{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 48.608996539792386,
  "global_step": 7000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 3.47,
      "learning_rate": 0.00015,
      "loss": 5.7552,
      "step": 500
    },
    {
      "epoch": 3.47,
      "eval_cer": 0.9804542678163042,
      "eval_loss": 3.6432340145111084,
      "eval_runtime": 117.6193,
      "eval_samples_per_second": 14.283,
      "step": 500
    },
    {
      "epoch": 6.94,
      "learning_rate": 0.0003,
      "loss": 1.9259,
      "step": 1000
    },
    {
      "epoch": 6.94,
      "eval_cer": 0.1655705336723533,
      "eval_loss": 0.387952595949173,
      "eval_runtime": 117.5642,
      "eval_samples_per_second": 14.29,
      "step": 1000
    },
    {
      "epoch": 10.42,
      "learning_rate": 0.0002758064516129032,
      "loss": 0.372,
      "step": 1500
    },
    {
      "epoch": 10.42,
      "eval_cer": 0.15922101166209748,
      "eval_loss": 0.35147520899772644,
      "eval_runtime": 117.7259,
      "eval_samples_per_second": 14.27,
      "step": 1500
    },
    {
      "epoch": 13.89,
      "learning_rate": 0.00025161290322580645,
      "loss": 0.2859,
      "step": 2000
    },
    {
      "epoch": 13.89,
      "eval_cer": 0.16001647750678277,
      "eval_loss": 0.3656272292137146,
      "eval_runtime": 116.8438,
      "eval_samples_per_second": 14.378,
      "step": 2000
    },
    {
      "epoch": 17.36,
      "learning_rate": 0.00022741935483870966,
      "loss": 0.2384,
      "step": 2500
    },
    {
      "epoch": 17.36,
      "eval_cer": 0.15967556357334622,
      "eval_loss": 0.39938414096832275,
      "eval_runtime": 117.3198,
      "eval_samples_per_second": 14.32,
      "step": 2500
    },
    {
      "epoch": 20.83,
      "learning_rate": 0.00020322580645161287,
      "loss": 0.2047,
      "step": 3000
    },
    {
      "epoch": 20.83,
      "eval_cer": 0.16220400857966732,
      "eval_loss": 0.41744646430015564,
      "eval_runtime": 117.2101,
      "eval_samples_per_second": 14.333,
      "step": 3000
    },
    {
      "epoch": 24.3,
      "learning_rate": 0.00017903225806451613,
      "loss": 0.1699,
      "step": 3500
    },
    {
      "epoch": 24.3,
      "eval_cer": 0.16042841517635192,
      "eval_loss": 0.4581703841686249,
      "eval_runtime": 117.0983,
      "eval_samples_per_second": 14.347,
      "step": 3500
    },
    {
      "epoch": 27.78,
      "learning_rate": 0.00015483870967741934,
      "loss": 0.1457,
      "step": 4000
    },
    {
      "epoch": 27.78,
      "eval_cer": 0.1622466228213469,
      "eval_loss": 0.5069878101348877,
      "eval_runtime": 117.1696,
      "eval_samples_per_second": 14.338,
      "step": 4000
    },
    {
      "epoch": 31.25,
      "learning_rate": 0.00013064516129032258,
      "loss": 0.1233,
      "step": 4500
    },
    {
      "epoch": 31.25,
      "eval_cer": 0.16633759002258555,
      "eval_loss": 0.5665308237075806,
      "eval_runtime": 117.8469,
      "eval_samples_per_second": 14.256,
      "step": 4500
    },
    {
      "epoch": 34.72,
      "learning_rate": 0.0001064516129032258,
      "loss": 0.1077,
      "step": 5000
    },
    {
      "epoch": 34.72,
      "eval_cer": 0.16429210642196623,
      "eval_loss": 0.5954398512840271,
      "eval_runtime": 117.3401,
      "eval_samples_per_second": 14.317,
      "step": 5000
    },
    {
      "epoch": 38.19,
      "learning_rate": 8.225806451612902e-05,
      "loss": 0.0944,
      "step": 5500
    },
    {
      "epoch": 38.19,
      "eval_cer": 0.16356766431341355,
      "eval_loss": 0.623413622379303,
      "eval_runtime": 116.7948,
      "eval_samples_per_second": 14.384,
      "step": 5500
    },
    {
      "epoch": 41.66,
      "learning_rate": 5.806451612903225e-05,
      "loss": 0.08,
      "step": 6000
    },
    {
      "epoch": 41.66,
      "eval_cer": 0.16270117473259563,
      "eval_loss": 0.6403368711471558,
      "eval_runtime": 116.8584,
      "eval_samples_per_second": 14.376,
      "step": 6000
    },
    {
      "epoch": 45.14,
      "learning_rate": 3.387096774193548e-05,
      "loss": 0.0746,
      "step": 6500
    },
    {
      "epoch": 45.14,
      "eval_cer": 0.16326936462165656,
      "eval_loss": 0.6701669096946716,
      "eval_runtime": 117.2831,
      "eval_samples_per_second": 14.324,
      "step": 6500
    },
    {
      "epoch": 48.61,
      "learning_rate": 9.677419354838709e-06,
      "loss": 0.0664,
      "step": 7000
    },
    {
      "epoch": 48.61,
      "eval_cer": 0.16285742695208738,
      "eval_loss": 0.684356153011322,
      "eval_runtime": 117.2078,
      "eval_samples_per_second": 14.334,
      "step": 7000
    }
  ],
  "max_steps": 7200,
  "num_train_epochs": 50,
  "total_flos": 2.1349156888097853e+19,
  "trial_name": null,
  "trial_params": null
}