{
  "best_metric": 0.3029824197292328,
  "best_model_checkpoint": "./convnext-base-3e-5-batch-8/checkpoint-8792",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 21980,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05,
      "grad_norm": 20.464111328125,
      "learning_rate": 2.999846786074732e-05,
      "loss": 2.7431,
      "step": 100
    },
    {
      "epoch": 0.09,
      "grad_norm": 26.436141967773438,
      "learning_rate": 2.999387175598269e-05,
      "loss": 1.6104,
      "step": 200
    },
    {
      "epoch": 0.14,
      "grad_norm": 54.28024673461914,
      "learning_rate": 2.998621262462245e-05,
      "loss": 1.2012,
      "step": 300
    },
    {
      "epoch": 0.18,
      "grad_norm": 53.153194427490234,
      "learning_rate": 2.9975492031314045e-05,
      "loss": 0.974,
      "step": 400
    },
    {
      "epoch": 0.23,
      "grad_norm": 38.65731430053711,
      "learning_rate": 2.996171216611638e-05,
      "loss": 0.8435,
      "step": 500
    },
    {
      "epoch": 0.27,
      "grad_norm": 23.02427101135254,
      "learning_rate": 2.994487584405244e-05,
      "loss": 0.7449,
      "step": 600
    },
    {
      "epoch": 0.32,
      "grad_norm": 19.776226043701172,
      "learning_rate": 2.992498650453421e-05,
      "loss": 0.6626,
      "step": 700
    },
    {
      "epoch": 0.36,
      "grad_norm": 19.967565536499023,
      "learning_rate": 2.990204821066006e-05,
      "loss": 0.5485,
      "step": 800
    },
    {
      "epoch": 0.41,
      "grad_norm": 32.89971923828125,
      "learning_rate": 2.9876065648384715e-05,
      "loss": 0.6233,
      "step": 900
    },
    {
      "epoch": 0.45,
      "grad_norm": 23.158945083618164,
      "learning_rate": 2.984704412556199e-05,
      "loss": 0.595,
      "step": 1000
    },
    {
      "epoch": 0.5,
      "grad_norm": 33.01531219482422,
      "learning_rate": 2.981498957086044e-05,
      "loss": 0.5844,
      "step": 1100
    },
    {
      "epoch": 0.55,
      "grad_norm": 36.02910614013672,
      "learning_rate": 2.977990853255228e-05,
      "loss": 0.4787,
      "step": 1200
    },
    {
      "epoch": 0.59,
      "grad_norm": 25.670297622680664,
      "learning_rate": 2.974180817717561e-05,
      "loss": 0.6183,
      "step": 1300
    },
    {
      "epoch": 0.64,
      "grad_norm": 16.535247802734375,
      "learning_rate": 2.970069628807043e-05,
      "loss": 0.509,
      "step": 1400
    },
    {
      "epoch": 0.68,
      "grad_norm": 20.26650047302246,
      "learning_rate": 2.965658126378862e-05,
      "loss": 0.5619,
      "step": 1500
    },
    {
      "epoch": 0.73,
      "grad_norm": 51.08666229248047,
      "learning_rate": 2.9609472116378222e-05,
      "loss": 0.4814,
      "step": 1600
    },
    {
      "epoch": 0.77,
      "grad_norm": 13.978100776672363,
      "learning_rate": 2.955937846954242e-05,
      "loss": 0.4695,
      "step": 1700
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.855776071548462,
      "learning_rate": 2.9506310556673573e-05,
      "loss": 0.5061,
      "step": 1800
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.468057155609131,
      "learning_rate": 2.945027921876265e-05,
      "loss": 0.5227,
      "step": 1900
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.113645315170288,
      "learning_rate": 2.9391295902184625e-05,
      "loss": 0.4312,
      "step": 2000
    },
    {
      "epoch": 0.96,
      "grad_norm": 37.98606872558594,
      "learning_rate": 2.93293726563601e-05,
      "loss": 0.5851,
      "step": 2100
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8918489065606362,
      "eval_loss": 0.38079071044921875,
      "eval_runtime": 114.5109,
      "eval_samples_per_second": 21.963,
      "eval_steps_per_second": 2.751,
      "step": 2198
    },
    {
      "epoch": 1.0,
      "grad_norm": 20.0971622467041,
      "learning_rate": 2.9264522131293818e-05,
      "loss": 0.5127,
      "step": 2200
    },
    {
      "epoch": 1.05,
      "grad_norm": 26.824159622192383,
      "learning_rate": 2.919675757499045e-05,
      "loss": 0.3676,
      "step": 2300
    },
    {
      "epoch": 1.09,
      "grad_norm": 14.176966667175293,
      "learning_rate": 2.9126092830748217e-05,
      "loss": 0.3944,
      "step": 2400
    },
    {
      "epoch": 1.14,
      "grad_norm": 31.566120147705078,
      "learning_rate": 2.9052542334330916e-05,
      "loss": 0.3678,
      "step": 2500
    },
    {
      "epoch": 1.18,
      "grad_norm": 30.576416015625,
      "learning_rate": 2.897612111101888e-05,
      "loss": 0.465,
      "step": 2600
    },
    {
      "epoch": 1.23,
      "grad_norm": 52.94912338256836,
      "learning_rate": 2.889684477253959e-05,
      "loss": 0.3741,
      "step": 2700
    },
    {
      "epoch": 1.27,
