{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002,
      "grad_norm": 3.529070725879476,
      "learning_rate": 9.999984208641271e-06,
      "loss": 0.4748,
      "num_input_tokens_seen": 396688,
      "step": 5,
      "train_runtime": 75.1268,
      "train_tokens_per_second": 5280.244
    },
    {
      "epoch": 0.004,
      "grad_norm": 1.1725318777053566,
      "learning_rate": 9.999920056417385e-06,
      "loss": 0.3853,
      "num_input_tokens_seen": 780032,
      "step": 10,
      "train_runtime": 138.2073,
      "train_tokens_per_second": 5643.926
    },
    {
      "epoch": 0.006,
      "grad_norm": 1.050069609189703,
      "learning_rate": 9.999806557001092e-06,
      "loss": 0.3416,
      "num_input_tokens_seen": 1239888,
      "step": 15,
      "train_runtime": 205.0125,
      "train_tokens_per_second": 6047.864
    },
    {
      "epoch": 0.008,
      "grad_norm": 1.0659218188772097,
      "learning_rate": 9.999643711512586e-06,
      "loss": 0.3701,
      "num_input_tokens_seen": 1603472,
      "step": 20,
      "train_runtime": 265.8186,
      "train_tokens_per_second": 6032.204
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.820657214467373,
      "learning_rate": 9.999431521559081e-06,
      "loss": 0.3371,
      "num_input_tokens_seen": 1963184,
      "step": 25,
      "train_runtime": 325.7553,
      "train_tokens_per_second": 6026.56
    },
    {
      "epoch": 0.012,
      "grad_norm": 1.1800961461261863,
      "learning_rate": 9.999169989234815e-06,
      "loss": 0.3436,
      "num_input_tokens_seen": 2326256,
      "step": 30,
      "train_runtime": 387.5961,
      "train_tokens_per_second": 6001.753
    },
    {
      "epoch": 0.014,
      "grad_norm": 1.231264145354163,
      "learning_rate": 9.998859117121e-06,
      "loss": 0.3498,
      "num_input_tokens_seen": 2713648,
      "step": 35,
      "train_runtime": 444.7317,
      "train_tokens_per_second": 6101.765
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.9000852106140734,
      "learning_rate": 9.99849890828582e-06,
      "loss": 0.3205,
      "num_input_tokens_seen": 3106864,
      "step": 40,
      "train_runtime": 510.8204,
      "train_tokens_per_second": 6082.106
    },
    {
      "epoch": 0.018,
      "grad_norm": 0.8813007763453813,
      "learning_rate": 9.998089366284392e-06,
      "loss": 0.3272,
      "num_input_tokens_seen": 3513728,
      "step": 45,
      "train_runtime": 573.2378,
      "train_tokens_per_second": 6129.617
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.842498340169096,
      "learning_rate": 9.997630495158728e-06,
      "loss": 0.3279,
      "num_input_tokens_seen": 3950720,
      "step": 50,
      "train_runtime": 651.4821,
      "train_tokens_per_second": 6064.203
    },
    {
      "epoch": 0.022,
      "grad_norm": 0.7406379773223285,
      "learning_rate": 9.9971222994377e-06,
      "loss": 0.304,
      "num_input_tokens_seen": 4341968,
      "step": 55,
      "train_runtime": 723.0612,
      "train_tokens_per_second": 6004.98
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.9246450042837577,
      "learning_rate": 9.996564784137e-06,
      "loss": 0.2791,
      "num_input_tokens_seen": 4716832,
      "step": 60,
      "train_runtime": 789.7507,
      "train_tokens_per_second": 5972.558
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.6813867833298154,
      "learning_rate": 9.995957954759073e-06,
      "loss": 0.3242,
      "num_input_tokens_seen": 5190560,
      "step": 65,
      "train_runtime": 868.1415,
      "train_tokens_per_second": 5978.933
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.958297020853992,
      "learning_rate": 9.995301817293084e-06,
      "loss": 0.3044,
      "num_input_tokens_seen": 5563120,
      "step": 70,
      "train_runtime": 931.2634,
      "train_tokens_per_second": 5973.734
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.8856568117076815,
      "learning_rate": 9.99459637821484e-06,
      "loss": 0.3096,
      "num_input_tokens_seen": 5970768,
      "step": 75,
      "train_runtime": 998.4455,
      "train_tokens_per_second": 5980.064
    },
    {
      "epoch": 0.032,
      "grad_norm": 1.1073162962080754,
      "learning_rate": 9.993841644486747e-06,
      "loss": 0.3166,
      "num_input_tokens_seen": 6259376,
      "step": 80,
      "train_runtime": 1039.407,
      "train_tokens_per_second": 6022.065
    },
    {
      "epoch": 0.034,
      "grad_norm": 0.9231847695523737,
      "learning_rate": 9.993037623557716e-06,
      "loss": 0.3017,
      "num_input_tokens_seen": 6654464,
      "step": 85,
      "train_runtime": 1110.0658,
      "train_tokens_per_second": 5994.657
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.8821941113621354,
      "learning_rate": 9.992184323363112e-06,
      "loss": 0.2731,
      "num_input_tokens_seen": 7023760,
      "step": 90,
      "train_runtime": 1183.6772,
      "train_tokens_per_second": 5933.848
    },
    {
      "epoch": 0.038,
      "grad_norm": 1.0797634407658638,
      "learning_rate": 9.991281752324664e-06,
      "loss": 0.3023,
      "num_input_tokens_seen": 7436112,
      "step": 95,
      "train_runtime": 1257.9906,
      "train_tokens_per_second": 5911.103
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.103353756296292,
      "learning_rate": 9.990329919350382e-06,
      "loss": 0.2834,
      "num_input_tokens_seen": 7791888,
      "step": 100,
      "train_runtime": 1322.6092,
      "train_tokens_per_second": 5891.301
    },
    {
      "epoch": 0.042,
      "grad_norm": 0.8278175812097224,
      "learning_rate": 9.989328833834472e-06,
      "loss": 0.3008,
      "num_input_tokens_seen": 8185936,
      "step": 105,
      "train_runtime": 1380.3591,
      "train_tokens_per_second": 5930.294
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.7830140246513998,
      "learning_rate": 9.988278505657247e-06,
      "loss": 0.2933,
      "num_input_tokens_seen": 8598768,
      "step": 110,
      "train_runtime": 1447.4638,
      "train_tokens_per_second": 5940.576
    },
    {
      "epoch": 0.046,
      "grad_norm": 0.6319254381860551,
      "learning_rate": 9.987178945185019e-06,
      "loss": 0.2986,
      "num_input_tokens_seen": 9017264,
      "step": 115,
      "train_runtime": 1515.8302,
      "train_tokens_per_second": 5948.73
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.7754207741241279,
      "learning_rate": 9.986030163270011e-06,
      "loss": 0.3024,
      "num_input_tokens_seen": 9449872,
      "step": 120,
      "train_runtime": 1593.8688,
      "train_tokens_per_second": 5928.889
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6941365154998115,
      "learning_rate": 9.98483217125023e-06,
      "loss": 0.2553,
      "num_input_tokens_seen": 9874192,
      "step": 125,
      "train_runtime": 1663.0592,
      "train_tokens_per_second": 5937.366
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.8267427353263731,
      "learning_rate": 9.98358498094938e-06,
      "loss": 0.3012,
      "num_input_tokens_seen": 10278272,
      "step": 130,
      "train_runtime": 1727.2891,
      "train_tokens_per_second": 5950.522
    },
    {
      "epoch": 0.054,
      "grad_norm": 2.1153588465265982,
      "learning_rate": 9.982288604676719e-06,
      "loss": 0.2758,
      "num_input_tokens_seen": 10695136,
      "step": 135,
      "train_runtime": 1793.2202,
      "train_tokens_per_second": 5964.207
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.8604952494397087,
      "learning_rate": 9.980943055226964e-06,
      "loss": 0.3056,
      "num_input_tokens_seen": 11039680,
      "step": 140,
      "train_runtime": 1854.7104,
      "train_tokens_per_second": 5952.239
    },
    {
      "epoch": 0.058,
      "grad_norm": 0.8382901602209842,
      "learning_rate": 9.979548345880142e-06,
      "loss": 0.2943,
      "num_input_tokens_seen": 11454128,
      "step": 145,
      "train_runtime": 1926.181,
      "train_tokens_per_second": 5946.548
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8875123598888999,
      "learning_rate": 9.978104490401468e-06,
      "loss": 0.2913,
      "num_input_tokens_seen": 11794496,
      "step": 150,
      "train_runtime": 1978.6218,
      "train_tokens_per_second": 5960.965
    },
    {
      "epoch": 0.062,
      "grad_norm": 1.0539384040695126,
      "learning_rate": 9.976611503041218e-06,
      "loss": 0.3153,
      "num_input_tokens_seen": 12169264,
      "step": 155,
      "train_runtime": 2051.3054,
      "train_tokens_per_second": 5932.449
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.8524180493974752,
      "learning_rate": 9.975069398534574e-06,
      "loss": 0.2993,
      "num_input_tokens_seen": 12515168,
      "step": 160,
      "train_runtime": 2113.3806,
      "train_tokens_per_second": 5921.871
    },
    {
      "epoch": 0.066,
      "grad_norm": 1.0205477764931508,
      "learning_rate": 9.97347819210148e-06,
      "loss": 0.2946,
      "num_input_tokens_seen": 12905488,
      "step": 165,
      "train_runtime": 2174.4797,
      "train_tokens_per_second": 5934.978
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.7838766165442029,
      "learning_rate": 9.971837899446505e-06,
      "loss": 0.3075,
      "num_input_tokens_seen": 13358672,
      "step": 170,
      "train_runtime": 2252.5741,
      "train_tokens_per_second": 5930.403
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8892411786342892,
      "learning_rate": 9.970148536758678e-06,
      "loss": 0.2952,
      "num_input_tokens_seen": 13779472,
      "step": 175,
      "train_runtime": 2315.3569,
      "train_tokens_per_second": 5951.338
    },
    {
      "epoch": 0.072,
      "grad_norm": 1.0296532392116329,
      "learning_rate": 9.968410120711321e-06,
      "loss": 0.2709,
      "num_input_tokens_seen": 14093728,
      "step": 180,
      "train_runtime": 2371.6187,
      "train_tokens_per_second": 5942.662
    },
    {
      "epoch": 0.074,
      "grad_norm": 0.8880922770210048,
      "learning_rate": 9.966622668461899e-06,
      "loss": 0.2928,
      "num_input_tokens_seen": 14595840,
      "step": 185,
      "train_runtime": 2444.1035,
      "train_tokens_per_second": 5971.858
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.9979494314336124,
      "learning_rate": 9.964786197651848e-06,
      "loss": 0.2898,
      "num_input_tokens_seen": 14919280,
      "step": 190,
      "train_runtime": 2497.4922,
      "train_tokens_per_second": 5973.704
    },
    {
      "epoch": 0.078,
      "grad_norm": 0.9769457891685653,
      "learning_rate": 9.96290072640639e-06,
      "loss": 0.2992,
      "num_input_tokens_seen": 15268256,
      "step": 195,
      "train_runtime": 2567.8553,
      "train_tokens_per_second": 5945.918
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.9296142727078585,
      "learning_rate": 9.96096627333437e-06,
      "loss": 0.3001,
      "num_input_tokens_seen": 15678432,
      "step": 200,
      "train_runtime": 2635.5512,
      "train_tokens_per_second": 5948.825
    },
    {
      "epoch": 0.082,
      "grad_norm": 1.2434185369871753,
      "learning_rate": 9.958982857528053e-06,
      "loss": 0.2888,
      "num_input_tokens_seen": 16161600,
      "step": 205,
      "train_runtime": 2714.3312,
      "train_tokens_per_second": 5954.174
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.7574454494180245,
      "learning_rate": 9.956950498562954e-06,
      "loss": 0.2696,
      "num_input_tokens_seen": 16572832,
      "step": 210,
      "train_runtime": 2789.5683,
      "train_tokens_per_second": 5941.002
    },
    {
      "epoch": 0.086,
      "grad_norm": 1.0339850849304288,
      "learning_rate": 9.954869216497636e-06,
      "loss": 0.298,
      "num_input_tokens_seen": 16966064,
      "step": 215,
      "train_runtime": 2858.0332,
      "train_tokens_per_second": 5936.272
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.9072760950057258,
      "learning_rate": 9.952739031873513e-06,
      "loss": 0.2919,
      "num_input_tokens_seen": 17326464,
      "step": 220,
      "train_runtime": 2923.5635,
      "train_tokens_per_second": 5926.488
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.9033068608878783,
      "learning_rate": 9.950559965714647e-06,
      "loss": 0.29,
      "num_input_tokens_seen": 17798944,
      "step": 225,
      "train_runtime": 2998.1702,
      "train_tokens_per_second": 5936.602
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.7265860170186728,
      "learning_rate": 9.948332039527541e-06,
      "loss": 0.281,
      "num_input_tokens_seen": 18203216,
      "step": 230,
      "train_runtime": 3064.0023,
      "train_tokens_per_second": 5940.993
    },
    {
      "epoch": 0.094,
      "grad_norm": 0.7719043892731912,
      "learning_rate": 9.946055275300929e-06,
      "loss": 0.2857,
      "num_input_tokens_seen": 18595600,
      "step": 235,
      "train_runtime": 3122.8974,
      "train_tokens_per_second": 5954.598
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.684701998859407,
      "learning_rate": 9.943729695505552e-06,
      "loss": 0.276,
      "num_input_tokens_seen": 19005936,
      "step": 240,
      "train_runtime": 3188.4373,
      "train_tokens_per_second": 5960.894
    },
    {
      "epoch": 0.098,
      "grad_norm": 0.946898379784884,
      "learning_rate": 9.941355323093944e-06,
      "loss": 0.297,
      "num_input_tokens_seen": 19412496,
      "step": 245,
      "train_runtime": 3255.2016,
      "train_tokens_per_second": 5963.531
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.9553375335195206,
      "learning_rate": 9.938932181500206e-06,
      "loss": 0.2649,
      "num_input_tokens_seen": 19830960,
      "step": 250,
      "train_runtime": 3333.374,
      "train_tokens_per_second": 5949.215
    },
    {
      "epoch": 0.102,
      "grad_norm": 0.8135687801680254,
      "learning_rate": 9.93646029463976e-06,
      "loss": 0.252,
      "num_input_tokens_seen": 20210288,
      "step": 255,
      "train_runtime": 3401.5027,
      "train_tokens_per_second": 5941.576
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.7875655466453482,
      "learning_rate": 9.933939686909137e-06,
      "loss": 0.2986,
      "num_input_tokens_seen": 20651728,
      "step": 260,
      "train_runtime": 3475.6609,
      "train_tokens_per_second": 5941.813
    },
    {
      "epoch": 0.106,
      "grad_norm": 1.270712900867129,
      "learning_rate": 9.931370383185717e-06,
      "loss": 0.2679,
      "num_input_tokens_seen": 21084048,
      "step": 265,
      "train_runtime": 3549.0279,
      "train_tokens_per_second": 5940.795
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.9393845827359342,
      "learning_rate": 9.92875240882749e-06,
      "loss": 0.2861,
      "num_input_tokens_seen": 21420560,
      "step": 270,
      "train_runtime": 3601.2535,
      "train_tokens_per_second": 5948.084
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7444755169661749,
      "learning_rate": 9.926085789672806e-06,
      "loss": 0.2649,
      "num_input_tokens_seen": 21794608,
      "step": 275,
      "train_runtime": 3667.1267,
      "train_tokens_per_second": 5943.238
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.7145025641256342,
      "learning_rate": 9.923370552040117e-06,
      "loss": 0.3103,
      "num_input_tokens_seen": 22200176,
      "step": 280,
      "train_runtime": 3735.5676,
      "train_tokens_per_second": 5942.919
    },
    {
      "epoch": 0.114,
      "grad_norm": 0.8512241574919769,
      "learning_rate": 9.920606722727726e-06,
      "loss": 0.2771,
      "num_input_tokens_seen": 22588416,
      "step": 285,
      "train_runtime": 3792.0445,
      "train_tokens_per_second": 5956.791
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.6266092029054396,
      "learning_rate": 9.917794329013511e-06,
      "loss": 0.2872,
      "num_input_tokens_seen": 22996016,
      "step": 290,
      "train_runtime": 3857.7998,
      "train_tokens_per_second": 5960.915
    },
    {
      "epoch": 0.118,
      "grad_norm": 0.751809560103108,
      "learning_rate": 9.914933398654663e-06,
      "loss": 0.2847,
      "num_input_tokens_seen": 23433456,
      "step": 295,
      "train_runtime": 3927.9056,
      "train_tokens_per_second": 5965.891
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6755889847224258,
      "learning_rate": 9.912023959887408e-06,
      "loss": 0.2537,
      "num_input_tokens_seen": 23781600,
      "step": 300,
      "train_runtime": 3982.1935,
      "train_tokens_per_second": 5971.985
    },
    {
      "epoch": 0.122,
      "grad_norm": 0.943323280572443,
      "learning_rate": 9.909066041426733e-06,
      "loss": 0.2843,
      "num_input_tokens_seen": 24225024,
      "step": 305,
      "train_runtime": 4055.2554,
      "train_tokens_per_second": 5973.736
    },
    {
      "epoch": 0.124,
      "grad_norm": 1.2175580370359331,
      "learning_rate": 9.9060596724661e-06,
      "loss": 0.2864,
      "num_input_tokens_seen": 24592480,
      "step": 310,
      "train_runtime": 4115.2286,
      "train_tokens_per_second": 5975.969
    },
    {
      "epoch": 0.126,
      "grad_norm": 1.0147370785297811,
      "learning_rate": 9.903004882677157e-06,
      "loss": 0.2892,
      "num_input_tokens_seen": 24957024,
      "step": 315,
      "train_runtime": 4168.4655,
      "train_tokens_per_second": 5987.101
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.7344408173379581,
