{
  "best_metric": 0.38556817173957825,
  "best_model_checkpoint": "./mixstral/05-04-24-Weni-WeniGPT-Agents-Mixstral-Instruct-2.0.0-KTO_KTO with Agents 1.2.0 dataset and Mixstral model-3_max_steps-145_batch_16_2024-04-05_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": 2.368553638458252,
      "kl": 0.5817955732345581,
      "learning_rate": 0.00018,
      "logps/chosen": -227.94122314453125,
      "logps/rejected": -240.64300537109375,
      "loss": 0.475,
      "rewards/chosen": -0.784079372882843,
      "rewards/margins": 0.5924594402313232,
      "rewards/rejected": -1.3401262760162354,
      "step": 20
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.441540241241455,
      "kl": 10.41219425201416,
      "learning_rate": 0.00015142857142857143,
      "logps/chosen": -233.67567443847656,
      "logps/rejected": -243.45187377929688,
      "loss": 0.4403,
      "rewards/chosen": 0.15312156081199646,
      "rewards/margins": 0.6628362536430359,
      "rewards/rejected": -0.5719818472862244,
      "step": 40
    },
    {
      "epoch": 0.34,
      "eval_kl": 0.5120495557785034,
      "eval_logps/chosen": -248.310302734375,
      "eval_logps/rejected": -253.4099578857422,
      "eval_loss": 0.4195210039615631,
      "eval_rewards/chosen": -1.1683257818222046,
      "eval_rewards/margins": 2.6137375831604004,
      "eval_rewards/rejected": -3.615262269973755,
      "eval_runtime": 357.6046,
      "eval_samples_per_second": 0.839,
      "eval_steps_per_second": 0.21,
      "step": 50
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.9933063983917236,
      "kl": 1.537040114402771,
      "learning_rate": 0.00012285714285714287,
      "logps/chosen": -244.6122283935547,
      "logps/rejected": -288.9703674316406,
      "loss": 0.3877,
      "rewards/chosen": -0.880365252494812,
      "rewards/margins": 3.219341516494751,
      "rewards/rejected": -4.040163516998291,
      "step": 60
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.7866162061691284,
      "kl": 8.153600692749023,
      "learning_rate": 9.428571428571429e-05,
      "logps/chosen": -193.49400329589844,
      "logps/rejected": -239.42953491210938,
      "loss": 0.3612,
      "rewards/chosen": 1.8910856246948242,
      "rewards/margins": 2.9180142879486084,
      "rewards/rejected": -1.0155872106552124,
      "step": 80
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.8967958688735962,
      "kl": 5.099704742431641,
      "learning_rate": 6.714285714285714e-05,
      "logps/chosen": -210.8452911376953,
      "logps/rejected": -259.0009460449219,
      "loss": 0.3518,
      "rewards/chosen": 1.323478102684021,
      "rewards/margins": 3.5613787174224854,
      "rewards/rejected": -2.316351890563965,
      "step": 100
    },
    {
      "epoch": 0.68,
      "eval_kl": 0.0,
      "eval_logps/chosen": -263.7559509277344,
      "eval_logps/rejected": -282.9854431152344,
      "eval_loss": 0.38556817173957825,
      "eval_rewards/chosen": -2.7128894329071045,
      "eval_rewards/margins": 3.9057483673095703,
      "eval_rewards/rejected": -6.572808265686035,
      "eval_runtime": 356.8194,
      "eval_samples_per_second": 0.841,
      "eval_steps_per_second": 0.21,
      "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
}