{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 7092,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.021150592216582064,
      "grad_norm": 8.332942008972168,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 11.4949,
      "step": 100
    },
    {
      "epoch": 0.04230118443316413,
      "grad_norm": 6.6360764503479,
      "learning_rate": 3.266666666666667e-05,
      "loss": 9.8459,
      "step": 200
    },
    {
      "epoch": 0.06345177664974619,
      "grad_norm": 8.05850601196289,
      "learning_rate": 4.933333333333334e-05,
      "loss": 8.5587,
      "step": 300
    },
    {
      "epoch": 0.08460236886632826,
      "grad_norm": 8.233019828796387,
      "learning_rate": 4.9654278305963706e-05,
      "loss": 7.976,
      "step": 400
    },
    {
      "epoch": 0.10575296108291032,
      "grad_norm": 7.412795066833496,
      "learning_rate": 4.929415154134255e-05,
      "loss": 7.5136,
      "step": 500
    },
    {
      "epoch": 0.12690355329949238,
      "grad_norm": 8.026376724243164,
      "learning_rate": 4.893402477672141e-05,
      "loss": 7.2849,
      "step": 600
    },
    {
      "epoch": 0.14805414551607446,
      "grad_norm": 13.290921211242676,
      "learning_rate": 4.857389801210026e-05,
      "loss": 7.1981,
      "step": 700
    },
    {
      "epoch": 0.1692047377326565,
      "grad_norm": 9.613779067993164,
      "learning_rate": 4.821377124747912e-05,
      "loss": 6.9762,
      "step": 800
    },
    {
      "epoch": 0.19035532994923857,
      "grad_norm": 10.849081993103027,
      "learning_rate": 4.785364448285797e-05,
      "loss": 6.986,
      "step": 900
    },
    {
      "epoch": 0.21150592216582065,
      "grad_norm": 9.172639846801758,
      "learning_rate": 4.7493517718236826e-05,
      "loss": 6.9641,
      "step": 1000
    },
    {
      "epoch": 0.2326565143824027,
      "grad_norm": 13.746626853942871,
      "learning_rate": 4.713339095361568e-05,
      "loss": 6.7526,
      "step": 1100
    },
    {
      "epoch": 0.25380710659898476,
      "grad_norm": 9.715788841247559,
      "learning_rate": 4.677326418899453e-05,
      "loss": 6.7382,
      "step": 1200
    },
    {
      "epoch": 0.2749576988155668,
      "grad_norm": 8.96385383605957,
      "learning_rate": 4.641313742437338e-05,
      "loss": 6.6046,
      "step": 1300
    },
    {
      "epoch": 0.2961082910321489,
      "grad_norm": 7.643514156341553,
      "learning_rate": 4.605301065975223e-05,
      "loss": 6.4929,
      "step": 1400
    },
    {
      "epoch": 0.31725888324873097,
      "grad_norm": 8.421730041503906,
      "learning_rate": 4.569288389513109e-05,
      "loss": 6.4243,
      "step": 1500
    },
    {
      "epoch": 0.338409475465313,
      "grad_norm": 21.085466384887695,
      "learning_rate": 4.533275713050994e-05,
      "loss": 6.3454,
      "step": 1600
    },
    {
      "epoch": 0.3595600676818951,
      "grad_norm": 9.470216751098633,
      "learning_rate": 4.4972630365888796e-05,
      "loss": 6.2785,
      "step": 1700
    },
    {
      "epoch": 0.38071065989847713,
      "grad_norm": 9.164347648620605,
      "learning_rate": 4.461250360126765e-05,
      "loss": 6.4042,
      "step": 1800
    },
    {
      "epoch": 0.40186125211505924,
      "grad_norm": 8.708257675170898,
      "learning_rate": 4.4252376836646505e-05,
      "loss": 6.1933,
      "step": 1900
    },
    {
      "epoch": 0.4230118443316413,
      "grad_norm": 10.921855926513672,
      "learning_rate": 4.3892250072025356e-05,
      "loss": 5.9611,
      "step": 2000
    },
    {
      "epoch": 0.8883248730964467,
      "grad_norm": 6.086740493774414,
      "learning_rate": 3.677856301531213e-05,
      "loss": 5.9772,
      "step": 2100
    },
    {
      "epoch": 0.9306260575296108,
      "grad_norm": 5.576613426208496,
      "learning_rate": 3.604240282685513e-05,
      "loss": 5.8948,
      "step": 2200
    },
    {
      "epoch": 0.9729272419627749,
      "grad_norm": 6.084731578826904,
      "learning_rate": 3.530624263839812e-05,
      "loss": 5.8356,
      "step": 2300
    },
    {
      "epoch": 1.015228426395939,
