8629882097244264,
      "epoch": 0.3292730779184999,
      "grad_norm": 0.30690711736679077,
      "learning_rate": 4.2440640347668066e-05,
      "loss": 0.2237,
      "mean_token_accuracy": 0.9463497452437878,
      "num_tokens": 36782323.0,
      "step": 4930
    },
    {
      "entropy": 1.9088253237307071,
      "epoch": 0.32994097462827376,
      "grad_norm": 0.25188544392585754,
      "learning_rate": 4.239887671208136e-05,
      "loss": 0.2165,
      "mean_token_accuracy": 0.9470059532672167,
      "num_tokens": 36858200.0,
      "step": 4940
    },
    {
      "entropy": 1.8869219616055488,
      "epoch": 0.33060887133804756,
      "grad_norm": 0.3319534957408905,
      "learning_rate": 4.2357018748133364e-05,
      "loss": 0.2578,
      "mean_token_accuracy": 0.9386107049882412,
      "num_tokens": 36931357.0,
      "step": 4950
    },
    {
      "entropy": 1.8974449284374715,
      "epoch": 0.3312767680478214,
      "grad_norm": 0.3015380799770355,
      "learning_rate": 4.231506668341189e-05,
      "loss": 0.2375,
      "mean_token_accuracy": 0.9412260301411152,
      "num_tokens": 37005948.0,
      "step": 4960
    },
    {
      "entropy": 1.8834066174924373,
      "epoch": 0.33194466475759526,
      "grad_norm": 0.2521904706954956,
      "learning_rate": 4.22730207460164e-05,
      "loss": 0.228,
      "mean_token_accuracy": 0.9452286325395107,
      "num_tokens": 37080267.0,
      "step": 4970
    },
    {
      "entropy": 1.8718890838325024,
      "epoch": 0.33261256146736906,
      "grad_norm": 0.258169561624527,
      "learning_rate": 4.223088116455679e-05,
      "loss": 0.2306,
      "mean_token_accuracy": 0.9449196413159371,
      "num_tokens": 37156854.0,
      "step": 4980
    },
    {
      "entropy": 1.8671221904456616,
      "epoch": 0.3332804581771429,
      "grad_norm": 0.3138870596885681,
      "learning_rate": 4.218864816815205e-05,
      "loss": 0.2305,
      "mean_token_accuracy": 0.9467808868736028,
      "num_tokens": 37230708.0,
      "step": 4990
    },
    {
      "entropy": 1.875975839793682,
      "epoch": 0.33394835488691677,
      "grad_norm": 0.26833420991897583,
      "learning_rate": 4.214632198642911e-05,
      "loss": 0.218,
      "mean_token_accuracy": 0.9458157274872064,
      "num_tokens": 37305729.0,
      "step": 5000
    },
    {
      "epoch": 0.33394835488691677,
      "eval_entropy": 1.8628402607917787,
      "eval_loss": 0.247247114777565,
      "eval_mean_token_accuracy": 0.9443248253226281,
      "eval_num_tokens": 37305729.0,
      "eval_runtime": 552.3326,
      "eval_samples_per_second": 36.21,
      "eval_steps_per_second": 9.053,
      "step": 5000
    },
    {
      "entropy": 1.8699682176113128,
      "epoch": 0.33461625159669056,
      "grad_norm": 0.3509800136089325,
      "learning_rate": 4.2103902849521585e-05,
      "loss": 0.2361,
      "mean_token_accuracy": 0.9422809060662984,
      "num_tokens": 37378348.0,
      "step": 5010
    },
    {
      "entropy": 1.8798143222928048,
      "epoch": 0.3352841483064644,
      "grad_norm": 0.35147419571876526,
      "learning_rate": 4.2061390988068476e-05,
      "loss": 0.2646,
      "mean_token_accuracy": 0.9363960832357406,
      "num_tokens": 37452337.0,
      "step": 5020
    },
    {
      "entropy": 1.8594155445694924,
      "epoch": 0.33595204501623827,
      "grad_norm": 0.30872824788093567,
      "learning_rate": 4.201878663321295e-05,
      "loss": 0.2289,
      "mean_token_accuracy": 0.943534504249692,
      "num_tokens": 37526211.0,
      "step": 5030
    },
    {
      "entropy": 1.9377403013408183,
      "epoch": 0.33661994172601206,
      "grad_norm": 0.31443890929222107,
      "learning_rate": 4.197609001660106e-05,
      "loss": 0.2231,
      "mean_token_accuracy": 0.9447415392845869,
      "num_tokens": 37602869.0,
      "step": 5040
    },
    {
      "entropy": 1.9212867490947247,
      "epoch": 0.3372878384357859,
      "grad_norm": 0.3028574585914612,
      "learning_rate": 4.19333013703805e-05,
      "loss": 0.2336,
      "mean_token_accuracy": 0.9436277154833078,
      "num_tokens": 37675465.0,
      "step": 5050
    },
    {
      "entropy": 1.8429759316146375,
      "epoch": 0.33795573514555977,
      "grad_norm": 0.3305186331272125,
      "learning_rate": 4.189042092719937e-05,
      "loss": 0.2194,
      "mean_token_accuracy": 0.9478144496679306,
      "num_tokens": 37750154.0,
      "step": 5060
    },
    {
      "entropy": 1.8589183568954468,
      "epoch": 0.33862363185533356,
      "grad_norm": 0.3074212670326233,
      "learning_rate": 4.184744892020485e-05,
      "loss": 0.2249,
      "mean_token_accuracy": 0.9462077233940362,
      "num_tokens": 37828307.0,
      "step": 5070
    },
    {
      "entropy": 1.8754634030163289,
      "epoch": 0.3392915285651074,
      "grad_norm": 0.2615527808666229,
      "learning_rate": 4.1804385583042e-05,
      "loss": 0.2258,
      "mean_token_accuracy": 0.9431425549089909,
      "num_tokens": 37900994.0,
      "step": 5080
    },
    {
      "entropy": 1.8587918348610402,
      "epoch": 0.33995942527488127,
      "grad_norm": 0.3230578899383545,
      "learning_rate": 4.1761231149852414e-05,
      "loss": 0.2283,
      "mean_token_accuracy": 0.9427317231893539,
      "num_tokens": 37977847.0,
      "step": 5090
    },
    {
      "entropy": 1.8816634833812713,
      "epoch": 0.34062732198465506,
      "grad_norm": 0.28146666288375854,
      "learning_rate": 4.171798585527302e-05,
      "loss": 0.2411,
      "mean_token_accuracy": 0.9417926289141179,
      "num_tokens": 38052360.0,
      "step": 5100
    },
    {
      "entropy": 1.8864408075809478,
      "epoch": 0.3412952186944289,
      "grad_norm": 0.3060280680656433,
      "learning_rate": 4.1674649934434757e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9435448095202446,
      "num_tokens": 38129125.0,
      "step": 5110
    },
    {
      "entropy": 1.8035241551697254,
      "epoch": 0.34196311540420277,
      "grad_norm": 0.27368125319480896,
      "learning_rate": 4.163122362296133e-05,
      "loss": 0.2173,
      "mean_token_accuracy": 0.9452866896986961,
      "num_tokens": 38204358.0,
      "step": 5120
    },
    {
      "entropy": 1.8450581595301627,
      "epoch": 0.34263101211397656,
      "grad_norm": 0.2824119031429291,
      "learning_rate": 4.15877071569679e-05,
      "loss": 0.2329,
      "mean_token_accuracy": 0.9438250198960304,
      "num_tokens": 38278942.0,
      "step": 5130
    },
    {
      "entropy": 1.8859374925494194,
      "epoch": 0.3432989088237504,
      "grad_norm": 0.3053126931190491,
      "learning_rate": 4.154410077305979e-05,
      "loss": 0.2362,
      "mean_token_accuracy": 0.9420538257807494,
      "num_tokens": 38353763.0,
      "step": 5140
    },
    {
      "entropy": 1.87126986682415,
      "epoch": 0.34396680553352427,
      "grad_norm": 0.31491461396217346,
      "learning_rate": 4.1500404708331266e-05,
      "loss": 0.2361,
      "mean_token_accuracy": 0.9447729744017124,
      "num_tokens": 38427115.0,
      "step": 5150
    },
    {
      "entropy": 1.8601775549352169,
      "epoch": 0.34463470224329806,
      "grad_norm": 0.24703249335289001,
      "learning_rate": 4.145661920036415e-05,
      "loss": 0.2139,
      "mean_token_accuracy": 0.9474674869328737,
      "num_tokens": 38502357.0,
      "step": 5160
    },
    {
      "entropy": 1.8727555222809316,
      "epoch": 0.3453025989530719,
      "grad_norm": 0.23795869946479797,
      "learning_rate": 4.141274448722661e-05,
      "loss": 0.2156,
      "mean_token_accuracy": 0.9438554402440786,
      "num_tokens": 38578874.0,
      "step": 5170
    },
    {
      "entropy": 1.8991989843547343,
      "epoch": 0.34597049566284577,
      "grad_norm": 0.34368664026260376,
      "learning_rate": 4.1368780807471855e-05,
      "loss": 0.2336,
      "mean_token_accuracy": 0.9409043125808239,
      "num_tokens": 38651842.0,
      "step": 5180
    },
    {
      "entropy": 1.8837950490415096,
      "epoch": 0.34663839237261956,
      "grad_norm": 0.33735916018486023,
      "learning_rate": 4.132472840013677e-05,
      "loss": 0.2321,
      "mean_token_accuracy": 0.9454182650893926,
      "num_tokens": 38726411.0,
      "step": 5190
    },
    {
      "entropy": 1.9352252803742886,
      "epoch": 0.3473062890823934,
      "grad_norm": 0.3296850621700287,
      "learning_rate": 4.1280587504740696e-05,
      "loss": 0.2243,
      "mean_token_accuracy": 0.9450970035046339,
      "num_tokens": 38800458.0,
      "step": 5200
    },
    {
      "entropy": 1.899579556286335,
      "epoch": 0.34797418579216727,
      "grad_norm": 0.35265713930130005,
      "learning_rate": 4.1236358361284084e-05,
      "loss": 0.2473,
      "mean_token_accuracy": 0.9409931804984808,
      "num_tokens": 38876462.0,
      "step": 5210
    },
    {
      "entropy": 1.8974872305989265,
      "epoch": 0.34864208250194106,
      "grad_norm": 0.29030710458755493,
      "learning_rate": 4.119204121024721e-05,
      "loss": 0.2325,
      "mean_token_accuracy": 0.9438703712075949,
      "num_tokens": 38952029.0,
      "step": 5220
    },
    {
      "entropy": 1.8944323100149632,
      "epoch": 0.3493099792117149,
      "grad_norm": 0.2948436439037323,
      "learning_rate": 4.1147636292588874e-05,
      "loss": 0.2438,
      "mean_token_accuracy": 0.9398472532629967,
      "num_tokens": 39027554.0,
      "step": 5230
    },
    {
      "entropy": 1.8910538606345653,
      "epoch": 0.34997787592148877,
      "grad_norm": 0.24279309809207916,
      "learning_rate": 4.110314384974504e-05,
      "loss": 0.2214,
      "mean_token_accuracy": 0.947320806607604,
      "num_tokens": 39103963.0,
      "step": 5240
    },
    {
      "entropy": 1.8806976824998856,
      "epoch": 0.35064577263126256,
      "grad_norm": 0.3154253363609314,
      "learning_rate": 4.105856412362759e-05,
      "loss": 0.2126,
      "mean_token_accuracy": 0.9457191519439221,
      "num_tokens": 39180417.0,
      "step": 5250
    },
    {
      "entropy": 1.8734717287123204,
      "epoch": 0.3513136693410364,
      "grad_norm": 0.34275585412979126,
      "learning_rate": 4.101389735662296e-05,
      "loss": 0.219,
      "mean_token_accuracy": 0.9458412297070027,
      "num_tokens": 39259315.0,
      "step": 5260
    },
    {
      "entropy": 1.8300933994352817,
      "epoch": 0.35198156605081027,
      "grad_norm": 0.2978379726409912,
      "learning_rate": 4.0969143791590844e-05,
      "loss": 0.226,
      "mean_token_accuracy": 0.9459999747574329,
      "num_tokens": 39334551.0,
      "step": 5270
    },
    {
      "entropy": 1.8873732060194015,
      "epoch": 0.35264946276058406,
      "grad_norm": 0.2864626348018646,
      "learning_rate": 4.0924303671862894e-05,
      "loss": 0.2409,
      "mean_token_accuracy": 0.9399205647408962,
      "num_tokens": 39409350.0,
      "step": 5280
    },
    {
      "entropy": 1.8625124730169773,
      "epoch": 0.3533173594703579,
      "grad_norm": 0.2921641170978546,
      "learning_rate": 4.0879377241241317e-05,
      "loss": 0.2266,
      "mean_token_accuracy": 0.9438492223620415,
      "num_tokens": 39483174.0,
      "step": 5290
    },
    {
      "entropy": 1.8765275306999682,
      "epoch": 0.35398525618013177,
      "grad_norm": 0.2676922082901001,
      "learning_rate": 4.083436474399766e-05,
      "loss": 0.2279,
      "mean_token_accuracy": 0.9452288087457419,
      "num_tokens": 39558669.0,
      "step": 5300
    },
    {
      "entropy": 1.9189662396907807,
      "epoch": 0.35465315288990557,
      "grad_norm": 0.3305489122867584,
      "learning_rate": 4.0789266424871394e-05,
      "loss": 0.2352,
      "mean_token_accuracy": 0.9428542200475931,
      "num_tokens": 39632021.0,
      "step": 5310
    },
    {
      "entropy": 1.879033349454403,
      "epoch": 0.3553210495996794,
      "grad_norm": 0.28791821002960205,
      "learning_rate": 4.074408252906864e-05,
      "loss": 0.2307,
      "mean_token_accuracy": 0.9449928559362888,
      "num_tokens": 39709464.0,
      "step": 5320
    },
    {
      "entropy": 1.8804195158183574,
      "epoch": 0.35598894630945327,
      "grad_norm": 0.303653359413147,
      "learning_rate": 4.06988133022608e-05,
      "loss": 0.2386,
      "mean_token_accuracy": 0.9407651532441378,
      "num_tokens": 39783333.0,
      "step": 5330
    },
    {
      "entropy": 1.9090037167072296,
      "epoch": 0.35665684301922707,
      "grad_norm": 0.32017049193382263,
      "learning_rate": 4.0653458990583224e-05,
      "loss": 0.2307,
      "mean_token_accuracy": 0.9425357978790998,
      "num_tokens": 39854691.0,
      "step": 5340
    },
    {
      "entropy": 1.8525074176490306,
      "epoch": 0.3573247397290009,
      "grad_norm": 0.28389784693717957,
      "learning_rate": 4.060801984063388e-05,
      "loss": 0.2246,
      "mean_token_accuracy": 0.944192036613822,
      "num_tokens": 39931201.0,
      "step": 5350
    },
    {
      "entropy": 1.8636192552745343,
      "epoch": 0.35799263643877477,
      "grad_norm": 0.2678419351577759,
      "learning_rate": 4.056249609947204e-05,
      "loss": 0.235,
      "mean_token_accuracy": 0.9450872488319874,
      "num_tokens": 40007127.0,
      "step": 5360
    },
    {
      "entropy": 1.8853872954845428,
      "epoch": 0.35866053314854857,
      "grad_norm": 0.2968935966491699,
      "learning_rate": 4.051688801461688e-05,
      "loss": 0.2126,
      "mean_token_accuracy": 0.9481700476258993,
      "num_tokens": 40082387.0,
      "step": 5370
    },
    {
      "entropy": 1.8639925502240657,
      "epoch": 0.3593284298583224,
      "grad_norm": 0.3370371460914612,
      "learning_rate": 4.04711958340462e-05,
      "loss": 0.2279,
      "mean_token_accuracy": 0.9431246116757392,
      "num_tokens": 40156022.0,
      "step": 5380
    },
    {
      "entropy": 1.873419537395239,
      "epoch": 0.35999632656809627,
      "grad_norm": 0.35656043887138367,
      "learning_rate": 4.042541980619499e-05,
      "loss": 0.2302,
      "mean_token_accuracy": 0.9458245374262333,
      "num_tokens": 40229621.0,
      "step": 5390
    },
    {
      "entropy": 1.8834988854825496,
      "epoch": 0.36066422327787007,
      "grad_norm": 0.3486693799495697,
      "learning_rate": 4.037956017995418e-05,
      "loss": 0.2259,
      "mean_token_accuracy": 0.9466361660510302,
      "num_tokens": 40303939.0,
      "step": 5400
    },
    {
      "entropy": 1.8507674120366573,
      "epoch": 0.3613321199876439,
      "grad_norm": 0.242417111992836,
      "learning_rate": 4.03336172046692e-05,
      "loss": 0.212,
      "mean_token_accuracy": 0.9497998952865601,
      "num_tokens": 40378777.0,
      "step": 5410
    },
    {
      "entropy": 1.8961491219699382,
      "epoch": 0.36200001669741777,
      "grad_norm": 0.3118096590042114,
      "learning_rate": 4.028759113013868e-05,
      "loss": 0.2414,
      "mean_token_accuracy": 0.9415444631129504,
      "num_tokens": 40451732.0,
      "step": 5420
    },
    {
      "entropy": 1.8738715805113315,
      "epoch": 0.36266791340719157,
      "grad_norm": 0.3759858310222626,
      "learning_rate": 4.024148220661309e-05,
      "loss": 0.2274,
      "mean_token_accuracy": 0.9449708998203278,
      "num_tokens": 40524296.0,
      "step": 5430
    },
    {
      "entropy": 1.8486330926418304,
      "epoch": 0.3633358101169654,
      "grad_norm": 0.30139195919036865,
      "learning_rate": 4.019529068479332e-05,
      "loss": 0.2212,
      "mean_token_accuracy": 0.9490444637835026,
      "num_tokens": 40599827.0,
      "step": 5440
    },
    {
      "entropy": 1.8444453194737433,
      "epoch": 0.36400370682673927,
      "grad_norm": 0.3065475821495056,
      "learning_rate": 4.014901681582938e-05,
      "loss": 0.2205,
      "mean_token_accuracy": 0.9450596127659082,
      "num_tokens": 40675038.0,
      "step": 5450
    },
    {
      "entropy": 1.8432600557804109,
      "epoch": 0.36467160353651307,
      "grad_norm": 0.35564085841178894,
      "learning_rate": 4.010266085131902e-05,
      "loss": 0.215,
      "mean_token_accuracy": 0.9459651198238135,
      "num_tokens": 40751590.0,
      "step": 5460
    },
    {
      "entropy": 1.8573269747197627,
      "epoch": 0.3653395002462869,
      "grad_norm": 0.300121545791626,
      "learning_rate": 4.005622304330635e-05,
      "loss": 0.2185,
      "mean_token_accuracy": 0.9460552595555782,
      "num_tokens": 40828149.0,
      "step": 5470
    },
    {
      "entropy": 1.90805182158947,
      "epoch": 0.36600739695606077,
      "grad_norm": 0.33883431553840637,
      "learning_rate": 4.0009703644280465e-05,
      "loss": 0.24,
      "mean_token_accuracy": 0.9444262105971575,
      "num_tokens": 40906033.0,
      "step": 5480
    },
    {
      "entropy": 1.9340757705271243,
      "epoch": 0.36667529366583457,
      "grad_norm": 0.2954683005809784,
      "learning_rate": 3.996310290717411e-05,
      "loss": 0.2483,
      "mean_token_accuracy": 0.9406178243458271,
      "num_tokens": 40979735.0,
      "step": 5490
    },
    {
      "entropy": 1.887842309474945,
      "epoch": 0.3673431903756084,
      "grad_norm": 0.2348015457391739,
      "learning_rate": 3.991642108536224e-05,
      "loss": 0.226,
      "mean_token_accuracy": 0.9455922391265631,
      "num_tokens": 41056378.0,
      "step": 5500
    },
    {
      "entropy": 1.894931075721979,
      "epoch": 0.36801108708538227,
      "grad_norm": 0.28648632764816284,
      "learning_rate": 3.9869658432660696e-05,
      "loss": 0.2359,
      "mean_token_accuracy": 0.9441044371575117,
      "num_tokens": 41130044.0,
      "step": 5510
    },
    {
      "entropy": 1.89505675137043,
      "epoch": 0.36867898379515607,
      "grad_norm": 0.4146939218044281,
      "learning_rate": 3.982281520332481e-05,
      "loss": 0.2475,
      "mean_token_accuracy": 0.9394860856235028,
      "num_tokens": 41206835.0,
      "step": 5520
    },
    {
      "entropy": 1.9290433168411254,
      "epoch": 0.3693468805049299,
      "grad_norm": 0.40325871109962463,
      "learning_rate": 3.977589165204804e-05,
      "loss": 0.2172,
      "mean_token_accuracy": 0.9462254717946053,
      "num_tokens": 41285784.0,
      "step": 5530
    },
    {
      "entropy": 1.8551645264029504,
      "epoch": 0.37001477721470377,
      "grad_norm": 0.2665170133113861,
      "learning_rate": 3.9728888033960525e-05,
      "loss": 0.2085,
      "mean_token_accuracy": 0.9506573807448149,
      "num_tokens": 41363196.0,
      "step": 5540
    },
    {
      "entropy": 1.885726773738861,
      "epoch": 0.37068267392447757,
      "grad_norm": 0.3068387508392334,
      "learning_rate": 3.9681804604627765e-05,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9449964951723814,
      "num_tokens": 41438447.0,
      "step": 5550
    },
    {
      "entropy": 1.898128830641508,
      "epoch": 0.3713505706342514,
      "grad_norm": 0.2888367772102356,
      "learning_rate": 3.96346416200492e-05,
      "loss": 0.245,
      "mean_token_accuracy": 0.939350976422429,
      "num_tokens": 41513874.0,
      "step": 5560
    },
    {
      "entropy": 1.920731969177723,
      "epoch": 0.37201846734402527,
      "grad_norm": 0.28700539469718933,
      "learning_rate": 3.958739933665683e-05,
      "loss": 0.2327,
      "mean_token_accuracy": 0.9404160708189011,
      "num_tokens": 41589871.0,
      "step": 5570
    },
    {
      "entropy": 1.8654952719807625,
      "epoch": 0.37268636405379907,
      "grad_norm": 0.30696895718574524,
      "learning_rate": 3.954007801131381e-05,
      "loss": 0.2215,
      "mean_token_accuracy": 0.943479948490858,
      "num_tokens": 41664208.0,
      "step": 5580
    },
    {
      "entropy": 1.8375459298491479,
      "epoch": 0.3733542607635729,
      "grad_norm": 0.28250840306282043,
      "learning_rate": 3.949267790131305e-05,
      "loss": 0.2085,
      "mean_token_accuracy": 0.94823301397264,
      "num_tokens": 41740079.0,
      "step": 5590
    },
    {
      "entropy": 1.8448408506810665,
      "epoch": 0.3740221574733468,
      "grad_norm": 0.2599450647830963,
      "learning_rate": 3.9445199264375826e-05,
      "loss": 0.232,
      "mean_token_accuracy": 0.9453869223594665,
      "num_tokens": 41816937.0,
      "step": 5600
    },
    {
      "entropy": 1.8887805461883544,
      "epoch": 0.37469005418312057,
      "grad_norm": 0.2883719503879547,
      "learning_rate": 3.939764235865039e-05,
      "loss": 0.2432,
      "mean_token_accuracy": 0.9409231137484312,
      "num_tokens": 41890371.0,
      "step": 5610
    },
    {
      "entropy": 1.9057791784405709,
      "epoch": 0.3753579508928944,
      "grad_norm": 0.26792749762535095,
      "learning_rate": 3.935000744271053e-05,
      "loss": 0.2279,
      "mean_token_accuracy": 0.9451626099646091,
      "num_tokens": 41967896.0,
      "step": 5620
    },
    {
      "entropy": 1.8630941912531853,
      "epoch": 0.3760258476026683,
      "grad_norm": 0.3250010311603546,
      "learning_rate": 3.930229477555421e-05,
      "loss": 0.2308,
      "mean_token_accuracy": 0.9438260331749916,
      "num_tokens": 42041454.0,
      "step": 5630
    },
    {
      "entropy": 1.8576290264725686,
      "epoch": 0.37669374431244207,
      "grad_norm": 0.34566250443458557,
      "learning_rate": 3.925450461660212e-05,
      "loss": 0.2274,
      "mean_token_accuracy": 0.9449837531894445,
      "num_tokens": 42116184.0,
      "step": 5640
    },
    {
      "entropy": 1.9070193007588387,
      "epoch": 0.3773616410222159,
      "grad_norm": 0.29610711336135864,
      "learning_rate": 3.92066372256963e-05,
      "loss": 0.2246,
      "mean_token_accuracy": 0.9449803315103054,
      "num_tokens": 42191456.0,
      "step": 5650
    },
    {
      "entropy": 1.8656429432332515,
      "epoch": 0.3780295377319898,
      "grad_norm": 0.2796553373336792,
      "learning_rate": 3.915869286309869e-05,
      "loss": 0.2207,
      "mean_token_accuracy": 0.945483610779047,
      "num_tokens": 42267930.0,
      "step": 5660
    },
    {
      "entropy": 1.8829548478126525,
      "epoch": 0.37869743444176357,
      "grad_norm": 0.26579412817955017,
      "learning_rate": 3.911067178948977e-05,
      "loss": 0.2277,
      "mean_token_accuracy": 0.9436649471521378,
      "num_tokens": 42343385.0,
      "step": 5670
    },
    {
      "entropy": 1.893177405744791,
      "epoch": 0.3793653311515374,
      "grad_norm": 0.3126572072505951,
      "learning_rate": 3.9062574265967075e-05,
      "loss": 0.2455,
      "mean_token_accuracy": 0.9410084024071693,
      "num_tokens": 42416351.0,
      "step": 5680
    },
    {
      "entropy": 1.8615103863179683,
      "epoch": 0.3800332278613113,
      "grad_norm": 0.3685029447078705,
      "learning_rate": 3.901440055404383e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9446178272366523,
      "num_tokens": 42492057.0,
      "step": 5690
    },
    {
      "entropy": 1.8649059802293777,
      "epoch": 0.38070112457108507,
      "grad_norm": 0.2746874988079071,
      "learning_rate": 3.8966150915647484e-05,
      "loss": 0.2312,
      "mean_token_accuracy": 0.9415227670222521,
      "num_tokens": 42566103.0,
      "step": 5700
    },
    {
      "entropy": 1.8582611083984375,
      "epoch": 0.3813690212808589,
      "grad_norm": 0.3445000648498535,
      "learning_rate": 3.8917825613118345e-05,
      "loss": 0.2248,
      "mean_token_accuracy": 0.9451813399791718,
      "num_tokens": 42641148.0,
      "step": 5710
    },
    {
      "entropy": 1.8385150879621506,
      "epoch": 0.3820369179906328,
      "grad_norm": 0.27385690808296204,
      "learning_rate": 3.886942490920809e-05,
      "loss": 0.2353,
      "mean_token_accuracy": 0.9439975164830685,
      "num_tokens": 42714158.0,
      "step": 5720
    },
    {
      "entropy": 1.8803850308060646,
      "epoch": 0.38270481470040657,
      "grad_norm": 0.2621834874153137,
      "learning_rate": 3.882094906707838e-05,
      "loss": 0.2353,
      "mean_token_accuracy": 0.9418150596320629,
      "num_tokens": 42789330.0,
      "step": 5730
    },
    {
      "entropy": 1.9232642516493796,
      "epoch": 0.3833727114101804,
      "grad_norm": 0.3616364598274231,
      "learning_rate": 3.8772398350299396e-05,
      "loss": 0.2575,
      "mean_token_accuracy": 0.9402707811444998,
      "num_tokens": 42860054.0,
      "step": 5740
    },
    {
      "entropy": 1.91171570494771,
      "epoch": 0.3840406081199543,
      "grad_norm": 0.389970064163208,
      "learning_rate": 3.872377302284843e-05,
      "loss": 0.2115,
      "mean_token_accuracy": 0.9457026407122612,
      "num_tokens": 42935941.0,
      "step": 5750
    },
    {
      "entropy": 1.9102558188140393,
      "epoch": 0.38470850482972807,
      "grad_norm": 0.2997053265571594,
      "learning_rate": 3.867507334910846e-05,
      "loss": 0.2391,
      "mean_token_accuracy": 0.9421985533088446,
      "num_tokens": 43010176.0,
      "step": 5760
    },
    {
      "entropy": 1.9124287240207196,
      "epoch": 0.3853764015395019,
      "grad_norm": 0.2375487983226776,
      "learning_rate": 3.8626299593866675e-05,
      "loss": 0.2154,
      "mean_token_accuracy": 0.9474853195250035,
      "num_tokens": 43086208.0,
      "step": 5770
    },
    {
      "entropy": 1.916738434880972,
      "epoch": 0.3860442982492758,
      "grad_norm": 0.2900070548057556,
      "learning_rate": 3.8577452022313074e-05,
      "loss": 0.2213,
      "mean_token_accuracy": 0.9462229896336793,
      "num_tokens": 43160191.0,
      "step": 5780
    },
    {
      "entropy": 1.8875885412096978,
      "epoch": 0.38671219495904957,
      "grad_norm": 0.2704409062862396,
      "learning_rate": 3.852853090003897e-05,
      "loss": 0.2305,
      "mean_token_accuracy": 0.9444563537836075,
      "num_tokens": 43236556.0,
      "step": 5790
    },
    {
      "entropy": 1.913160105794668,
      "epoch": 0.3873800916688234,
      "grad_norm": 0.344618558883667,
      "learning_rate": 3.8479536493035634e-05,
      "loss": 0.2252,
      "mean_token_accuracy": 0.9445569045841694,
      "num_tokens": 43309911.0,
      "step": 5800
    },
    {
      "entropy": 1.8723728090524674,
      "epoch": 0.3880479883785973,
      "grad_norm": 0.2803967297077179,
      "learning_rate": 3.843046906769275e-05,
      "loss": 0.2165,
      "mean_token_accuracy": 0.9457434058189392,
      "num_tokens": 43384311.0,
      "step": 5810
    },
    {
      "entropy": 1.9039440669119359,
      "epoch": 0.38871588508837107,
      "grad_norm": 0.2550482451915741,
      "learning_rate": 3.838132889079704e-05,
      "loss": 0.2238,
      "mean_token_accuracy": 0.9465091783553362,
      "num_tokens": 43461111.0,
      "step": 5820
    },
    {
      "entropy": 1.9514834955334663,
      "epoch": 0.3893837817981449,
      "grad_norm": 0.3363208770751953,
      "learning_rate": 3.833211622953078e-05,
      "loss": 0.2201,
      "mean_token_accuracy": 0.9457925707101822,
      "num_tokens": 43537573.0,
      "step": 5830
    },
    {
      "entropy": 1.9034170739352703,
      "epoch": 0.3900516785079188,
      "grad_norm": 0.2836158573627472,
      "learning_rate": 3.828283135147036e-05,
      "loss": 0.2155,
      "mean_token_accuracy": 0.946396753564477,
      "num_tokens": 43611669.0,
      "step": 5840
    },
    {
      "entropy": 1.9066494800150395,
      "epoch": 0.39071957521769257,
      "grad_norm": 0.23912501335144043,
      "learning_rate": 3.82334745245848e-05,
      "loss": 0.229,
      "mean_token_accuracy": 0.9444659076631069,
      "num_tokens": 43687255.0,
      "step": 5850
    },
    {
      "entropy": 1.9668964512646199,
      "epoch": 0.3913874719274664,
      "grad_norm": 0.29459482431411743,
      "learning_rate": 3.8184046017234354e-05,
      "loss": 0.2378,
      "mean_token_accuracy": 0.9422699268907309,
      "num_tokens": 43759618.0,
      "step": 5860
    },
    {
      "entropy": 1.9583055019378661,
      "epoch": 0.3920553686372403,
      "grad_norm": 0.3198317587375641,
      "learning_rate": 3.8134546098168975e-05,
      "loss": 0.2327,
      "mean_token_accuracy": 0.9433270651847124,
      "num_tokens": 43832131.0,
      "step": 5870
    },
    {
      "entropy": 1.9342499651014804,
      "epoch": 0.39272326534701407,
      "grad_norm": 0.2366897016763687,
      "learning_rate": 3.8084975036526924e-05,
      "loss": 0.2379,
      "mean_token_accuracy": 0.9422232013195753,
      "num_tokens": 43908279.0,
      "step": 5880
    },
    {
      "entropy": 1.9105512894690038,
      "epoch": 0.3933911620567879,
      "grad_norm": 0.24744364619255066,
      "learning_rate": 3.8035333101833244e-05,
      "loss": 0.2064,
      "mean_token_accuracy": 0.9485102720558644,
      "num_tokens": 43984882.0,
      "step": 5890
    },
    {
      "entropy": 1.883164818584919,
      "epoch": 0.3940590587665618,
      "grad_norm": 0.2920915186405182,
      "learning_rate": 3.798562056399835e-05,
      "loss": 0.219,
      "mean_token_accuracy": 0.9483438994735479,
      "num_tokens": 44060062.0,
      "step": 5900
    },
    {
      "entropy": 1.9149408027529717,
      "epoch": 0.3947269554763356,
      "grad_norm": 0.23199617862701416,
      "learning_rate": 3.793583769331653e-05,
      "loss": 0.2279,
      "mean_token_accuracy": 0.9456027213484048,
      "num_tokens": 44136277.0,
      "step": 5910
    },
    {
      "entropy": 1.9140928737819194,
      "epoch": 0.3953948521861094,
      "grad_norm": 0.2765490412712097,
      "learning_rate": 3.788598476046448e-05,
      "loss": 0.2311,
      "mean_token_accuracy": 0.9434793140739203,
      "num_tokens": 44210326.0,
      "step": 5920
    },
    {
      "entropy": 1.9026202492415905,
      "epoch": 0.3960627488958833,
      "grad_norm": 0.37410467863082886,
      "learning_rate": 3.783606203649981e-05,
      "loss": 0.2249,
      "mean_token_accuracy": 0.9440909586846828,
      "num_tokens": 44286586.0,
      "step": 5930
    },
    {
      "entropy": 1.8660123124718666,
      "epoch": 0.3967306456056571,
      "grad_norm": 0.3082118630409241,
      "learning_rate": 3.778606979285965e-05,
      "loss": 0.2229,
      "mean_token_accuracy": 0.9430748876184225,
      "num_tokens": 44361043.0,
      "step": 5940
    },
    {
      "entropy": 1.8822263188660144,
      "epoch": 0.3973985423154309,
      "grad_norm": 0.32805633544921875,
      "learning_rate": 3.7736008301359064e-05,
      "loss": 0.2441,
      "mean_token_accuracy": 0.9426589887589216,
      "num_tokens": 44436727.0,
      "step": 5950
    },
    {
      "entropy": 1.9247616775333882,
      "epoch": 0.3980664390252048,
      "grad_norm": 0.35823896527290344,
      "learning_rate": 3.768587783418967e-05,
      "loss": 0.2398,
      "mean_token_accuracy": 0.9381592474877835,
      "num_tokens": 44510360.0,
      "step": 5960
    },
    {
      "entropy": 1.8901113018393516,
      "epoch": 0.3987343357349786,
      "grad_norm": 0.26516541838645935,
      "learning_rate": 3.763567866391807e-05,
      "loss": 0.2216,
      "mean_token_accuracy": 0.9455227475613356,
      "num_tokens": 44585669.0,
      "step": 5970
    },
    {
      "entropy": 1.926313439756632,
      "epoch": 0.3994022324447524,
      "grad_norm": 0.26676836609840393,
      "learning_rate": 3.758541106348446e-05,
      "loss": 0.2494,
      "mean_token_accuracy": 0.9367516294121743,
      "num_tokens": 44659978.0,
      "step": 5980
    },
    {
      "entropy": 1.917150054126978,
      "epoch": 0.4000701291545263,
      "grad_norm": 0.3763718903064728,
      "learning_rate": 3.753507530620107e-05,
      "loss": 0.2362,
      "mean_token_accuracy": 0.9436690431088209,
      "num_tokens": 44733512.0,
      "step": 5990
    },
    {
      "entropy": 1.8905444204807282,
      "epoch": 0.4007380258643001,
      "grad_norm": 0.3248978853225708,
      "learning_rate": 3.748467166575072e-05,
      "loss": 0.2101,
      "mean_token_accuracy": 0.9461406122893095,
      "num_tokens": 44809175.0,
      "step": 6000
    },
    {
      "epoch": 0.4007380258643001,
      "eval_entropy": 1.8695236456871032,
      "eval_loss": 0.24391567707061768,
      "eval_mean_token_accuracy": 0.944749633693695,
      "eval_num_tokens": 44809175.0,
      "eval_runtime": 552.2428,
      "eval_samples_per_second": 36.216,
      "eval_steps_per_second": 9.054,
      "step": 6000
    },
    {
      "entropy": 1.9028013505041599,
      "epoch": 0.4014059225740739,
      "grad_norm": 0.34846240282058716,
      "learning_rate": 3.743420041618534e-05,
      "loss": 0.2434,
      "mean_token_accuracy": 0.9411895230412484,
      "num_tokens": 44880134.0,
      "step": 6010
    },
    {
      "entropy": 1.8826014414429664,
      "epoch": 0.4020738192838478,
      "grad_norm": 0.27955442667007446,
      "learning_rate": 3.738366183192441e-05,
      "loss": 0.212,
      "mean_token_accuracy": 0.9476633965969086,
      "num_tokens": 44956579.0,
      "step": 6020
    },
    {
      "entropy": 1.9236787579953671,
      "epoch": 0.4027417159936216,
      "grad_norm": 0.2702292203903198,
      "learning_rate": 3.733305618775357e-05,
      "loss": 0.2275,
      "mean_token_accuracy": 0.9423356868326664,
      "num_tokens": 45031454.0,
      "step": 6030
    },
    {
      "entropy": 1.8541994221508502,
      "epoch": 0.4034096127033954,
      "grad_norm": 0.3068564832210541,
      "learning_rate": 3.728238375882305e-05,
      "loss": 0.2267,
      "mean_token_accuracy": 0.942199082672596,
      "num_tokens": 45102674.0,
      "step": 6040
    },
    {
      "entropy": 1.8715508930385112,
      "epoch": 0.4040775094131693,
      "grad_norm": 0.2965051829814911,
      "learning_rate": 3.72316448206462e-05,
      "loss": 0.2245,
      "mean_token_accuracy": 0.9468481604009866,
      "num_tokens": 45175986.0,
      "step": 6050
    },
    {
      "entropy": 1.8987006291747093,
      "epoch": 0.4047454061229431,
      "grad_norm": 0.33186593651771545,
      "learning_rate": 3.7180839649098005e-05,
      "loss": 0.2235,
      "mean_token_accuracy": 0.9451838321983814,
      "num_tokens": 45248123.0,
      "step": 6060
    },
    {
      "entropy": 1.9166165970265865,
      "epoch": 0.4054133028327169,
      "grad_norm": 0.31308844685554504,
      "learning_rate": 3.712996852041353e-05,
      "loss": 0.2279,
      "mean_token_accuracy": 0.9453062769025564,
      "num_tokens": 45321976.0,
      "step": 6070
    },
    {
      "entropy": 1.9069445885717868,
      "epoch": 0.4060811995424908,
      "grad_norm": 0.28432127833366394,
      "learning_rate": 3.707903171118652e-05,
      "loss": 0.2479,
      "mean_token_accuracy": 0.9419134773314,
      "num_tokens": 45396684.0,
      "step": 6080
    },
    {
      "entropy": 1.9531974382698536,
      "epoch": 0.4067490962522646,
      "grad_norm": 0.33367905020713806,
      "learning_rate": 3.7028029498367784e-05,
      "loss": 0.2515,
      "mean_token_accuracy": 0.9403557687997818,
      "num_tokens": 45471213.0,
      "step": 6090
    },
    {
      "entropy": 1.9115024484694003,
      "epoch": 0.4074169929620384,
      "grad_norm": 0.3217218518257141,
      "learning_rate": 3.697696215926375e-05,
      "loss": 0.2171,
      "mean_token_accuracy": 0.9454292073845864,
      "num_tokens": 45547231.0,
      "step": 6100
    },
    {
      "entropy": 1.8814896136522292,
      "epoch": 0.4080848896718123,
      "grad_norm": 0.29801493883132935,
      "learning_rate": 3.692582997153497e-05,
      "loss": 0.2352,
      "mean_token_accuracy": 0.9412771929055452,
      "num_tokens": 45622938.0,
      "step": 6110
    },
    {
      "entropy": 1.8915587805211544,
      "epoch": 0.4087527863815861,
      "grad_norm": 0.2785051167011261,
      "learning_rate": 3.6874633213194565e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9448517318814993,
      "num_tokens": 45698068.0,
      "step": 6120
    },
    {
      "entropy": 1.8883536547422408,
      "epoch": 0.4094206830913599,
      "grad_norm": 0.31130892038345337,
      "learning_rate": 3.6823372162606745e-05,
      "loss": 0.2189,
      "mean_token_accuracy": 0.9458511333912611,
      "num_tokens": 45774861.0,
      "step": 6130
    },
    {
      "entropy": 1.928010568022728,
      "epoch": 0.4100885798011338,
      "grad_norm": 0.3430582880973816,
      "learning_rate": 3.6772047098485274e-05,
      "loss": 0.2323,
      "mean_token_accuracy": 0.9417624350637197,
      "num_tokens": 45848688.0,
      "step": 6140
    },
    {
      "entropy": 1.9242531947791577,
      "epoch": 0.4107564765109076,
      "grad_norm": 0.308286190032959,
      "learning_rate": 3.6720658299891995e-05,
      "loss": 0.2302,
      "mean_token_accuracy": 0.9458325952291489,
      "num_tokens": 45923913.0,
      "step": 6150
    },
    {
      "entropy": 1.9068501241505147,
      "epoch": 0.4114243732206814,
      "grad_norm": 0.28455981612205505,
      "learning_rate": 3.666920604623525e-05,
      "loss": 0.236,
      "mean_token_accuracy": 0.9448170073330402,
      "num_tokens": 45999355.0,
      "step": 6160
    },
    {
      "entropy": 1.9273151695728301,
      "epoch": 0.4120922699304553,
      "grad_norm": 0.2953311502933502,
      "learning_rate": 3.6617690617268416e-05,
      "loss": 0.2493,
      "mean_token_accuracy": 0.9371218640357256,
      "num_tokens": 46074070.0,
      "step": 6170
    },
    {
