{
  "best_metric": 0.010238240472972393,
  "best_model_checkpoint": "./results/checkpoint-14000",
  "epoch": 0.8173273395995097,
  "global_step": 14000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "learning_rate": 1.9708097378714462e-05,
      "loss": 0.014,
      "step": 500
    },
    {
      "epoch": 0.03,
      "eval_accuracy": 0.9974897107329461,
      "eval_loss": 0.014051680453121662,
      "eval_runtime": 272.9865,
      "eval_samples_per_second": 125.497,
      "eval_steps_per_second": 15.689,
      "step": 500
    },
    {
      "epoch": 0.06,
      "learning_rate": 1.9416194757428922e-05,
      "loss": 0.0068,
      "step": 1000
    },
    {
      "epoch": 0.06,
      "eval_accuracy": 0.9936658980122012,
      "eval_loss": 0.03880644589662552,
      "eval_runtime": 271.4393,
      "eval_samples_per_second": 126.212,
      "eval_steps_per_second": 15.779,
      "step": 1000
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.9124292136143383e-05,
      "loss": 0.0065,
      "step": 1500
    },
    {
      "epoch": 0.09,
      "eval_accuracy": 0.9963221343296652,
      "eval_loss": 0.034014396369457245,
      "eval_runtime": 271.2782,
      "eval_samples_per_second": 126.287,
      "eval_steps_per_second": 15.788,
      "step": 1500
    },
    {
      "epoch": 0.12,
      "learning_rate": 1.8832389514857846e-05,
      "loss": 0.0088,
      "step": 2000
    },
    {
      "epoch": 0.12,
      "eval_accuracy": 0.9975189001430281,
      "eval_loss": 0.01912350207567215,
      "eval_runtime": 271.5387,
      "eval_samples_per_second": 126.166,
      "eval_steps_per_second": 15.773,
      "step": 2000
    },
    {
      "epoch": 0.15,
      "learning_rate": 1.8540486893572303e-05,
      "loss": 0.0081,
      "step": 2500
    },
    {
      "epoch": 0.15,
      "eval_accuracy": 0.9977816048337663,
      "eval_loss": 0.01667490415275097,
      "eval_runtime": 271.3282,
      "eval_samples_per_second": 126.264,
      "eval_steps_per_second": 15.785,
      "step": 2500
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.8248584272286767e-05,
      "loss": 0.0142,
      "step": 3000
    },
    {
      "epoch": 0.18,
      "eval_accuracy": 0.9969351119413876,
      "eval_loss": 0.017350222915410995,
      "eval_runtime": 271.4826,
      "eval_samples_per_second": 126.192,
      "eval_steps_per_second": 15.776,
      "step": 3000
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.7956681651001227e-05,
      "loss": 0.0066,
      "step": 3500
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.9967891648909776,
      "eval_loss": 0.025246594101190567,
      "eval_runtime": 271.289,
      "eval_samples_per_second": 126.282,
      "eval_steps_per_second": 15.788,
      "step": 3500
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.7664779029715688e-05,
      "loss": 0.0078,
      "step": 4000
    },
    {
      "epoch": 0.23,
      "eval_accuracy": 0.9965848390204034,
      "eval_loss": 0.02241995558142662,
      "eval_runtime": 271.2726,
      "eval_samples_per_second": 126.29,
      "eval_steps_per_second": 15.789,
      "step": 4000
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.7372876408430148e-05,
      "loss": 0.0062,
      "step": 4500
    },
    {
      "epoch": 0.26,
      "eval_accuracy": 0.9968183543010596,
      "eval_loss": 0.021609827876091003,
      "eval_runtime": 271.3057,
      "eval_samples_per_second": 126.275,
      "eval_steps_per_second": 15.787,
      "step": 4500
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.7080973787144612e-05,
      "loss": 0.0047,
      "step": 5000
    },
    {
      "epoch": 0.29,
      "eval_accuracy": 0.9971978166321258,
      "eval_loss": 0.01993330754339695,
      "eval_runtime": 271.1898,
      "eval_samples_per_second": 126.328,
      "eval_steps_per_second": 15.793,
      "step": 5000
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.678907116585907e-05,
      "loss": 0.0035,
      "step": 5500
    },
    {
      "epoch": 0.32,
      "eval_accuracy": 0.9968767331212236,
      "eval_loss": 0.025054221972823143,
      "eval_runtime": 272.0376,
      "eval_samples_per_second": 125.935,
      "eval_steps_per_second": 15.744,
      "step": 5500
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.6497168544573532e-05,
      "loss": 0.0147,
      "step": 6000
    },
    {
      "epoch": 0.35,
      "eval_accuracy": 0.997343763682536,
      "eval_loss": 0.017792223021388054,
      "eval_runtime": 271.3522,
      "eval_samples_per_second": 126.253,
      "eval_steps_per_second": 15.784,
      "step": 6000
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.6205265923287993e-05,
      "loss": 0.0093,
      "step": 6500
    },
    {
      "epoch": 0.38,
      "eval_accuracy": 0.9971102484018798,
      "eval_loss": 0.019385505467653275,
      "eval_runtime": 270.7817,
      "eval_samples_per_second": 126.519,
      "eval_steps_per_second": 15.817,
      "step": 6500
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.5913363302002453e-05,
      "loss": 0.0084,
      "step": 7000
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.9974897107329461,
      "eval_loss": 0.01369693223387003,
      "eval_runtime": 270.7077,
      "eval_samples_per_second": 126.553,
      "eval_steps_per_second": 15.821,
      "step": 7000
