{
  "best_metric": 5.272278467814128,
  "best_model_checkpoint": "./results/checkpoint-5496",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 6412,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1091703056768559,
      "grad_norm": 25.480337142944336,
      "learning_rate": 2.959061135371179e-05,
      "loss": 56.923,
      "step": 100
    },
    {
      "epoch": 0.2183406113537118,
      "grad_norm": 39.87223815917969,
      "learning_rate": 2.918122270742358e-05,
      "loss": 46.6475,
      "step": 200
    },
    {
      "epoch": 0.32751091703056767,
      "grad_norm": 48.05048751831055,
      "learning_rate": 2.877183406113537e-05,
      "loss": 33.6867,
      "step": 300
    },
    {
      "epoch": 0.4366812227074236,
      "grad_norm": 31.941883087158203,
      "learning_rate": 2.8362445414847164e-05,
      "loss": 21.1084,
      "step": 400
    },
    {
      "epoch": 0.5458515283842795,
      "grad_norm": 55.025856018066406,
      "learning_rate": 2.7953056768558954e-05,
      "loss": 12.9495,
      "step": 500
    },
    {
      "epoch": 0.6550218340611353,
      "grad_norm": 34.957523345947266,
      "learning_rate": 2.7543668122270742e-05,
      "loss": 10.0745,
      "step": 600
    },
    {
      "epoch": 0.7641921397379913,
      "grad_norm": 24.020906448364258,
      "learning_rate": 2.7134279475982533e-05,
      "loss": 8.3541,
      "step": 700
    },
    {
      "epoch": 0.8733624454148472,
      "grad_norm": 32.709571838378906,
      "learning_rate": 2.6724890829694323e-05,
      "loss": 7.5128,
      "step": 800
    },
    {
      "epoch": 0.982532751091703,
      "grad_norm": 38.94672393798828,
      "learning_rate": 2.6315502183406114e-05,
      "loss": 7.2241,
      "step": 900
    },
    {
      "epoch": 1.0,
      "eval_avg_mae": 7.529487609863281,
      "eval_loss": 7.529487609863281,
      "eval_mae_lex": 6.992014408111572,
      "eval_mae_sem": 5.432034492492676,
      "eval_mae_syn": 10.164413452148438,
      "eval_runtime": 27.1764,
      "eval_samples_per_second": 269.609,
      "eval_steps_per_second": 8.426,
      "step": 916
    },
    {
      "epoch": 1.091703056768559,
      "grad_norm": 26.391666412353516,
      "learning_rate": 2.5906113537117905e-05,
      "loss": 6.7699,
      "step": 1000
    },
    {
      "epoch": 1.2008733624454149,
      "grad_norm": 22.994029998779297,
      "learning_rate": 2.5496724890829696e-05,
      "loss": 6.5552,
      "step": 1100
    },
    {
      "epoch": 1.3100436681222707,
      "grad_norm": 20.722883224487305,
      "learning_rate": 2.5087336244541486e-05,
      "loss": 6.5897,
      "step": 1200
    },
    {
      "epoch": 1.4192139737991267,
      "grad_norm": 32.02668380737305,
      "learning_rate": 2.4677947598253277e-05,
      "loss": 6.5073,
      "step": 1300
    },
    {
      "epoch": 1.5283842794759825,
      "grad_norm": 32.40359115600586,
      "learning_rate": 2.4268558951965064e-05,
      "loss": 6.4684,
      "step": 1400
    },
    {
      "epoch": 1.6375545851528384,
      "grad_norm": 47.73025131225586,
      "learning_rate": 2.3859170305676855e-05,
      "loss": 6.3165,
      "step": 1500
    },
    {
      "epoch": 1.7467248908296944,
      "grad_norm": 47.35511016845703,
      "learning_rate": 2.344978165938865e-05,
      "loss": 6.2866,
      "step": 1600
    },
    {
      "epoch": 1.8558951965065502,
      "grad_norm": 44.51765441894531,
      "learning_rate": 2.3040393013100437e-05,
      "loss": 6.3404,
      "step": 1700
