{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.581755593803787,
  "eval_steps": 500,
  "global_step": 12000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "grad_norm": 1.281562089920044,
      "learning_rate": 4e-05,
      "loss": 0.5044,
      "step": 100
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.3552560806274414,
      "learning_rate": 8e-05,
      "loss": 0.4867,
      "step": 200
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8318453431129456,
      "learning_rate": 0.00012,
      "loss": 0.4359,
      "step": 300
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.3478361368179321,
      "learning_rate": 0.00016,
      "loss": 0.4475,
      "step": 400
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.1437219381332397,
      "learning_rate": 0.0002,
      "loss": 0.4524,
      "step": 500
    },
    {
      "epoch": 0.11,
      "eval_loss": 0.3725915253162384,
      "eval_runtime": 90.4733,
      "eval_samples_per_second": 11.053,
      "eval_steps_per_second": 2.763,
      "step": 500
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.8374431133270264,
      "learning_rate": 0.00019897435897435898,
      "loss": 0.4569,
      "step": 600
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6386727094650269,
      "learning_rate": 0.00019794871794871796,
      "loss": 0.4584,
      "step": 700
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.4089267253875732,
      "learning_rate": 0.00019692307692307696,
      "loss": 0.4625,
      "step": 800
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.269509792327881,
      "learning_rate": 0.0001958974358974359,
      "loss": 0.464,
      "step": 900
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0896354913711548,
      "learning_rate": 0.00019487179487179487,
      "loss": 0.4807,
      "step": 1000
    },
    {
      "epoch": 0.22,
      "eval_loss": 0.3314511179924011,
      "eval_runtime": 90.3867,
      "eval_samples_per_second": 11.064,
      "eval_steps_per_second": 2.766,
      "step": 1000
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0933148860931396,
      "learning_rate": 0.00019384615384615385,
      "loss": 0.4032,
      "step": 1100
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.555145740509033,
      "learning_rate": 0.00019282051282051282,
      "loss": 0.4353,
      "step": 1200
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.245523452758789,
      "learning_rate": 0.00019179487179487182,
      "loss": 0.409,
      "step": 1300
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.084887981414795,
      "learning_rate": 0.0001907692307692308,
      "loss": 0.4214,
      "step": 1400
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.2279912233352661,
      "learning_rate": 0.00018974358974358974,
      "loss": 0.4507,
      "step": 1500
    },
    {
      "epoch": 0.32,
      "eval_loss": 0.312484472990036,
      "eval_runtime": 90.4149,
      "eval_samples_per_second": 11.06,
      "eval_steps_per_second": 2.765,
      "step": 1500
    },
    {
      "epoch": 0.34,
      "grad_norm": 4.321805477142334,
      "learning_rate": 0.0001887179487179487,
      "loss": 0.4523,
      "step": 1600
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.376891016960144,
      "learning_rate": 0.0001876923076923077,
      "loss": 0.441,
      "step": 1700
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.077620267868042,
      "learning_rate": 0.0001866666666666667,
      "loss": 0.4392,
      "step": 1800
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.546964168548584,
      "learning_rate": 0.00018564102564102566,
      "loss": 0.4749,
      "step": 1900
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.035524845123291,
      "learning_rate": 0.00018461538461538463,
      "loss": 0.4116,
      "step": 2000
    },
    {
      "epoch": 0.43,
      "eval_loss": 0.2960582375526428,
      "eval_runtime": 90.8091,
      "eval_samples_per_second": 11.012,
      "eval_steps_per_second": 2.753,
      "step": 2000
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.3167320489883423,
      "learning_rate": 0.00018358974358974358,
      "loss": 0.4135,
      "step": 2100
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.908482074737549,
      "learning_rate": 0.00018256410256410258,
      "loss": 0.4092,
      "step": 2200
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.8161234855651855,
      "learning_rate": 0.00018153846153846155,
      "loss": 0.4364,
      "step": 2300
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.0435675382614136,
      "learning_rate": 0.00018051282051282052,
      "loss": 0.4258,
      "step": 2400
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.7152374982833862,
      "learning_rate": 0.0001794871794871795,
      "loss": 0.4208,
      "step": 2500
    },
    {