      "entropy": 1.923949245363474,
      "epoch": 0.4127601666402291,
      "grad_norm": 0.32243552803993225,
      "learning_rate": 3.656611229308837e-05,
      "loss": 0.2259,
      "mean_token_accuracy": 0.9464456308633089,
      "num_tokens": 46149589.0,
      "step": 6180
    },
    {
      "entropy": 1.9378270879387856,
      "epoch": 0.41342806335000293,
      "grad_norm": 0.2911277711391449,
      "learning_rate": 3.6514471354133936e-05,
      "loss": 0.2309,
      "mean_token_accuracy": 0.9456627234816551,
      "num_tokens": 46222113.0,
      "step": 6190
    },
    {
      "entropy": 1.9103655092418195,
      "epoch": 0.4140959600597768,
      "grad_norm": 0.2618463635444641,
      "learning_rate": 3.646276808118439e-05,
      "loss": 0.219,
      "mean_token_accuracy": 0.9453224431723356,
      "num_tokens": 46297486.0,
      "step": 6200
    },
    {
      "entropy": 1.937165941298008,
      "epoch": 0.4147638567695506,
      "grad_norm": 0.34674936532974243,
      "learning_rate": 3.641100275535795e-05,
      "loss": 0.2323,
      "mean_token_accuracy": 0.9436026971787215,
      "num_tokens": 46372503.0,
      "step": 6210
    },
    {
      "entropy": 1.8939755447208881,
      "epoch": 0.41543175347932443,
      "grad_norm": 0.28311529755592346,
      "learning_rate": 3.63591756581102e-05,
      "loss": 0.2312,
      "mean_token_accuracy": 0.9440379265695811,
      "num_tokens": 46446897.0,
      "step": 6220
    },
    {
      "entropy": 1.9137402102351189,
      "epoch": 0.4160996501890983,
      "grad_norm": 0.3030659556388855,
      "learning_rate": 3.630728707123259e-05,
      "loss": 0.2235,
      "mean_token_accuracy": 0.943624097108841,
      "num_tokens": 46523563.0,
      "step": 6230
    },
    {
      "entropy": 1.9250687412917613,
      "epoch": 0.4167675468988721,
      "grad_norm": 0.3375629782676697,
      "learning_rate": 3.62553372768509e-05,
      "loss": 0.2353,
      "mean_token_accuracy": 0.9426155641674996,
      "num_tokens": 46597863.0,
      "step": 6240
    },
    {
      "entropy": 1.9081721894443036,
      "epoch": 0.41743544360864593,
      "grad_norm": 0.3522295355796814,
      "learning_rate": 3.620332655742371e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9423638172447681,
      "num_tokens": 46674309.0,
      "step": 6250
    },
    {
      "entropy": 1.9130800761282445,
      "epoch": 0.4181033403184198,
      "grad_norm": 0.3143600523471832,
      "learning_rate": 3.615125519574084e-05,
      "loss": 0.2353,
      "mean_token_accuracy": 0.9443993449211121,
      "num_tokens": 46749360.0,
      "step": 6260
    },
    {
      "entropy": 1.9183715388178826,
      "epoch": 0.4187712370281936,
      "grad_norm": 0.3188576102256775,
      "learning_rate": 3.6099123474921864e-05,
      "loss": 0.2217,
      "mean_token_accuracy": 0.9451267641037703,
      "num_tokens": 46822602.0,
      "step": 6270
    },
    {
      "entropy": 1.9148419328033923,
      "epoch": 0.41943913373796743,
      "grad_norm": 0.3026546239852905,
      "learning_rate": 3.6046931678414485e-05,
      "loss": 0.2263,
      "mean_token_accuracy": 0.9442095790058375,
      "num_tokens": 46895405.0,
      "step": 6280
    },
    {
      "entropy": 1.9719792023301124,
      "epoch": 0.4201070304477413,
      "grad_norm": 0.31735241413116455,
      "learning_rate": 3.59946800899931e-05,
      "loss": 0.2405,
      "mean_token_accuracy": 0.941047714278102,
      "num_tokens": 46973170.0,
      "step": 6290
    },
    {
      "entropy": 1.9277971625328063,
      "epoch": 0.4207749271575151,
      "grad_norm": 0.30713051557540894,
      "learning_rate": 3.594236899375719e-05,
      "loss": 0.2209,
      "mean_token_accuracy": 0.9454218167811632,
      "num_tokens": 47046074.0,
      "step": 6300
    },
    {
      "entropy": 1.9037809073925018,
      "epoch": 0.42144282386728893,
      "grad_norm": 0.3050428330898285,
      "learning_rate": 3.588999867412975e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9419395077973605,
      "num_tokens": 47120609.0,
      "step": 6310
    },
    {
      "entropy": 1.931861577183008,
      "epoch": 0.4221107205770628,
      "grad_norm": 0.283743292093277,
      "learning_rate": 3.5837569415855825e-05,
      "loss": 0.226,
      "mean_token_accuracy": 0.9440953090786934,
      "num_tokens": 47194666.0,
      "step": 6320
    },
    {
      "entropy": 1.9015536345541477,
      "epoch": 0.4227786172868366,
      "grad_norm": 0.2638198435306549,
      "learning_rate": 3.578508150400092e-05,
      "loss": 0.2221,
      "mean_token_accuracy": 0.9482598397880793,
      "num_tokens": 47271163.0,
      "step": 6330
    },
    {
      "entropy": 1.8965262487530707,
      "epoch": 0.42344651399661043,
      "grad_norm": 0.28280046582221985,
      "learning_rate": 3.57325352239494e-05,
      "loss": 0.214,
      "mean_token_accuracy": 0.9467114940285682,
      "num_tokens": 47347255.0,
      "step": 6340
    },
    {
      "entropy": 1.8840107902884484,
      "epoch": 0.4241144107063843,
      "grad_norm": 0.33287447690963745,
      "learning_rate": 3.5679930861403036e-05,
      "loss": 0.2053,
      "mean_token_accuracy": 0.9492501009255647,
      "num_tokens": 47421584.0,
      "step": 6350
    },
    {
      "entropy": 1.8743778616189957,
      "epoch": 0.4247823074161581,
      "grad_norm": 0.26861509680747986,
      "learning_rate": 3.562726870237939e-05,
      "loss": 0.2384,
      "mean_token_accuracy": 0.9431404061615467,
      "num_tokens": 47497066.0,
      "step": 6360
    },
    {
      "entropy": 1.8955380499362946,
      "epoch": 0.42545020412593193,
      "grad_norm": 0.2798490822315216,
      "learning_rate": 3.557454903321026e-05,
      "loss": 0.2278,
      "mean_token_accuracy": 0.9434572167694568,
      "num_tokens": 47574882.0,
      "step": 6370
    },
    {
      "entropy": 1.9196617253124715,
      "epoch": 0.4261181008357058,
      "grad_norm": 0.3466589152812958,
      "learning_rate": 3.5521772140540126e-05,
      "loss": 0.237,
      "mean_token_accuracy": 0.9427095215767622,
      "num_tokens": 47649258.0,
      "step": 6380
    },
    {
      "entropy": 1.8840134643018245,
      "epoch": 0.4267859975454796,
      "grad_norm": 0.3197396993637085,
      "learning_rate": 3.546893831132464e-05,
      "loss": 0.2144,
      "mean_token_accuracy": 0.9482835672795773,
      "num_tokens": 47722587.0,
      "step": 6390
    },
    {
      "entropy": 1.8524423643946648,
      "epoch": 0.42745389425525343,
      "grad_norm": 0.36200129985809326,
      "learning_rate": 3.541604783282899e-05,
      "loss": 0.2338,
      "mean_token_accuracy": 0.9426054045557976,
      "num_tokens": 47796769.0,
      "step": 6400
    },
    {
      "entropy": 1.855437734723091,
      "epoch": 0.4281217909650273,
      "grad_norm": 0.3258037567138672,
      "learning_rate": 3.536310099262638e-05,
      "loss": 0.2188,
      "mean_token_accuracy": 0.9453244972974062,
      "num_tokens": 47871403.0,
      "step": 6410
    },
    {
      "entropy": 1.8843614600598813,
      "epoch": 0.4287896876748011,
      "grad_norm": 0.3072390556335449,
      "learning_rate": 3.531009807859648e-05,
      "loss": 0.2436,
      "mean_token_accuracy": 0.9394047453999519,
      "num_tokens": 47944732.0,
      "step": 6420
    },
    {
      "entropy": 1.8818963214755058,
      "epoch": 0.42945758438457493,
      "grad_norm": 0.2918820381164551,
      "learning_rate": 3.525703937892381e-05,
      "loss": 0.2245,
      "mean_token_accuracy": 0.9446279395371675,
      "num_tokens": 48020628.0,
      "step": 6430
    },
    {
      "entropy": 1.8776385493576526,
      "epoch": 0.4301254810943488,
      "grad_norm": 0.30202630162239075,
      "learning_rate": 3.520392518209624e-05,
      "loss": 0.2364,
      "mean_token_accuracy": 0.9430569637566805,
      "num_tokens": 48095543.0,
      "step": 6440
    },
    {
      "entropy": 1.919315802305937,
      "epoch": 0.4307933778041226,
      "grad_norm": 0.33853879570961,
      "learning_rate": 3.515075577690337e-05,
      "loss": 0.2138,
      "mean_token_accuracy": 0.9462950874119997,
      "num_tokens": 48168243.0,
      "step": 6450
    },
    {
      "entropy": 1.8829366579651832,
      "epoch": 0.43146127451389643,
      "grad_norm": 0.274326890707016,
      "learning_rate": 3.509753145243495e-05,
      "loss": 0.227,
      "mean_token_accuracy": 0.944806321337819,
      "num_tokens": 48243203.0,
      "step": 6460
    },
    {
      "entropy": 1.8829902492463588,
      "epoch": 0.4321291712236703,
      "grad_norm": 0.2924010753631592,
      "learning_rate": 3.504425249807938e-05,
      "loss": 0.2403,
      "mean_token_accuracy": 0.9419176794588566,
      "num_tokens": 48318494.0,
      "step": 6470
    },
    {
      "entropy": 1.9198862127959728,
      "epoch": 0.4327970679334441,
      "grad_norm": 0.316094309091568,
      "learning_rate": 3.4990919203522084e-05,
      "loss": 0.2325,
      "mean_token_accuracy": 0.9449028253555298,
      "num_tokens": 48391396.0,
      "step": 6480
    },
    {
      "entropy": 1.9263180814683438,
      "epoch": 0.43346496464321793,
      "grad_norm": 0.27902543544769287,
      "learning_rate": 3.4937531858743896e-05,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9459644220769405,
      "num_tokens": 48465129.0,
      "step": 6490
    },
    {
      "entropy": 1.9368058815598488,
      "epoch": 0.4341328613529918,
      "grad_norm": 0.31840354204177856,
      "learning_rate": 3.4884090754019575e-05,
      "loss": 0.2297,
      "mean_token_accuracy": 0.9452568523585796,
      "num_tokens": 48538972.0,
      "step": 6500
    },
    {
      "entropy": 1.9461675800383091,
      "epoch": 0.4348007580627656,
      "grad_norm": 0.346287339925766,
      "learning_rate": 3.483059617991616e-05,
      "loss": 0.2176,
      "mean_token_accuracy": 0.9455453310161829,
      "num_tokens": 48612056.0,
      "step": 6510
    },
    {
      "entropy": 1.9228738784790038,
      "epoch": 0.43546865477253943,
      "grad_norm": 0.2979698181152344,
      "learning_rate": 3.477704842729142e-05,
      "loss": 0.2376,
      "mean_token_accuracy": 0.9440739706158638,
      "num_tokens": 48687725.0,
      "step": 6520
    },
    {
      "entropy": 1.9221594251692296,
      "epoch": 0.4361365514823133,
      "grad_norm": 0.26502421498298645,
      "learning_rate": 3.472344778729227e-05,
      "loss": 0.2202,
      "mean_token_accuracy": 0.9462688464671374,
      "num_tokens": 48762673.0,
      "step": 6530
    },
    {
      "entropy": 1.9363283306360244,
      "epoch": 0.4368044481920871,
      "grad_norm": 0.37736600637435913,
      "learning_rate": 3.4669794551353154e-05,
      "loss": 0.2322,
      "mean_token_accuracy": 0.9412756383419036,
      "num_tokens": 48835051.0,
      "step": 6540
    },
    {
      "entropy": 1.9189944982528686,
      "epoch": 0.43747234490186093,
      "grad_norm": 0.24977274239063263,
      "learning_rate": 3.461608901119452e-05,
      "loss": 0.2281,
      "mean_token_accuracy": 0.9436902448534965,
      "num_tokens": 48909132.0,
      "step": 6550
    },
    {
      "entropy": 1.8862173050642013,
      "epoch": 0.4381402416116348,
      "grad_norm": 0.32071566581726074,
      "learning_rate": 3.4562331458821164e-05,
      "loss": 0.2211,
      "mean_token_accuracy": 0.9463409461081028,
      "num_tokens": 48984980.0,
      "step": 6560
    },
    {
      "entropy": 1.895054043084383,
      "epoch": 0.4388081383214086,
      "grad_norm": 0.2953815460205078,
      "learning_rate": 3.4508522186520724e-05,
      "loss": 0.2198,
      "mean_token_accuracy": 0.9452447794377804,
      "num_tokens": 49060013.0,
      "step": 6570
    },
    {
      "entropy": 1.8868306331336497,
      "epoch": 0.43947603503118243,
      "grad_norm": 0.34867608547210693,
      "learning_rate": 3.445466148686203e-05,
      "loss": 0.2394,
      "mean_token_accuracy": 0.9442655451595783,
      "num_tokens": 49135941.0,
      "step": 6580
    },
    {
      "entropy": 1.9204572565853595,
      "epoch": 0.4401439317409563,
      "grad_norm": 0.3229929804801941,
      "learning_rate": 3.44007496526935e-05,
      "loss": 0.2332,
      "mean_token_accuracy": 0.9417555730789899,
      "num_tokens": 49209215.0,
      "step": 6590
    },
    {
      "entropy": 1.9017700776457787,
      "epoch": 0.4408118284507301,
      "grad_norm": 0.318572074174881,
      "learning_rate": 3.434678697714162e-05,
      "loss": 0.2349,
      "mean_token_accuracy": 0.9422289121896028,
      "num_tokens": 49285006.0,
      "step": 6600
    },
    {
      "entropy": 1.894771010428667,
      "epoch": 0.44147972516050393,
      "grad_norm": 0.26797112822532654,
      "learning_rate": 3.429277375360928e-05,
      "loss": 0.2439,
      "mean_token_accuracy": 0.941288698464632,
      "num_tokens": 49358810.0,
      "step": 6610
    },
    {
      "entropy": 1.902489211410284,
      "epoch": 0.4421476218702778,
      "grad_norm": 0.2798771858215332,
      "learning_rate": 3.4238710275774245e-05,
      "loss": 0.2208,
      "mean_token_accuracy": 0.9456051144748926,
      "num_tokens": 49433063.0,
      "step": 6620
    },
    {
      "entropy": 1.919239255040884,
      "epoch": 0.4428155185800516,
      "grad_norm": 0.32063212990760803,
      "learning_rate": 3.418459683758746e-05,
      "loss": 0.2277,
      "mean_token_accuracy": 0.9427162941545248,
      "num_tokens": 49507216.0,
      "step": 6630
    },
    {
      "entropy": 1.8541309811174869,
      "epoch": 0.44348341528982543,
      "grad_norm": 0.26682358980178833,
      "learning_rate": 3.4130433733271555e-05,
      "loss": 0.2217,
      "mean_token_accuracy": 0.9454412367194891,
      "num_tokens": 49584691.0,
      "step": 6640
    },
    {
      "entropy": 1.850600217282772,
      "epoch": 0.4441513119995993,
      "grad_norm": 0.3180466592311859,
      "learning_rate": 3.4076221257319206e-05,
      "loss": 0.2172,
      "mean_token_accuracy": 0.9476006656885148,
      "num_tokens": 49662014.0,
      "step": 6650
    },
    {
      "entropy": 1.8940122798085213,
      "epoch": 0.4448192087093731,
      "grad_norm": 0.30723103880882263,
      "learning_rate": 3.4021959704491493e-05,
      "loss": 0.234,
      "mean_token_accuracy": 0.9394343450665474,
      "num_tokens": 49733624.0,
      "step": 6660
    },
    {
      "entropy": 1.880489770323038,
      "epoch": 0.44548710541914693,
      "grad_norm": 0.30570298433303833,
      "learning_rate": 3.3967649369816364e-05,
      "loss": 0.2321,
      "mean_token_accuracy": 0.945229908823967,
      "num_tokens": 49806263.0,
      "step": 6670
    },
    {
      "entropy": 1.8641036197543144,
      "epoch": 0.4461550021289208,
      "grad_norm": 0.35547709465026855,
      "learning_rate": 3.3913290548587e-05,
      "loss": 0.2219,
      "mean_token_accuracy": 0.9454311702400446,
      "num_tokens": 49878510.0,
      "step": 6680
    },
    {
      "entropy": 1.862290508300066,
      "epoch": 0.4468228988386946,
      "grad_norm": 0.34313470125198364,
      "learning_rate": 3.38588835363602e-05,
      "loss": 0.2443,
      "mean_token_accuracy": 0.9395472008734942,
      "num_tokens": 49950671.0,
      "step": 6690
    },
    {
      "entropy": 1.892933504283428,
      "epoch": 0.44749079554846843,
      "grad_norm": 0.26510775089263916,
      "learning_rate": 3.380442862895478e-05,
      "loss": 0.2398,
      "mean_token_accuracy": 0.9421889808028936,
      "num_tokens": 50027718.0,
      "step": 6700
    },
    {
      "entropy": 1.8904846854507924,
      "epoch": 0.4481586922582423,
      "grad_norm": 0.2967054843902588,
      "learning_rate": 3.374992612244999e-05,
      "loss": 0.218,
      "mean_token_accuracy": 0.9464267261326313,
      "num_tokens": 50103437.0,
      "step": 6710
    },
    {
      "entropy": 1.8867049098014832,
      "epoch": 0.4488265889680161,
      "grad_norm": 0.32129737734794617,
      "learning_rate": 3.3695376313183856e-05,
      "loss": 0.2191,
      "mean_token_accuracy": 0.9464637506753206,
      "num_tokens": 50179359.0,
      "step": 6720
    },
    {
      "entropy": 1.8691880777478218,
      "epoch": 0.44949448567778993,
      "grad_norm": 0.2949567437171936,
      "learning_rate": 3.364077949775162e-05,
      "loss": 0.2226,
      "mean_token_accuracy": 0.9465657882392406,
      "num_tokens": 50254559.0,
      "step": 6730
    },
    {
      "entropy": 1.8490160331130028,
      "epoch": 0.4501623823875638,
      "grad_norm": 0.33848509192466736,
      "learning_rate": 3.3586135973004096e-05,
      "loss": 0.224,
      "mean_token_accuracy": 0.9438275959342718,
      "num_tokens": 50328081.0,
      "step": 6740
    },
    {
      "entropy": 1.8460321567952633,
      "epoch": 0.4508302790973376,
      "grad_norm": 0.3190463185310364,
      "learning_rate": 3.353144603604606e-05,
      "loss": 0.2218,
      "mean_token_accuracy": 0.9459980715066194,
      "num_tokens": 50403646.0,
      "step": 6750
    },
    {
      "entropy": 1.814967006444931,
      "epoch": 0.45149817580711143,
      "grad_norm": 0.2492627054452896,
      "learning_rate": 3.347670998423463e-05,
      "loss": 0.2251,
      "mean_token_accuracy": 0.94491036683321,
      "num_tokens": 50480554.0,
      "step": 6760
    },
    {
      "entropy": 1.8196490325033665,
      "epoch": 0.4521660725168853,
      "grad_norm": 0.34555625915527344,
      "learning_rate": 3.342192811517767e-05,
      "loss": 0.2252,
      "mean_token_accuracy": 0.9440069407224655,
      "num_tokens": 50555603.0,
      "step": 6770
    },
    {
      "entropy": 1.8382070384919644,
      "epoch": 0.4528339692266591,
      "grad_norm": 0.27213919162750244,
      "learning_rate": 3.3367100726732166e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.9474160704761744,
      "num_tokens": 50628568.0,
      "step": 6780
    },
    {
      "entropy": 1.8790905468165875,
      "epoch": 0.45350186593643294,
      "grad_norm": 0.2994993031024933,
      "learning_rate": 3.3312228117002564e-05,
      "loss": 0.234,
      "mean_token_accuracy": 0.9440126392990351,
      "num_tokens": 50700871.0,
      "step": 6790
    },
    {
      "entropy": 1.855829305201769,
      "epoch": 0.4541697626462068,
      "grad_norm": 0.3038565218448639,
      "learning_rate": 3.3257310584339236e-05,
      "loss": 0.2231,
      "mean_token_accuracy": 0.9469724778085947,
      "num_tokens": 50775395.0,
      "step": 6800
    },
    {
      "entropy": 1.9227498710155486,
      "epoch": 0.4548376593559806,
      "grad_norm": 0.46543049812316895,
      "learning_rate": 3.320234842733675e-05,
      "loss": 0.2275,
      "mean_token_accuracy": 0.9463567581027746,
      "num_tokens": 50850265.0,
      "step": 6810
    },
    {
      "entropy": 1.8953380823135375,
      "epoch": 0.45550555606575444,
      "grad_norm": 0.2792391777038574,
      "learning_rate": 3.3147341944832367e-05,
      "loss": 0.2243,
      "mean_token_accuracy": 0.9461825627833604,
      "num_tokens": 50923483.0,
      "step": 6820
    },
    {
      "entropy": 1.8902866654098034,
      "epoch": 0.4561734527755283,
      "grad_norm": 0.2938770651817322,
      "learning_rate": 3.30922914359043e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9462299484759569,
      "num_tokens": 50999320.0,
      "step": 6830
    },
    {
      "entropy": 1.8849791131913662,
      "epoch": 0.4568413494853021,
      "grad_norm": 0.3157196342945099,
      "learning_rate": 3.303719719987015e-05,
      "loss": 0.2137,
      "mean_token_accuracy": 0.947416216135025,
      "num_tokens": 51074433.0,
      "step": 6840
    },
    {
      "entropy": 1.892227377742529,
      "epoch": 0.45750924619507594,
      "grad_norm": 0.350421667098999,
      "learning_rate": 3.2982059536285295e-05,
      "loss": 0.2217,
      "mean_token_accuracy": 0.947562987357378,
      "num_tokens": 51150388.0,
      "step": 6850
    },
    {
      "entropy": 1.9007094338536263,
      "epoch": 0.4581771429048498,
      "grad_norm": 0.29797685146331787,
      "learning_rate": 3.292687874494121e-05,
      "loss": 0.2346,
      "mean_token_accuracy": 0.9450659941881895,
      "num_tokens": 51223896.0,
      "step": 6860
    },
    {
      "entropy": 1.9213022746145725,
      "epoch": 0.4588450396146236,
      "grad_norm": 0.270723819732666,
      "learning_rate": 3.287165512586386e-05,
      "loss": 0.2445,
      "mean_token_accuracy": 0.9438197415322065,
      "num_tokens": 51298201.0,
      "step": 6870
    },
    {
      "entropy": 1.8529110699892044,
      "epoch": 0.45951293632439744,
      "grad_norm": 0.3196532130241394,
      "learning_rate": 3.2816388979312084e-05,
      "loss": 0.2224,
      "mean_token_accuracy": 0.9463130470365286,
      "num_tokens": 51372610.0,
      "step": 6880
    },
    {
      "entropy": 1.8407471835613252,
      "epoch": 0.4601808330341713,
      "grad_norm": 0.2640169858932495,
      "learning_rate": 3.276108060577595e-05,
      "loss": 0.2247,
      "mean_token_accuracy": 0.9447287756949663,
      "num_tokens": 51447201.0,
      "step": 6890
    },
    {
      "entropy": 1.8109105303883553,
      "epoch": 0.4608487297439451,
      "grad_norm": 0.2655593454837799,
      "learning_rate": 3.270573030597511e-05,
      "loss": 0.2065,
      "mean_token_accuracy": 0.9496787924319505,
      "num_tokens": 51522173.0,
      "step": 6900
    },
    {
      "entropy": 1.8473745577037335,
      "epoch": 0.46151662645371894,
      "grad_norm": 0.30133378505706787,
      "learning_rate": 3.265033838085716e-05,
      "loss": 0.2398,
      "mean_token_accuracy": 0.9436753828078508,
      "num_tokens": 51594639.0,
      "step": 6910
    },
    {
      "entropy": 1.858110285550356,
      "epoch": 0.4621845231634928,
      "grad_norm": 0.36376380920410156,
      "learning_rate": 3.259490513159608e-05,
      "loss": 0.2367,
      "mean_token_accuracy": 0.9416541632264852,
      "num_tokens": 51669740.0,
      "step": 6920
    },
    {
      "entropy": 1.884439592808485,
      "epoch": 0.4628524198732666,
      "grad_norm": 0.3309226334095001,
      "learning_rate": 3.2539430859590466e-05,
      "loss": 0.2174,
      "mean_token_accuracy": 0.9472479324787855,
      "num_tokens": 51742364.0,
      "step": 6930
    },
    {
      "entropy": 1.8767762765288354,
      "epoch": 0.46352031658304044,
      "grad_norm": 0.37698131799697876,
      "learning_rate": 3.2483915866461976e-05,
      "loss": 0.2174,
      "mean_token_accuracy": 0.9452353831380605,
      "num_tokens": 51818060.0,
      "step": 6940
    },
    {
      "entropy": 1.9130305364727973,
      "epoch": 0.4641882132928143,
      "grad_norm": 0.28122803568840027,
      "learning_rate": 3.242836045405372e-05,
      "loss": 0.2415,
      "mean_token_accuracy": 0.942621236294508,
      "num_tokens": 51890991.0,
      "step": 6950
    },
    {
      "entropy": 1.921229613572359,
      "epoch": 0.4648561100025881,
      "grad_norm": 0.2826037108898163,
      "learning_rate": 3.237276492442852e-05,
      "loss": 0.2262,
      "mean_token_accuracy": 0.9483342789113521,
      "num_tokens": 51966020.0,
      "step": 6960
    },
    {
      "entropy": 1.8836260102689266,
      "epoch": 0.46552400671236194,
      "grad_norm": 0.30362722277641296,
      "learning_rate": 3.231712957986735e-05,
      "loss": 0.2411,
      "mean_token_accuracy": 0.9425248399376869,
      "num_tokens": 52041765.0,
      "step": 6970
    },
    {
      "entropy": 1.8544265240430833,
      "epoch": 0.4661919034221358,
      "grad_norm": 0.35201749205589294,
      "learning_rate": 3.226145472286766e-05,
      "loss": 0.1992,
      "mean_token_accuracy": 0.9497203458100557,
      "num_tokens": 52117393.0,
      "step": 6980
    },
    {
      "entropy": 1.8515958763659,
      "epoch": 0.4668598001319096,
      "grad_norm": 0.3822620213031769,
      "learning_rate": 3.2205740656141704e-05,
      "loss": 0.2357,
      "mean_token_accuracy": 0.9414434704929591,
      "num_tokens": 52189668.0,
      "step": 6990
    },
    {
      "entropy": 1.8938551887869834,
      "epoch": 0.46752769684168344,
      "grad_norm": 0.36982882022857666,
      "learning_rate": 3.214998768261498e-05,
      "loss": 0.2308,
      "mean_token_accuracy": 0.9430800579488278,
      "num_tokens": 52264870.0,
      "step": 7000
    },
    {
      "epoch": 0.46752769684168344,
      "eval_entropy": 1.884410347867012,
      "eval_loss": 0.24157805740833282,
      "eval_mean_token_accuracy": 0.9451566737055779,
      "eval_num_tokens": 52264870.0,
      "eval_runtime": 553.2883,
      "eval_samples_per_second": 36.148,
      "eval_steps_per_second": 9.037,
      "step": 7000
    },
    {
      "entropy": 1.8968149848282336,
      "epoch": 0.4681955935514573,
      "grad_norm": 0.33856382966041565,
      "learning_rate": 3.2094196105424476e-05,
      "loss": 0.2344,
      "mean_token_accuracy": 0.9420758977532386,
      "num_tokens": 52336884.0,
      "step": 7010
    },
    {
      "entropy": 1.8971335098147393,
      "epoch": 0.4688634902612311,
      "grad_norm": 0.3559689521789551,
      "learning_rate": 3.203836622791711e-05,
      "loss": 0.2327,
      "mean_token_accuracy": 0.9416660256683826,
      "num_tokens": 52408974.0,
      "step": 7020
    },
    {
      "entropy": 1.8569822050631046,
      "epoch": 0.46953138697100494,
      "grad_norm": 0.2902000844478607,
      "learning_rate": 3.198249835364804e-05,
      "loss": 0.2201,
      "mean_token_accuracy": 0.9460431203246117,
      "num_tokens": 52482004.0,
      "step": 7030
    },
    {
      "entropy": 1.8399673692882061,
      "epoch": 0.4701992836807788,
      "grad_norm": 0.30542469024658203,
      "learning_rate": 3.1926592786378997e-05,
      "loss": 0.2337,
      "mean_token_accuracy": 0.9438430979847908,
      "num_tokens": 52558037.0,
      "step": 7040
    },
    {
      "entropy": 1.8313255891203881,
      "epoch": 0.4708671803905526,
      "grad_norm": 0.2980274558067322,
      "learning_rate": 3.187064983007666e-05,
      "loss": 0.2273,
      "mean_token_accuracy": 0.9442933153361082,
      "num_tokens": 52632581.0,
      "step": 7050
    },
    {
      "entropy": 1.8898407019674779,
      "epoch": 0.47153507710032644,
      "grad_norm": 0.29755520820617676,
      "learning_rate": 3.181466978891102e-05,
      "loss": 0.2392,
      "mean_token_accuracy": 0.9428853433579206,
      "num_tokens": 52704292.0,
      "step": 7060
    },
    {
      "entropy": 1.9019944280385972,
      "epoch": 0.4722029738101003,
      "grad_norm": 0.30816546082496643,
      "learning_rate": 3.175865296725367e-05,
      "loss": 0.2394,
      "mean_token_accuracy": 0.9419173806905746,
      "num_tokens": 52778940.0,
      "step": 7070
    },
    {
      "entropy": 1.9086784988641738,
      "epoch": 0.4728708705198741,
      "grad_norm": 0.2939527928829193,
      "learning_rate": 3.170259966967622e-05,
      "loss": 0.2224,
      "mean_token_accuracy": 0.9432874158024788,
      "num_tokens": 52852578.0,
      "step": 7080
    },
    {
      "entropy": 1.8858931235969068,
      "epoch": 0.47353876722964794,
      "grad_norm": 0.2692057490348816,
      "learning_rate": 3.164651020094857e-05,
      "loss": 0.2296,
      "mean_token_accuracy": 0.943479648604989,
      "num_tokens": 52928111.0,
      "step": 7090
    },
    {
      "entropy": 1.9161983594298362,
      "epoch": 0.4742066639394218,
      "grad_norm": 0.3036022186279297,
      "learning_rate": 3.159038486603733e-05,
      "loss": 0.2191,
      "mean_token_accuracy": 0.9462693601846695,
      "num_tokens": 53004598.0,
      "step": 7100
    },
    {
      "entropy": 1.905858301371336,
      "epoch": 0.4748745606491956,
      "grad_norm": 0.27847298979759216,
      "learning_rate": 3.153422397010405e-05,
      "loss": 0.2182,
      "mean_token_accuracy": 0.9456410579383373,
      "num_tokens": 53080309.0,
      "step": 7110
    },
    {
      "entropy": 1.8999128639698029,
      "epoch": 0.47554245735896944,
      "grad_norm": 0.2785731554031372,
      "learning_rate": 3.147802781850371e-05,
      "loss": 0.224,
      "mean_token_accuracy": 0.947356840595603,
      "num_tokens": 53157053.0,
      "step": 7120
    },
    {
      "entropy": 1.8780075281858444,
      "epoch": 0.4762103540687433,
      "grad_norm": 0.2808716595172882,
      "learning_rate": 3.1421796716782936e-05,
      "loss": 0.2152,
      "mean_token_accuracy": 0.949623067677021,
      "num_tokens": 53231403.0,
      "step": 7130
    },
    {
      "entropy": 1.9050028070807457,
      "epoch": 0.4768782507785171,
      "grad_norm": 0.3088010251522064,
      "learning_rate": 3.1365530970678384e-05,
      "loss": 0.2276,
      "mean_token_accuracy": 0.9450955573469401,
      "num_tokens": 53304287.0,
      "step": 7140
    },
    {
      "entropy": 1.9440035305917263,
      "epoch": 0.47754614748829094,
      "grad_norm": 0.3013440668582916,
      "learning_rate": 3.13092308861151e-05,
      "loss": 0.2232,
      "mean_token_accuracy": 0.9448355570435524,
      "num_tokens": 53377127.0,
      "step": 7150
    },
    {
      "entropy": 1.903311026096344,
      "epoch": 0.4782140441980648,
      "grad_norm": 0.31735163927078247,
      "learning_rate": 3.125289676920479e-05,
      "loss": 0.225,
      "mean_token_accuracy": 0.9447057921439409,
      "num_tokens": 53452268.0,
      "step": 7160
    },
    {
      "entropy": 1.867805613577366,
      "epoch": 0.4788819409078386,
      "grad_norm": 0.29131561517715454,
      "learning_rate": 3.119652892624426e-05,
      "loss": 0.233,
      "mean_token_accuracy": 0.9428677547723054,
      "num_tokens": 53522599.0,
      "step": 7170
    },
    {
      "entropy": 1.857548114657402,
      "epoch": 0.47954983761761244,
      "grad_norm": 0.2810252904891968,
      "learning_rate": 3.114012766371363e-05,
      "loss": 0.2249,
      "mean_token_accuracy": 0.9437949914485216,
      "num_tokens": 53596596.0,
      "step": 7180
    },
    {
      "entropy": 1.8477457262575627,
      "epoch": 0.4802177343273863,
      "grad_norm": 0.2858681082725525,
      "learning_rate": 3.108369328827476e-05,
      "loss": 0.2273,
      "mean_token_accuracy": 0.9447164285928011,
      "num_tokens": 53671157.0,
      "step": 7190
    },
    {
      "entropy": 1.8654329650104047,
      "epoch": 0.4808856310371601,
      "grad_norm": 0.30987459421157837,
      "learning_rate": 3.1027226106769544e-05,
      "loss": 0.2179,
      "mean_token_accuracy": 0.9447156686335803,
      "num_tokens": 53746197.0,
      "step": 7200
    },
    {
      "entropy": 1.8650679022073746,
      "epoch": 0.48155352774693394,
      "grad_norm": 0.33569371700286865,
      "learning_rate": 3.097072642621825e-05,
      "loss": 0.2166,
      "mean_token_accuracy": 0.9470722407102585,
      "num_tokens": 53822332.0,
      "step": 7210
    },
    {
      "entropy": 1.8911114886403084,
      "epoch": 0.4822214244567078,
      "grad_norm": 0.27788618206977844,
      "learning_rate": 3.091419455381782e-05,
      "loss": 0.2115,
      "mean_token_accuracy": 0.9469106510281563,
      "num_tokens": 53895170.0,
      "step": 7220
    },
    {
      "entropy": 1.843412983417511,
      "epoch": 0.4828893211664816,
      "grad_norm": 0.27476271986961365,
      "learning_rate": 3.085763079694029e-05,
      "loss": 0.224,
      "mean_token_accuracy": 0.9419576160609722,
      "num_tokens": 53969783.0,
      "step": 7230
    },
    {
      "entropy": 1.8810128442943097,
      "epoch": 0.48355721787625544,
      "grad_norm": 0.29250454902648926,
      "learning_rate": 3.0801035463130986e-05,
      "loss": 0.21,
      "mean_token_accuracy": 0.9495118394494056,
      "num_tokens": 54046213.0,
      "step": 7240
    },
    {
      "entropy": 1.8746795989573002,
      "epoch": 0.4842251145860293,
      "grad_norm": 0.3428109288215637,
      "learning_rate": 3.074440886010697e-05,
      "loss": 0.217,
      "mean_token_accuracy": 0.946789625659585,
      "num_tokens": 54121485.0,
      "step": 7250
    },
    {
      "entropy": 1.911435604840517,
      "epoch": 0.4848930112958031,
      "grad_norm": 0.3250446021556854,
      "learning_rate": 3.068775129575528e-05,
      "loss": 0.2476,
      "mean_token_accuracy": 0.940335764363408,
      "num_tokens": 54194329.0,
      "step": 7260
    },
    {
      "entropy": 1.9061202213168145,
      "epoch": 0.48556090800557694,
      "grad_norm": 0.2643176019191742,
      "learning_rate": 3.0631063078131353e-05,
      "loss": 0.2158,
      "mean_token_accuracy": 0.9485023591667414,
      "num_tokens": 54269080.0,
      "step": 7270
    },
    {
      "entropy": 1.8695035621523857,
      "epoch": 0.4862288047153508,
      "grad_norm": 0.37113991379737854,
      "learning_rate": 3.0574344515457235e-05,
      "loss": 0.2404,
      "mean_token_accuracy": 0.9416939731687307,
      "num_tokens": 54345512.0,
      "step": 7280
    },
    {
      "entropy": 1.8922847397625446,
      "epoch": 0.4868967014251246,
      "grad_norm": 0.32886573672294617,
      "learning_rate": 3.051759591612e-05,
      "loss": 0.2203,
      "mean_token_accuracy": 0.9482366230338812,
      "num_tokens": 54420805.0,
      "step": 7290
    },
    {
      "entropy": 1.8674429669976234,
      "epoch": 0.48756459813489844,
      "grad_norm": 0.31715741753578186,
      "learning_rate": 3.0460817588670004e-05,
      "loss": 0.2265,
      "mean_token_accuracy": 0.9446107037365437,
      "num_tokens": 54494417.0,
      "step": 7300
    },
    {
      "entropy": 1.8811933800578118,
      "epoch": 0.4882324948446723,
      "grad_norm": 0.22023308277130127,
      "learning_rate": 3.0404009841819263e-05,
      "loss": 0.2113,
      "mean_token_accuracy": 0.9497490409761667,
      "num_tokens": 54572044.0,
      "step": 7310
    },
    {
      "entropy": 1.896532792598009,
      "epoch": 0.4889003915544461,
      "grad_norm": 0.3073304295539856,
      "learning_rate": 3.0347172984439753e-05,
      "loss": 0.2327,
      "mean_token_accuracy": 0.9447525463998317,
      "num_tokens": 54649823.0,
      "step": 7320
    },
    {
      "entropy": 1.9122162185609342,
      "epoch": 0.48956828826421994,
      "grad_norm": 0.3749728202819824,
      "learning_rate": 3.029030732556171e-05,
      "loss": 0.23,
      "mean_token_accuracy": 0.9446923159062862,
      "num_tokens": 54725767.0,
      "step": 7330
    },
    {
      "entropy": 1.8501642137765884,
      "epoch": 0.4902361849739938,
      "grad_norm": 0.31925761699676514,
      "learning_rate": 3.0233413174371984e-05,
      "loss": 0.2079,
      "mean_token_accuracy": 0.9485565960407257,
      "num_tokens": 54801038.0,
      "step": 7340
    },
    {
      "entropy": 1.890178529918194,
      "epoch": 0.4909040816837676,
      "grad_norm": 0.29928600788116455,
      "learning_rate": 3.0176490840212317e-05,
      "loss": 0.2247,
      "mean_token_accuracy": 0.9467503130435944,
      "num_tokens": 54874858.0,
      "step": 7350
    },
    {
      "entropy": 1.8896193079650403,
      "epoch": 0.49157197839354144,
      "grad_norm": 0.2945694625377655,
      "learning_rate": 3.0119540632577725e-05,
      "loss": 0.2289,
      "mean_token_accuracy": 0.9461097911000251,
      "num_tokens": 54947814.0,
      "step": 7360
    },
    {
      "entropy": 1.9230032838881015,
      "epoch": 0.4922398751033153,
      "grad_norm": 0.32664257287979126,
      "learning_rate": 3.006256286111474e-05,
      "loss": 0.2383,
      "mean_token_accuracy": 0.9404773727059365,
      "num_tokens": 55022623.0,
      "step": 7370
    },
    {
      "entropy": 1.927004227042198,
      "epoch": 0.4929077718130891,
      "grad_norm": 0.32149356603622437,
      "learning_rate": 3.0005557835619787e-05,
      "loss": 0.2259,
      "mean_token_accuracy": 0.9433298457413912,
      "num_tokens": 55097924.0,
      "step": 7380
    },
    {
      "entropy": 1.929019882529974,
      "epoch": 0.49357566852286294,
      "grad_norm": 0.2955763638019562,
      "learning_rate": 2.9948525866037447e-05,
      "loss": 0.2388,
      "mean_token_accuracy": 0.942776921018958,
      "num_tokens": 55171100.0,
      "step": 7390
    },
    {
      "entropy": 1.8914112061262132,
      "epoch": 0.4942435652326368,
      "grad_norm": 0.2722773551940918,
      "learning_rate": 2.9891467262458844e-05,
      "loss": 0.234,
      "mean_token_accuracy": 0.9418485786765813,
      "num_tokens": 55245366.0,
      "step": 7400
    },
    {
      "entropy": 1.8359092064201832,
      "epoch": 0.4949114619424106,
      "grad_norm": 0.27552735805511475,
      "learning_rate": 2.9834382335119876e-05,
      "loss": 0.2159,
      "mean_token_accuracy": 0.9463882490992546,
      "num_tokens": 55321714.0,
      "step": 7410
    },
    {
      "entropy": 1.8465522609651088,
      "epoch": 0.49557935865218444,
      "grad_norm": 0.32382237911224365,
      "learning_rate": 2.9777271394399575e-05,
      "loss": 0.2197,
      "mean_token_accuracy": 0.9491492222994566,
      "num_tokens": 55394885.0,
      "step": 7420
    },
    {
      "entropy": 1.836780784279108,
      "epoch": 0.4962472553619583,
      "grad_norm": 0.35757163166999817,
      "learning_rate": 2.9720134750818434e-05,
      "loss": 0.2169,
      "mean_token_accuracy": 0.9473698291927576,
      "num_tokens": 55466291.0,
      "step": 7430
    },
    {
      "entropy": 1.8885270342230798,
      "epoch": 0.4969151520717321,