      "grad_norm": 26.88067054748535,
      "learning_rate": 2.8814729513878365e-05,
      "loss": 0.3993,
      "step": 2800
    },
    {
      "epoch": 1.32,
      "grad_norm": 26.82840919494629,
      "learning_rate": 2.8729792109970015e-05,
      "loss": 0.3801,
      "step": 2900
    },
    {
      "epoch": 1.36,
      "grad_norm": 52.661399841308594,
      "learning_rate": 2.864204991227195e-05,
      "loss": 0.326,
      "step": 3000
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.977348327636719,
      "learning_rate": 2.855152084521953e-05,
      "loss": 0.3646,
      "step": 3100
    },
    {
      "epoch": 1.46,
      "grad_norm": 28.014543533325195,
      "learning_rate": 2.8458223402564366e-05,
      "loss": 0.3409,
      "step": 3200
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.3806567192077637,
      "learning_rate": 2.836217664359634e-05,
      "loss": 0.467,
      "step": 3300
    },
    {
      "epoch": 1.55,
      "grad_norm": 33.71268844604492,
      "learning_rate": 2.826340018925006e-05,
      "loss": 0.4337,
      "step": 3400
    },
    {
      "epoch": 1.59,
      "grad_norm": 34.983802795410156,
      "learning_rate": 2.8161914218096568e-05,
      "loss": 0.3762,
      "step": 3500
    },
    {
      "epoch": 1.64,
      "grad_norm": 23.66938591003418,
      "learning_rate": 2.8057739462221215e-05,
      "loss": 0.4465,
      "step": 3600
    },
    {
      "epoch": 1.68,
      "grad_norm": 16.07572364807129,
      "learning_rate": 2.7950897202988338e-05,
      "loss": 0.3359,
      "step": 3700
    },
    {
      "epoch": 1.73,
      "grad_norm": 25.094499588012695,
      "learning_rate": 2.7841409266693838e-05,
      "loss": 0.4376,
      "step": 3800
    },
    {
      "epoch": 1.77,
      "grad_norm": 44.122901916503906,
      "learning_rate": 2.7729298020106363e-05,
      "loss": 0.3529,
      "step": 3900
    },
    {
      "epoch": 1.82,
      "grad_norm": 29.057430267333984,
      "learning_rate": 2.761458636589813e-05,
      "loss": 0.3423,
      "step": 4000
    },
    {
      "epoch": 1.87,
      "grad_norm": 39.68132019042969,
      "learning_rate": 2.7497297737966217e-05,
      "loss": 0.3745,
      "step": 4100
    },
    {
      "epoch": 1.91,
      "grad_norm": 25.461807250976562,
      "learning_rate": 2.7377456096645395e-05,
      "loss": 0.4214,
      "step": 4200
    },
    {
      "epoch": 1.96,
      "grad_norm": 34.991153717041016,
      "learning_rate": 2.725508592381337e-05,
      "loss": 0.3975,
      "step": 4300
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9093439363817097,
      "eval_loss": 0.3232283592224121,
      "eval_runtime": 114.6327,
      "eval_samples_per_second": 21.94,
      "eval_steps_per_second": 2.748,
      "step": 4396
    },
    {
      "epoch": 2.0,
      "grad_norm": 23.15580177307129,
      "learning_rate": 2.7130212217889484e-05,
      "loss": 0.3681,
      "step": 4400
    },
    {
      "epoch": 2.05,
      "grad_norm": 28.531185150146484,
      "learning_rate": 2.7002860488727944e-05,
      "loss": 0.3262,
      "step": 4500
    },
    {
      "epoch": 2.09,
      "grad_norm": 38.52848434448242,
      "learning_rate": 2.6873056752406504e-05,
      "loss": 0.3097,
      "step": 4600
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.2146570682525635,
      "learning_rate": 2.6740827525911766e-05,
      "loss": 0.4274,
      "step": 4700
    },
    {
      "epoch": 2.18,
      "grad_norm": 19.732980728149414,
      "learning_rate": 2.6606199821722166e-05,
      "loss": 0.268,
      "step": 4800
    },
    {
      "epoch": 2.23,
      "grad_norm": 15.875640869140625,
      "learning_rate": 2.646920114228972e-05,
      "loss": 0.2734,
      "step": 4900
    },
    {
      "epoch": 2.27,
      "grad_norm": 18.1243953704834,
      "learning_rate": 2.632985947442167e-05,
      "loss": 0.2739,
      "step": 5000
    },
    {
      "epoch": 2.32,
      "grad_norm": 17.237104415893555,
      "learning_rate": 2.6188203283563198e-05,
      "loss": 0.2917,
      "step": 5100
    },
    {
      "epoch": 2.37,
      "grad_norm": 18.371570587158203,
      "learning_rate": 2.6044261507982356e-05,
      "loss": 0.3551,
      "step": 5200
    },
    {
      "epoch": 2.41,
      "grad_norm": 14.930416107177734,
      "learning_rate": 2.589806355285841e-05,
      "loss": 0.3264,
      "step": 5300
    },
    {
      "epoch": 2.46,
      "grad_norm": 30.236698150634766,
      "learning_rate": 2.5749639284274782e-05,
      "loss": 0.3388,
      "step": 5400
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.13715609908103943,
      "learning_rate": 2.5599019023117872e-05,
      "loss": 0.2575,
      "step": 5500
    },
    {
      "epoch": 2.55,
      "grad_norm": 6.84519624710083,
      "learning_rate": 2.5446233538882924e-05,