      "epoch": 0.54,
      "eval_loss": 0.3006269633769989,
      "eval_runtime": 90.5246,
      "eval_samples_per_second": 11.047,
      "eval_steps_per_second": 2.762,
      "step": 2500
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.79283607006073,
      "learning_rate": 0.00017846153846153847,
      "loss": 0.4209,
      "step": 2600
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5817480087280273,
      "learning_rate": 0.00017743589743589744,
      "loss": 0.415,
      "step": 2700
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.306807279586792,
      "learning_rate": 0.00017641025641025642,
      "loss": 0.4231,
      "step": 2800
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.0432254076004028,
      "learning_rate": 0.0001753846153846154,
      "loss": 0.38,
      "step": 2900
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.0641353130340576,
      "learning_rate": 0.00017435897435897436,
      "loss": 0.4671,
      "step": 3000
    },
    {
      "epoch": 0.65,
      "eval_loss": 0.3091895282268524,
      "eval_runtime": 90.8377,
      "eval_samples_per_second": 11.009,
      "eval_steps_per_second": 2.752,
      "step": 3000
    },
    {
      "epoch": 0.67,
      "grad_norm": 2.6519484519958496,
      "learning_rate": 0.00017333333333333334,
      "loss": 0.4148,
      "step": 3100
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.851606547832489,
      "learning_rate": 0.00017230769230769234,
      "loss": 0.37,
      "step": 3200
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.1799293756484985,
      "learning_rate": 0.0001712923076923077,
      "loss": 0.3687,
      "step": 3300
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.2787513732910156,
      "learning_rate": 0.0001702666666666667,
      "loss": 0.3997,
      "step": 3400
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.2793489694595337,
      "learning_rate": 0.00016924102564102564,
      "loss": 0.3936,
      "step": 3500
    },
    {
      "epoch": 0.75,
      "eval_loss": 0.2866547703742981,
      "eval_runtime": 90.8477,
      "eval_samples_per_second": 11.007,
      "eval_steps_per_second": 2.752,
      "step": 3500
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.7189435362815857,
      "learning_rate": 0.0001682153846153846,
      "loss": 0.4191,
      "step": 3600
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9801158308982849,
      "learning_rate": 0.0001671897435897436,
      "loss": 0.3978,
      "step": 3700
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.41176438331604,
      "learning_rate": 0.0001661641025641026,
      "loss": 0.3762,
      "step": 3800
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.9064908027648926,
      "learning_rate": 0.00016513846153846156,
      "loss": 0.3746,
      "step": 3900
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.3439726829528809,
      "learning_rate": 0.00016411282051282053,
      "loss": 0.3823,
      "step": 4000
    },
    {
      "epoch": 0.86,
      "eval_loss": 0.2893534004688263,
      "eval_runtime": 90.8714,
      "eval_samples_per_second": 11.005,
      "eval_steps_per_second": 2.751,
      "step": 4000
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6646331548690796,
      "learning_rate": 0.00016308717948717948,
      "loss": 0.4029,
      "step": 4100
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.9874048829078674,
      "learning_rate": 0.00016206153846153845,
      "loss": 0.3933,
      "step": 4200
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.0010713338851929,
      "learning_rate": 0.00016103589743589745,
      "loss": 0.41,
      "step": 4300
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.9607357978820801,
      "learning_rate": 0.00016001025641025642,
      "loss": 0.3659,
      "step": 4400
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.3652217388153076,
      "learning_rate": 0.0001589846153846154,
      "loss": 0.3794,
      "step": 4500
    },
    {
      "epoch": 0.97,
      "eval_loss": 0.2667659819126129,
      "eval_runtime": 90.9201,
      "eval_samples_per_second": 10.999,
      "eval_steps_per_second": 2.75,
      "step": 4500
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.056810736656189,
      "learning_rate": 0.00015795897435897437,
      "loss": 0.3878,
      "step": 4600
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.7765600085258484,
      "learning_rate": 0.00015693333333333334,
      "loss": 0.3396,
      "step": 4700
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.0664465427398682,
      "learning_rate": 0.00015590769230769232,
      "loss": 0.2813,
      "step": 4800
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.6342141628265381,
      "learning_rate": 0.0001548820512820513,
      "loss": 0.309,
      "step": 4900
