{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 31710,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0946073793755913,
      "grad_norm": 26.54265022277832,
      "learning_rate": 2.9526963103122042e-05,
      "loss": 1.3126,
      "step": 500
    },
    {
      "epoch": 0.1892147587511826,
      "grad_norm": 16.264204025268555,
      "learning_rate": 2.905392620624409e-05,
      "loss": 0.9908,
      "step": 1000
    },
    {
      "epoch": 0.28382213812677387,
      "grad_norm": 15.488258361816406,
      "learning_rate": 2.8580889309366132e-05,
      "loss": 0.924,
      "step": 1500
    },
    {
      "epoch": 0.3784295175023652,
      "grad_norm": 17.02219581604004,
      "learning_rate": 2.8107852412488173e-05,
      "loss": 0.8811,
      "step": 2000
    },
    {
      "epoch": 0.47303689687795647,
      "grad_norm": 20.8419132232666,
      "learning_rate": 2.763481551561022e-05,
      "loss": 0.8707,
      "step": 2500
    },
    {
      "epoch": 0.5676442762535477,
      "grad_norm": 14.816078186035156,
      "learning_rate": 2.7161778618732263e-05,
      "loss": 0.8418,
      "step": 3000
    },
    {
      "epoch": 0.6622516556291391,
      "grad_norm": 16.095823287963867,
      "learning_rate": 2.6688741721854305e-05,
      "loss": 0.8021,
      "step": 3500
    },
    {
      "epoch": 0.7568590350047304,
      "grad_norm": 16.451648712158203,
      "learning_rate": 2.6215704824976346e-05,
      "loss": 0.8166,
      "step": 4000
    },
    {
      "epoch": 0.8514664143803217,
      "grad_norm": 23.100683212280273,
      "learning_rate": 2.5742667928098394e-05,
      "loss": 0.7894,
      "step": 4500
    },
    {
      "epoch": 0.9460737937559129,
      "grad_norm": 16.882429122924805,
      "learning_rate": 2.5269631031220436e-05,
      "loss": 0.7754,
      "step": 5000
    },
    {
      "epoch": 1.0406811731315042,
      "grad_norm": 14.48630142211914,
      "learning_rate": 2.4796594134342477e-05,
      "loss": 0.7216,
      "step": 5500
    },
    {
      "epoch": 1.1352885525070955,
      "grad_norm": 16.525306701660156,
      "learning_rate": 2.4323557237464526e-05,
      "loss": 0.6325,
      "step": 6000
    },
    {
      "epoch": 1.2298959318826868,
      "grad_norm": 19.755531311035156,
      "learning_rate": 2.3850520340586567e-05,
      "loss": 0.6503,
      "step": 6500
    },
    {
      "epoch": 1.3245033112582782,
      "grad_norm": 9.933408737182617,
      "learning_rate": 2.337748344370861e-05,
      "loss": 0.6369,
      "step": 7000
    },
    {
      "epoch": 1.4191106906338695,
      "grad_norm": 21.490463256835938,
      "learning_rate": 2.2904446546830654e-05,
      "loss": 0.6411,
      "step": 7500
    },
    {
      "epoch": 1.5137180700094608,
      "grad_norm": 16.791980743408203,
      "learning_rate": 2.24314096499527e-05,
      "loss": 0.6319,
      "step": 8000
    },
    {
      "epoch": 1.608325449385052,
      "grad_norm": 14.076953887939453,
      "learning_rate": 2.195837275307474e-05,
      "loss": 0.6282,
      "step": 8500
    },
    {
      "epoch": 1.7029328287606433,
      "grad_norm": 23.48201560974121,
      "learning_rate": 2.148533585619678e-05,
      "loss": 0.6325,
      "step": 9000
    },
    {
      "epoch": 1.7975402081362346,
      "grad_norm": 32.473262786865234,
      "learning_rate": 2.101229895931883e-05,
      "loss": 0.6411,
      "step": 9500
    },
    {
      "epoch": 1.8921475875118259,
      "grad_norm": 17.878799438476562,
      "learning_rate": 2.053926206244087e-05,
      "loss": 0.6314,
      "step": 10000
    },
    {
      "epoch": 1.9867549668874172,
      "grad_norm": 11.421870231628418,
      "learning_rate": 2.0066225165562913e-05,
      "loss": 0.6474,
      "step": 10500
    },
    {
      "epoch": 2.0813623462630084,
      "grad_norm": 23.016128540039062,