    },
    {
      "epoch": 1.965065502183406,
      "grad_norm": 26.496959686279297,
      "learning_rate": 2.2631004366812227e-05,
      "loss": 6.1681,
      "step": 1800
    },
    {
      "epoch": 2.0,
      "eval_avg_mae": 6.067600250244141,
      "eval_loss": 6.067600250244141,
      "eval_mae_lex": 5.595421314239502,
      "eval_mae_sem": 4.1164045333862305,
      "eval_mae_syn": 8.490975379943848,
      "eval_runtime": 27.2193,
      "eval_samples_per_second": 269.184,
      "eval_steps_per_second": 8.413,
      "step": 1832
    },
    {
      "epoch": 2.074235807860262,
      "grad_norm": 36.847415924072266,
      "learning_rate": 2.2221615720524018e-05,
      "loss": 5.9928,
      "step": 1900
    },
    {
      "epoch": 2.183406113537118,
      "grad_norm": 37.08506393432617,
      "learning_rate": 2.181222707423581e-05,
      "loss": 5.9648,
      "step": 2000
    },
    {
      "epoch": 2.2925764192139737,
      "grad_norm": 35.595909118652344,
      "learning_rate": 2.1402838427947596e-05,
      "loss": 5.8648,
      "step": 2100
    },
    {
      "epoch": 2.4017467248908297,
      "grad_norm": 23.82405662536621,
      "learning_rate": 2.099344978165939e-05,
      "loss": 5.9043,
      "step": 2200
    },
    {
      "epoch": 2.5109170305676853,
      "grad_norm": 30.872852325439453,
      "learning_rate": 2.058406113537118e-05,
      "loss": 5.8428,
      "step": 2300
    },
    {
      "epoch": 2.6200873362445414,
      "grad_norm": 42.079261779785156,
      "learning_rate": 2.0174672489082972e-05,
      "loss": 5.8529,
      "step": 2400
    },
    {
      "epoch": 2.7292576419213974,
      "grad_norm": 23.549190521240234,
      "learning_rate": 1.976528384279476e-05,
      "loss": 5.8328,
      "step": 2500
    },
    {
      "epoch": 2.8384279475982535,
      "grad_norm": 32.223079681396484,
      "learning_rate": 1.935589519650655e-05,
      "loss": 5.8484,
      "step": 2600
    },
    {
      "epoch": 2.947598253275109,
      "grad_norm": 25.67125129699707,
      "learning_rate": 1.894650655021834e-05,
      "loss": 5.5861,
      "step": 2700
    },
    {
      "epoch": 3.0,
      "eval_avg_mae": 5.645811716715495,
      "eval_loss": 5.645811080932617,
      "eval_mae_lex": 4.994715213775635,
      "eval_mae_sem": 3.6993861198425293,
      "eval_mae_syn": 8.24333381652832,
      "eval_runtime": 27.1472,
      "eval_samples_per_second": 269.899,
      "eval_steps_per_second": 8.435,
      "step": 2748
    },
    {
      "epoch": 3.056768558951965,
      "grad_norm": 31.40544319152832,
      "learning_rate": 1.8537117903930135e-05,
      "loss": 5.4937,
      "step": 2800
    },
    {
      "epoch": 3.165938864628821,
      "grad_norm": 28.67197608947754,
      "learning_rate": 1.8127729257641922e-05,
      "loss": 5.5573,
      "step": 2900
    },
    {
      "epoch": 3.2751091703056767,
      "grad_norm": 26.671180725097656,
      "learning_rate": 1.7718340611353713e-05,
      "loss": 5.6147,
      "step": 3000
    },
    {
      "epoch": 3.3842794759825328,
      "grad_norm": 32.73609924316406,
      "learning_rate": 1.7308951965065504e-05,
      "loss": 5.2704,
      "step": 3100
    },
    {
      "epoch": 3.493449781659389,
      "grad_norm": 26.268295288085938,
      "learning_rate": 1.689956331877729e-05,
      "loss": 5.4946,
      "step": 3200
    },
    {
      "epoch": 3.6026200873362444,
      "grad_norm": 24.3873233795166,
      "learning_rate": 1.649017467248908e-05,
      "loss": 5.5757,
      "step": 3300