      "grad_norm": 0.2864765524864197,
      "learning_rate": 2.9662972715036663e-05,
      "loss": 0.2139,
      "mean_token_accuracy": 0.9477612286806106,
      "num_tokens": 55540456.0,
      "step": 7440
    },
    {
      "entropy": 1.8900952532887458,
      "epoch": 0.49758304878150594,
      "grad_norm": 0.3118791878223419,
      "learning_rate": 2.9605785597852547e-05,
      "loss": 0.2269,
      "mean_token_accuracy": 0.9429914653301239,
      "num_tokens": 55613429.0,
      "step": 7450
    },
    {
      "entropy": 1.891358818113804,
      "epoch": 0.4982509454912798,
      "grad_norm": 0.363603413105011,
      "learning_rate": 2.9548573710200754e-05,
      "loss": 0.2337,
      "mean_token_accuracy": 0.9432939335703849,
      "num_tokens": 55688127.0,
      "step": 7460
    },
    {
      "entropy": 1.9004506438970565,
      "epoch": 0.4989188422010536,
      "grad_norm": 0.2596246898174286,
      "learning_rate": 2.9491337363150622e-05,
      "loss": 0.2263,
      "mean_token_accuracy": 0.9436323195695877,
      "num_tokens": 55763700.0,
      "step": 7470
    },
    {
      "entropy": 1.8800619326531887,
      "epoch": 0.49958673891082744,
      "grad_norm": 0.2418212592601776,
      "learning_rate": 2.943407686790448e-05,
      "loss": 0.2177,
      "mean_token_accuracy": 0.9471217170357704,
      "num_tokens": 55837299.0,
      "step": 7480
    },
    {
      "entropy": 1.8909110814332961,
      "epoch": 0.5002546356206012,
      "grad_norm": 0.3288103938102722,
      "learning_rate": 2.937679253579595e-05,
      "loss": 0.2356,
      "mean_token_accuracy": 0.942650742828846,
      "num_tokens": 55911171.0,
      "step": 7490
    },
    {
      "entropy": 1.8849252589046954,
      "epoch": 0.5009225323303751,
      "grad_norm": 0.2833203375339508,
      "learning_rate": 2.9319484678288262e-05,
      "loss": 0.2301,
      "mean_token_accuracy": 0.945120845735073,
      "num_tokens": 55986308.0,
      "step": 7500
    },
    {
      "entropy": 1.9096172168850898,
      "epoch": 0.5015904290401489,
      "grad_norm": 0.32700464129447937,
      "learning_rate": 2.926215360697255e-05,
      "loss": 0.2226,
      "mean_token_accuracy": 0.9419653285294771,
      "num_tokens": 56059138.0,
      "step": 7510
    },
    {
      "entropy": 1.8952309794723987,
      "epoch": 0.5022583257499228,
      "grad_norm": 0.32460543513298035,
      "learning_rate": 2.9204799633566183e-05,
      "loss": 0.2288,
      "mean_token_accuracy": 0.944747943803668,
      "num_tokens": 56130710.0,
      "step": 7520
    },
    {
      "entropy": 1.9046694047749042,
      "epoch": 0.5029262224596966,
      "grad_norm": 0.2865072786808014,
      "learning_rate": 2.914742306991104e-05,
      "loss": 0.2377,
      "mean_token_accuracy": 0.9400522530078887,
      "num_tokens": 56205557.0,
      "step": 7530
    },
    {
      "entropy": 1.9235623486340045,
      "epoch": 0.5035941191694704,
      "grad_norm": 0.2817114293575287,
      "learning_rate": 2.909002422797182e-05,
      "loss": 0.2239,
      "mean_token_accuracy": 0.945854339376092,
      "num_tokens": 56283915.0,
      "step": 7540
    },
    {
      "entropy": 1.8534704700112343,
      "epoch": 0.5042620158792442,
      "grad_norm": 0.28292497992515564,
      "learning_rate": 2.903260341983437e-05,
      "loss": 0.2099,
      "mean_token_accuracy": 0.9479105699807405,
      "num_tokens": 56358765.0,
      "step": 7550
    },
    {
      "entropy": 1.8678145796060561,
      "epoch": 0.5049299125890181,
      "grad_norm": 0.27667731046676636,
      "learning_rate": 2.8975160957703952e-05,
      "loss": 0.2286,
      "mean_token_accuracy": 0.9462063115090131,
      "num_tokens": 56432537.0,
      "step": 7560
    },
    {
      "entropy": 1.9020389944314957,
      "epoch": 0.505597809298792,
      "grad_norm": 0.3124503195285797,
      "learning_rate": 2.8917697153903567e-05,
      "loss": 0.2189,
      "mean_token_accuracy": 0.9463117547333241,
      "num_tokens": 56505356.0,
      "step": 7570
    },
    {
      "entropy": 1.8646268159151078,
      "epoch": 0.5062657060085658,
      "grad_norm": 0.34951356053352356,
      "learning_rate": 2.8860212320872264e-05,
      "loss": 0.2129,
      "mean_token_accuracy": 0.9468833450227976,
      "num_tokens": 56578292.0,
      "step": 7580
    },
    {
      "entropy": 1.8221788071095943,
      "epoch": 0.5069336027183396,
      "grad_norm": 0.2863594591617584,
      "learning_rate": 2.8802706771163428e-05,
      "loss": 0.2093,
      "mean_token_accuracy": 0.9479325834661723,
      "num_tokens": 56654498.0,
      "step": 7590
    },
    {
      "entropy": 1.8536480709910392,
      "epoch": 0.5076014994281134,
      "grad_norm": 0.327137291431427,
      "learning_rate": 2.874518081744309e-05,
      "loss": 0.2421,
      "mean_token_accuracy": 0.9376656226813793,
      "num_tokens": 56728279.0,
      "step": 7600
    },
    {
      "entropy": 1.8634555891156197,
      "epoch": 0.5082693961378872,
      "grad_norm": 0.2959681749343872,
      "learning_rate": 2.8687634772488186e-05,
      "loss": 0.2316,
      "mean_token_accuracy": 0.9440638281404972,
      "num_tokens": 56802952.0,
      "step": 7610
    },
    {
      "entropy": 1.871063868701458,
      "epoch": 0.5089372928476611,
      "grad_norm": 0.32574859261512756,
      "learning_rate": 2.863006894918494e-05,
      "loss": 0.2126,
      "mean_token_accuracy": 0.9461169857531786,
      "num_tokens": 56878476.0,
      "step": 7620
    },
    {
      "entropy": 1.8941338248550892,
      "epoch": 0.509605189557435,
      "grad_norm": 0.30667027831077576,
      "learning_rate": 2.8572483660527078e-05,
      "loss": 0.2394,
      "mean_token_accuracy": 0.9433797467499971,
      "num_tokens": 56954176.0,
      "step": 7630
    },
    {
      "entropy": 1.8805853471159935,
      "epoch": 0.5102730862672088,
      "grad_norm": 0.29259997606277466,
      "learning_rate": 2.8514879219614182e-05,
      "loss": 0.2106,
      "mean_token_accuracy": 0.9467896539717913,
      "num_tokens": 57029462.0,
      "step": 7640
    },
    {
      "entropy": 1.8603178225457668,
      "epoch": 0.5109409829769826,
      "grad_norm": 0.27425578236579895,
      "learning_rate": 2.845725593964996e-05,
      "loss": 0.2095,
      "mean_token_accuracy": 0.9481328092515469,
      "num_tokens": 57104841.0,
      "step": 7650
    },
    {
      "entropy": 1.9001431725919247,
      "epoch": 0.5116088796867564,
      "grad_norm": 0.31165963411331177,
      "learning_rate": 2.8399614133940537e-05,
      "loss": 0.2211,
      "mean_token_accuracy": 0.9453007891774178,
      "num_tokens": 57177654.0,
      "step": 7660
    },
    {
      "entropy": 1.8816741175949574,
      "epoch": 0.5122767763965302,
      "grad_norm": 0.26656800508499146,
      "learning_rate": 2.8341954115892798e-05,
      "loss": 0.2147,
      "mean_token_accuracy": 0.9465901501476764,
      "num_tokens": 57253170.0,
      "step": 7670
    },
    {
      "entropy": 1.8403189226984977,
      "epoch": 0.5129446731063041,
      "grad_norm": 0.2890680432319641,
      "learning_rate": 2.8284276199012618e-05,
      "loss": 0.2263,
      "mean_token_accuracy": 0.9464253015816212,
      "num_tokens": 57326768.0,
      "step": 7680
    },
    {
      "entropy": 1.8553779326379298,
      "epoch": 0.513612569816078,
      "grad_norm": 0.40375080704689026,
      "learning_rate": 2.822658069690321e-05,
      "loss": 0.2322,
      "mean_token_accuracy": 0.943346194177866,
      "num_tokens": 57400329.0,
      "step": 7690
    },
    {
      "entropy": 1.8662341050803661,
      "epoch": 0.5142804665258518,
      "grad_norm": 0.30424338579177856,
      "learning_rate": 2.81688679232634e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.9460494916886091,
      "num_tokens": 57476630.0,
      "step": 7700
    },
    {
      "entropy": 1.8503398597240448,
      "epoch": 0.5149483632356257,
      "grad_norm": 0.2923349440097809,
      "learning_rate": 2.81111381918859e-05,
      "loss": 0.2006,
      "mean_token_accuracy": 0.9492053009569645,
      "num_tokens": 57553020.0,
      "step": 7710
    },
    {
      "entropy": 1.8587808765470981,
      "epoch": 0.5156162599453994,
      "grad_norm": 0.26285168528556824,
      "learning_rate": 2.8053391816655656e-05,
      "loss": 0.2294,
      "mean_token_accuracy": 0.9411367610096931,
      "num_tokens": 57629074.0,
      "step": 7720
    },
    {
      "entropy": 1.8495754942297935,
      "epoch": 0.5162841566551732,
      "grad_norm": 0.31382155418395996,
      "learning_rate": 2.799562911154809e-05,
      "loss": 0.2256,
      "mean_token_accuracy": 0.9439142808318138,
      "num_tokens": 57702819.0,
      "step": 7730
    },
    {
      "entropy": 1.837379440665245,
      "epoch": 0.5169520533649471,
      "grad_norm": 0.26100775599479675,
      "learning_rate": 2.7937850390627406e-05,
      "loss": 0.2157,
      "mean_token_accuracy": 0.9464406784623861,
      "num_tokens": 57779718.0,
      "step": 7740
    },
    {
      "entropy": 1.8983632288873196,
      "epoch": 0.517619950074721,
      "grad_norm": 0.31474626064300537,
      "learning_rate": 2.788005596804492e-05,
      "loss": 0.2282,
      "mean_token_accuracy": 0.9455306556075811,
      "num_tokens": 57855193.0,
      "step": 7750
    },
    {
      "entropy": 1.8795831046998501,
      "epoch": 0.5182878467844948,
      "grad_norm": 0.29401010274887085,
      "learning_rate": 2.7822246158037268e-05,
      "loss": 0.2362,
      "mean_token_accuracy": 0.9417175475507975,
      "num_tokens": 57933506.0,
      "step": 7760
    },
    {
      "entropy": 1.8830607444047929,
      "epoch": 0.5189557434942687,
      "grad_norm": 0.30570951104164124,
      "learning_rate": 2.7764421274924803e-05,
      "loss": 0.2239,
      "mean_token_accuracy": 0.9436976905912161,
      "num_tokens": 58008336.0,
      "step": 7770
    },
    {
      "entropy": 1.916810903698206,
      "epoch": 0.5196236402040424,
      "grad_norm": 0.2321866899728775,
      "learning_rate": 2.77065816331098e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9460313744843006,
      "num_tokens": 58081812.0,
      "step": 7780
    },
    {
      "entropy": 1.8622855216264724,
      "epoch": 0.5202915369138162,
      "grad_norm": 0.33274272084236145,
      "learning_rate": 2.764872754707479e-05,
      "loss": 0.2403,
      "mean_token_accuracy": 0.9432096824049949,
      "num_tokens": 58156913.0,
      "step": 7790
    },
    {
      "entropy": 1.8577060863375663,
      "epoch": 0.5209594336235901,
      "grad_norm": 0.3021269738674164,
      "learning_rate": 2.7590859331380842e-05,
      "loss": 0.2162,
      "mean_token_accuracy": 0.9461881406605244,
      "num_tokens": 58234003.0,
      "step": 7800
    },
    {
      "entropy": 1.8902149342000485,
      "epoch": 0.521627330333364,
      "grad_norm": 0.3278430998325348,
      "learning_rate": 2.7532977300665836e-05,
      "loss": 0.2392,
      "mean_token_accuracy": 0.9436495702713728,
      "num_tokens": 58306444.0,
      "step": 7810
    },
    {
      "entropy": 1.8972839921712876,
      "epoch": 0.5222952270431378,
      "grad_norm": 0.35462868213653564,
      "learning_rate": 2.747508176964278e-05,
      "loss": 0.2375,
      "mean_token_accuracy": 0.9418607074767351,
      "num_tokens": 58382798.0,
      "step": 7820
    },
    {
      "entropy": 1.8793871521949768,
      "epoch": 0.5229631237529117,
      "grad_norm": 0.26488596200942993,
      "learning_rate": 2.741717305309808e-05,
      "loss": 0.2029,
      "mean_token_accuracy": 0.9513823010027409,
      "num_tokens": 58458235.0,
      "step": 7830
    },
    {
      "entropy": 1.884317271411419,
      "epoch": 0.5236310204626854,
      "grad_norm": 0.3493448495864868,
      "learning_rate": 2.735925146588984e-05,
      "loss": 0.2266,
      "mean_token_accuracy": 0.9436733886599541,
      "num_tokens": 58532396.0,
      "step": 7840
    },
    {
      "entropy": 1.8265122584998608,
      "epoch": 0.5242989171724592,
      "grad_norm": 0.3416137993335724,
      "learning_rate": 2.730131732294614e-05,
      "loss": 0.2035,
      "mean_token_accuracy": 0.9487059112638235,
      "num_tokens": 58608437.0,
      "step": 7850
    },
    {
      "entropy": 1.8567959085106849,
      "epoch": 0.5249668138822331,
      "grad_norm": 0.30284300446510315,
      "learning_rate": 2.7243370939263314e-05,
      "loss": 0.2228,
      "mean_token_accuracy": 0.9484416019171477,
      "num_tokens": 58685069.0,
      "step": 7860
    },
    {
      "entropy": 1.9211342953145505,
      "epoch": 0.525634710592007,
      "grad_norm": 0.26501908898353577,
      "learning_rate": 2.7185412629904266e-05,
      "loss": 0.2307,
      "mean_token_accuracy": 0.9424899592995644,
      "num_tokens": 58759659.0,
      "step": 7870
    },
    {
      "entropy": 1.9020105704665184,
      "epoch": 0.5263026073017808,
      "grad_norm": 0.2968217134475708,
      "learning_rate": 2.7127442709996754e-05,
      "loss": 0.2282,
      "mean_token_accuracy": 0.9454322118312121,
      "num_tokens": 58835859.0,
      "step": 7880
    },
    {
      "entropy": 1.8932685352861882,
      "epoch": 0.5269705040115547,
      "grad_norm": 0.33987903594970703,
      "learning_rate": 2.7069461494731625e-05,
      "loss": 0.2353,
      "mean_token_accuracy": 0.9406254064291716,
      "num_tokens": 58909083.0,
      "step": 7890
    },
    {
      "entropy": 1.8616687595844268,
      "epoch": 0.5276384007213284,
      "grad_norm": 0.2925051152706146,
      "learning_rate": 2.7011469299361185e-05,
      "loss": 0.2273,
      "mean_token_accuracy": 0.9443053800612689,
      "num_tokens": 58982603.0,
      "step": 7900
    },
    {
      "entropy": 1.886468544602394,
      "epoch": 0.5283062974311022,
      "grad_norm": 0.25922802090644836,
      "learning_rate": 2.69534664391974e-05,
      "loss": 0.2325,
      "mean_token_accuracy": 0.941967923566699,
      "num_tokens": 59058616.0,
      "step": 7910
    },
    {
      "entropy": 1.8770767435431481,
      "epoch": 0.5289741941408761,
      "grad_norm": 0.30758416652679443,
      "learning_rate": 2.689545322961025e-05,
      "loss": 0.2136,
      "mean_token_accuracy": 0.9472320355474949,
      "num_tokens": 59131392.0,
      "step": 7920
    },
    {
      "entropy": 1.8575767748057843,
      "epoch": 0.52964209085065,
      "grad_norm": 0.2764682173728943,
      "learning_rate": 2.6837429986025986e-05,
      "loss": 0.2223,
      "mean_token_accuracy": 0.9451409664005042,
      "num_tokens": 59204405.0,
      "step": 7930
    },
    {
      "entropy": 1.8431726060807705,
      "epoch": 0.5303099875604238,
      "grad_norm": 0.29136666655540466,
      "learning_rate": 2.677939702392539e-05,
      "loss": 0.2388,
      "mean_token_accuracy": 0.9441762343049049,
      "num_tokens": 59279492.0,
      "step": 7940
    },
    {
      "entropy": 1.866943745315075,
      "epoch": 0.5309778842701977,
      "grad_norm": 0.3011800944805145,
      "learning_rate": 2.6721354658842117e-05,
      "loss": 0.2235,
      "mean_token_accuracy": 0.9448039442300796,
      "num_tokens": 59352696.0,
      "step": 7950
    },
    {
      "entropy": 1.8778190918266773,
      "epoch": 0.5316457809799714,
      "grad_norm": 0.29560622572898865,
      "learning_rate": 2.6663303206360917e-05,
      "loss": 0.2319,
      "mean_token_accuracy": 0.9429251346737146,
      "num_tokens": 59425908.0,
      "step": 7960
    },
    {
      "entropy": 1.835938535630703,
      "epoch": 0.5323136776897452,
      "grad_norm": 0.33805063366889954,
      "learning_rate": 2.6605242982115975e-05,
      "loss": 0.2255,
      "mean_token_accuracy": 0.9455182150006294,
      "num_tokens": 59500064.0,
      "step": 7970
    },
    {
      "entropy": 1.86594817340374,
      "epoch": 0.5329815743995191,
      "grad_norm": 0.2938584089279175,
      "learning_rate": 2.6547174301789163e-05,
      "loss": 0.2194,
      "mean_token_accuracy": 0.9445056736469268,
      "num_tokens": 59573575.0,
      "step": 7980
    },
    {
      "entropy": 1.856761346757412,
      "epoch": 0.533649471109293,
      "grad_norm": 0.3130123019218445,
      "learning_rate": 2.6489097481108316e-05,
      "loss": 0.215,
      "mean_token_accuracy": 0.9471686478704214,
      "num_tokens": 59650851.0,
      "step": 7990
    },
    {
      "entropy": 1.8781053856015206,
      "epoch": 0.5343173678190668,
      "grad_norm": 0.2852950096130371,
      "learning_rate": 2.6431012835845554e-05,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9456228621304035,
      "num_tokens": 59727481.0,
      "step": 8000
    },
    {
      "epoch": 0.5343173678190668,
      "eval_entropy": 1.8662060023546219,
      "eval_loss": 0.2400374561548233,
      "eval_mean_token_accuracy": 0.9455185312509536,
      "eval_num_tokens": 59727481.0,
      "eval_runtime": 551.793,
      "eval_samples_per_second": 36.245,
      "eval_steps_per_second": 9.061,
      "step": 8000
    },
    {
      "entropy": 1.8653590969741345,
      "epoch": 0.5349852645288407,
      "grad_norm": 0.3038094639778137,
      "learning_rate": 2.6372920681815506e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.9466887358576059,
      "num_tokens": 59803300.0,
      "step": 8010
    },
    {
      "entropy": 1.8486957401037216,
      "epoch": 0.5356531612386144,
      "grad_norm": 0.35459452867507935,
      "learning_rate": 2.631482133487365e-05,
      "loss": 0.2,
      "mean_token_accuracy": 0.951388519257307,
      "num_tokens": 59879357.0,
      "step": 8020
    },
    {
      "entropy": 1.8877455413341522,
      "epoch": 0.5363210579483882,
      "grad_norm": 0.3583106994628906,
      "learning_rate": 2.6256715110914594e-05,
      "loss": 0.2288,
      "mean_token_accuracy": 0.9436018981039525,
      "num_tokens": 59952207.0,
      "step": 8030
    },
    {
      "entropy": 1.8483912728726863,
      "epoch": 0.5369889546581621,
      "grad_norm": 0.2703855633735657,
      "learning_rate": 2.619860232587028e-05,
      "loss": 0.2179,
      "mean_token_accuracy": 0.9464797597378493,
      "num_tokens": 60028845.0,
      "step": 8040
    },
    {
      "entropy": 1.8448423579335214,
      "epoch": 0.537656851367936,
      "grad_norm": 0.35421881079673767,
      "learning_rate": 2.614048329570838e-05,
      "loss": 0.2272,
      "mean_token_accuracy": 0.9448291774839163,
      "num_tokens": 60103064.0,
      "step": 8050
    },
    {
      "entropy": 1.8547725677490234,
      "epoch": 0.5383247480777098,
      "grad_norm": 0.2990838587284088,
      "learning_rate": 2.6082358336430483e-05,
      "loss": 0.238,
      "mean_token_accuracy": 0.9419344145804643,
      "num_tokens": 60176068.0,
      "step": 8060
    },
    {
      "entropy": 1.8373032554984092,
      "epoch": 0.5389926447874837,
      "grad_norm": 0.31900662183761597,
      "learning_rate": 2.6024227764070445e-05,
      "loss": 0.2195,
      "mean_token_accuracy": 0.9477964285761118,
      "num_tokens": 60250295.0,
      "step": 8070
    },
    {
      "entropy": 1.8211630158126355,
      "epoch": 0.5396605414972574,
      "grad_norm": 0.3096672296524048,
      "learning_rate": 2.5966091894692622e-05,
      "loss": 0.2116,
      "mean_token_accuracy": 0.9478581257164478,
      "num_tokens": 60325809.0,
      "step": 8080
    },
    {
      "entropy": 1.8228729732334614,
      "epoch": 0.5403284382070312,
      "grad_norm": 0.32388898730278015,
      "learning_rate": 2.590795104439017e-05,
      "loss": 0.2298,
      "mean_token_accuracy": 0.9449355151504278,
      "num_tokens": 60402637.0,
      "step": 8090
    },
    {
      "entropy": 1.8285527899861336,
      "epoch": 0.5409963349168051,
      "grad_norm": 0.34106579422950745,
      "learning_rate": 2.5849805529283345e-05,
      "loss": 0.2125,
      "mean_token_accuracy": 0.9474385660141706,
      "num_tokens": 60475072.0,
      "step": 8100
    },
    {
      "entropy": 1.8566961660981178,
      "epoch": 0.541664231626579,
      "grad_norm": 0.31970328092575073,
      "learning_rate": 2.579165566551774e-05,
      "loss": 0.2201,
      "mean_token_accuracy": 0.9445961304008961,
      "num_tokens": 60548308.0,
      "step": 8110
    },
    {
      "entropy": 1.8654835544526578,
      "epoch": 0.5423321283363528,
      "grad_norm": 0.32882964611053467,
      "learning_rate": 2.5733501769262603e-05,
      "loss": 0.2187,
      "mean_token_accuracy": 0.9471868388354778,
      "num_tokens": 60623089.0,
      "step": 8120
    },
    {
      "entropy": 1.8858015693724155,
      "epoch": 0.5430000250461267,
      "grad_norm": 0.30182120203971863,
      "learning_rate": 2.567534415670914e-05,
      "loss": 0.2367,
      "mean_token_accuracy": 0.9412665236741304,
      "num_tokens": 60695316.0,
      "step": 8130
    },
    {
      "entropy": 1.8781175762414932,
      "epoch": 0.5436679217559004,
      "grad_norm": 0.30538490414619446,
      "learning_rate": 2.5617183144068696e-05,
      "loss": 0.2378,
      "mean_token_accuracy": 0.9420166041702032,
      "num_tokens": 60769035.0,
      "step": 8140
    },
    {
      "entropy": 1.873946213722229,
      "epoch": 0.5443358184656742,
      "grad_norm": 0.27864402532577515,
      "learning_rate": 2.5559019047571158e-05,
      "loss": 0.2174,
      "mean_token_accuracy": 0.9482248164713383,
      "num_tokens": 60843937.0,
      "step": 8150
    },
    {
      "entropy": 1.8589042626321315,
      "epoch": 0.5450037151754481,
      "grad_norm": 0.2768818438053131,
      "learning_rate": 2.5500852183463176e-05,
      "loss": 0.2276,
      "mean_token_accuracy": 0.9468735165894031,
      "num_tokens": 60918638.0,
      "step": 8160
    },
    {
      "entropy": 1.8665156617760659,
      "epoch": 0.545671611885222,
      "grad_norm": 0.27443212270736694,
      "learning_rate": 2.54426828680064e-05,
      "loss": 0.2081,
      "mean_token_accuracy": 0.9479584507644176,
      "num_tokens": 60993406.0,
      "step": 8170
    },
    {
      "entropy": 1.9043079972267152,
      "epoch": 0.5463395085949958,
      "grad_norm": 0.27907121181488037,
      "learning_rate": 2.5384511417475876e-05,
      "loss": 0.2197,
      "mean_token_accuracy": 0.9454525198787451,
      "num_tokens": 61070200.0,
      "step": 8180
    },
    {
      "entropy": 1.858091312646866,
      "epoch": 0.5470074053047697,
      "grad_norm": 0.31660377979278564,
      "learning_rate": 2.532633814815821e-05,
      "loss": 0.2192,
      "mean_token_accuracy": 0.9451952267438173,
      "num_tokens": 61145664.0,
      "step": 8190
    },
    {
      "entropy": 1.8549466118216515,
      "epoch": 0.5476753020145434,
      "grad_norm": 0.3399874269962311,
      "learning_rate": 2.5268163376349917e-05,
      "loss": 0.2187,
      "mean_token_accuracy": 0.947270717844367,
      "num_tokens": 61220071.0,
      "step": 8200
    },
    {
      "entropy": 1.8488216981291772,
      "epoch": 0.5483431987243172,
      "grad_norm": 0.31718650460243225,
      "learning_rate": 2.520998741835567e-05,
      "loss": 0.2441,
      "mean_token_accuracy": 0.9422694411128759,
      "num_tokens": 61292349.0,
      "step": 8210
    },
    {
      "entropy": 1.8508421078324317,
      "epoch": 0.5490110954340911,
      "grad_norm": 0.3083728551864624,
      "learning_rate": 2.5151810590486603e-05,
      "loss": 0.2384,
      "mean_token_accuracy": 0.9440817829221487,
      "num_tokens": 61369176.0,
      "step": 8220
    },
    {
      "entropy": 1.8460893273353576,
      "epoch": 0.549678992143865,
      "grad_norm": 0.2717324495315552,
      "learning_rate": 2.509363320905858e-05,
      "loss": 0.2152,
      "mean_token_accuracy": 0.9461930736899375,
      "num_tokens": 61440801.0,
      "step": 8230
    },
    {
      "entropy": 1.837957539409399,
      "epoch": 0.5503468888536388,
      "grad_norm": 0.27208277583122253,
      "learning_rate": 2.5035455590390466e-05,
      "loss": 0.212,
      "mean_token_accuracy": 0.9456074565649033,
      "num_tokens": 61516268.0,
      "step": 8240
    },
    {
      "entropy": 1.857341504096985,
      "epoch": 0.5510147855634127,
      "grad_norm": 0.30795618891716003,
      "learning_rate": 2.4977278050802416e-05,
      "loss": 0.2108,
      "mean_token_accuracy": 0.9469363793730736,
      "num_tokens": 61591889.0,
      "step": 8250
    },
    {
      "entropy": 1.859669104218483,
      "epoch": 0.5516826822731864,
      "grad_norm": 0.26675838232040405,
      "learning_rate": 2.4919100906614166e-05,
      "loss": 0.2215,
      "mean_token_accuracy": 0.9426109079271555,
      "num_tokens": 61665263.0,
      "step": 8260
    },
    {
      "entropy": 1.883779937773943,
      "epoch": 0.5523505789829602,
      "grad_norm": 0.3161466121673584,
      "learning_rate": 2.486092447414328e-05,
      "loss": 0.2296,
      "mean_token_accuracy": 0.9453294552862644,
      "num_tokens": 61741405.0,
      "step": 8270
    },
    {
      "entropy": 1.9008271612226963,
      "epoch": 0.5530184756927341,
      "grad_norm": 0.3709556758403778,
      "learning_rate": 2.4802749069703486e-05,
      "loss": 0.2378,
      "mean_token_accuracy": 0.9416403114795685,
      "num_tokens": 61816289.0,
      "step": 8280
    },
    {
      "entropy": 1.9084829442203044,
      "epoch": 0.553686372402508,
      "grad_norm": 0.3552102744579315,
      "learning_rate": 2.474457500960289e-05,
      "loss": 0.2207,
      "mean_token_accuracy": 0.944573924690485,
      "num_tokens": 61890748.0,
      "step": 8290
    },
    {
      "entropy": 1.9128666058182717,
      "epoch": 0.5543542691122818,
      "grad_norm": 0.2954138517379761,
      "learning_rate": 2.4686402610142308e-05,
      "loss": 0.23,
      "mean_token_accuracy": 0.9433212172240019,
      "num_tokens": 61963216.0,
      "step": 8300
    },
    {
      "entropy": 1.8967509530484676,
      "epoch": 0.5550221658220557,
      "grad_norm": 0.329120397567749,
      "learning_rate": 2.462823218761352e-05,
      "loss": 0.2344,
      "mean_token_accuracy": 0.9463879529386758,
      "num_tokens": 62037313.0,
      "step": 8310
    },
    {
      "entropy": 1.8885513134300709,
      "epoch": 0.5556900625318294,
      "grad_norm": 0.28542664647102356,
      "learning_rate": 2.457006405829755e-05,
      "loss": 0.2084,
      "mean_token_accuracy": 0.9479470700025558,
      "num_tokens": 62113956.0,
      "step": 8320
    },
    {
      "entropy": 1.9086256839334965,
      "epoch": 0.5563579592416033,
      "grad_norm": 0.324243426322937,
      "learning_rate": 2.451189853846298e-05,
      "loss": 0.2385,
      "mean_token_accuracy": 0.9402048338204623,
      "num_tokens": 62186155.0,
      "step": 8330
    },
    {
      "entropy": 1.8833911538124084,
      "epoch": 0.5570258559513771,
      "grad_norm": 0.28702181577682495,
      "learning_rate": 2.4453735944364173e-05,
      "loss": 0.2223,
      "mean_token_accuracy": 0.9454862978309393,
      "num_tokens": 62261828.0,
      "step": 8340
    },
    {
      "entropy": 1.8757649213075638,
      "epoch": 0.557693752661151,
      "grad_norm": 0.37363365292549133,
      "learning_rate": 2.4395576592239597e-05,
      "loss": 0.2186,
      "mean_token_accuracy": 0.947131934016943,
      "num_tokens": 62337417.0,
      "step": 8350
    },
    {
      "entropy": 1.8477208264172078,
      "epoch": 0.5583616493709248,
      "grad_norm": 0.2944476008415222,
      "learning_rate": 2.4337420798310094e-05,
      "loss": 0.213,
      "mean_token_accuracy": 0.9485928833484649,
      "num_tokens": 62411953.0,
      "step": 8360
    },
    {
      "entropy": 1.8999232091009617,
      "epoch": 0.5590295460806987,
      "grad_norm": 0.4096059501171112,
      "learning_rate": 2.427926887877716e-05,
      "loss": 0.2299,
      "mean_token_accuracy": 0.9444582555443048,
      "num_tokens": 62487166.0,
      "step": 8370
    },
    {
      "entropy": 1.8844099454581738,
      "epoch": 0.5596974427904724,
      "grad_norm": 0.2626202702522278,
      "learning_rate": 2.422112114982122e-05,
      "loss": 0.2093,
      "mean_token_accuracy": 0.9487072411924601,
      "num_tokens": 62561941.0,
      "step": 8380
    },
    {
      "entropy": 1.8820263959467411,
      "epoch": 0.5603653395002463,
      "grad_norm": 0.27855777740478516,
      "learning_rate": 2.4162977927599922e-05,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9470836147665977,
      "num_tokens": 62636692.0,
      "step": 8390
    },
    {
      "entropy": 1.885035403072834,
      "epoch": 0.5610332362100201,
      "grad_norm": 0.3330121338367462,
      "learning_rate": 2.4104839528246412e-05,
      "loss": 0.2168,
      "mean_token_accuracy": 0.9466604709625244,
      "num_tokens": 62708450.0,
      "step": 8400
    },
    {
      "entropy": 1.8940217658877372,
      "epoch": 0.561701132919794,
      "grad_norm": 0.39527344703674316,
      "learning_rate": 2.40467062678676e-05,
      "loss": 0.2635,
      "mean_token_accuracy": 0.9377460069954395,
      "num_tokens": 62780463.0,
      "step": 8410
    },
    {
      "entropy": 1.8828867062926293,
      "epoch": 0.5623690296295678,
      "grad_norm": 0.27089861035346985,
      "learning_rate": 2.398857846254246e-05,
      "loss": 0.2335,
      "mean_token_accuracy": 0.9446988385170698,
      "num_tokens": 62853909.0,
      "step": 8420
    },
    {
      "entropy": 1.8660736046731472,
      "epoch": 0.5630369263393417,
      "grad_norm": 0.2980462610721588,
      "learning_rate": 2.393045642832031e-05,
      "loss": 0.2139,
      "mean_token_accuracy": 0.9481238935142755,
      "num_tokens": 62927668.0,
      "step": 8430
    },
    {
      "entropy": 1.880422654747963,
      "epoch": 0.5637048230491154,
      "grad_norm": 0.32885050773620605,
      "learning_rate": 2.387234048121908e-05,
      "loss": 0.2418,
      "mean_token_accuracy": 0.9407288305461406,
      "num_tokens": 63002537.0,
      "step": 8440
    },
    {
      "entropy": 1.854230558872223,
      "epoch": 0.5643727197588893,
      "grad_norm": 0.29425862431526184,
      "learning_rate": 2.3814230937223625e-05,
      "loss": 0.2074,
      "mean_token_accuracy": 0.949595658481121,
      "num_tokens": 63079226.0,
      "step": 8450
    },
    {
      "entropy": 1.8476892814040184,
      "epoch": 0.5650406164686631,
      "grad_norm": 0.3095671236515045,
      "learning_rate": 2.375612811228397e-05,
      "loss": 0.2082,
      "mean_token_accuracy": 0.9486114624887705,
      "num_tokens": 63153413.0,
      "step": 8460
    },
    {
      "entropy": 1.852500356733799,
      "epoch": 0.565708513178437,
      "grad_norm": 0.38513025641441345,
      "learning_rate": 2.3698032322313597e-05,
      "loss": 0.2328,
      "mean_token_accuracy": 0.9416500601917506,
      "num_tokens": 63227570.0,
      "step": 8470
    },
    {
      "entropy": 1.8510682493448258,
      "epoch": 0.5663764098882108,
      "grad_norm": 0.27778077125549316,
      "learning_rate": 2.363994388318777e-05,
      "loss": 0.2101,
      "mean_token_accuracy": 0.9455769184976817,
      "num_tokens": 63299693.0,
      "step": 8480
    },
    {
      "entropy": 1.875589492917061,
      "epoch": 0.5670443065979847,
      "grad_norm": 0.3203050196170807,
      "learning_rate": 2.358186311074175e-05,
      "loss": 0.2358,
      "mean_token_accuracy": 0.9414773181080818,
      "num_tokens": 63375898.0,
      "step": 8490
    },
    {
      "entropy": 1.8759691730141639,
      "epoch": 0.5677122033077584,
      "grad_norm": 0.3363718092441559,
      "learning_rate": 2.352379032076913e-05,
      "loss": 0.2382,
      "mean_token_accuracy": 0.9420657128095626,
      "num_tokens": 63451921.0,
      "step": 8500
    },
    {
      "entropy": 1.8908679261803627,
      "epoch": 0.5683801000175323,
      "grad_norm": 0.34495604038238525,
      "learning_rate": 2.3465725829020107e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9428388562053442,
      "num_tokens": 63527033.0,
      "step": 8510
    },
    {
      "entropy": 1.890539725869894,
      "epoch": 0.5690479967273061,
      "grad_norm": 0.31119126081466675,
      "learning_rate": 2.3407669951199746e-05,
      "loss": 0.2267,
      "mean_token_accuracy": 0.943016241490841,
      "num_tokens": 63601512.0,
      "step": 8520
    },
    {
      "entropy": 1.9125523708760739,
      "epoch": 0.56971589343708,
      "grad_norm": 0.3263043463230133,
      "learning_rate": 2.334962300296629e-05,
      "loss": 0.2271,
      "mean_token_accuracy": 0.9413581982254982,
      "num_tokens": 63678361.0,
      "step": 8530
    },
    {
      "entropy": 1.857296995073557,
      "epoch": 0.5703837901468538,
      "grad_norm": 0.33531224727630615,
      "learning_rate": 2.3291585299929408e-05,
      "loss": 0.2118,
      "mean_token_accuracy": 0.9479674693197012,
      "num_tokens": 63754152.0,
      "step": 8540
    },
    {
      "entropy": 1.8451183453202247,
      "epoch": 0.5710516868566277,
      "grad_norm": 0.3622986078262329,
      "learning_rate": 2.323355715764854e-05,
      "loss": 0.2231,
      "mean_token_accuracy": 0.9451228737831116,
      "num_tokens": 63829944.0,
      "step": 8550
    },
    {
      "entropy": 1.8715076744556427,
      "epoch": 0.5717195835664014,
      "grad_norm": 0.27118003368377686,
      "learning_rate": 2.3175538891631098e-05,
      "loss": 0.2252,
      "mean_token_accuracy": 0.9428528670221568,
      "num_tokens": 63907417.0,
      "step": 8560
    },
    {
      "entropy": 1.8734317906200886,
      "epoch": 0.5723874802761753,
      "grad_norm": 0.34238213300704956,
      "learning_rate": 2.3117530817330828e-05,
      "loss": 0.2284,
      "mean_token_accuracy": 0.944256753847003,
      "num_tokens": 63983175.0,
      "step": 8570
    },
    {
      "entropy": 1.8877094030380248,
      "epoch": 0.5730553769859491,
      "grad_norm": 0.25891923904418945,
      "learning_rate": 2.3059533250146053e-05,
      "loss": 0.2155,
      "mean_token_accuracy": 0.9459561046212912,
      "num_tokens": 64059041.0,
      "step": 8580
    },
    {
      "entropy": 1.8511331334710122,
      "epoch": 0.573723273695723,
      "grad_norm": 0.25855767726898193,
      "learning_rate": 2.3001546505417955e-05,
      "loss": 0.2165,
      "mean_token_accuracy": 0.9472135379910469,
      "num_tokens": 64133932.0,
      "step": 8590
    },
    {
      "entropy": 1.8539586246013642,
      "epoch": 0.5743911704054968,
      "grad_norm": 0.3344471752643585,
      "learning_rate": 2.29435708984289e-05,
      "loss": 0.2169,
      "mean_token_accuracy": 0.9478591121733189,
      "num_tokens": 64207767.0,
      "step": 8600
    },
    {
      "entropy": 1.8822246044874191,
      "epoch": 0.5750590671152707,
      "grad_norm": 0.33001628518104553,
      "learning_rate": 2.2885606744400665e-05,
      "loss": 0.2332,
      "mean_token_accuracy": 0.9429154712706804,
      "num_tokens": 64283244.0,
      "step": 8610
    },
    {
      "entropy": 1.8829777382314206,
      "epoch": 0.5757269638250444,
      "grad_norm": 0.33794325590133667,
      "learning_rate": 2.2827654358492776e-05,
      "loss": 0.2273,
      "mean_token_accuracy": 0.9459087539464235,
      "num_tokens": 64358844.0,
      "step": 8620
    },
    {
      "entropy": 1.8854925595223904,
      "epoch": 0.5763948605348183,
      "grad_norm": 0.2960599660873413,
      "learning_rate": 2.276971405580077e-05,
      "loss": 0.2492,
      "mean_token_accuracy": 0.940941645950079,
      "num_tokens": 64432317.0,
      "step": 8630
    },
    {
      "entropy": 1.8776641704142094,
      "epoch": 0.5770627572445921,
      "grad_norm": 0.30550751090049744,
      "learning_rate": 2.271178615135449e-05,
      "loss": 0.2197,
      "mean_token_accuracy": 0.9465542431920767,
      "num_tokens": 64506050.0,
      "step": 8640
    },
    {
      "entropy": 1.8687704667448997,
      "epoch": 0.577730653954366,
      "grad_norm": 0.30258113145828247,
      "learning_rate": 2.2653870960116343e-05,
      "loss": 0.2244,
      "mean_token_accuracy": 0.9450478479266167,
      "num_tokens": 64580978.0,
      "step": 8650
    },
    {
      "entropy": 1.875409981608391,
      "epoch": 0.5783985506641398,
      "grad_norm": 0.26282933354377747,
      "learning_rate": 2.2595968796979647e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9448274232447147,
      "num_tokens": 64655287.0,
      "step": 8660
    },
    {
      "entropy": 1.8955595210194587,
      "epoch": 0.5790664473739137,
      "grad_norm": 0.2973979115486145,
      "learning_rate": 2.253807997676686e-05,
      "loss": 0.2078,
      "mean_token_accuracy": 0.9478851392865181,
      "num_tokens": 64728315.0,
      "step": 8670
    },
    {
      "entropy": 1.8781954944133759,
      "epoch": 0.5797343440836874,
      "grad_norm": 0.3379066288471222,
      "learning_rate": 2.2480204814227916e-05,
      "loss": 0.2306,
      "mean_token_accuracy": 0.9437055505812169,
      "num_tokens": 64803924.0,
      "step": 8680
    },
    {
      "entropy": 1.8334377713501453,
      "epoch": 0.5804022407934613,
      "grad_norm": 0.24142281711101532,
      "learning_rate": 2.2422343624038454e-05,
      "loss": 0.2344,
      "mean_token_accuracy": 0.9435370784252882,
      "num_tokens": 64878201.0,
      "step": 8690
    },
    {
      "entropy": 1.8821712143719196,
      "epoch": 0.5810701375032351,
      "grad_norm": 0.34564438462257385,
      "learning_rate": 2.2364496720798182e-05,
