trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 22860,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 500
    },
    {
      "epoch": 0.17,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 1000
    },
    {
      "epoch": 0.26,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 1500
    },
    {
      "epoch": 0.35,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 2000
    },
    {
      "epoch": 0.44,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 2500
    },
    {
      "epoch": 0.52,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 3000
    },
    {
      "epoch": 0.61,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 3500
    },
    {
      "epoch": 0.7,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 4000
    },
    {
      "epoch": 0.79,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 4500
    },
    {
      "epoch": 0.87,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 5000
    },
    {
      "epoch": 0.96,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 5500
    },
    {
      "epoch": 1.05,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 6000
    },
    {
      "epoch": 1.14,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 6500
    },
    {
      "epoch": 1.22,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 7000
    },
    {
      "epoch": 1.31,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 7500
    },
    {
      "epoch": 1.4,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 8000
    },
    {
      "epoch": 1.49,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 8500
    },
    {
      "epoch": 1.57,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 9000
    },
    {
      "epoch": 1.66,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 9500
    },
    {
      "epoch": 1.75,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 10000
    },
    {
      "epoch": 1.84,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 10500
    },
    {
      "epoch": 1.92,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 11000
    },
    {
      "epoch": 2.01,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 11500
    },
    {
      "epoch": 2.1,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 12000
    },
    {
      "epoch": 2.19,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 12500
    },
    {
      "epoch": 2.27,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 13000
    },
    {
      "epoch": 2.36,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 13500
    },
    {
      "epoch": 2.45,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 14000
    },
    {
      "epoch": 2.54,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 14500
    },
    {
      "epoch": 2.62,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 15000
    },
    {
      "epoch": 2.71,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 15500
    },
    {
      "epoch": 2.8,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 16000
    },
    {
      "epoch": 2.89,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 16500
    },
    {
      "epoch": 2.97,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 17000
    },
    {
      "epoch": 3.06,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 17500
    },
    {
      "epoch": 3.15,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 18000
    },
    {
      "epoch": 3.24,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 18500
    },
    {
      "epoch": 3.32,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 19000
    },
    {
      "epoch": 3.41,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 19500
    },
    {
      "epoch": 3.5,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 20000
    },
    {
      "epoch": 3.59,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 20500
    },
    {
      "epoch": 3.67,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 21000
    },
    {
      "epoch": 3.76,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 21500
    },
    {
      "epoch": 3.85,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 22000
    },
    {
      "epoch": 3.94,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 22500
    },
    {
      "epoch": 4.0,
      "step": 22860,
      "total_flos": 2.1912235278336e+17,
      "train_loss": 0.0,
      "train_runtime": 7438.5423,
      "train_samples_per_second": 21.51,
      "train_steps_per_second": 3.073
    }
  ],
  "logging_steps": 500,
  "max_steps": 22860,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 4000,
  "total_flos": 2.1912235278336e+17,
  "train_batch_size": 7,
  "trial_name": null,
  "trial_params": null
}