checkpoint-100/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.03980099502487562,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03980099502487562,
      "grad_norm": 1.5459058284759521,
      "learning_rate": 4.980474401576887e-07,
      "logits/chosen": -1.4838274717330933,
      "logits/rejected": -1.5693057775497437,
      "logps/chosen": -190.30685424804688,
      "logps/rejected": -190.12179565429688,
      "loss": 0.7107,
      "rewards/accuracies": 0.49312499165534973,
      "rewards/chosen": -0.8337415456771851,
      "rewards/margins": 0.0017802910879254341,
      "rewards/rejected": -0.8355217576026917,
      "step": 100
    }
  ],
  "logging_steps": 100,
  "max_steps": 2512,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}