trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.6923076923076925,
  "eval_steps": 500,
  "global_step": 36,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.20512820512820512,
      "grad_norm": 3.4090847969055176,
      "learning_rate": 0.00018888888888888888,
      "loss": 8.6815,
      "step": 2
    },
    {
      "epoch": 0.41025641025641024,
      "grad_norm": 3.7809109687805176,
      "learning_rate": 0.00017777777777777779,
      "loss": 8.2204,
      "step": 4
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 5.441247463226318,
      "learning_rate": 0.0001666666666666667,
      "loss": 7.2478,
      "step": 6
    },
    {
      "epoch": 0.8205128205128205,
      "grad_norm": 3.1511142253875732,
      "learning_rate": 0.00015555555555555556,
      "loss": 6.0805,
      "step": 8
    },
    {
      "epoch": 1.0256410256410255,
      "grad_norm": 3.6586670875549316,
      "learning_rate": 0.00014444444444444444,
      "loss": 5.8499,
      "step": 10
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 3.667562961578369,
      "learning_rate": 0.00013333333333333334,
      "loss": 5.6406,
      "step": 12
    },
    {
      "epoch": 1.435897435897436,
      "grad_norm": 4.54071044921875,
      "learning_rate": 0.00012222222222222224,
      "loss": 4.7872,
      "step": 14
    },
    {
      "epoch": 1.641025641025641,
      "grad_norm": 4.788601875305176,
      "learning_rate": 0.00011111111111111112,
      "loss": 6.0102,
      "step": 16
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 2.731595516204834,
      "learning_rate": 0.0001,
      "loss": 5.3443,
      "step": 18
    },
    {
      "epoch": 2.051282051282051,
      "grad_norm": 2.9969563484191895,
      "learning_rate": 8.888888888888889e-05,
      "loss": 5.0464,
      "step": 20
    },
    {
      "epoch": 2.2564102564102564,
      "grad_norm": 2.7965035438537598,
      "learning_rate": 7.777777777777778e-05,
      "loss": 5.0523,
      "step": 22
    },
    {
      "epoch": 2.4615384615384617,
      "grad_norm": 2.5996360778808594,
      "learning_rate": 6.666666666666667e-05,
      "loss": 5.1417,
      "step": 24
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 2.2516720294952393,
      "learning_rate": 5.555555555555556e-05,
      "loss": 4.571,
      "step": 26
    },
    {
      "epoch": 2.871794871794872,
      "grad_norm": 3.5645246505737305,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 4.9004,
      "step": 28
    },
    {
      "epoch": 3.076923076923077,
      "grad_norm": 1.808791995048523,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 4.6747,
      "step": 30
    },
    {
      "epoch": 3.282051282051282,
      "grad_norm": 2.8537707328796387,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 3.9459,
      "step": 32
    },
    {
      "epoch": 3.4871794871794872,
      "grad_norm": 3.786196708679199,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 5.4513,
      "step": 34
    },
    {
      "epoch": 3.6923076923076925,
      "grad_norm": 0.8137506246566772,
      "learning_rate": 0.0,
      "loss": 4.9191,
      "step": 36
    },
    {
      "epoch": 3.6923076923076925,
      "step": 36,
      "total_flos": 73188581212248.0,
      "train_loss": 5.642510652542114,
      "train_runtime": 91.2089,
      "train_samples_per_second": 3.421,
      "train_steps_per_second": 0.395
    }
  ],
  "logging_steps": 2,
  "max_steps": 36,
  "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": 73188581212248.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}