checkpoint-90/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.931506849315069,
  "eval_steps": 500,
  "global_step": 90,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.55,
      "grad_norm": 0.4512141942977905,
      "learning_rate": 0.00019753086419753085,
      "loss": 1.2665,
      "step": 10
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.32605287432670593,
      "learning_rate": 0.0001728395061728395,
      "loss": 0.739,
      "step": 20
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.3033522069454193,
      "learning_rate": 0.00014814814814814815,
      "loss": 0.6175,
      "step": 30
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.2529822885990143,
      "learning_rate": 0.0001234567901234568,
      "loss": 0.4745,
      "step": 40
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.3419560492038727,
      "learning_rate": 9.876543209876543e-05,
      "loss": 0.366,
      "step": 50
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.24991345405578613,
      "learning_rate": 7.407407407407407e-05,
      "loss": 0.3953,
      "step": 60
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.19746990501880646,
      "learning_rate": 4.938271604938271e-05,
      "loss": 0.3283,
      "step": 70
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.23590216040611267,
      "learning_rate": 2.4691358024691357e-05,
      "loss": 0.353,
      "step": 80
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.17705382406711578,
      "learning_rate": 0.0,
      "loss": 0.3079,
      "step": 90
    }
  ],
  "logging_steps": 10,
  "max_steps": 90,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 1.463294116233216e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}