    },
    {
      "epoch": 3.7117903930131004,
      "grad_norm": 26.872316360473633,
      "learning_rate": 1.6080786026200872e-05,
      "loss": 5.3305,
      "step": 3400
    },
    {
      "epoch": 3.8209606986899565,
      "grad_norm": 31.78321647644043,
      "learning_rate": 1.5671397379912666e-05,
      "loss": 5.4091,
      "step": 3500
    },
    {
      "epoch": 3.930131004366812,
      "grad_norm": 37.95060729980469,
      "learning_rate": 1.5262008733624454e-05,
      "loss": 5.464,
      "step": 3600
    },
    {
      "epoch": 4.0,
      "eval_avg_mae": 5.841625213623047,
      "eval_loss": 5.841624736785889,
      "eval_mae_lex": 5.572142124176025,
      "eval_mae_sem": 3.7872631549835205,
      "eval_mae_syn": 8.1654691696167,
      "eval_runtime": 27.1382,
      "eval_samples_per_second": 269.988,
      "eval_steps_per_second": 8.438,
      "step": 3664
    },
    {
      "epoch": 4.039301310043668,
      "grad_norm": 37.76191329956055,
      "learning_rate": 1.4852620087336245e-05,
      "loss": 5.4072,
      "step": 3700
    },
    {
      "epoch": 4.148471615720524,
      "grad_norm": 33.29827117919922,
      "learning_rate": 1.4443231441048035e-05,
      "loss": 5.0985,
      "step": 3800
    },
    {
      "epoch": 4.25764192139738,
      "grad_norm": 47.478206634521484,
      "learning_rate": 1.4033842794759826e-05,
      "loss": 5.1541,
      "step": 3900
    },
    {
      "epoch": 4.366812227074236,
      "grad_norm": 31.66642189025879,
      "learning_rate": 1.3624454148471617e-05,
      "loss": 5.106,
      "step": 4000
    },
    {
      "epoch": 4.475982532751091,
      "grad_norm": 27.389015197753906,
      "learning_rate": 1.3215065502183406e-05,
      "loss": 5.1793,
      "step": 4100
    },
    {
      "epoch": 4.585152838427947,
      "grad_norm": 26.702226638793945,
      "learning_rate": 1.2805676855895198e-05,
      "loss": 5.0975,
      "step": 4200
    },
    {
      "epoch": 4.6943231441048034,
      "grad_norm": 31.537691116333008,
      "learning_rate": 1.2396288209606987e-05,
      "loss": 4.9613,
      "step": 4300
    },
    {
      "epoch": 4.8034934497816595,
      "grad_norm": 26.946945190429688,
      "learning_rate": 1.1986899563318778e-05,
      "loss": 5.1799,
      "step": 4400
    },
    {
      "epoch": 4.9126637554585155,
      "grad_norm": 27.92361068725586,
      "learning_rate": 1.1577510917030569e-05,
      "loss": 5.0393,
      "step": 4500
    },
    {
      "epoch": 5.0,
      "eval_avg_mae": 5.481770197550456,
      "eval_loss": 5.481770038604736,
      "eval_mae_lex": 4.809901714324951,
      "eval_mae_sem": 3.8779022693634033,
      "eval_mae_syn": 7.757506370544434,
      "eval_runtime": 27.0994,
      "eval_samples_per_second": 270.375,
      "eval_steps_per_second": 8.45,
      "step": 4580
    },
    {
      "epoch": 5.021834061135372,
      "grad_norm": 25.387714385986328,
      "learning_rate": 1.1168122270742358e-05,
      "loss": 5.0585,
      "step": 4600
    },
    {
      "epoch": 5.131004366812227,
      "grad_norm": 33.1529655456543,
      "learning_rate": 1.0758733624454149e-05,
      "loss": 4.774,
      "step": 4700
    },
    {
      "epoch": 5.240174672489083,
      "grad_norm": 31.700637817382812,
      "learning_rate": 1.034934497816594e-05,
      "loss": 4.762,
      "step": 4800
    },
    {
      "epoch": 5.349344978165939,
      "grad_norm": 34.32217025756836,
      "learning_rate": 9.93995633187773e-06,
      "loss": 4.8645,
      "step": 4900