      "loss": 0.3263,
      "step": 5600
    },
    {
      "epoch": 2.59,
      "grad_norm": 58.51634979248047,
      "learning_rate": 2.5291314043388295e-05,
      "loss": 0.3716,
      "step": 5700
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.961408257484436,
      "learning_rate": 2.513429218439932e-05,
      "loss": 0.3012,
      "step": 5800
    },
    {
      "epoch": 2.68,
      "grad_norm": 13.59073257446289,
      "learning_rate": 2.497520003916316e-05,
      "loss": 0.2911,
      "step": 5900
    },
    {
      "epoch": 2.73,
      "grad_norm": 12.26279354095459,
      "learning_rate": 2.4814070107855878e-05,
      "loss": 0.279,
      "step": 6000
    },
    {
      "epoch": 2.78,
      "grad_norm": 30.789432525634766,
      "learning_rate": 2.465093530694315e-05,
      "loss": 0.29,
      "step": 6100
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.28742390871047974,
      "learning_rate": 2.448582896245591e-05,
      "loss": 0.3051,
      "step": 6200
    },
    {
      "epoch": 2.87,
      "grad_norm": 2.97086501121521,
      "learning_rate": 2.4318784803182317e-05,
      "loss": 0.2735,
      "step": 6300
    },
    {
      "epoch": 2.91,
      "grad_norm": 32.147727966308594,
      "learning_rate": 2.4149836953777488e-05,
      "loss": 0.2992,
      "step": 6400
    },
    {
      "epoch": 2.96,
      "grad_norm": 28.64170265197754,
      "learning_rate": 2.3979019927792315e-05,
      "loss": 0.3337,
      "step": 6500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.925248508946322,
      "eval_loss": 0.3209765553474426,
      "eval_runtime": 113.6693,
      "eval_samples_per_second": 22.126,
      "eval_steps_per_second": 2.771,
      "step": 6594
    },
    {
      "epoch": 3.0,
      "grad_norm": 14.05492115020752,
      "learning_rate": 2.3806368620622876e-05,
      "loss": 0.3353,
      "step": 6600
    },
    {
      "epoch": 3.05,
      "grad_norm": 5.791962623596191,
      "learning_rate": 2.3631918302381803e-05,
      "loss": 0.2148,
      "step": 6700
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.8230623602867126,
      "learning_rate": 2.345570461069312e-05,
      "loss": 0.267,
      "step": 6800
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.2485540211200714,
      "learning_rate": 2.327776354341202e-05,
      "loss": 0.2472,
      "step": 6900
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.06697408854961395,
      "learning_rate": 2.3098131451271016e-05,
      "loss": 0.2568,
      "step": 7000
    },
    {
      "epoch": 3.23,
      "grad_norm": 5.22568941116333,
      "learning_rate": 2.291684503045402e-05,
      "loss": 0.2392,
      "step": 7100
    },
    {
      "epoch": 3.28,
      "grad_norm": 58.771671295166016,
      "learning_rate": 2.2733941315099883e-05,
      "loss": 0.2867,
      "step": 7200
    },
    {
      "epoch": 3.32,
      "grad_norm": 21.552099227905273,
      "learning_rate": 2.2549457669736836e-05,
      "loss": 0.2652,
      "step": 7300
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.265632152557373,
      "learning_rate": 2.2363431781649483e-05,
      "loss": 0.2284,
      "step": 7400
    },
    {
      "epoch": 3.41,
      "grad_norm": 5.533514499664307,
      "learning_rate": 2.2175901653179847e-05,
      "loss": 0.232,
      "step": 7500
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.05318501219153404,
      "learning_rate": 2.1986905593964048e-05,
      "loss": 0.2094,
      "step": 7600
    },
    {
      "epoch": 3.5,
      "grad_norm": 20.877737045288086,
      "learning_rate": 2.1796482213106203e-05,
      "loss": 0.3036,
      "step": 7700
    },
    {
      "epoch": 3.55,
      "grad_norm": 43.25149917602539,
      "learning_rate": 2.1604670411291174e-05,
      "loss": 0.2388,
      "step": 7800
    },
    {
      "epoch": 3.59,
      "grad_norm": 15.92627239227295,
      "learning_rate": 2.1411509372837724e-05,
      "loss": 0.3357,
      "step": 7900
    },
    {
      "epoch": 3.64,
      "grad_norm": 12.71852970123291,
      "learning_rate": 2.121703855769373e-05,
      "loss": 0.2069,
      "step": 8000
    },
    {
      "epoch": 3.69,
      "grad_norm": 12.913580894470215,
      "learning_rate": 2.102129769337511e-05,
      "loss": 0.2867,
      "step": 8100
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.06732411682605743,
      "learning_rate": 2.0824326766850072e-05,
      "loss": 0.28,
      "step": 8200
    },
    {
      "epoch": 3.78,
      "grad_norm": 12.845504760742188,
      "learning_rate": 2.0626166016370375e-05,
      "loss": 0.2245,
      "step": 8300
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.00928166788071394,
      "learning_rate": 2.042685592325123e-05,
      "loss": 0.2359,
      "step": 8400
    },