      "loss": 0.2189,
      "mean_token_accuracy": 0.9451495166867971,
      "num_tokens": 64950541.0,
      "step": 8700
    },
    {
      "entropy": 1.861238856613636,
      "epoch": 0.581738034213009,
      "grad_norm": 0.27618059515953064,
      "learning_rate": 2.230666441902911e-05,
      "loss": 0.2093,
      "mean_token_accuracy": 0.9489861331880093,
      "num_tokens": 65025332.0,
      "step": 8710
    },
    {
      "entropy": 1.8650747872889042,
      "epoch": 0.5824059309227828,
      "grad_norm": 0.34626632928848267,
      "learning_rate": 2.224884703317385e-05,
      "loss": 0.2168,
      "mean_token_accuracy": 0.9464887123554945,
      "num_tokens": 65103052.0,
      "step": 8720
    },
    {
      "entropy": 1.8383357889950276,
      "epoch": 0.5830738276325567,
      "grad_norm": 0.26530808210372925,
      "learning_rate": 2.219104487759393e-05,
      "loss": 0.218,
      "mean_token_accuracy": 0.9458564352244139,
      "num_tokens": 65179654.0,
      "step": 8730
    },
    {
      "entropy": 1.8766178764402865,
      "epoch": 0.5837417243423304,
      "grad_norm": 0.3435424566268921,
      "learning_rate": 2.2133258266568054e-05,
      "loss": 0.2272,
      "mean_token_accuracy": 0.9450260020792485,
      "num_tokens": 65258149.0,
      "step": 8740
    },
    {
      "entropy": 1.886311811953783,
      "epoch": 0.5844096210521043,
      "grad_norm": 0.37311139702796936,
      "learning_rate": 2.207548751429042e-05,
      "loss": 0.2348,
      "mean_token_accuracy": 0.9424681596457958,
      "num_tokens": 65330251.0,
      "step": 8750
    },
    {
      "entropy": 1.8896283090114594,
      "epoch": 0.5850775177618781,
      "grad_norm": 0.3530047535896301,
      "learning_rate": 2.2017732934869e-05,
      "loss": 0.2324,
      "mean_token_accuracy": 0.9447159558534622,
      "num_tokens": 65405042.0,
      "step": 8760
    },
    {
      "entropy": 1.8673791393637658,
      "epoch": 0.585745414471652,
      "grad_norm": 0.28683504462242126,
      "learning_rate": 2.1959994842323817e-05,
      "loss": 0.207,
      "mean_token_accuracy": 0.9495283108204603,
      "num_tokens": 65482208.0,
      "step": 8770
    },
    {
      "entropy": 1.9134384475648403,
      "epoch": 0.5864133111814258,
      "grad_norm": 0.25487807393074036,
      "learning_rate": 2.1902273550585268e-05,
      "loss": 0.2235,
      "mean_token_accuracy": 0.9441673211753369,
      "num_tokens": 65553720.0,
      "step": 8780
    },
    {
      "entropy": 1.8667981952428818,
      "epoch": 0.5870812078911997,
      "grad_norm": 0.31790125370025635,
      "learning_rate": 2.184456937349239e-05,
      "loss": 0.2117,
      "mean_token_accuracy": 0.9489233203232288,
      "num_tokens": 65627833.0,
      "step": 8790
    },
    {
      "entropy": 1.8735299795866012,
      "epoch": 0.5877491046009734,
      "grad_norm": 0.2702594995498657,
      "learning_rate": 2.1786882624791184e-05,
      "loss": 0.2279,
      "mean_token_accuracy": 0.9412693019956351,
      "num_tokens": 65700658.0,
      "step": 8800
    },
    {
      "entropy": 1.8720247872173785,
      "epoch": 0.5884170013107473,
      "grad_norm": 0.33199939131736755,
      "learning_rate": 2.172921361813286e-05,
      "loss": 0.2355,
      "mean_token_accuracy": 0.943020797893405,
      "num_tokens": 65773535.0,
      "step": 8810
    },
    {
      "entropy": 1.93101811632514,
      "epoch": 0.5890848980205211,
      "grad_norm": 0.312505304813385,
      "learning_rate": 2.1671562667072192e-05,
      "loss": 0.2224,
      "mean_token_accuracy": 0.946207469701767,
      "num_tokens": 65846820.0,
      "step": 8820
    },
    {
      "entropy": 1.9124097622931004,
      "epoch": 0.589752794730295,
      "grad_norm": 0.2954193353652954,
      "learning_rate": 2.161393008506577e-05,
      "loss": 0.2277,
      "mean_token_accuracy": 0.9451969265937805,
      "num_tokens": 65921770.0,
      "step": 8830
    },
    {
      "entropy": 1.903800082951784,
      "epoch": 0.5904206914400688,
      "grad_norm": 0.3304347097873688,
      "learning_rate": 2.15563161854703e-05,
      "loss": 0.2292,
      "mean_token_accuracy": 0.9432186961174012,
      "num_tokens": 65997086.0,
      "step": 8840
    },
    {
      "entropy": 1.9177384577691554,
      "epoch": 0.5910885881498427,
      "grad_norm": 0.28481411933898926,
      "learning_rate": 2.1498721281540933e-05,
      "loss": 0.228,
      "mean_token_accuracy": 0.9438555292785168,
      "num_tokens": 66070291.0,
      "step": 8850
    },
    {
      "entropy": 1.905167955905199,
      "epoch": 0.5917564848596164,
      "grad_norm": 0.3228171169757843,
      "learning_rate": 2.1441145686429504e-05,
      "loss": 0.2188,
      "mean_token_accuracy": 0.946232233941555,
      "num_tokens": 66147691.0,
      "step": 8860
    },
    {
      "entropy": 1.888053610920906,
      "epoch": 0.5924243815693903,
      "grad_norm": 0.28190138936042786,
      "learning_rate": 2.13835897131829e-05,
      "loss": 0.2098,
      "mean_token_accuracy": 0.9469707697629929,
      "num_tokens": 66220838.0,
      "step": 8870
    },
    {
      "entropy": 1.9335898198187351,
      "epoch": 0.5930922782791641,
      "grad_norm": 0.2705141603946686,
      "learning_rate": 2.132605367474129e-05,
      "loss": 0.2328,
      "mean_token_accuracy": 0.9434020731598138,
      "num_tokens": 66296329.0,
      "step": 8880
    },
    {
      "entropy": 1.9194143183529377,
      "epoch": 0.593760174988938,
      "grad_norm": 0.2994906008243561,
      "learning_rate": 2.1268537883936475e-05,
      "loss": 0.2121,
      "mean_token_accuracy": 0.9483712319284677,
      "num_tokens": 66370210.0,
      "step": 8890
    },
    {
      "entropy": 1.8665772572159767,
      "epoch": 0.5944280716987118,
      "grad_norm": 0.2874792516231537,
      "learning_rate": 2.1211042653490153e-05,
      "loss": 0.2155,
      "mean_token_accuracy": 0.9487036559730768,
      "num_tokens": 66447967.0,
      "step": 8900
    },
    {
      "entropy": 1.8565380208194255,
      "epoch": 0.5950959684084857,
      "grad_norm": 0.2950595021247864,
      "learning_rate": 2.115356829601225e-05,
      "loss": 0.223,
      "mean_token_accuracy": 0.9464364275336266,
      "num_tokens": 66526240.0,
      "step": 8910
    },
    {
      "entropy": 1.9438115067780017,
      "epoch": 0.5957638651182594,
      "grad_norm": 0.2848196029663086,
      "learning_rate": 2.109611512399918e-05,
      "loss": 0.2436,
      "mean_token_accuracy": 0.9396346066147089,
      "num_tokens": 66599836.0,
      "step": 8920
    },
    {
      "entropy": 1.8883643224835396,
      "epoch": 0.5964317618280333,
      "grad_norm": 0.3058909773826599,
      "learning_rate": 2.1038683449832187e-05,
      "loss": 0.2224,
      "mean_token_accuracy": 0.945417657494545,
      "num_tokens": 66677293.0,
      "step": 8930
    },
    {
      "entropy": 1.9076268143951893,
      "epoch": 0.5970996585378071,
      "grad_norm": 0.24468903243541718,
      "learning_rate": 2.0981273585775614e-05,
      "loss": 0.2233,
      "mean_token_accuracy": 0.9450872246176004,
      "num_tokens": 66754489.0,
      "step": 8940
    },
    {
      "entropy": 1.8957035705447196,
      "epoch": 0.597767555247581,
      "grad_norm": 0.38515496253967285,
      "learning_rate": 2.092388584397523e-05,
      "loss": 0.2283,
      "mean_token_accuracy": 0.9423262067139149,
      "num_tokens": 66831836.0,
      "step": 8950
    },
    {
      "entropy": 1.9043849639594554,
      "epoch": 0.5984354519573548,
      "grad_norm": 0.2943609058856964,
      "learning_rate": 2.0866520536456525e-05,
      "loss": 0.2174,
      "mean_token_accuracy": 0.9477788418531418,
      "num_tokens": 66904632.0,
      "step": 8960
    },
    {
      "entropy": 1.9066061675548553,
      "epoch": 0.5991033486671287,
      "grad_norm": 0.27270668745040894,
      "learning_rate": 2.0809177975122995e-05,
      "loss": 0.2272,
      "mean_token_accuracy": 0.9435584831982851,
      "num_tokens": 66977461.0,
      "step": 8970
    },
    {
      "entropy": 1.8915106266736985,
      "epoch": 0.5997712453769024,
      "grad_norm": 0.38080450892448425,
      "learning_rate": 2.0751858471754477e-05,
      "loss": 0.2302,
      "mean_token_accuracy": 0.9453171331435442,
      "num_tokens": 67051417.0,
      "step": 8980
    },
    {
      "entropy": 1.8840370245277882,
      "epoch": 0.6004391420866763,
      "grad_norm": 0.2830585241317749,
      "learning_rate": 2.0694562338005424e-05,
      "loss": 0.2254,
      "mean_token_accuracy": 0.946152427047491,
      "num_tokens": 67126731.0,
      "step": 8990
    },
    {
      "entropy": 1.9158569969236852,
      "epoch": 0.6011070387964501,
      "grad_norm": 0.3386516869068146,
      "learning_rate": 2.063728988540324e-05,
      "loss": 0.2268,
      "mean_token_accuracy": 0.9459601316601038,
      "num_tokens": 67199356.0,
      "step": 9000
    },
    {
      "epoch": 0.6011070387964501,
      "eval_entropy": 1.9054617335557937,
      "eval_loss": 0.23829054832458496,
      "eval_mean_token_accuracy": 0.9458289430618286,
      "eval_num_tokens": 67199356.0,
      "eval_runtime": 552.7364,
      "eval_samples_per_second": 36.184,
      "eval_steps_per_second": 9.046,
      "step": 9000
    },
    {
      "entropy": 1.8930243149399757,
      "epoch": 0.601774935506224,
      "grad_norm": 0.3108629584312439,
      "learning_rate": 2.0580041425346568e-05,
      "loss": 0.2216,
      "mean_token_accuracy": 0.9460563812404871,
      "num_tokens": 67274255.0,
      "step": 9010
    },
    {
      "entropy": 1.8935005687177182,
      "epoch": 0.6024428322159978,
      "grad_norm": 0.30720093846321106,
      "learning_rate": 2.0522817269103586e-05,
      "loss": 0.2106,
      "mean_token_accuracy": 0.9482952229678631,
      "num_tokens": 67346807.0,
      "step": 9020
    },
    {
      "entropy": 1.8668569043278693,
      "epoch": 0.6031107289257717,
      "grad_norm": 0.31876540184020996,
      "learning_rate": 2.0465617727810345e-05,
      "loss": 0.2111,
      "mean_token_accuracy": 0.9487547673285007,
      "num_tokens": 67423456.0,
      "step": 9030
    },
    {
      "entropy": 1.8769634239375592,
      "epoch": 0.6037786256355455,
      "grad_norm": 0.36163222789764404,
      "learning_rate": 2.0408443112469055e-05,
      "loss": 0.2346,
      "mean_token_accuracy": 0.9438016287982464,
      "num_tokens": 67498565.0,
      "step": 9040
    },
    {
      "entropy": 1.9187280721962452,
      "epoch": 0.6044465223453193,
      "grad_norm": 0.30067041516304016,
      "learning_rate": 2.03512937339464e-05,
      "loss": 0.2119,
      "mean_token_accuracy": 0.9475815940648318,
      "num_tokens": 67575572.0,
      "step": 9050
    },
    {
      "entropy": 1.9150158524513246,
      "epoch": 0.6051144190550931,
      "grad_norm": 0.3035737872123718,
      "learning_rate": 2.029416990297184e-05,
      "loss": 0.2176,
      "mean_token_accuracy": 0.9453137967735529,
      "num_tokens": 67649413.0,
      "step": 9060
    },
    {
      "entropy": 1.9148891039192677,
      "epoch": 0.605782315764867,
      "grad_norm": 0.3508419990539551,
      "learning_rate": 2.0237071930135946e-05,
      "loss": 0.2106,
      "mean_token_accuracy": 0.947960839420557,
      "num_tokens": 67723583.0,
      "step": 9070
    },
    {
      "entropy": 1.8879984691739082,
      "epoch": 0.6064502124746408,
      "grad_norm": 0.2787244915962219,
      "learning_rate": 2.0180000125888686e-05,
      "loss": 0.2144,
      "mean_token_accuracy": 0.947710957750678,
      "num_tokens": 67798263.0,
      "step": 9080
    },
    {
      "entropy": 1.9052270293235778,
      "epoch": 0.6071181091844147,
      "grad_norm": 0.35291776061058044,
      "learning_rate": 2.012295480053774e-05,
      "loss": 0.2167,
      "mean_token_accuracy": 0.946917486935854,
      "num_tokens": 67875758.0,
      "step": 9090
    },
    {
      "entropy": 1.926875291019678,
      "epoch": 0.6077860058941885,
      "grad_norm": 0.3031468689441681,
      "learning_rate": 2.0065936264246838e-05,
      "loss": 0.2251,
      "mean_token_accuracy": 0.9462259538471699,
      "num_tokens": 67951984.0,
      "step": 9100
    },
    {
      "entropy": 1.90597188398242,
      "epoch": 0.6084539026039623,
      "grad_norm": 0.34051406383514404,
      "learning_rate": 2.0008944827034016e-05,
      "loss": 0.2223,
      "mean_token_accuracy": 0.9454433262348175,
      "num_tokens": 68025754.0,
      "step": 9110
    },
    {
      "entropy": 1.891847651451826,
      "epoch": 0.6091217993137361,
      "grad_norm": 0.28218090534210205,
      "learning_rate": 1.995198079877e-05,
      "loss": 0.224,
      "mean_token_accuracy": 0.9456320758908987,
      "num_tokens": 68102227.0,
      "step": 9120
    },
    {
      "entropy": 1.906551606953144,
      "epoch": 0.60978969602351,
      "grad_norm": 0.32842859625816345,
      "learning_rate": 1.9895044489176478e-05,
      "loss": 0.2116,
      "mean_token_accuracy": 0.9474180229008198,
      "num_tokens": 68178543.0,
      "step": 9130
    },
    {
      "entropy": 1.9225084953010083,
      "epoch": 0.6104575927332838,
      "grad_norm": 0.3469751179218292,
      "learning_rate": 1.983813620782445e-05,
      "loss": 0.2199,
      "mean_token_accuracy": 0.9457121465355158,
      "num_tokens": 68255718.0,
      "step": 9140
    },
    {
      "entropy": 1.9057083517313003,
      "epoch": 0.6111254894430577,
      "grad_norm": 0.26819270849227905,
      "learning_rate": 1.978125626413248e-05,
      "loss": 0.2226,
      "mean_token_accuracy": 0.9458238828927279,
      "num_tokens": 68328639.0,
      "step": 9150
    },
    {
      "entropy": 1.90034921169281,
      "epoch": 0.6117933861528315,
      "grad_norm": 0.3128930628299713,
      "learning_rate": 1.9724404967365102e-05,
      "loss": 0.2177,
      "mean_token_accuracy": 0.9472173463553191,
      "num_tokens": 68401721.0,
      "step": 9160
    },
    {
      "entropy": 1.875298696011305,
      "epoch": 0.6124612828626053,
      "grad_norm": 0.33448532223701477,
      "learning_rate": 1.9667582626631046e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9450565122067929,
      "num_tokens": 68477289.0,
      "step": 9170
    },
    {
      "entropy": 1.895997653156519,
      "epoch": 0.6131291795723791,
      "grad_norm": 0.3845632076263428,
      "learning_rate": 1.9610789550881645e-05,
      "loss": 0.2358,
      "mean_token_accuracy": 0.9430301874876023,
      "num_tokens": 68550763.0,
      "step": 9180
    },
    {
      "entropy": 1.9016799658536911,
      "epoch": 0.613797076282153,
      "grad_norm": 0.32598045468330383,
      "learning_rate": 1.9554026048909092e-05,
      "loss": 0.2117,
      "mean_token_accuracy": 0.9479414414614439,
      "num_tokens": 68625310.0,
      "step": 9190
    },
    {
      "entropy": 1.8913936592638492,
      "epoch": 0.6144649729919268,
      "grad_norm": 0.34851107001304626,
      "learning_rate": 1.9497292429344784e-05,
      "loss": 0.2198,
      "mean_token_accuracy": 0.945693937316537,
      "num_tokens": 68703488.0,
      "step": 9200
    },
    {
      "entropy": 1.8864638805389404,
      "epoch": 0.6151328697017007,
      "grad_norm": 0.28284165263175964,
      "learning_rate": 1.9440589000657648e-05,
      "loss": 0.2304,
      "mean_token_accuracy": 0.9435976024717092,
      "num_tokens": 68779271.0,
      "step": 9210
    },
    {
      "entropy": 1.8724992901086808,
      "epoch": 0.6158007664114745,
      "grad_norm": 0.2989194691181183,
      "learning_rate": 1.938391607115246e-05,
      "loss": 0.223,
      "mean_token_accuracy": 0.9453818060457706,
      "num_tokens": 68853160.0,
      "step": 9220
    },
    {
      "entropy": 1.8866638407111167,
      "epoch": 0.6164686631212483,
      "grad_norm": 0.30060428380966187,
      "learning_rate": 1.9327273948968175e-05,
      "loss": 0.2109,
      "mean_token_accuracy": 0.9489407524466514,
      "num_tokens": 68926654.0,
      "step": 9230
    },
    {
      "entropy": 1.8895188309252262,
      "epoch": 0.6171365598310221,
      "grad_norm": 0.2604714632034302,
      "learning_rate": 1.927066294207621e-05,
      "loss": 0.2372,
      "mean_token_accuracy": 0.9458476543426514,
      "num_tokens": 69000867.0,
      "step": 9240
    },
    {
      "entropy": 1.9009686149656773,
      "epoch": 0.617804456540796,
      "grad_norm": 0.3436797261238098,
      "learning_rate": 1.9214083358278846e-05,
      "loss": 0.2339,
      "mean_token_accuracy": 0.9434888288378716,
      "num_tokens": 69074583.0,
      "step": 9250
    },
    {
      "entropy": 1.8971208855509758,
      "epoch": 0.6184723532505698,
      "grad_norm": 0.2961723506450653,
      "learning_rate": 1.9157535505207495e-05,
      "loss": 0.2172,
      "mean_token_accuracy": 0.9440949592739344,
      "num_tokens": 69149262.0,
      "step": 9260
    },
    {
      "entropy": 1.8461739234626293,
      "epoch": 0.6191402499603437,
      "grad_norm": 0.3290032148361206,
      "learning_rate": 1.9101019690321035e-05,
      "loss": 0.2203,
      "mean_token_accuracy": 0.9460521291941404,
      "num_tokens": 69225923.0,
      "step": 9270
    },
    {
      "entropy": 1.8542075894773007,
      "epoch": 0.6198081466701175,
      "grad_norm": 0.26672402024269104,
      "learning_rate": 1.9044536220904162e-05,
      "loss": 0.2113,
      "mean_token_accuracy": 0.9467404522001743,
      "num_tokens": 69301932.0,
      "step": 9280
    },
    {
      "entropy": 1.8576584227383137,
      "epoch": 0.6204760433798913,
      "grad_norm": 0.3438381254673004,
      "learning_rate": 1.8988085404065695e-05,
      "loss": 0.2033,
      "mean_token_accuracy": 0.9490065883845091,
      "num_tokens": 69374963.0,
      "step": 9290
    },
    {
      "entropy": 1.8618731454014779,
      "epoch": 0.6211439400896651,
      "grad_norm": 0.2744857370853424,
      "learning_rate": 1.8931667546736925e-05,
      "loss": 0.2213,
      "mean_token_accuracy": 0.9437948763370514,
      "num_tokens": 69449043.0,
      "step": 9300
    },
    {
      "entropy": 1.8539308667182923,
      "epoch": 0.621811836799439,
      "grad_norm": 0.3019644618034363,
      "learning_rate": 1.8875282955669926e-05,
      "loss": 0.2193,
      "mean_token_accuracy": 0.9427087072283029,
      "num_tokens": 69524625.0,
      "step": 9310
    },
    {
      "entropy": 1.893137027323246,
      "epoch": 0.6224797335092128,
      "grad_norm": 0.2724396884441376,
      "learning_rate": 1.8818931937435908e-05,
      "loss": 0.2431,
      "mean_token_accuracy": 0.94328334890306,
      "num_tokens": 69597493.0,
      "step": 9320
    },
    {
      "entropy": 1.8799325823783875,
      "epoch": 0.6231476302189867,
      "grad_norm": 0.27898383140563965,
      "learning_rate": 1.876261479842355e-05,
      "loss": 0.2278,
      "mean_token_accuracy": 0.9454366229474545,
      "num_tokens": 69673405.0,
      "step": 9330
    },
    {
      "entropy": 1.8927228800952434,
      "epoch": 0.6238155269287605,
      "grad_norm": 0.2509833872318268,
      "learning_rate": 1.8706331844837298e-05,
      "loss": 0.2055,
      "mean_token_accuracy": 0.9507538910955191,
      "num_tokens": 69748795.0,
      "step": 9340
    },
    {
      "entropy": 1.9088469952344895,
      "epoch": 0.6244834236385343,
      "grad_norm": 0.33189061284065247,
      "learning_rate": 1.8650083382695736e-05,
      "loss": 0.2229,
      "mean_token_accuracy": 0.9470415290445089,
      "num_tokens": 69822471.0,
      "step": 9350
    },
    {
      "entropy": 1.9100259378552438,
      "epoch": 0.6251513203483081,
      "grad_norm": 0.3775787055492401,
      "learning_rate": 1.859386971782994e-05,
      "loss": 0.2318,
      "mean_token_accuracy": 0.9430906638503075,
      "num_tokens": 69896670.0,
      "step": 9360
    },
    {
      "entropy": 1.8997727893292904,
      "epoch": 0.625819217058082,
      "grad_norm": 0.29177314043045044,
      "learning_rate": 1.8537691155881742e-05,
      "loss": 0.2178,
      "mean_token_accuracy": 0.9444642275571823,
      "num_tokens": 69973519.0,
      "step": 9370
    },
    {
      "entropy": 1.8985559575259685,
      "epoch": 0.6264871137678558,
      "grad_norm": 0.3505083918571472,
      "learning_rate": 1.8481548002302158e-05,
      "loss": 0.2292,
      "mean_token_accuracy": 0.9444281063973904,
      "num_tokens": 70048731.0,
      "step": 9380
    },
    {
      "entropy": 1.874289197474718,
      "epoch": 0.6271550104776297,
      "grad_norm": 0.3835582733154297,
      "learning_rate": 1.8425440562349657e-05,
      "loss": 0.2126,
      "mean_token_accuracy": 0.9475785404443741,
      "num_tokens": 70126718.0,
      "step": 9390
    },
    {
      "entropy": 1.8724834360182285,
      "epoch": 0.6278229071874035,
      "grad_norm": 0.32895326614379883,
      "learning_rate": 1.8369369141088532e-05,
      "loss": 0.2128,
      "mean_token_accuracy": 0.9477534204721451,
      "num_tokens": 70203946.0,
      "step": 9400
    },
    {
      "entropy": 1.9146189346909523,
      "epoch": 0.6284908038971773,
      "grad_norm": 0.3408152461051941,
      "learning_rate": 1.8313334043387253e-05,
      "loss": 0.2377,
      "mean_token_accuracy": 0.9416464570909738,
      "num_tokens": 70277262.0,
      "step": 9410
    },
    {
      "entropy": 1.90368904620409,
      "epoch": 0.6291587006069511,
      "grad_norm": 0.34073203802108765,
      "learning_rate": 1.825733557391677e-05,
      "loss": 0.2271,
      "mean_token_accuracy": 0.9401124123483896,
      "num_tokens": 70351864.0,
      "step": 9420
    },
    {
      "entropy": 1.8903987191617488,
      "epoch": 0.629826597316725,
      "grad_norm": 0.3101690411567688,
      "learning_rate": 1.820137403714889e-05,
      "loss": 0.2162,
      "mean_token_accuracy": 0.9469412431120873,
      "num_tokens": 70424867.0,
      "step": 9430
    },
    {
      "entropy": 1.9022814162075519,
      "epoch": 0.6304944940264988,
      "grad_norm": 0.3661165237426758,
      "learning_rate": 1.8145449737354623e-05,
      "loss": 0.2246,
      "mean_token_accuracy": 0.9438300043344497,
      "num_tokens": 70500190.0,
      "step": 9440
    },
    {
      "entropy": 1.8979595214128495,
      "epoch": 0.6311623907362727,
      "grad_norm": 0.36789000034332275,
      "learning_rate": 1.8089562978602493e-05,
      "loss": 0.223,
      "mean_token_accuracy": 0.9425424974411726,
      "num_tokens": 70576208.0,
      "step": 9450
    },
    {
      "entropy": 1.8826049737632276,
      "epoch": 0.6318302874460465,
      "grad_norm": 0.28903254866600037,
      "learning_rate": 1.8033714064756934e-05,
      "loss": 0.2207,
      "mean_token_accuracy": 0.9426618255674839,
      "num_tokens": 70648440.0,
      "step": 9460
    },
    {
      "entropy": 1.895417645573616,
      "epoch": 0.6324981841558203,
      "grad_norm": 0.3099706470966339,
      "learning_rate": 1.7977903299476584e-05,
      "loss": 0.2186,
      "mean_token_accuracy": 0.9456299874931574,
      "num_tokens": 70723830.0,
      "step": 9470
    },
    {
      "entropy": 1.8829485289752483,
      "epoch": 0.6331660808655941,
      "grad_norm": 0.309557169675827,
      "learning_rate": 1.7922130986212696e-05,
      "loss": 0.222,
      "mean_token_accuracy": 0.9443500895053148,
      "num_tokens": 70798345.0,
      "step": 9480
    },
    {
      "entropy": 1.899361400306225,
      "epoch": 0.633833977575368,
      "grad_norm": 0.29913508892059326,
      "learning_rate": 1.786639742820741e-05,
      "loss": 0.2135,
      "mean_token_accuracy": 0.9478815961629152,
      "num_tokens": 70874617.0,
      "step": 9490
    },
    {
      "entropy": 1.8868928365409374,
      "epoch": 0.6345018742851418,
      "grad_norm": 0.24733923375606537,
      "learning_rate": 1.781070292849218e-05,
      "loss": 0.2152,
      "mean_token_accuracy": 0.9475489914417267,
      "num_tokens": 70949437.0,
      "step": 9500
    },
    {
      "entropy": 1.8958236001431943,
      "epoch": 0.6351697709949157,
      "grad_norm": 0.302013635635376,
      "learning_rate": 1.77550477898861e-05,
      "loss": 0.2297,
      "mean_token_accuracy": 0.9406222008168698,
      "num_tokens": 71023015.0,
      "step": 9510
    },
    {
      "entropy": 1.8743866764008998,
      "epoch": 0.6358376677046895,
      "grad_norm": 0.29208335280418396,
      "learning_rate": 1.7699432314994226e-05,
      "loss": 0.2089,
      "mean_token_accuracy": 0.946915852278471,
      "num_tokens": 71098694.0,
      "step": 9520
    },
    {
      "entropy": 1.8999532632529736,
      "epoch": 0.6365055644144633,
      "grad_norm": 0.27738434076309204,
      "learning_rate": 1.764385680620597e-05,
      "loss": 0.2275,
      "mean_token_accuracy": 0.9449752297252416,
      "num_tokens": 71173168.0,
      "step": 9530
    },
    {
      "entropy": 1.8916003756225108,
      "epoch": 0.6371734611242371,
      "grad_norm": 0.36834222078323364,
      "learning_rate": 1.7588321565693444e-05,
      "loss": 0.2174,
      "mean_token_accuracy": 0.9484237350523472,
      "num_tokens": 71247516.0,
      "step": 9540
    },
    {
      "entropy": 1.8881634168326855,
      "epoch": 0.637841357834011,
      "grad_norm": 0.36912626028060913,
      "learning_rate": 1.7532826895409815e-05,
      "loss": 0.2359,
      "mean_token_accuracy": 0.9426255002617836,
      "num_tokens": 71321309.0,
      "step": 9550
    },
    {
      "entropy": 1.9114818081259728,
      "epoch": 0.6385092545437848,
      "grad_norm": 0.32032787799835205,
      "learning_rate": 1.7477373097087664e-05,
      "loss": 0.2164,
      "mean_token_accuracy": 0.9461173076182604,
      "num_tokens": 71395536.0,
      "step": 9560
    },
    {
      "entropy": 1.8963179916143418,
      "epoch": 0.6391771512535587,
      "grad_norm": 0.30320611596107483,
      "learning_rate": 1.7421960472237325e-05,
      "loss": 0.2284,
      "mean_token_accuracy": 0.9446917619556189,
      "num_tokens": 71469799.0,
      "step": 9570
    },
    {
      "entropy": 1.8971133276820182,
      "epoch": 0.6398450479633325,
      "grad_norm": 0.3104685842990875,
      "learning_rate": 1.7366589322145292e-05,
      "loss": 0.2192,
      "mean_token_accuracy": 0.9449749819934368,
      "num_tokens": 71543703.0,
      "step": 9580
    },
    {
      "entropy": 1.8932307437062263,
      "epoch": 0.6405129446731063,
      "grad_norm": 0.28371432423591614,
      "learning_rate": 1.7311259947872564e-05,
      "loss": 0.2338,
      "mean_token_accuracy": 0.9452110283076763,
      "num_tokens": 71617039.0,
      "step": 9590
    },
    {
      "entropy": 1.8857156790792942,
      "epoch": 0.6411808413828801,
      "grad_norm": 0.3060430884361267,
      "learning_rate": 1.725597265025296e-05,
      "loss": 0.2062,
      "mean_token_accuracy": 0.9494408696889878,
      "num_tokens": 71691536.0,
      "step": 9600
    },
    {
      "entropy": 1.914458305388689,
      "epoch": 0.641848738092654,
      "grad_norm": 0.2998836040496826,
      "learning_rate": 1.7200727729891557e-05,
      "loss": 0.2105,
      "mean_token_accuracy": 0.9502223901450634,
      "num_tokens": 71770840.0,
      "step": 9610
    },
    {
      "entropy": 1.9375434875488282,
      "epoch": 0.6425166348024278,
      "grad_norm": 0.33531689643859863,
      "learning_rate": 1.7145525487162996e-05,
      "loss": 0.2405,
      "mean_token_accuracy": 0.9410932116210461,
      "num_tokens": 71844026.0,
      "step": 9620
    },
    {
      "entropy": 1.900812315195799,
      "epoch": 0.6431845315122017,
      "grad_norm": 0.2553114891052246,
      "learning_rate": 1.709036622220989e-05,
      "loss": 0.2257,
      "mean_token_accuracy": 0.9443730093538761,
      "num_tokens": 71918643.0,
      "step": 9630
    },
    {
      "entropy": 1.9463625065982342,
      "epoch": 0.6438524282219755,
      "grad_norm": 0.37230777740478516,
      "learning_rate": 1.703525023494118e-05,
      "loss": 0.2435,
      "mean_token_accuracy": 0.9423607751727104,
      "num_tokens": 71990166.0,
      "step": 9640
    },
    {
      "entropy": 1.9112353555858135,
      "epoch": 0.6445203249317493,
      "grad_norm": 0.2643808126449585,
      "learning_rate": 1.6980177825030468e-05,
      "loss": 0.2187,
      "mean_token_accuracy": 0.9439635403454304,
      "num_tokens": 72063734.0,
      "step": 9650
    },
    {
      "entropy": 1.9036917440593242,
      "epoch": 0.6451882216415231,
      "grad_norm": 0.32149404287338257,
      "learning_rate": 1.6925149291914465e-05,
      "loss": 0.2355,
      "mean_token_accuracy": 0.942261739820242,
      "num_tokens": 72135959.0,
      "step": 9660
    },
    {
      "entropy": 1.8577286154031754,
      "epoch": 0.645856118351297,
      "grad_norm": 0.32772204279899597,
      "learning_rate": 1.6870164934791278e-05,
      "loss": 0.2036,
      "mean_token_accuracy": 0.9486184068024158,
      "num_tokens": 72211938.0,
      "step": 9670
    },
    {
      "entropy": 1.913424578309059,
      "epoch": 0.6465240150610708,
      "grad_norm": 0.32256755232810974,
      "learning_rate": 1.681522505261885e-05,
      "loss": 0.2432,
      "mean_token_accuracy": 0.942617866396904,
      "num_tokens": 72284128.0,
      "step": 9680
    },
    {
      "entropy": 1.9126008175313474,
      "epoch": 0.6471919117708447,
      "grad_norm": 0.33424055576324463,
      "learning_rate": 1.6760329944113286e-05,
      "loss": 0.2256,
      "mean_token_accuracy": 0.9433588128536939,
      "num_tokens": 72356840.0,
      "step": 9690
    },
    {
      "entropy": 1.9473415218293666,
      "epoch": 0.6478598084806185,
      "grad_norm": 0.33258989453315735,
      "learning_rate": 1.6705479907747266e-05,
      "loss": 0.2369,
      "mean_token_accuracy": 0.9426125854253768,
      "num_tokens": 72430720.0,
      "step": 9700
    },
    {
      "entropy": 1.8974012099206448,
      "epoch": 0.6485277051903923,
      "grad_norm": 0.29704323410987854,
      "learning_rate": 1.6650675241748402e-05,
      "loss": 0.2324,
      "mean_token_accuracy": 0.9434922259300947,
      "num_tokens": 72503072.0,
      "step": 9710
    },
    {
      "entropy": 1.9120988190174102,
      "epoch": 0.6491956019001661,
      "grad_norm": 0.2917936146259308,
      "learning_rate": 1.6595916244097618e-05,
      "loss": 0.2373,
      "mean_token_accuracy": 0.9428652595728636,
      "num_tokens": 72579059.0,
      "step": 9720
    },
    {
      "entropy": 1.9111275486648083,
      "epoch": 0.64986349860994,
      "grad_norm": 0.32185760140419006,
      "learning_rate": 1.6541203212527534e-05,
      "loss": 0.2209,
      "mean_token_accuracy": 0.9451362181454896,
      "num_tokens": 72656873.0,
      "step": 9730
    },
    {
      "entropy": 1.9101318418979645,
      "epoch": 0.6505313953197138,
      "grad_norm": 0.33374306559562683,
      "learning_rate": 1.648653644452084e-05,
      "loss": 0.2309,
      "mean_token_accuracy": 0.9426966361701489,
      "num_tokens": 72731989.0,
      "step": 9740
    },
    {
      "entropy": 1.8926914855837822,
      "epoch": 0.6511992920294877,
      "grad_norm": 0.30848056077957153,
      "learning_rate": 1.6431916237308703e-05,
      "loss": 0.2193,
      "mean_token_accuracy": 0.9459914352744818,
      "num_tokens": 72804092.0,
      "step": 9750
    },
    {
      "entropy": 1.8973691657185554,
      "epoch": 0.6518671887392615,
      "grad_norm": 0.32435324788093567,
      "learning_rate": 1.6377342887869105e-05,
      "loss": 0.2523,
      "mean_token_accuracy": 0.941319564729929,
      "num_tokens": 72876634.0,
      "step": 9760
    },
    {
      "entropy": 1.892369518429041,
      "epoch": 0.6525350854490353,
      "grad_norm": 0.28423383831977844,
      "learning_rate": 1.6322816692925273e-05,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9461352091282607,
      "num_tokens": 72951549.0,
      "step": 9770
    },
    {
      "entropy": 1.917740623652935,
      "epoch": 0.6532029821588091,
      "grad_norm": 0.3266948461532593,
      "learning_rate": 1.6268337948944054e-05,
      "loss": 0.2308,
      "mean_token_accuracy": 0.9406094893813133,
      "num_tokens": 73023602.0,
      "step": 9780
    },
    {
      "entropy": 1.922014880180359,
      "epoch": 0.653870878868583,
      "grad_norm": 0.35251879692077637,
      "learning_rate": 1.6213906952134283e-05,
      "loss": 0.2335,
      "mean_token_accuracy": 0.9410470966249704,
      "num_tokens": 73094850.0,
      "step": 9790
    },
    {
      "entropy": 1.891479628533125,
      "epoch": 0.6545387755783568,
      "grad_norm": 0.25936606526374817,
      "learning_rate": 1.6159523998445182e-05,
      "loss": 0.2125,
      "mean_token_accuracy": 0.9463196698576212,
      "num_tokens": 73168575.0,
      "step": 9800
    },
    {
      "entropy": 1.883493823558092,
      "epoch": 0.6552066722881307,
      "grad_norm": 0.3399842083454132,
      "learning_rate": 1.6105189383564782e-05,
      "loss": 0.222,
      "mean_token_accuracy": 0.9449453841894865,
      "num_tokens": 73242763.0,
      "step": 9810
    },
    {
      "entropy": 1.8951681099832058,
      "epoch": 0.6558745689979045,
      "grad_norm": 0.34557831287384033,
      "learning_rate": 1.6050903402918253e-05,
      "loss": 0.2298,
      "mean_token_accuracy": 0.9446959998458624,
      "num_tokens": 73317960.0,
      "step": 9820
    },
    {
      "entropy": 1.8923872597515583,
      "epoch": 0.6565424657076783,
      "grad_norm": 0.3376432955265045,
      "learning_rate": 1.5996666351666364e-05,
      "loss": 0.2068,
      "mean_token_accuracy": 0.9481453604996204,
      "num_tokens": 73391636.0,
      "step": 9830
    },
    {
      "entropy": 1.891785554587841,
      "epoch": 0.6572103624174521,
      "grad_norm": 0.30720290541648865,
      "learning_rate": 1.5942478524703838e-05,
      "loss": 0.2323,
      "mean_token_accuracy": 0.9421215292066336,
      "num_tokens": 73464577.0,
      "step": 9840
    },
    {
      "entropy": 1.8919872418045998,
      "epoch": 0.657878259127226,
      "grad_norm": 0.26905012130737305,
      "learning_rate": 1.5888340216657754e-05,
      "loss": 0.2087,
      "mean_token_accuracy": 0.9481104116886854,
      "num_tokens": 73539071.0,
      "step": 9850
    },
    {
      "entropy": 1.9074813894927503,
      "epoch": 0.6585461558369998,
      "grad_norm": 0.28303661942481995,
      "learning_rate": 1.5834251721885955e-05,
      "loss": 0.2331,
      "mean_token_accuracy": 0.9425707191228867,
      "num_tokens": 73611514.0,
      "step": 9860
    },
    {
      "entropy": 1.8842093892395497,
      "epoch": 0.6592140525467737,
      "grad_norm": 0.3321745991706848,
      "learning_rate": 1.578021333447543e-05,
      "loss": 0.2116,
      "mean_token_accuracy": 0.9476104851812124,
      "num_tokens": 73685360.0,
      "step": 9870
    },
    {
      "entropy": 1.9000260949134826,
      "epoch": 0.6598819492565475,
      "grad_norm": 0.2837125062942505,
      "learning_rate": 1.572622534824075e-05,
      "loss": 0.2301,
      "mean_token_accuracy": 0.9446627900004387,
      "num_tokens": 73758057.0,
      "step": 9880
    },
    {
      "entropy": 1.9003470219671725,
      "epoch": 0.6605498459663213,
      "grad_norm": 0.26702362298965454,
      "learning_rate": 1.567228805672242e-05,
      "loss": 0.209,
      "mean_token_accuracy": 0.9485840313136578,
      "num_tokens": 73833189.0,
      "step": 9890
    },
    {
      "entropy": 1.8759419411420821,
      "epoch": 0.6612177426760951,
      "grad_norm": 0.2915317416191101,
      "learning_rate": 1.5618401753185338e-05,
      "loss": 0.2112,
      "mean_token_accuracy": 0.9462038677185773,
      "num_tokens": 73907938.0,
      "step": 9900
    },
    {
      "entropy": 1.897911613434553,
      "epoch": 0.661885639385869,
      "grad_norm": 0.34235090017318726,
      "learning_rate": 1.5564566730617153e-05,
      "loss": 0.2359,
      "mean_token_accuracy": 0.9429119873791934,
      "num_tokens": 73979605.0,
      "step": 9910
    },
    {
      "entropy": 1.9387646824121476,
      "epoch": 0.6625535360956428,
      "grad_norm": 0.3523752689361572,
      "learning_rate": 1.5510783281726693e-05,
      "loss": 0.2124,
      "mean_token_accuracy": 0.9452576469630003,
      "num_tokens": 74053270.0,
      "step": 9920
    },
    {
      "entropy": 1.879902134090662,
      "epoch": 0.6632214328054167,
      "grad_norm": 0.29795941710472107,
      "learning_rate": 1.5457051698942387e-05,
      "loss": 0.2144,
      "mean_token_accuracy": 0.9486784808337688,
      "num_tokens": 74128527.0,
      "step": 9930
    },
    {
      "entropy": 1.8928095526993274,
      "epoch": 0.6638893295151905,
      "grad_norm": 0.3268255293369293,
      "learning_rate": 1.540337227441064e-05,
      "loss": 0.207,
      "mean_token_accuracy": 0.9484345391392708,
      "num_tokens": 74203714.0,
      "step": 9940
    },
    {
      "entropy": 1.9202770486474037,
      "epoch": 0.6645572262249643,
      "grad_norm": 0.3489333391189575,
      "learning_rate": 1.5349745299994285e-05,
      "loss": 0.2509,
      "mean_token_accuracy": 0.9413472600281239,
      "num_tokens": 74279169.0,
      "step": 9950
    },
    {
      "entropy": 1.9356919780373574,
      "epoch": 0.6652251229347381,
      "grad_norm": 0.32012757658958435,
      "learning_rate": 1.529617106727097e-05,
      "loss": 0.2331,
