checkpoint-100/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.005243151133831433,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 2e-05,
      "loss": 1.5236,
      "step": 10
    },
    {
      "epoch": 0.0,
      "learning_rate": 4e-05,
      "loss": 1.3172,
      "step": 20
    },
    {
      "epoch": 0.0,
      "learning_rate": 6e-05,
      "loss": 1.2325,
      "step": 30
    },
    {
      "epoch": 0.0,
      "learning_rate": 8e-05,
      "loss": 1.0105,
      "step": 40
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0001,
      "loss": 0.9889,
      "step": 50
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.00012,
      "loss": 0.9394,
      "step": 60
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.00014,
      "loss": 0.9342,
      "step": 70
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.00016,
      "loss": 0.9496,
      "step": 80
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.00018,
      "loss": 1.0416,
      "step": 90
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9842,
      "step": 100
    }
  ],
  "logging_steps": 10,
  "max_steps": 3000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 7173598526595072.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}