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.1380507946014404,
      "learning_rate": 0.00015385641025641026,
      "loss": 0.2888,
      "step": 5000
    },
    {
      "epoch": 1.08,
      "eval_loss": 0.27630186080932617,
      "eval_runtime": 90.9762,
      "eval_samples_per_second": 10.992,
      "eval_steps_per_second": 2.748,
      "step": 5000
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.6043438911437988,
      "learning_rate": 0.00015283076923076924,
      "loss": 0.3836,
      "step": 5100
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.2961331307888031,
      "learning_rate": 0.0001518051282051282,
      "loss": 0.2964,
      "step": 5200
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.0106085538864136,
      "learning_rate": 0.00015077948717948718,
      "loss": 0.2837,
      "step": 5300
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.2187350988388062,
      "learning_rate": 0.00014975384615384616,
      "loss": 0.3093,
      "step": 5400
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.4018324613571167,
      "learning_rate": 0.00014872820512820513,
      "loss": 0.3029,
      "step": 5500
    },
    {
      "epoch": 1.18,
      "eval_loss": 0.2620677351951599,
      "eval_runtime": 90.8592,
      "eval_samples_per_second": 11.006,
      "eval_steps_per_second": 2.752,
      "step": 5500
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.8634820580482483,
      "learning_rate": 0.00014771282051282051,
      "loss": 0.3046,
      "step": 5600
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.6309552788734436,
      "learning_rate": 0.0001466871794871795,
      "loss": 0.3041,
      "step": 5700
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.2985124588012695,
      "learning_rate": 0.00014566153846153846,
      "loss": 0.312,
      "step": 5800
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.9580160975456238,
      "learning_rate": 0.00014463589743589746,
      "loss": 0.3306,
      "step": 5900
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.0316152572631836,
      "learning_rate": 0.00014361025641025643,
      "loss": 0.2979,
      "step": 6000
    },
    {
      "epoch": 1.29,
      "eval_loss": 0.26443469524383545,
      "eval_runtime": 90.9026,
      "eval_samples_per_second": 11.001,
      "eval_steps_per_second": 2.75,
      "step": 6000
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.9922102689743042,
      "learning_rate": 0.00014258461538461538,
      "loss": 0.3105,
      "step": 6100
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.274574875831604,
      "learning_rate": 0.00014155897435897435,
      "loss": 0.3174,
      "step": 6200
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.1152174472808838,
      "learning_rate": 0.00014053333333333335,
      "loss": 0.3094,
      "step": 6300
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.4428844451904297,
      "learning_rate": 0.00013950769230769233,
      "loss": 0.3136,
      "step": 6400
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.418609380722046,
      "learning_rate": 0.0001384820512820513,
      "loss": 0.3425,
      "step": 6500
    },
    {
      "epoch": 1.4,
      "eval_loss": 0.299164742231369,
      "eval_runtime": 90.9837,
      "eval_samples_per_second": 10.991,
      "eval_steps_per_second": 2.748,
      "step": 6500
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.309278130531311,
      "learning_rate": 0.00013745641025641027,
      "loss": 0.3318,
      "step": 6600
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.0984652042388916,
      "learning_rate": 0.00013643076923076922,
      "loss": 0.3221,
      "step": 6700
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.1776598691940308,
      "learning_rate": 0.00013540512820512822,
      "loss": 0.3151,
      "step": 6800
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.1536751985549927,
      "learning_rate": 0.0001343794871794872,
      "loss": 0.3471,
      "step": 6900
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.5731265544891357,
      "learning_rate": 0.00013335384615384616,
      "loss": 0.3342,
      "step": 7000
    },
    {
      "epoch": 1.51,
      "eval_loss": 0.25467580556869507,
      "eval_runtime": 90.9083,
      "eval_samples_per_second": 11.0,
      "eval_steps_per_second": 2.75,
      "step": 7000
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.5424981117248535,
      "learning_rate": 0.00013233846153846155,
      "loss": 0.3337,
      "step": 7100
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.2443273067474365,
      "learning_rate": 0.00013131282051282052,
      "loss": 0.2955,
      "step": 7200
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.7886701822280884,
      "learning_rate": 0.0001302871794871795,
      "loss": 0.3017,
      "step": 7300