      "learning_rate": 9.899901702209445e-06,
      "loss": 0.2789,
      "num_input_tokens_seen": 25374832,
      "step": 320,
      "train_runtime": 4244.2432,
      "train_tokens_per_second": 5978.647
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.3103019223544174,
      "learning_rate": 9.8967501616901e-06,
      "loss": 0.2766,
      "num_input_tokens_seen": 25810848,
      "step": 325,
      "train_runtime": 4317.4147,
      "train_tokens_per_second": 5978.311
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.8685597289161924,
      "learning_rate": 9.89355029222356e-06,
      "loss": 0.293,
      "num_input_tokens_seen": 26237024,
      "step": 330,
      "train_runtime": 4380.5436,
      "train_tokens_per_second": 5989.445
    },
    {
      "epoch": 0.134,
      "grad_norm": 0.8404098804337911,
      "learning_rate": 9.89030212539124e-06,
      "loss": 0.286,
      "num_input_tokens_seen": 26629216,
      "step": 335,
      "train_runtime": 4437.1469,
      "train_tokens_per_second": 6001.428
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.7386985305075902,
      "learning_rate": 9.88700569325124e-06,
      "loss": 0.3048,
      "num_input_tokens_seen": 27042592,
      "step": 340,
      "train_runtime": 4503.3805,
      "train_tokens_per_second": 6004.954
    },
    {
      "epoch": 0.138,
      "grad_norm": 0.8859553305112097,
      "learning_rate": 9.883661028338009e-06,
      "loss": 0.2555,
      "num_input_tokens_seen": 27433280,
      "step": 345,
      "train_runtime": 4577.0851,
      "train_tokens_per_second": 5993.614
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7480812603367697,
      "learning_rate": 9.880268163662043e-06,
      "loss": 0.2678,
      "num_input_tokens_seen": 27796240,
      "step": 350,
      "train_runtime": 4640.6632,
      "train_tokens_per_second": 5989.713
    },
    {
      "epoch": 0.142,
      "grad_norm": 1.1794465477248353,
      "learning_rate": 9.876827132709545e-06,
      "loss": 0.2818,
      "num_input_tokens_seen": 28242816,
      "step": 355,
      "train_runtime": 4716.0159,
      "train_tokens_per_second": 5988.702
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.8608826793810502,
      "learning_rate": 9.873337969442102e-06,
      "loss": 0.2624,
      "num_input_tokens_seen": 28623824,
      "step": 360,
      "train_runtime": 4784.382,
      "train_tokens_per_second": 5982.763
    },
    {
      "epoch": 0.146,
      "grad_norm": 0.842112418190514,
      "learning_rate": 9.869800708296347e-06,
      "loss": 0.2888,
      "num_input_tokens_seen": 29059536,
      "step": 365,
      "train_runtime": 4860.2713,
      "train_tokens_per_second": 5978.995
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.7582805292047462,
      "learning_rate": 9.86621538418362e-06,
      "loss": 0.2907,
      "num_input_tokens_seen": 29474832,
      "step": 370,
      "train_runtime": 4921.9495,
      "train_tokens_per_second": 5988.447
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7906066031029222,
      "learning_rate": 9.862582032489621e-06,
      "loss": 0.2706,
      "num_input_tokens_seen": 29872512,
      "step": 375,
      "train_runtime": 4981.8012,
      "train_tokens_per_second": 5996.328
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.7636592190389889,
      "learning_rate": 9.858900689074065e-06,
      "loss": 0.2743,
      "num_input_tokens_seen": 30343872,
      "step": 380,
      "train_runtime": 5056.2219,
      "train_tokens_per_second": 6001.294
    },
    {
      "epoch": 0.154,
      "grad_norm": 0.7503467243251319,
      "learning_rate": 9.855171390270325e-06,
      "loss": 0.2688,
      "num_input_tokens_seen": 30848768,
      "step": 385,
      "train_runtime": 5131.3151,
      "train_tokens_per_second": 6011.864
    },
    {
      "epoch": 0.156,
      "grad_norm": 2.8570072621011606,
      "learning_rate": 9.851394172885075e-06,
      "loss": 0.2894,
      "num_input_tokens_seen": 31282352,
      "step": 390,
      "train_runtime": 5203.7543,
      "train_tokens_per_second": 6011.497
    },
    {
      "epoch": 0.158,
      "grad_norm": 0.8255350410433486,
      "learning_rate": 9.847569074197927e-06,
      "loss": 0.2886,
      "num_input_tokens_seen": 31685040,
      "step": 395,
      "train_runtime": 5273.3386,
      "train_tokens_per_second": 6008.535
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7055444954194724,
      "learning_rate": 9.843696131961058e-06,
      "loss": 0.273,
      "num_input_tokens_seen": 32051808,
      "step": 400,
      "train_runtime": 5349.3981,
      "train_tokens_per_second": 5991.666
    },
    {
      "epoch": 0.162,
      "grad_norm": 0.7738430122073711,
      "learning_rate": 9.839775384398846e-06,
      "loss": 0.2698,
      "num_input_tokens_seen": 32439248,
      "step": 405,
      "train_runtime": 5424.8322,
      "train_tokens_per_second": 5979.77
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.8597473971332746,
      "learning_rate": 9.835806870207487e-06,
      "loss": 0.2574,
      "num_input_tokens_seen": 32860528,
      "step": 410,
      "train_runtime": 5488.1995,
      "train_tokens_per_second": 5987.488
    },
    {
      "epoch": 0.166,
      "grad_norm": 0.9548490042158907,
      "learning_rate": 9.831790628554613e-06,
      "loss": 0.2709,
      "num_input_tokens_seen": 33238192,
      "step": 415,
      "train_runtime": 5558.4847,
      "train_tokens_per_second": 5979.722
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.6176099320311793,
      "learning_rate": 9.827726699078907e-06,
      "loss": 0.2656,
      "num_input_tokens_seen": 33690160,
      "step": 420,
      "train_runtime": 5625.9505,
      "train_tokens_per_second": 5988.35
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.8127343411925148,
      "learning_rate": 9.823615121889716e-06,
      "loss": 0.29,
      "num_input_tokens_seen": 34073536,
      "step": 425,
      "train_runtime": 5692.2387,
      "train_tokens_per_second": 5985.964
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.7691896317723499,
      "learning_rate": 9.819455937566642e-06,
      "loss": 0.2823,
      "num_input_tokens_seen": 34424192,
      "step": 430,
      "train_runtime": 5754.8576,
      "train_tokens_per_second": 5981.763
    },
    {
      "epoch": 0.174,
      "grad_norm": 1.0082130619949137,
      "learning_rate": 9.815249187159158e-06,
      "loss": 0.2496,
      "num_input_tokens_seen": 34809680,
      "step": 435,
      "train_runtime": 5832.4698,
      "train_tokens_per_second": 5968.257
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.7539701356352378,
      "learning_rate": 9.81099491218619e-06,
      "loss": 0.2627,
      "num_input_tokens_seen": 35219808,
      "step": 440,
      "train_runtime": 5895.6451,
      "train_tokens_per_second": 5973.868
    },
    {
      "epoch": 0.178,
      "grad_norm": 0.8592295102624296,
      "learning_rate": 9.806693154635719e-06,
      "loss": 0.2774,
      "num_input_tokens_seen": 35565248,
      "step": 445,
      "train_runtime": 5941.5778,
      "train_tokens_per_second": 5985.826
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6582951154075225,
      "learning_rate": 9.802343956964348e-06,
      "loss": 0.2588,
      "num_input_tokens_seen": 36022288,
      "step": 450,
      "train_runtime": 6003.9751,
      "train_tokens_per_second": 5999.74
    },
    {
      "epoch": 0.182,
      "grad_norm": 0.9061477622621547,
      "learning_rate": 9.797947362096909e-06,
      "loss": 0.2588,
      "num_input_tokens_seen": 36379920,
      "step": 455,
      "train_runtime": 6055.0509,
      "train_tokens_per_second": 6008.194
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.8275916426654469,
      "learning_rate": 9.793503413426016e-06,
      "loss": 0.283,
      "num_input_tokens_seen": 36772016,
      "step": 460,
      "train_runtime": 6126.2574,
      "train_tokens_per_second": 6002.362
    },
    {
      "epoch": 0.186,
      "grad_norm": 0.666351274255967,
      "learning_rate": 9.789012154811648e-06,
      "loss": 0.2506,
      "num_input_tokens_seen": 37179920,
      "step": 465,
      "train_runtime": 6201.0957,
      "train_tokens_per_second": 5995.702
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.7350566549608656,
      "learning_rate": 9.784473630580713e-06,
      "loss": 0.2597,
      "num_input_tokens_seen": 37564736,
      "step": 470,
      "train_runtime": 6266.0186,
      "train_tokens_per_second": 5994.993
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.8759339634430292,
      "learning_rate": 9.779887885526616e-06,
      "loss": 0.2517,
      "num_input_tokens_seen": 37917184,
      "step": 475,
      "train_runtime": 6331.4411,
      "train_tokens_per_second": 5988.713
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.7821815325835307,
      "learning_rate": 9.775254964908807e-06,
      "loss": 0.2538,
      "num_input_tokens_seen": 38358080,
      "step": 480,
      "train_runtime": 6403.8247,
      "train_tokens_per_second": 5989.87
    },
    {
      "epoch": 0.194,
      "grad_norm": 0.7670902572150325,
      "learning_rate": 9.770574914452343e-06,
      "loss": 0.26,
      "num_input_tokens_seen": 38766496,
      "step": 485,
      "train_runtime": 6470.9129,
      "train_tokens_per_second": 5990.885
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.7521204078415982,
      "learning_rate": 9.765847780347433e-06,
      "loss": 0.3009,
      "num_input_tokens_seen": 39138720,
      "step": 490,
      "train_runtime": 6532.3628,
      "train_tokens_per_second": 5991.51
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.8185418550494987,
      "learning_rate": 9.761073609248981e-06,
      "loss": 0.268,
      "num_input_tokens_seen": 39469344,
      "step": 495,
      "train_runtime": 6593.0053,
      "train_tokens_per_second": 5986.548
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.65232833162797,
      "learning_rate": 9.756252448276128e-06,
      "loss": 0.2887,
      "num_input_tokens_seen": 39864896,
      "step": 500,
      "train_runtime": 6658.7107,
      "train_tokens_per_second": 5986.879
    },
    {
      "epoch": 0.202,
      "grad_norm": 0.8062005254862974,
      "learning_rate": 9.751384345011787e-06,
      "loss": 0.2467,
      "num_input_tokens_seen": 40238784,
      "step": 505,
      "train_runtime": 6729.641,
      "train_tokens_per_second": 5979.336
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.8767149003459634,
      "learning_rate": 9.746469347502174e-06,
      "loss": 0.2698,
      "num_input_tokens_seen": 40666368,
      "step": 510,
      "train_runtime": 6799.6733,
      "train_tokens_per_second": 5980.636
    },
    {
      "epoch": 0.206,
      "grad_norm": 0.8711260304261372,
      "learning_rate": 9.741507504256327e-06,
      "loss": 0.2785,
      "num_input_tokens_seen": 41071520,
      "step": 515,
      "train_runtime": 6866.2851,
      "train_tokens_per_second": 5981.622
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.6363360313886239,
      "learning_rate": 9.736498864245638e-06,
      "loss": 0.2527,
      "num_input_tokens_seen": 41512256,
      "step": 520,
      "train_runtime": 6939.7657,
      "train_tokens_per_second": 5981.795
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.8109179802915915,
      "learning_rate": 9.73144347690336e-06,
      "loss": 0.29,
      "num_input_tokens_seen": 41873264,
      "step": 525,
      "train_runtime": 6999.3465,
      "train_tokens_per_second": 5982.453
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.8539497377334406,
      "learning_rate": 9.726341392124127e-06,
      "loss": 0.2566,
      "num_input_tokens_seen": 42217568,
      "step": 530,
      "train_runtime": 7053.5565,
      "train_tokens_per_second": 5985.288
    },
    {
      "epoch": 0.214,
      "grad_norm": 0.7599514956902396,
      "learning_rate": 9.721192660263454e-06,
      "loss": 0.2762,
      "num_input_tokens_seen": 42533184,
      "step": 535,
      "train_runtime": 7107.1623,
      "train_tokens_per_second": 5984.552
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.7228195853216592,
      "learning_rate": 9.715997332137248e-06,
      "loss": 0.2783,
      "num_input_tokens_seen": 42919952,
      "step": 540,
      "train_runtime": 7177.2775,
      "train_tokens_per_second": 5979.977
    },
    {
      "epoch": 0.218,
      "grad_norm": 0.8834406791399088,
      "learning_rate": 9.710755459021297e-06,
      "loss": 0.2809,
      "num_input_tokens_seen": 43338336,
      "step": 545,
      "train_runtime": 7248.3874,
      "train_tokens_per_second": 5979.031
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.8666526531262357,
      "learning_rate": 9.705467092650775e-06,
      "loss": 0.2603,
      "num_input_tokens_seen": 43671568,
      "step": 550,
      "train_runtime": 7301.6838,
      "train_tokens_per_second": 5981.027
    },
    {
      "epoch": 0.222,
      "grad_norm": 1.3601258560189031,
      "learning_rate": 9.700132285219724e-06,
      "loss": 0.2591,
      "num_input_tokens_seen": 44031712,
      "step": 555,
      "train_runtime": 7371.2006,
      "train_tokens_per_second": 5973.479
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.7394669684887641,
      "learning_rate": 9.694751089380536e-06,
      "loss": 0.2417,
      "num_input_tokens_seen": 44454032,
      "step": 560,
      "train_runtime": 7434.7623,
      "train_tokens_per_second": 5979.214
    },
    {
      "epoch": 0.226,
      "grad_norm": 0.9392531588494222,
      "learning_rate": 9.689323558243446e-06,
      "loss": 0.2764,
      "num_input_tokens_seen": 44824640,
      "step": 565,
      "train_runtime": 7492.0783,
      "train_tokens_per_second": 5982.938
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.9030061133696439,
      "learning_rate": 9.683849745375991e-06,
      "loss": 0.2747,
      "num_input_tokens_seen": 45255888,
      "step": 570,
      "train_runtime": 7561.0701,
      "train_tokens_per_second": 5985.381
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.7295413257969424,
      "learning_rate": 9.678329704802495e-06,
      "loss": 0.2908,
      "num_input_tokens_seen": 45694304,
      "step": 575,
      "train_runtime": 7636.2571,
      "train_tokens_per_second": 5983.861
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.7831225605388562,
      "learning_rate": 9.672763491003531e-06,
      "loss": 0.2688,
      "num_input_tokens_seen": 46114448,
      "step": 580,
      "train_runtime": 7706.713,
      "train_tokens_per_second": 5983.673
    },
    {
      "epoch": 0.234,
      "grad_norm": 0.7815429105705216,
      "learning_rate": 9.667151158915382e-06,
      "loss": 0.2664,
      "num_input_tokens_seen": 46491280,
      "step": 585,
      "train_runtime": 7768.7784,
      "train_tokens_per_second": 5984.375
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.8060541161386678,
      "learning_rate": 9.661492763929502e-06,
      "loss": 0.2569,
      "num_input_tokens_seen": 46979712,
      "step": 590,
      "train_runtime": 7846.7326,
      "train_tokens_per_second": 5987.169
    },
    {
      "epoch": 0.238,
      "grad_norm": 1.152392275434152,
      "learning_rate": 9.65578836189196e-06,
      "loss": 0.2726,
      "num_input_tokens_seen": 47322192,
      "step": 595,
      "train_runtime": 7910.5668,
      "train_tokens_per_second": 5982.149
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6887772490533477,
      "learning_rate": 9.650038009102905e-06,
      "loss": 0.2834,
      "num_input_tokens_seen": 47679712,
      "step": 600,
      "train_runtime": 7966.1483,
      "train_tokens_per_second": 5985.291
    },
    {
      "epoch": 0.242,
      "grad_norm": 0.6130384062661248,
      "learning_rate": 9.644241762315995e-06,
      "loss": 0.2423,
      "num_input_tokens_seen": 48067456,
      "step": 605,
      "train_runtime": 8031.7985,
      "train_tokens_per_second": 5984.644
    },
    {
      "epoch": 0.244,
      "grad_norm": 1.2446167359065596,
      "learning_rate": 9.63839967873785e-06,
      "loss": 0.2658,
      "num_input_tokens_seen": 48471152,
      "step": 610,
      "train_runtime": 8102.9608,
      "train_tokens_per_second": 5981.906
    },
    {
      "epoch": 0.246,
      "grad_norm": 0.7294863740740954,
      "learning_rate": 9.63251181602747e-06,
      "loss": 0.2777,
      "num_input_tokens_seen": 48821504,
      "step": 615,
      "train_runtime": 8163.5213,
      "train_tokens_per_second": 5980.447
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.7657155324916852,
      "learning_rate": 9.626578232295689e-06,
      "loss": 0.2731,
      "num_input_tokens_seen": 49249872,
      "step": 620,
      "train_runtime": 8238.4067,
      "train_tokens_per_second": 5978.082
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.8492612505753264,
      "learning_rate": 9.620598986104578e-06,
      "loss": 0.2416,
      "num_input_tokens_seen": 49722272,
      "step": 625,
      "train_runtime": 8312.1519,
      "train_tokens_per_second": 5981.877
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.5904583667356289,
      "learning_rate": 9.61457413646689e-06,
      "loss": 0.2351,
      "num_input_tokens_seen": 50162672,
      "step": 630,
      "train_runtime": 8385.8082,
      "train_tokens_per_second": 5981.853
    },
    {
      "epoch": 0.254,
      "grad_norm": 0.8152781635195294,