      "mean_token_accuracy": 0.9424417037516832,
      "num_tokens": 74353994.0,
      "step": 9960
    },
    {
      "entropy": 1.931555875390768,
      "epoch": 0.665893019644512,
      "grad_norm": 0.31553220748901367,
      "learning_rate": 1.5242649867531566e-05,
      "loss": 0.223,
      "mean_token_accuracy": 0.9461064361035824,
      "num_tokens": 74428290.0,
      "step": 9970
    },
    {
      "entropy": 1.8998085968196392,
      "epoch": 0.6665609163542858,
      "grad_norm": 0.33924880623817444,
      "learning_rate": 1.5189181991778614e-05,
      "loss": 0.2085,
      "mean_token_accuracy": 0.9488823279738426,
      "num_tokens": 74503465.0,
      "step": 9980
    },
    {
      "entropy": 1.8661912620067596,
      "epoch": 0.6672288130640597,
      "grad_norm": 0.31078508496284485,
      "learning_rate": 1.513576773072471e-05,
      "loss": 0.2265,
      "mean_token_accuracy": 0.9466509778052569,
      "num_tokens": 74579565.0,
      "step": 9990
    },
    {
      "entropy": 1.8821153879165649,
      "epoch": 0.6678967097738335,
      "grad_norm": 0.31515589356422424,
      "learning_rate": 1.5082407374790951e-05,
      "loss": 0.2144,
      "mean_token_accuracy": 0.9475899521261454,
      "num_tokens": 74654639.0,
      "step": 10000
    },
    {
      "epoch": 0.6678967097738335,
      "eval_entropy": 1.8767498084783554,
      "eval_loss": 0.236663356423378,
      "eval_mean_token_accuracy": 0.9460465997934342,
      "eval_num_tokens": 74654639.0,
      "eval_runtime": 552.1022,
      "eval_samples_per_second": 36.225,
      "eval_steps_per_second": 9.056,
      "step": 10000
    },
    {
      "entropy": 1.8770928032696248,
      "epoch": 0.6685646064836073,
      "grad_norm": 0.2725706994533539,
      "learning_rate": 1.5029101214105334e-05,
      "loss": 0.22,
      "mean_token_accuracy": 0.9455895718187094,
      "num_tokens": 74731174.0,
      "step": 10010
    },
    {
      "entropy": 1.8732770144939423,
      "epoch": 0.6692325031933811,
      "grad_norm": 0.32018646597862244,
      "learning_rate": 1.4975849538501205e-05,
      "loss": 0.1976,
      "mean_token_accuracy": 0.9508919149637223,
      "num_tokens": 74808245.0,
      "step": 10020
    },
    {
      "entropy": 1.8560627780854702,
      "epoch": 0.669900399903155,
      "grad_norm": 0.3270680010318756,
      "learning_rate": 1.4922652637515635e-05,
      "loss": 0.2318,
      "mean_token_accuracy": 0.9454466596245765,
      "num_tokens": 74883563.0,
      "step": 10030
    },
    {
      "entropy": 1.8667067840695382,
      "epoch": 0.6705682966129288,
      "grad_norm": 0.29106202721595764,
      "learning_rate": 1.4869510800387918e-05,
      "loss": 0.2131,
      "mean_token_accuracy": 0.9496993191540242,
      "num_tokens": 74959220.0,
      "step": 10040
    },
    {
      "entropy": 1.8844712659716607,
      "epoch": 0.6712361933227027,
      "grad_norm": 0.32260608673095703,
      "learning_rate": 1.481642431605792e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9459765709936618,
      "num_tokens": 75032123.0,
      "step": 10050
    },
    {
      "entropy": 1.8944692999124526,
      "epoch": 0.6719040900324765,
      "grad_norm": 0.269023060798645,
      "learning_rate": 1.4763393473164577e-05,
      "loss": 0.2193,
      "mean_token_accuracy": 0.946693105995655,
      "num_tokens": 75109592.0,
      "step": 10060
    },
    {
      "entropy": 1.8866144388914108,
      "epoch": 0.6725719867422503,
      "grad_norm": 0.31930679082870483,
      "learning_rate": 1.471041856004427e-05,
      "loss": 0.2273,
      "mean_token_accuracy": 0.9449324276298284,
      "num_tokens": 75185468.0,
      "step": 10070
    },
    {
      "entropy": 1.9378182969987392,
      "epoch": 0.6732398834520241,
      "grad_norm": 0.30855849385261536,
      "learning_rate": 1.4657499864729299e-05,
      "loss": 0.2447,
      "mean_token_accuracy": 0.9389895871281624,
      "num_tokens": 75260197.0,
      "step": 10080
    },
    {
      "entropy": 1.9235576070845126,
      "epoch": 0.673907780161798,
      "grad_norm": 0.36566832661628723,
      "learning_rate": 1.460463767494629e-05,
      "loss": 0.2222,
      "mean_token_accuracy": 0.9441259548068046,
      "num_tokens": 75333137.0,
      "step": 10090
    },
    {
      "entropy": 1.8855039738118649,
      "epoch": 0.6745756768715718,
      "grad_norm": 0.24700982868671417,
      "learning_rate": 1.4551832278114638e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9484541725367308,
      "num_tokens": 75409681.0,
      "step": 10100
    },
    {
      "entropy": 1.8987921811640263,
      "epoch": 0.6752435735813457,
      "grad_norm": 0.279258131980896,
      "learning_rate": 1.4499083961344961e-05,
      "loss": 0.2014,
      "mean_token_accuracy": 0.9481212608516216,
      "num_tokens": 75483582.0,
      "step": 10110
    },
    {
      "entropy": 1.9032227523624896,
      "epoch": 0.6759114702911195,
      "grad_norm": 0.32155728340148926,
      "learning_rate": 1.4446393011437525e-05,
      "loss": 0.2219,
      "mean_token_accuracy": 0.9450641572475433,
      "num_tokens": 75559399.0,
      "step": 10120
    },
    {
      "entropy": 1.8590004473924637,
      "epoch": 0.6765793670008933,
      "grad_norm": 0.32483652234077454,
      "learning_rate": 1.4393759714880644e-05,
      "loss": 0.2175,
      "mean_token_accuracy": 0.9464751157909632,
      "num_tokens": 75633898.0,
      "step": 10130
    },
    {
      "entropy": 1.8997207395732403,
      "epoch": 0.6772472637106671,
      "grad_norm": 0.27143311500549316,
      "learning_rate": 1.4341184357849213e-05,
      "loss": 0.2211,
      "mean_token_accuracy": 0.9461022023111582,
      "num_tokens": 75708829.0,
      "step": 10140
    },
    {
      "entropy": 1.9080256782472134,
      "epoch": 0.677915160420441,
      "grad_norm": 0.2840427756309509,
      "learning_rate": 1.4288667226203077e-05,
      "loss": 0.2286,
      "mean_token_accuracy": 0.9443538438528776,
      "num_tokens": 75782624.0,
      "step": 10150
    },
    {
      "entropy": 1.93028691932559,
      "epoch": 0.6785830571302148,
      "grad_norm": 0.2752184271812439,
      "learning_rate": 1.4236208605485501e-05,
      "loss": 0.2378,
      "mean_token_accuracy": 0.943329131603241,
      "num_tokens": 75857633.0,
      "step": 10160
    },
    {
      "entropy": 1.9177371695637704,
      "epoch": 0.6792509538399887,
      "grad_norm": 0.3550587594509125,
      "learning_rate": 1.418380878092164e-05,
      "loss": 0.2316,
      "mean_token_accuracy": 0.9445929065346718,
      "num_tokens": 75932074.0,
      "step": 10170
    },
    {
      "entropy": 1.9028411082923413,
      "epoch": 0.6799188505497625,
      "grad_norm": 0.2528437674045563,
      "learning_rate": 1.4131468037416917e-05,
      "loss": 0.204,
      "mean_token_accuracy": 0.9501767240464687,
      "num_tokens": 76008562.0,
      "step": 10180
    },
    {
      "entropy": 1.8954525627195835,
      "epoch": 0.6805867472595363,
      "grad_norm": 0.282778799533844,
      "learning_rate": 1.4079186659555571e-05,
      "loss": 0.2258,
      "mean_token_accuracy": 0.9455604698508978,
      "num_tokens": 76084297.0,
      "step": 10190
    },
    {
      "entropy": 1.8905874237418174,
      "epoch": 0.6812546439693101,
      "grad_norm": 0.28031420707702637,
      "learning_rate": 1.402696493159904e-05,
      "loss": 0.2285,
      "mean_token_accuracy": 0.9434068106114865,
      "num_tokens": 76157111.0,
      "step": 10200
    },
    {
      "entropy": 1.9129234299063682,
      "epoch": 0.681922540679084,
      "grad_norm": 0.3190227150917053,
      "learning_rate": 1.3974803137484444e-05,
      "loss": 0.2242,
      "mean_token_accuracy": 0.9443533584475517,
      "num_tokens": 76228777.0,
      "step": 10210
    },
    {
      "entropy": 1.9016528375446797,
      "epoch": 0.6825904373888578,
      "grad_norm": 0.2575254738330841,
      "learning_rate": 1.3922701560823031e-05,
      "loss": 0.2237,
      "mean_token_accuracy": 0.9442144501954317,
      "num_tokens": 76302041.0,
      "step": 10220
    },
    {
      "entropy": 1.8829542085528375,
      "epoch": 0.6832583340986317,
      "grad_norm": 0.3498046398162842,
      "learning_rate": 1.3870660484898627e-05,
      "loss": 0.2075,
      "mean_token_accuracy": 0.946049178764224,
      "num_tokens": 76376733.0,
      "step": 10230
    },
    {
      "entropy": 1.8748780526220798,
      "epoch": 0.6839262308084055,
      "grad_norm": 0.27811399102211,
      "learning_rate": 1.3818680192666133e-05,
      "loss": 0.2067,
      "mean_token_accuracy": 0.9452836539596319,
      "num_tokens": 76455409.0,
      "step": 10240
    },
    {
      "entropy": 1.8928939312696458,
      "epoch": 0.6845941275181793,
      "grad_norm": 0.3447349965572357,
      "learning_rate": 1.3766760966749925e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9453567314893008,
      "num_tokens": 76532469.0,
      "step": 10250
    },
    {
      "entropy": 1.8949653565883637,
      "epoch": 0.6852620242279531,
      "grad_norm": 0.30839037895202637,
      "learning_rate": 1.3714903089442378e-05,
      "loss": 0.2219,
      "mean_token_accuracy": 0.9457360960543155,
      "num_tokens": 76606123.0,
      "step": 10260
    },
    {
      "entropy": 1.9129513271152974,
      "epoch": 0.685929920937727,
      "grad_norm": 0.30321478843688965,
      "learning_rate": 1.3663106842702303e-05,
      "loss": 0.2128,
      "mean_token_accuracy": 0.9456833682954311,
      "num_tokens": 76679441.0,
      "step": 10270
    },
    {
      "entropy": 1.8930385418236255,
      "epoch": 0.6865978176475008,
      "grad_norm": 0.314028799533844,
      "learning_rate": 1.3611372508153422e-05,
      "loss": 0.2177,
      "mean_token_accuracy": 0.947524844110012,
      "num_tokens": 76751750.0,
      "step": 10280
    },
    {
      "entropy": 1.912549114227295,
      "epoch": 0.6872657143572747,
      "grad_norm": 0.3038738965988159,
      "learning_rate": 1.3559700367082804e-05,
      "loss": 0.2166,
      "mean_token_accuracy": 0.9463405251502991,
      "num_tokens": 76824812.0,
      "step": 10290
    },
    {
      "entropy": 1.8967111505568028,
      "epoch": 0.6879336110670485,
      "grad_norm": 0.36032551527023315,
      "learning_rate": 1.350809070043939e-05,
      "loss": 0.2339,
      "mean_token_accuracy": 0.9436006501317025,
      "num_tokens": 76898782.0,
      "step": 10300
    },
    {
      "entropy": 1.9159264162182807,
      "epoch": 0.6886015077768223,
      "grad_norm": 0.31301388144493103,
      "learning_rate": 1.3456543788832429e-05,
      "loss": 0.2395,
      "mean_token_accuracy": 0.9434678621590138,
      "num_tokens": 76972774.0,
      "step": 10310
    },
    {
      "entropy": 1.8853584058582782,
      "epoch": 0.6892694044865961,
      "grad_norm": 0.2590923011302948,
      "learning_rate": 1.3405059912529974e-05,
      "loss": 0.2186,
      "mean_token_accuracy": 0.9430528480559588,
      "num_tokens": 77047570.0,
      "step": 10320
    },
    {
      "entropy": 1.9312948137521744,
      "epoch": 0.68993730119637,
      "grad_norm": 0.3591325879096985,
      "learning_rate": 1.3353639351457314e-05,
      "loss": 0.221,
      "mean_token_accuracy": 0.9452490895986557,
      "num_tokens": 77123320.0,
      "step": 10330
    },
    {
      "entropy": 1.9148299671709537,
      "epoch": 0.6906051979061438,
      "grad_norm": 0.37425604462623596,
      "learning_rate": 1.3302282385195516e-05,
      "loss": 0.2394,
      "mean_token_accuracy": 0.9393338304013014,
      "num_tokens": 77198843.0,
      "step": 10340
    },
    {
      "entropy": 1.9022190049290657,
      "epoch": 0.6912730946159177,
      "grad_norm": 0.35374850034713745,
      "learning_rate": 1.3250989292979859e-05,
      "loss": 0.2113,
      "mean_token_accuracy": 0.9470495611429215,
      "num_tokens": 77274202.0,
      "step": 10350
    },
    {
      "entropy": 1.883647397905588,
      "epoch": 0.6919409913256915,
      "grad_norm": 0.31867679953575134,
      "learning_rate": 1.3199760353698334e-05,
      "loss": 0.2138,
      "mean_token_accuracy": 0.9473440900444985,
      "num_tokens": 77350488.0,
      "step": 10360
    },
    {
      "entropy": 1.920442698150873,
      "epoch": 0.6926088880354653,
      "grad_norm": 0.34543076157569885,
      "learning_rate": 1.3148595845890126e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9456390470266343,
      "num_tokens": 77425469.0,
      "step": 10370
    },
    {
      "entropy": 1.913692919164896,
      "epoch": 0.6932767847452391,
      "grad_norm": 0.2782209813594818,
      "learning_rate": 1.3097496047744081e-05,
      "loss": 0.2105,
      "mean_token_accuracy": 0.9458931490778923,
      "num_tokens": 77501130.0,
      "step": 10380
    },
    {
      "entropy": 1.9360430128872395,
      "epoch": 0.693944681455013,
      "grad_norm": 0.3282945156097412,
      "learning_rate": 1.3046461237097227e-05,
      "loss": 0.2162,
      "mean_token_accuracy": 0.9447419974952936,
      "num_tokens": 77576687.0,
      "step": 10390
    },
    {
      "entropy": 1.9158230133354663,
      "epoch": 0.6946125781647868,
      "grad_norm": 0.3610730767250061,
      "learning_rate": 1.2995491691433229e-05,
      "loss": 0.2389,
      "mean_token_accuracy": 0.9394503276795149,
      "num_tokens": 77649955.0,
      "step": 10400
    },
    {
      "entropy": 1.9148353688418864,
      "epoch": 0.6952804748745607,
      "grad_norm": 0.325503408908844,
      "learning_rate": 1.2944587687880919e-05,
      "loss": 0.2458,
      "mean_token_accuracy": 0.9413210216909647,
      "num_tokens": 77722460.0,
      "step": 10410
    },
    {
      "entropy": 1.9086212821304798,
      "epoch": 0.6959483715843345,
      "grad_norm": 0.2849908471107483,
      "learning_rate": 1.2893749503212754e-05,
      "loss": 0.205,
      "mean_token_accuracy": 0.9504839979112149,
      "num_tokens": 77798379.0,
      "step": 10420
    },
    {
      "entropy": 1.911132575571537,
      "epoch": 0.6966162682941083,
      "grad_norm": 0.3763129711151123,
      "learning_rate": 1.2842977413843311e-05,
      "loss": 0.2078,
      "mean_token_accuracy": 0.9516026563942432,
      "num_tokens": 77876895.0,
      "step": 10430
    },
    {
      "entropy": 1.8872507065534592,
      "epoch": 0.6972841650038821,
      "grad_norm": 0.3012331426143646,
      "learning_rate": 1.279227169582782e-05,
      "loss": 0.1955,
      "mean_token_accuracy": 0.9522138494998217,
      "num_tokens": 77949297.0,
      "step": 10440
    },
    {
      "entropy": 1.9060810387134552,
      "epoch": 0.697952061713656,
      "grad_norm": 0.3150883913040161,
      "learning_rate": 1.2741632624860628e-05,
      "loss": 0.2114,
      "mean_token_accuracy": 0.9443527456372977,
      "num_tokens": 78023998.0,
      "step": 10450
    },
    {
      "entropy": 1.864063174277544,
      "epoch": 0.6986199584234298,
      "grad_norm": 0.27818211913108826,
      "learning_rate": 1.269106047627372e-05,
      "loss": 0.1976,
      "mean_token_accuracy": 0.9494179770350456,
      "num_tokens": 78098766.0,
      "step": 10460
    },
    {
      "entropy": 1.9028538651764393,
      "epoch": 0.6992878551332037,
      "grad_norm": 0.2586256265640259,
      "learning_rate": 1.2640555525035213e-05,
      "loss": 0.2244,
      "mean_token_accuracy": 0.9441565670073032,
      "num_tokens": 78173065.0,
      "step": 10470
    },
    {
      "entropy": 1.9098172560334206,
      "epoch": 0.6999557518429775,
      "grad_norm": 0.3565428853034973,
      "learning_rate": 1.2590118045747854e-05,
      "loss": 0.2302,
      "mean_token_accuracy": 0.9427990823984146,
      "num_tokens": 78245662.0,
      "step": 10480
    },
    {
      "entropy": 1.9347416557371617,
      "epoch": 0.7006236485527513,
      "grad_norm": 0.2804452180862427,
      "learning_rate": 1.253974831264753e-05,
      "loss": 0.2081,
      "mean_token_accuracy": 0.9460851203650236,
      "num_tokens": 78320231.0,
      "step": 10490
    },
    {
      "entropy": 1.9094208985567094,
      "epoch": 0.7012915452625251,
      "grad_norm": 0.3190443813800812,
      "learning_rate": 1.2489446599601801e-05,
      "loss": 0.237,
      "mean_token_accuracy": 0.9449720848351717,
      "num_tokens": 78394690.0,
      "step": 10500
    },
    {
      "entropy": 1.8809963166713715,
      "epoch": 0.701959441972299,
      "grad_norm": 0.29107731580734253,
      "learning_rate": 1.2439213180108385e-05,
      "loss": 0.2052,
      "mean_token_accuracy": 0.9492700982838869,
      "num_tokens": 78471645.0,
      "step": 10510
    },
    {
      "entropy": 1.9181700721383095,
      "epoch": 0.7026273386820728,
      "grad_norm": 0.2929648458957672,
      "learning_rate": 1.2389048327293672e-05,
      "loss": 0.2152,
      "mean_token_accuracy": 0.9482477650046348,
      "num_tokens": 78546909.0,
      "step": 10520
    },
    {
      "entropy": 1.97821604013443,
      "epoch": 0.7032952353918467,
      "grad_norm": 0.29727914929389954,
      "learning_rate": 1.2338952313911264e-05,
      "loss": 0.2439,
      "mean_token_accuracy": 0.9414060182869435,
      "num_tokens": 78620557.0,
      "step": 10530
    },
    {
      "entropy": 1.9196878917515279,
      "epoch": 0.7039631321016205,
      "grad_norm": 0.3136504888534546,
      "learning_rate": 1.2288925412340437e-05,
      "loss": 0.2102,
      "mean_token_accuracy": 0.9470030102878809,
      "num_tokens": 78694063.0,
      "step": 10540
    },
    {
      "entropy": 1.9015464268624782,
      "epoch": 0.7046310288113943,
      "grad_norm": 0.3344622850418091,
      "learning_rate": 1.2238967894584729e-05,
      "loss": 0.2145,
      "mean_token_accuracy": 0.9453800730407238,
      "num_tokens": 78769097.0,
      "step": 10550
    },
    {
      "entropy": 1.9156240098178388,
      "epoch": 0.7052989255211681,
      "grad_norm": 0.2830991744995117,
      "learning_rate": 1.218908003227041e-05,
      "loss": 0.2349,
      "mean_token_accuracy": 0.9418215606361627,
      "num_tokens": 78843831.0,
      "step": 10560
    },
    {
      "entropy": 1.9218591064214707,
      "epoch": 0.705966822230942,
      "grad_norm": 0.28117820620536804,
      "learning_rate": 1.2139262096645047e-05,
      "loss": 0.214,
      "mean_token_accuracy": 0.946132193505764,
      "num_tokens": 78919302.0,
      "step": 10570
    },
    {
      "entropy": 1.9197048082947732,
      "epoch": 0.7066347189407158,
      "grad_norm": 0.31449583172798157,
      "learning_rate": 1.2089514358575969e-05,
      "loss": 0.2343,
      "mean_token_accuracy": 0.9422256987541914,
      "num_tokens": 78993131.0,
      "step": 10580
    },
    {
      "entropy": 1.9210989877581597,
      "epoch": 0.7073026156504897,
      "grad_norm": 0.38509050011634827,
      "learning_rate": 1.2039837088548854e-05,
      "loss": 0.2285,
      "mean_token_accuracy": 0.9447336968034505,
      "num_tokens": 79067877.0,
      "step": 10590
    },
    {
      "entropy": 1.9332555092871189,
      "epoch": 0.7079705123602635,
      "grad_norm": 0.3056189715862274,
      "learning_rate": 1.1990230556666236e-05,
      "loss": 0.2211,
      "mean_token_accuracy": 0.94359893463552,
      "num_tokens": 79144098.0,
      "step": 10600
    },
    {
      "entropy": 1.923354833573103,
      "epoch": 0.7086384090700373,
      "grad_norm": 0.28786858916282654,
      "learning_rate": 1.1940695032646027e-05,
      "loss": 0.2089,
      "mean_token_accuracy": 0.9515276707708835,
      "num_tokens": 79221011.0,
      "step": 10610
    },
    {
      "entropy": 1.9306829623878001,
      "epoch": 0.7093063057798111,
      "grad_norm": 0.2754662334918976,
      "learning_rate": 1.1891230785820074e-05,
      "loss": 0.2198,
      "mean_token_accuracy": 0.9464119967073202,
      "num_tokens": 79296624.0,
      "step": 10620
    },
    {
      "entropy": 1.908396078646183,
      "epoch": 0.709974202489585,
      "grad_norm": 0.29287096858024597,
      "learning_rate": 1.1841838085132647e-05,
      "loss": 0.2256,
      "mean_token_accuracy": 0.9482095073908567,
      "num_tokens": 79373238.0,
      "step": 10630
    },
    {
      "entropy": 1.9073860429227352,
      "epoch": 0.7106420991993588,
      "grad_norm": 0.3153214156627655,
      "learning_rate": 1.1792517199139041e-05,
      "loss": 0.2175,
      "mean_token_accuracy": 0.9462470531463623,
      "num_tokens": 79448905.0,
      "step": 10640
    },
    {
      "entropy": 1.925882250070572,
      "epoch": 0.7113099959091327,
      "grad_norm": 0.4114491641521454,
      "learning_rate": 1.1743268396004077e-05,
      "loss": 0.2259,
      "mean_token_accuracy": 0.9453411553055048,
      "num_tokens": 79523396.0,
      "step": 10650
    },
    {
      "entropy": 1.9333959378302097,
      "epoch": 0.7119778926189065,
      "grad_norm": 0.3115549087524414,
      "learning_rate": 1.1694091943500641e-05,
      "loss": 0.2114,
      "mean_token_accuracy": 0.9485629417002202,
      "num_tokens": 79599257.0,
      "step": 10660
    },
    {
      "entropy": 1.9398183852434159,
      "epoch": 0.7126457893286803,
      "grad_norm": 0.30406397581100464,
      "learning_rate": 1.1644988109008256e-05,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9486421644687653,
      "num_tokens": 79673245.0,
      "step": 10670
    },
    {
      "entropy": 1.897850428521633,
      "epoch": 0.7133136860384541,
      "grad_norm": 0.31005430221557617,
      "learning_rate": 1.1595957159511582e-05,
      "loss": 0.2093,
      "mean_token_accuracy": 0.9467908661812544,
      "num_tokens": 79747565.0,
      "step": 10680
    },
    {
      "entropy": 1.9372698307037353,
      "epoch": 0.713981582748228,
      "grad_norm": 0.36624276638031006,
      "learning_rate": 1.1546999361599016e-05,
      "loss": 0.2316,
      "mean_token_accuracy": 0.9452444355934858,
      "num_tokens": 79821979.0,
      "step": 10690
    },
    {
      "entropy": 1.902782541513443,
      "epoch": 0.7146494794580018,
      "grad_norm": 0.34434133768081665,
      "learning_rate": 1.1498114981461226e-05,
      "loss": 0.225,
      "mean_token_accuracy": 0.9443479336798191,
      "num_tokens": 79896916.0,
      "step": 10700
    },
    {
      "entropy": 1.9020543485879897,
      "epoch": 0.7153173761677757,
      "grad_norm": 0.3101283609867096,
      "learning_rate": 1.1449304284889657e-05,
      "loss": 0.2123,
      "mean_token_accuracy": 0.9482042495161295,
      "num_tokens": 79974405.0,
      "step": 10710
    },
    {
      "entropy": 1.91372000426054,
      "epoch": 0.7159852728775495,
      "grad_norm": 0.3263084888458252,
      "learning_rate": 1.1400567537275169e-05,
      "loss": 0.2068,
      "mean_token_accuracy": 0.9487948149442673,
      "num_tokens": 80050472.0,
      "step": 10720
    },
    {
      "entropy": 1.8923282079398631,
      "epoch": 0.7166531695873233,
      "grad_norm": 0.3215576708316803,
      "learning_rate": 1.1351905003606545e-05,
      "loss": 0.2156,
      "mean_token_accuracy": 0.9463704567402601,
      "num_tokens": 80124647.0,
      "step": 10730
    },
    {
      "entropy": 1.9246387995779515,
      "epoch": 0.7173210662970971,
      "grad_norm": 0.29945650696754456,
      "learning_rate": 1.130331694846902e-05,
      "loss": 0.2229,
      "mean_token_accuracy": 0.9477593809366226,
      "num_tokens": 80199444.0,
      "step": 10740
    },
    {
      "entropy": 1.9335595175623894,
      "epoch": 0.717988963006871,
      "grad_norm": 0.31465283036231995,
      "learning_rate": 1.1254803636042916e-05,
      "loss": 0.2155,
      "mean_token_accuracy": 0.9464696452021599,
      "num_tokens": 80273836.0,
      "step": 10750
    },
    {
      "entropy": 1.902474556863308,
      "epoch": 0.7186568597166448,
      "grad_norm": 0.3238000273704529,
      "learning_rate": 1.1206365330102156e-05,
      "loss": 0.2131,
      "mean_token_accuracy": 0.9495478864759207,
      "num_tokens": 80350461.0,
      "step": 10760
    },
    {
      "entropy": 1.9139510154724122,
      "epoch": 0.7193247564264187,
      "grad_norm": 0.36762136220932007,
      "learning_rate": 1.1158002294012838e-05,
      "loss": 0.2165,
      "mean_token_accuracy": 0.9491926331073046,
      "num_tokens": 80424054.0,
      "step": 10770
    },
    {
      "entropy": 1.8873869948089124,
      "epoch": 0.7199926531361925,
      "grad_norm": 0.3309178948402405,
      "learning_rate": 1.1109714790731826e-05,
      "loss": 0.2185,
      "mean_token_accuracy": 0.9451238356530667,
      "num_tokens": 80501494.0,
      "step": 10780
    },
    {
      "entropy": 1.9332789048552512,
      "epoch": 0.7206605498459663,
      "grad_norm": 0.2736462354660034,
      "learning_rate": 1.1061503082805262e-05,
      "loss": 0.2211,
      "mean_token_accuracy": 0.9458402171730995,
      "num_tokens": 80575730.0,
      "step": 10790
    },
    {
      "entropy": 1.911842057108879,
      "epoch": 0.7213284465557401,
      "grad_norm": 0.27536875009536743,
      "learning_rate": 1.1013367432367209e-05,
      "loss": 0.226,
      "mean_token_accuracy": 0.944275650382042,
      "num_tokens": 80650372.0,
      "step": 10800
    },
    {
      "entropy": 1.9260739140212535,
      "epoch": 0.721996343265514,
      "grad_norm": 0.24608100950717926,
      "learning_rate": 1.0965308101138192e-05,
      "loss": 0.2076,
      "mean_token_accuracy": 0.948920963704586,
      "num_tokens": 80727426.0,
      "step": 10810
    },
    {
      "entropy": 1.9246294118463994,
      "epoch": 0.7226642399752878,
      "grad_norm": 0.3301178216934204,
      "learning_rate": 1.0917325350423776e-05,
      "loss": 0.2364,
      "mean_token_accuracy": 0.9442116837948561,
      "num_tokens": 80802556.0,
      "step": 10820
    },
    {
      "entropy": 1.9289285562932492,
      "epoch": 0.7233321366850617,
      "grad_norm": 0.2902006506919861,
      "learning_rate": 1.0869419441113127e-05,
      "loss": 0.221,
      "mean_token_accuracy": 0.9450737927109003,
      "num_tokens": 80876470.0,
      "step": 10830
    },
    {
      "entropy": 1.9085661433637142,
      "epoch": 0.7240000333948355,
      "grad_norm": 0.3091467320919037,
      "learning_rate": 1.0821590633677639e-05,
      "loss": 0.2183,
      "mean_token_accuracy": 0.946364825963974,
      "num_tokens": 80950736.0,
      "step": 10840
    },
    {
      "entropy": 1.930638635903597,
      "epoch": 0.7246679301046093,
      "grad_norm": 0.33393821120262146,
      "learning_rate": 1.0773839188169479e-05,
      "loss": 0.2143,
      "mean_token_accuracy": 0.9487283498048782,
      "num_tokens": 81024404.0,
      "step": 10850
    },
    {
      "entropy": 1.888561774045229,
      "epoch": 0.7253358268143831,
      "grad_norm": 0.3741356134414673,
      "learning_rate": 1.0726165364220188e-05,
      "loss": 0.2147,
      "mean_token_accuracy": 0.9482394717633724,
      "num_tokens": 81099544.0,
      "step": 10860
    },
    {
      "entropy": 1.900745715945959,
      "epoch": 0.726003723524157,
      "grad_norm": 0.2864074409008026,
      "learning_rate": 1.067856942103928e-05,
      "loss": 0.2161,
      "mean_token_accuracy": 0.9452195778489113,
      "num_tokens": 81175181.0,
      "step": 10870
    },
    {
      "entropy": 1.9070482835173608,
      "epoch": 0.7266716202339308,
      "grad_norm": 0.28629112243652344,
      "learning_rate": 1.0631051617412784e-05,
      "loss": 0.2209,
      "mean_token_accuracy": 0.9458576466888189,
      "num_tokens": 81249993.0,
      "step": 10880
    },
    {
      "entropy": 1.931928315013647,
      "epoch": 0.7273395169437047,
      "grad_norm": 0.2930496037006378,
      "learning_rate": 1.0583612211701913e-05,
      "loss": 0.2365,
      "mean_token_accuracy": 0.9416486214846372,
      "num_tokens": 81326535.0,
      "step": 10890
    },
    {
      "entropy": 1.90124344304204,
      "epoch": 0.7280074136534785,
      "grad_norm": 0.2984851002693176,
      "learning_rate": 1.0536251461841596e-05,
      "loss": 0.2117,
      "mean_token_accuracy": 0.945863937959075,
      "num_tokens": 81400821.0,
      "step": 10900
    },
    {
      "entropy": 1.920573817193508,
      "epoch": 0.7286753103632523,
      "grad_norm": 0.30207282304763794,
      "learning_rate": 1.0488969625339113e-05,
      "loss": 0.2233,
      "mean_token_accuracy": 0.9451156925410032,
      "num_tokens": 81475823.0,
      "step": 10910
    },
    {
      "entropy": 1.9225694991648197,
      "epoch": 0.7293432070730261,
      "grad_norm": 0.3032468259334564,
      "learning_rate": 1.0441766959272676e-05,
      "loss": 0.2197,
      "mean_token_accuracy": 0.9471049934625626,
      "num_tokens": 81553280.0,
      "step": 10920
    },
    {
      "entropy": 1.921476662904024,
      "epoch": 0.7300111037828,
      "grad_norm": 0.2499317079782486,
      "learning_rate": 1.0394643720290028e-05,
      "loss": 0.214,
      "mean_token_accuracy": 0.9472498957067728,
      "num_tokens": 81625523.0,
      "step": 10930
    },
    {
      "entropy": 1.9392130978405475,
      "epoch": 0.7306790004925738,
      "grad_norm": 0.36099153757095337,
      "learning_rate": 1.0347600164607046e-05,
      "loss": 0.2306,
      "mean_token_accuracy": 0.9440038096159696,
      "num_tokens": 81699770.0,
      "step": 10940
    },
    {
      "entropy": 1.9041217111051083,
      "epoch": 0.7313468972023477,
      "grad_norm": 0.26680734753608704,
      "learning_rate": 1.0300636548006386e-05,
      "loss": 0.2023,
      "mean_token_accuracy": 0.9486195117235183,
      "num_tokens": 81774648.0,
      "step": 10950
    },
    {
      "entropy": 1.9059810176491738,
      "epoch": 0.7320147939121215,
      "grad_norm": 0.2969484329223633,
      "learning_rate": 1.025375312583605e-05,
      "loss": 0.2231,
      "mean_token_accuracy": 0.9443758942186833,
      "num_tokens": 81851430.0,
      "step": 10960
    },
    {
      "entropy": 1.918486201763153,
      "epoch": 0.7326826906218953,
      "grad_norm": 0.35175079107284546,
      "learning_rate": 1.0206950153008005e-05,
      "loss": 0.2316,
      "mean_token_accuracy": 0.9463289577513934,
      "num_tokens": 81923444.0,
      "step": 10970
    },
    {
      "entropy": 1.9131188362836837,
      "epoch": 0.7333505873316691,
      "grad_norm": 0.30851465463638306,
      "learning_rate": 1.0160227883996823e-05,
      "loss": 0.231,
      "mean_token_accuracy": 0.9428766563534736,
      "num_tokens": 81999590.0,
      "step": 10980
    },
    {
      "entropy": 1.9351812101900578,
      "epoch": 0.734018484041443,
      "grad_norm": 0.24056492745876312,
      "learning_rate": 1.0113586572838244e-05,
      "loss": 0.2143,
      "mean_token_accuracy": 0.9471086826175451,
      "num_tokens": 82071641.0,
      "step": 10990
    },
    {
      "entropy": 1.9139883778989315,
      "epoch": 0.7346863807512168,
      "grad_norm": 0.28496241569519043,
      "learning_rate": 1.0067026473127857e-05,
      "loss": 0.2126,
      "mean_token_accuracy": 0.947807178273797,
      "num_tokens": 82147075.0,
      "step": 11000
    },
    {
      "epoch": 0.7346863807512168,
      "eval_entropy": 1.9059199431180953,
      "eval_loss": 0.2356305718421936,
      "eval_mean_token_accuracy": 0.9463012171268463,
      "eval_num_tokens": 82147075.0,
      "eval_runtime": 554.1746,
      "eval_samples_per_second": 36.09,
      "eval_steps_per_second": 9.022,
      "step": 11000
    },
    {
      "entropy": 1.9091171011328698,
      "epoch": 0.7353542774609907,
      "grad_norm": 0.34441131353378296,
      "learning_rate": 1.0020547838019682e-05,
      "loss": 0.2062,
      "mean_token_accuracy": 0.9468987006694078,
      "num_tokens": 82224421.0,
      "step": 11010
    },
    {
      "entropy": 1.8974645651876927,
      "epoch": 0.7360221741707645,
      "grad_norm": 0.3231317698955536,
      "learning_rate": 9.974150920224805e-06,
      "loss": 0.2306,
      "mean_token_accuracy": 0.9416075870394707,
      "num_tokens": 82298797.0,
      "step": 11020
    },
    {
      "entropy": 1.9342247806489468,
      "epoch": 0.7366900708805383,
      "grad_norm": 0.35956835746765137,
      "learning_rate": 9.927835972010017e-06,
      "loss": 0.2467,
      "mean_token_accuracy": 0.9416825965046882,
      "num_tokens": 82369569.0,
      "step": 11030
    },
    {
      "entropy": 1.8589746229350568,
      "epoch": 0.7373579675903121,
      "grad_norm": 0.2973991334438324,
      "learning_rate": 9.881603245196394e-06,
      "loss": 0.2015,
      "mean_token_accuracy": 0.9503379218280316,
      "num_tokens": 82446934.0,
      "step": 11040
    },
    {
      "entropy": 1.864432605355978,
      "epoch": 0.738025864300086,
      "grad_norm": 0.2519768178462982,
      "learning_rate": 9.835452991157993e-06,
      "loss": 0.2082,
      "mean_token_accuracy": 0.9485573947429657,
      "num_tokens": 82520340.0,
      "step": 11050
    },
    {
      "entropy": 1.8844397969543933,
      "epoch": 0.7386937610098598,
      "grad_norm": 0.4154239892959595,
      "learning_rate": 9.789385460820447e-06,
      "loss": 0.2284,
      "mean_token_accuracy": 0.944726001471281,
      "num_tokens": 82593202.0,
      "step": 11060
    },
    {
      "entropy": 1.8970091849565507,
      "epoch": 0.7393616577196337,
      "grad_norm": 0.3660540282726288,
      "learning_rate": 9.743400904659613e-06,
      "loss": 0.2036,
      "mean_token_accuracy": 0.9501904439181089,
      "num_tokens": 82668771.0,
      "step": 11070
    },
    {
      "entropy": 1.8937359876930713,
      "epoch": 0.7400295544294075,
      "grad_norm": 0.3173390030860901,
      "learning_rate": 9.697499572700177e-06,
      "loss": 0.2148,
      "mean_token_accuracy": 0.9467160064727068,
      "num_tokens": 82741509.0,
      "step": 11080
    },
    {
      "entropy": 1.8948323883116245,
      "epoch": 0.7406974511391813,
      "grad_norm": 0.33909356594085693,
      "learning_rate": 9.651681714514362e-06,
      "loss": 0.2117,
      "mean_token_accuracy": 0.9481560476124287,
      "num_tokens": 82816825.0,
      "step": 11090
    },
    {
      "entropy": 1.8804951347410679,
      "epoch": 0.7413653478489551,
      "grad_norm": 0.3089444935321808,
      "learning_rate": 9.60594757922051e-06,
      "loss": 0.2116,
      "mean_token_accuracy": 0.9482202224433423,
      "num_tokens": 82894242.0,
      "step": 11100
    },
    {
      "entropy": 1.9098335951566696,
      "epoch": 0.742033244558729,
      "grad_norm": 0.28488823771476746,
      "learning_rate": 9.560297415481752e-06,
      "loss": 0.2254,
      "mean_token_accuracy": 0.9422137197107077,
      "num_tokens": 82967463.0,
      "step": 11110
    },
    {
      "entropy": 1.9448677077889442,
      "epoch": 0.7427011412685028,
      "grad_norm": 0.3729751408100128,
      "learning_rate": 9.514731471504668e-06,
      "loss": 0.2478,
      "mean_token_accuracy": 0.940930238738656,
      "num_tokens": 83040541.0,
      "step": 11120
    },
    {
      "entropy": 1.9311937920749187,
      "epoch": 0.7433690379782767,
      "grad_norm": 0.3697790503501892,
      "learning_rate": 9.469249995037892e-06,
      "loss": 0.2286,
      "mean_token_accuracy": 0.9461524050682784,
      "num_tokens": 83113719.0,
      "step": 11130
    },
    {
      "entropy": 1.9159417890012265,
      "epoch": 0.7440369346880505,
      "grad_norm": 0.29803135991096497,
      "learning_rate": 9.423853233370822e-06,
      "loss": 0.2185,
      "mean_token_accuracy": 0.9477081406861544,
      "num_tokens": 83187726.0,
      "step": 11140
    },
    {
      "entropy": 1.893703680485487,
      "epoch": 0.7447048313978243,
      "grad_norm": 0.3220638036727905,
      "learning_rate": 9.378541433332241e-06,
      "loss": 0.1938,
      "mean_token_accuracy": 0.9507091857492924,
      "num_tokens": 83264504.0,
      "step": 11150
    },
    {
      "entropy": 1.9050804533064365,
      "epoch": 0.7453727281075981,
      "grad_norm": 1.331411600112915,
      "learning_rate": 9.333314841288996e-06,
      "loss": 0.213,
      "mean_token_accuracy": 0.9479123786091804,
      "num_tokens": 83338216.0,
      "step": 11160
    },
    {
      "entropy": 1.9048821784555912,
      "epoch": 0.746040624817372,
      "grad_norm": 0.2998942732810974,
      "learning_rate": 9.28817370314461e-06,
      "loss": 0.2265,
      "mean_token_accuracy": 0.9448808398097753,
      "num_tokens": 83413860.0,
      "step": 11170
    },
    {
      "entropy": 1.9363364651799202,
      "epoch": 0.7467085215271458,
      "grad_norm": 0.29526156187057495,
      "learning_rate": 9.243118264338028e-06,
      "loss": 0.2308,
      "mean_token_accuracy": 0.9450293987989425,
      "num_tokens": 83486314.0,
      "step": 11180
    },
    {
      "entropy": 1.9200671084225178,
      "epoch": 0.7473764182369197,
      "grad_norm": 0.2819261848926544,
      "learning_rate": 9.198148769842195e-06,
      "loss": 0.2247,
      "mean_token_accuracy": 0.9459836155176162,
      "num_tokens": 83561342.0,
      "step": 11190
    },
    {
      "entropy": 1.9335692524909973,
      "epoch": 0.7480443149466935,
      "grad_norm": 0.2717633843421936,
      "learning_rate": 9.153265464162785e-06,
      "loss": 0.2117,
      "mean_token_accuracy": 0.9457612648606301,
      "num_tokens": 83635250.0,
      "step": 11200
    },
    {
      "entropy": 1.9017163820564746,
      "epoch": 0.7487122116564673,
