{
  "best_metric": 0.544333279132843,
  "best_model_checkpoint": "./zephyr/09-04-24-Weni-WeniGPT-Agents-Zephyr-1.0.9-KTO_Hyperparameter search, altering desired and undesired weights for KTO task.-2_max_steps-145_batch_16_2024-04-09_ppid_9/checkpoint-100",
  "epoch": 0.684931506849315,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.14,
      "grad_norm": 0.0,
      "kl": 0.9986292123794556,
      "learning_rate": 0.00018,
      "logps/chosen": -876.9887084960938,
      "logps/rejected": -891.7728271484375,
      "loss": 0.5066,
      "rewards/chosen": -59.218875885009766,
      "rewards/margins": 1.2524851560592651,
      "rewards/rejected": -60.5962028503418,
      "step": 20
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.0,
      "kl": 0.0,
      "learning_rate": 0.00015142857142857143,
      "logps/chosen": -1901.8116455078125,
      "logps/rejected": -1940.32470703125,
      "loss": 0.5606,
      "rewards/chosen": -161.08740234375,
      "rewards/margins": 2.580688953399658,
      "rewards/rejected": -164.8165283203125,
      "step": 40
    },
    {
      "epoch": 0.34,
      "eval_kl": 0.0,
      "eval_logps/chosen": -1797.90625,
      "eval_logps/rejected": -1778.623291015625,
      "eval_loss": 0.5443333983421326,
      "eval_rewards/chosen": -151.05047607421875,
      "eval_rewards/margins": 2.7896132469177246,
      "eval_rewards/rejected": -151.03567504882812,
      "eval_runtime": 138.2181,
      "eval_samples_per_second": 2.17,
      "eval_steps_per_second": 0.543,
      "step": 50
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.0,
      "kl": 0.0,
      "learning_rate": 0.00012285714285714287,
      "logps/chosen": -1852.9610595703125,
      "logps/rejected": -1893.4521484375,
      "loss": 0.5211,
      "rewards/chosen": -157.99769592285156,
      "rewards/margins": 3.8418266773223877,
      "rewards/rejected": -160.36117553710938,
      "step": 60
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.0,
      "kl": 0.0,
      "learning_rate": 9.428571428571429e-05,
      "logps/chosen": -2051.344482421875,
      "logps/rejected": -1892.0853271484375,
      "loss": 0.5714,
      "rewards/chosen": -176.026611328125,
      "rewards/margins": -20.468326568603516,
      "rewards/rejected": -161.7165069580078,
      "step": 80
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.0,
      "kl": 0.0,
      "learning_rate": 6.571428571428571e-05,
      "logps/chosen": -2035.6630859375,
      "logps/rejected": -1939.001220703125,
      "loss": 0.4995,
      "rewards/chosen": -175.16189575195312,
      "rewards/margins": -12.196085929870605,
      "rewards/rejected": -165.23004150390625,
      "step": 100
    },
    {
      "epoch": 0.68,
      "eval_kl": 0.0,
      "eval_logps/chosen": -1859.810791015625,
      "eval_logps/rejected": -1831.84814453125,
      "eval_loss": 0.544333279132843,
      "eval_rewards/chosen": -157.24095153808594,
      "eval_rewards/margins": 2.0474841594696045,
      "eval_rewards/rejected": -156.35816955566406,
      "eval_runtime": 138.0332,
      "eval_samples_per_second": 2.173,
      "eval_steps_per_second": 0.543,
      "step": 100
    }
  ],
  "logging_steps": 20,
  "max_steps": 145,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}