checkpoint-10/trainer_state.json

{
  "best_metric": 0.31577983498573303,
  "best_model_checkpoint": "./Zephyr/28-03-24-Weni-WeniGPT-QA-Zephyr-7B-5.0.0-KTO_WeniGPT Experiment using KTO trainer with no collator-2_max_steps-522_batch_48_2024-03-28_ppid_3151/checkpoint-10",
  "epoch": 0.056457304163726185,
  "eval_steps": 5,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "eval_kl": 0.5473354458808899,
      "eval_logps/chosen": -187.94171142578125,
      "eval_logps/rejected": -207.13204956054688,
      "eval_loss": 0.468790739774704,
      "eval_rewards/chosen": 0.1763239949941635,
      "eval_rewards/margins": 0.2541382610797882,
      "eval_rewards/rejected": -0.07781429588794708,
      "eval_runtime": 485.556,
      "eval_samples_per_second": 1.071,
      "eval_steps_per_second": 0.535,
      "step": 5
    },
    {
      "epoch": 0.06,
      "eval_kl": 0.13708601891994476,
      "eval_logps/chosen": -181.25985717773438,
      "eval_logps/rejected": -218.45681762695312,
      "eval_loss": 0.31577983498573303,
      "eval_rewards/chosen": 0.8445124626159668,
      "eval_rewards/margins": 2.0548036098480225,
      "eval_rewards/rejected": -1.2102913856506348,
      "eval_runtime": 485.4275,
      "eval_samples_per_second": 1.071,
      "eval_steps_per_second": 0.536,
      "step": 10
    }
  ],
  "logging_steps": 20,
  "max_steps": 522,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 5,
  "total_flos": 0.0,
  "train_batch_size": 3,
  "trial_name": null,
  "trial_params": null
}