{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 116.27799227799228,
  "global_step": 15000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 3.87,
      "learning_rate": 1e-05,
      "learning_rate_embeddings": 1e-05,
      "loss": 8.5314,
      "step": 500
    },
    {
      "epoch": 7.75,
      "learning_rate": 2e-05,
      "learning_rate_embeddings": 2e-05,
      "loss": 5.6865,
      "step": 1000
    },
    {
      "epoch": 11.63,
      "learning_rate": 3e-05,
      "learning_rate_embeddings": 3e-05,
      "loss": 4.756,
      "step": 1500
    },
    {
      "epoch": 15.5,
      "learning_rate": 4e-05,
      "learning_rate_embeddings": 4e-05,
      "loss": 4.3111,
      "step": 2000
    },
    {
      "epoch": 19.38,
      "learning_rate": 5e-05,
      "learning_rate_embeddings": 5e-05,
      "loss": 4.021,
      "step": 2500
    },
    {
      "epoch": 19.38,
      "eval_loss": 3.8212602138519287,
      "eval_runtime": 23.2233,
      "eval_samples_per_second": 315.33,
      "eval_steps_per_second": 2.497,
      "step": 2500
    },
    {
      "epoch": 23.25,
      "learning_rate": 6e-05,
      "learning_rate_embeddings": 6e-05,
      "loss": 3.829,
      "step": 3000
    },
    {
      "epoch": 27.13,
      "learning_rate": 7.000000000000001e-05,
      "learning_rate_embeddings": 7.000000000000001e-05,
      "loss": 3.6784,
      "step": 3500
    },
    {
      "epoch": 31.01,
      "learning_rate": 8e-05,
      "learning_rate_embeddings": 8e-05,
      "loss": 3.5455,
      "step": 4000
    },
    {
      "epoch": 34.88,
      "learning_rate": 8.999999999999999e-05,
      "learning_rate_embeddings": 8.999999999999999e-05,
      "loss": 3.4304,
      "step": 4500
    },
    {
      "epoch": 38.76,
      "learning_rate": 0.0001,
      "learning_rate_embeddings": 0.0001,
      "loss": 3.3348,
      "step": 5000
    },
    {
      "epoch": 38.76,
      "eval_loss": 3.238213300704956,
      "eval_runtime": 9.9786,
      "eval_samples_per_second": 733.872,
      "eval_steps_per_second": 5.812,
      "step": 5000
    },
    {
      "epoch": 42.63,
      "learning_rate": 0.00011,
      "learning_rate_embeddings": 0.00011,
      "loss": 3.2411,
      "step": 5500
    },
    {
      "epoch": 46.51,
      "learning_rate": 0.00012,
      "learning_rate_embeddings": 0.00012,
      "loss": 3.1542,
      "step": 6000
    },
    {
      "epoch": 50.39,
      "learning_rate": 0.00013000000000000002,
      "learning_rate_embeddings": 0.00013000000000000002,
      "loss": 3.0781,
      "step": 6500
    },
    {
      "epoch": 54.26,
      "learning_rate": 0.00014000000000000001,
      "learning_rate_embeddings": 0.00014000000000000001,
      "loss": 3.0012,
      "step": 7000
    },
    {
      "epoch": 58.14,
      "learning_rate": 0.00015,
      "learning_rate_embeddings": 0.00015,
      "loss": 2.9349,
      "step": 7500
    },
    {
      "epoch": 58.14,
      "eval_loss": 2.92873477935791,
      "eval_runtime": 9.9086,
      "eval_samples_per_second": 739.053,
      "eval_steps_per_second": 5.853,
      "step": 7500
    },
    {
      "epoch": 62.02,
      "learning_rate": 0.00016,
      "learning_rate_embeddings": 0.00016,
      "loss": 2.8707,
      "step": 8000
    },
    {
      "epoch": 65.89,
      "learning_rate": 0.00017,
      "learning_rate_embeddings": 0.00017,
      "loss": 2.8075,
      "step": 8500
    },
    {
      "epoch": 69.76,
      "learning_rate": 0.00017999999999999998,
      "learning_rate_embeddings": 0.00017999999999999998,
      "loss": 2.7504,
      "step": 9000
    },
    {
      "epoch": 73.64,
      "learning_rate": 0.00019,
      "learning_rate_embeddings": 0.00019,
      "loss": 2.695,
      "step": 9500
    },
    {
      "epoch": 77.52,
      "learning_rate": 0.0002,
      "learning_rate_embeddings": 0.0002,
      "loss": 2.6372,
      "step": 10000
    },
    {
      "epoch": 77.52,
      "eval_loss": 2.7580177783966064,
      "eval_runtime": 9.9239,
      "eval_samples_per_second": 737.915,
      "eval_steps_per_second": 5.844,
      "step": 10000
    },
    {
      "epoch": 81.39,
      "learning_rate": 0.00021,
      "learning_rate_embeddings": 0.00021,
      "loss": 2.5911,
      "step": 10500
    },
    {
      "epoch": 85.27,
      "learning_rate": 0.00022,
      "learning_rate_embeddings": 0.00022,
      "loss": 2.5392,
      "step": 11000
    },
    {
      "epoch": 89.15,
      "learning_rate": 0.00023,
      "learning_rate_embeddings": 0.00023,
      "loss": 2.4921,
      "step": 11500
    },
    {
      "epoch": 93.02,
      "learning_rate": 0.00024,
      "learning_rate_embeddings": 0.00024,
      "loss": 2.4477,
      "step": 12000
    },
    {
      "epoch": 96.9,
      "learning_rate": 0.00025,
      "learning_rate_embeddings": 0.00025,
      "loss": 2.3984,
      "step": 12500
    },
    {
      "epoch": 96.9,
      "eval_loss": 2.669841766357422,
      "eval_runtime": 9.9424,
      "eval_samples_per_second": 736.544,
      "eval_steps_per_second": 5.834,
      "step": 12500
    },
    {
      "epoch": 100.77,
      "learning_rate": 0.00026000000000000003,
      "learning_rate_embeddings": 0.00026000000000000003,
      "loss": 2.3548,
      "step": 13000
    },
    {
      "epoch": 104.65,
      "learning_rate": 0.00027,
      "learning_rate_embeddings": 0.00027,
      "loss": 2.3141,
      "step": 13500
    },
    {
      "epoch": 108.53,
      "learning_rate": 0.00028000000000000003,
      "learning_rate_embeddings": 0.00028000000000000003,
      "loss": 2.2728,
      "step": 14000
    },
    {
      "epoch": 112.4,
      "learning_rate": 0.00029,
      "learning_rate_embeddings": 0.00029,
      "loss": 2.233,
      "step": 14500
    },
    {
      "epoch": 116.28,
      "learning_rate": 0.0003,
      "learning_rate_embeddings": 0.0003,
      "loss": 2.1912,
      "step": 15000
    },
    {
      "epoch": 116.28,
      "eval_loss": 2.6404244899749756,
      "eval_runtime": 9.8742,
      "eval_samples_per_second": 741.631,
      "eval_steps_per_second": 5.874,
      "step": 15000
    }
  ],
  "max_steps": 250000,
  "num_train_epochs": 1938,
  "total_flos": 2.51810266742784e+17,
  "trial_name": null,
  "trial_params": null
}