{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 50.0,
  "eval_steps": 500,
  "global_step": 12950,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.9305019305019306,
      "grad_norm": 2.738110065460205,
      "learning_rate": 9.613899613899614e-06,
      "loss": 0.0888,
      "step": 500
    },
    {
      "epoch": 1.9305019305019306,
      "eval_runtime": 667.5755,
      "eval_samples_per_second": 264.418,
      "eval_steps_per_second": 16.527,
      "step": 500
    },
    {
      "epoch": 3.861003861003861,
      "grad_norm": 4.450708389282227,
      "learning_rate": 9.227799227799229e-06,
      "loss": 0.0558,
      "step": 1000
    },
    {
      "epoch": 3.861003861003861,
      "eval_runtime": 651.8038,
      "eval_samples_per_second": 270.816,
      "eval_steps_per_second": 16.927,
      "step": 1000
    },
    {
      "epoch": 5.7915057915057915,
      "grad_norm": 5.049165725708008,
      "learning_rate": 8.841698841698842e-06,
      "loss": 0.0409,
      "step": 1500
    },
    {
      "epoch": 5.7915057915057915,
      "eval_runtime": 650.4821,
      "eval_samples_per_second": 271.366,
      "eval_steps_per_second": 16.961,
      "step": 1500
    },
    {
      "epoch": 7.722007722007722,
      "grad_norm": 3.824125051498413,
      "learning_rate": 8.455598455598457e-06,
      "loss": 0.0319,
      "step": 2000
    },
    {
      "epoch": 7.722007722007722,
      "eval_runtime": 652.7164,
      "eval_samples_per_second": 270.438,
      "eval_steps_per_second": 16.903,
      "step": 2000
    },
    {
      "epoch": 9.652509652509652,
      "grad_norm": 3.7455878257751465,
      "learning_rate": 8.06949806949807e-06,
      "loss": 0.0277,
      "step": 2500
    },
    {
      "epoch": 9.652509652509652,
      "eval_runtime": 652.469,
      "eval_samples_per_second": 270.54,
      "eval_steps_per_second": 16.91,
      "step": 2500
    },
    {
      "epoch": 11.583011583011583,
      "grad_norm": 2.825524091720581,
      "learning_rate": 7.683397683397685e-06,
      "loss": 0.0249,
      "step": 3000
    },
    {
      "epoch": 11.583011583011583,
      "eval_runtime": 724.8262,
      "eval_samples_per_second": 243.533,
      "eval_steps_per_second": 15.222,
      "step": 3000
    },
    {
      "epoch": 13.513513513513514,
      "grad_norm": 1.5855698585510254,
      "learning_rate": 7.297297297297298e-06,
      "loss": 0.0233,
      "step": 3500
    },
    {
      "epoch": 13.513513513513514,
      "eval_runtime": 721.7805,
      "eval_samples_per_second": 244.56,
      "eval_steps_per_second": 15.286,
      "step": 3500
    },
    {
      "epoch": 15.444015444015443,
      "grad_norm": 2.516690254211426,
      "learning_rate": 6.911196911196911e-06,
      "loss": 0.0217,
      "step": 4000
    },
    {
      "epoch": 15.444015444015443,
      "eval_runtime": 723.8804,
      "eval_samples_per_second": 243.851,
      "eval_steps_per_second": 15.241,
      "step": 4000
    },
    {
      "epoch": 17.374517374517374,
      "grad_norm": 2.6874454021453857,
      "learning_rate": 6.525096525096526e-06,
      "loss": 0.0207,
      "step": 4500
    },
    {
      "epoch": 17.374517374517374,
      "eval_runtime": 726.2467,
      "eval_samples_per_second": 243.057,
      "eval_steps_per_second": 15.192,
      "step": 4500
    },
    {
      "epoch": 19.305019305019304,
      "grad_norm": 2.5405662059783936,
      "learning_rate": 6.13899613899614e-06,
      "loss": 0.0198,
      "step": 5000
    },
    {
      "epoch": 19.305019305019304,
      "eval_runtime": 725.831,
      "eval_samples_per_second": 243.196,
      "eval_steps_per_second": 15.201,
      "step": 5000
    },
    {
      "epoch": 21.235521235521237,
      "grad_norm": 2.5149359703063965,
      "learning_rate": 5.752895752895753e-06,
      "loss": 0.0191,
      "step": 5500
    },
    {
      "epoch": 21.235521235521237,
      "eval_runtime": 724.4353,
      "eval_samples_per_second": 243.664,
      "eval_steps_per_second": 15.23,
      "step": 5500
    },
    {
      "epoch": 23.166023166023166,
      "grad_norm": 1.5971148014068604,
      "learning_rate": 5.366795366795368e-06,
      "loss": 0.0183,
      "step": 6000
    },
    {
      "epoch": 23.166023166023166,
      "eval_runtime": 725.1569,
      "eval_samples_per_second": 243.422,
      "eval_steps_per_second": 15.215,
      "step": 6000
    },
    {
      "epoch": 25.096525096525095,
      "grad_norm": 2.0284860134124756,
      "learning_rate": 4.980694980694981e-06,
      "loss": 0.0178,
      "step": 6500
    },
    {
      "epoch": 25.096525096525095,
      "eval_runtime": 726.8832,
      "eval_samples_per_second": 242.844,
      "eval_steps_per_second": 15.179,
      "step": 6500
    },