    {
      "epoch": 3.87,
      "grad_norm": 9.65197467803955,
      "learning_rate": 2.0226437203601602e-05,
      "loss": 0.2984,
      "step": 8500
    },
    {
      "epoch": 3.91,
      "grad_norm": 25.082759857177734,
      "learning_rate": 2.0024950800006463e-05,
      "loss": 0.2164,
      "step": 8600
    },
    {
      "epoch": 3.96,
      "grad_norm": 9.887031555175781,
      "learning_rate": 1.9822437873162863e-05,
      "loss": 0.2279,
      "step": 8700
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9308151093439364,
      "eval_loss": 0.3029824197292328,
      "eval_runtime": 112.1873,
      "eval_samples_per_second": 22.418,
      "eval_steps_per_second": 2.808,
      "step": 8792
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.438531398773193,
      "learning_rate": 1.961893979347137e-05,
      "loss": 0.2683,
      "step": 8800
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.2623905837535858,
      "learning_rate": 1.9414498132584773e-05,
      "loss": 0.2297,
      "step": 8900
    },
    {
      "epoch": 4.09,
      "grad_norm": 13.5099515914917,
      "learning_rate": 1.9209154654915524e-05,
      "loss": 0.1985,
      "step": 9000
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.006642199121415615,
      "learning_rate": 1.900295130910396e-05,
      "loss": 0.2145,
      "step": 9100
    },
    {
      "epoch": 4.19,
      "grad_norm": 41.91756057739258,
      "learning_rate": 1.879593021944875e-05,
      "loss": 0.2221,
      "step": 9200
    },
    {
      "epoch": 4.23,
      "grad_norm": 21.074771881103516,
      "learning_rate": 1.8588133677301595e-05,
      "loss": 0.2129,
      "step": 9300
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.3562418520450592,
      "learning_rate": 1.837960413242765e-05,
      "loss": 0.2438,
      "step": 9400
    },
    {
      "epoch": 4.32,
      "grad_norm": 40.30086898803711,
      "learning_rate": 1.817038418433373e-05,
      "loss": 0.2036,
      "step": 9500
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.014717689715325832,
      "learning_rate": 1.796051657356582e-05,
      "loss": 0.2003,
      "step": 9600
    },
    {
      "epoch": 4.41,
      "grad_norm": 17.778074264526367,
      "learning_rate": 1.7750044172977838e-05,
      "loss": 0.2179,
      "step": 9700
    },
    {
      "epoch": 4.46,
      "grad_norm": 5.801839351654053,
      "learning_rate": 1.7539009978973312e-05,
      "loss": 0.2142,
      "step": 9800
    },
    {
      "epoch": 4.5,
      "grad_norm": 5.123954772949219,
      "learning_rate": 1.7327457102721887e-05,
      "loss": 0.1686,
      "step": 9900
    },
    {
      "epoch": 4.55,
      "grad_norm": 4.904812812805176,
      "learning_rate": 1.711542876135233e-05,
      "loss": 0.2073,
      "step": 10000
    },
    {
      "epoch": 4.6,
      "grad_norm": 15.447718620300293,
      "learning_rate": 1.6902968269123902e-05,
      "loss": 0.2276,
      "step": 10100
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.12395219504833221,
      "learning_rate": 1.669011902857791e-05,
      "loss": 0.1882,
      "step": 10200
    },
    {
      "epoch": 4.69,
      "grad_norm": 17.49830436706543,
      "learning_rate": 1.6476924521671194e-05,
      "loss": 0.2127,
      "step": 10300
    },
    {
      "epoch": 4.73,
      "grad_norm": 10.882052421569824,
      "learning_rate": 1.6263428300893422e-05,
      "loss": 0.2202,
      "step": 10400
    },
    {
      "epoch": 4.78,
      "grad_norm": 37.238731384277344,
      "learning_rate": 1.604967398036996e-05,
      "loss": 0.1663,
      "step": 10500
    },
    {
      "epoch": 4.82,
      "grad_norm": 9.180378913879395,
      "learning_rate": 1.5835705226952112e-05,
      "loss": 0.2547,
      "step": 10600
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.04440607875585556,
      "learning_rate": 1.5621565751296676e-05,
      "loss": 0.2186,
      "step": 10700
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.13343572616577148,
      "learning_rate": 1.540729929893649e-05,
      "loss": 0.2779,
      "step": 10800
    },
    {
      "epoch": 4.96,
      "grad_norm": 1.3157250881195068,
      "learning_rate": 1.5192949641343834e-05,
      "loss": 0.1696,
      "step": 10900
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9292246520874752,
      "eval_loss": 0.34775686264038086,
      "eval_runtime": 111.4119,
      "eval_samples_per_second": 22.574,
      "eval_steps_per_second": 2.827,
      "step": 10990
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.44680455327033997,
      "learning_rate": 1.4978560566988603e-05,
      "loss": 0.23,
      "step": 11000
    },
    {
      "epoch": 5.05,
      "grad_norm": 0.041424062103033066,
      "learning_rate": 1.4764175872392958e-05,