      "grad_norm": 6.4574198722839355,
      "learning_rate": 3.457008244994111e-05,
      "loss": 5.8376,
      "step": 2400
    },
    {
      "epoch": 1.0575296108291032,
      "grad_norm": 6.7739691734313965,
      "learning_rate": 3.38339222614841e-05,
      "loss": 5.7059,
      "step": 2500
    },
    {
      "epoch": 1.0998307952622675,
      "grad_norm": 6.258169174194336,
      "learning_rate": 3.30977620730271e-05,
      "loss": 5.7126,
      "step": 2600
    },
    {
      "epoch": 1.1421319796954315,
      "grad_norm": 6.656664848327637,
      "learning_rate": 3.236160188457009e-05,
      "loss": 5.6466,
      "step": 2700
    },
    {
      "epoch": 1.1844331641285957,
      "grad_norm": 9.155718803405762,
      "learning_rate": 3.162544169611308e-05,
      "loss": 5.5292,
      "step": 2800
    },
    {
      "epoch": 1.2267343485617597,
      "grad_norm": 6.5999345779418945,
      "learning_rate": 3.088928150765606e-05,
      "loss": 5.4244,
      "step": 2900
    },
    {
      "epoch": 1.2690355329949239,
      "grad_norm": 7.881080150604248,
      "learning_rate": 3.0153121319199055e-05,
      "loss": 5.4133,
      "step": 3000
    },
    {
      "epoch": 1.3113367174280879,
      "grad_norm": 6.419029235839844,
      "learning_rate": 2.941696113074205e-05,
      "loss": 5.42,
      "step": 3100
    },
    {
      "epoch": 1.353637901861252,
      "grad_norm": 7.26897668838501,
      "learning_rate": 2.868080094228504e-05,
      "loss": 5.3998,
      "step": 3200
    },
    {
      "epoch": 1.3959390862944163,
      "grad_norm": 7.908596515655518,
      "learning_rate": 2.7944640753828033e-05,
      "loss": 5.3602,
      "step": 3300
    },
    {
      "epoch": 1.4382402707275803,
      "grad_norm": 6.74896240234375,
      "learning_rate": 2.7208480565371023e-05,
      "loss": 5.341,
      "step": 3400
    },
    {
      "epoch": 1.4805414551607445,
      "grad_norm": 6.694516658782959,
      "learning_rate": 2.6472320376914017e-05,
      "loss": 5.3151,
      "step": 3500
    },
    {
      "epoch": 1.5228426395939088,
      "grad_norm": 8.561161994934082,
      "learning_rate": 2.5736160188457007e-05,
      "loss": 5.242,
      "step": 3600
    },
    {
      "epoch": 1.5651438240270727,
      "grad_norm": 6.73771333694458,
      "learning_rate": 2.5e-05,
      "loss": 5.3066,
      "step": 3700
    },
    {
      "epoch": 1.6074450084602367,
      "grad_norm": 5.139662265777588,
      "learning_rate": 2.4263839811542995e-05,
      "loss": 5.2239,
      "step": 3800
    },
    {
      "epoch": 1.649746192893401,
      "grad_norm": 9.075642585754395,
      "learning_rate": 2.3527679623085985e-05,
      "loss": 5.2235,
      "step": 3900
    },
    {
      "epoch": 1.6920473773265652,
      "grad_norm": 7.239570140838623,
      "learning_rate": 2.279151943462898e-05,
      "loss": 5.1589,
      "step": 4000
    },
    {
      "epoch": 1.7343485617597292,
      "grad_norm": 6.387134552001953,
      "learning_rate": 2.205535924617197e-05,
      "loss": 5.0827,
      "step": 4100
    },
    {
      "epoch": 1.7766497461928934,
      "grad_norm": 6.9777913093566895,
      "learning_rate": 2.131919905771496e-05,
      "loss": 5.0286,
      "step": 4200
    },
    {
      "epoch": 1.8189509306260576,
      "grad_norm": 6.586590766906738,
      "learning_rate": 2.058303886925795e-05,
      "loss": 5.0558,
      "step": 4300
    },
    {
      "epoch": 1.8612521150592216,
      "grad_norm": 8.922995567321777,
      "learning_rate": 1.9846878680800944e-05,
      "loss": 5.0571,
      "step": 4400
    },
    {
      "epoch": 1.9035532994923858,
      "grad_norm": 15.648545265197754,
      "learning_rate": 1.9110718492343934e-05,
      "loss": 5.0074,
      "step": 4500
    },
    {
      "epoch": 1.94585448392555,
      "grad_norm": 9.67465877532959,
      "learning_rate": 1.8374558303886928e-05,
      "loss": 4.9584,
      "step": 4600
    },
    {
      "epoch": 1.988155668358714,
      "grad_norm": 8.49885082244873,
      "learning_rate": 1.7638398115429918e-05,
      "loss": 4.9303,
      "step": 4700
    },
    {
      "epoch": 2.030456852791878,