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.6013621091842651,
      "learning_rate": 0.00012926153846153847,
      "loss": 0.2976,
      "step": 7400
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.495753288269043,
      "learning_rate": 0.00012823589743589744,
      "loss": 0.3086,
      "step": 7500
    },
    {
      "epoch": 1.61,
      "eval_loss": 0.25845155119895935,
      "eval_runtime": 90.8943,
      "eval_samples_per_second": 11.002,
      "eval_steps_per_second": 2.75,
      "step": 7500
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.181015968322754,
      "learning_rate": 0.00012721025641025641,
      "loss": 0.2657,
      "step": 7600
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.2074403762817383,
      "learning_rate": 0.0001261846153846154,
      "loss": 0.3149,
      "step": 7700
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.216676950454712,
      "learning_rate": 0.00012515897435897436,
      "loss": 0.2999,
      "step": 7800
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.8093730211257935,
      "learning_rate": 0.00012413333333333333,
      "loss": 0.3144,
      "step": 7900
    },
    {
      "epoch": 1.72,
      "grad_norm": 3.7325637340545654,
      "learning_rate": 0.00012310769230769233,
      "loss": 0.3326,
      "step": 8000
    },
    {
      "epoch": 1.72,
      "eval_loss": 0.23783066868782043,
      "eval_runtime": 90.8877,
      "eval_samples_per_second": 11.003,
      "eval_steps_per_second": 2.751,
      "step": 8000
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.6969020366668701,
      "learning_rate": 0.00012208205128205128,
      "loss": 0.2935,
      "step": 8100
    },
    {
      "epoch": 1.76,
      "grad_norm": 2.1927125453948975,
      "learning_rate": 0.00012105641025641025,
      "loss": 0.3091,
      "step": 8200
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.8521186113357544,
      "learning_rate": 0.00012003076923076924,
      "loss": 0.2517,
      "step": 8300
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.5349504947662354,
      "learning_rate": 0.00011900512820512821,
      "loss": 0.2794,
      "step": 8400
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.6325456500053406,
      "learning_rate": 0.00011797948717948718,
      "loss": 0.2912,
      "step": 8500
    },
    {
      "epoch": 1.83,
      "eval_loss": 0.23375801742076874,
      "eval_runtime": 90.8478,
      "eval_samples_per_second": 11.007,
      "eval_steps_per_second": 2.752,
      "step": 8500
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.615515112876892,
      "learning_rate": 0.00011695384615384617,
      "loss": 0.2953,
      "step": 8600
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.2424674034118652,
      "learning_rate": 0.00011592820512820513,
      "loss": 0.3047,
      "step": 8700
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.2125675678253174,
      "learning_rate": 0.0001149025641025641,
      "loss": 0.3209,
      "step": 8800
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.6464908123016357,
      "learning_rate": 0.00011387692307692308,
      "loss": 0.3214,
      "step": 8900
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.300310730934143,
      "learning_rate": 0.00011285128205128206,
      "loss": 0.2965,
      "step": 9000
    },
    {
      "epoch": 1.94,
      "eval_loss": 0.2334287166595459,
      "eval_runtime": 90.9451,
      "eval_samples_per_second": 10.996,
      "eval_steps_per_second": 2.749,
      "step": 9000
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.7211843729019165,
      "learning_rate": 0.00011182564102564104,
      "loss": 0.2813,
      "step": 9100
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.5777404308319092,
      "learning_rate": 0.00011080000000000001,
      "loss": 0.2951,
      "step": 9200
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.3048077821731567,
      "learning_rate": 0.00010977435897435897,
      "loss": 0.301,
      "step": 9300
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.1753878593444824,
      "learning_rate": 0.00010874871794871794,
      "loss": 0.1944,
      "step": 9400
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.8378590941429138,
      "learning_rate": 0.00010772307692307693,
      "loss": 0.2041,
      "step": 9500
    },
    {
      "epoch": 2.04,
      "eval_loss": 0.2733120024204254,
      "eval_runtime": 90.9249,
      "eval_samples_per_second": 10.998,
      "eval_steps_per_second": 2.75,
      "step": 9500
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.4123268127441406,
      "learning_rate": 0.0001066974358974359,
      "loss": 0.2206,
      "step": 9600
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.5923714637756348,
      "learning_rate": 0.00010567179487179489,
      "loss": 0.2331,
      "step": 9700