      "loss": 0.1913,
      "step": 11100
    },
    {
      "epoch": 5.1,
      "grad_norm": 2.6414008140563965,
      "learning_rate": 1.454983935318433e-05,
      "loss": 0.1374,
      "step": 11200
    },
    {
      "epoch": 5.14,
      "grad_norm": 0.396395206451416,
      "learning_rate": 1.433559479514864e-05,
      "loss": 0.1807,
      "step": 11300
    },
    {
      "epoch": 5.19,
      "grad_norm": 8.14771842956543,
      "learning_rate": 1.4121485965285485e-05,
      "loss": 0.1888,
      "step": 11400
    },
    {
      "epoch": 5.23,
      "grad_norm": 0.3801160454750061,
      "learning_rate": 1.3907556602867213e-05,
      "loss": 0.1838,
      "step": 11500
    },
    {
      "epoch": 5.28,
      "grad_norm": 1.0655605792999268,
      "learning_rate": 1.3693850410503614e-05,
      "loss": 0.1483,
      "step": 11600
    },
    {
      "epoch": 5.32,
      "grad_norm": 0.37717050313949585,
      "learning_rate": 1.3480411045214147e-05,
      "loss": 0.1635,
      "step": 11700
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.017707068473100662,
      "learning_rate": 1.326728210950942e-05,
      "loss": 0.1871,
      "step": 11800
    },
    {
      "epoch": 5.41,
      "grad_norm": 0.01935901865363121,
      "learning_rate": 1.3054507142483875e-05,
      "loss": 0.1479,
      "step": 11900
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.2556498050689697,
      "learning_rate": 1.2842129610921378e-05,
      "loss": 0.1754,
      "step": 12000
    },
    {
      "epoch": 5.51,
      "grad_norm": 1.742725133895874,
      "learning_rate": 1.2630192900415582e-05,
      "loss": 0.2713,
      "step": 12100
    },
    {
      "epoch": 5.55,
      "grad_norm": 40.924644470214844,
      "learning_rate": 1.2418740306506923e-05,
      "loss": 0.1909,
      "step": 12200
    },
    {
      "epoch": 5.6,
      "grad_norm": 49.58342361450195,
      "learning_rate": 1.2207815025837977e-05,
      "loss": 0.159,
      "step": 12300
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.03460393100976944,
      "learning_rate": 1.1997460147328984e-05,
      "loss": 0.1559,
      "step": 12400
    },
    {
      "epoch": 5.69,
      "grad_norm": 0.0873652920126915,
      "learning_rate": 1.178771864337546e-05,
      "loss": 0.1713,
      "step": 12500
    },
    {
      "epoch": 5.73,
      "grad_norm": 0.1891886442899704,
      "learning_rate": 1.1578633361069559e-05,
      "loss": 0.2004,
      "step": 12600
    },
    {
      "epoch": 5.78,
      "grad_norm": 0.3393521010875702,
      "learning_rate": 1.1370247013447035e-05,
      "loss": 0.1993,
      "step": 12700
    },
    {
      "epoch": 5.82,
      "grad_norm": 3.934659719467163,
      "learning_rate": 1.1162602170761611e-05,
      "loss": 0.1507,
      "step": 12800
    },
    {
      "epoch": 5.87,
      "grad_norm": 1.6851441860198975,
      "learning_rate": 1.095574125178849e-05,
      "loss": 0.1649,
      "step": 12900
    },
    {
      "epoch": 5.91,
      "grad_norm": 74.29869079589844,
      "learning_rate": 1.0749706515158863e-05,
      "loss": 0.2056,
      "step": 13000
    },
    {
      "epoch": 5.96,
      "grad_norm": 23.88290786743164,
      "learning_rate": 1.0544540050727048e-05,
      "loss": 0.1658,
      "step": 13100
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9427435387673956,
      "eval_loss": 0.3084171712398529,
      "eval_runtime": 111.2598,
      "eval_samples_per_second": 22.605,
      "eval_steps_per_second": 2.831,
      "step": 13188
    },
    {
      "epoch": 6.01,
      "grad_norm": 15.39907169342041,
      "learning_rate": 1.0340283770972167e-05,
      "loss": 0.2241,
      "step": 13200
    },
    {
      "epoch": 6.05,
      "grad_norm": 0.004868438933044672,
      "learning_rate": 1.0136979402436069e-05,
      "loss": 0.1529,
      "step": 13300
    },
    {
      "epoch": 6.1,
      "grad_norm": 45.70091247558594,
      "learning_rate": 9.93466847719919e-06,
      "loss": 0.1543,
      "step": 13400
    },
    {
      "epoch": 6.14,
      "grad_norm": 0.8413438200950623,
      "learning_rate": 9.733392324396167e-06,
      "loss": 0.1236,
      "step": 13500
    },
    {
      "epoch": 6.19,
      "grad_norm": 13.94447135925293,
      "learning_rate": 9.533192061772919e-06,
      "loss": 0.1151,
      "step": 13600
    },
    {
      "epoch": 6.23,
      "grad_norm": 0.031886328011751175,
      "learning_rate": 9.334108587286877e-06,
      "loss": 0.1213,
      "step": 13700
    },
    {
      "epoch": 6.28,
      "grad_norm": 1.3224759101867676,
      "learning_rate": 9.136182570752153e-06,
      "loss": 0.1151,
      "step": 13800
    },
    {
      "epoch": 6.32,
      "grad_norm": 0.02312690019607544,
      "learning_rate": 8.93945444553128e-06,
      "loss": 0.1601,