      "learning_rate": 9.60850374284545e-06,
      "loss": 0.2645,
      "num_input_tokens_seen": 50561280,
      "step": 635,
      "train_runtime": 8456.8003,
      "train_tokens_per_second": 5978.772
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.8578030045027673,
      "learning_rate": 9.602387865152597e-06,
      "loss": 0.2591,
      "num_input_tokens_seen": 50972256,
      "step": 640,
      "train_runtime": 8523.5289,
      "train_tokens_per_second": 5980.182
    },
    {
      "epoch": 0.258,
      "grad_norm": 0.6780597416159073,
      "learning_rate": 9.596226563749575e-06,
      "loss": 0.2685,
      "num_input_tokens_seen": 51396304,
      "step": 645,
      "train_runtime": 8597.0686,
      "train_tokens_per_second": 5978.352
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.102448723613161,
      "learning_rate": 9.59001989944594e-06,
      "loss": 0.2784,
      "num_input_tokens_seen": 51798816,
      "step": 650,
      "train_runtime": 8668.8465,
      "train_tokens_per_second": 5975.284
    },
    {
      "epoch": 0.262,
      "grad_norm": 0.8841959640928475,
      "learning_rate": 9.583767933498964e-06,
      "loss": 0.2514,
      "num_input_tokens_seen": 52139568,
      "step": 655,
      "train_runtime": 8710.9548,
      "train_tokens_per_second": 5985.517
    },
    {
      "epoch": 0.264,
      "grad_norm": 1.037638223962085,
      "learning_rate": 9.577470727613025e-06,
      "loss": 0.2774,
      "num_input_tokens_seen": 52493312,
      "step": 660,
      "train_runtime": 8765.6718,
      "train_tokens_per_second": 5988.51
    },
    {
      "epoch": 0.266,
      "grad_norm": 0.8858274237729105,
      "learning_rate": 9.571128343939006e-06,
      "loss": 0.2814,
      "num_input_tokens_seen": 52863632,
      "step": 665,
      "train_runtime": 8843.4286,
      "train_tokens_per_second": 5977.73
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.8391980666108796,
      "learning_rate": 9.56474084507367e-06,
      "loss": 0.261,
      "num_input_tokens_seen": 53215856,
      "step": 670,
      "train_runtime": 8907.6923,
      "train_tokens_per_second": 5974.146
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.8495736711346393,
      "learning_rate": 9.558308294059055e-06,
      "loss": 0.2676,
      "num_input_tokens_seen": 53562768,
      "step": 675,
      "train_runtime": 8965.163,
      "train_tokens_per_second": 5974.545
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.941473355481309,
      "learning_rate": 9.55183075438184e-06,
      "loss": 0.2357,
      "num_input_tokens_seen": 54030624,
      "step": 680,
      "train_runtime": 9040.6249,
      "train_tokens_per_second": 5976.426
    },
    {
      "epoch": 0.274,
      "grad_norm": 0.7012350033428856,
      "learning_rate": 9.545308289972727e-06,
      "loss": 0.2402,
      "num_input_tokens_seen": 54343888,
      "step": 685,
      "train_runtime": 9086.5654,
      "train_tokens_per_second": 5980.685
    },
    {
      "epoch": 0.276,
      "grad_norm": 0.626498976221356,
      "learning_rate": 9.538740965205809e-06,
      "loss": 0.2372,
      "num_input_tokens_seen": 54668576,
      "step": 690,
      "train_runtime": 9140.102,
      "train_tokens_per_second": 5981.178
    },
    {
      "epoch": 0.278,
      "grad_norm": 0.6826557552174048,
      "learning_rate": 9.532128844897928e-06,
      "loss": 0.2741,
      "num_input_tokens_seen": 55095472,
      "step": 695,
      "train_runtime": 9209.0632,
      "train_tokens_per_second": 5982.744
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.8107331496725793,
      "learning_rate": 9.52547199430804e-06,
      "loss": 0.2364,
      "num_input_tokens_seen": 55413664,
      "step": 700,
      "train_runtime": 9262.2508,
      "train_tokens_per_second": 5982.743
    },
    {
      "epoch": 0.282,
      "grad_norm": 0.7874944184568773,
      "learning_rate": 9.51877047913658e-06,
      "loss": 0.2586,
      "num_input_tokens_seen": 55779360,
      "step": 705,
      "train_runtime": 9320.6217,
      "train_tokens_per_second": 5984.511
    },
    {
      "epoch": 0.284,
      "grad_norm": 1.8567686550412206,
      "learning_rate": 9.512024365524788e-06,
      "loss": 0.2655,
      "num_input_tokens_seen": 56130272,
      "step": 710,
      "train_runtime": 9375.9114,
      "train_tokens_per_second": 5986.647
    },
    {
      "epoch": 0.286,
      "grad_norm": 0.7851630152910232,
      "learning_rate": 9.505233720054086e-06,
      "loss": 0.2579,
      "num_input_tokens_seen": 56500464,
      "step": 715,
      "train_runtime": 9440.7193,
      "train_tokens_per_second": 5984.763
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.8612089302332713,
      "learning_rate": 9.498398609745405e-06,
      "loss": 0.2478,
      "num_input_tokens_seen": 56844928,
      "step": 720,
      "train_runtime": 9500.9395,
      "train_tokens_per_second": 5983.085
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7600175238797224,
      "learning_rate": 9.491519102058523e-06,
      "loss": 0.2765,
      "num_input_tokens_seen": 57261328,
      "step": 725,
      "train_runtime": 9573.8278,
      "train_tokens_per_second": 5981.028
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.9375851471978519,
      "learning_rate": 9.484595264891403e-06,
      "loss": 0.2419,
      "num_input_tokens_seen": 57632624,
      "step": 730,
      "train_runtime": 9644.209,
      "train_tokens_per_second": 5975.879
    },
    {
      "epoch": 0.294,
      "grad_norm": 0.9027653994149234,
      "learning_rate": 9.477627166579523e-06,
      "loss": 0.271,
      "num_input_tokens_seen": 58025408,
      "step": 735,
      "train_runtime": 9710.5708,
      "train_tokens_per_second": 5975.489
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.8578883410124413,
      "learning_rate": 9.4706148758952e-06,
      "loss": 0.2637,
      "num_input_tokens_seen": 58396208,
      "step": 740,
      "train_runtime": 9777.666,
      "train_tokens_per_second": 5972.408
    },
    {
      "epoch": 0.298,
      "grad_norm": 0.8776112915445362,
      "learning_rate": 9.463558462046912e-06,
      "loss": 0.2727,
      "num_input_tokens_seen": 58707376,
      "step": 745,
      "train_runtime": 9819.0785,
      "train_tokens_per_second": 5978.909
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.5840963430287371,
      "learning_rate": 9.456457994678616e-06,
      "loss": 0.2322,
      "num_input_tokens_seen": 59145392,
      "step": 750,
      "train_runtime": 9890.3227,
      "train_tokens_per_second": 5980.128
    },
    {
      "epoch": 0.302,
      "grad_norm": 0.7317905873414259,
      "learning_rate": 9.449313543869056e-06,
      "loss": 0.2846,
      "num_input_tokens_seen": 59450400,
      "step": 755,
      "train_runtime": 9940.4488,
      "train_tokens_per_second": 5980.656
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.7403200227430362,
      "learning_rate": 9.44212518013108e-06,
      "loss": 0.2741,
      "num_input_tokens_seen": 59840944,
      "step": 760,
      "train_runtime": 10013.8692,
      "train_tokens_per_second": 5975.806
    },
    {
      "epoch": 0.306,
      "grad_norm": 0.7622642170344124,
      "learning_rate": 9.434892974410932e-06,
      "loss": 0.2604,
      "num_input_tokens_seen": 60216688,
      "step": 765,
      "train_runtime": 10074.7344,
      "train_tokens_per_second": 5977.0
    },
    {
      "epoch": 0.308,
      "grad_norm": 0.8785789117135203,
      "learning_rate": 9.427616998087568e-06,
      "loss": 0.2431,
      "num_input_tokens_seen": 60638976,
      "step": 770,
      "train_runtime": 10146.9679,
      "train_tokens_per_second": 5976.069
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.8249894805799513,
      "learning_rate": 9.420297322971934e-06,
      "loss": 0.2526,
      "num_input_tokens_seen": 61134256,
      "step": 775,
      "train_runtime": 10225.8927,
      "train_tokens_per_second": 5978.378
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.6744756944455406,
      "learning_rate": 9.412934021306267e-06,
      "loss": 0.2512,
      "num_input_tokens_seen": 61553648,
      "step": 780,
      "train_runtime": 10286.5901,
      "train_tokens_per_second": 5983.873
    },
    {
      "epoch": 0.314,
      "grad_norm": 0.8062812974951873,
      "learning_rate": 9.405527165763384e-06,
      "loss": 0.286,
      "num_input_tokens_seen": 61957648,
      "step": 785,
      "train_runtime": 10352.746,
      "train_tokens_per_second": 5984.658
    },
    {
      "epoch": 0.316,
      "grad_norm": 0.7356106568881381,
      "learning_rate": 9.398076829445958e-06,
      "loss": 0.2711,
      "num_input_tokens_seen": 62288880,
      "step": 790,
      "train_runtime": 10408.5909,
      "train_tokens_per_second": 5984.372
    },
    {
      "epoch": 0.318,
      "grad_norm": 0.7610857265233825,
      "learning_rate": 9.3905830858858e-06,
      "loss": 0.2685,
      "num_input_tokens_seen": 62663632,
      "step": 795,
      "train_runtime": 10467.9413,
      "train_tokens_per_second": 5986.242
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.7966898610801778,
      "learning_rate": 9.383046009043134e-06,
      "loss": 0.2693,
      "num_input_tokens_seen": 63057792,
      "step": 800,
      "train_runtime": 10539.3948,
      "train_tokens_per_second": 5983.056
    },
    {
      "epoch": 0.322,
      "grad_norm": 0.7289146876069654,
      "learning_rate": 9.37546567330587e-06,
      "loss": 0.2542,
      "num_input_tokens_seen": 63428432,
      "step": 805,
      "train_runtime": 10606.3414,
      "train_tokens_per_second": 5980.237
    },
    {
      "epoch": 0.324,
      "grad_norm": 0.7123973284012891,
      "learning_rate": 9.367842153488853e-06,
      "loss": 0.2482,
      "num_input_tokens_seen": 63837776,
      "step": 810,
      "train_runtime": 10684.1996,
      "train_tokens_per_second": 5974.97
    },
    {
      "epoch": 0.326,
      "grad_norm": 0.6116848677069046,
      "learning_rate": 9.360175524833153e-06,
      "loss": 0.2393,
      "num_input_tokens_seen": 64207744,
      "step": 815,
      "train_runtime": 10746.2594,
      "train_tokens_per_second": 5974.892
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.7090177624564499,
      "learning_rate": 9.352465863005295e-06,
      "loss": 0.2464,
      "num_input_tokens_seen": 64527552,
      "step": 820,
      "train_runtime": 10794.1158,
      "train_tokens_per_second": 5978.03
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9228460888788435,
      "learning_rate": 9.344713244096533e-06,
      "loss": 0.2576,
      "num_input_tokens_seen": 64870000,
      "step": 825,
      "train_runtime": 10861.5895,
      "train_tokens_per_second": 5972.422
    },
    {
      "epoch": 0.332,
      "grad_norm": 0.7749537888998517,
      "learning_rate": 9.336917744622083e-06,
      "loss": 0.255,
      "num_input_tokens_seen": 65269280,
      "step": 830,
      "train_runtime": 10929.169,
      "train_tokens_per_second": 5972.026
    },
    {
      "epoch": 0.334,
      "grad_norm": 0.7288672113983271,
      "learning_rate": 9.329079441520377e-06,
      "loss": 0.2511,
      "num_input_tokens_seen": 65627792,
      "step": 835,
      "train_runtime": 10988.928,
      "train_tokens_per_second": 5972.174
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.601500012457241,
      "learning_rate": 9.321198412152303e-06,
      "loss": 0.2454,
      "num_input_tokens_seen": 66037840,
      "step": 840,
      "train_runtime": 11058.4902,
      "train_tokens_per_second": 5971.687
    },
    {
      "epoch": 0.338,
      "grad_norm": 0.8144980181728814,
      "learning_rate": 9.31327473430044e-06,
      "loss": 0.2629,
      "num_input_tokens_seen": 66426768,
      "step": 845,
      "train_runtime": 11133.0882,
      "train_tokens_per_second": 5966.608
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.7166518808881578,
      "learning_rate": 9.305308486168288e-06,
      "loss": 0.2955,
      "num_input_tokens_seen": 66772240,
      "step": 850,
      "train_runtime": 11191.3696,
      "train_tokens_per_second": 5966.405
    },
    {
      "epoch": 0.342,
      "grad_norm": 0.7897465503734388,
      "learning_rate": 9.297299746379503e-06,
      "loss": 0.2835,
      "num_input_tokens_seen": 67111024,
      "step": 855,
      "train_runtime": 11250.8182,
      "train_tokens_per_second": 5964.991
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.7396912203991731,
      "learning_rate": 9.28924859397711e-06,
      "loss": 0.2637,
      "num_input_tokens_seen": 67482144,
      "step": 860,
      "train_runtime": 11318.7263,
      "train_tokens_per_second": 5961.991
    },
    {
      "epoch": 0.346,
      "grad_norm": 0.6992972676350053,
      "learning_rate": 9.281155108422732e-06,
      "loss": 0.237,
      "num_input_tokens_seen": 67876032,
      "step": 865,
      "train_runtime": 11382.7481,
      "train_tokens_per_second": 5963.062
    },
    {
      "epoch": 0.348,
      "grad_norm": 0.6948771027671538,
      "learning_rate": 9.27301936959581e-06,
      "loss": 0.2797,
      "num_input_tokens_seen": 68282368,
      "step": 870,
      "train_runtime": 11454.6041,
      "train_tokens_per_second": 5961.129
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.7539839679695668,
      "learning_rate": 9.264841457792795e-06,
      "loss": 0.2602,
      "num_input_tokens_seen": 68701008,
      "step": 875,
      "train_runtime": 11524.7372,
      "train_tokens_per_second": 5961.178
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.7952906428828391,
      "learning_rate": 9.25662145372638e-06,
      "loss": 0.2757,
      "num_input_tokens_seen": 69139648,
      "step": 880,
      "train_runtime": 11594.0901,
      "train_tokens_per_second": 5963.353
    },
    {
      "epoch": 0.354,
      "grad_norm": 0.5875437368411229,
      "learning_rate": 9.248359438524683e-06,
      "loss": 0.2738,
      "num_input_tokens_seen": 69591040,
      "step": 885,
      "train_runtime": 11669.2233,
      "train_tokens_per_second": 5963.639
    },
    {
      "epoch": 0.356,
      "grad_norm": 0.8362276418870209,
      "learning_rate": 9.24005549373046e-06,
      "loss": 0.2693,
      "num_input_tokens_seen": 69919248,
      "step": 890,
      "train_runtime": 11720.5537,
      "train_tokens_per_second": 5965.524
    },
    {
      "epoch": 0.358,
      "grad_norm": 0.8689160927298022,
      "learning_rate": 9.231709701300293e-06,
      "loss": 0.2522,
      "num_input_tokens_seen": 70352544,
      "step": 895,
      "train_runtime": 11791.7221,
      "train_tokens_per_second": 5966.265
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.7199820818597936,
      "learning_rate": 9.223322143603786e-06,
      "loss": 0.2546,
      "num_input_tokens_seen": 70743040,
      "step": 900,
      "train_runtime": 11864.5227,
      "train_tokens_per_second": 5962.569
    },
    {
      "epoch": 0.362,
      "grad_norm": 0.6472872435071626,
      "learning_rate": 9.214892903422745e-06,
      "loss": 0.2453,
      "num_input_tokens_seen": 71217216,
      "step": 905,
      "train_runtime": 11942.6153,
      "train_tokens_per_second": 5963.285
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.6936535531088693,
      "learning_rate": 9.206422063950368e-06,
      "loss": 0.27,
      "num_input_tokens_seen": 71603280,
      "step": 910,
      "train_runtime": 12014.7541,
      "train_tokens_per_second": 5959.613
    },
    {
      "epoch": 0.366,
      "grad_norm": 0.7343273673132791,
      "learning_rate": 9.19790970879042e-06,
      "loss": 0.2542,
      "num_input_tokens_seen": 72051488,
      "step": 915,
      "train_runtime": 12090.529,
      "train_tokens_per_second": 5959.333
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.917360247584837,
      "learning_rate": 9.189355921956414e-06,
      "loss": 0.2589,
      "num_input_tokens_seen": 72527008,
      "step": 920,
      "train_runtime": 12168.4904,
      "train_tokens_per_second": 5960.231
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7350510206851814,
      "learning_rate": 9.180760787870766e-06,
      "loss": 0.2867,
      "num_input_tokens_seen": 72948304,
      "step": 925,
      "train_runtime": 12242.6135,
      "train_tokens_per_second": 5958.556
    },
    {
      "epoch": 0.372,
      "grad_norm": 0.7541549801390238,
      "learning_rate": 9.172124391363986e-06,
      "loss": 0.2782,
      "num_input_tokens_seen": 73288864,
      "step": 930,
      "train_runtime": 12299.826,
      "train_tokens_per_second": 5958.529
    },
    {
      "epoch": 0.374,
      "grad_norm": 0.8317564298674665,
      "learning_rate": 9.163446817673817e-06,
      "loss": 0.2686,
      "num_input_tokens_seen": 73737984,
      "step": 935,
      "train_runtime": 12370.4346,
      "train_tokens_per_second": 5960.824
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.6607759087260642,
      "learning_rate": 9.154728152444408e-06,
      "loss": 0.2233,
      "num_input_tokens_seen": 74213840,
      "step": 940,
      "train_runtime": 12443.1991,
      "train_tokens_per_second": 5964.209
    },
    {
      "epoch": 0.378,
      "grad_norm": 0.6062097980101698,
      "learning_rate": 9.145968481725466e-06,
      "loss": 0.2441,
      "num_input_tokens_seen": 74626672,
      "step": 945,
      "train_runtime": 12519.7182,
      "train_tokens_per_second": 5960.731
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.6654792002985258,
