trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 8583,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.17,
      "learning_rate": 4.708726552487476e-05,
      "loss": 2.685,
      "step": 500
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.4174531049749504e-05,
      "loss": 2.1249,
      "step": 1000
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.126179657462426e-05,
      "loss": 1.9314,
      "step": 1500
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.834906209949901e-05,
      "loss": 1.8212,
      "step": 2000
    },
    {
      "epoch": 0.87,
      "learning_rate": 3.543632762437377e-05,
      "loss": 1.7248,
      "step": 2500
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.2523593149248515e-05,
      "loss": 1.6022,
      "step": 3000
    },
    {
      "epoch": 1.22,
      "learning_rate": 2.961085867412327e-05,
      "loss": 1.4302,
      "step": 3500
    },
    {
      "epoch": 1.4,
      "learning_rate": 2.669812419899802e-05,
      "loss": 1.4135,
      "step": 4000
    },
    {
      "epoch": 1.57,
      "learning_rate": 2.3785389723872772e-05,
      "loss": 1.3846,
      "step": 4500
    },
    {
      "epoch": 1.75,
      "learning_rate": 2.0872655248747523e-05,
      "loss": 1.3618,
      "step": 5000
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.7959920773622277e-05,
      "loss": 1.343,
      "step": 5500
    },
    {
      "epoch": 2.1,
      "learning_rate": 1.5047186298497029e-05,
      "loss": 1.21,
      "step": 6000
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.2134451823371781e-05,
      "loss": 1.1325,
      "step": 6500
    },
    {
      "epoch": 2.45,
      "learning_rate": 9.221717348246536e-06,
      "loss": 1.1156,
      "step": 7000
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.308982873121287e-06,
      "loss": 1.1092,
      "step": 7500
    },
    {
      "epoch": 2.8,
      "learning_rate": 3.3962483979960386e-06,
      "loss": 1.0911,
      "step": 8000
    },
    {
      "epoch": 2.97,
      "learning_rate": 4.835139228707911e-07,
      "loss": 1.0756,
      "step": 8500
    },
    {
      "epoch": 3.0,
      "step": 8583,
      "total_flos": 1.5341061280432128e+16,
      "train_loss": 1.4993489574444374,
      "train_runtime": 2601.28,
      "train_samples_per_second": 32.988,
      "train_steps_per_second": 3.3
    }
  ],
  "max_steps": 8583,
  "num_train_epochs": 3,
  "total_flos": 1.5341061280432128e+16,
  "trial_name": null,
  "trial_params": null
}