{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9936305732484076,
  "eval_steps": 500,
  "global_step": 39,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.025477707006369428,
      "grad_norm": 9.697940219413395,
      "learning_rate": 1.25e-07,
      "logits/chosen": -2.9487781524658203,
      "logits/rejected": -2.858518123626709,
      "logps/chosen": -304.58709716796875,
      "logps/pi_response": -66.34120178222656,
      "logps/ref_response": -66.34120178222656,
      "logps/rejected": -133.11962890625,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.25477707006369427,
      "grad_norm": 7.564884688170802,
      "learning_rate": 4.646121984004665e-07,
      "logits/chosen": -2.7672486305236816,
      "logits/rejected": -2.7376208305358887,
      "logps/chosen": -257.03143310546875,
      "logps/pi_response": -73.40153503417969,
      "logps/ref_response": -72.56087493896484,
      "logps/rejected": -174.0777130126953,
      "loss": 0.6872,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": 0.01721888780593872,
      "rewards/margins": 0.016090132296085358,
      "rewards/rejected": 0.0011287559755146503,
      "step": 10
    },
    {
      "epoch": 0.5095541401273885,
      "grad_norm": 6.239131230843375,
      "learning_rate": 2.8355831645441387e-07,
      "logits/chosen": -2.693856716156006,
      "logits/rejected": -2.6693053245544434,
      "logps/chosen": -235.26699829101562,
      "logps/pi_response": -83.85284423828125,
      "logps/ref_response": -70.75907897949219,
      "logps/rejected": -182.71365356445312,
      "loss": 0.6539,
      "rewards/accuracies": 0.746874988079071,
      "rewards/chosen": 0.017035722732543945,
      "rewards/margins": 0.07849643379449844,
      "rewards/rejected": -0.0614607110619545,
      "step": 20
    },
    {
      "epoch": 0.7643312101910829,
      "grad_norm": 7.307349699720556,
      "learning_rate": 7.723433775328384e-08,
      "logits/chosen": -2.678931713104248,
      "logits/rejected": -2.6359119415283203,
      "logps/chosen": -260.82794189453125,
      "logps/pi_response": -106.23026275634766,
      "logps/ref_response": -70.76898193359375,
      "logps/rejected": -200.829345703125,
      "loss": 0.6205,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.08356819301843643,
      "rewards/margins": 0.1813538521528244,
      "rewards/rejected": -0.26492205262184143,
      "step": 30
    },
    {
      "epoch": 0.9936305732484076,
      "step": 39,
      "total_flos": 0.0,
      "train_loss": 0.6438931807493552,
      "train_runtime": 1776.8849,
      "train_samples_per_second": 5.628,
      "train_steps_per_second": 0.022
    }
  ],
  "logging_steps": 10,
  "max_steps": 39,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}