      "learning_rate": 9.137167891971407e-06,
      "loss": 0.2461,
      "num_input_tokens_seen": 75031472,
      "step": 950,
      "train_runtime": 12593.3703,
      "train_tokens_per_second": 5958.014
    },
    {
      "epoch": 0.382,
      "grad_norm": 0.8388912617650379,
      "learning_rate": 9.128326470040495e-06,
      "loss": 0.2524,
      "num_input_tokens_seen": 75420768,
      "step": 955,
      "train_runtime": 12664.1715,
      "train_tokens_per_second": 5955.444
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.6889457837790466,
      "learning_rate": 9.119444303193997e-06,
      "loss": 0.2614,
      "num_input_tokens_seen": 75773776,
      "step": 960,
      "train_runtime": 12727.9061,
      "train_tokens_per_second": 5953.358
    },
    {
      "epoch": 0.386,
      "grad_norm": 0.8297814546544214,
      "learning_rate": 9.110521479095314e-06,
      "loss": 0.2407,
      "num_input_tokens_seen": 76196496,
      "step": 965,
      "train_runtime": 12791.1909,
      "train_tokens_per_second": 5956.951
    },
    {
      "epoch": 0.388,
      "grad_norm": 0.6413044681177218,
      "learning_rate": 9.101558085809114e-06,
      "loss": 0.2573,
      "num_input_tokens_seen": 76591056,
      "step": 970,
      "train_runtime": 12855.6204,
      "train_tokens_per_second": 5957.788
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.7210732824974905,
      "learning_rate": 9.092554211800476e-06,
      "loss": 0.2697,
      "num_input_tokens_seen": 76955088,
      "step": 975,
      "train_runtime": 12909.069,
      "train_tokens_per_second": 5961.32
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.7245964043841143,
      "learning_rate": 9.083509945933996e-06,
      "loss": 0.2575,
      "num_input_tokens_seen": 77376480,
      "step": 980,
      "train_runtime": 12984.986,
      "train_tokens_per_second": 5958.919
    },
    {
      "epoch": 0.394,
      "grad_norm": 0.9126407724225012,
      "learning_rate": 9.074425377472932e-06,
      "loss": 0.2609,
      "num_input_tokens_seen": 77724816,
      "step": 985,
      "train_runtime": 13042.6743,
      "train_tokens_per_second": 5959.27
    },
    {
      "epoch": 0.396,
      "grad_norm": 0.7128423265110231,
      "learning_rate": 9.065300596078304e-06,
      "loss": 0.2565,
      "num_input_tokens_seen": 78153632,
      "step": 990,
      "train_runtime": 13120.8794,
      "train_tokens_per_second": 5956.432
    },
    {
      "epoch": 0.398,
      "grad_norm": 0.6067421606247867,
      "learning_rate": 9.056135691808019e-06,
      "loss": 0.2617,
      "num_input_tokens_seen": 78580512,
      "step": 995,
      "train_runtime": 13191.8231,
      "train_tokens_per_second": 5956.759
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.0171109792998725,
      "learning_rate": 9.046930755115986e-06,
      "loss": 0.2443,
      "num_input_tokens_seen": 79016832,
      "step": 1000,
      "train_runtime": 13262.4875,
      "train_tokens_per_second": 5957.919
    },
    {
      "epoch": 0.402,
      "grad_norm": 0.624788109000857,
      "learning_rate": 9.037685876851211e-06,
      "loss": 0.2496,
      "num_input_tokens_seen": 79339488,
      "step": 1005,
      "train_runtime": 13317.1987,
      "train_tokens_per_second": 5957.671
    },
    {
      "epoch": 0.404,
      "grad_norm": 1.3437666934516999,
      "learning_rate": 9.028401148256911e-06,
      "loss": 0.2595,
      "num_input_tokens_seen": 79734112,
      "step": 1010,
      "train_runtime": 13384.2845,
      "train_tokens_per_second": 5957.294
    },
    {
      "epoch": 0.406,
      "grad_norm": 0.9331183492160343,
      "learning_rate": 9.01907666096961e-06,
      "loss": 0.2649,
      "num_input_tokens_seen": 80113424,
      "step": 1015,
      "train_runtime": 13448.7587,
      "train_tokens_per_second": 5956.938
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.8916393321532121,
      "learning_rate": 9.00971250701823e-06,
      "loss": 0.2566,
      "num_input_tokens_seen": 80485408,
      "step": 1020,
      "train_runtime": 13514.594,
      "train_tokens_per_second": 5955.444
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6466488488459519,
      "learning_rate": 9.000308778823196e-06,
      "loss": 0.2335,
      "num_input_tokens_seen": 80804000,
      "step": 1025,
      "train_runtime": 13566.7577,
      "train_tokens_per_second": 5956.029
    },
    {
      "epoch": 0.412,
      "grad_norm": 0.7558803329595326,
      "learning_rate": 8.990865569195502e-06,
      "loss": 0.27,
      "num_input_tokens_seen": 81218816,
      "step": 1030,
      "train_runtime": 13639.7747,
      "train_tokens_per_second": 5954.557
    },
    {
      "epoch": 0.414,
      "grad_norm": 0.7131746675468659,
      "learning_rate": 8.98138297133582e-06,
      "loss": 0.2453,
      "num_input_tokens_seen": 81613792,
      "step": 1035,
      "train_runtime": 13706.1897,
      "train_tokens_per_second": 5954.521
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.6901488903873247,
      "learning_rate": 8.971861078833558e-06,
      "loss": 0.2493,
      "num_input_tokens_seen": 82060224,
      "step": 1040,
      "train_runtime": 13784.5271,
      "train_tokens_per_second": 5953.068
    },
    {
      "epoch": 0.418,
      "grad_norm": 0.6646198187655422,
      "learning_rate": 8.962299985665955e-06,
      "loss": 0.2342,
      "num_input_tokens_seen": 82494944,
      "step": 1045,
      "train_runtime": 13849.5372,
      "train_tokens_per_second": 5956.513
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.6253491898077594,
      "learning_rate": 8.952699786197137e-06,
      "loss": 0.2373,
      "num_input_tokens_seen": 82808768,
      "step": 1050,
      "train_runtime": 13914.8709,
      "train_tokens_per_second": 5951.099
    },
    {
      "epoch": 0.422,
      "grad_norm": 0.7708996437720635,
      "learning_rate": 8.943060575177197e-06,
      "loss": 0.243,
      "num_input_tokens_seen": 83196256,
      "step": 1055,
      "train_runtime": 13976.3704,
      "train_tokens_per_second": 5952.637
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.9176135630737275,
      "learning_rate": 8.93338244774126e-06,
      "loss": 0.2866,
      "num_input_tokens_seen": 83661872,
      "step": 1060,
      "train_runtime": 14050.0901,
      "train_tokens_per_second": 5954.543
    },
    {
      "epoch": 0.426,
      "grad_norm": 0.7219954539715555,
      "learning_rate": 8.923665499408535e-06,
      "loss": 0.2466,
      "num_input_tokens_seen": 84013360,
      "step": 1065,
      "train_runtime": 14119.6235,
      "train_tokens_per_second": 5950.113
    },
    {
      "epoch": 0.428,
      "grad_norm": 0.7629186621884477,
      "learning_rate": 8.91390982608138e-06,
      "loss": 0.2593,
      "num_input_tokens_seen": 84398000,
      "step": 1070,
      "train_runtime": 14195.2042,
      "train_tokens_per_second": 5945.529
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.7305819771780363,
      "learning_rate": 8.904115524044349e-06,
      "loss": 0.2553,
      "num_input_tokens_seen": 84711792,
      "step": 1075,
      "train_runtime": 14240.4426,
      "train_tokens_per_second": 5948.677
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.8171300968987977,
      "learning_rate": 8.894282689963252e-06,
      "loss": 0.2611,
      "num_input_tokens_seen": 85117264,
      "step": 1080,
      "train_runtime": 14298.5417,
      "train_tokens_per_second": 5952.863
    },
    {
      "epoch": 0.434,
      "grad_norm": 0.7999063479153911,
      "learning_rate": 8.88441142088419e-06,
      "loss": 0.2605,
      "num_input_tokens_seen": 85480624,
      "step": 1085,
      "train_runtime": 14353.0025,
      "train_tokens_per_second": 5955.592
    },
    {
      "epoch": 0.436,
      "grad_norm": 0.6760703286202087,
      "learning_rate": 8.874501814232603e-06,
      "loss": 0.2476,
      "num_input_tokens_seen": 85832000,
      "step": 1090,
      "train_runtime": 14408.823,
      "train_tokens_per_second": 5956.906
    },
    {
      "epoch": 0.438,
      "grad_norm": 0.6222891384021169,
      "learning_rate": 8.86455396781231e-06,
      "loss": 0.24,
      "num_input_tokens_seen": 86208160,
      "step": 1095,
      "train_runtime": 14471.0092,
      "train_tokens_per_second": 5957.301
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.8225171871864566,
      "learning_rate": 8.854567979804538e-06,
      "loss": 0.2628,
      "num_input_tokens_seen": 86613904,
      "step": 1100,
      "train_runtime": 14541.9038,
      "train_tokens_per_second": 5956.16
    },
    {
      "epoch": 0.442,
      "grad_norm": 0.7386320533997327,
      "learning_rate": 8.844543948766958e-06,
      "loss": 0.2561,
      "num_input_tokens_seen": 87012032,
      "step": 1105,
      "train_runtime": 14616.3222,
      "train_tokens_per_second": 5953.073
    },
    {
      "epoch": 0.444,
      "grad_norm": 0.8014432527357591,
      "learning_rate": 8.834481973632709e-06,
      "loss": 0.2612,
      "num_input_tokens_seen": 87397408,
      "step": 1110,
      "train_runtime": 14677.529,
      "train_tokens_per_second": 5954.504
    },
    {
      "epoch": 0.446,
      "grad_norm": 0.930443798550076,
      "learning_rate": 8.824382153709423e-06,
      "loss": 0.2647,
      "num_input_tokens_seen": 87762928,
      "step": 1115,
      "train_runtime": 14750.6792,
      "train_tokens_per_second": 5949.755
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.6435095826798272,
      "learning_rate": 8.814244588678245e-06,
      "loss": 0.242,
      "num_input_tokens_seen": 88121120,
      "step": 1120,
      "train_runtime": 14807.3984,
      "train_tokens_per_second": 5951.155
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.7653907584800824,
      "learning_rate": 8.80406937859285e-06,
      "loss": 0.2395,
      "num_input_tokens_seen": 88525008,
      "step": 1125,
      "train_runtime": 14870.9967,
      "train_tokens_per_second": 5952.863
    },
    {
      "epoch": 0.452,
      "grad_norm": 0.7573515883491185,
      "learning_rate": 8.793856623878453e-06,
      "loss": 0.2401,
      "num_input_tokens_seen": 88958160,
      "step": 1130,
      "train_runtime": 14945.3854,
      "train_tokens_per_second": 5952.216
    },
    {
      "epoch": 0.454,
      "grad_norm": 1.1357409156529634,
      "learning_rate": 8.78360642533082e-06,
      "loss": 0.2585,
      "num_input_tokens_seen": 89252000,
      "step": 1135,
      "train_runtime": 14990.5393,
      "train_tokens_per_second": 5953.889
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.8644413546151707,
      "learning_rate": 8.773318884115273e-06,
      "loss": 0.2189,
      "num_input_tokens_seen": 89628016,
      "step": 1140,
      "train_runtime": 15058.5277,
      "train_tokens_per_second": 5951.977
    },
    {
      "epoch": 0.458,
      "grad_norm": 0.7389022088863988,
      "learning_rate": 8.76299410176569e-06,
      "loss": 0.2649,
      "num_input_tokens_seen": 90034128,
      "step": 1145,
      "train_runtime": 15122.4941,
      "train_tokens_per_second": 5953.656
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.7710676631866474,
      "learning_rate": 8.752632180183504e-06,
      "loss": 0.2562,
      "num_input_tokens_seen": 90366512,
      "step": 1150,
      "train_runtime": 15178.7113,
      "train_tokens_per_second": 5953.504
    },
    {
      "epoch": 0.462,
      "grad_norm": 0.9633116613760667,
      "learning_rate": 8.7422332216367e-06,
      "loss": 0.2366,
      "num_input_tokens_seen": 90739280,
      "step": 1155,
      "train_runtime": 15251.2362,
      "train_tokens_per_second": 5949.634
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.8014531807399391,
      "learning_rate": 8.7317973287588e-06,
      "loss": 0.2465,
      "num_input_tokens_seen": 91111104,
      "step": 1160,
      "train_runtime": 15326.9003,
      "train_tokens_per_second": 5944.523
    },
    {
      "epoch": 0.466,
      "grad_norm": 0.8222390185564179,
      "learning_rate": 8.721324604547851e-06,
      "loss": 0.2689,
      "num_input_tokens_seen": 91526176,
      "step": 1165,
      "train_runtime": 15404.9048,
      "train_tokens_per_second": 5941.366
    },
    {
      "epoch": 0.468,
      "grad_norm": 1.0652869678290553,
      "learning_rate": 8.710815152365416e-06,
      "loss": 0.2943,
      "num_input_tokens_seen": 91904368,
      "step": 1170,
      "train_runtime": 15478.9534,
      "train_tokens_per_second": 5937.376
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.6030089904636131,
      "learning_rate": 8.700269075935542e-06,
      "loss": 0.2485,
      "num_input_tokens_seen": 92358160,
      "step": 1175,
      "train_runtime": 15556.3907,
      "train_tokens_per_second": 5936.992
    },
    {
      "epoch": 0.472,
      "grad_norm": 0.9092828633416254,
      "learning_rate": 8.689686479343747e-06,
      "loss": 0.2464,
      "num_input_tokens_seen": 92821904,
      "step": 1180,
      "train_runtime": 15632.5857,
      "train_tokens_per_second": 5937.719
    },
    {
      "epoch": 0.474,
      "grad_norm": 0.7894642154049413,
      "learning_rate": 8.679067467035989e-06,
      "loss": 0.2504,
      "num_input_tokens_seen": 93187600,
      "step": 1185,
      "train_runtime": 15697.5108,
      "train_tokens_per_second": 5936.457
    },
    {
      "epoch": 0.476,
      "grad_norm": 0.8786028772279872,
      "learning_rate": 8.66841214381763e-06,
      "loss": 0.2568,
      "num_input_tokens_seen": 93487696,
      "step": 1190,
      "train_runtime": 15741.8142,
      "train_tokens_per_second": 5938.813
    },
    {
      "epoch": 0.478,
      "grad_norm": 0.7893669505556468,
      "learning_rate": 8.657720614852412e-06,
      "loss": 0.2278,
      "num_input_tokens_seen": 93876672,
      "step": 1195,
      "train_runtime": 15806.9191,
      "train_tokens_per_second": 5938.961
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8565601256929846,
      "learning_rate": 8.646992985661404e-06,
      "loss": 0.2417,
      "num_input_tokens_seen": 94238000,
      "step": 1200,
      "train_runtime": 15871.1761,
      "train_tokens_per_second": 5937.682
    },
    {
      "epoch": 0.482,
      "grad_norm": 0.9164857605094122,
      "learning_rate": 8.636229362121979e-06,
      "loss": 0.2496,
      "num_input_tokens_seen": 94567520,
      "step": 1205,
      "train_runtime": 15931.6796,
      "train_tokens_per_second": 5935.816
    },
    {
      "epoch": 0.484,
      "grad_norm": 0.7077259117523806,
      "learning_rate": 8.625429850466756e-06,
      "loss": 0.2244,
      "num_input_tokens_seen": 94893184,
      "step": 1210,
      "train_runtime": 15990.0473,
      "train_tokens_per_second": 5934.516
    },
    {
      "epoch": 0.486,
      "grad_norm": 0.6485501307737555,
      "learning_rate": 8.614594557282553e-06,
      "loss": 0.2397,
      "num_input_tokens_seen": 95276080,
      "step": 1215,
      "train_runtime": 16055.5553,
      "train_tokens_per_second": 5934.15
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.6447256273387733,
      "learning_rate": 8.603723589509342e-06,
      "loss": 0.2477,
      "num_input_tokens_seen": 95722112,
      "step": 1220,
      "train_runtime": 16124.7208,
      "train_tokens_per_second": 5936.358
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.7012472368965309,
      "learning_rate": 8.592817054439184e-06,
      "loss": 0.2625,
      "num_input_tokens_seen": 96141024,
      "step": 1225,
      "train_runtime": 16196.1707,
      "train_tokens_per_second": 5936.034
    },
    {
      "epoch": 0.492,
      "grad_norm": 0.6324888494194018,
      "learning_rate": 8.581875059715177e-06,
      "loss": 0.2503,
      "num_input_tokens_seen": 96587392,
      "step": 1230,
      "train_runtime": 16262.1697,
      "train_tokens_per_second": 5939.391
    },
    {
      "epoch": 0.494,
      "grad_norm": 0.7054612824038707,
      "learning_rate": 8.570897713330392e-06,
      "loss": 0.2599,
      "num_input_tokens_seen": 97023264,
      "step": 1235,
      "train_runtime": 16336.3737,
      "train_tokens_per_second": 5939.094
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.8286039820576346,
      "learning_rate": 8.559885123626806e-06,
      "loss": 0.2513,
      "num_input_tokens_seen": 97430560,
      "step": 1240,
      "train_runtime": 16397.8473,
      "train_tokens_per_second": 5941.668
    },
    {
      "epoch": 0.498,
      "grad_norm": 0.564446015093804,
      "learning_rate": 8.548837399294235e-06,
      "loss": 0.2637,
      "num_input_tokens_seen": 97827472,
      "step": 1245,
      "train_runtime": 16455.0469,
      "train_tokens_per_second": 5945.135
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7169858114687652,
      "learning_rate": 8.537754649369256e-06,
      "loss": 0.2419,
      "num_input_tokens_seen": 98229376,
      "step": 1250,
      "train_runtime": 16532.9815,
      "train_tokens_per_second": 5941.419
    },
    {
      "epoch": 0.502,
      "grad_norm": 0.753834648508727,
      "learning_rate": 8.526636983234135e-06,
      "loss": 0.2418,
      "num_input_tokens_seen": 98599872,
      "step": 1255,
      "train_runtime": 16586.3065,
      "train_tokens_per_second": 5944.655
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.6582526707978466,
      "learning_rate": 8.515484510615753e-06,
      "loss": 0.2554,
      "num_input_tokens_seen": 99026864,
      "step": 1260,
      "train_runtime": 16653.3937,
