trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 9012,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.11,
      "learning_rate": 4.722592099422992e-05,
      "loss": 1.4475,
      "step": 500
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.4451841988459835e-05,
      "loss": 1.176,
      "step": 1000
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.167776298268975e-05,
      "loss": 1.1202,
      "step": 1500
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.890368397691966e-05,
      "loss": 1.0756,
      "step": 2000
    },
    {
      "epoch": 0.55,
      "learning_rate": 3.612960497114958e-05,
      "loss": 1.0282,
      "step": 2500
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.3355525965379494e-05,
      "loss": 1.0141,
      "step": 3000
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.058144695960941e-05,
      "loss": 0.9844,
      "step": 3500
    },
    {
      "epoch": 0.89,
      "learning_rate": 2.7807367953839327e-05,
      "loss": 0.9714,
      "step": 4000
    },
    {
      "epoch": 1.0,
      "learning_rate": 2.5033288948069243e-05,
      "loss": 0.9554,
      "step": 4500
    },
    {
      "epoch": 1.11,
      "learning_rate": 2.225920994229916e-05,
      "loss": 0.864,
      "step": 5000
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.9485130936529072e-05,
      "loss": 0.8439,
      "step": 5500
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.671105193075899e-05,
      "loss": 0.84,
      "step": 6000
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.3936972924988904e-05,
      "loss": 0.8381,
      "step": 6500
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.116289391921882e-05,
      "loss": 0.83,
      "step": 7000
    },
    {
      "epoch": 1.66,
      "learning_rate": 8.388814913448735e-06,
      "loss": 0.8192,
      "step": 7500
    },
    {
      "epoch": 1.78,
      "learning_rate": 5.614735907678651e-06,
      "loss": 0.8102,
      "step": 8000
    },
    {
      "epoch": 1.89,
      "learning_rate": 2.8406569019085665e-06,
      "loss": 0.805,
      "step": 8500
    },
    {
      "epoch": 2.0,
      "learning_rate": 6.657789613848203e-08,
      "loss": 0.8009,
      "step": 9000
    },
    {
      "epoch": 2.0,
      "step": 9012,
      "total_flos": 2.0306395703938253e+17,
      "train_loss": 0.9567329946010842,
      "train_runtime": 10006.773,
      "train_samples_per_second": 14.409,
      "train_steps_per_second": 0.901
    }
  ],
  "logging_steps": 500,
  "max_steps": 9012,
  "num_train_epochs": 2,
  "save_steps": 4500,
  "total_flos": 2.0306395703938253e+17,
  "trial_name": null,
  "trial_params": null
}