      "grad_norm": 0.300033837556839,
      "learning_rate": 9.10846859133685e-06,
      "loss": 0.2221,
      "mean_token_accuracy": 0.9455062616616488,
      "num_tokens": 83709246.0,
      "step": 11210
    },
    {
      "entropy": 1.898740069568157,
      "epoch": 0.7493801083662411,
      "grad_norm": 0.30894923210144043,
      "learning_rate": 9.063758394931486e-06,
      "loss": 0.218,
      "mean_token_accuracy": 0.9431704297661782,
      "num_tokens": 83783894.0,
      "step": 11220
    },
    {
      "entropy": 1.9074263170361518,
      "epoch": 0.750048005076015,
      "grad_norm": 0.38207611441612244,
      "learning_rate": 9.019135118042536e-06,
      "loss": 0.2119,
      "mean_token_accuracy": 0.9463793229311704,
      "num_tokens": 83856472.0,
      "step": 11230
    },
    {
      "entropy": 1.8768710792064667,
      "epoch": 0.7507159017857888,
      "grad_norm": 0.3123346269130707,
      "learning_rate": 8.974599003293216e-06,
      "loss": 0.2055,
      "mean_token_accuracy": 0.9483194727450609,
      "num_tokens": 83932722.0,
      "step": 11240
    },
    {
      "entropy": 1.8639044649899006,
      "epoch": 0.7513837984955627,
      "grad_norm": 0.303500771522522,
      "learning_rate": 8.930150292832855e-06,
      "loss": 0.2085,
      "mean_token_accuracy": 0.9482971403747797,
      "num_tokens": 84008255.0,
      "step": 11250
    },
    {
      "entropy": 1.90454136505723,
      "epoch": 0.7520516952053365,
      "grad_norm": 0.3632895052433014,
      "learning_rate": 8.885789228335544e-06,
      "loss": 0.209,
      "mean_token_accuracy": 0.9486837543547153,
      "num_tokens": 84085090.0,
      "step": 11260
    },
    {
      "entropy": 1.913232085108757,
      "epoch": 0.7527195919151103,
      "grad_norm": 0.3618438243865967,
      "learning_rate": 8.841516050998843e-06,
      "loss": 0.226,
      "mean_token_accuracy": 0.9446337543427944,
      "num_tokens": 84159161.0,
      "step": 11270
    },
    {
      "entropy": 1.8702809795737267,
      "epoch": 0.7533874886248841,
      "grad_norm": 0.30103811621665955,
      "learning_rate": 8.797331001542423e-06,
      "loss": 0.2065,
      "mean_token_accuracy": 0.949903305247426,
      "num_tokens": 84235783.0,
      "step": 11280
    },
    {
      "entropy": 1.8931458652019502,
      "epoch": 0.754055385334658,
      "grad_norm": 0.2811412215232849,
      "learning_rate": 8.753234320206822e-06,
      "loss": 0.2332,
      "mean_token_accuracy": 0.9413031421601772,
      "num_tokens": 84312456.0,
      "step": 11290
    },
    {
      "entropy": 1.8912273250520228,
      "epoch": 0.7547232820444318,
      "grad_norm": 0.32122495770454407,
      "learning_rate": 8.709226246752101e-06,
      "loss": 0.2211,
      "mean_token_accuracy": 0.9437547899782658,
      "num_tokens": 84386846.0,
      "step": 11300
    },
    {
      "entropy": 1.9023813880980014,
      "epoch": 0.7553911787542057,
      "grad_norm": 0.3737948536872864,
      "learning_rate": 8.665307020456538e-06,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9444470826536417,
      "num_tokens": 84462405.0,
      "step": 11310
    },
    {
      "entropy": 1.9102597139775752,
      "epoch": 0.7560590754639795,
      "grad_norm": 0.27867555618286133,
      "learning_rate": 8.621476880115357e-06,
      "loss": 0.2182,
      "mean_token_accuracy": 0.9484806664288044,
      "num_tokens": 84535000.0,
      "step": 11320
    },
    {
      "entropy": 1.8945983208715915,
      "epoch": 0.7567269721737533,
      "grad_norm": 0.3151988387107849,
      "learning_rate": 8.577736064039376e-06,
      "loss": 0.2394,
      "mean_token_accuracy": 0.9415588092058897,
      "num_tokens": 84607288.0,
      "step": 11330
    },
    {
      "entropy": 1.8958983808755874,
      "epoch": 0.7573948688835271,
      "grad_norm": 0.3211052715778351,
      "learning_rate": 8.53408481005377e-06,
      "loss": 0.2269,
      "mean_token_accuracy": 0.9449847292155027,
      "num_tokens": 84680202.0,
      "step": 11340
    },
    {
      "entropy": 1.9088520348072051,
      "epoch": 0.758062765593301,
      "grad_norm": 0.3037714660167694,
      "learning_rate": 8.490523355496742e-06,
      "loss": 0.2288,
      "mean_token_accuracy": 0.942930956184864,
      "num_tokens": 84757110.0,
      "step": 11350
    },
    {
      "entropy": 1.8966107808053494,
      "epoch": 0.7587306623030748,
      "grad_norm": 0.2812509536743164,
      "learning_rate": 8.447051937218244e-06,
      "loss": 0.2261,
      "mean_token_accuracy": 0.9439510587602854,
      "num_tokens": 84831847.0,
      "step": 11360
    },
    {
      "entropy": 1.895159225165844,
      "epoch": 0.7593985590128487,
      "grad_norm": 0.2748472988605499,
      "learning_rate": 8.403670791578702e-06,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9463084604591131,
      "num_tokens": 84905932.0,
      "step": 11370
    },
    {
      "entropy": 1.8941810354590416,
      "epoch": 0.7600664557226225,
      "grad_norm": 0.389934241771698,
      "learning_rate": 8.36038015444768e-06,
      "loss": 0.2334,
      "mean_token_accuracy": 0.9450410585850477,
      "num_tokens": 84983069.0,
      "step": 11380
    },
    {
      "entropy": 1.9433298885822297,
      "epoch": 0.7607343524323963,
      "grad_norm": 0.335888147354126,
      "learning_rate": 8.31718026120268e-06,
      "loss": 0.2218,
      "mean_token_accuracy": 0.9459588959813118,
      "num_tokens": 85055435.0,
      "step": 11390
    },
    {
      "entropy": 1.9400506347417832,
      "epoch": 0.7614022491421701,
      "grad_norm": 0.3339567482471466,
      "learning_rate": 8.274071346727768e-06,
      "loss": 0.2199,
      "mean_token_accuracy": 0.9434316240251064,
      "num_tokens": 85130097.0,
      "step": 11400
    },
    {
      "entropy": 1.906027965247631,
      "epoch": 0.762070145851944,
      "grad_norm": 0.3136909008026123,
      "learning_rate": 8.231053645412385e-06,
      "loss": 0.2228,
      "mean_token_accuracy": 0.945674081146717,
      "num_tokens": 85205660.0,
      "step": 11410
    },
    {
      "entropy": 1.8981819480657578,
      "epoch": 0.7627380425617178,
      "grad_norm": 0.36558428406715393,
      "learning_rate": 8.188127391150016e-06,
      "loss": 0.2122,
      "mean_token_accuracy": 0.9477827738970518,
      "num_tokens": 85279802.0,
      "step": 11420
    },
    {
      "entropy": 1.923303385823965,
      "epoch": 0.7634059392714917,
      "grad_norm": 0.2768043577671051,
      "learning_rate": 8.145292817336944e-06,
      "loss": 0.2092,
      "mean_token_accuracy": 0.9491335444152356,
      "num_tokens": 85355867.0,
      "step": 11430
    },
    {
      "entropy": 1.8939982995390892,
      "epoch": 0.7640738359812655,
      "grad_norm": 0.41168367862701416,
      "learning_rate": 8.102550156870952e-06,
      "loss": 0.222,
      "mean_token_accuracy": 0.946155945956707,
      "num_tokens": 85430653.0,
      "step": 11440
    },
    {
      "entropy": 1.9339173465967179,
      "epoch": 0.7647417326910393,
      "grad_norm": 0.2979883551597595,
      "learning_rate": 8.059899642150095e-06,
      "loss": 0.2184,
      "mean_token_accuracy": 0.945008360594511,
      "num_tokens": 85504074.0,
      "step": 11450
    },
    {
      "entropy": 1.9269521936774254,
      "epoch": 0.7654096294008131,
      "grad_norm": 0.3283374607563019,
      "learning_rate": 8.017341505071417e-06,
      "loss": 0.2245,
      "mean_token_accuracy": 0.9450273986905813,
      "num_tokens": 85576352.0,
      "step": 11460
    },
    {
      "entropy": 1.945946456491947,
      "epoch": 0.766077526110587,
      "grad_norm": 0.33567506074905396,
      "learning_rate": 7.974875977029684e-06,
      "loss": 0.2256,
      "mean_token_accuracy": 0.94437888674438,
      "num_tokens": 85648112.0,
      "step": 11470
    },
    {
      "entropy": 1.9503719687461853,
      "epoch": 0.7667454228203608,
      "grad_norm": 0.2934599816799164,
      "learning_rate": 7.932503288916149e-06,
      "loss": 0.213,
      "mean_token_accuracy": 0.9469171226024627,
      "num_tokens": 85723171.0,
      "step": 11480
    },
    {
      "entropy": 1.925863005965948,
      "epoch": 0.7674133195301347,
      "grad_norm": 0.2816382646560669,
      "learning_rate": 7.890223671117247e-06,
      "loss": 0.2131,
      "mean_token_accuracy": 0.94874166585505,
      "num_tokens": 85802098.0,
      "step": 11490
    },
    {
      "entropy": 1.9346450693905353,
      "epoch": 0.7680812162399085,
      "grad_norm": 0.3477094769477844,
      "learning_rate": 7.848037353513413e-06,
      "loss": 0.221,
      "mean_token_accuracy": 0.9449240598827601,
      "num_tokens": 85873887.0,
      "step": 11500
    },
    {
      "entropy": 1.9228808254003524,
      "epoch": 0.7687491129496823,
      "grad_norm": 0.29014065861701965,
      "learning_rate": 7.805944565477774e-06,
      "loss": 0.2255,
      "mean_token_accuracy": 0.9460843607783318,
      "num_tokens": 85947648.0,
      "step": 11510
    },
    {
      "entropy": 1.9236449711024761,
      "epoch": 0.7694170096594561,
      "grad_norm": 0.3228459060192108,
      "learning_rate": 7.763945535874938e-06,
      "loss": 0.2146,
      "mean_token_accuracy": 0.94769237190485,
      "num_tokens": 86023484.0,
      "step": 11520
    },
    {
      "entropy": 1.9296578779816627,
      "epoch": 0.77008490636923,
      "grad_norm": 0.3200177848339081,
      "learning_rate": 7.722040493059714e-06,
      "loss": 0.2229,
      "mean_token_accuracy": 0.9456036038696766,
      "num_tokens": 86098122.0,
      "step": 11530
    },
    {
      "entropy": 1.9002298817038537,
      "epoch": 0.7707528030790038,
      "grad_norm": 0.28655269742012024,
      "learning_rate": 7.68022966487591e-06,
      "loss": 0.2073,
      "mean_token_accuracy": 0.9483430609107018,
      "num_tokens": 86172617.0,
      "step": 11540
    },
    {
      "entropy": 1.9428872235119343,
      "epoch": 0.7714206997887777,
      "grad_norm": 0.31794604659080505,
      "learning_rate": 7.638513278655068e-06,
      "loss": 0.2274,
      "mean_token_accuracy": 0.9445866737514734,
      "num_tokens": 86246206.0,
      "step": 11550
    },
    {
      "entropy": 1.9186835162341596,
      "epoch": 0.7720885964985515,
      "grad_norm": 0.2955509126186371,
      "learning_rate": 7.596891561215235e-06,
      "loss": 0.2101,
      "mean_token_accuracy": 0.9489998806267976,
      "num_tokens": 86321585.0,
      "step": 11560
    },
    {
      "entropy": 1.9407105453312397,
      "epoch": 0.7727564932083253,
      "grad_norm": 0.3108774423599243,
      "learning_rate": 7.55536473885974e-06,
      "loss": 0.2202,
      "mean_token_accuracy": 0.9468152474611997,
      "num_tokens": 86396296.0,
      "step": 11570
    },
    {
      "entropy": 1.9339792892336845,
      "epoch": 0.7734243899180991,
      "grad_norm": 0.29260650277137756,
      "learning_rate": 7.513933037375928e-06,
      "loss": 0.2296,
      "mean_token_accuracy": 0.9442907094955444,
      "num_tokens": 86471896.0,
      "step": 11580
    },
    {
      "entropy": 1.9268646262586118,
      "epoch": 0.774092286627873,
      "grad_norm": 0.2828223407268524,
      "learning_rate": 7.472596682033987e-06,
      "loss": 0.2149,
      "mean_token_accuracy": 0.9474092531949282,
      "num_tokens": 86547339.0,
      "step": 11590
    },
    {
      "entropy": 1.9156761407852172,
      "epoch": 0.7747601833376468,
      "grad_norm": 0.3413717746734619,
      "learning_rate": 7.431355897585676e-06,
      "loss": 0.2384,
      "mean_token_accuracy": 0.9410256911069155,
      "num_tokens": 86622026.0,
      "step": 11600
    },
    {
      "entropy": 1.9566381104290484,
      "epoch": 0.7754280800474207,
      "grad_norm": 0.3666268587112427,
      "learning_rate": 7.39021090826313e-06,
      "loss": 0.237,
      "mean_token_accuracy": 0.9418189655989408,
      "num_tokens": 86694498.0,
      "step": 11610
    },
    {
      "entropy": 1.9231278404593468,
      "epoch": 0.7760959767571946,
      "grad_norm": 0.2729465663433075,
      "learning_rate": 7.3491619377776366e-06,
      "loss": 0.2196,
      "mean_token_accuracy": 0.9478519674390554,
      "num_tokens": 86770968.0,
      "step": 11620
    },
    {
      "entropy": 1.90982995480299,
      "epoch": 0.7767638734669683,
      "grad_norm": 0.30833160877227783,
      "learning_rate": 7.308209209318402e-06,
      "loss": 0.2068,
      "mean_token_accuracy": 0.9494866393506527,
      "num_tokens": 86846327.0,
      "step": 11630
    },
    {
      "entropy": 1.9449258789420127,
      "epoch": 0.7774317701767421,
      "grad_norm": 0.3366851210594177,
      "learning_rate": 7.267352945551344e-06,
      "loss": 0.2217,
      "mean_token_accuracy": 0.9434635043144226,
      "num_tokens": 86922728.0,
      "step": 11640
    },
    {
      "entropy": 1.914618470519781,
      "epoch": 0.778099666886516,
      "grad_norm": 0.3299575448036194,
      "learning_rate": 7.226593368617911e-06,
      "loss": 0.2346,
      "mean_token_accuracy": 0.9417701933532954,
      "num_tokens": 86998649.0,
      "step": 11650
    },
    {
      "entropy": 1.9171494588255882,
      "epoch": 0.7787675635962898,
      "grad_norm": 0.32758399844169617,
      "learning_rate": 7.185930700133839e-06,
      "loss": 0.2077,
      "mean_token_accuracy": 0.950836981460452,
      "num_tokens": 87071544.0,
      "step": 11660
    },
    {
      "entropy": 1.9156634002923965,
      "epoch": 0.7794354603060637,
      "grad_norm": 0.2863110601902008,
      "learning_rate": 7.145365161187959e-06,
      "loss": 0.2218,
      "mean_token_accuracy": 0.9448548093438148,
      "num_tokens": 87145390.0,
      "step": 11670
    },
    {
      "entropy": 1.9434235371649264,
      "epoch": 0.7801033570158376,
      "grad_norm": 0.30168262124061584,
      "learning_rate": 7.104896972341008e-06,
      "loss": 0.2215,
      "mean_token_accuracy": 0.9473841827362776,
      "num_tokens": 87224108.0,
      "step": 11680
    },
    {
      "entropy": 1.957365720719099,
      "epoch": 0.7807712537256113,
      "grad_norm": 0.3578754663467407,
      "learning_rate": 7.0645263536243785e-06,
      "loss": 0.2302,
      "mean_token_accuracy": 0.941792643815279,
      "num_tokens": 87298779.0,
      "step": 11690
    },
    {
      "entropy": 1.9179133355617524,
      "epoch": 0.7814391504353851,
      "grad_norm": 0.4008285105228424,
      "learning_rate": 7.024253524539003e-06,
      "loss": 0.2228,
      "mean_token_accuracy": 0.9455786202102899,
      "num_tokens": 87371678.0,
      "step": 11700
    },
    {
      "entropy": 1.9348440304398538,
      "epoch": 0.782107047145159,
      "grad_norm": 0.3001181483268738,
      "learning_rate": 6.984078704054092e-06,
      "loss": 0.2387,
      "mean_token_accuracy": 0.943041504919529,
      "num_tokens": 87445081.0,
      "step": 11710
    },
    {
      "entropy": 1.942325195670128,
      "epoch": 0.7827749438549328,
      "grad_norm": 0.34287431836128235,
      "learning_rate": 6.944002110605984e-06,
      "loss": 0.2373,
      "mean_token_accuracy": 0.9439044259488583,
      "num_tokens": 87519273.0,
      "step": 11720
    },
    {
      "entropy": 1.930584143102169,
      "epoch": 0.7834428405647067,
      "grad_norm": 0.25890690088272095,
      "learning_rate": 6.904023962096937e-06,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9463085103780031,
      "num_tokens": 87593384.0,
      "step": 11730
    },
    {
      "entropy": 1.921193440258503,
      "epoch": 0.7841107372744806,
      "grad_norm": 0.31245142221450806,
      "learning_rate": 6.864144475893937e-06,
      "loss": 0.2156,
      "mean_token_accuracy": 0.9478344406932593,
      "num_tokens": 87667522.0,
      "step": 11740
    },
    {
      "entropy": 1.8906810037791728,
      "epoch": 0.7847786339842543,
      "grad_norm": 0.3278355598449707,
      "learning_rate": 6.824363868827548e-06,
      "loss": 0.2125,
      "mean_token_accuracy": 0.9487185422331095,
      "num_tokens": 87746625.0,
      "step": 11750
    },
    {
      "entropy": 1.9176104918122292,
      "epoch": 0.7854465306940281,
      "grad_norm": 0.30767709016799927,
      "learning_rate": 6.784682357190701e-06,
      "loss": 0.2318,
      "mean_token_accuracy": 0.9401469543576241,
      "num_tokens": 87820796.0,
      "step": 11760
    },
    {
      "entropy": 1.91853381767869,
      "epoch": 0.786114427403802,
      "grad_norm": 0.2556558847427368,
      "learning_rate": 6.745100156737548e-06,
      "loss": 0.2072,
      "mean_token_accuracy": 0.9488288678228856,
      "num_tokens": 87896771.0,
      "step": 11770
    },
    {
      "entropy": 1.9115279085934163,
      "epoch": 0.7867823241135758,
      "grad_norm": 0.2916814088821411,
      "learning_rate": 6.7056174826822405e-06,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9439541105180979,
      "num_tokens": 87971536.0,
      "step": 11780
    },
    {
      "entropy": 1.9110812567174436,
      "epoch": 0.7874502208233497,
      "grad_norm": 0.2770801782608032,
      "learning_rate": 6.6662345496978184e-06,
      "loss": 0.1919,
      "mean_token_accuracy": 0.951653453707695,
      "num_tokens": 88046920.0,
      "step": 11790
    },
    {
      "entropy": 1.8751946441829204,
      "epoch": 0.7881181175331236,
      "grad_norm": 0.25386354327201843,
      "learning_rate": 6.626951571915007e-06,
      "loss": 0.2033,
      "mean_token_accuracy": 0.949246945977211,
      "num_tokens": 88122550.0,
      "step": 11800
    },
    {
      "entropy": 1.9106215745210648,
      "epoch": 0.7887860142428973,
      "grad_norm": 0.3737806975841522,
      "learning_rate": 6.587768762921058e-06,
      "loss": 0.2366,
      "mean_token_accuracy": 0.9426119856536388,
      "num_tokens": 88198512.0,
      "step": 11810
    },
    {
      "entropy": 1.8998952463269234,
      "epoch": 0.7894539109526711,
      "grad_norm": 0.29235225915908813,
      "learning_rate": 6.5486863357586e-06,
      "loss": 0.2112,
      "mean_token_accuracy": 0.9466220244765282,
      "num_tokens": 88273588.0,
      "step": 11820
    },
    {
      "entropy": 1.9093636505305767,
      "epoch": 0.790121807662445,
      "grad_norm": 0.341068297624588,
      "learning_rate": 6.509704502924448e-06,
      "loss": 0.2191,
      "mean_token_accuracy": 0.9445913273841142,
      "num_tokens": 88348515.0,
      "step": 11830
    },
    {
      "entropy": 1.92974793612957,
      "epoch": 0.7907897043722188,
      "grad_norm": 0.297319233417511,
      "learning_rate": 6.470823476368496e-06,
      "loss": 0.2297,
      "mean_token_accuracy": 0.9433461457490921,
      "num_tokens": 88423411.0,
      "step": 11840
    },
    {
      "entropy": 1.9271962210536002,
      "epoch": 0.7914576010819927,
      "grad_norm": 0.33146753907203674,
      "learning_rate": 6.432043467492533e-06,
      "loss": 0.2197,
      "mean_token_accuracy": 0.9456533152610064,
      "num_tokens": 88500680.0,
      "step": 11850
    },
    {
      "entropy": 1.9142560467123986,
      "epoch": 0.7921254977917666,
      "grad_norm": 0.2693401575088501,
      "learning_rate": 6.393364687149085e-06,
      "loss": 0.2248,
      "mean_token_accuracy": 0.9466971836984157,
      "num_tokens": 88573332.0,
      "step": 11860
    },
    {
      "entropy": 1.899408396333456,
      "epoch": 0.7927933945015403,
      "grad_norm": 0.260791540145874,
      "learning_rate": 6.354787345640301e-06,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9448356207460165,
      "num_tokens": 88649494.0,
      "step": 11870
    },
    {
      "entropy": 1.8895390599966049,
      "epoch": 0.7934612912113141,
      "grad_norm": 0.2668962776660919,
      "learning_rate": 6.316311652716799e-06,
      "loss": 0.2049,
      "mean_token_accuracy": 0.9486229028552771,
      "num_tokens": 88725208.0,
      "step": 11880
    },
    {
      "entropy": 1.919366579502821,
      "epoch": 0.794129187921088,
      "grad_norm": 0.3220490515232086,
      "learning_rate": 6.2779378175764905e-06,
      "loss": 0.238,
      "mean_token_accuracy": 0.9420530818402767,
      "num_tokens": 88797715.0,
      "step": 11890
    },
    {
      "entropy": 1.9052425637841224,
      "epoch": 0.7947970846308618,
      "grad_norm": 0.25093507766723633,
      "learning_rate": 6.239666048863495e-06,
      "loss": 0.2089,
      "mean_token_accuracy": 0.9471659358590842,
      "num_tokens": 88871536.0,
      "step": 11900
    },
    {
      "entropy": 1.9246492482721806,
      "epoch": 0.7954649813406357,
      "grad_norm": 0.3261584937572479,
      "learning_rate": 6.201496554666975e-06,
      "loss": 0.2203,
      "mean_token_accuracy": 0.9443686258047819,
      "num_tokens": 88945601.0,
      "step": 11910
    },
    {
      "entropy": 1.9137545451521873,
      "epoch": 0.7961328780504096,
      "grad_norm": 0.35194119811058044,
      "learning_rate": 6.163429542520008e-06,
      "loss": 0.2451,
      "mean_token_accuracy": 0.9419476479291916,
      "num_tokens": 89019858.0,
      "step": 11920
    },
    {
      "entropy": 1.9139618828892708,
      "epoch": 0.7968007747601833,
      "grad_norm": 0.32016295194625854,
      "learning_rate": 6.125465219398479e-06,
      "loss": 0.2376,
      "mean_token_accuracy": 0.9402606707066298,
      "num_tokens": 89092021.0,
      "step": 11930
    },
    {
      "entropy": 1.8968940548598767,
      "epoch": 0.7974686714699571,
      "grad_norm": 0.27857133746147156,
      "learning_rate": 6.087603791719909e-06,
      "loss": 0.2152,
      "mean_token_accuracy": 0.9455442499369383,
      "num_tokens": 89167921.0,
      "step": 11940
    },
    {
      "entropy": 1.9391052447259427,
      "epoch": 0.798136568179731,
      "grad_norm": 0.290174663066864,
      "learning_rate": 6.049845465342379e-06,
      "loss": 0.2353,
      "mean_token_accuracy": 0.942417386546731,
      "num_tokens": 89240609.0,
      "step": 11950
    },
    {
      "entropy": 1.9463041372597218,
      "epoch": 0.7988044648895049,
      "grad_norm": 0.30286282300949097,
      "learning_rate": 6.012190445563396e-06,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9446444526314736,
      "num_tokens": 89313669.0,
      "step": 11960
    },
    {
      "entropy": 1.9215276122093201,
      "epoch": 0.7994723615992787,
      "grad_norm": 0.28920048475265503,
      "learning_rate": 5.9746389371187655e-06,
      "loss": 0.2265,
      "mean_token_accuracy": 0.9458133615553379,
      "num_tokens": 89388749.0,
      "step": 11970
    },
    {
      "entropy": 1.8859002590179443,
      "epoch": 0.8001402583090526,
      "grad_norm": 0.43147513270378113,
      "learning_rate": 5.937191144181502e-06,
      "loss": 0.219,
      "mean_token_accuracy": 0.9450708650052547,
      "num_tokens": 89465190.0,
      "step": 11980
    },
    {
      "entropy": 1.912620224058628,
      "epoch": 0.8008081550188263,
      "grad_norm": 0.2776070535182953,
      "learning_rate": 5.899847270360674e-06,
      "loss": 0.2235,
      "mean_token_accuracy": 0.9443076837807893,
      "num_tokens": 89539439.0,
      "step": 11990
    },
    {
      "entropy": 1.927357778698206,
      "epoch": 0.8014760517286001,
      "grad_norm": 0.35332590341567993,
      "learning_rate": 5.862607518700353e-06,
      "loss": 0.2212,
      "mean_token_accuracy": 0.9499105460941791,
      "num_tokens": 89614509.0,
      "step": 12000
    },
    {
      "epoch": 0.8014760517286001,
      "eval_entropy": 1.9077495398283004,
      "eval_loss": 0.23475231230258942,
      "eval_mean_token_accuracy": 0.9464578078866005,
      "eval_num_tokens": 89614509.0,
      "eval_runtime": 552.6917,
      "eval_samples_per_second": 36.187,
      "eval_steps_per_second": 9.047,
      "step": 12000
    },
    {
      "entropy": 1.93703440874815,
      "epoch": 0.802143948438374,
      "grad_norm": 0.38874146342277527,
      "learning_rate": 5.825472091678475e-06,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9452136412262917,
      "num_tokens": 89690978.0,
      "step": 12010
    },
    {
      "entropy": 1.9058509640395642,
      "epoch": 0.8028118451481479,
      "grad_norm": 0.317646861076355,
      "learning_rate": 5.788441191205756e-06,
      "loss": 0.232,
      "mean_token_accuracy": 0.9432979535311461,
      "num_tokens": 89765550.0,
      "step": 12020
    },
    {
      "entropy": 1.9148917861282826,
      "epoch": 0.8034797418579217,
      "grad_norm": 0.2909880578517914,
      "learning_rate": 5.751515018624561e-06,
      "loss": 0.2018,
      "mean_token_accuracy": 0.9504593200981617,
      "num_tokens": 89839392.0,
      "step": 12030
    },
    {
      "entropy": 1.9205414779484271,
      "epoch": 0.8041476385676956,
      "grad_norm": 0.45270484685897827,
      "learning_rate": 5.714693774707862e-06,
      "loss": 0.2163,
      "mean_token_accuracy": 0.9461734648793936,
      "num_tokens": 89911436.0,
      "step": 12040
    },
    {
      "entropy": 1.9065644189715385,
      "epoch": 0.8048155352774693,
      "grad_norm": 0.3243974447250366,
      "learning_rate": 5.6779776596581065e-06,
      "loss": 0.2329,
      "mean_token_accuracy": 0.9426402520388365,
      "num_tokens": 89983728.0,
      "step": 12050
    },
    {
      "entropy": 1.8841743096709251,
      "epoch": 0.8054834319872431,
      "grad_norm": 0.284850150346756,
      "learning_rate": 5.6413668731061395e-06,
      "loss": 0.1984,
      "mean_token_accuracy": 0.949528944119811,
      "num_tokens": 90059480.0,
      "step": 12060
    },
    {
      "entropy": 1.9222078263759612,
      "epoch": 0.806151328697017,
      "grad_norm": 0.33089596033096313,
      "learning_rate": 5.604861614110132e-06,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9441944282501936,
      "num_tokens": 90133896.0,
      "step": 12070
    },
    {
      "entropy": 1.933812576532364,
      "epoch": 0.8068192254067909,
      "grad_norm": 0.2974908947944641,
      "learning_rate": 5.5684620811544684e-06,
      "loss": 0.2337,
      "mean_token_accuracy": 0.9459289290010929,
      "num_tokens": 90208380.0,
      "step": 12080
    },
    {
      "entropy": 1.9106913723051548,
      "epoch": 0.8074871221165647,
      "grad_norm": 0.3189352750778198,
      "learning_rate": 5.53216847214869e-06,
      "loss": 0.23,
      "mean_token_accuracy": 0.9435811273753643,
      "num_tokens": 90282721.0,
      "step": 12090
    },
    {
      "entropy": 1.9149140909314155,
      "epoch": 0.8081550188263386,
      "grad_norm": 0.3116757571697235,
      "learning_rate": 5.495980984426421e-06,
      "loss": 0.2239,
      "mean_token_accuracy": 0.944933907687664,
      "num_tokens": 90356135.0,
      "step": 12100
    },
    {
      "entropy": 1.8974915623664856,
      "epoch": 0.8088229155361123,
      "grad_norm": 0.29558226466178894,
      "learning_rate": 5.459899814744286e-06,
      "loss": 0.2247,
      "mean_token_accuracy": 0.946451997756958,
      "num_tokens": 90432327.0,
      "step": 12110
    },
    {
      "entropy": 1.9107790172100068,
      "epoch": 0.8094908122458861,
      "grad_norm": 0.31916290521621704,
      "learning_rate": 5.423925159280845e-06,
      "loss": 0.2183,
      "mean_token_accuracy": 0.9461866691708565,
      "num_tokens": 90506458.0,
      "step": 12120
    },
    {
      "entropy": 1.906505247950554,
      "epoch": 0.81015870895566,
      "grad_norm": 0.31288862228393555,
      "learning_rate": 5.388057213635526e-06,
      "loss": 0.2162,
      "mean_token_accuracy": 0.946871118247509,
      "num_tokens": 90582548.0,
      "step": 12130
    },
    {
      "entropy": 1.9007952421903611,
      "epoch": 0.8108266056654339,
      "grad_norm": 0.30547887086868286,
      "learning_rate": 5.352296172827547e-06,
      "loss": 0.2105,
      "mean_token_accuracy": 0.9467106480151415,
      "num_tokens": 90658389.0,
      "step": 12140
    },
    {
      "entropy": 1.952115386724472,
      "epoch": 0.8114945023752077,
      "grad_norm": 0.3439217805862427,
      "learning_rate": 5.316642231294882e-06,
      "loss": 0.2446,
      "mean_token_accuracy": 0.9407574523240327,
      "num_tokens": 90733808.0,
      "step": 12150
    },
    {
      "entropy": 1.9200212970376014,
      "epoch": 0.8121623990849816,
      "grad_norm": 0.30591586232185364,
      "learning_rate": 5.281095582893189e-06,
      "loss": 0.2336,
      "mean_token_accuracy": 0.9443939696997404,
      "num_tokens": 90805914.0,
      "step": 12160
    },
    {
      "entropy": 1.9278677806258202,
      "epoch": 0.8128302957947553,
      "grad_norm": 0.31881797313690186,
      "learning_rate": 5.245656420894755e-06,
      "loss": 0.2209,
      "mean_token_accuracy": 0.9470836572349072,
      "num_tokens": 90881121.0,
      "step": 12170
    },
    {
      "entropy": 1.90970710888505,
      "epoch": 0.8134981925045291,
      "grad_norm": 0.2699946463108063,
      "learning_rate": 5.210324937987452e-06,
      "loss": 0.2129,
      "mean_token_accuracy": 0.9496553532779217,
      "num_tokens": 90955665.0,
      "step": 12180
    },
    {
      "entropy": 1.9190218165516852,
      "epoch": 0.814166089214303,
      "grad_norm": 0.28772151470184326,
      "learning_rate": 5.175101326273678e-06,
      "loss": 0.2164,
      "mean_token_accuracy": 0.9481560837477445,
      "num_tokens": 91029252.0,
      "step": 12190
    },
    {
      "entropy": 1.9169261828064919,
      "epoch": 0.8148339859240769,
      "grad_norm": 0.28831717371940613,
      "learning_rate": 5.139985777269319e-06,
      "loss": 0.2198,
      "mean_token_accuracy": 0.9456579513847828,
      "num_tokens": 91101994.0,
      "step": 12200
    },
    {
      "entropy": 1.8855088956654071,
      "epoch": 0.8155018826338507,
      "grad_norm": 0.24257218837738037,
      "learning_rate": 5.104978481902719e-06,
      "loss": 0.2005,
      "mean_token_accuracy": 0.9510732624679804,
      "num_tokens": 91180249.0,
      "step": 12210
    },
    {
      "entropy": 1.911930686980486,
      "epoch": 0.8161697793436246,
      "grad_norm": 0.3006708323955536,
      "learning_rate": 5.0700796305136255e-06,
      "loss": 0.2231,
      "mean_token_accuracy": 0.945105604082346,
      "num_tokens": 91257831.0,
      "step": 12220
    },
    {
      "entropy": 1.9221073605120182,
      "epoch": 0.8168376760533983,
      "grad_norm": 0.3255353271961212,
      "learning_rate": 5.0352894128521634e-06,
      "loss": 0.2364,
      "mean_token_accuracy": 0.939706540107727,
      "num_tokens": 91332692.0,
      "step": 12230
    },
    {
      "entropy": 1.9181040100753308,
      "epoch": 0.8175055727631722,
      "grad_norm": 0.2864046096801758,
      "learning_rate": 5.0006080180777865e-06,
      "loss": 0.2156,
      "mean_token_accuracy": 0.9446999654173851,
      "num_tokens": 91405687.0,
      "step": 12240
    },
    {
      "entropy": 1.9098816119134425,
      "epoch": 0.818173469472946,
      "grad_norm": 0.2883688509464264,
      "learning_rate": 4.9660356347582764e-06,
      "loss": 0.2213,
      "mean_token_accuracy": 0.9463308542966843,
      "num_tokens": 91480106.0,
      "step": 12250
    },
    {
      "entropy": 1.9178796097636224,
      "epoch": 0.8188413661827199,
      "grad_norm": 0.2966650128364563,
      "learning_rate": 4.931572450868704e-06,
      "loss": 0.2223,
      "mean_token_accuracy": 0.9446263413876295,
      "num_tokens": 91552284.0,
      "step": 12260
    },
    {
      "entropy": 1.892970908433199,
      "epoch": 0.8195092628924937,
      "grad_norm": 0.31876063346862793,
      "learning_rate": 4.8972186537904055e-06,
      "loss": 0.214,
      "mean_token_accuracy": 0.945142986625433,
      "num_tokens": 91625080.0,
      "step": 12270
    },
    {
      "entropy": 1.9201669298112392,
      "epoch": 0.8201771596022676,
      "grad_norm": 0.296844482421875,
      "learning_rate": 4.862974430309952e-06,
      "loss": 0.2212,
      "mean_token_accuracy": 0.9442084044218063,
      "num_tokens": 91700662.0,
      "step": 12280
    },
    {
      "entropy": 1.9088800638914107,
      "epoch": 0.8208450563120413,
      "grad_norm": 0.27243557572364807,
      "learning_rate": 4.828839966618166e-06,
      "loss": 0.2263,
      "mean_token_accuracy": 0.9437139257788658,
      "num_tokens": 91773943.0,
      "step": 12290
    },
    {
      "entropy": 1.9159510530531407,
      "epoch": 0.8215129530218152,
      "grad_norm": 0.35071954131126404,
      "learning_rate": 4.79481544830909e-06,
      "loss": 0.2198,
      "mean_token_accuracy": 0.9465903297066689,
      "num_tokens": 91851143.0,
      "step": 12300
    },
    {
      "entropy": 1.9333980567753315,
      "epoch": 0.822180849731589,
      "grad_norm": 0.25562265515327454,
      "learning_rate": 4.7609010603789495e-06,
      "loss": 0.2198,
      "mean_token_accuracy": 0.9474662080407142,
      "num_tokens": 91926345.0,
      "step": 12310
    },
    {
      "entropy": 1.9161305882036685,
      "epoch": 0.8228487464413629,
      "grad_norm": 0.3357928991317749,
      "learning_rate": 4.7270969872252085e-06,
      "loss": 0.2257,
      "mean_token_accuracy": 0.9430730618536473,
      "num_tokens": 92002132.0,
      "step": 12320
    },
    {
      "entropy": 1.9020003490149975,
      "epoch": 0.8235166431511367,
      "grad_norm": 0.2852373421192169,
      "learning_rate": 4.693403412645523e-06,
      "loss": 0.2228,
      "mean_token_accuracy": 0.9454558629542589,
      "num_tokens": 92077437.0,
      "step": 12330
    },
    {
      "entropy": 1.9694581009447574,
      "epoch": 0.8241845398609106,
      "grad_norm": 0.29638954997062683,
      "learning_rate": 4.6598205198367375e-06,
      "loss": 0.229,
      "mean_token_accuracy": 0.9444546531885862,
      "num_tokens": 92151258.0,
      "step": 12340
    },
    {
      "entropy": 1.9170562088489533,
      "epoch": 0.8248524365706843,
      "grad_norm": 0.29127585887908936,
      "learning_rate": 4.626348491393925e-06,
      "loss": 0.2213,
      "mean_token_accuracy": 0.9467327196151019,
      "num_tokens": 92225354.0,
      "step": 12350
    },
    {
      "entropy": 1.9222478553652764,
      "epoch": 0.8255203332804582,
      "grad_norm": 0.3237292766571045,
      "learning_rate": 4.592987509309356e-06,
      "loss": 0.2212,
      "mean_token_accuracy": 0.947163712978363,
      "num_tokens": 92300008.0,
      "step": 12360
    },
    {
      "entropy": 1.877466270327568,
      "epoch": 0.826188229990232,
      "grad_norm": 0.2953042685985565,
      "learning_rate": 4.559737754971541e-06,
      "loss": 0.2083,
      "mean_token_accuracy": 0.947779618948698,
      "num_tokens": 92375963.0,
      "step": 12370
    },
    {
      "entropy": 1.8960151240229606,
      "epoch": 0.8268561267000059,
      "grad_norm": 0.3921753764152527,
      "learning_rate": 4.526599409164226e-06,
      "loss": 0.2056,
      "mean_token_accuracy": 0.9488136760890484,
      "num_tokens": 92453021.0,
      "step": 12380
    },
    {
      "entropy": 1.9146280325949192,
      "epoch": 0.8275240234097797,
      "grad_norm": 0.29990142583847046,
      "learning_rate": 4.493572652065391e-06,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9465552151203156,
      "num_tokens": 92526342.0,
      "step": 12390
    },
    {
      "entropy": 1.9248810023069383,
      "epoch": 0.8281919201195536,
      "grad_norm": 0.32960548996925354,
      "learning_rate": 4.460657663246322e-06,
      "loss": 0.2199,
      "mean_token_accuracy": 0.9430462893098592,
      "num_tokens": 92602652.0,
      "step": 12400
    },
    {
      "entropy": 1.9142834946513176,
      "epoch": 0.8288598168293273,
      "grad_norm": 0.26918256282806396,
      "learning_rate": 4.427854621670583e-06,
      "loss": 0.2264,
      "mean_token_accuracy": 0.9485366005450487,
      "num_tokens": 92677039.0,
      "step": 12410
    },
    {
      "entropy": 1.9398212164640427,
      "epoch": 0.8295277135391012,
      "grad_norm": 0.28655320405960083,
      "learning_rate": 4.395163705693083e-06,
      "loss": 0.2316,
      "mean_token_accuracy": 0.9404951419681311,
      "num_tokens": 92749309.0,
      "step": 12420
    },
    {
      "entropy": 1.9200722940266133,
      "epoch": 0.830195610248875,
      "grad_norm": 0.3873966634273529,
      "learning_rate": 4.362585093059077e-06,
      "loss": 0.2307,
      "mean_token_accuracy": 0.9440862823277711,
      "num_tokens": 92825786.0,
      "step": 12430
    },
    {
      "entropy": 1.9236732959747314,
      "epoch": 0.8308635069586489,
      "grad_norm": 0.3685950040817261,
      "learning_rate": 4.330118960903201e-06,
      "loss": 0.2276,
      "mean_token_accuracy": 0.9434462014585734,
      "num_tokens": 92900267.0,
      "step": 12440
    },
    {
      "entropy": 1.8944612585008145,
      "epoch": 0.8315314036684227,
      "grad_norm": 0.30650460720062256,
      "learning_rate": 4.297765485748536e-06,
      "loss": 0.2193,
      "mean_token_accuracy": 0.9466446179896593,
      "num_tokens": 92975421.0,
      "step": 12450
    },
    {
      "entropy": 1.9075546264648438,
      "epoch": 0.8321993003781966,
      "grad_norm": 0.3539635241031647,
      "learning_rate": 4.265524843505623e-06,
      "loss": 0.2328,
      "mean_token_accuracy": 0.9422083970159292,
      "num_tokens": 93050007.0,
      "step": 12460
    },
    {
      "entropy": 1.9267285540699959,
      "epoch": 0.8328671970879703,
      "grad_norm": 0.318481981754303,