      "step": 13900
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.02013530395925045,
      "learning_rate": 8.743964400275304e-06,
      "loss": 0.133,
      "step": 14000
    },
    {
      "epoch": 6.41,
      "grad_norm": 0.02437993884086609,
      "learning_rate": 8.549752370713798e-06,
      "loss": 0.1754,
      "step": 14100
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.005770612042397261,
      "learning_rate": 8.356858031496596e-06,
      "loss": 0.1421,
      "step": 14200
    },
    {
      "epoch": 6.51,
      "grad_norm": 0.008364029228687286,
      "learning_rate": 8.165320788088888e-06,
      "loss": 0.1988,
      "step": 14300
    },
    {
      "epoch": 6.55,
      "grad_norm": 0.06003904342651367,
      "learning_rate": 7.975179768721187e-06,
      "loss": 0.073,
      "step": 14400
    },
    {
      "epoch": 6.6,
      "grad_norm": 13.212935447692871,
      "learning_rate": 7.78647381639607e-06,
      "loss": 0.1858,
      "step": 14500
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.03764009103178978,
      "learning_rate": 7.599241480953112e-06,
      "loss": 0.1373,
      "step": 14600
    },
    {
      "epoch": 6.69,
      "grad_norm": 0.01994572952389717,
      "learning_rate": 7.413521011193705e-06,
      "loss": 0.1165,
      "step": 14700
    },
    {
      "epoch": 6.73,
      "grad_norm": 36.3884162902832,
      "learning_rate": 7.229350347067426e-06,
      "loss": 0.1989,
      "step": 14800
    },
    {
      "epoch": 6.78,
      "grad_norm": 0.005731828045099974,
      "learning_rate": 7.046767111921425e-06,
      "loss": 0.0917,
      "step": 14900
    },
    {
      "epoch": 6.82,
      "grad_norm": 48.87177658081055,
      "learning_rate": 6.865808604814564e-06,
      "loss": 0.1263,
      "step": 15000
    },
    {
      "epoch": 6.87,
      "grad_norm": 29.0344295501709,
      "learning_rate": 6.686511792897767e-06,
      "loss": 0.1724,
      "step": 15100
    },
    {
      "epoch": 6.92,
      "grad_norm": 0.013815321959555149,
      "learning_rate": 6.508913303862144e-06,
      "loss": 0.1795,
      "step": 15200
    },
    {
      "epoch": 6.96,
      "grad_norm": 29.06551742553711,
      "learning_rate": 6.333049418456533e-06,
      "loss": 0.1383,
      "step": 15300
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9391650099403579,
      "eval_loss": 0.3318726718425751,
      "eval_runtime": 111.0425,
      "eval_samples_per_second": 22.649,
      "eval_steps_per_second": 2.837,
      "step": 15386
    },
    {
      "epoch": 7.01,
      "grad_norm": 54.77537536621094,
      "learning_rate": 6.1589560630758656e-06,
      "loss": 0.102,
      "step": 15400
    },
    {
      "epoch": 7.05,
      "grad_norm": 1.445946455001831,
      "learning_rate": 5.986668802421924e-06,
      "loss": 0.1061,
      "step": 15500
    },
    {
      "epoch": 7.1,
      "grad_norm": 30.50276756286621,
      "learning_rate": 5.8162228322380155e-06,
      "loss": 0.1004,
      "step": 15600
    },
    {
      "epoch": 7.14,
      "grad_norm": 0.28580981492996216,
      "learning_rate": 5.647652972118998e-06,
      "loss": 0.095,
      "step": 15700
    },
    {
      "epoch": 7.19,
      "grad_norm": 8.433425903320312,
      "learning_rate": 5.480993658398129e-06,
      "loss": 0.2191,
      "step": 15800
    },
    {
      "epoch": 7.23,
      "grad_norm": 23.46518898010254,
      "learning_rate": 5.316278937112267e-06,
      "loss": 0.1039,
      "step": 15900
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.1998785138130188,
      "learning_rate": 5.153542457046737e-06,
      "loss": 0.1161,
      "step": 16000
    },
    {
      "epoch": 7.32,
      "grad_norm": 0.5174553990364075,
      "learning_rate": 4.992817462861397e-06,
      "loss": 0.1093,
      "step": 16100
    },
    {
      "epoch": 7.37,
      "grad_norm": 10.745183944702148,
      "learning_rate": 4.834136788299248e-06,
      "loss": 0.152,
      "step": 16200
    },
    {
      "epoch": 7.42,
      "grad_norm": 0.035963889211416245,
      "learning_rate": 4.67753284947898e-06,
      "loss": 0.0887,
      "step": 16300
    },
    {
      "epoch": 7.46,
      "grad_norm": 0.02713492140173912,
      "learning_rate": 4.523037638272822e-06,
      "loss": 0.0964,
      "step": 16400
    },
    {
      "epoch": 7.51,
      "grad_norm": 0.04027700796723366,
      "learning_rate": 4.370682715771108e-06,
      "loss": 0.0898,
      "step": 16500
    },
    {
      "epoch": 7.55,
      "grad_norm": 32.831809997558594,
      "learning_rate": 4.220499205834783e-06,
      "loss": 0.1376,
      "step": 16600
    },
    {
      "epoch": 7.6,
      "grad_norm": 1.195786952972412,
      "learning_rate": 4.072517788737264e-06,
      "loss": 0.1386,
      "step": 16700