      "train_tokens_per_second": 5946.347
    },
    {
      "epoch": 0.506,
      "grad_norm": 0.6945432775129547,
      "learning_rate": 8.504297341584509e-06,
      "loss": 0.2658,
      "num_input_tokens_seen": 99395904,
      "step": 1265,
      "train_runtime": 16722.2133,
      "train_tokens_per_second": 5943.944
    },
    {
      "epoch": 0.508,
      "grad_norm": 0.7918732739548086,
      "learning_rate": 8.493075586553245e-06,
      "loss": 0.2281,
      "num_input_tokens_seen": 99825616,
      "step": 1270,
      "train_runtime": 16787.7864,
      "train_tokens_per_second": 5946.324
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.7696512289103038,
      "learning_rate": 8.481819356276155e-06,
      "loss": 0.2434,
      "num_input_tokens_seen": 100220256,
      "step": 1275,
      "train_runtime": 16855.7722,
      "train_tokens_per_second": 5945.753
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.7508764829190959,
      "learning_rate": 8.470528761847684e-06,
      "loss": 0.2503,
      "num_input_tokens_seen": 100620576,
      "step": 1280,
      "train_runtime": 16926.3709,
      "train_tokens_per_second": 5944.604
    },
    {
      "epoch": 0.514,
      "grad_norm": 0.9707421200986698,
      "learning_rate": 8.459203914701444e-06,
      "loss": 0.2738,
      "num_input_tokens_seen": 101019904,
      "step": 1285,
      "train_runtime": 16986.4734,
      "train_tokens_per_second": 5947.079
    },
    {
      "epoch": 0.516,
      "grad_norm": 0.8459354981861116,
      "learning_rate": 8.447844926609103e-06,
      "loss": 0.2742,
      "num_input_tokens_seen": 101476336,
      "step": 1290,
      "train_runtime": 17064.6621,
      "train_tokens_per_second": 5946.578
    },
    {
      "epoch": 0.518,
      "grad_norm": 0.6264057337370963,
      "learning_rate": 8.436451909679286e-06,
      "loss": 0.2436,
      "num_input_tokens_seen": 101840464,
      "step": 1295,
      "train_runtime": 17121.8045,
      "train_tokens_per_second": 5947.998
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.8953106899143487,
      "learning_rate": 8.425024976356474e-06,
      "loss": 0.2449,
      "num_input_tokens_seen": 102215776,
      "step": 1300,
      "train_runtime": 17186.2906,
      "train_tokens_per_second": 5947.518
    },
    {
      "epoch": 0.522,
      "grad_norm": 0.9429866958899052,
      "learning_rate": 8.413564239419883e-06,
      "loss": 0.2516,
      "num_input_tokens_seen": 102620256,
      "step": 1305,
      "train_runtime": 17260.7388,
      "train_tokens_per_second": 5945.299
    },
    {
      "epoch": 0.524,
      "grad_norm": 0.7083743370162111,
      "learning_rate": 8.40206981198236e-06,
      "loss": 0.2496,
      "num_input_tokens_seen": 102981744,
      "step": 1310,
      "train_runtime": 17315.691,
      "train_tokens_per_second": 5947.308
    },
    {
      "epoch": 0.526,
      "grad_norm": 0.7049439211819539,
      "learning_rate": 8.390541807489266e-06,
      "loss": 0.2612,
      "num_input_tokens_seen": 103420880,
      "step": 1315,
      "train_runtime": 17394.0332,
      "train_tokens_per_second": 5945.768
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.6884521726121087,
      "learning_rate": 8.378980339717348e-06,
      "loss": 0.2544,
      "num_input_tokens_seen": 103881808,
      "step": 1320,
      "train_runtime": 17469.4075,
      "train_tokens_per_second": 5946.499
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.0516635133353769,
      "learning_rate": 8.367385522773625e-06,
      "loss": 0.2626,
      "num_input_tokens_seen": 104250576,
      "step": 1325,
      "train_runtime": 17525.1657,
      "train_tokens_per_second": 5948.621
    },
    {
      "epoch": 0.532,
      "grad_norm": 0.7098039380682222,
      "learning_rate": 8.355757471094263e-06,
      "loss": 0.2524,
      "num_input_tokens_seen": 104591808,
      "step": 1330,
      "train_runtime": 17586.521,
      "train_tokens_per_second": 5947.271
    },
    {
      "epoch": 0.534,
      "grad_norm": 0.5159589865497483,
      "learning_rate": 8.344096299443434e-06,
      "loss": 0.2493,
      "num_input_tokens_seen": 104999392,
      "step": 1335,
      "train_runtime": 17651.5705,
      "train_tokens_per_second": 5948.445
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.904008052443577,
      "learning_rate": 8.332402122912198e-06,
      "loss": 0.2369,
      "num_input_tokens_seen": 105360928,
      "step": 1340,
      "train_runtime": 17714.3169,
      "train_tokens_per_second": 5947.784
    },
    {
      "epoch": 0.538,
      "grad_norm": 0.6651472179253896,
      "learning_rate": 8.320675056917353e-06,
      "loss": 0.2566,
      "num_input_tokens_seen": 105742624,
      "step": 1345,
      "train_runtime": 17778.4474,
      "train_tokens_per_second": 5947.799
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.7927826502831012,
      "learning_rate": 8.308915217200305e-06,
      "loss": 0.2643,
      "num_input_tokens_seen": 106120016,
      "step": 1350,
      "train_runtime": 17837.0793,
      "train_tokens_per_second": 5949.405
    },
    {
      "epoch": 0.542,
      "grad_norm": 0.8565986406816929,
      "learning_rate": 8.297122719825928e-06,
      "loss": 0.2274,
      "num_input_tokens_seen": 106498992,
      "step": 1355,
      "train_runtime": 17911.1244,
      "train_tokens_per_second": 5945.969
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.6638912658832349,
      "learning_rate": 8.285297681181408e-06,
      "loss": 0.233,
      "num_input_tokens_seen": 106906656,
      "step": 1360,
      "train_runtime": 17973.95,
      "train_tokens_per_second": 5947.867
    },
    {
      "epoch": 0.546,
      "grad_norm": 0.8715953996990423,
      "learning_rate": 8.273440217975103e-06,
      "loss": 0.2527,
      "num_input_tokens_seen": 107212272,
      "step": 1365,
      "train_runtime": 18023.9906,
      "train_tokens_per_second": 5948.309
    },
    {
      "epoch": 0.548,
      "grad_norm": 0.7473907734015952,
      "learning_rate": 8.261550447235389e-06,
      "loss": 0.2356,
      "num_input_tokens_seen": 107567200,
      "step": 1370,
      "train_runtime": 18080.5797,
      "train_tokens_per_second": 5949.323
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.7915564307590499,
      "learning_rate": 8.2496284863095e-06,
      "loss": 0.2204,
      "num_input_tokens_seen": 108001632,
      "step": 1375,
      "train_runtime": 18151.1849,
      "train_tokens_per_second": 5950.115
    },
    {
      "epoch": 0.552,
      "grad_norm": 0.690764307158169,
      "learning_rate": 8.23767445286238e-06,
      "loss": 0.2518,
      "num_input_tokens_seen": 108400032,
      "step": 1380,
      "train_runtime": 18216.9813,
      "train_tokens_per_second": 5950.494
    },
    {
      "epoch": 0.554,
      "grad_norm": 0.6612957959493211,
      "learning_rate": 8.225688464875514e-06,
      "loss": 0.254,
      "num_input_tokens_seen": 108793600,
      "step": 1385,
      "train_runtime": 18294.9254,
      "train_tokens_per_second": 5946.654
    },
    {
      "epoch": 0.556,
      "grad_norm": 0.8146239119318059,
      "learning_rate": 8.213670640645762e-06,
      "loss": 0.2616,
      "num_input_tokens_seen": 109239856,
      "step": 1390,
      "train_runtime": 18361.5892,
      "train_tokens_per_second": 5949.368
    },
    {
      "epoch": 0.558,
      "grad_norm": 0.9407259867063503,
      "learning_rate": 8.201621098784198e-06,
      "loss": 0.2497,
      "num_input_tokens_seen": 109586848,
      "step": 1395,
      "train_runtime": 18423.9904,
      "train_tokens_per_second": 5948.052
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.6787557953668363,
      "learning_rate": 8.189539958214934e-06,
      "loss": 0.2422,
      "num_input_tokens_seen": 110004576,
      "step": 1400,
      "train_runtime": 18486.0657,
      "train_tokens_per_second": 5950.675
    },
    {
      "epoch": 0.562,
      "grad_norm": 0.6738263355281091,
      "learning_rate": 8.177427338173955e-06,
      "loss": 0.2511,
      "num_input_tokens_seen": 110425520,
      "step": 1405,
      "train_runtime": 18556.9515,
      "train_tokens_per_second": 5950.628
    },
    {
      "epoch": 0.564,
      "grad_norm": 0.6654146458771333,
      "learning_rate": 8.165283358207924e-06,
      "loss": 0.2545,
      "num_input_tokens_seen": 110798416,
      "step": 1410,
      "train_runtime": 18617.3657,
      "train_tokens_per_second": 5951.348
    },
    {
      "epoch": 0.566,
      "grad_norm": 0.7124097479742371,
      "learning_rate": 8.153108138173027e-06,
      "loss": 0.243,
      "num_input_tokens_seen": 111221456,
      "step": 1415,
      "train_runtime": 18693.9621,
      "train_tokens_per_second": 5949.592
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.7722193156964303,
      "learning_rate": 8.140901798233766e-06,
      "loss": 0.2481,
      "num_input_tokens_seen": 111584784,
      "step": 1420,
      "train_runtime": 18761.6948,
      "train_tokens_per_second": 5947.479
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.708503545830723,
      "learning_rate": 8.12866445886179e-06,
      "loss": 0.2444,
      "num_input_tokens_seen": 111922640,
      "step": 1425,
      "train_runtime": 18826.1559,
      "train_tokens_per_second": 5945.061
    },
    {
      "epoch": 0.572,
      "grad_norm": 0.858693182825798,
      "learning_rate": 8.116396240834699e-06,
      "loss": 0.2414,
      "num_input_tokens_seen": 112287760,
      "step": 1430,
      "train_runtime": 18895.7567,
      "train_tokens_per_second": 5942.485
    },
    {
      "epoch": 0.574,
      "grad_norm": 0.6950450411170785,
      "learning_rate": 8.10409726523485e-06,
      "loss": 0.2424,
      "num_input_tokens_seen": 112661728,
      "step": 1435,
      "train_runtime": 18947.4324,
      "train_tokens_per_second": 5946.016
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.9025950371365907,
      "learning_rate": 8.091767653448169e-06,
      "loss": 0.2563,
      "num_input_tokens_seen": 113064880,
      "step": 1440,
      "train_runtime": 19013.0998,
      "train_tokens_per_second": 5946.683
    },
    {
      "epoch": 0.578,
      "grad_norm": 0.783582958375125,
      "learning_rate": 8.079407527162944e-06,
      "loss": 0.2386,
      "num_input_tokens_seen": 113464160,
      "step": 1445,
      "train_runtime": 19080.6528,
      "train_tokens_per_second": 5946.555
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.6682602674979189,
      "learning_rate": 8.067017008368632e-06,
      "loss": 0.2558,
      "num_input_tokens_seen": 113867952,
      "step": 1450,
      "train_runtime": 19145.0648,
      "train_tokens_per_second": 5947.64
    },
    {
      "epoch": 0.582,
      "grad_norm": 0.707706565040942,
      "learning_rate": 8.054596219354655e-06,
      "loss": 0.2554,
      "num_input_tokens_seen": 114273424,
      "step": 1455,
      "train_runtime": 19222.6331,
      "train_tokens_per_second": 5944.733
    },
    {
      "epoch": 0.584,
      "grad_norm": 0.7233732054254851,
      "learning_rate": 8.042145282709181e-06,
      "loss": 0.2601,
      "num_input_tokens_seen": 114717424,
      "step": 1460,
      "train_runtime": 19289.3055,
      "train_tokens_per_second": 5947.203
    },
    {
      "epoch": 0.586,
      "grad_norm": 0.6422217738433834,
      "learning_rate": 8.029664321317932e-06,
      "loss": 0.2601,
      "num_input_tokens_seen": 115135984,
      "step": 1465,
      "train_runtime": 19354.3487,
      "train_tokens_per_second": 5948.843
    },
    {
      "epoch": 0.588,
      "grad_norm": 0.7333788682022774,
      "learning_rate": 8.017153458362957e-06,
      "loss": 0.2652,
      "num_input_tokens_seen": 115545440,
      "step": 1470,
      "train_runtime": 19414.5638,
      "train_tokens_per_second": 5951.483
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.6290152315681616,
      "learning_rate": 8.004612817321419e-06,
      "loss": 0.2579,
      "num_input_tokens_seen": 115943440,
      "step": 1475,
      "train_runtime": 19481.8184,
      "train_tokens_per_second": 5951.366
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.7113000835362475,
      "learning_rate": 7.99204252196439e-06,
      "loss": 0.2586,
      "num_input_tokens_seen": 116366400,
      "step": 1480,
      "train_runtime": 19550.9137,
      "train_tokens_per_second": 5951.967
    },
    {
      "epoch": 0.594,
      "grad_norm": 0.676866820935986,
      "learning_rate": 7.979442696355601e-06,
      "loss": 0.2324,
      "num_input_tokens_seen": 116736672,
      "step": 1485,
      "train_runtime": 19619.6574,
      "train_tokens_per_second": 5949.985
    },
    {
      "epoch": 0.596,
      "grad_norm": 0.63597242705262,
      "learning_rate": 7.966813464850252e-06,
      "loss": 0.2577,
      "num_input_tokens_seen": 117116064,
      "step": 1490,
      "train_runtime": 19679.1949,
      "train_tokens_per_second": 5951.263
    },
    {
      "epoch": 0.598,
      "grad_norm": 0.5934642312626544,
      "learning_rate": 7.954154952093754e-06,
      "loss": 0.2366,
      "num_input_tokens_seen": 117565696,
      "step": 1495,
      "train_runtime": 19754.6802,
      "train_tokens_per_second": 5951.283
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8246530677768509,
      "learning_rate": 7.941467283020521e-06,
      "loss": 0.2458,
      "num_input_tokens_seen": 117980528,
      "step": 1500,
      "train_runtime": 19825.7152,
      "train_tokens_per_second": 5950.884
    },
    {
      "epoch": 0.602,
      "grad_norm": 0.6803775472920268,
      "learning_rate": 7.928750582852722e-06,
      "loss": 0.2354,
      "num_input_tokens_seen": 118392000,
      "step": 1505,
      "train_runtime": 19894.375,
      "train_tokens_per_second": 5951.029
    },
    {
      "epoch": 0.604,
      "grad_norm": 0.6829702832729062,
      "learning_rate": 7.916004977099054e-06,
      "loss": 0.2437,
      "num_input_tokens_seen": 118773680,
      "step": 1510,
      "train_runtime": 19959.575,
      "train_tokens_per_second": 5950.712
    },
    {
      "epoch": 0.606,
      "grad_norm": 0.5796988428999044,
      "learning_rate": 7.903230591553504e-06,
      "loss": 0.2312,
      "num_input_tokens_seen": 119115104,
      "step": 1515,
      "train_runtime": 20011.3737,
      "train_tokens_per_second": 5952.37
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.7848124831820052,
      "learning_rate": 7.890427552294093e-06,
      "loss": 0.2591,
      "num_input_tokens_seen": 119460384,
      "step": 1520,
      "train_runtime": 20074.942,
      "train_tokens_per_second": 5950.721
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7615681932253368,
      "learning_rate": 7.877595985681656e-06,
      "loss": 0.2653,
      "num_input_tokens_seen": 119858560,
      "step": 1525,
      "train_runtime": 20141.7147,
      "train_tokens_per_second": 5950.762
    },
    {
      "epoch": 0.612,
      "grad_norm": 0.6772425523098504,
      "learning_rate": 7.864736018358571e-06,
      "loss": 0.2344,
      "num_input_tokens_seen": 120298352,
      "step": 1530,
      "train_runtime": 20211.3072,
      "train_tokens_per_second": 5952.032
    },
    {
      "epoch": 0.614,
      "grad_norm": 0.7585007734797548,
      "learning_rate": 7.851847777247528e-06,
      "loss": 0.2657,
      "num_input_tokens_seen": 120735536,
      "step": 1535,
      "train_runtime": 20289.6603,
      "train_tokens_per_second": 5950.594
    },
    {
      "epoch": 0.616,
      "grad_norm": 0.9257964242960042,
      "learning_rate": 7.83893138955026e-06,
      "loss": 0.2744,
      "num_input_tokens_seen": 121111328,
      "step": 1540,
      "train_runtime": 20354.3359,
      "train_tokens_per_second": 5950.149
    },
    {
      "epoch": 0.618,
      "grad_norm": 0.7584226438302087,
      "learning_rate": 7.8259869827463e-06,
      "loss": 0.2471,
      "num_input_tokens_seen": 121505792,
      "step": 1545,
      "train_runtime": 20414.9433,
      "train_tokens_per_second": 5951.806
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.7804373473000645,
      "learning_rate": 7.813014684591718e-06,
      "loss": 0.2406,
      "num_input_tokens_seen": 121888000,
      "step": 1550,
      "train_runtime": 20492.5707,
      "train_tokens_per_second": 5947.912
    },
    {
      "epoch": 0.622,
      "grad_norm": 0.6855534375843669,
      "learning_rate": 7.800014623117858e-06,
      "loss": 0.2424,
      "num_input_tokens_seen": 122302912,
      "step": 1555,
      "train_runtime": 20566.4692,
      "train_tokens_per_second": 5946.714
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.7810986327996574,
      "learning_rate": 7.786986926630079e-06,
      "loss": 0.2585,
      "num_input_tokens_seen": 122677712,
      "step": 1560,
      "train_runtime": 20618.1097,
      "train_tokens_per_second": 5949.998
    },
    {
      "epoch": 0.626,
      "grad_norm": 0.7759520825635546,
      "learning_rate": 7.773931723706487e-06,
      "loss": 0.2532,
      "num_input_tokens_seen": 123054656,
      "step": 1565,
      "train_runtime": 20678.9508,
      "train_tokens_per_second": 5950.72
    },
    {
      "epoch": 0.628,
      "grad_norm": 0.7987781132107461,
      "learning_rate": 7.760849143196664e-06,
      "loss": 0.2555,
      "num_input_tokens_seen": 123475648,
      "step": 1570,
      "train_runtime": 20756.4498,
      "train_tokens_per_second": 5948.785
