trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 9375,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.16,
      "grad_norm": 287471.90625,
      "learning_rate": 4.7843665768194077e-05,
      "loss": 3.6892,
      "step": 500
    },
    {
      "epoch": 0.32,
      "grad_norm": 146946.265625,
      "learning_rate": 4.5148247978436656e-05,
      "loss": 1.3412,
      "step": 1000
    },
    {
      "epoch": 0.48,
      "grad_norm": 62433.12109375,
      "learning_rate": 4.245283018867925e-05,
      "loss": 1.2162,
      "step": 1500
    },
    {
      "epoch": 0.64,
      "grad_norm": 62861.984375,
      "learning_rate": 3.9757412398921835e-05,
      "loss": 1.1654,
      "step": 2000
    },
    {
      "epoch": 0.8,
      "grad_norm": 50599.45703125,
      "learning_rate": 3.706199460916443e-05,
      "loss": 1.1287,
      "step": 2500
    },
    {
      "epoch": 0.96,
      "grad_norm": 54249.3203125,
      "learning_rate": 3.4366576819407007e-05,
      "loss": 1.1063,
      "step": 3000
    },
    {
      "epoch": 1.12,
      "grad_norm": 55991.08984375,
      "learning_rate": 3.16711590296496e-05,
      "loss": 1.0831,
      "step": 3500
    },
    {
      "epoch": 1.28,
      "grad_norm": 96158.0625,
      "learning_rate": 2.8975741239892185e-05,
      "loss": 1.0572,
      "step": 4000
    },
    {
      "epoch": 1.44,
      "grad_norm": 48621.35546875,
      "learning_rate": 2.6280323450134768e-05,
      "loss": 1.0496,
      "step": 4500
    },
    {
      "epoch": 1.6,
      "grad_norm": 47878.76953125,
      "learning_rate": 2.358490566037736e-05,
      "loss": 1.0389,
      "step": 5000
    },
    {
      "epoch": 1.76,
      "grad_norm": 62109.3984375,
      "learning_rate": 2.0889487870619947e-05,
      "loss": 1.0359,
      "step": 5500
    },
    {
      "epoch": 1.92,
      "grad_norm": 72237.4140625,
      "learning_rate": 1.8194070080862536e-05,
      "loss": 1.0198,
      "step": 6000
    },
    {
      "epoch": 2.08,
      "grad_norm": 116479.84375,
      "learning_rate": 1.5498652291105122e-05,
      "loss": 1.0144,
      "step": 6500
    },
    {
      "epoch": 2.24,
      "grad_norm": 53538.13671875,
      "learning_rate": 1.2803234501347711e-05,
      "loss": 1.009,
      "step": 7000
    },
    {
      "epoch": 2.4,
      "grad_norm": 42354.23828125,
      "learning_rate": 1.0107816711590297e-05,
      "loss": 1.0089,
      "step": 7500
    },
    {
      "epoch": 2.56,
      "grad_norm": 50047.01953125,
      "learning_rate": 7.412398921832884e-06,
      "loss": 0.9998,
      "step": 8000
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 40101.76171875,
      "learning_rate": 4.716981132075472e-06,
      "loss": 1.0038,
      "step": 8500
    },
    {
      "epoch": 2.88,
      "grad_norm": 59145.29296875,
      "learning_rate": 2.0215633423180592e-06,
      "loss": 0.9942,
      "step": 9000
    }
  ],
  "logging_steps": 500,
  "max_steps": 9375,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 30000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.083549155328e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}