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.0985260009765625,
      "learning_rate": 3.926768694896931e-06,
      "loss": 0.1031,
      "step": 16800
    },
    {
      "epoch": 7.69,
      "grad_norm": 66.71458435058594,
      "learning_rate": 3.783281698701482e-06,
      "loss": 0.1584,
      "step": 16900
    },
    {
      "epoch": 7.73,
      "grad_norm": 0.06048920005559921,
      "learning_rate": 3.6420861124254607e-06,
      "loss": 0.1557,
      "step": 17000
    },
    {
      "epoch": 7.78,
      "grad_norm": 35.74193572998047,
      "learning_rate": 3.5032107802422107e-06,
      "loss": 0.1509,
      "step": 17100
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.14910955727100372,
      "learning_rate": 3.3666840723314145e-06,
      "loss": 0.1421,
      "step": 17200
    },
    {
      "epoch": 7.87,
      "grad_norm": 44.823760986328125,
      "learning_rate": 3.232533879083511e-06,
      "loss": 0.1181,
      "step": 17300
    },
    {
      "epoch": 7.92,
      "grad_norm": 7.532151699066162,
      "learning_rate": 3.1007876054020724e-06,
      "loss": 0.1385,
      "step": 17400
    },
    {
      "epoch": 7.96,
      "grad_norm": 28.50664520263672,
      "learning_rate": 2.9714721651054e-06,
      "loss": 0.1222,
      "step": 17500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9479125248508946,
      "eval_loss": 0.3132196068763733,
      "eval_runtime": 111.2173,
      "eval_samples_per_second": 22.613,
      "eval_steps_per_second": 2.832,
      "step": 17584
    },
    {
      "epoch": 8.01,
      "grad_norm": 0.00245882966555655,
      "learning_rate": 2.8446139754284486e-06,
      "loss": 0.0782,
      "step": 17600
    },
    {
      "epoch": 8.05,
      "grad_norm": 0.3096585273742676,
      "learning_rate": 2.7202389516261346e-06,
      "loss": 0.1482,
      "step": 17700
    },
    {
      "epoch": 8.1,
      "grad_norm": 38.84019088745117,
      "learning_rate": 2.5983725016792574e-06,
      "loss": 0.1191,
      "step": 17800
    },
    {
      "epoch": 8.14,
      "grad_norm": 0.24393437802791595,
      "learning_rate": 2.4790395211040296e-06,
      "loss": 0.1095,
      "step": 17900
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.32781603932380676,
      "learning_rate": 2.36226438786627e-06,
      "loss": 0.121,
      "step": 18000
    },
    {
      "epoch": 8.23,
      "grad_norm": 8.179335594177246,
      "learning_rate": 2.2480709574013637e-06,
      "loss": 0.0564,
      "step": 18100
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.03705134242773056,
      "learning_rate": 2.1364825577409424e-06,
      "loss": 0.1466,
      "step": 18200
    },
    {
      "epoch": 8.33,
      "grad_norm": 0.8985774517059326,
      "learning_rate": 2.0275219847473026e-06,
      "loss": 0.1097,
      "step": 18300
    },
    {
      "epoch": 8.37,
      "grad_norm": 7.916272163391113,
      "learning_rate": 1.9212114974565664e-06,
      "loss": 0.1213,
      "step": 18400
    },
    {
      "epoch": 8.42,
      "grad_norm": 0.16036684811115265,
      "learning_rate": 1.8175728135314707e-06,
      "loss": 0.0848,
      "step": 18500
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.01501951552927494,
      "learning_rate": 1.7166271048247796e-06,
      "loss": 0.089,
      "step": 18600
    },
    {
      "epoch": 8.51,
      "grad_norm": 80.24330139160156,
      "learning_rate": 1.6183949930541898e-06,
      "loss": 0.1014,
      "step": 18700
    },
    {
      "epoch": 8.55,
      "grad_norm": 0.014790402725338936,
      "learning_rate": 1.5228965455896054e-06,
      "loss": 0.1042,
      "step": 18800
    },
    {
      "epoch": 8.6,
      "grad_norm": 0.005228283815085888,
      "learning_rate": 1.4301512713536873e-06,
      "loss": 0.1056,
      "step": 18900
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.0038773128762841225,
      "learning_rate": 1.3401781168364591e-06,
      "loss": 0.0963,
      "step": 19000
    },
    {
      "epoch": 8.69,
      "grad_norm": 0.009148034267127514,
      "learning_rate": 1.2529954622248114e-06,
      "loss": 0.109,
      "step": 19100
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.10844717919826508,
      "learning_rate": 1.1686211176477208e-06,
      "loss": 0.1075,
      "step": 19200
    },
    {
      "epoch": 8.78,
      "grad_norm": 0.017569424584507942,
      "learning_rate": 1.0870723195378852e-06,
      "loss": 0.1228,
      "step": 19300
    },
    {
      "epoch": 8.83,
      "grad_norm": 39.32170104980469,
      "learning_rate": 1.00836572711058e-06,
      "loss": 0.1414,
      "step": 19400
    },
    {
      "epoch": 8.87,
      "grad_norm": 28.396100997924805,
      "learning_rate": 9.325174189604346e-07,
      "loss": 0.129,
      "step": 19500
    },
    {