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.7418203966884972,
      "learning_rate": 7.747739314220398e-06,
      "loss": 0.2416,
      "num_input_tokens_seen": 123922816,
      "step": 1575,
      "train_runtime": 20831.9196,
      "train_tokens_per_second": 5948.699
    },
    {
      "epoch": 0.632,
      "grad_norm": 0.6174717984031014,
      "learning_rate": 7.734602366166406e-06,
      "loss": 0.2552,
      "num_input_tokens_seen": 124277408,
      "step": 1580,
      "train_runtime": 20888.719,
      "train_tokens_per_second": 5949.499
    },
    {
      "epoch": 0.634,
      "grad_norm": 0.587009610351487,
      "learning_rate": 7.721438428691065e-06,
      "loss": 0.2438,
      "num_input_tokens_seen": 124687888,
      "step": 1585,
      "train_runtime": 20962.2577,
      "train_tokens_per_second": 5948.209
    },
    {
      "epoch": 0.636,
      "grad_norm": 0.7006930473267186,
      "learning_rate": 7.708247631717122e-06,
      "loss": 0.2349,
      "num_input_tokens_seen": 125035328,
      "step": 1590,
      "train_runtime": 21019.6479,
      "train_tokens_per_second": 5948.498
    },
    {
      "epoch": 0.638,
      "grad_norm": 0.7599865290838017,
      "learning_rate": 7.695030105432417e-06,
      "loss": 0.2621,
      "num_input_tokens_seen": 125398688,
      "step": 1595,
      "train_runtime": 21076.0616,
      "train_tokens_per_second": 5949.816
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.6599089458820105,
      "learning_rate": 7.681785980288601e-06,
      "loss": 0.2424,
      "num_input_tokens_seen": 125792304,
      "step": 1600,
      "train_runtime": 21138.3773,
      "train_tokens_per_second": 5950.897
    },
    {
      "epoch": 0.642,
      "grad_norm": 0.6818631558687155,
      "learning_rate": 7.668515386999837e-06,
      "loss": 0.2416,
      "num_input_tokens_seen": 126242176,
      "step": 1605,
      "train_runtime": 21213.8268,
      "train_tokens_per_second": 5950.938
    },
    {
      "epoch": 0.644,
      "grad_norm": 0.6131556907902158,
      "learning_rate": 7.65521845654153e-06,
      "loss": 0.2298,
      "num_input_tokens_seen": 126620048,
      "step": 1610,
      "train_runtime": 21289.2762,
      "train_tokens_per_second": 5947.598
    },
    {
      "epoch": 0.646,
      "grad_norm": 0.7008878416577765,
      "learning_rate": 7.641895320149008e-06,
      "loss": 0.2582,
      "num_input_tokens_seen": 127024944,
      "step": 1615,
      "train_runtime": 21353.3578,
      "train_tokens_per_second": 5948.711
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.7032794542013124,
      "learning_rate": 7.628546109316257e-06,
      "loss": 0.2617,
      "num_input_tokens_seen": 127455024,
      "step": 1620,
      "train_runtime": 21427.4186,
      "train_tokens_per_second": 5948.221
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9036139861758402,
      "learning_rate": 7.615170955794592e-06,
      "loss": 0.2782,
      "num_input_tokens_seen": 127834640,
      "step": 1625,
      "train_runtime": 21504.2807,
      "train_tokens_per_second": 5944.614
    },
    {
      "epoch": 0.652,
      "grad_norm": 0.8227192501807431,
      "learning_rate": 7.60176999159138e-06,
      "loss": 0.25,
      "num_input_tokens_seen": 128188304,
      "step": 1630,
      "train_runtime": 21559.1839,
      "train_tokens_per_second": 5945.879
    },
    {
      "epoch": 0.654,
      "grad_norm": 0.6968370902757749,
      "learning_rate": 7.588343348968728e-06,
      "loss": 0.2343,
      "num_input_tokens_seen": 128624560,
      "step": 1635,
      "train_runtime": 21637.7464,
      "train_tokens_per_second": 5944.453
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.6928512624740695,
      "learning_rate": 7.574891160442179e-06,
      "loss": 0.2391,
      "num_input_tokens_seen": 128979536,
      "step": 1640,
      "train_runtime": 21696.0196,
      "train_tokens_per_second": 5944.848
    },
    {
      "epoch": 0.658,
      "grad_norm": 0.6282910568416581,
      "learning_rate": 7.561413558779401e-06,
      "loss": 0.2453,
      "num_input_tokens_seen": 129395280,
      "step": 1645,
      "train_runtime": 21773.3461,
      "train_tokens_per_second": 5942.829
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5387647591500422,
      "learning_rate": 7.547910676998883e-06,
      "loss": 0.2445,
      "num_input_tokens_seen": 129790480,
      "step": 1650,
      "train_runtime": 21839.9147,
      "train_tokens_per_second": 5942.811
    },
    {
      "epoch": 0.662,
      "grad_norm": 0.7777098347909889,
      "learning_rate": 7.534382648368617e-06,
      "loss": 0.2519,
      "num_input_tokens_seen": 130179568,
      "step": 1655,
      "train_runtime": 21901.6622,
      "train_tokens_per_second": 5943.821
    },
    {
      "epoch": 0.664,
      "grad_norm": 0.769633021051581,
      "learning_rate": 7.520829606404781e-06,
      "loss": 0.2464,
      "num_input_tokens_seen": 130546960,
      "step": 1660,
      "train_runtime": 21956.2789,
      "train_tokens_per_second": 5945.769
    },
    {
      "epoch": 0.666,
      "grad_norm": 0.798499009796827,
      "learning_rate": 7.507251684870433e-06,
      "loss": 0.2613,
      "num_input_tokens_seen": 131000944,
      "step": 1665,
      "train_runtime": 22027.5074,
      "train_tokens_per_second": 5947.152
    },
    {
      "epoch": 0.668,
      "grad_norm": 0.7054809636104681,
      "learning_rate": 7.493649017774172e-06,
      "loss": 0.2391,
      "num_input_tokens_seen": 131408096,
      "step": 1670,
      "train_runtime": 22092.0472,
      "train_tokens_per_second": 5948.208
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.7480944347369234,
      "learning_rate": 7.480021739368831e-06,
      "loss": 0.238,
      "num_input_tokens_seen": 131842384,
      "step": 1675,
      "train_runtime": 22165.098,
      "train_tokens_per_second": 5948.198
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.7749303371068519,
      "learning_rate": 7.46636998415015e-06,
      "loss": 0.2482,
      "num_input_tokens_seen": 132275008,
      "step": 1680,
      "train_runtime": 22241.2007,
      "train_tokens_per_second": 5947.296
    },
    {
      "epoch": 0.674,
      "grad_norm": 0.5451444378369477,
      "learning_rate": 7.452693886855438e-06,
      "loss": 0.2456,
      "num_input_tokens_seen": 132563520,
      "step": 1685,
      "train_runtime": 22282.3979,
      "train_tokens_per_second": 5949.248
    },
    {
      "epoch": 0.676,
      "grad_norm": 0.7569897610310852,
      "learning_rate": 7.438993582462255e-06,
      "loss": 0.2424,
      "num_input_tokens_seen": 132997392,
      "step": 1690,
      "train_runtime": 22352.6253,
      "train_tokens_per_second": 5949.967
    },
    {
      "epoch": 0.678,
      "grad_norm": 1.0505987211800965,
      "learning_rate": 7.425269206187076e-06,
      "loss": 0.2544,
      "num_input_tokens_seen": 133426496,
      "step": 1695,
      "train_runtime": 22430.7241,
      "train_tokens_per_second": 5948.381
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9802178584484053,
      "learning_rate": 7.411520893483952e-06,
      "loss": 0.2391,
      "num_input_tokens_seen": 133870544,
      "step": 1700,
      "train_runtime": 22509.2718,
      "train_tokens_per_second": 5947.351
    },
    {
      "epoch": 0.682,
      "grad_norm": 0.6878690951367751,
      "learning_rate": 7.397748780043179e-06,
      "loss": 0.2578,
      "num_input_tokens_seen": 134297232,
      "step": 1705,
      "train_runtime": 22573.9014,
      "train_tokens_per_second": 5949.226
    },
    {
      "epoch": 0.684,
      "grad_norm": 0.5834750130153558,
      "learning_rate": 7.38395300178996e-06,
      "loss": 0.2515,
      "num_input_tokens_seen": 134663632,
      "step": 1710,
      "train_runtime": 22644.4422,
      "train_tokens_per_second": 5946.873
    },
    {
      "epoch": 0.686,
      "grad_norm": 0.9249623959814148,
      "learning_rate": 7.370133694883052e-06,
      "loss": 0.2333,
      "num_input_tokens_seen": 135047920,
      "step": 1715,
      "train_runtime": 22705.4724,
      "train_tokens_per_second": 5947.814
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.6098108074100824,
      "learning_rate": 7.356290995713436e-06,
      "loss": 0.2569,
      "num_input_tokens_seen": 135533312,
      "step": 1720,
      "train_runtime": 22783.9238,
      "train_tokens_per_second": 5948.638
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.7622642918846533,
      "learning_rate": 7.342425040902967e-06,
      "loss": 0.2587,
      "num_input_tokens_seen": 135924224,
      "step": 1725,
      "train_runtime": 22851.1503,
      "train_tokens_per_second": 5948.244
    },
    {
      "epoch": 0.692,
      "grad_norm": 0.7997119352799948,
      "learning_rate": 7.32853596730302e-06,
      "loss": 0.2396,
      "num_input_tokens_seen": 136269040,
      "step": 1730,
      "train_runtime": 22918.3706,
      "train_tokens_per_second": 5945.843
    },
    {
      "epoch": 0.694,
      "grad_norm": 0.617079296944827,
      "learning_rate": 7.314623911993143e-06,
      "loss": 0.2483,
      "num_input_tokens_seen": 136639296,
      "step": 1735,
      "train_runtime": 22978.1982,
      "train_tokens_per_second": 5946.476
    },
    {
      "epoch": 0.696,
      "grad_norm": 0.7038872747532084,
      "learning_rate": 7.300689012279706e-06,
      "loss": 0.2255,
      "num_input_tokens_seen": 137049552,
      "step": 1740,
      "train_runtime": 23047.5997,
      "train_tokens_per_second": 5946.37
    },
    {
      "epoch": 0.698,
      "grad_norm": 0.700125410633313,
      "learning_rate": 7.286731405694544e-06,
      "loss": 0.2147,
      "num_input_tokens_seen": 137441952,
      "step": 1745,
      "train_runtime": 23110.0137,
      "train_tokens_per_second": 5947.29
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.7270805438257301,
      "learning_rate": 7.272751229993598e-06,
      "loss": 0.2395,
      "num_input_tokens_seen": 137846304,
      "step": 1750,
      "train_runtime": 23172.4464,
      "train_tokens_per_second": 5948.716
    },
    {
      "epoch": 0.702,
      "grad_norm": 0.9542369589000788,
      "learning_rate": 7.258748623155558e-06,
      "loss": 0.2618,
      "num_input_tokens_seen": 138221600,
      "step": 1755,
      "train_runtime": 23231.6747,
      "train_tokens_per_second": 5949.705
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.7708838947852765,
      "learning_rate": 7.244723723380504e-06,
      "loss": 0.2495,
      "num_input_tokens_seen": 138551408,
      "step": 1760,
      "train_runtime": 23285.9836,
      "train_tokens_per_second": 5949.992
    },
    {
      "epoch": 0.706,
      "grad_norm": 0.7712799091118461,
      "learning_rate": 7.23067666908853e-06,
      "loss": 0.2567,
      "num_input_tokens_seen": 138961472,
      "step": 1765,
      "train_runtime": 23355.8988,
      "train_tokens_per_second": 5949.738
    },
    {
      "epoch": 0.708,
      "grad_norm": 0.7569151862311946,
      "learning_rate": 7.216607598918392e-06,
      "loss": 0.2481,
      "num_input_tokens_seen": 139357456,
      "step": 1770,
      "train_runtime": 23423.3842,
      "train_tokens_per_second": 5949.501
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.6715017049823879,
      "learning_rate": 7.202516651726135e-06,
      "loss": 0.2582,
      "num_input_tokens_seen": 139772704,
      "step": 1775,
      "train_runtime": 23495.5051,
      "train_tokens_per_second": 5948.912
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.7171565711417337,
      "learning_rate": 7.1884039665837165e-06,
      "loss": 0.2319,
      "num_input_tokens_seen": 140175600,
      "step": 1780,
      "train_runtime": 23566.3719,
      "train_tokens_per_second": 5948.12
    },
    {
      "epoch": 0.714,
      "grad_norm": 0.7456884832299608,
      "learning_rate": 7.1742696827776415e-06,
      "loss": 0.2401,
      "num_input_tokens_seen": 140581648,
      "step": 1785,
      "train_runtime": 23637.9719,
      "train_tokens_per_second": 5947.28
    },
    {
      "epoch": 0.716,
      "grad_norm": 0.7525544880376253,
      "learning_rate": 7.160113939807587e-06,
      "loss": 0.239,
      "num_input_tokens_seen": 140968624,
      "step": 1790,
      "train_runtime": 23704.7362,
      "train_tokens_per_second": 5946.855
    },
    {
      "epoch": 0.718,
      "grad_norm": 0.7879553824914558,
      "learning_rate": 7.145936877385019e-06,
      "loss": 0.2254,
      "num_input_tokens_seen": 141456880,
      "step": 1795,
      "train_runtime": 23780.4742,
      "train_tokens_per_second": 5948.447
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.6311955732363271,
      "learning_rate": 7.131738635431822e-06,
      "loss": 0.2468,
      "num_input_tokens_seen": 141757984,
      "step": 1800,
      "train_runtime": 23831.672,
      "train_tokens_per_second": 5948.302
    },
    {
      "epoch": 0.722,
      "grad_norm": 0.6989839986980433,
      "learning_rate": 7.11751935407891e-06,
      "loss": 0.2488,
      "num_input_tokens_seen": 142217936,
      "step": 1805,
      "train_runtime": 23894.4573,
      "train_tokens_per_second": 5951.922
    },
    {
      "epoch": 0.724,
      "grad_norm": 0.7097400213735673,
      "learning_rate": 7.103279173664851e-06,
      "loss": 0.2444,
      "num_input_tokens_seen": 142634320,
      "step": 1810,
      "train_runtime": 23967.6077,
      "train_tokens_per_second": 5951.129
    },
    {
      "epoch": 0.726,
      "grad_norm": 0.8803147982554296,
      "learning_rate": 7.089018234734476e-06,
      "loss": 0.2165,
      "num_input_tokens_seen": 142950304,
      "step": 1815,
      "train_runtime": 24025.9246,
      "train_tokens_per_second": 5949.836
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.7666276980489303,
      "learning_rate": 7.074736678037495e-06,
      "loss": 0.2464,
      "num_input_tokens_seen": 143344704,
      "step": 1820,
      "train_runtime": 24103.6543,
      "train_tokens_per_second": 5947.011
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.7010307195980472,
      "learning_rate": 7.060434644527105e-06,
      "loss": 0.2115,
      "num_input_tokens_seen": 143742144,
      "step": 1825,
      "train_runtime": 24177.3942,
      "train_tokens_per_second": 5945.312
    },
    {
      "epoch": 0.732,
      "grad_norm": 0.6425733515884736,
      "learning_rate": 7.046112275358607e-06,
      "loss": 0.234,
      "num_input_tokens_seen": 144077152,
      "step": 1830,
      "train_runtime": 24227.8337,
      "train_tokens_per_second": 5946.762
    },
    {
      "epoch": 0.734,
      "grad_norm": 0.7067376629815052,
      "learning_rate": 7.031769711887999e-06,
      "loss": 0.2143,
      "num_input_tokens_seen": 144484576,
      "step": 1835,
      "train_runtime": 24292.9239,
      "train_tokens_per_second": 5947.599
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.7142769637706615,
      "learning_rate": 7.017407095670594e-06,
      "loss": 0.2273,
      "num_input_tokens_seen": 144852640,
      "step": 1840,
      "train_runtime": 24349.9516,
      "train_tokens_per_second": 5948.786
    },
    {
      "epoch": 0.738,
      "grad_norm": 0.7450024424388124,
      "learning_rate": 7.003024568459614e-06,
      "loss": 0.2441,
      "num_input_tokens_seen": 145182656,
      "step": 1845,
      "train_runtime": 24400.0657,
      "train_tokens_per_second": 5950.093
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.583393807541574,
      "learning_rate": 6.988622272204799e-06,
      "loss": 0.2422,
      "num_input_tokens_seen": 145621648,
      "step": 1850,
      "train_runtime": 24475.6918,
      "train_tokens_per_second": 5949.644
    },
    {
      "epoch": 0.742,
      "grad_norm": 0.6180993512501228,
      "learning_rate": 6.974200349050996e-06,
      "loss": 0.235,
      "num_input_tokens_seen": 146013264,
      "step": 1855,
      "train_runtime": 24541.8438,
      "train_tokens_per_second": 5949.564
    },
    {
      "epoch": 0.744,
      "grad_norm": 0.7147341091774608,
      "learning_rate": 6.959758941336762e-06,
      "loss": 0.2716,
      "num_input_tokens_seen": 146365616,
      "step": 1860,
      "train_runtime": 24601.9867,
      "train_tokens_per_second": 5949.341
    },
    {
      "epoch": 0.746,
      "grad_norm": 0.86917031416008,
      "learning_rate": 6.945298191592967e-06,
      "loss": 0.2341,
      "num_input_tokens_seen": 146754320,
      "step": 1865,
      "train_runtime": 24667.9934,
      "train_tokens_per_second": 5949.179
    },
    {
      "epoch": 0.748,
      "grad_norm": 0.7262798973273732,
      "learning_rate": 6.930818242541368e-06,
      "loss": 0.2514,
      "num_input_tokens_seen": 147171456,
      "step": 1870,
      "train_runtime": 24727.888,
      "train_tokens_per_second": 5951.639
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.6370198483444947,
      "learning_rate": 6.916319237093219e-06,
      "loss": 0.2189,
      "num_input_tokens_seen": 147646320,
      "step": 1875,
      "train_runtime": 24806.7081,
      "train_tokens_per_second": 5951.871
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.8726760293713298,
      "learning_rate": 6.901801318347848e-06,
      "loss": 0.2585,
      "num_input_tokens_seen": 147975568,
      "step": 1880,
      "train_runtime": 24864.9692,
      "train_tokens_per_second": 5951.166
    },
    {
      "epoch": 0.754,
      "grad_norm": 0.7702378297000733,
      "learning_rate": 6.887264629591254e-06,