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.31000018119812,
      "learning_rate": 0.00010465641025641026,
      "loss": 0.2564,
      "step": 9800
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.4770272970199585,
      "learning_rate": 0.00010363076923076925,
      "loss": 0.2009,
      "step": 9900
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.5393586158752441,
      "learning_rate": 0.00010260512820512822,
      "loss": 0.2168,
      "step": 10000
    },
    {
      "epoch": 2.15,
      "eval_loss": 0.24773281812667847,
      "eval_runtime": 90.901,
      "eval_samples_per_second": 11.001,
      "eval_steps_per_second": 2.75,
      "step": 10000
    },
    {
      "epoch": 2.17,
      "grad_norm": 2.3399605751037598,
      "learning_rate": 0.00010157948717948718,
      "loss": 0.2275,
      "step": 10100
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.417143702507019,
      "learning_rate": 0.00010055384615384615,
      "loss": 0.2093,
      "step": 10200
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.7041810750961304,
      "learning_rate": 9.952820512820513e-05,
      "loss": 0.2105,
      "step": 10300
    },
    {
      "epoch": 2.24,
      "grad_norm": 3.7192060947418213,
      "learning_rate": 9.850256410256411e-05,
      "loss": 0.2078,
      "step": 10400
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.7868184447288513,
      "learning_rate": 9.747692307692307e-05,
      "loss": 0.2058,
      "step": 10500
    },
    {
      "epoch": 2.26,
      "eval_loss": 0.22978341579437256,
      "eval_runtime": 90.9187,
      "eval_samples_per_second": 10.999,
      "eval_steps_per_second": 2.75,
      "step": 10500
    },
    {
      "epoch": 2.28,
      "grad_norm": 2.572187662124634,
      "learning_rate": 9.645128205128206e-05,
      "loss": 0.2304,
      "step": 10600
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.3671247959136963,
      "learning_rate": 9.542564102564103e-05,
      "loss": 0.1962,
      "step": 10700
    },
    {
      "epoch": 2.32,
      "grad_norm": 2.6237735748291016,
      "learning_rate": 9.44e-05,
      "loss": 0.2201,
      "step": 10800
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.1776219606399536,
      "learning_rate": 9.337435897435898e-05,
      "loss": 0.1972,
      "step": 10900
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.236425757408142,
      "learning_rate": 9.234871794871795e-05,
      "loss": 0.2126,
      "step": 11000
    },
    {
      "epoch": 2.37,
      "eval_loss": 0.24023191630840302,
      "eval_runtime": 90.9406,
      "eval_samples_per_second": 10.996,
      "eval_steps_per_second": 2.749,
      "step": 11000
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.0826618671417236,
      "learning_rate": 9.132307692307692e-05,
      "loss": 0.2168,
      "step": 11100
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.8385189771652222,
      "learning_rate": 9.02974358974359e-05,
      "loss": 0.2029,
      "step": 11200
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.7595863342285156,
      "learning_rate": 8.927179487179488e-05,
      "loss": 0.1902,
      "step": 11300
    },
    {
      "epoch": 2.45,
      "grad_norm": 2.0246148109436035,
      "learning_rate": 8.824615384615384e-05,
      "loss": 0.2276,
      "step": 11400
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.2247196435928345,
      "learning_rate": 8.722051282051283e-05,
      "loss": 0.2,
      "step": 11500
    },
    {
      "epoch": 2.47,
      "eval_loss": 0.25648975372314453,
      "eval_runtime": 90.8579,
      "eval_samples_per_second": 11.006,
      "eval_steps_per_second": 2.752,
      "step": 11500
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.0143483877182007,
      "learning_rate": 8.61948717948718e-05,
      "loss": 0.2179,
      "step": 11600
    },
    {
      "epoch": 2.52,
      "grad_norm": 2.7020885944366455,
      "learning_rate": 8.516923076923076e-05,
      "loss": 0.2214,
      "step": 11700
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.8533117771148682,
      "learning_rate": 8.414358974358975e-05,
      "loss": 0.2074,
      "step": 11800
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.7365753650665283,
      "learning_rate": 8.311794871794872e-05,
      "loss": 0.2396,
      "step": 11900
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.8160982131958008,
      "learning_rate": 8.209230769230771e-05,
      "loss": 0.1786,
      "step": 12000
    },
    {
      "epoch": 2.58,
      "eval_loss": 0.2251870185136795,
      "eval_runtime": 90.8167,
      "eval_samples_per_second": 11.011,
      "eval_steps_per_second": 2.753,
      "step": 12000
    }
  ],
  "logging_steps": 100,
  "max_steps": 20000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 3.86418819603628e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}