{
  "best_metric": 0.3698837161064148,
  "best_model_checkpoint": "./mixstral/05-04-24-Weni-WeniGPT-Agents-Mixstral-Instruct-2.0.1-KTO_KTO with Agents 1.2.0 dataset and Mixstral model, with tokenization zephyr chat template-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": 4.052237033843994,
      "kl": 10.023146629333496,
      "learning_rate": 0.0001785714285714286,
      "logps/chosen": -232.72720336914062,
      "logps/rejected": -241.88575744628906,
      "loss": 0.4316,
      "rewards/chosen": 1.2224007844924927,
      "rewards/margins": 1.0828871726989746,
      "rewards/rejected": 0.279308944940567,
      "step": 20
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.734194278717041,
      "kl": 10.954643249511719,
      "learning_rate": 0.00015000000000000001,
      "logps/chosen": -228.13992309570312,
      "logps/rejected": -270.3298034667969,
      "loss": 0.4077,
      "rewards/chosen": 1.687286615371704,
      "rewards/margins": 1.9949144124984741,
      "rewards/rejected": -0.2780403196811676,
      "step": 40
    },
    {
      "epoch": 0.34,
      "eval_kl": 0.0,
      "eval_logps/chosen": -315.1575927734375,
      "eval_logps/rejected": -328.9866027832031,
      "eval_loss": 0.466037392616272,
      "eval_rewards/chosen": -7.166321754455566,
      "eval_rewards/margins": 2.8236684799194336,
      "eval_rewards/rejected": -10.133459091186523,
      "eval_runtime": 353.626,
      "eval_samples_per_second": 0.848,
      "eval_steps_per_second": 0.212,
      "step": 50
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.9602097272872925,
      "kl": 1.5916956663131714,
      "learning_rate": 0.00012142857142857143,
      "logps/chosen": -268.18023681640625,
      "logps/rejected": -304.6438903808594,
      "loss": 0.4028,
      "rewards/chosen": -2.1896169185638428,
      "rewards/margins": 4.930688858032227,
      "rewards/rejected": -6.980587482452393,
      "step": 60
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.9362258911132812,
      "kl": 3.6249618530273438,
      "learning_rate": 9.285714285714286e-05,
      "logps/chosen": -263.2818603515625,
      "logps/rejected": -308.0414733886719,
      "loss": 0.3873,
      "rewards/chosen": -0.1703629493713379,
      "rewards/margins": 5.487633228302002,
      "rewards/rejected": -5.359496116638184,
      "step": 80
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.322431802749634,
      "kl": 0.9517351388931274,
      "learning_rate": 6.428571428571429e-05,
      "logps/chosen": -240.21408081054688,
      "logps/rejected": -329.3759765625,
      "loss": 0.3351,
      "rewards/chosen": 0.027771174907684326,
      "rewards/margins": 7.602840423583984,
      "rewards/rejected": -7.7202911376953125,
      "step": 100
    },
    {
      "epoch": 0.68,
      "eval_kl": 0.48766180872917175,
      "eval_logps/chosen": -261.78826904296875,
      "eval_logps/rejected": -295.7093505859375,
      "eval_loss": 0.3698837161064148,
      "eval_rewards/chosen": -1.8293884992599487,
      "eval_rewards/margins": 4.808248043060303,
      "eval_rewards/rejected": -6.805734634399414,
      "eval_runtime": 352.0516,
      "eval_samples_per_second": 0.852,
      "eval_steps_per_second": 0.213,
      "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
}