      "loss": 0.2672,
      "num_input_tokens_seen": 148400576,
      "step": 1885,
      "train_runtime": 24934.885,
      "train_tokens_per_second": 5951.524
    },
    {
      "epoch": 0.756,
      "grad_norm": 0.8206586292419491,
      "learning_rate": 6.872709314294685e-06,
      "loss": 0.2799,
      "num_input_tokens_seen": 148783072,
      "step": 1890,
      "train_runtime": 25003.0005,
      "train_tokens_per_second": 5950.609
    },
    {
      "epoch": 0.758,
      "grad_norm": 0.6874374001615647,
      "learning_rate": 6.858135516113226e-06,
      "loss": 0.2466,
      "num_input_tokens_seen": 149130848,
      "step": 1895,
      "train_runtime": 25065.5397,
      "train_tokens_per_second": 5949.636
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8392979612325142,
      "learning_rate": 6.8435433788843865e-06,
      "loss": 0.2659,
      "num_input_tokens_seen": 149563296,
      "step": 1900,
      "train_runtime": 25136.1861,
      "train_tokens_per_second": 5950.119
    },
    {
      "epoch": 0.762,
      "grad_norm": 0.8786055292411358,
      "learning_rate": 6.8289330466266635e-06,
      "loss": 0.2309,
      "num_input_tokens_seen": 149916624,
      "step": 1905,
      "train_runtime": 25194.5613,
      "train_tokens_per_second": 5950.357
    },
    {
      "epoch": 0.764,
      "grad_norm": 0.8701214273853315,
      "learning_rate": 6.814304663538142e-06,
      "loss": 0.2335,
      "num_input_tokens_seen": 150286112,
      "step": 1910,
      "train_runtime": 25261.7882,
      "train_tokens_per_second": 5949.148
    },
    {
      "epoch": 0.766,
      "grad_norm": 0.7772777271877992,
      "learning_rate": 6.799658373995054e-06,
      "loss": 0.2577,
      "num_input_tokens_seen": 150645008,
      "step": 1915,
      "train_runtime": 25327.1219,
      "train_tokens_per_second": 5947.972
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.6789678615174672,
      "learning_rate": 6.784994322550367e-06,
      "loss": 0.2399,
      "num_input_tokens_seen": 151060512,
      "step": 1920,
      "train_runtime": 25397.5162,
      "train_tokens_per_second": 5947.846
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8342583329233456,
      "learning_rate": 6.770312653932346e-06,
      "loss": 0.2615,
      "num_input_tokens_seen": 151368336,
      "step": 1925,
      "train_runtime": 25449.7373,
      "train_tokens_per_second": 5947.737
    },
    {
      "epoch": 0.772,
      "grad_norm": 0.7715569011388002,
      "learning_rate": 6.755613513043136e-06,
      "loss": 0.2686,
      "num_input_tokens_seen": 151745056,
      "step": 1930,
      "train_runtime": 25520.8898,
      "train_tokens_per_second": 5945.916
    },
    {
      "epoch": 0.774,
      "grad_norm": 0.5625454378176744,
      "learning_rate": 6.740897044957322e-06,
      "loss": 0.2523,
      "num_input_tokens_seen": 152255616,
      "step": 1935,
      "train_runtime": 25590.1646,
      "train_tokens_per_second": 5949.771
    },
    {
      "epoch": 0.776,
      "grad_norm": 0.6661207653183532,
      "learning_rate": 6.726163394920503e-06,
      "loss": 0.2364,
      "num_input_tokens_seen": 152666400,
      "step": 1940,
      "train_runtime": 25660.9655,
      "train_tokens_per_second": 5949.363
    },
    {
      "epoch": 0.778,
      "grad_norm": 0.6388285351639795,
      "learning_rate": 6.711412708347857e-06,
      "loss": 0.2455,
      "num_input_tokens_seen": 153036000,
      "step": 1945,
      "train_runtime": 25720.8358,
      "train_tokens_per_second": 5949.884
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8766843893276587,
      "learning_rate": 6.696645130822704e-06,
      "loss": 0.2646,
      "num_input_tokens_seen": 153381872,
      "step": 1950,
      "train_runtime": 25775.8047,
      "train_tokens_per_second": 5950.614
    },
    {
      "epoch": 0.782,
      "grad_norm": 0.7564699089666967,
      "learning_rate": 6.681860808095074e-06,
      "loss": 0.2373,
      "num_input_tokens_seen": 153796576,
      "step": 1955,
      "train_runtime": 25843.2925,
      "train_tokens_per_second": 5951.122
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.6125793792916168,
      "learning_rate": 6.667059886080263e-06,
      "loss": 0.2367,
      "num_input_tokens_seen": 154185392,
      "step": 1960,
      "train_runtime": 25902.8859,
      "train_tokens_per_second": 5952.441
    },
    {
      "epoch": 0.786,
      "grad_norm": 0.7264229713494964,
      "learning_rate": 6.652242510857395e-06,
      "loss": 0.2566,
      "num_input_tokens_seen": 154549680,
      "step": 1965,
      "train_runtime": 25956.6843,
      "train_tokens_per_second": 5954.138
    },
    {
      "epoch": 0.788,
      "grad_norm": 0.6457872880391132,
      "learning_rate": 6.637408828667982e-06,
      "loss": 0.2559,
      "num_input_tokens_seen": 154973808,
      "step": 1970,
      "train_runtime": 26025.2685,
      "train_tokens_per_second": 5954.744
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.7336303068433705,
      "learning_rate": 6.622558985914478e-06,
      "loss": 0.2273,
      "num_input_tokens_seen": 155365328,
      "step": 1975,
      "train_runtime": 26095.6873,
      "train_tokens_per_second": 5953.678
    },
    {
      "epoch": 0.792,
      "grad_norm": 0.7703023606949606,
      "learning_rate": 6.6076931291588375e-06,
      "loss": 0.2382,
      "num_input_tokens_seen": 155713088,
      "step": 1980,
      "train_runtime": 26156.9246,
      "train_tokens_per_second": 5953.035
    },
    {
      "epoch": 0.794,
      "grad_norm": 0.6634539968504769,
      "learning_rate": 6.592811405121064e-06,
      "loss": 0.2294,
      "num_input_tokens_seen": 156104880,
      "step": 1985,
      "train_runtime": 26224.5285,
      "train_tokens_per_second": 5952.629
    },
    {
      "epoch": 0.796,
      "grad_norm": 0.7228410141497181,
      "learning_rate": 6.577913960677766e-06,
      "loss": 0.2317,
      "num_input_tokens_seen": 156474112,
      "step": 1990,
      "train_runtime": 26280.5612,
      "train_tokens_per_second": 5953.987
    },
    {
      "epoch": 0.798,
      "grad_norm": 0.7453160213506707,
      "learning_rate": 6.5630009428607065e-06,
      "loss": 0.2452,
      "num_input_tokens_seen": 156803472,
      "step": 1995,
      "train_runtime": 26338.1332,
      "train_tokens_per_second": 5953.477
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.7218884021194475,
      "learning_rate": 6.54807249885535e-06,
      "loss": 0.2334,
      "num_input_tokens_seen": 157110992,
      "step": 2000,
      "train_runtime": 26383.3178,
      "train_tokens_per_second": 5954.937
    },
    {
      "epoch": 0.802,
      "grad_norm": 0.5333416528634705,
      "learning_rate": 6.533128775999411e-06,
      "loss": 0.2265,
      "num_input_tokens_seen": 157500272,
      "step": 2005,
      "train_runtime": 26454.7029,
      "train_tokens_per_second": 5953.583
    },
    {
      "epoch": 0.804,
      "grad_norm": 0.6743123682865275,
      "learning_rate": 6.5181699217814025e-06,
      "loss": 0.2163,
      "num_input_tokens_seen": 157834992,
      "step": 2010,
      "train_runtime": 26522.5739,
      "train_tokens_per_second": 5950.968
    },
    {
      "epoch": 0.806,
      "grad_norm": 0.6875769892486772,
      "learning_rate": 6.503196083839175e-06,
      "loss": 0.2625,
      "num_input_tokens_seen": 158224720,
      "step": 2015,
      "train_runtime": 26587.824,
      "train_tokens_per_second": 5951.022
    },
    {
      "epoch": 0.808,
      "grad_norm": 0.6570377982154686,
      "learning_rate": 6.488207409958466e-06,
      "loss": 0.2608,
      "num_input_tokens_seen": 158592160,
      "step": 2020,
      "train_runtime": 26642.0577,
      "train_tokens_per_second": 5952.699
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.7948682216403246,
      "learning_rate": 6.473204048071433e-06,
      "loss": 0.253,
      "num_input_tokens_seen": 158955232,
      "step": 2025,
      "train_runtime": 26698.2361,
      "train_tokens_per_second": 5953.773
    },
    {
      "epoch": 0.812,
      "grad_norm": 0.5825323695015356,
      "learning_rate": 6.458186146255203e-06,
      "loss": 0.2568,
      "num_input_tokens_seen": 159277888,
      "step": 2030,
      "train_runtime": 26754.9701,
      "train_tokens_per_second": 5953.207
    },
    {
      "epoch": 0.814,
      "grad_norm": 0.7005050901769546,
      "learning_rate": 6.443153852730404e-06,
      "loss": 0.229,
      "num_input_tokens_seen": 159700576,
      "step": 2035,
      "train_runtime": 26818.2772,
      "train_tokens_per_second": 5954.916
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.8854758398921162,
      "learning_rate": 6.428107315859702e-06,
      "loss": 0.2511,
      "num_input_tokens_seen": 160017808,
      "step": 2040,
      "train_runtime": 26880.0385,
      "train_tokens_per_second": 5953.035
    },
    {
      "epoch": 0.818,
      "grad_norm": 0.685561132332679,
      "learning_rate": 6.413046684146343e-06,
      "loss": 0.2343,
      "num_input_tokens_seen": 160373024,
      "step": 2045,
      "train_runtime": 26950.2096,
      "train_tokens_per_second": 5950.715
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.7082441136939686,
      "learning_rate": 6.3979721062326815e-06,
      "loss": 0.2231,
      "num_input_tokens_seen": 160775872,
      "step": 2050,
      "train_runtime": 27019.1206,
      "train_tokens_per_second": 5950.448
    },
    {
      "epoch": 0.822,
      "grad_norm": 0.5096417611075408,
      "learning_rate": 6.382883730898717e-06,
      "loss": 0.2134,
      "num_input_tokens_seen": 161209616,
      "step": 2055,
      "train_runtime": 27091.3587,
      "train_tokens_per_second": 5950.592
    },
    {
      "epoch": 0.824,
      "grad_norm": 0.8009013794002183,
      "learning_rate": 6.36778170706062e-06,
      "loss": 0.2574,
      "num_input_tokens_seen": 161526656,
      "step": 2060,
      "train_runtime": 27139.6623,
      "train_tokens_per_second": 5951.683
    },
    {
      "epoch": 0.826,
      "grad_norm": 0.6785197101100082,
      "learning_rate": 6.352666183769269e-06,
      "loss": 0.2264,
      "num_input_tokens_seen": 162002192,
      "step": 2065,
      "train_runtime": 27214.9638,
      "train_tokens_per_second": 5952.688
    },
    {
      "epoch": 0.828,
      "grad_norm": 0.7744495254464379,
      "learning_rate": 6.337537310208779e-06,
      "loss": 0.2579,
      "num_input_tokens_seen": 162370528,
      "step": 2070,
      "train_runtime": 27284.1663,
      "train_tokens_per_second": 5951.09
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.7760494440011119,
      "learning_rate": 6.322395235695022e-06,
      "loss": 0.2442,
      "num_input_tokens_seen": 162742848,
      "step": 2075,
      "train_runtime": 27350.3482,
      "train_tokens_per_second": 5950.303
    },
    {
      "epoch": 0.832,
      "grad_norm": 1.0101069822281035,
      "learning_rate": 6.307240109674162e-06,
      "loss": 0.2468,
      "num_input_tokens_seen": 163114144,
      "step": 2080,
      "train_runtime": 27402.3481,
      "train_tokens_per_second": 5952.561
    },
    {
      "epoch": 0.834,
      "grad_norm": 0.6818940132851085,
      "learning_rate": 6.292072081721173e-06,
      "loss": 0.2531,
      "num_input_tokens_seen": 163504080,
      "step": 2085,
      "train_runtime": 27466.6183,
      "train_tokens_per_second": 5952.829
    },
    {
      "epoch": 0.836,
      "grad_norm": 0.6574555343522533,
      "learning_rate": 6.2768913015383696e-06,
      "loss": 0.2476,
      "num_input_tokens_seen": 163972864,
      "step": 2090,
      "train_runtime": 27543.1965,
      "train_tokens_per_second": 5953.298
    },
    {
      "epoch": 0.838,
      "grad_norm": 0.878884816070722,
      "learning_rate": 6.261697918953922e-06,
      "loss": 0.2503,
      "num_input_tokens_seen": 164394080,
      "step": 2095,
      "train_runtime": 27608.9053,
      "train_tokens_per_second": 5954.386
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.708190951091209,
      "learning_rate": 6.2464920839203805e-06,
      "loss": 0.2337,
      "num_input_tokens_seen": 164803104,
      "step": 2100,
      "train_runtime": 27675.9486,
      "train_tokens_per_second": 5954.741
    },
    {
      "epoch": 0.842,
      "grad_norm": 0.6729880610621995,
      "learning_rate": 6.231273946513201e-06,
      "loss": 0.2475,
      "num_input_tokens_seen": 165165728,
      "step": 2105,
      "train_runtime": 27742.8541,
      "train_tokens_per_second": 5953.451
    },
    {
      "epoch": 0.844,
      "grad_norm": 1.0390349425994523,
      "learning_rate": 6.216043656929254e-06,
      "loss": 0.226,
      "num_input_tokens_seen": 165552752,
      "step": 2110,
      "train_runtime": 27816.0039,
      "train_tokens_per_second": 5951.709
    },
    {
      "epoch": 0.846,
      "grad_norm": 0.6678856927176096,
      "learning_rate": 6.2008013654853505e-06,
      "loss": 0.239,
      "num_input_tokens_seen": 165952016,
      "step": 2115,
      "train_runtime": 27876.694,
      "train_tokens_per_second": 5953.074
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.6209290845661071,
      "learning_rate": 6.1855472226167525e-06,
      "loss": 0.2259,
      "num_input_tokens_seen": 166406208,
      "step": 2120,
      "train_runtime": 27952.9765,
      "train_tokens_per_second": 5953.077
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.9264111697053214,
      "learning_rate": 6.170281378875692e-06,
      "loss": 0.2217,
      "num_input_tokens_seen": 166793568,
      "step": 2125,
      "train_runtime": 28026.8287,
      "train_tokens_per_second": 5951.211
    },
    {
      "epoch": 0.852,
      "grad_norm": 0.8117944248091487,
      "learning_rate": 6.155003984929883e-06,
      "loss": 0.2205,
      "num_input_tokens_seen": 167110880,
      "step": 2130,
      "train_runtime": 28083.294,
      "train_tokens_per_second": 5950.544
    },
    {
      "epoch": 0.854,
      "grad_norm": 0.7750515953804475,
      "learning_rate": 6.139715191561038e-06,
      "loss": 0.2505,
      "num_input_tokens_seen": 167529632,
      "step": 2135,
      "train_runtime": 28151.1623,
      "train_tokens_per_second": 5951.073
    },
    {
      "epoch": 0.856,
      "grad_norm": 0.9646210816534532,
      "learning_rate": 6.124415149663374e-06,
      "loss": 0.2498,
      "num_input_tokens_seen": 167902176,
      "step": 2140,
      "train_runtime": 28212.803,
      "train_tokens_per_second": 5951.276
    },
    {
      "epoch": 0.858,
      "grad_norm": 0.637371204897658,
      "learning_rate": 6.109104010242127e-06,
      "loss": 0.2226,
      "num_input_tokens_seen": 168281184,
      "step": 2145,
      "train_runtime": 28272.597,
      "train_tokens_per_second": 5952.095
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.7820949153505182,
      "learning_rate": 6.093781924412063e-06,
      "loss": 0.223,
      "num_input_tokens_seen": 168648480,
      "step": 2150,
      "train_runtime": 28336.1128,
      "train_tokens_per_second": 5951.715
    },
    {
      "epoch": 0.862,
      "grad_norm": 0.7820636325924925,
      "learning_rate": 6.078449043395982e-06,
      "loss": 0.2435,
      "num_input_tokens_seen": 169014928,
      "step": 2155,
      "train_runtime": 28402.0916,
      "train_tokens_per_second": 5950.792
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.658338208972596,
      "learning_rate": 6.06310551852323e-06,
      "loss": 0.2333,
      "num_input_tokens_seen": 169395632,
      "step": 2160,
      "train_runtime": 28465.5346,
      "train_tokens_per_second": 5950.903
    },
    {
      "epoch": 0.866,
      "grad_norm": 0.7066861787934736,
      "learning_rate": 6.047751501228203e-06,
      "loss": 0.2464,
      "num_input_tokens_seen": 169811824,
      "step": 2165,
      "train_runtime": 28543.6007,
      "train_tokens_per_second": 5949.208
    },
    {
      "epoch": 0.868,
      "grad_norm": 0.781375321039837,
      "learning_rate": 6.032387143048853e-06,
      "loss": 0.231,
      "num_input_tokens_seen": 170151744,
      "step": 2170,
      "train_runtime": 28602.3354,
      "train_tokens_per_second": 5948.876
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.93859473763133,
      "learning_rate": 6.0170125956251935e-06,
      "loss": 0.2206,
      "num_input_tokens_seen": 170534368,
      "step": 2175,
      "train_runtime": 28668.7905,
      "train_tokens_per_second": 5948.433
    },
    {
      "epoch": 0.872,
      "grad_norm": 0.7865219290835191,
      "learning_rate": 6.0016280106978e-06,
      "loss": 0.2541,
      "num_input_tokens_seen": 170939264,
      "step": 2180,
      "train_runtime": 28739.4293,
      "train_tokens_per_second": 5947.9
    },
    {
      "epoch": 0.874,
      "grad_norm": 1.0981142367719585,
      "learning_rate": 5.986233540106315e-06,
      "loss": 0.2198,
      "num_input_tokens_seen": 171327728,
      "step": 2185,
      "train_runtime": 28810.52,
      "train_tokens_per_second": 5946.707
    },
    {
      "epoch": 0.876,
      "grad_norm": 0.6189848711597874,
      "learning_rate": 5.970829335787946e-06,
      "loss": 0.2446,
      "num_input_tokens_seen": 171683792,
      "step": 2190,
      "train_runtime": 28867.3719,
      "train_tokens_per_second": 5947.33
    },
    {
      "epoch": 0.878,
      "grad_norm": 0.8024094051926371,
      "learning_rate": 5.955415549775975e-06,
      "loss": 0.2166,