      "learning_rate": 1.959318826868496e-05,
      "loss": 0.4893,
      "step": 11000
    },
    {
      "epoch": 2.1759697256385997,
      "grad_norm": 14.312499046325684,
      "learning_rate": 1.9120151371807002e-05,
      "loss": 0.4706,
      "step": 11500
    },
    {
      "epoch": 2.270577105014191,
      "grad_norm": 16.817712783813477,
      "learning_rate": 1.8647114474929044e-05,
      "loss": 0.4662,
      "step": 12000
    },
    {
      "epoch": 2.3651844843897822,
      "grad_norm": 25.88548469543457,
      "learning_rate": 1.8174077578051085e-05,
      "loss": 0.4881,
      "step": 12500
    },
    {
      "epoch": 2.4597918637653735,
      "grad_norm": 12.041430473327637,
      "learning_rate": 1.7701040681173134e-05,
      "loss": 0.4904,
      "step": 13000
    },
    {
      "epoch": 2.5543992431409652,
      "grad_norm": 14.665155410766602,
      "learning_rate": 1.7228003784295175e-05,
      "loss": 0.4981,
      "step": 13500
    },
    {
      "epoch": 2.6490066225165565,
      "grad_norm": 29.060611724853516,
      "learning_rate": 1.6754966887417217e-05,
      "loss": 0.4832,
      "step": 14000
    },
    {
      "epoch": 2.7436140018921478,
      "grad_norm": 20.540754318237305,
      "learning_rate": 1.6281929990539265e-05,
      "loss": 0.4713,
      "step": 14500
    },
    {
      "epoch": 2.838221381267739,
      "grad_norm": 24.385042190551758,
      "learning_rate": 1.5808893093661306e-05,
      "loss": 0.4775,
      "step": 15000
    },
    {
      "epoch": 2.9328287606433303,
      "grad_norm": 12.536005020141602,
      "learning_rate": 1.5335856196783348e-05,
      "loss": 0.4776,
      "step": 15500
    },
    {
      "epoch": 3.0274361400189216,
      "grad_norm": 14.910382270812988,
      "learning_rate": 1.4862819299905393e-05,
      "loss": 0.4355,
      "step": 16000
    },
    {
      "epoch": 3.122043519394513,
      "grad_norm": 14.342169761657715,
      "learning_rate": 1.4389782403027438e-05,
      "loss": 0.3343,
      "step": 16500
    },
    {
      "epoch": 3.216650898770104,
      "grad_norm": 16.1813907623291,
      "learning_rate": 1.3916745506149479e-05,
      "loss": 0.3452,
      "step": 17000
    },
    {
      "epoch": 3.3112582781456954,
      "grad_norm": 9.025822639465332,
      "learning_rate": 1.3443708609271524e-05,
      "loss": 0.3522,
      "step": 17500
    },
    {
      "epoch": 3.4058656575212867,
      "grad_norm": 25.14023780822754,
      "learning_rate": 1.2970671712393567e-05,
      "loss": 0.3408,
      "step": 18000
    },
    {
      "epoch": 3.500473036896878,
      "grad_norm": 29.443084716796875,
      "learning_rate": 1.249763481551561e-05,
      "loss": 0.3414,
      "step": 18500
    },
    {
      "epoch": 3.595080416272469,
      "grad_norm": 30.374174118041992,
      "learning_rate": 1.2024597918637655e-05,
      "loss": 0.3511,
      "step": 19000
    },
    {
      "epoch": 3.6896877956480605,
      "grad_norm": 19.67069435119629,
      "learning_rate": 1.1551561021759697e-05,
      "loss": 0.3505,
      "step": 19500
    },
    {
      "epoch": 3.7842951750236518,
      "grad_norm": 14.601593971252441,
      "learning_rate": 1.1078524124881742e-05,
      "loss": 0.3415,
      "step": 20000
    },
    {
      "epoch": 3.878902554399243,
      "grad_norm": 19.330163955688477,
      "learning_rate": 1.0605487228003785e-05,
      "loss": 0.3509,
      "step": 20500
    },
    {
      "epoch": 3.9735099337748343,
      "grad_norm": 19.882015228271484,
      "learning_rate": 1.0132450331125828e-05,
      "loss": 0.3486,
      "step": 21000
    },
    {
      "epoch": 4.068117313150426,
      "grad_norm": 8.843262672424316,
      "learning_rate": 9.659413434247873e-06,
      "loss": 0.259,
      "step": 21500
    },
    {
      "epoch": 4.162724692526017,
      "grad_norm": 38.07181930541992,
