trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.84,
  "eval_steps": 500,
  "global_step": 24,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.32,
      "grad_norm": 1.0148409605026245,
      "learning_rate": 0.00018333333333333334,
      "loss": 9.5085,
      "step": 2
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6703659296035767,
      "learning_rate": 0.0001666666666666667,
      "loss": 9.1769,
      "step": 4
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.8155523538589478,
      "learning_rate": 0.00015000000000000001,
      "loss": 8.7379,
      "step": 6
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.9680312871932983,
      "learning_rate": 0.00013333333333333334,
      "loss": 8.2451,
      "step": 8
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.053583860397339,
      "learning_rate": 0.00011666666666666668,
      "loss": 7.9159,
      "step": 10
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.528686285018921,
      "learning_rate": 0.0001,
      "loss": 7.6275,
      "step": 12
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.1224510669708252,
      "learning_rate": 8.333333333333334e-05,
      "loss": 7.4755,
      "step": 14
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.065908670425415,
      "learning_rate": 6.666666666666667e-05,
      "loss": 7.3267,
      "step": 16
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.025452971458435,
      "learning_rate": 5e-05,
      "loss": 7.2462,
      "step": 18
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.8463966846466064,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 7.2758,
      "step": 20
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.6603997945785522,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 7.3307,
      "step": 22
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.8327799439430237,
      "learning_rate": 0.0,
      "loss": 7.1545,
      "step": 24
    },
    {
      "epoch": 3.84,
      "step": 24,
      "total_flos": 108923677416360.0,
      "train_loss": 7.9184350570042925,
      "train_runtime": 637.0525,
      "train_samples_per_second": 0.628,
      "train_steps_per_second": 0.038
    }
  ],
  "logging_steps": 2,
  "max_steps": 24,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 108923677416360.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}