trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.739130434782608,
  "eval_steps": 500,
  "global_step": 66,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 3.6891603469848633,
      "learning_rate": 3.571428571428572e-05,
      "loss": 3.0307,
      "step": 10
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 1.072988748550415,
      "learning_rate": 4.837540606713538e-05,
      "loss": 0.7724,
      "step": 20
    },
    {
      "epoch": 2.608695652173913,
      "grad_norm": 0.9131895899772644,
      "learning_rate": 3.920161866827889e-05,
      "loss": 0.4177,
      "step": 30
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 0.7992828488349915,
      "learning_rate": 2.5e-05,
      "loss": 0.2625,
      "step": 40
    },
    {
      "epoch": 4.3478260869565215,
      "grad_norm": 0.6741700768470764,
      "learning_rate": 1.0798381331721109e-05,
      "loss": 0.1684,
      "step": 50
    },
    {
      "epoch": 5.217391304347826,
      "grad_norm": 0.4760473966598511,
      "learning_rate": 1.624593932864632e-06,
      "loss": 0.1069,
      "step": 60
    },
    {
      "epoch": 5.739130434782608,
      "step": 66,
      "total_flos": 1872683801247744.0,
      "train_loss": 0.7294434074199561,
      "train_runtime": 422.8198,
      "train_samples_per_second": 1.291,
      "train_steps_per_second": 0.156
    }
  ],
  "logging_steps": 10,
  "max_steps": 66,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1872683801247744.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}