{
  "best_metric": 0.47915583848953247,
  "best_model_checkpoint": "./zephyr/10-04-24-Weni-WeniGPT-Agents-Zephyr-1.0.27-KTO_Experiment with a new tokenizer configuration for chat template of zephyr-2_max_steps-147_batch_16_2024-04-10_ppid_9/checkpoint-100",
  "epoch": 0.6600660066006601,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.13,
      "grad_norm": 87.22119140625,
      "kl": 0.0514632947742939,
      "learning_rate": 1.8732394366197181e-06,
      "logps/chosen": NaN,
      "logps/rejected": NaN,
      "loss": 0.7523,
      "rewards/chosen": NaN,
      "rewards/margins": NaN,
      "rewards/rejected": NaN,
      "step": 20
    },
    {
      "epoch": 0.26,
      "grad_norm": 75.8567123413086,
      "kl": 0.3588915467262268,
      "learning_rate": 1.591549295774648e-06,
      "logps/chosen": NaN,
      "logps/rejected": NaN,
      "loss": 0.5807,
      "rewards/chosen": NaN,
      "rewards/margins": NaN,
      "rewards/rejected": NaN,
      "step": 40
    },
    {
      "epoch": 0.33,
      "eval_kl": 0.4327344000339508,
      "eval_logps/chosen": -279.24505615234375,
      "eval_logps/rejected": -238.83538818359375,
      "eval_loss": 0.48886266350746155,
      "eval_rewards/chosen": 0.024170810356736183,
      "eval_rewards/margins": 0.09017277508974075,
      "eval_rewards/rejected": -0.06600195914506912,
      "eval_runtime": 169.6418,
      "eval_samples_per_second": 2.063,
      "eval_steps_per_second": 0.519,
      "step": 50
    },
    {
      "epoch": 0.4,
      "grad_norm": 86.48399353027344,
      "kl": 0.4763767719268799,
      "learning_rate": 1.3098591549295774e-06,
      "logps/chosen": NaN,
      "logps/rejected": NaN,
      "loss": 0.6521,
      "rewards/chosen": NaN,
      "rewards/margins": NaN,
      "rewards/rejected": NaN,
      "step": 60
    },
    {
      "epoch": 0.53,
      "grad_norm": 89.11936950683594,
      "kl": 0.35531342029571533,
      "learning_rate": 1.028169014084507e-06,
      "logps/chosen": NaN,
      "logps/rejected": NaN,
      "loss": 0.7257,
      "rewards/chosen": NaN,
      "rewards/margins": NaN,
      "rewards/rejected": NaN,
      "step": 80
    },
    {
      "epoch": 0.66,
      "grad_norm": 83.82781219482422,
      "kl": 0.20901036262512207,
      "learning_rate": 7.464788732394366e-07,
      "logps/chosen": NaN,
      "logps/rejected": NaN,
      "loss": 0.6536,
      "rewards/chosen": NaN,
      "rewards/margins": NaN,
      "rewards/rejected": NaN,
      "step": 100
    },
    {
      "epoch": 0.66,
      "eval_kl": 0.20153704285621643,
      "eval_logps/chosen": -280.2303771972656,
      "eval_logps/rejected": -240.73260498046875,
      "eval_loss": 0.47915583848953247,
      "eval_rewards/chosen": -0.07436122000217438,
      "eval_rewards/margins": 0.18136295676231384,
      "eval_rewards/rejected": -0.255724161863327,
      "eval_runtime": 169.8539,
      "eval_samples_per_second": 2.061,
      "eval_steps_per_second": 0.518,
      "step": 100
    }
  ],
  "logging_steps": 20,
  "max_steps": 147,
  "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
}