      "learning_rate": 4.233397209471513e-06,
      "loss": 0.234,
      "mean_token_accuracy": 0.9424730904400349,
      "num_tokens": 93124474.0,
      "step": 12470
    },
    {
      "entropy": 1.9024205163121224,
      "epoch": 0.8335350937977442,
      "grad_norm": 0.32240915298461914,
      "learning_rate": 4.201382758328825e-06,
      "loss": 0.2117,
      "mean_token_accuracy": 0.9464437153190375,
      "num_tokens": 93201345.0,
      "step": 12480
    },
    {
      "entropy": 1.9199492737650872,
      "epoch": 0.834202990507518,
      "grad_norm": 0.288259893655777,
      "learning_rate": 4.169481664144765e-06,
      "loss": 0.2358,
      "mean_token_accuracy": 0.9420751929283142,
      "num_tokens": 93274253.0,
      "step": 12490
    },
    {
      "entropy": 1.9158204957842826,
      "epoch": 0.8348708872172919,
      "grad_norm": 0.27644994854927063,
      "learning_rate": 4.137694100370218e-06,
      "loss": 0.2193,
      "mean_token_accuracy": 0.9459875266999006,
      "num_tokens": 93351974.0,
      "step": 12500
    },
    {
      "entropy": 1.9240369744598866,
      "epoch": 0.8355387839270657,
      "grad_norm": 0.29895466566085815,
      "learning_rate": 4.1060202398387816e-06,
      "loss": 0.2472,
      "mean_token_accuracy": 0.9408362310379743,
      "num_tokens": 93426423.0,
      "step": 12510
    },
    {
      "entropy": 1.9075682036578656,
      "epoch": 0.8362066806368396,
      "grad_norm": 0.3330353796482086,
      "learning_rate": 4.074460254765844e-06,
      "loss": 0.2124,
      "mean_token_accuracy": 0.94851104170084,
      "num_tokens": 93501709.0,
      "step": 12520
    },
    {
      "entropy": 1.909358086436987,
      "epoch": 0.8368745773466133,
      "grad_norm": 0.28923672437667847,
      "learning_rate": 4.043014316747614e-06,
      "loss": 0.2163,
      "mean_token_accuracy": 0.9456011939793825,
      "num_tokens": 93577318.0,
      "step": 12530
    },
    {
      "entropy": 1.899984909594059,
      "epoch": 0.8375424740563872,
      "grad_norm": 0.2875509262084961,
      "learning_rate": 4.011682596760221e-06,
      "loss": 0.2292,
      "mean_token_accuracy": 0.9445662580430507,
      "num_tokens": 93652909.0,
      "step": 12540
    },
    {
      "entropy": 1.906442727893591,
      "epoch": 0.838210370766161,
      "grad_norm": 0.35136204957962036,
      "learning_rate": 3.980465265158779e-06,
      "loss": 0.2164,
      "mean_token_accuracy": 0.9453341763466596,
      "num_tokens": 93726521.0,
      "step": 12550
    },
    {
      "entropy": 1.905178725719452,
      "epoch": 0.8388782674759349,
      "grad_norm": 0.23824502527713776,
      "learning_rate": 3.949362491676454e-06,
      "loss": 0.2187,
      "mean_token_accuracy": 0.9441476624459029,
      "num_tokens": 93800759.0,
      "step": 12560
    },
    {
      "entropy": 1.8736170925199986,
      "epoch": 0.8395461641857087,
      "grad_norm": 0.26656362414360046,
      "learning_rate": 3.918374445423536e-06,
      "loss": 0.2015,
      "mean_token_accuracy": 0.9506821319460869,
      "num_tokens": 93877628.0,
      "step": 12570
    },
    {
      "entropy": 1.8988473922014237,
      "epoch": 0.8402140608954826,
      "grad_norm": 0.3266134560108185,
      "learning_rate": 3.8875012948865406e-06,
      "loss": 0.2191,
      "mean_token_accuracy": 0.9459964755922556,
      "num_tokens": 93951331.0,
      "step": 12580
    },
    {
      "entropy": 1.8885542929172516,
      "epoch": 0.8408819576052563,
      "grad_norm": 0.2893097996711731,
      "learning_rate": 3.8567432079272575e-06,
      "loss": 0.2026,
      "mean_token_accuracy": 0.9486055232584476,
      "num_tokens": 94024591.0,
      "step": 12590
    },
    {
      "entropy": 1.9056413725018502,
      "epoch": 0.8415498543150302,
      "grad_norm": 0.33855727314949036,
      "learning_rate": 3.826100351781877e-06,
      "loss": 0.2062,
      "mean_token_accuracy": 0.9487339001148939,
      "num_tokens": 94099122.0,
      "step": 12600
    },
    {
      "entropy": 1.9189244344830514,
      "epoch": 0.842217751024804,
      "grad_norm": 0.3240315616130829,
      "learning_rate": 3.7955728930600537e-06,
      "loss": 0.2217,
      "mean_token_accuracy": 0.9466369111090899,
      "num_tokens": 94171605.0,
      "step": 12610
    },
    {
      "entropy": 1.9320820726454258,
      "epoch": 0.8428856477345779,
      "grad_norm": 0.3500101566314697,
      "learning_rate": 3.7651609977440132e-06,
      "loss": 0.2376,
      "mean_token_accuracy": 0.9429630015045405,
      "num_tokens": 94245534.0,
      "step": 12620
    },
    {
      "entropy": 1.899515087902546,
      "epoch": 0.8435535444443517,
      "grad_norm": 0.2735781669616699,
      "learning_rate": 3.734864831187655e-06,
      "loss": 0.2134,
      "mean_token_accuracy": 0.9473102346062661,
      "num_tokens": 94321308.0,
      "step": 12630
    },
    {
      "entropy": 1.879213922470808,
      "epoch": 0.8442214411541256,
      "grad_norm": 0.38318684697151184,
      "learning_rate": 3.704684558115619e-06,
      "loss": 0.2158,
      "mean_token_accuracy": 0.9502399206161499,
      "num_tokens": 94396459.0,
      "step": 12640
    },
    {
      "entropy": 1.914496787637472,
      "epoch": 0.8448893378638993,
      "grad_norm": 0.28392308950424194,
      "learning_rate": 3.6746203426224394e-06,
      "loss": 0.2036,
      "mean_token_accuracy": 0.9497009128332138,
      "num_tokens": 94469224.0,
      "step": 12650
    },
    {
      "entropy": 1.911051009595394,
      "epoch": 0.8455572345736732,
      "grad_norm": 0.3505001962184906,
      "learning_rate": 3.6446723481716195e-06,
      "loss": 0.2386,
      "mean_token_accuracy": 0.9406479235738516,
      "num_tokens": 94541886.0,
      "step": 12660
    },
    {
      "entropy": 1.9401042722165585,
      "epoch": 0.846225131283447,
      "grad_norm": 0.3758535087108612,
      "learning_rate": 3.614840737594754e-06,
      "loss": 0.2085,
      "mean_token_accuracy": 0.9472375724464654,
      "num_tokens": 94616074.0,
      "step": 12670
    },
    {
      "entropy": 1.9013698816299438,
      "epoch": 0.8468930279932209,
      "grad_norm": 0.3473319411277771,
      "learning_rate": 3.585125673090644e-06,
      "loss": 0.2088,
      "mean_token_accuracy": 0.9465346351265908,
      "num_tokens": 94690957.0,
      "step": 12680
    },
    {
      "entropy": 1.9082304663956164,
      "epoch": 0.8475609247029947,
      "grad_norm": 0.2587113678455353,
      "learning_rate": 3.555527316224398e-06,
      "loss": 0.2067,
      "mean_token_accuracy": 0.9490227155387402,
      "num_tokens": 94767652.0,
      "step": 12690
    },
    {
      "entropy": 1.8926838703453541,
      "epoch": 0.8482288214127686,
      "grad_norm": 0.31046247482299805,
      "learning_rate": 3.526045827926585e-06,
      "loss": 0.2177,
      "mean_token_accuracy": 0.9462960693985224,
      "num_tokens": 94843529.0,
      "step": 12700
    },
    {
      "entropy": 1.9144454523921013,
      "epoch": 0.8488967181225423,
      "grad_norm": 0.32953667640686035,
      "learning_rate": 3.4966813684923384e-06,
      "loss": 0.2272,
      "mean_token_accuracy": 0.9424541935324668,
      "num_tokens": 94916688.0,
      "step": 12710
    },
    {
      "entropy": 1.913926786929369,
      "epoch": 0.8495646148323162,
      "grad_norm": 0.3478092849254608,
      "learning_rate": 3.467434097580493e-06,
      "loss": 0.2153,
      "mean_token_accuracy": 0.9467847377061844,
      "num_tokens": 94991692.0,
      "step": 12720
    },
    {
      "entropy": 1.896678277105093,
      "epoch": 0.85023251154209,
      "grad_norm": 0.3317002058029175,
      "learning_rate": 3.438304174212693e-06,
      "loss": 0.2207,
      "mean_token_accuracy": 0.9460543278604746,
      "num_tokens": 95068804.0,
      "step": 12730
    },
    {
      "entropy": 1.9101593144237996,
      "epoch": 0.8509004082518639,
      "grad_norm": 0.30426764488220215,
      "learning_rate": 3.40929175677257e-06,
      "loss": 0.227,
      "mean_token_accuracy": 0.9424716554582119,
      "num_tokens": 95144528.0,
      "step": 12740
    },
    {
      "entropy": 1.9011964708566667,
      "epoch": 0.8515683049616377,
      "grad_norm": 0.2705959379673004,
      "learning_rate": 3.3803970030048434e-06,
      "loss": 0.2013,
      "mean_token_accuracy": 0.947756539285183,
      "num_tokens": 95218276.0,
      "step": 12750
    },
    {
      "entropy": 1.9015478432178496,
      "epoch": 0.8522362016714116,
      "grad_norm": 0.2622295916080475,
      "learning_rate": 3.351620070014497e-06,
      "loss": 0.2072,
      "mean_token_accuracy": 0.9473953269422054,
      "num_tokens": 95293257.0,
      "step": 12760
    },
    {
      "entropy": 1.9096700623631477,
      "epoch": 0.8529040983811853,
      "grad_norm": 0.3143070340156555,
      "learning_rate": 3.3229611142658747e-06,
      "loss": 0.2255,
      "mean_token_accuracy": 0.9448044616729021,
      "num_tokens": 95368100.0,
      "step": 12770
    },
    {
      "entropy": 1.9495323419570922,
      "epoch": 0.8535719950909592,
      "grad_norm": 0.30113184452056885,
      "learning_rate": 3.294420291581897e-06,
      "loss": 0.2192,
      "mean_token_accuracy": 0.9458323460072279,
      "num_tokens": 95442181.0,
      "step": 12780
    },
    {
      "entropy": 1.9024573378264904,
      "epoch": 0.854239891800733,
      "grad_norm": 0.337389200925827,
      "learning_rate": 3.2659977571431473e-06,
      "loss": 0.2146,
      "mean_token_accuracy": 0.9440126322209835,
      "num_tokens": 95515226.0,
      "step": 12790
    },
    {
      "entropy": 1.8989201940596103,
      "epoch": 0.8549077885105069,
      "grad_norm": 0.2654149830341339,
      "learning_rate": 3.2376936654870746e-06,
      "loss": 0.2219,
      "mean_token_accuracy": 0.9450605168938637,
      "num_tokens": 95590563.0,
      "step": 12800
    },
    {
      "entropy": 1.897411823272705,
      "epoch": 0.8555756852202807,
      "grad_norm": 0.29004570841789246,
      "learning_rate": 3.209508170507135e-06,
      "loss": 0.2006,
      "mean_token_accuracy": 0.9493835728615523,
      "num_tokens": 95667791.0,
      "step": 12810
    },
    {
      "entropy": 1.9329859793186188,
      "epoch": 0.8562435819300546,
      "grad_norm": 0.31763482093811035,
      "learning_rate": 3.1814414254519602e-06,
      "loss": 0.2144,
      "mean_token_accuracy": 0.9449219807982445,
      "num_tokens": 95743162.0,
      "step": 12820
    },
    {
      "entropy": 1.95360249504447,
      "epoch": 0.8569114786398283,
      "grad_norm": 0.35177379846572876,
      "learning_rate": 3.1534935829245176e-06,
      "loss": 0.2137,
      "mean_token_accuracy": 0.9477232187986374,
      "num_tokens": 95819891.0,
      "step": 12830
    },
    {
      "entropy": 1.914057232439518,
      "epoch": 0.8575793753496022,
      "grad_norm": 0.3635084927082062,
      "learning_rate": 3.1256647948812793e-06,
      "loss": 0.2234,
      "mean_token_accuracy": 0.94504219815135,
      "num_tokens": 95894739.0,
      "step": 12840
    },
    {
      "entropy": 1.9173757053911686,
      "epoch": 0.858247272059376,
      "grad_norm": 0.2878390848636627,
      "learning_rate": 3.097955212631406e-06,
      "loss": 0.234,
      "mean_token_accuracy": 0.9414523094892502,
      "num_tokens": 95969087.0,
      "step": 12850
    },
    {
      "entropy": 1.920248629897833,
      "epoch": 0.8589151687691499,
      "grad_norm": 0.37370526790618896,
      "learning_rate": 3.0703649868359225e-06,
      "loss": 0.2305,
      "mean_token_accuracy": 0.943334087356925,
      "num_tokens": 96044766.0,
      "step": 12860
    },
    {
      "entropy": 1.9216134034097194,
      "epoch": 0.8595830654789237,
      "grad_norm": 0.34205037355422974,
      "learning_rate": 3.042894267506891e-06,
      "loss": 0.2413,
      "mean_token_accuracy": 0.9426057361066341,
      "num_tokens": 96119491.0,
      "step": 12870
    },
    {
      "entropy": 1.9309019044041633,
      "epoch": 0.8602509621886976,
      "grad_norm": 0.3354184627532959,
      "learning_rate": 3.015543204006604e-06,
      "loss": 0.2353,
      "mean_token_accuracy": 0.9419073890894651,
      "num_tokens": 96193036.0,
      "step": 12880
    },
    {
      "entropy": 1.8997986532747746,
      "epoch": 0.8609188588984713,
      "grad_norm": 0.2701426148414612,
      "learning_rate": 2.9883119450467557e-06,
      "loss": 0.2171,
      "mean_token_accuracy": 0.947497657686472,
      "num_tokens": 96267893.0,
      "step": 12890
    },
    {
      "entropy": 1.8968277364969253,
      "epoch": 0.8615867556082452,
      "grad_norm": 0.2995654344558716,
      "learning_rate": 2.9612006386876613e-06,
      "loss": 0.2063,
      "mean_token_accuracy": 0.9469488143920899,
      "num_tokens": 96342369.0,
      "step": 12900
    },
    {
      "entropy": 1.8971393764019013,
      "epoch": 0.862254652318019,
      "grad_norm": 0.33524850010871887,
      "learning_rate": 2.9342094323374267e-06,
      "loss": 0.2232,
      "mean_token_accuracy": 0.9447850570082664,
      "num_tokens": 96416594.0,
      "step": 12910
    },
    {
      "entropy": 1.918541420251131,
      "epoch": 0.8629225490277929,
      "grad_norm": 0.2766539752483368,
      "learning_rate": 2.9073384727511606e-06,
      "loss": 0.2187,
      "mean_token_accuracy": 0.9467494156211614,
      "num_tokens": 96490004.0,
      "step": 12920
    },
    {
      "entropy": 1.907379398494959,
      "epoch": 0.8635904457375667,
      "grad_norm": 0.341058611869812,
      "learning_rate": 2.8805879060301734e-06,
      "loss": 0.2259,
      "mean_token_accuracy": 0.9431953445076943,
      "num_tokens": 96564373.0,
      "step": 12930
    },
    {
      "entropy": 1.8784002639353274,
      "epoch": 0.8642583424473406,
      "grad_norm": 0.3375591039657593,
      "learning_rate": 2.8539578776211743e-06,
      "loss": 0.2052,
      "mean_token_accuracy": 0.9507540252059699,
      "num_tokens": 96639314.0,
      "step": 12940
    },
    {
      "entropy": 1.9218274757266045,
      "epoch": 0.8649262391571143,
      "grad_norm": 0.3263833820819855,
      "learning_rate": 2.827448532315497e-06,
      "loss": 0.2044,
      "mean_token_accuracy": 0.9460774265229702,
      "num_tokens": 96712931.0,
      "step": 12950
    },
    {
      "entropy": 1.895472052693367,
      "epoch": 0.8655941358668882,
      "grad_norm": 0.32461920380592346,
      "learning_rate": 2.8010600142482977e-06,
      "loss": 0.2131,
      "mean_token_accuracy": 0.9485759288072586,
      "num_tokens": 96784654.0,
      "step": 12960
    },
    {
      "entropy": 1.897088672965765,
      "epoch": 0.866262032576662,
      "grad_norm": 0.2537863850593567,
      "learning_rate": 2.774792466897791e-06,
      "loss": 0.2108,
      "mean_token_accuracy": 0.9478284027427435,
      "num_tokens": 96860264.0,
      "step": 12970
    },
    {
      "entropy": 1.8856997050344944,
      "epoch": 0.8669299292864359,
      "grad_norm": 0.278632253408432,
      "learning_rate": 2.7486460330844283e-06,
      "loss": 0.2048,
      "mean_token_accuracy": 0.9465761728584766,
      "num_tokens": 96936431.0,
      "step": 12980
    },
    {
      "entropy": 1.8715303860604764,
      "epoch": 0.8675978259962097,
      "grad_norm": 0.26264268159866333,
      "learning_rate": 2.7226208549701793e-06,
      "loss": 0.2012,
      "mean_token_accuracy": 0.9532707251608372,
      "num_tokens": 97012887.0,
      "step": 12990
    },
    {
      "entropy": 1.884982492774725,
      "epoch": 0.8682657227059836,
      "grad_norm": 0.2999236583709717,
      "learning_rate": 2.696717074057707e-06,
      "loss": 0.214,
      "mean_token_accuracy": 0.9472188141196967,
      "num_tokens": 97091533.0,
      "step": 13000
    },
    {
      "epoch": 0.8682657227059836,
      "eval_entropy": 1.891543303322792,
      "eval_loss": 0.23422756791114807,
      "eval_mean_token_accuracy": 0.9465205826520919,
      "eval_num_tokens": 97091533.0,
      "eval_runtime": 553.8499,
      "eval_samples_per_second": 36.111,
      "eval_steps_per_second": 9.028,
      "step": 13000
    },
    {
      "entropy": 1.92298434227705,
      "epoch": 0.8689336194157573,
      "grad_norm": 0.34577375650405884,
      "learning_rate": 2.670934831189634e-06,
      "loss": 0.2283,
      "mean_token_accuracy": 0.9452300623059273,
      "num_tokens": 97165811.0,
      "step": 13010
    },
    {
      "entropy": 1.8975606806576253,
      "epoch": 0.8696015161255312,
      "grad_norm": 0.2593967020511627,
      "learning_rate": 2.6452742665477545e-06,
      "loss": 0.2227,
      "mean_token_accuracy": 0.9417459007352591,
      "num_tokens": 97240731.0,
      "step": 13020
    },
    {
      "entropy": 1.8891439154744147,
      "epoch": 0.870269412835305,
      "grad_norm": 0.3155575394630432,
      "learning_rate": 2.619735519652291e-06,
      "loss": 0.2073,
      "mean_token_accuracy": 0.9488394863903522,
      "num_tokens": 97316521.0,
      "step": 13030
    },
    {
      "entropy": 1.946111287176609,
      "epoch": 0.8709373095450789,
      "grad_norm": 0.35275089740753174,
      "learning_rate": 2.594318729361106e-06,
      "loss": 0.2192,
      "mean_token_accuracy": 0.9465547289699316,
      "num_tokens": 97393093.0,
      "step": 13040
    },
    {
      "entropy": 1.8990215368568897,
      "epoch": 0.8716052062548527,
      "grad_norm": 0.3260514736175537,
      "learning_rate": 2.5690240338689856e-06,
      "loss": 0.2132,
      "mean_token_accuracy": 0.9487630125135184,
      "num_tokens": 97466457.0,
      "step": 13050
    },
    {
      "entropy": 1.8967769339680671,
      "epoch": 0.8722731029646266,
      "grad_norm": 0.3189840614795685,
      "learning_rate": 2.543851570706858e-06,
      "loss": 0.2137,
      "mean_token_accuracy": 0.9483622338622808,
      "num_tokens": 97539486.0,
      "step": 13060
    },
    {
      "entropy": 1.8951519280672073,
      "epoch": 0.8729409996744003,
      "grad_norm": 0.3604692220687866,
      "learning_rate": 2.5188014767410647e-06,
      "loss": 0.212,
      "mean_token_accuracy": 0.942385621741414,
      "num_tokens": 97613622.0,
      "step": 13070
    },
    {
      "entropy": 1.8882951810956001,
      "epoch": 0.8736088963841742,
      "grad_norm": 0.31977880001068115,
      "learning_rate": 2.493873888172605e-06,
      "loss": 0.2083,
      "mean_token_accuracy": 0.947973096370697,
      "num_tokens": 97688284.0,
      "step": 13080
    },
    {
      "entropy": 1.917658195644617,
      "epoch": 0.874276793093948,
      "grad_norm": 0.29543742537498474,
      "learning_rate": 2.4690689405363868e-06,
      "loss": 0.2386,
      "mean_token_accuracy": 0.9422120042145252,
      "num_tokens": 97760135.0,
      "step": 13090
    },
    {
      "entropy": 1.8869654476642608,
      "epoch": 0.8749446898037219,
      "grad_norm": 0.39594390988349915,
      "learning_rate": 2.4443867687005175e-06,
      "loss": 0.2197,
      "mean_token_accuracy": 0.9461137805134058,
      "num_tokens": 97835639.0,
      "step": 13100
    },
    {
      "entropy": 1.9309095345437526,
      "epoch": 0.8756125865134957,
      "grad_norm": 0.3087207078933716,
      "learning_rate": 2.419827506865549e-06,
      "loss": 0.2163,
      "mean_token_accuracy": 0.9479531478136778,
      "num_tokens": 97909477.0,
      "step": 13110
    },
    {
      "entropy": 1.9147302009165288,
      "epoch": 0.8762804832232696,
      "grad_norm": 0.32227468490600586,
      "learning_rate": 2.3953912885637522e-06,
      "loss": 0.2091,
      "mean_token_accuracy": 0.9483824409544468,
      "num_tokens": 97983235.0,
      "step": 13120
    },
    {
      "entropy": 1.8944000601768494,
      "epoch": 0.8769483799330433,
      "grad_norm": 0.27947521209716797,
      "learning_rate": 2.371078246658399e-06,
      "loss": 0.2183,
      "mean_token_accuracy": 0.9452764131128788,
      "num_tokens": 98057045.0,
      "step": 13130
    },
    {
      "entropy": 1.8982866704463959,
      "epoch": 0.8776162766428172,
      "grad_norm": 0.3334547281265259,
      "learning_rate": 2.3468885133430154e-06,
      "loss": 0.2088,
      "mean_token_accuracy": 0.9467113155871629,
      "num_tokens": 98131077.0,
      "step": 13140
    },
    {
      "entropy": 1.8906438417732716,
      "epoch": 0.878284173352591,
      "grad_norm": 0.31850579380989075,
      "learning_rate": 2.3228222201407016e-06,
      "loss": 0.2198,
      "mean_token_accuracy": 0.946795203909278,
      "num_tokens": 98202713.0,
      "step": 13150
    },
    {
      "entropy": 1.8986133173108102,
      "epoch": 0.8789520700623649,
      "grad_norm": 0.2782582938671112,
      "learning_rate": 2.2988794979033888e-06,
      "loss": 0.2228,
      "mean_token_accuracy": 0.9429177559912205,
      "num_tokens": 98278027.0,
      "step": 13160
    },
    {
      "entropy": 1.8995852150022983,
      "epoch": 0.8796199667721387,
      "grad_norm": 0.29466956853866577,
      "learning_rate": 2.2750604768111266e-06,
      "loss": 0.2209,
      "mean_token_accuracy": 0.9461796466261149,
      "num_tokens": 98353802.0,
      "step": 13170
    },
    {
      "entropy": 1.8842368714511395,
      "epoch": 0.8802878634819126,
      "grad_norm": 0.3136873245239258,
      "learning_rate": 2.2513652863714043e-06,
      "loss": 0.223,
      "mean_token_accuracy": 0.9453888233751059,
      "num_tokens": 98426818.0,
      "step": 13180
    },
    {
      "entropy": 1.9025919929146766,
      "epoch": 0.8809557601916863,
      "grad_norm": 0.28327301144599915,
      "learning_rate": 2.2277940554183955e-06,
      "loss": 0.2232,
      "mean_token_accuracy": 0.9449585396796465,
      "num_tokens": 98503440.0,
      "step": 13190
    },
    {
      "entropy": 1.9027289628982544,
      "epoch": 0.8816236569014602,
      "grad_norm": 0.28458476066589355,
      "learning_rate": 2.204346912112313e-06,
      "loss": 0.2128,
      "mean_token_accuracy": 0.9478524778038263,
      "num_tokens": 98576337.0,
      "step": 13200
    },
    {
      "entropy": 1.9297608852386474,
      "epoch": 0.882291553611234,
      "grad_norm": 0.3180808424949646,
      "learning_rate": 2.181023983938681e-06,
      "loss": 0.225,
      "mean_token_accuracy": 0.9438484117388726,
      "num_tokens": 98649622.0,
      "step": 13210
    },
    {
      "entropy": 1.9100798867642879,
      "epoch": 0.8829594503210079,
      "grad_norm": 0.28057873249053955,
      "learning_rate": 2.157825397707644e-06,
      "loss": 0.2185,
      "mean_token_accuracy": 0.9445180971175432,
      "num_tokens": 98727263.0,
      "step": 13220
    },
    {
      "entropy": 1.8764438286423684,
      "epoch": 0.8836273470307817,
      "grad_norm": 0.33722302317619324,
      "learning_rate": 2.1347512795532856e-06,
      "loss": 0.2113,
      "mean_token_accuracy": 0.9472705338150262,
      "num_tokens": 98803249.0,
      "step": 13230
    },
    {
      "entropy": 1.902427864074707,
      "epoch": 0.8842952437405556,
      "grad_norm": 0.3187564015388489,
      "learning_rate": 2.1118017549329277e-06,
      "loss": 0.2358,
      "mean_token_accuracy": 0.9436456378549337,
      "num_tokens": 98879488.0,
      "step": 13240
    },
    {
      "entropy": 1.8838684670627117,
      "epoch": 0.8849631404503293,
      "grad_norm": 0.2611239552497864,
      "learning_rate": 2.0889769486264703e-06,
      "loss": 0.22,
      "mean_token_accuracy": 0.9462853536009789,
      "num_tokens": 98954630.0,
      "step": 13250
    },
    {
      "entropy": 1.9105690382421017,
      "epoch": 0.8856310371601032,
      "grad_norm": 0.4400027096271515,
      "learning_rate": 2.0662769847357053e-06,
      "loss": 0.2317,
      "mean_token_accuracy": 0.9410594008862972,
      "num_tokens": 99027957.0,
      "step": 13260
    },
    {
      "entropy": 1.9107958316802978,
      "epoch": 0.886298933869877,
      "grad_norm": 0.32987266778945923,
      "learning_rate": 2.0437019866836244e-06,
      "loss": 0.2197,
      "mean_token_accuracy": 0.9442878343164921,
      "num_tokens": 99100160.0,
      "step": 13270
    },
    {
      "entropy": 1.891150862723589,
      "epoch": 0.8869668305796509,
      "grad_norm": 0.40508028864860535,
      "learning_rate": 2.0212520772137756e-06,
      "loss": 0.2149,
      "mean_token_accuracy": 0.9451200991868973,
      "num_tokens": 99173770.0,
      "step": 13280
    },
    {
      "entropy": 1.8941858768463136,
      "epoch": 0.8876347272894247,
      "grad_norm": 0.31875917315483093,
      "learning_rate": 1.9989273783895704e-06,
      "loss": 0.2163,
      "mean_token_accuracy": 0.9474922582507134,
      "num_tokens": 99249994.0,
      "step": 13290
    },
    {
      "entropy": 1.8973739825189113,
      "epoch": 0.8883026239991986,
      "grad_norm": 0.3059801161289215,
      "learning_rate": 1.976728011593642e-06,
      "loss": 0.2121,
      "mean_token_accuracy": 0.9487315006554127,
      "num_tokens": 99321787.0,
      "step": 13300
    },
    {
      "entropy": 1.905952476710081,
      "epoch": 0.8889705207089723,
      "grad_norm": 0.33482030034065247,
      "learning_rate": 1.9546540975271644e-06,
      "loss": 0.2203,
      "mean_token_accuracy": 0.9448102749884129,
      "num_tokens": 99394023.0,
      "step": 13310
    },
    {
      "entropy": 1.8998353369534016,
      "epoch": 0.8896384174187462,
      "grad_norm": 0.26776888966560364,
      "learning_rate": 1.9327057562092183e-06,
      "loss": 0.2092,
      "mean_token_accuracy": 0.9486879654228687,
      "num_tokens": 99467128.0,
      "step": 13320
    },
    {
      "entropy": 1.8985233709216118,
      "epoch": 0.89030631412852,
      "grad_norm": 0.3306865096092224,
      "learning_rate": 1.910883106976125e-06,
      "loss": 0.2202,
      "mean_token_accuracy": 0.945213158801198,
      "num_tokens": 99541382.0,
      "step": 13330
    },
    {
      "entropy": 1.891828079521656,
      "epoch": 0.8909742108382939,
      "grad_norm": 0.32416030764579773,
      "learning_rate": 1.8891862684807859e-06,
      "loss": 0.2108,
      "mean_token_accuracy": 0.9457236006855965,
      "num_tokens": 99616447.0,
      "step": 13340
    },
    {
      "entropy": 1.8976975105702878,
      "epoch": 0.8916421075480677,
      "grad_norm": 0.3189708888530731,
      "learning_rate": 1.8676153586920695e-06,
      "loss": 0.2236,
      "mean_token_accuracy": 0.9464722644537688,
      "num_tokens": 99690491.0,
      "step": 13350
    },
    {
      "entropy": 1.8865771420300006,
      "epoch": 0.8923100042578416,
      "grad_norm": 0.3577705919742584,
      "learning_rate": 1.8461704948941404e-06,
      "loss": 0.2407,
      "mean_token_accuracy": 0.9435342576354742,
      "num_tokens": 99763480.0,
      "step": 13360
    },
    {
      "entropy": 1.8979328125715256,
      "epoch": 0.8929779009676153,
      "grad_norm": 0.31675660610198975,
      "learning_rate": 1.824851793685838e-06,
      "loss": 0.2152,
      "mean_token_accuracy": 0.9487810526043177,
      "num_tokens": 99839534.0,
      "step": 13370
    },
    {
      "entropy": 1.906707002967596,
      "epoch": 0.8936457976773892,
      "grad_norm": 0.2757381200790405,
      "learning_rate": 1.8036593709800376e-06,
      "loss": 0.2114,
      "mean_token_accuracy": 0.9467173650860786,
      "num_tokens": 99915392.0,
      "step": 13380
    },
    {
      "entropy": 1.905162078142166,
      "epoch": 0.894313694387163,
      "grad_norm": 0.2738649547100067,
      "learning_rate": 1.7825933420030136e-06,
      "loss": 0.2232,
      "mean_token_accuracy": 0.947211854159832,
      "num_tokens": 99989599.0,
      "step": 13390
    },
    {
      "entropy": 1.95468819886446,
      "epoch": 0.8949815910969369,
      "grad_norm": 0.3508623540401459,
      "learning_rate": 1.7616538212938284e-06,
      "loss": 0.2273,
      "mean_token_accuracy": 0.9445957690477371,
      "num_tokens": 100063496.0,
      "step": 13400
    },
    {
      "entropy": 1.9230702228844165,
      "epoch": 0.8956494878067107,
      "grad_norm": 0.2760393023490906,
      "learning_rate": 1.7408409227036918e-06,
      "loss": 0.2305,
      "mean_token_accuracy": 0.9437668323516846,
      "num_tokens": 100139068.0,
      "step": 13410
    },
    {
      "entropy": 1.9089729115366936,
      "epoch": 0.8963173845164846,
      "grad_norm": 0.33417776226997375,
      "learning_rate": 1.7201547593953556e-06,
      "loss": 0.2148,
      "mean_token_accuracy": 0.9468063410371542,
      "num_tokens": 100212345.0,
      "step": 13420
    },
    {
      "entropy": 1.8938631422817707,
      "epoch": 0.8969852812262583,
      "grad_norm": 0.2712090313434601,
      "learning_rate": 1.6995954438425e-06,
      "loss": 0.2049,
      "mean_token_accuracy": 0.948897211253643,
      "num_tokens": 100288882.0,
      "step": 13430
    },
    {
      "entropy": 1.8983085237443447,
      "epoch": 0.8976531779360322,
      "grad_norm": 0.28794336318969727,
      "learning_rate": 1.679163087829097e-06,
      "loss": 0.2281,
      "mean_token_accuracy": 0.9439247619360686,
      "num_tokens": 100365523.0,
      "step": 13440
    },
    {
      "entropy": 1.8961265586316585,
      "epoch": 0.898321074645806,
      "grad_norm": 0.3383331000804901,
      "learning_rate": 1.6588578024488408e-06,
      "loss": 0.2145,
      "mean_token_accuracy": 0.9469839949160814,
      "num_tokens": 100438344.0,
      "step": 13450
    },
    {
      "entropy": 1.8918723858892919,
      "epoch": 0.8989889713555799,
      "grad_norm": 0.3040715754032135,
      "learning_rate": 1.6386796981045117e-06,
      "loss": 0.2038,
      "mean_token_accuracy": 0.9499476838856935,
      "num_tokens": 100516650.0,
      "step": 13460
    },
    {
      "entropy": 1.901743783801794,
      "epoch": 0.8996568680653537,
      "grad_norm": 0.3100825548171997,
      "learning_rate": 1.6186288845073929e-06,
      "loss": 0.2243,
      "mean_token_accuracy": 0.9449595987796784,
      "num_tokens": 100594588.0,
      "step": 13470
    },
    {
      "entropy": 1.8686980076134205,
      "epoch": 0.9003247647751276,
      "grad_norm": 0.3108823001384735,
      "learning_rate": 1.598705470676672e-06,
      "loss": 0.1911,
      "mean_token_accuracy": 0.9540467452257871,
      "num_tokens": 100669922.0,
      "step": 13480
    },
    {
      "entropy": 1.912983275949955,
      "epoch": 0.9009926614849013,
      "grad_norm": 0.34020450711250305,
      "learning_rate": 1.5789095649388359e-06,
      "loss": 0.2441,
      "mean_token_accuracy": 0.9437492664903402,
      "num_tokens": 100743601.0,
      "step": 13490
    },
    {
      "entropy": 1.8800438351929187,
      "epoch": 0.9016605581946752,
      "grad_norm": 0.3228403329849243,
      "learning_rate": 1.559241274927103e-06,
      "loss": 0.2115,
      "mean_token_accuracy": 0.9504777606576681,
      "num_tokens": 100819877.0,
      "step": 13500
    },
    {
      "entropy": 1.9302244670689106,
      "epoch": 0.902328454904449,
      "grad_norm": 0.2962304651737213,
      "learning_rate": 1.5397007075808213e-06,
      "loss": 0.2172,
      "mean_token_accuracy": 0.9446598574519157,
      "num_tokens": 100893443.0,
      "step": 13510
    },
    {
      "entropy": 1.9175043262541294,
      "epoch": 0.9029963516142229,
      "grad_norm": 0.2838495075702667,
      "learning_rate": 1.5202879691448898e-06,
      "loss": 0.2176,
      "mean_token_accuracy": 0.9475421570241451,
      "num_tokens": 100967830.0,
      "step": 13520
    },
    {
      "entropy": 1.9105577133595943,
      "epoch": 0.9036642483239967,
      "grad_norm": 0.3758520185947418,
      "learning_rate": 1.5010031651691909e-06,
      "loss": 0.2144,
      "mean_token_accuracy": 0.9449237577617169,
      "num_tokens": 101040690.0,
      "step": 13530
    },
    {
      "entropy": 1.9058876417577266,
      "epoch": 0.9043321450337706,
      "grad_norm": 0.2541896402835846,
      "learning_rate": 1.4818464005079978e-06,
      "loss": 0.2207,
      "mean_token_accuracy": 0.9438064724206925,
      "num_tokens": 101115522.0,
      "step": 13540
    },
    {
      "entropy": 1.9006869301199913,
      "epoch": 0.9050000417435443,
      "grad_norm": 0.25745898485183716,
      "learning_rate": 1.4628177793194219e-06,
      "loss": 0.2026,
      "mean_token_accuracy": 0.9499967951327563,
      "num_tokens": 101188806.0,
      "step": 13550
    },
    {
      "entropy": 1.924962904304266,
      "epoch": 0.9056679384533182,
      "grad_norm": 0.28642842173576355,
      "learning_rate": 1.4439174050648389e-06,
      "loss": 0.2299,
      "mean_token_accuracy": 0.9418151676654816,
      "num_tokens": 101263232.0,
      "step": 13560
    },
    {
      "entropy": 1.9254615865647793,
      "epoch": 0.906335835163092,
      "grad_norm": 0.3092668950557709,
      "learning_rate": 1.4251453805083297e-06,
      "loss": 0.2272,
      "mean_token_accuracy": 0.9459533967077732,
      "num_tokens": 101337802.0,
      "step": 13570
    },
    {
      "entropy": 1.9103708907961845,
      "epoch": 0.9070037318728659,
      "grad_norm": 0.2696140706539154,
      "learning_rate": 1.4065018077161197e-06,
      "loss": 0.2055,
      "mean_token_accuracy": 0.9483724005520344,
      "num_tokens": 101413467.0,
      "step": 13580
    },
    {
      "entropy": 1.897906582802534,
      "epoch": 0.9076716285826397,
      "grad_norm": 0.3275409936904907,
      "learning_rate": 1.3879867880560133e-06,
      "loss": 0.2032,
      "mean_token_accuracy": 0.9504713468253613,
      "num_tokens": 101487348.0,
      "step": 13590
    },
    {
      "entropy": 1.9108335733413697,
      "epoch": 0.9083395252924136,
      "grad_norm": 0.2650662362575531,
      "learning_rate": 1.3696004221968643e-06,
      "loss": 0.224,
      "mean_token_accuracy": 0.9450354319065809,
      "num_tokens": 101564194.0,
      "step": 13600
    },
    {
      "entropy": 1.901456455886364,
      "epoch": 0.9090074220021873,
      "grad_norm": 0.30371785163879395,
      "learning_rate": 1.3513428101080117e-06,
      "loss": 0.202,
      "mean_token_accuracy": 0.9506184365600348,
      "num_tokens": 101642504.0,
      "step": 13610
    },
    {
      "entropy": 1.9014749571681022,
      "epoch": 0.9096753187119612,
      "grad_norm": 0.2971813678741455,
      "learning_rate": 1.333214051058747e-06,
      "loss": 0.2208,
      "mean_token_accuracy": 0.9452633641660213,
      "num_tokens": 101716721.0,
      "step": 13620
    },
    {
      "entropy": 1.9070034220814704,
      "epoch": 0.910343215421735,
      "grad_norm": 0.2824757993221283,
      "learning_rate": 1.3152142436177634e-06,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9453223545104266,
      "num_tokens": 101788718.0,
      "step": 13630
    },
    {
      "entropy": 1.8911651529371738,
      "epoch": 0.9110111121315089,
      "grad_norm": 0.30897578597068787,
      "learning_rate": 1.2973434856526215e-06,
      "loss": 0.2075,
      "mean_token_accuracy": 0.9477046266198158,
      "num_tokens": 101860138.0,
      "step": 13640
    },
    {
      "entropy": 1.9213181003928184,
      "epoch": 0.9116790088412827,
      "grad_norm": 0.3536868095397949,
      "learning_rate": 1.279601874329233e-06,
      "loss": 0.2215,
      "mean_token_accuracy": 0.9447006203234196,
      "num_tokens": 101934680.0,
      "step": 13650
    },
    {
      "entropy": 1.9152114614844322,
      "epoch": 0.9123469055510566,
      "grad_norm": 0.35017839074134827,
      "learning_rate": 1.2619895061113113e-06,
      "loss": 0.214,
      "mean_token_accuracy": 0.9472366467118263,
      "num_tokens": 102010265.0,
      "step": 13660
    },
    {
      "entropy": 1.8986652418971062,
      "epoch": 0.9130148022608303,
      "grad_norm": 0.28530582785606384,
      "learning_rate": 1.2445064767598626e-06,
      "loss": 0.2168,
      "mean_token_accuracy": 0.9429144073277712,
      "num_tokens": 102084591.0,
      "step": 13670
    },
    {
      "entropy": 1.902928152680397,
      "epoch": 0.9136826989706042,
      "grad_norm": 0.2830169200897217,
      "learning_rate": 1.227152881332655e-06,
      "loss": 0.2022,
      "mean_token_accuracy": 0.9487363889813423,
      "num_tokens": 102159440.0,
      "step": 13680
    },
    {
      "entropy": 1.9146923258900643,
      "epoch": 0.914350595680378,
      "grad_norm": 0.3174077570438385,
      "learning_rate": 1.2099288141837036e-06,
      "loss": 0.2161,
      "mean_token_accuracy": 0.9469379331916571,
      "num_tokens": 102236383.0,
      "step": 13690
    },
    {
      "entropy": 1.886171568185091,
      "epoch": 0.9150184923901519,
      "grad_norm": 0.34835055470466614,
      "learning_rate": 1.1928343689627603e-06,
      "loss": 0.2272,
      "mean_token_accuracy": 0.9440027885138988,
      "num_tokens": 102311299.0,
      "step": 13700
    },
    {
      "entropy": 1.930020059645176,
      "epoch": 0.9156863890999257,
      "grad_norm": 0.33307892084121704,
      "learning_rate": 1.175869638614807e-06,
      "loss": 0.2214,
      "mean_token_accuracy": 0.9430679339915514,
      "num_tokens": 102386799.0,
      "step": 13710
    },
    {
      "entropy": 1.8897304765880107,
      "epoch": 0.9163542858096996,
      "grad_norm": 0.30179741978645325,
      "learning_rate": 1.1590347153795486e-06,
      "loss": 0.1948,
      "mean_token_accuracy": 0.9499950230121612,
      "num_tokens": 102461963.0,
      "step": 13720
    },
    {
      "entropy": 1.9021281115710735,