    {
      "epoch": 27.027027027027028,
      "grad_norm": 2.005959987640381,
      "learning_rate": 4.594594594594596e-06,
      "loss": 0.0174,
      "step": 7000
    },
    {
      "epoch": 27.027027027027028,
      "eval_runtime": 725.3449,
      "eval_samples_per_second": 243.359,
      "eval_steps_per_second": 15.211,
      "step": 7000
    },
    {
      "epoch": 28.957528957528957,
      "grad_norm": 1.96770441532135,
      "learning_rate": 4.208494208494209e-06,
      "loss": 0.0168,
      "step": 7500
    },
    {
      "epoch": 28.957528957528957,
      "eval_runtime": 765.8326,
      "eval_samples_per_second": 230.493,
      "eval_steps_per_second": 14.407,
      "step": 7500
    },
    {
      "epoch": 30.888030888030887,
      "grad_norm": 1.844897747039795,
      "learning_rate": 3.822393822393823e-06,
      "loss": 0.0165,
      "step": 8000
    },
    {
      "epoch": 30.888030888030887,
      "eval_runtime": 819.9764,
      "eval_samples_per_second": 215.273,
      "eval_steps_per_second": 13.455,
      "step": 8000
    },
    {
      "epoch": 32.818532818532816,
      "grad_norm": 1.615881323814392,
      "learning_rate": 3.4362934362934363e-06,
      "loss": 0.0162,
      "step": 8500
    },
    {
      "epoch": 32.818532818532816,
      "eval_runtime": 821.2267,
      "eval_samples_per_second": 214.946,
      "eval_steps_per_second": 13.435,
      "step": 8500
    },
    {
      "epoch": 34.74903474903475,
      "grad_norm": 1.6435213088989258,
      "learning_rate": 3.0501930501930503e-06,
      "loss": 0.0158,
      "step": 9000
    },
    {
      "epoch": 34.74903474903475,
      "eval_runtime": 819.0429,
      "eval_samples_per_second": 215.519,
      "eval_steps_per_second": 13.471,
      "step": 9000
    },
    {
      "epoch": 36.67953667953668,
      "grad_norm": 1.7934831380844116,
      "learning_rate": 2.6640926640926647e-06,
      "loss": 0.0155,
      "step": 9500
    },
    {
      "epoch": 36.67953667953668,
      "eval_runtime": 820.1169,
      "eval_samples_per_second": 215.236,
      "eval_steps_per_second": 13.453,
      "step": 9500
    },
    {
      "epoch": 38.61003861003861,
      "grad_norm": 1.1239484548568726,
      "learning_rate": 2.2779922779922782e-06,
      "loss": 0.0152,
      "step": 10000
    },
    {
      "epoch": 38.61003861003861,
      "eval_runtime": 821.2669,
      "eval_samples_per_second": 214.935,
      "eval_steps_per_second": 13.434,
      "step": 10000
    },
    {
      "epoch": 40.54054054054054,
      "grad_norm": 1.256516456604004,
      "learning_rate": 1.8918918918918922e-06,
      "loss": 0.015,
      "step": 10500
    },
    {
      "epoch": 40.54054054054054,
      "eval_runtime": 818.9208,
      "eval_samples_per_second": 215.551,
      "eval_steps_per_second": 13.473,
      "step": 10500
    },
    {
      "epoch": 42.47104247104247,
      "grad_norm": 0.4177967607975006,
      "learning_rate": 1.505791505791506e-06,
      "loss": 0.0148,
      "step": 11000
    },
    {
      "epoch": 42.47104247104247,
      "eval_runtime": 823.9199,
      "eval_samples_per_second": 214.243,
      "eval_steps_per_second": 13.391,
      "step": 11000
    },
    {
      "epoch": 44.4015444015444,
      "grad_norm": 0.3761753737926483,
      "learning_rate": 1.1196911196911197e-06,
      "loss": 0.0146,
      "step": 11500
    },
    {
      "epoch": 44.4015444015444,
      "eval_runtime": 821.1987,
      "eval_samples_per_second": 214.953,
      "eval_steps_per_second": 13.435,
      "step": 11500
    },
    {
      "epoch": 46.33204633204633,
      "grad_norm": 0.4973774552345276,
      "learning_rate": 7.335907335907337e-07,
      "loss": 0.0145,
      "step": 12000
    },
    {
      "epoch": 46.33204633204633,
      "eval_runtime": 821.7593,
      "eval_samples_per_second": 214.806,
      "eval_steps_per_second": 13.426,
      "step": 12000
    },
    {
      "epoch": 48.262548262548265,
      "grad_norm": 0.3773488998413086,
      "learning_rate": 3.474903474903475e-07,
      "loss": 0.0144,
      "step": 12500
    },
    {
      "epoch": 48.262548262548265,
      "eval_runtime": 824.3743,
      "eval_samples_per_second": 214.125,
      "eval_steps_per_second": 13.383,
      "step": 12500
    },
    {
      "epoch": 50.0,
      "step": 12950,
      "total_flos": 5.544539360447693e+19,
      "train_loss": 0.023562463738283135,
      "train_runtime": 98807.7466,
      "train_samples_per_second": 267.972,
      "train_steps_per_second": 0.131
    }
  ],
  "logging_steps": 500,
  "max_steps": 12950,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 100,
  "total_flos": 5.544539360447693e+19,
  "train_batch_size": 2048,
  "trial_name": null,
  "trial_params": null
}