      "learning_rate": 9.186376537369914e-06,
      "loss": 0.2325,
      "step": 22000
    },
    {
      "epoch": 4.257332071901608,
      "grad_norm": 16.061630249023438,
      "learning_rate": 8.713339640491959e-06,
      "loss": 0.2346,
      "step": 22500
    },
    {
      "epoch": 4.351939451277199,
      "grad_norm": 14.632463455200195,
      "learning_rate": 8.240302743614002e-06,
      "loss": 0.2365,
      "step": 23000
    },
    {
      "epoch": 4.446546830652791,
      "grad_norm": 9.404703140258789,
      "learning_rate": 7.767265846736046e-06,
      "loss": 0.2347,
      "step": 23500
    },
    {
      "epoch": 4.541154210028382,
      "grad_norm": 6.847620964050293,
      "learning_rate": 7.2942289498580895e-06,
      "loss": 0.2326,
      "step": 24000
    },
    {
      "epoch": 4.635761589403973,
      "grad_norm": 14.490519523620605,
      "learning_rate": 6.821192052980133e-06,
      "loss": 0.2321,
      "step": 24500
    },
    {
      "epoch": 4.7303689687795645,
      "grad_norm": 27.82660675048828,
      "learning_rate": 6.348155156102176e-06,
      "loss": 0.2459,
      "step": 25000
    },
    {
      "epoch": 4.824976348155156,
      "grad_norm": 22.924592971801758,
      "learning_rate": 5.87511825922422e-06,
      "loss": 0.2418,
      "step": 25500
    },
    {
      "epoch": 4.919583727530747,
      "grad_norm": 22.808650970458984,
      "learning_rate": 5.402081362346263e-06,
      "loss": 0.2272,
      "step": 26000
    },
    {
      "epoch": 5.014191106906338,
      "grad_norm": 8.198173522949219,
      "learning_rate": 4.929044465468307e-06,
      "loss": 0.2201,
      "step": 26500
    },
    {
      "epoch": 5.10879848628193,
      "grad_norm": 20.825037002563477,
      "learning_rate": 4.45600756859035e-06,
      "loss": 0.1464,
      "step": 27000
    },
    {
      "epoch": 5.203405865657521,
      "grad_norm": 21.507328033447266,
      "learning_rate": 3.9829706717123935e-06,
      "loss": 0.1451,
      "step": 27500
    },
    {
      "epoch": 5.298013245033113,
      "grad_norm": 11.167850494384766,
      "learning_rate": 3.509933774834437e-06,
      "loss": 0.1417,
      "step": 28000
    },
    {
      "epoch": 5.392620624408704,
      "grad_norm": 20.28238868713379,
      "learning_rate": 3.0368968779564807e-06,
      "loss": 0.1527,
      "step": 28500
    },
    {
      "epoch": 5.4872280037842955,
      "grad_norm": 13.64089584350586,
      "learning_rate": 2.5638599810785243e-06,
      "loss": 0.148,
      "step": 29000
    },
    {
      "epoch": 5.581835383159887,
      "grad_norm": 16.22063446044922,
      "learning_rate": 2.0908230842005675e-06,
      "loss": 0.1446,
      "step": 29500
    },
    {
      "epoch": 5.676442762535478,
      "grad_norm": 25.130605697631836,
      "learning_rate": 1.6177861873226113e-06,
      "loss": 0.1401,
      "step": 30000
    },
    {
      "epoch": 5.771050141911069,
      "grad_norm": 8.35947036743164,
      "learning_rate": 1.1447492904446547e-06,
      "loss": 0.1488,
      "step": 30500
    },
    {
      "epoch": 5.865657521286661,
      "grad_norm": 15.456842422485352,
      "learning_rate": 6.717123935666982e-07,
      "loss": 0.1388,
      "step": 31000
    },
    {
      "epoch": 5.960264900662252,
      "grad_norm": 25.6274356842041,
      "learning_rate": 1.9867549668874173e-07,
      "loss": 0.1381,
      "step": 31500
    },
    {
      "epoch": 6.0,
      "step": 31710,
      "total_flos": 1.6805564390366208e+16,
      "train_loss": 0.4557977186772621,
      "train_runtime": 8236.6278,
      "train_samples_per_second": 123.188,
      "train_steps_per_second": 3.85
    }
  ],
  "logging_steps": 500,
  "max_steps": 31710,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.6805564390366208e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}