      "num_input_tokens_seen": 171980720,
      "step": 2195,
      "train_runtime": 28917.9366,
      "train_tokens_per_second": 5947.199
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.700283931520735,
      "learning_rate": 5.939992334198242e-06,
      "loss": 0.2486,
      "num_input_tokens_seen": 172406304,
      "step": 2200,
      "train_runtime": 28995.7249,
      "train_tokens_per_second": 5945.922
    },
    {
      "epoch": 0.882,
      "grad_norm": 0.6156776119471373,
      "learning_rate": 5.924559841275661e-06,
      "loss": 0.2502,
      "num_input_tokens_seen": 172786272,
      "step": 2205,
      "train_runtime": 29058.6595,
      "train_tokens_per_second": 5946.12
    },
    {
      "epoch": 0.884,
      "grad_norm": 0.7100376433954627,
      "learning_rate": 5.9091182233207075e-06,
      "loss": 0.2422,
      "num_input_tokens_seen": 173227264,
      "step": 2210,
      "train_runtime": 29131.5652,
      "train_tokens_per_second": 5946.377
    },
    {
      "epoch": 0.886,
      "grad_norm": 0.8147996007145993,
      "learning_rate": 5.893667632735915e-06,
      "loss": 0.2352,
      "num_input_tokens_seen": 173598624,
      "step": 2215,
      "train_runtime": 29202.1961,
      "train_tokens_per_second": 5944.711
    },
    {
      "epoch": 0.888,
      "grad_norm": 0.5311637083968614,
      "learning_rate": 5.878208222012377e-06,
      "loss": 0.2289,
      "num_input_tokens_seen": 173997648,
      "step": 2220,
      "train_runtime": 29260.9457,
      "train_tokens_per_second": 5946.412
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.6006361048911637,
      "learning_rate": 5.8627401437282334e-06,
      "loss": 0.2277,
      "num_input_tokens_seen": 174440288,
      "step": 2225,
      "train_runtime": 29338.854,
      "train_tokens_per_second": 5945.709
    },
    {
      "epoch": 0.892,
      "grad_norm": 0.5915073750833558,
      "learning_rate": 5.847263550547174e-06,
      "loss": 0.2295,
      "num_input_tokens_seen": 174846016,
      "step": 2230,
      "train_runtime": 29404.7298,
      "train_tokens_per_second": 5946.187
    },
    {
      "epoch": 0.894,
      "grad_norm": 0.8132516342886842,
      "learning_rate": 5.8317785952169245e-06,
      "loss": 0.2405,
      "num_input_tokens_seen": 175197376,
      "step": 2235,
      "train_runtime": 29465.5737,
      "train_tokens_per_second": 5945.833
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.9838730375194601,
      "learning_rate": 5.816285430567743e-06,
      "loss": 0.2464,
      "num_input_tokens_seen": 175543776,
      "step": 2240,
      "train_runtime": 29512.3984,
      "train_tokens_per_second": 5948.137
    },
    {
      "epoch": 0.898,
      "grad_norm": 0.5852209016193204,
      "learning_rate": 5.80078420951091e-06,
      "loss": 0.2315,
      "num_input_tokens_seen": 175959728,
      "step": 2245,
      "train_runtime": 29585.2542,
      "train_tokens_per_second": 5947.548
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.8236650897027592,
      "learning_rate": 5.785275085037218e-06,
      "loss": 0.2683,
      "num_input_tokens_seen": 176298768,
      "step": 2250,
      "train_runtime": 29641.8108,
      "train_tokens_per_second": 5947.638
    },
    {
      "epoch": 0.902,
      "grad_norm": 0.5611644053571633,
      "learning_rate": 5.769758210215466e-06,
      "loss": 0.2277,
      "num_input_tokens_seen": 176757936,
      "step": 2255,
      "train_runtime": 29716.4289,
      "train_tokens_per_second": 5948.155
    },
    {
      "epoch": 0.904,
      "grad_norm": 0.6457725745525457,
      "learning_rate": 5.754233738190942e-06,
      "loss": 0.2306,
      "num_input_tokens_seen": 177153168,
      "step": 2260,
      "train_runtime": 29779.1106,
      "train_tokens_per_second": 5948.907
    },
    {
      "epoch": 0.906,
      "grad_norm": 0.7742797303479264,
      "learning_rate": 5.7387018221839195e-06,
      "loss": 0.2447,
      "num_input_tokens_seen": 177558336,
      "step": 2265,
      "train_runtime": 29847.2103,
      "train_tokens_per_second": 5948.909
    },
    {
      "epoch": 0.908,
      "grad_norm": 0.6731925005917008,
      "learning_rate": 5.723162615488137e-06,
      "loss": 0.2224,
      "num_input_tokens_seen": 177970112,
      "step": 2270,
      "train_runtime": 29915.2381,
      "train_tokens_per_second": 5949.146
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.600224184362238,
      "learning_rate": 5.707616271469293e-06,
      "loss": 0.2213,
      "num_input_tokens_seen": 178365456,
      "step": 2275,
      "train_runtime": 29987.3778,
      "train_tokens_per_second": 5948.018
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.6578513225026819,
      "learning_rate": 5.692062943563525e-06,
      "loss": 0.2283,
      "num_input_tokens_seen": 178744496,
      "step": 2280,
      "train_runtime": 30060.8882,
      "train_tokens_per_second": 5946.082
    },
    {
      "epoch": 0.914,
      "grad_norm": 0.6546013647844027,
      "learning_rate": 5.6765027852759015e-06,
      "loss": 0.2273,
      "num_input_tokens_seen": 179089216,
      "step": 2285,
      "train_runtime": 30116.3719,
      "train_tokens_per_second": 5946.573
    },
    {
      "epoch": 0.916,
      "grad_norm": 0.6436125861923324,
      "learning_rate": 5.660935950178904e-06,
      "loss": 0.2114,
      "num_input_tokens_seen": 179463584,
      "step": 2290,
      "train_runtime": 30185.9557,
      "train_tokens_per_second": 5945.268
    },
    {
      "epoch": 0.918,
      "grad_norm": 1.1106079072044501,
      "learning_rate": 5.645362591910908e-06,
      "loss": 0.2317,
      "num_input_tokens_seen": 179834016,
      "step": 2295,
      "train_runtime": 30244.7067,
      "train_tokens_per_second": 5945.967
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.7625223505139128,
      "learning_rate": 5.629782864174672e-06,
      "loss": 0.2542,
      "num_input_tokens_seen": 180259712,
      "step": 2300,
      "train_runtime": 30314.1024,
      "train_tokens_per_second": 5946.398
    },
    {
      "epoch": 0.922,
      "grad_norm": 0.8050923063608859,
      "learning_rate": 5.614196920735822e-06,
      "loss": 0.253,
      "num_input_tokens_seen": 180769440,
      "step": 2305,
      "train_runtime": 30390.5223,
      "train_tokens_per_second": 5948.218
    },
    {
      "epoch": 0.924,
      "grad_norm": 0.838730652939852,
      "learning_rate": 5.598604915421324e-06,
      "loss": 0.2404,
      "num_input_tokens_seen": 181141040,
      "step": 2310,
      "train_runtime": 30446.9371,
      "train_tokens_per_second": 5949.401
    },
    {
      "epoch": 0.926,
      "grad_norm": 0.7311635967485809,
      "learning_rate": 5.5830070021179785e-06,
      "loss": 0.2475,
      "num_input_tokens_seen": 181512128,
      "step": 2315,
      "train_runtime": 30511.8838,
      "train_tokens_per_second": 5948.899
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.8333701668205385,
      "learning_rate": 5.567403334770891e-06,
      "loss": 0.241,
      "num_input_tokens_seen": 181898720,
      "step": 2320,
      "train_runtime": 30575.5824,
      "train_tokens_per_second": 5949.15
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.4982827092355415,
      "learning_rate": 5.551794067381959e-06,
      "loss": 0.2296,
      "num_input_tokens_seen": 182339792,
      "step": 2325,
      "train_runtime": 30653.1374,
      "train_tokens_per_second": 5948.487
    },
    {
      "epoch": 0.932,
      "grad_norm": 0.8970555784936441,
      "learning_rate": 5.536179354008351e-06,
      "loss": 0.2179,
      "num_input_tokens_seen": 182758928,
      "step": 2330,
      "train_runtime": 30726.7154,
      "train_tokens_per_second": 5947.884
    },
    {
      "epoch": 0.934,
      "grad_norm": 0.7859900791455534,
      "learning_rate": 5.520559348760984e-06,
      "loss": 0.2383,
      "num_input_tokens_seen": 183131360,
      "step": 2335,
      "train_runtime": 30796.714,
      "train_tokens_per_second": 5946.458
    },
    {
      "epoch": 0.936,
      "grad_norm": 0.9506801273249746,
      "learning_rate": 5.504934205803002e-06,
      "loss": 0.2529,
      "num_input_tokens_seen": 183537936,
      "step": 2340,
      "train_runtime": 30870.769,
      "train_tokens_per_second": 5945.363
    },
    {
      "epoch": 0.938,
      "grad_norm": 0.7170187918020063,
      "learning_rate": 5.489304079348259e-06,
      "loss": 0.2478,
      "num_input_tokens_seen": 183862896,
      "step": 2345,
      "train_runtime": 30933.3538,
      "train_tokens_per_second": 5943.84
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.6021121497604376,
      "learning_rate": 5.473669123659793e-06,
      "loss": 0.2532,
      "num_input_tokens_seen": 184259888,
      "step": 2350,
      "train_runtime": 30996.337,
      "train_tokens_per_second": 5944.57
    },
    {
      "epoch": 0.942,
      "grad_norm": 0.7051727487587609,
      "learning_rate": 5.458029493048303e-06,
      "loss": 0.2123,
      "num_input_tokens_seen": 184656352,
      "step": 2355,
      "train_runtime": 31067.0032,
      "train_tokens_per_second": 5943.81
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.6635146242890422,
      "learning_rate": 5.442385341870633e-06,
      "loss": 0.2118,
      "num_input_tokens_seen": 185019024,
      "step": 2360,
      "train_runtime": 31131.4898,
      "train_tokens_per_second": 5943.147
    },
    {
      "epoch": 0.946,
      "grad_norm": 0.6724672853288053,
      "learning_rate": 5.426736824528236e-06,
      "loss": 0.2506,
      "num_input_tokens_seen": 185474944,
      "step": 2365,
      "train_runtime": 31205.1483,
      "train_tokens_per_second": 5943.729
    },
    {
      "epoch": 0.948,
      "grad_norm": 0.7639385973705092,
      "learning_rate": 5.411084095465661e-06,
      "loss": 0.253,
      "num_input_tokens_seen": 185960416,
      "step": 2370,
      "train_runtime": 31275.112,
      "train_tokens_per_second": 5945.955
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.6731404031064729,
      "learning_rate": 5.3954273091690245e-06,
      "loss": 0.2305,
      "num_input_tokens_seen": 186388384,
      "step": 2375,
      "train_runtime": 31345.6404,
      "train_tokens_per_second": 5946.23
    },
    {
      "epoch": 0.952,
      "grad_norm": 0.7944137119655447,
      "learning_rate": 5.379766620164488e-06,
      "loss": 0.242,
      "num_input_tokens_seen": 186822368,
      "step": 2380,
      "train_runtime": 31414.4063,
      "train_tokens_per_second": 5947.028
    },
    {
      "epoch": 0.954,
      "grad_norm": 0.7009221618788651,
      "learning_rate": 5.36410218301673e-06,
      "loss": 0.231,
      "num_input_tokens_seen": 187241824,
      "step": 2385,
      "train_runtime": 31487.9223,
      "train_tokens_per_second": 5946.465
    },
    {
      "epoch": 0.956,
      "grad_norm": 0.7826192747499381,
      "learning_rate": 5.348434152327418e-06,
      "loss": 0.2367,
      "num_input_tokens_seen": 187634528,
      "step": 2390,
      "train_runtime": 31557.8152,
      "train_tokens_per_second": 5945.739
    },
    {
      "epoch": 0.958,
      "grad_norm": 0.8568472463615774,
      "learning_rate": 5.3327626827336906e-06,
      "loss": 0.23,
      "num_input_tokens_seen": 188055872,
      "step": 2395,
      "train_runtime": 31621.8896,
      "train_tokens_per_second": 5947.016
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.9212451337260199,
      "learning_rate": 5.3170879289066265e-06,
      "loss": 0.2423,
      "num_input_tokens_seen": 188417040,
      "step": 2400,
      "train_runtime": 31681.7657,
      "train_tokens_per_second": 5947.176
    },
    {
      "epoch": 0.962,
      "grad_norm": 0.6202368278565747,
      "learning_rate": 5.301410045549719e-06,
      "loss": 0.2283,
      "num_input_tokens_seen": 188757392,
      "step": 2405,
      "train_runtime": 31735.9795,
      "train_tokens_per_second": 5947.741
    },
    {
      "epoch": 0.964,
      "grad_norm": 0.8197217478906974,
      "learning_rate": 5.285729187397344e-06,
      "loss": 0.2321,
      "num_input_tokens_seen": 189109872,
      "step": 2410,
      "train_runtime": 31803.2473,
      "train_tokens_per_second": 5946.244
    },
    {
      "epoch": 0.966,
      "grad_norm": 0.8149889759430892,
      "learning_rate": 5.270045509213244e-06,
      "loss": 0.2244,
      "num_input_tokens_seen": 189461472,
      "step": 2415,
      "train_runtime": 31871.1936,
      "train_tokens_per_second": 5944.599
    },
    {
      "epoch": 0.968,
      "grad_norm": 0.6823638505459694,
      "learning_rate": 5.25435916578899e-06,
      "loss": 0.2451,
      "num_input_tokens_seen": 189885568,
      "step": 2420,
      "train_runtime": 31945.3988,
      "train_tokens_per_second": 5944.066
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.6838081564497045,
      "learning_rate": 5.238670311942459e-06,
      "loss": 0.2443,
      "num_input_tokens_seen": 190242304,
      "step": 2425,
      "train_runtime": 31993.066,
      "train_tokens_per_second": 5946.361
    },
    {
      "epoch": 0.972,
      "grad_norm": 0.6769627193812837,
      "learning_rate": 5.222979102516304e-06,
      "loss": 0.2422,
      "num_input_tokens_seen": 190598224,
      "step": 2430,
      "train_runtime": 32054.5311,
      "train_tokens_per_second": 5946.062
    },
    {
      "epoch": 0.974,
      "grad_norm": 0.8137418632037458,
      "learning_rate": 5.207285692376427e-06,
      "loss": 0.2372,
      "num_input_tokens_seen": 190981888,
      "step": 2435,
      "train_runtime": 32120.4466,
      "train_tokens_per_second": 5945.804
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.8709294531660275,
      "learning_rate": 5.191590236410451e-06,
      "loss": 0.2312,
      "num_input_tokens_seen": 191325504,
      "step": 2440,
      "train_runtime": 32190.2973,
      "train_tokens_per_second": 5943.577
    },
    {
      "epoch": 0.978,
      "grad_norm": 0.8141333292380133,
      "learning_rate": 5.175892889526189e-06,
      "loss": 0.2086,
      "num_input_tokens_seen": 191691824,
      "step": 2445,
      "train_runtime": 32250.6545,
      "train_tokens_per_second": 5943.812
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.6568883957821576,
      "learning_rate": 5.16019380665012e-06,
      "loss": 0.2279,
      "num_input_tokens_seen": 192164848,
      "step": 2450,
      "train_runtime": 32329.2932,
      "train_tokens_per_second": 5943.985
    },
    {
      "epoch": 0.982,
      "grad_norm": 0.8113314122650447,
      "learning_rate": 5.144493142725851e-06,
      "loss": 0.2391,
      "num_input_tokens_seen": 192517376,
      "step": 2455,
      "train_runtime": 32392.0021,
      "train_tokens_per_second": 5943.361
    },
    {
      "epoch": 0.984,
      "grad_norm": 0.8866398589116381,
      "learning_rate": 5.128791052712597e-06,
      "loss": 0.236,
      "num_input_tokens_seen": 192975344,
      "step": 2460,
      "train_runtime": 32467.0335,
      "train_tokens_per_second": 5943.732
    },
    {
      "epoch": 0.986,
      "grad_norm": 0.6305031077039938,
      "learning_rate": 5.1130876915836495e-06,
      "loss": 0.2355,
      "num_input_tokens_seen": 193448704,
      "step": 2465,
      "train_runtime": 32543.6576,
      "train_tokens_per_second": 5944.283
    },
    {
      "epoch": 0.988,
      "grad_norm": 0.7628934806649452,
      "learning_rate": 5.0973832143248405e-06,
      "loss": 0.2254,
      "num_input_tokens_seen": 193822592,
      "step": 2470,
      "train_runtime": 32598.0646,
      "train_tokens_per_second": 5945.831
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.70672706494417,
      "learning_rate": 5.0816777759330215e-06,
      "loss": 0.2537,
      "num_input_tokens_seen": 194232336,
      "step": 2475,
      "train_runtime": 32670.2703,
      "train_tokens_per_second": 5945.232
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.7385436359837361,
      "learning_rate": 5.065971531414528e-06,
      "loss": 0.2267,
      "num_input_tokens_seen": 194573072,
      "step": 2480,
      "train_runtime": 32732.6925,
      "train_tokens_per_second": 5944.304
    },
    {
      "epoch": 0.994,
      "grad_norm": 0.8708620218303826,
      "learning_rate": 5.050264635783654e-06,
      "loss": 0.2445,
      "num_input_tokens_seen": 194995520,
      "step": 2485,
      "train_runtime": 32808.6939,
      "train_tokens_per_second": 5943.41
    },
    {
      "epoch": 0.996,
      "grad_norm": 0.8837777853923816,
      "learning_rate": 5.034557244061117e-06,
      "loss": 0.2221,
      "num_input_tokens_seen": 195391968,
      "step": 2490,
      "train_runtime": 32866.1399,
      "train_tokens_per_second": 5945.084
    },
    {
      "epoch": 0.998,
      "grad_norm": 2.625974706230731,
      "learning_rate": 5.018849511272532e-06,
      "loss": 0.2541,
      "num_input_tokens_seen": 195720288,
      "step": 2495,
      "train_runtime": 32917.1453,
      "train_tokens_per_second": 5945.846
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.7009528907653958,
      "learning_rate": 5.0031415924468816e-06,
      "loss": 0.2342,
      "num_input_tokens_seen": 196169248,
      "step": 2500,
      "train_runtime": 32991.3291,
      "train_tokens_per_second": 5946.085
    }
  ],
  "logging_steps": 5,
  "max_steps": 5000,
  "num_input_tokens_seen": 196169248,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 362784421937152.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}