    },
    {
      "epoch": 5.458515283842795,
      "grad_norm": 52.338130950927734,
      "learning_rate": 9.530567685589519e-06,
      "loss": 4.9913,
      "step": 5000
    },
    {
      "epoch": 5.567685589519651,
      "grad_norm": 27.761211395263672,
      "learning_rate": 9.12117903930131e-06,
      "loss": 4.9047,
      "step": 5100
    },
    {
      "epoch": 5.676855895196507,
      "grad_norm": 36.54159164428711,
      "learning_rate": 8.7117903930131e-06,
      "loss": 4.7824,
      "step": 5200
    },
    {
      "epoch": 5.786026200873362,
      "grad_norm": 31.954957962036133,
      "learning_rate": 8.302401746724891e-06,
      "loss": 4.7555,
      "step": 5300
    },
    {
      "epoch": 5.895196506550218,
      "grad_norm": 33.35627365112305,
      "learning_rate": 7.89301310043668e-06,
      "loss": 4.8389,
      "step": 5400
    },
    {
      "epoch": 6.0,
      "eval_avg_mae": 5.272278467814128,
      "eval_loss": 5.272278308868408,
      "eval_mae_lex": 4.76518440246582,
      "eval_mae_sem": 3.384120225906372,
      "eval_mae_syn": 7.6675310134887695,
      "eval_runtime": 27.0821,
      "eval_samples_per_second": 270.548,
      "eval_steps_per_second": 8.456,
      "step": 5496
    },
    {
      "epoch": 6.004366812227074,
      "grad_norm": 29.006881713867188,
      "learning_rate": 7.483624454148472e-06,
      "loss": 4.8767,
      "step": 5500
    },
    {
      "epoch": 6.11353711790393,
      "grad_norm": 29.40102195739746,
      "learning_rate": 7.074235807860262e-06,
      "loss": 4.6832,
      "step": 5600
    },
    {
      "epoch": 6.222707423580786,
      "grad_norm": 32.63214111328125,
      "learning_rate": 6.664847161572053e-06,
      "loss": 4.6676,
      "step": 5700
    },
    {
      "epoch": 6.331877729257642,
      "grad_norm": 32.83290481567383,
      "learning_rate": 6.2554585152838425e-06,
      "loss": 4.6235,
      "step": 5800
    },
    {
      "epoch": 6.441048034934497,
      "grad_norm": 36.680030822753906,
      "learning_rate": 5.846069868995633e-06,
      "loss": 4.4681,
      "step": 5900
    },
    {
      "epoch": 6.550218340611353,
      "grad_norm": 37.740535736083984,
      "learning_rate": 5.436681222707424e-06,
      "loss": 4.518,
      "step": 6000
    },
    {
      "epoch": 6.6593886462882095,
      "grad_norm": 33.8775749206543,
      "learning_rate": 5.027292576419214e-06,
      "loss": 4.6268,
      "step": 6100
    },
    {
      "epoch": 6.7685589519650655,
      "grad_norm": 50.033729553222656,
      "learning_rate": 4.617903930131005e-06,
      "loss": 4.6026,
      "step": 6200
    },
    {
      "epoch": 6.877729257641922,
      "grad_norm": 34.454341888427734,
      "learning_rate": 4.208515283842795e-06,
      "loss": 4.618,
      "step": 6300
    },
    {
      "epoch": 6.986899563318778,
      "grad_norm": 41.41761779785156,
      "learning_rate": 3.799126637554585e-06,
      "loss": 4.6299,
      "step": 6400
    },
    {
      "epoch": 7.0,
      "eval_avg_mae": 5.449769337972005,
      "eval_loss": 5.449769496917725,
      "eval_mae_lex": 4.877564430236816,
      "eval_mae_sem": 3.8569977283477783,
      "eval_mae_syn": 7.614744663238525,
      "eval_runtime": 27.0562,
      "eval_samples_per_second": 270.806,
      "eval_steps_per_second": 8.464,
      "step": 6412
    }
  ],
  "logging_steps": 100,
  "max_steps": 7328,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.3494363633370368e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}