      "grad_norm": 7.5250396728515625,
      "learning_rate": 1.690223792697291e-05,
      "loss": 4.8864,
      "step": 4800
    },
    {
      "epoch": 2.0727580372250425,
      "grad_norm": 6.662505626678467,
      "learning_rate": 1.6166077738515902e-05,
      "loss": 4.8241,
      "step": 4900
    },
    {
      "epoch": 2.1150592216582065,
      "grad_norm": 13.685736656188965,
      "learning_rate": 1.5429917550058892e-05,
      "loss": 4.8014,
      "step": 5000
    },
    {
      "epoch": 2.1573604060913705,
      "grad_norm": 6.445871353149414,
      "learning_rate": 1.4693757361601884e-05,
      "loss": 4.8768,
      "step": 5100
    },
    {
      "epoch": 2.199661590524535,
      "grad_norm": 6.524866580963135,
      "learning_rate": 1.3957597173144876e-05,
      "loss": 4.8566,
      "step": 5200
    },
    {
      "epoch": 2.241962774957699,
      "grad_norm": 8.739165306091309,
      "learning_rate": 1.3221436984687868e-05,
      "loss": 4.7029,
      "step": 5300
    },
    {
      "epoch": 2.284263959390863,
      "grad_norm": 10.373656272888184,
      "learning_rate": 1.248527679623086e-05,
      "loss": 4.7585,
      "step": 5400
    },
    {
      "epoch": 2.326565143824027,
      "grad_norm": 9.050317764282227,
      "learning_rate": 1.1749116607773852e-05,
      "loss": 4.7496,
      "step": 5500
    },
    {
      "epoch": 2.3688663282571913,
      "grad_norm": 7.840747356414795,
      "learning_rate": 1.1012956419316844e-05,
      "loss": 4.7871,
      "step": 5600
    },
    {
      "epoch": 2.4111675126903553,
      "grad_norm": 7.422430515289307,
      "learning_rate": 1.0276796230859835e-05,
      "loss": 4.7949,
      "step": 5700
    },
    {
      "epoch": 2.4534686971235193,
      "grad_norm": 6.53894567489624,
      "learning_rate": 9.540636042402827e-06,
      "loss": 4.7628,
      "step": 5800
    },
    {
      "epoch": 2.495769881556684,
      "grad_norm": 10.329708099365234,
      "learning_rate": 8.804475853945819e-06,
      "loss": 4.6616,
      "step": 5900
    },
    {
      "epoch": 2.5380710659898478,
      "grad_norm": 8.34196949005127,
      "learning_rate": 8.06831566548881e-06,
      "loss": 4.8091,
      "step": 6000
    },
    {
      "epoch": 2.5803722504230118,
      "grad_norm": 6.8171706199646,
      "learning_rate": 7.332155477031802e-06,
      "loss": 4.6833,
      "step": 6100
    },
    {
      "epoch": 2.6226734348561758,
      "grad_norm": 6.502234935760498,
      "learning_rate": 6.595995288574794e-06,
      "loss": 4.579,
      "step": 6200
    },
    {
      "epoch": 2.66497461928934,
      "grad_norm": 9.155757904052734,
      "learning_rate": 5.859835100117786e-06,
      "loss": 4.6477,
      "step": 6300
    },
    {
      "epoch": 2.707275803722504,
      "grad_norm": 6.970887660980225,
      "learning_rate": 5.123674911660778e-06,
      "loss": 4.6742,
      "step": 6400
    },
    {
      "epoch": 2.749576988155668,
      "grad_norm": 11.221705436706543,
      "learning_rate": 4.387514723203769e-06,
      "loss": 4.6385,
      "step": 6500
    },
    {
      "epoch": 2.7918781725888326,
      "grad_norm": 10.325379371643066,
      "learning_rate": 3.6513545347467615e-06,
      "loss": 4.7299,
      "step": 6600
    },
    {
      "epoch": 2.8341793570219966,
      "grad_norm": 6.159130573272705,
      "learning_rate": 2.9151943462897527e-06,
      "loss": 4.6718,
      "step": 6700
    },
    {
      "epoch": 2.8764805414551606,
      "grad_norm": 7.569215774536133,
      "learning_rate": 2.1790341578327443e-06,
      "loss": 4.6594,
      "step": 6800
    },
    {
      "epoch": 2.9187817258883246,
      "grad_norm": 7.764076232910156,
      "learning_rate": 1.4428739693757363e-06,
      "loss": 4.6538,
      "step": 6900
    },
    {
      "epoch": 2.961082910321489,
      "grad_norm": 7.675158977508545,
      "learning_rate": 7.067137809187279e-07,
      "loss": 4.6145,
      "step": 7000
    }
  ],
  "logging_steps": 100,
  "max_steps": 7092,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 9190341574201800.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}