      "epoch": 0.9170221825194733,
      "grad_norm": 0.27086541056632996,
      "learning_rate": 1.1423296907909062e-06,
      "loss": 0.2387,
      "mean_token_accuracy": 0.9403638765215874,
      "num_tokens": 102536274.0,
      "step": 13730
    },
    {
      "entropy": 1.8921031907200814,
      "epoch": 0.9176900792292472,
      "grad_norm": 0.26246461272239685,
      "learning_rate": 1.125754655676519e-06,
      "loss": 0.2106,
      "mean_token_accuracy": 0.9463224306702613,
      "num_tokens": 102613924.0,
      "step": 13740
    },
    {
      "entropy": 1.9295793034136295,
      "epoch": 0.918357975939021,
      "grad_norm": 0.34940090775489807,
      "learning_rate": 1.1093097001572611e-06,
      "loss": 0.2419,
      "mean_token_accuracy": 0.9409036424010992,
      "num_tokens": 102687478.0,
      "step": 13750
    },
    {
      "entropy": 1.8949863009154797,
      "epoch": 0.9190258726487949,
      "grad_norm": 0.3371618986129761,
      "learning_rate": 1.0929949136467444e-06,
      "loss": 0.2333,
      "mean_token_accuracy": 0.9415029205381871,
      "num_tokens": 102760956.0,
      "step": 13760
    },
    {
      "entropy": 1.8929891839623452,
      "epoch": 0.9196937693585687,
      "grad_norm": 0.33426401019096375,
      "learning_rate": 1.0768103848508323e-06,
      "loss": 0.2093,
      "mean_token_accuracy": 0.949324493855238,
      "num_tokens": 102837543.0,
      "step": 13770
    },
    {
      "entropy": 1.909164971113205,
      "epoch": 0.9203616660683426,
      "grad_norm": 0.3457319438457489,
      "learning_rate": 1.0607562017671615e-06,
      "loss": 0.2293,
      "mean_token_accuracy": 0.9444771248847246,
      "num_tokens": 102908119.0,
      "step": 13780
    },
    {
      "entropy": 1.8979900188744068,
      "epoch": 0.9210295627781163,
      "grad_norm": 0.3295208513736725,
      "learning_rate": 1.0448324516846438e-06,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9469987910240889,
      "num_tokens": 102986304.0,
      "step": 13790
    },
    {
      "entropy": 1.8948651857674121,
      "epoch": 0.9216974594878902,
      "grad_norm": 0.2926713824272156,
      "learning_rate": 1.0290392211830323e-06,
      "loss": 0.2135,
      "mean_token_accuracy": 0.9480871453881263,
      "num_tokens": 103061776.0,
      "step": 13800
    },
    {
      "entropy": 1.8892649672925472,
      "epoch": 0.922365356197664,
      "grad_norm": 0.2630287706851959,
      "learning_rate": 1.013376596132405e-06,
      "loss": 0.2105,
      "mean_token_accuracy": 0.9479119196534157,
      "num_tokens": 103135807.0,
      "step": 13810
    },
    {
      "entropy": 1.90837921500206,
      "epoch": 0.9230332529074379,
      "grad_norm": 0.34752416610717773,
      "learning_rate": 9.97844661692735e-07,
      "loss": 0.213,
      "mean_token_accuracy": 0.9481565687805414,
      "num_tokens": 103208733.0,
      "step": 13820
    },
    {
      "entropy": 1.8790317371487617,
      "epoch": 0.9237011496172117,
      "grad_norm": 0.33678585290908813,
      "learning_rate": 9.82443502313403e-07,
      "loss": 0.2152,
      "mean_token_accuracy": 0.9483610242605209,
      "num_tokens": 103283759.0,
      "step": 13830
    },
    {
      "entropy": 1.90202339887619,
      "epoch": 0.9243690463269856,
      "grad_norm": 0.3287935256958008,
      "learning_rate": 9.671732017327448e-07,
      "loss": 0.2268,
      "mean_token_accuracy": 0.9440580252557993,
      "num_tokens": 103357120.0,
      "step": 13840
    },
    {
      "entropy": 1.914595778286457,
      "epoch": 0.9250369430367593,
      "grad_norm": 0.39622658491134644,
      "learning_rate": 9.520338429776044e-07,
      "loss": 0.2248,
      "mean_token_accuracy": 0.9448769468814134,
      "num_tokens": 103429041.0,
      "step": 13850
    },
    {
      "entropy": 1.9054407931864261,
      "epoch": 0.9257048397465332,
      "grad_norm": 0.3200603127479553,
      "learning_rate": 9.37025508362873e-07,
      "loss": 0.2116,
      "mean_token_accuracy": 0.9475846383720636,
      "num_tokens": 103501936.0,
      "step": 13860
    },
    {
      "entropy": 1.9195763342082501,
      "epoch": 0.926372736456307,
      "grad_norm": 0.31297188997268677,
      "learning_rate": 9.221482794910454e-07,
      "loss": 0.2088,
      "mean_token_accuracy": 0.9481914173811674,
      "num_tokens": 103577288.0,
      "step": 13870
    },
    {
      "entropy": 1.9043946728110313,
      "epoch": 0.9270406331660809,
      "grad_norm": 0.26709577441215515,
      "learning_rate": 9.074022372517754e-07,
      "loss": 0.2106,
      "mean_token_accuracy": 0.9488805778324604,
      "num_tokens": 103651484.0,
      "step": 13880
    },
    {
      "entropy": 1.8853257417678833,
      "epoch": 0.9277085298758547,
      "grad_norm": 0.32161444425582886,
      "learning_rate": 8.927874618214313e-07,
      "loss": 0.2109,
      "mean_token_accuracy": 0.9493244666606188,
      "num_tokens": 103728072.0,
      "step": 13890
    },
    {
      "entropy": 1.8881514206528665,
      "epoch": 0.9283764265856286,
      "grad_norm": 0.28794118762016296,
      "learning_rate": 8.783040326626754e-07,
      "loss": 0.2131,
      "mean_token_accuracy": 0.9478063277900219,
      "num_tokens": 103802010.0,
      "step": 13900
    },
    {
      "entropy": 1.891056627780199,
      "epoch": 0.9290443232954023,
      "grad_norm": 0.24881400167942047,
      "learning_rate": 8.639520285240077e-07,
      "loss": 0.2127,
      "mean_token_accuracy": 0.946513257175684,
      "num_tokens": 103877530.0,
      "step": 13910
    },
    {
      "entropy": 1.9036525025963784,
      "epoch": 0.9297122200051762,
      "grad_norm": 0.3208293914794922,
      "learning_rate": 8.497315274393595e-07,
      "loss": 0.2039,
      "mean_token_accuracy": 0.948467843979597,
      "num_tokens": 103952735.0,
      "step": 13920
    },
    {
      "entropy": 1.8826133012771606,
      "epoch": 0.93038011671495,
      "grad_norm": 0.3096645772457123,
      "learning_rate": 8.356426067276634e-07,
      "loss": 0.2145,
      "mean_token_accuracy": 0.9482813503593206,
      "num_tokens": 104030107.0,
      "step": 13930
    },
    {
      "entropy": 1.9399089455604552,
      "epoch": 0.9310480134247239,
      "grad_norm": 0.35808226466178894,
      "learning_rate": 8.216853429924286e-07,
      "loss": 0.2277,
      "mean_token_accuracy": 0.9447576142847538,
      "num_tokens": 104103384.0,
      "step": 13940
    },
    {
      "entropy": 1.8939123027026654,
      "epoch": 0.9317159101344977,
      "grad_norm": 0.313656747341156,
      "learning_rate": 8.078598121213162e-07,
      "loss": 0.2244,
      "mean_token_accuracy": 0.9436252113431692,
      "num_tokens": 104178860.0,
      "step": 13950
    },
    {
      "entropy": 1.9000036634504796,
      "epoch": 0.9323838068442716,
      "grad_norm": 0.3107992112636566,
      "learning_rate": 7.941660892857477e-07,
      "loss": 0.2307,
      "mean_token_accuracy": 0.9432311546057462,
      "num_tokens": 104251339.0,
      "step": 13960
    },
    {
      "entropy": 1.882425457239151,
      "epoch": 0.9330517035540453,
      "grad_norm": 0.3070521354675293,
      "learning_rate": 7.806042489404854e-07,
      "loss": 0.209,
      "mean_token_accuracy": 0.9472176652401686,
      "num_tokens": 104326846.0,
      "step": 13970
    },
    {
      "entropy": 1.9240444339811802,
      "epoch": 0.9337196002638192,
      "grad_norm": 0.35235562920570374,
      "learning_rate": 7.671743648232226e-07,
      "loss": 0.2192,
      "mean_token_accuracy": 0.9461739677935839,
      "num_tokens": 104401009.0,
      "step": 13980
    },
    {
      "entropy": 1.9051270239055156,
      "epoch": 0.934387496973593,
      "grad_norm": 0.30857589840888977,
      "learning_rate": 7.538765099541853e-07,
      "loss": 0.2146,
      "mean_token_accuracy": 0.9468018088489771,
      "num_tokens": 104476807.0,
      "step": 13990
    },
    {
      "entropy": 1.9361830152571202,
      "epoch": 0.9350553936833669,
      "grad_norm": 0.4242134690284729,
      "learning_rate": 7.407107566357463e-07,
      "loss": 0.2191,
      "mean_token_accuracy": 0.9460461411625147,
      "num_tokens": 104554010.0,
      "step": 14000
    },
    {
      "epoch": 0.9350553936833669,
      "eval_entropy": 1.8998861152410507,
      "eval_loss": 0.2339961975812912,
      "eval_mean_token_accuracy": 0.9466467488050461,
      "eval_num_tokens": 104554010.0,
      "eval_runtime": 553.4051,
      "eval_samples_per_second": 36.14,
      "eval_steps_per_second": 9.035,
      "step": 14000
    },
    {
      "entropy": 1.9053326681256295,
      "epoch": 0.9357232903931407,
      "grad_norm": 0.2986902892589569,
      "learning_rate": 7.276771764520137e-07,
      "loss": 0.223,
      "mean_token_accuracy": 0.9454509973526001,
      "num_tokens": 104629176.0,
      "step": 14010
    },
    {
      "entropy": 1.8850824356079101,
      "epoch": 0.9363911871029146,
      "grad_norm": 0.3177836835384369,
      "learning_rate": 7.14775840268458e-07,
      "loss": 0.2233,
      "mean_token_accuracy": 0.9440376691520214,
      "num_tokens": 104707245.0,
      "step": 14020
    },
    {
      "entropy": 1.9137109152972698,
      "epoch": 0.9370590838126883,
      "grad_norm": 0.29121869802474976,
      "learning_rate": 7.02006818231518e-07,
      "loss": 0.1946,
      "mean_token_accuracy": 0.9512248717248439,
      "num_tokens": 104784253.0,
      "step": 14030
    },
    {
      "entropy": 1.8909002617001534,
      "epoch": 0.9377269805224622,
      "grad_norm": 0.26387354731559753,
      "learning_rate": 6.893701797682138e-07,
      "loss": 0.2179,
      "mean_token_accuracy": 0.9464529417455196,
      "num_tokens": 104860652.0,
      "step": 14040
    },
    {
      "entropy": 1.909472694247961,
      "epoch": 0.938394877232236,
      "grad_norm": 0.2949415445327759,
      "learning_rate": 6.768659935857912e-07,
      "loss": 0.2243,
      "mean_token_accuracy": 0.9460593670606613,
      "num_tokens": 104934840.0,
      "step": 14050
    },
    {
      "entropy": 1.8926683761179448,
      "epoch": 0.9390627739420099,
      "grad_norm": 0.2625250220298767,
      "learning_rate": 6.644943276713187e-07,
      "loss": 0.2164,
      "mean_token_accuracy": 0.9448520917445421,
      "num_tokens": 105009642.0,
      "step": 14060
    },
    {
      "entropy": 1.906169418990612,
      "epoch": 0.9397306706517837,
      "grad_norm": 0.3173508942127228,
      "learning_rate": 6.522552492913448e-07,
      "loss": 0.228,
      "mean_token_accuracy": 0.9431961376219988,
      "num_tokens": 105082751.0,
      "step": 14070
    },
    {
      "entropy": 1.8969881288707255,
      "epoch": 0.9403985673615576,
      "grad_norm": 0.3377363085746765,
      "learning_rate": 6.401488249915154e-07,
      "loss": 0.2143,
      "mean_token_accuracy": 0.9458045039325953,
      "num_tokens": 105156833.0,
      "step": 14080
    },
    {
      "entropy": 1.8988959573209285,
      "epoch": 0.9410664640713313,
      "grad_norm": 0.3353278934955597,
      "learning_rate": 6.281751205962112e-07,
      "loss": 0.2285,
      "mean_token_accuracy": 0.9452743317931891,
      "num_tokens": 105231605.0,
      "step": 14090
    },
    {
      "entropy": 1.9253309786319732,
      "epoch": 0.9417343607811052,
      "grad_norm": 0.30926719307899475,
      "learning_rate": 6.163342012082096e-07,
      "loss": 0.2242,
      "mean_token_accuracy": 0.9434264447540045,
      "num_tokens": 105303215.0,
      "step": 14100
    },
    {
      "entropy": 1.9120075926184654,
      "epoch": 0.942402257490879,
      "grad_norm": 0.35029345750808716,
      "learning_rate": 6.046261312083021e-07,
      "loss": 0.2216,
      "mean_token_accuracy": 0.9450719900429249,
      "num_tokens": 105379463.0,
      "step": 14110
    },
    {
      "entropy": 1.929402296990156,
      "epoch": 0.9430701542006529,
      "grad_norm": 0.27098631858825684,
      "learning_rate": 5.930509742549684e-07,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9445500519126654,
      "num_tokens": 105454003.0,
      "step": 14120
    },
    {
      "entropy": 1.9110081180930139,
      "epoch": 0.9437380509104267,
      "grad_norm": 0.2694525420665741,
      "learning_rate": 5.81608793284015e-07,
      "loss": 0.2121,
      "mean_token_accuracy": 0.9466637037694454,
      "num_tokens": 105525907.0,
      "step": 14130
    },
    {
      "entropy": 1.9003278836607933,
      "epoch": 0.9444059476202006,
      "grad_norm": 0.30434873700141907,
      "learning_rate": 5.70299650508244e-07,
      "loss": 0.2227,
      "mean_token_accuracy": 0.9444978535175323,
      "num_tokens": 105602125.0,
      "step": 14140
    },
    {
      "entropy": 1.8926567010581494,
      "epoch": 0.9450738443299743,
      "grad_norm": 0.2583397626876831,
      "learning_rate": 5.591236074171032e-07,
      "loss": 0.2111,
      "mean_token_accuracy": 0.9485705241560936,
      "num_tokens": 105675849.0,
      "step": 14150
    },
    {
      "entropy": 1.8904980294406415,
      "epoch": 0.9457417410397482,
      "grad_norm": 0.3499576151371002,
      "learning_rate": 5.480807247763626e-07,
      "loss": 0.2315,
      "mean_token_accuracy": 0.9434334632009268,
      "num_tokens": 105749472.0,
      "step": 14160
    },
    {
      "entropy": 1.9075636386871337,
      "epoch": 0.946409637749522,
      "grad_norm": 0.31966209411621094,
      "learning_rate": 5.371710626277761e-07,
      "loss": 0.227,
      "mean_token_accuracy": 0.9443599779158831,
      "num_tokens": 105821787.0,
      "step": 14170
    },
    {
      "entropy": 1.9219057574868201,
      "epoch": 0.9470775344592959,
      "grad_norm": 0.631982684135437,
      "learning_rate": 5.263946802887687e-07,
      "loss": 0.228,
      "mean_token_accuracy": 0.9410583045333624,
      "num_tokens": 105898727.0,
      "step": 14180
    },
    {
      "entropy": 1.8956761822104453,
      "epoch": 0.9477454311690697,
      "grad_norm": 0.3022257089614868,
      "learning_rate": 5.157516363520846e-07,
      "loss": 0.2155,
      "mean_token_accuracy": 0.9445299483835697,
      "num_tokens": 105973432.0,
      "step": 14190
    },
    {
      "entropy": 1.9058410614728927,
      "epoch": 0.9484133278788436,
      "grad_norm": 0.29628491401672363,
      "learning_rate": 5.052419886855046e-07,
      "loss": 0.2136,
      "mean_token_accuracy": 0.9480999674648046,
      "num_tokens": 106048958.0,
      "step": 14200
    },
    {
      "entropy": 1.8964909233152867,
      "epoch": 0.9490812245886173,
      "grad_norm": 0.2666878402233124,
      "learning_rate": 4.948657944315083e-07,
      "loss": 0.2191,
      "mean_token_accuracy": 0.9467222951352596,
      "num_tokens": 106124757.0,
      "step": 14210
    },
    {
      "entropy": 1.8713796056807042,
      "epoch": 0.9497491212983912,
      "grad_norm": 0.3048289120197296,
      "learning_rate": 4.846231100069693e-07,
      "loss": 0.2316,
      "mean_token_accuracy": 0.9437967505306005,
      "num_tokens": 106202782.0,
      "step": 14220
    },
    {
      "entropy": 1.902660033106804,
      "epoch": 0.950417018008165,
      "grad_norm": 0.2985217869281769,
      "learning_rate": 4.7451399110285357e-07,
      "loss": 0.2268,
      "mean_token_accuracy": 0.9446544133126735,
      "num_tokens": 106275810.0,
      "step": 14230
    },
    {
      "entropy": 1.9055275343358518,
      "epoch": 0.9510849147179389,
      "grad_norm": 0.30049318075180054,
      "learning_rate": 4.645384926839029e-07,
      "loss": 0.2358,
      "mean_token_accuracy": 0.9416691426187753,
      "num_tokens": 106349309.0,
      "step": 14240
    },
    {
      "entropy": 1.918624348193407,
      "epoch": 0.9517528114277127,
      "grad_norm": 0.3278043270111084,
      "learning_rate": 4.546966689883507e-07,
      "loss": 0.2229,
      "mean_token_accuracy": 0.9437994439154863,
      "num_tokens": 106423947.0,
      "step": 14250
    },
    {
      "entropy": 1.9064877651631833,
      "epoch": 0.9524207081374866,
      "grad_norm": 0.3928927481174469,
      "learning_rate": 4.4498857352762723e-07,
      "loss": 0.216,
      "mean_token_accuracy": 0.9470358230173588,
      "num_tokens": 106498959.0,
      "step": 14260
    },
    {
      "entropy": 1.9171678505837917,
      "epoch": 0.9530886048472603,
      "grad_norm": 0.29642802476882935,
      "learning_rate": 4.3541425908605023e-07,
      "loss": 0.2236,
      "mean_token_accuracy": 0.9467587042599916,
      "num_tokens": 106571940.0,
      "step": 14270
    },
    {
      "entropy": 1.9286782294511795,
      "epoch": 0.9537565015570342,
      "grad_norm": 0.30293843150138855,
      "learning_rate": 4.259737777205614e-07,
      "loss": 0.2352,
      "mean_token_accuracy": 0.9423229865729809,
      "num_tokens": 106644657.0,
      "step": 14280
    },
    {
      "entropy": 1.8920840539038182,
      "epoch": 0.954424398266808,
      "grad_norm": 0.2970978021621704,
      "learning_rate": 4.1666718076042444e-07,
      "loss": 0.2023,
      "mean_token_accuracy": 0.950397364795208,
      "num_tokens": 106719816.0,
      "step": 14290
    },
    {
      "entropy": 1.9169700197875499,
      "epoch": 0.9550922949765819,
      "grad_norm": 0.32482442259788513,
      "learning_rate": 4.074945188069621e-07,
      "loss": 0.2229,
      "mean_token_accuracy": 0.943707600608468,
      "num_tokens": 106793048.0,
      "step": 14300
    },
    {
      "entropy": 1.9219482079148293,
      "epoch": 0.9557601916863557,
      "grad_norm": 0.2949027121067047,
      "learning_rate": 3.98455841733265e-07,
      "loss": 0.2221,
      "mean_token_accuracy": 0.9472261693328619,
      "num_tokens": 106866192.0,
      "step": 14310
    },
    {
      "entropy": 1.8989798329770564,
      "epoch": 0.9564280883961296,
      "grad_norm": 0.361895889043808,
      "learning_rate": 3.895511986839372e-07,
      "loss": 0.2213,
      "mean_token_accuracy": 0.9457820616662502,
      "num_tokens": 106940744.0,
      "step": 14320
    },
    {
      "entropy": 1.9364443197846413,
      "epoch": 0.9570959851059033,
      "grad_norm": 0.2804163992404938,
      "learning_rate": 3.807806380748162e-07,
      "loss": 0.2262,
      "mean_token_accuracy": 0.9443983431905508,
      "num_tokens": 107016144.0,
      "step": 14330
    },
    {
      "entropy": 1.9024449817836284,
      "epoch": 0.9577638818156772,
      "grad_norm": 0.3311755955219269,
      "learning_rate": 3.7214420759270687e-07,
      "loss": 0.2369,
      "mean_token_accuracy": 0.9418485555797815,
      "num_tokens": 107088045.0,
      "step": 14340
    },
    {
      "entropy": 1.8688890233635902,
      "epoch": 0.958431778525451,
      "grad_norm": 0.2731431722640991,
      "learning_rate": 3.63641954195138e-07,
      "loss": 0.2007,
      "mean_token_accuracy": 0.9524892080575228,
      "num_tokens": 107165448.0,
      "step": 14350
    },
    {
      "entropy": 1.91527017056942,
      "epoch": 0.9590996752352249,
      "grad_norm": 0.32071930170059204,
      "learning_rate": 3.5527392411010187e-07,
      "loss": 0.2024,
      "mean_token_accuracy": 0.9514481287449599,
      "num_tokens": 107243571.0,
      "step": 14360
    },
    {
      "entropy": 1.9192841432988643,
      "epoch": 0.9597675719449987,
      "grad_norm": 0.307857871055603,
      "learning_rate": 3.4704016283578527e-07,
      "loss": 0.2237,
      "mean_token_accuracy": 0.9444409117102623,
      "num_tokens": 107316773.0,
      "step": 14370
    },
    {
      "entropy": 1.8934790387749671,
      "epoch": 0.9604354686547726,
      "grad_norm": 0.30186915397644043,
      "learning_rate": 3.389407151403454e-07,
      "loss": 0.2104,
      "mean_token_accuracy": 0.9473525546491146,
      "num_tokens": 107391509.0,
      "step": 14380
    },
    {
      "entropy": 1.9183316327631474,
      "epoch": 0.9611033653645463,
      "grad_norm": 0.2908487915992737,
      "learning_rate": 3.3097562506165783e-07,
      "loss": 0.2122,
      "mean_token_accuracy": 0.9467424344271421,
      "num_tokens": 107466915.0,
      "step": 14390
    },
    {
      "entropy": 1.9257058031857013,
      "epoch": 0.9617712620743202,
      "grad_norm": 0.3677487373352051,
      "learning_rate": 3.2314493590706977e-07,
      "loss": 0.2384,
      "mean_token_accuracy": 0.9423711311072112,
      "num_tokens": 107539796.0,
      "step": 14400
    },
    {
      "entropy": 1.9166744612157345,
      "epoch": 0.962439158784094,
      "grad_norm": 0.2927500605583191,
      "learning_rate": 3.154486902531786e-07,
      "loss": 0.2183,
      "mean_token_accuracy": 0.9451176710426807,
      "num_tokens": 107616849.0,
      "step": 14410
    },
    {
      "entropy": 1.9063184700906277,
      "epoch": 0.9631070554938679,
      "grad_norm": 0.24640081822872162,
      "learning_rate": 3.078869299455853e-07,
      "loss": 0.2217,
      "mean_token_accuracy": 0.9478651534765958,
      "num_tokens": 107690799.0,
      "step": 14420
    },
    {
      "entropy": 1.9056262366473675,
      "epoch": 0.9637749522036417,
      "grad_norm": 0.313030481338501,
      "learning_rate": 3.00459696098684e-07,
      "loss": 0.2047,
      "mean_token_accuracy": 0.9487129893153906,
      "num_tokens": 107765207.0,
      "step": 14430
    },
    {
      "entropy": 1.9169410578906536,
      "epoch": 0.9644428489134156,
      "grad_norm": 0.32929080724716187,
      "learning_rate": 2.9316702909541794e-07,
      "loss": 0.2164,
      "mean_token_accuracy": 0.9461659785360098,
      "num_tokens": 107839380.0,
      "step": 14440
    },
    {
      "entropy": 1.8900426775217056,
      "epoch": 0.9651107456231893,
      "grad_norm": 0.24310599267482758,
      "learning_rate": 2.860089685870805e-07,
      "loss": 0.2194,
      "mean_token_accuracy": 0.945015249028802,
      "num_tokens": 107915972.0,
      "step": 14450
    },
    {
      "entropy": 1.9423066228628159,
      "epoch": 0.9657786423329632,
      "grad_norm": 0.32423171401023865,
      "learning_rate": 2.789855534930823e-07,
      "loss": 0.2207,
      "mean_token_accuracy": 0.9456261999905109,
      "num_tokens": 107989110.0,
      "step": 14460
    },
    {
      "entropy": 1.902545154839754,
      "epoch": 0.966446539042737,
      "grad_norm": 0.28054577112197876,
      "learning_rate": 2.720968220007514e-07,
      "loss": 0.2117,
      "mean_token_accuracy": 0.9457728024572134,
      "num_tokens": 108062731.0,
      "step": 14470
    },
    {
      "entropy": 1.899439264833927,
      "epoch": 0.9671144357525109,
      "grad_norm": 0.276172012090683,
      "learning_rate": 2.6534281156512614e-07,
      "loss": 0.2155,
      "mean_token_accuracy": 0.9465055733919143,
      "num_tokens": 108136353.0,
      "step": 14480
    },
    {
      "entropy": 1.8697900146245956,
      "epoch": 0.9677823324622847,
      "grad_norm": 0.2538411021232605,
      "learning_rate": 2.5872355890873333e-07,
      "loss": 0.22,
      "mean_token_accuracy": 0.9464835111051798,
      "num_tokens": 108212839.0,
      "step": 14490
    },
    {
      "entropy": 1.9229343213140964,
      "epoch": 0.9684502291720586,
      "grad_norm": 0.30318912863731384,
      "learning_rate": 2.52239100021419e-07,
      "loss": 0.2289,
      "mean_token_accuracy": 0.9440717861056328,
      "num_tokens": 108289585.0,
      "step": 14500
    },
    {
      "entropy": 1.8845447473227979,
      "epoch": 0.9691181258818323,
      "grad_norm": 0.25452524423599243,
      "learning_rate": 2.4588947016011874e-07,
      "loss": 0.1919,
      "mean_token_accuracy": 0.9521220989525319,
      "num_tokens": 108365302.0,
      "step": 14510
    },
    {
      "entropy": 1.92604451328516,
      "epoch": 0.9697860225916062,
      "grad_norm": 0.4214472770690918,
      "learning_rate": 2.3967470384869995e-07,
      "loss": 0.2447,
      "mean_token_accuracy": 0.9431238524615765,
      "num_tokens": 108437724.0,
      "step": 14520
    },
    {
      "entropy": 1.9152753211557865,
      "epoch": 0.97045391930138,
      "grad_norm": 0.35824206471443176,
      "learning_rate": 2.335948348777456e-07,
      "loss": 0.2196,
      "mean_token_accuracy": 0.9456470660865307,
      "num_tokens": 108511518.0,
      "step": 14530
    },
    {
      "entropy": 1.9206460803747176,
      "epoch": 0.9711218160111539,
      "grad_norm": 0.2893131971359253,
      "learning_rate": 2.2764989630438515e-07,
      "loss": 0.2185,
      "mean_token_accuracy": 0.9451597556471825,
      "num_tokens": 108587601.0,
      "step": 14540
    },
    {
      "entropy": 1.895983377099037,
      "epoch": 0.9717897127209277,
      "grad_norm": 0.32508042454719543,
      "learning_rate": 2.218399204521146e-07,
      "loss": 0.2178,
      "mean_token_accuracy": 0.9476647909730673,
      "num_tokens": 108660487.0,
      "step": 14550
    },
    {
      "entropy": 1.8673418171703815,
      "epoch": 0.9724576094307016,
      "grad_norm": 0.2992874085903168,
      "learning_rate": 2.1616493891061388e-07,
      "loss": 0.2047,
      "mean_token_accuracy": 0.9492248620837926,
      "num_tokens": 108737543.0,
      "step": 14560
    },
    {
      "entropy": 1.912288798391819,
      "epoch": 0.9731255061404753,
      "grad_norm": 0.30885881185531616,
      "learning_rate": 2.1062498253558577e-07,
      "loss": 0.2112,
      "mean_token_accuracy": 0.9461349155753851,
      "num_tokens": 108813388.0,
      "step": 14570
    },
    {
      "entropy": 1.92212390974164,
      "epoch": 0.9737934028502492,
      "grad_norm": 0.3620794713497162,
      "learning_rate": 2.0522008144857357e-07,
      "loss": 0.2407,
      "mean_token_accuracy": 0.9415260531008244,
      "num_tokens": 108885346.0,
      "step": 14580
    },
    {
      "entropy": 1.8799037128686904,
      "epoch": 0.974461299560023,
      "grad_norm": 0.41817936301231384,
      "learning_rate": 1.999502650368082e-07,
      "loss": 0.2079,
      "mean_token_accuracy": 0.9467885941267014,
      "num_tokens": 108961868.0,
      "step": 14590
    },
    {
      "entropy": 1.90477839037776,
      "epoch": 0.9751291962697969,
      "grad_norm": 0.2815674841403961,
      "learning_rate": 1.9481556195305088e-07,
      "loss": 0.2311,
      "mean_token_accuracy": 0.9456219393759966,
      "num_tokens": 109037259.0,
      "step": 14600
    },
    {
      "entropy": 1.8997058264911175,
      "epoch": 0.9757970929795707,
      "grad_norm": 0.24212032556533813,
      "learning_rate": 1.8981600011542368e-07,
      "loss": 0.2258,
      "mean_token_accuracy": 0.9447796110063791,
      "num_tokens": 109113235.0,
      "step": 14610
    },
    {
      "entropy": 1.903115016222,
      "epoch": 0.9764649896893446,
      "grad_norm": 0.2980491816997528,
      "learning_rate": 1.8495160670727132e-07,
      "loss": 0.2145,
      "mean_token_accuracy": 0.9470945868641139,
      "num_tokens": 109188055.0,
      "step": 14620
    },
    {
      "entropy": 1.9022199727594853,
      "epoch": 0.9771328863991183,
      "grad_norm": 0.3174128830432892,
      "learning_rate": 1.8022240817700867e-07,
      "loss": 0.2084,
      "mean_token_accuracy": 0.9467992827296257,
      "num_tokens": 109262767.0,
      "step": 14630
    },
    {
      "entropy": 1.8780999280512334,
      "epoch": 0.9778007831088922,
      "grad_norm": 0.2915668785572052,
      "learning_rate": 1.7562843023797403e-07,
      "loss": 0.2177,
      "mean_token_accuracy": 0.9480854034423828,
      "num_tokens": 109337773.0,
      "step": 14640
    },
    {
      "entropy": 1.9047947071492672,
      "epoch": 0.978468679818666,
      "grad_norm": 0.3345016837120056,
      "learning_rate": 1.7116969786829057e-07,
      "loss": 0.2154,
      "mean_token_accuracy": 0.9464879151433706,
      "num_tokens": 109413221.0,
      "step": 14650
    },
    {
      "entropy": 1.8854721903800964,
      "epoch": 0.9791365765284399,
      "grad_norm": 0.3091772794723511,
      "learning_rate": 1.6684623531073336e-07,
      "loss": 0.2181,
      "mean_token_accuracy": 0.9489821385592222,
      "num_tokens": 109487825.0,
      "step": 14660
    },
    {
      "entropy": 1.9225395686924458,
      "epoch": 0.9798044732382137,
      "grad_norm": 0.31313666701316833,
      "learning_rate": 1.626580660725965e-07,
      "loss": 0.2274,
      "mean_token_accuracy": 0.9445375587791205,
      "num_tokens": 109562805.0,
      "step": 14670
    },
    {
      "entropy": 1.9297599755227566,
      "epoch": 0.9804723699479876,
      "grad_norm": 0.29253339767456055,
      "learning_rate": 1.5860521292556837e-07,
      "loss": 0.2207,
      "mean_token_accuracy": 0.9450436372309923,
      "num_tokens": 109640239.0,
      "step": 14680
    },
    {
      "entropy": 1.9096783593297004,
      "epoch": 0.9811402666577613,
      "grad_norm": 0.33872610330581665,
      "learning_rate": 1.546876979055959e-07,
      "loss": 0.2149,
      "mean_token_accuracy": 0.9444828130304813,
      "num_tokens": 109715591.0,
      "step": 14690
    },
    {
      "entropy": 1.8925483390688895,
      "epoch": 0.9818081633675352,
      "grad_norm": 0.2907775938510895,
      "learning_rate": 1.509055423127739e-07,
      "loss": 0.2126,
      "mean_token_accuracy": 0.9496159307658673,
      "num_tokens": 109788423.0,
      "step": 14700
    },
    {
      "entropy": 1.8837422780692576,
      "epoch": 0.982476060077309,
      "grad_norm": 0.323563814163208,
      "learning_rate": 1.472587667112369e-07,
      "loss": 0.2109,
      "mean_token_accuracy": 0.9462040577083826,
      "num_tokens": 109863505.0,
      "step": 14710
    },
    {
      "entropy": 1.9297125674784184,
      "epoch": 0.9831439567870829,
      "grad_norm": 0.3640925884246826,
      "learning_rate": 1.437473909290317e-07,
      "loss": 0.2345,
      "mean_token_accuracy": 0.9424944326281548,
      "num_tokens": 109937996.0,
      "step": 14720
    },
    {
      "entropy": 1.8911085098981857,
      "epoch": 0.9838118534968567,
      "grad_norm": 0.27827513217926025,
      "learning_rate": 1.4037143405802059e-07,
      "loss": 0.199,
      "mean_token_accuracy": 0.9495569165796042,
      "num_tokens": 110013526.0,
      "step": 14730
    },
    {
      "entropy": 1.9321283496916295,
      "epoch": 0.9844797502066306,
      "grad_norm": 0.35959592461586,
      "learning_rate": 1.3713091445376483e-07,
      "loss": 0.2221,
      "mean_token_accuracy": 0.9471862558275461,
      "num_tokens": 110086477.0,
      "step": 14740
    },
    {
      "entropy": 1.8906516335904597,
      "epoch": 0.9851476469164043,
      "grad_norm": 0.3196421265602112,
      "learning_rate": 1.3402584973544442e-07,
      "loss": 0.2286,
      "mean_token_accuracy": 0.9432332340627909,
      "num_tokens": 110159790.0,
      "step": 14750
    },
    {
      "entropy": 1.9261316314339638,
      "epoch": 0.9858155436261782,
      "grad_norm": 0.3286134600639343,
      "learning_rate": 1.310562567857473e-07,
      "loss": 0.2232,
      "mean_token_accuracy": 0.9430730476975441,
      "num_tokens": 110234090.0,
      "step": 14760
    },
    {
      "entropy": 1.9303351260721684,
      "epoch": 0.986483440335952,
      "grad_norm": 0.2944607436656952,
      "learning_rate": 1.2822215175077788e-07,
      "loss": 0.2311,
      "mean_token_accuracy": 0.943125244602561,
      "num_tokens": 110307598.0,
      "step": 14770
    },
    {
      "entropy": 1.8846277087926864,
      "epoch": 0.9871513370457259,
      "grad_norm": 0.2993800938129425,
      "learning_rate": 1.2552355003998233e-07,
      "loss": 0.2171,
      "mean_token_accuracy": 0.9478538773953915,
      "num_tokens": 110383656.0,
      "step": 14780
    },
    {
      "entropy": 1.892453780770302,
      "epoch": 0.9878192337554997,
      "grad_norm": 0.2868650555610657,
      "learning_rate": 1.2296046632604316e-07,
      "loss": 0.2021,
      "mean_token_accuracy": 0.9499805081635714,
      "num_tokens": 110460658.0,
      "step": 14790
    },
    {
      "entropy": 1.8919722713530063,
      "epoch": 0.9884871304652736,
      "grad_norm": 0.29760387539863586,
      "learning_rate": 1.2053291454482413e-07,
      "loss": 0.2192,
      "mean_token_accuracy": 0.9454273480921984,
      "num_tokens": 110540082.0,
      "step": 14800
    },
    {
      "entropy": 1.869931484013796,
      "epoch": 0.9891550271750473,
      "grad_norm": 0.34532567858695984,
      "learning_rate": 1.1824090789527574e-07,
      "loss": 0.2063,
      "mean_token_accuracy": 0.9508180834352971,
      "num_tokens": 110616426.0,
      "step": 14810
    },
    {
      "entropy": 1.910967905819416,
      "epoch": 0.9898229238848212,
      "grad_norm": 0.30471616983413696,
      "learning_rate": 1.1608445883936888e-07,
      "loss": 0.2331,
      "mean_token_accuracy": 0.9438163720071315,
      "num_tokens": 110690791.0,
      "step": 14820
    },
    {
      "entropy": 1.9329442329704762,
      "epoch": 0.990490820594595,
      "grad_norm": 0.38030868768692017,
      "learning_rate": 1.1406357910202836e-07,
      "loss": 0.2279,
      "mean_token_accuracy": 0.9443112336099148,
      "num_tokens": 110763196.0,
      "step": 14830
    },
    {
      "entropy": 1.8962192341685296,
      "epoch": 0.9911587173043689,
      "grad_norm": 0.2726113200187683,
      "learning_rate": 1.1217827967107762e-07,
      "loss": 0.2244,
      "mean_token_accuracy": 0.9445026602596045,
      "num_tokens": 110837125.0,
      "step": 14840
    },
    {
      "entropy": 1.8790459103882313,
      "epoch": 0.9918266140141427,
      "grad_norm": 0.3209843337535858,
      "learning_rate": 1.1042857079715813e-07,
      "loss": 0.2182,
      "mean_token_accuracy": 0.9493644427508116,
      "num_tokens": 110912006.0,
      "step": 14850
    },
    {
      "entropy": 1.9133743420243263,
      "epoch": 0.9924945107239166,
      "grad_norm": 0.3805257976055145,
      "learning_rate": 1.0881446199369085e-07,
      "loss": 0.2132,
      "mean_token_accuracy": 0.948195218667388,
      "num_tokens": 110990216.0,
      "step": 14860
    },
    {
      "entropy": 1.9052971661090852,
      "epoch": 0.9931624074336903,
      "grad_norm": 0.3305679261684418,
      "learning_rate": 1.0733596203682072e-07,
      "loss": 0.2186,
      "mean_token_accuracy": 0.945969745516777,
      "num_tokens": 111063905.0,
      "step": 14870
    },
    {
      "entropy": 1.9224503308534622,
      "epoch": 0.9938303041434642,
      "grad_norm": 0.32435184717178345,
      "learning_rate": 1.0599307896536684e-07,
      "loss": 0.2376,
      "mean_token_accuracy": 0.9441445644944906,
      "num_tokens": 111136174.0,
      "step": 14880
    },
    {
      "entropy": 1.9025490529835225,
      "epoch": 0.994498200853238,
      "grad_norm": 0.26026588678359985,
      "learning_rate": 1.0478582008078089e-07,
      "loss": 0.2249,
      "mean_token_accuracy": 0.9466857232153416,
      "num_tokens": 111209499.0,
      "step": 14890
    },
    {
      "entropy": 1.9036887191236018,
      "epoch": 0.9951660975630119,
      "grad_norm": 0.29366225004196167,
      "learning_rate": 1.0371419194710279e-07,
      "loss": 0.22,
      "mean_token_accuracy": 0.9459269996732473,
      "num_tokens": 111284322.0,
      "step": 14900
    },
    {
      "entropy": 1.915351528674364,
      "epoch": 0.9958339942727857,
      "grad_norm": 0.24741236865520477,
      "learning_rate": 1.0277820039093034e-07,
      "loss": 0.2082,
      "mean_token_accuracy": 0.949078006669879,
      "num_tokens": 111361067.0,
      "step": 14910
    },
    {
      "entropy": 1.901783287525177,
      "epoch": 0.9965018909825596,
      "grad_norm": 0.3467392921447754,
      "learning_rate": 1.0197785050138585e-07,
      "loss": 0.2108,
      "mean_token_accuracy": 0.946212413161993,
      "num_tokens": 111435370.0,
      "step": 14920
    },
    {
      "entropy": 1.9154698580503464,
      "epoch": 0.9971697876923333,
      "grad_norm": 0.2944880723953247,
      "learning_rate": 1.0131314663008852e-07,
      "loss": 0.2073,
      "mean_token_accuracy": 0.9457383468747139,
      "num_tokens": 111509558.0,
      "step": 14930
    },
    {
      "entropy": 1.9072790823876857,
      "epoch": 0.9978376844021072,
      "grad_norm": 0.28679195046424866,
      "learning_rate": 1.0078409239112666e-07,
      "loss": 0.2134,
      "mean_token_accuracy": 0.9477059442549944,
      "num_tokens": 111584565.0,
      "step": 14940
    },
    {
      "entropy": 1.936252660304308,
      "epoch": 0.998505581111881,
      "grad_norm": 0.33688589930534363,
      "learning_rate": 1.0039069066104954e-07,
      "loss": 0.2143,
      "mean_token_accuracy": 0.9477420080453157,
      "num_tokens": 111657978.0,
      "step": 14950
    },
    {
      "entropy": 1.888141191750765,
      "epoch": 0.9991734778216549,
      "grad_norm": 0.28729745745658875,
      "learning_rate": 1.0013294357883669e-07,
      "loss": 0.2108,
      "mean_token_accuracy": 0.9503798212856054,
      "num_tokens": 111731287.0,
      "step": 14960
    },
    {
      "entropy": 1.8898568801581859,
      "epoch": 0.9998413745314287,
      "grad_norm": 0.34546494483947754,
      "learning_rate": 1.0001085254589811e-07,
      "loss": 0.2032,
      "mean_token_accuracy": 0.9500156525522471,
      "num_tokens": 111806500.0,
      "step": 14970
    }
  ],
  "logging_steps": 10,
  "max_steps": 14973,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.117928302156605e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}