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.5621460680716913e-05,
      "loss": 0.0135,
      "step": 7500
    },
    {
      "epoch": 0.44,
      "eval_accuracy": 0.9967307860708136,
      "eval_loss": 0.023163480684161186,
      "eval_runtime": 270.9905,
      "eval_samples_per_second": 126.421,
      "eval_steps_per_second": 15.805,
      "step": 7500
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.5329558059431374e-05,
      "loss": 0.0131,
      "step": 8000
    },
    {
      "epoch": 0.47,
      "eval_accuracy": 0.9974605213228641,
      "eval_loss": 0.015544239431619644,
      "eval_runtime": 270.8012,
      "eval_samples_per_second": 126.51,
      "eval_steps_per_second": 15.816,
      "step": 8000
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.5037655438145836e-05,
      "loss": 0.0196,
      "step": 8500
    },
    {
      "epoch": 0.5,
      "eval_accuracy": 0.9965556496103214,
      "eval_loss": 0.01619912125170231,
      "eval_runtime": 272.0032,
      "eval_samples_per_second": 125.951,
      "eval_steps_per_second": 15.746,
      "step": 8500
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.4745752816860298e-05,
      "loss": 0.0207,
      "step": 9000
    },
    {
      "epoch": 0.53,
      "eval_accuracy": 0.9978691730640123,
      "eval_loss": 0.011031306348741055,
      "eval_runtime": 270.862,
      "eval_samples_per_second": 126.481,
      "eval_steps_per_second": 15.812,
      "step": 9000
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.4453850195574756e-05,
      "loss": 0.0172,
      "step": 9500
    },
    {
      "epoch": 0.55,
      "eval_accuracy": 0.9969059225313056,
      "eval_loss": 0.013371887616813183,
      "eval_runtime": 270.9823,
      "eval_samples_per_second": 126.425,
      "eval_steps_per_second": 15.805,
      "step": 9500
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.4161947574289218e-05,
      "loss": 0.0147,
      "step": 10000
    },
    {
      "epoch": 0.58,
      "eval_accuracy": 0.9956799673078607,
      "eval_loss": 0.018292322754859924,
      "eval_runtime": 271.0579,
      "eval_samples_per_second": 126.39,
      "eval_steps_per_second": 15.801,
      "step": 10000
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.3870044953003679e-05,
      "loss": 0.0115,
      "step": 10500
    },
    {
      "epoch": 0.61,
      "eval_accuracy": 0.9976356577833562,
      "eval_loss": 0.013687117025256157,
      "eval_runtime": 270.8313,
      "eval_samples_per_second": 126.496,
      "eval_steps_per_second": 15.814,
      "step": 10500
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.357814233171814e-05,
      "loss": 0.0058,
      "step": 11000
    },
    {
      "epoch": 0.64,
      "eval_accuracy": 0.9978399836539303,
      "eval_loss": 0.01474874746054411,
      "eval_runtime": 270.6785,
      "eval_samples_per_second": 126.567,
      "eval_steps_per_second": 15.823,
      "step": 11000
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.32862397104326e-05,
      "loss": 0.0127,
      "step": 11500
    },
    {
      "epoch": 0.67,
      "eval_accuracy": 0.9975189001430281,
      "eval_loss": 0.014400546438992023,
      "eval_runtime": 271.0019,
      "eval_samples_per_second": 126.416,
      "eval_steps_per_second": 15.804,
      "step": 11500
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.2994337089147061e-05,
      "loss": 0.0158,
      "step": 12000
    },
    {
      "epoch": 0.7,
      "eval_accuracy": 0.9978399836539303,
      "eval_loss": 0.011529939249157906,
      "eval_runtime": 271.0144,
      "eval_samples_per_second": 126.41,
      "eval_steps_per_second": 15.804,
      "step": 12000
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.2702434467861522e-05,
      "loss": 0.0149,
      "step": 12500
    },
    {
      "epoch": 0.73,
      "eval_accuracy": 0.9978983624740944,
      "eval_loss": 0.011057616211473942,
      "eval_runtime": 270.832,
      "eval_samples_per_second": 126.495,
      "eval_steps_per_second": 15.814,
      "step": 12500
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.2410531846575984e-05,
      "loss": 0.0112,
      "step": 13000
    },
    {
      "epoch": 0.76,
      "eval_accuracy": 0.9975772789631921,
      "eval_loss": 0.013288214802742004,
      "eval_runtime": 270.8227,
      "eval_samples_per_second": 126.5,
      "eval_steps_per_second": 15.815,
      "step": 13000
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.2118629225290444e-05,
      "loss": 0.0163,
      "step": 13500
    },
    {
      "epoch": 0.79,
      "eval_accuracy": 0.997372953092618,
      "eval_loss": 0.012437746860086918,
      "eval_runtime": 270.7754,
      "eval_samples_per_second": 126.522,
      "eval_steps_per_second": 15.818,
      "step": 13500
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.1826726604004906e-05,
      "loss": 0.0107,
      "step": 14000
    },
    {
      "epoch": 0.82,
      "eval_accuracy": 0.9979859307043405,
      "eval_loss": 0.010238240472972393,
      "eval_runtime": 271.1151,
      "eval_samples_per_second": 126.363,
      "eval_steps_per_second": 15.798,
      "step": 14000
    }
  ],
  "max_steps": 34258,
  "num_train_epochs": 2,
  "total_flos": 1.4836348649472e+16,
  "trial_name": null,
  "trial_params": null
}