      "epoch": 8.92,
      "grad_norm": 0.031751640141010284,
      "learning_rate": 8.595428897768071e-07,
      "loss": 0.1163,
      "step": 19600
    },
    {
      "epoch": 8.96,
      "grad_norm": 1.8334925174713135,
      "learning_rate": 7.894570471784418e-07,
      "loss": 0.1196,
      "step": 19700
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9467196819085487,
      "eval_loss": 0.3136024475097656,
      "eval_runtime": 111.3615,
      "eval_samples_per_second": 22.584,
      "eval_steps_per_second": 2.829,
      "step": 19782
    },
    {
      "epoch": 9.01,
      "grad_norm": 9.599562644958496,
      "learning_rate": 7.222742086680756e-07,
      "loss": 0.1225,
      "step": 19800
    },
    {
      "epoch": 9.05,
      "grad_norm": 20.813501358032227,
      "learning_rate": 6.580080987075721e-07,
      "loss": 0.0845,
      "step": 19900
    },
    {
      "epoch": 9.1,
      "grad_norm": 0.00961952656507492,
      "learning_rate": 5.966718459142196e-07,
      "loss": 0.107,
      "step": 20000
    },
    {
      "epoch": 9.14,
      "grad_norm": 0.0361330471932888,
      "learning_rate": 5.382779803787579e-07,
      "loss": 0.0993,
      "step": 20100
    },
    {
      "epoch": 9.19,
      "grad_norm": 0.8669134378433228,
      "learning_rate": 4.82838431105655e-07,
      "loss": 0.0874,
      "step": 20200
    },
    {
      "epoch": 9.24,
      "grad_norm": 0.02739112637937069,
      "learning_rate": 4.303645235761866e-07,
      "loss": 0.1095,
      "step": 20300
    },
    {
      "epoch": 9.28,
      "grad_norm": 0.019733713939785957,
      "learning_rate": 3.808669774348167e-07,
      "loss": 0.1325,
      "step": 20400
    },
    {
      "epoch": 9.33,
      "grad_norm": 0.0024758102372288704,
      "learning_rate": 3.3435590429932493e-07,
      "loss": 0.0883,
      "step": 20500
    },
    {
      "epoch": 9.37,
      "grad_norm": 0.04354145750403404,
      "learning_rate": 2.908408056951578e-07,
      "loss": 0.0939,
      "step": 20600
    },
    {
      "epoch": 9.42,
      "grad_norm": 10.014009475708008,
      "learning_rate": 2.5033057111440106e-07,
      "loss": 0.0856,
      "step": 20700
    },
    {
      "epoch": 9.46,
      "grad_norm": 0.00877867080271244,
      "learning_rate": 2.1283347619979243e-07,
      "loss": 0.0977,
      "step": 20800
    },
    {
      "epoch": 9.51,
      "grad_norm": 0.0024496885016560555,
      "learning_rate": 1.7835718105413235e-07,
      "loss": 0.1331,
      "step": 20900
    },
    {
      "epoch": 9.55,
      "grad_norm": 0.2954551577568054,
      "learning_rate": 1.4690872867542892e-07,
      "loss": 0.126,
      "step": 21000
    },
    {
      "epoch": 9.6,
      "grad_norm": 0.06035974249243736,
      "learning_rate": 1.1849454351812394e-07,
      "loss": 0.077,
      "step": 21100
    },
    {
      "epoch": 9.65,
      "grad_norm": 0.7805526852607727,
      "learning_rate": 9.312043018067762e-08,
      "loss": 0.1167,
      "step": 21200
    },
    {
      "epoch": 9.69,
      "grad_norm": 0.09065116941928864,
      "learning_rate": 7.079157221975718e-08,
      "loss": 0.1658,
      "step": 21300
    },
    {
      "epoch": 9.74,
      "grad_norm": 0.00335172307677567,
      "learning_rate": 5.1512531091333914e-08,
      "loss": 0.126,
      "step": 21400
    },
    {
      "epoch": 9.78,
      "grad_norm": 1.6308414936065674,
      "learning_rate": 3.528724521882687e-08,
      "loss": 0.1088,
      "step": 21500
    },
    {
      "epoch": 9.83,
      "grad_norm": 22.293624877929688,
      "learning_rate": 2.211902918855313e-08,
      "loss": 0.1201,
      "step": 21600
    },
    {
      "epoch": 9.87,
      "grad_norm": 0.030308537185192108,
      "learning_rate": 1.2010573072602783e-08,
      "loss": 0.0758,
      "step": 21700
    },
    {
      "epoch": 9.92,
      "grad_norm": 5.30928897857666,
      "learning_rate": 4.963941879295164e-09,
      "loss": 0.0952,
      "step": 21800
    },
    {
      "epoch": 9.96,
      "grad_norm": 0.04648282751441002,
      "learning_rate": 9.805751313296529e-10,
      "loss": 0.1257,
      "step": 21900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.94831013916501,
      "eval_loss": 0.3119918704032898,
      "eval_runtime": 111.4853,
      "eval_samples_per_second": 22.559,
      "eval_steps_per_second": 2.825,
      "step": 21980
    },
    {
      "epoch": 10.0,
      "step": 21980,
      "total_flos": 4.09349935387607e+19,
      "train_loss": 0.2596938943081926,
      "train_runtime": 18130.6882,
      "train_samples_per_second": 9.697,
      "train_steps_per_second": 1.212
    }
  ],
  "logging_steps": 100,
  "max_steps": 21980,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 4.09349935387607e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}