{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9874476987447699,
  "eval_steps": 500,
  "global_step": 59,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.016736401673640166,
      "grad_norm": 126.2238527844094,
      "learning_rate": 1.6666666666666664e-08,
      "logits/chosen": -0.8470035195350647,
      "logits/rejected": -0.5631465315818787,
      "logps/chosen": -390.660400390625,
      "logps/pi_response": -197.14308166503906,
      "logps/ref_response": -197.14308166503906,
      "logps/rejected": -774.7786865234375,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.16736401673640167,
      "grad_norm": 136.14999220764295,
      "learning_rate": 9.860114570402053e-08,
      "logits/chosen": -0.74688321352005,
      "logits/rejected": -0.4899700880050659,
      "logps/chosen": -478.9698486328125,
      "logps/pi_response": -176.40945434570312,
      "logps/ref_response": -174.20945739746094,
      "logps/rejected": -738.7280883789062,
      "loss": 0.6789,
      "rewards/accuracies": 0.5520833134651184,
      "rewards/chosen": -0.008676827885210514,
      "rewards/margins": 0.021465951576828957,
      "rewards/rejected": -0.030142780393362045,
      "step": 10
    },
    {
      "epoch": 0.33472803347280333,
      "grad_norm": 112.25884202942129,
      "learning_rate": 8.374915007591053e-08,
      "logits/chosen": -0.6827807426452637,
      "logits/rejected": -0.48651519417762756,
      "logps/chosen": -466.18634033203125,
      "logps/pi_response": -313.5439758300781,
      "logps/ref_response": -195.1559295654297,
      "logps/rejected": -830.7267456054688,
      "loss": 0.5474,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": -0.20106668770313263,
      "rewards/margins": 0.9386097192764282,
      "rewards/rejected": -1.1396763324737549,
      "step": 20
    },
    {
      "epoch": 0.502092050209205,
      "grad_norm": 71.37245318602794,
      "learning_rate": 5.738232820012406e-08,
      "logits/chosen": -0.5754955410957336,
      "logits/rejected": -0.37761348485946655,
      "logps/chosen": -527.6432495117188,
      "logps/pi_response": -468.9288024902344,
      "logps/ref_response": -188.6576385498047,
      "logps/rejected": -963.0016479492188,
      "loss": 0.4406,
      "rewards/accuracies": 0.846875011920929,
      "rewards/chosen": -0.714775562286377,
      "rewards/margins": 2.008810520172119,
      "rewards/rejected": -2.723586320877075,
      "step": 30
    },
    {
      "epoch": 0.6694560669456067,
      "grad_norm": 91.06844156551222,
      "learning_rate": 2.8496739886173992e-08,
      "logits/chosen": -0.4327356219291687,
      "logits/rejected": -0.13516364991664886,
      "logps/chosen": -603.9708251953125,
      "logps/pi_response": -443.88800048828125,
      "logps/ref_response": -201.4982147216797,
      "logps/rejected": -1077.4971923828125,
      "loss": 0.4142,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": -1.268905520439148,
      "rewards/margins": 2.0063252449035645,
      "rewards/rejected": -3.275231122970581,
      "step": 40
    },
    {
      "epoch": 0.8368200836820083,
      "grad_norm": 102.29535857316833,
      "learning_rate": 6.947819411632222e-09,
      "logits/chosen": -0.4187712073326111,
      "logits/rejected": -0.18879997730255127,
      "logps/chosen": -592.0279541015625,
      "logps/pi_response": -473.74993896484375,
      "logps/ref_response": -189.38746643066406,
      "logps/rejected": -1088.6204833984375,
      "loss": 0.411,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": -1.303727388381958,
      "rewards/margins": 2.3152670860290527,
      "rewards/rejected": -3.6189944744110107,
      "step": 50
    },
    {
      "epoch": 0.9874476987447699,
      "step": 59,
      "total_flos": 0.0,
      "train_loss": 0.49046951229289426,
      "train_runtime": 2660.7207,
      "train_samples_per_second": 5.744,
      "train_steps_per_second": 0.022
    }
  ],
  "logging_steps": 10,
  "max_steps": 59,
  "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
}