{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.3929273084479371,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09823182711198428,
      "grad_norm": 4.68359375,
      "learning_rate": 5e-07,
      "logits/chosen": -1.882665991783142,
      "logits/rejected": -2.022334098815918,
      "logps/chosen": -629.2000122070312,
      "logps/rejected": -886.400146484375,
      "loss": 0.6197,
      "rewards/accuracies": 0.39500001072883606,
      "rewards/chosen": 1.2083027362823486,
      "rewards/margins": 0.816202700138092,
      "rewards/rejected": 0.392100065946579,
      "step": 50
    },
    {
      "epoch": 0.09823182711198428,
      "eval_logits/chosen": -1.8351960182189941,
      "eval_logits/rejected": -1.8982521295547485,
      "eval_logps/chosen": -827.6643676757812,
      "eval_logps/rejected": -769.7268676757812,
      "eval_loss": 0.21161487698554993,
      "eval_rewards/accuracies": 0.7060185074806213,
      "eval_rewards/chosen": 4.600923538208008,
      "eval_rewards/margins": 3.7983312606811523,
      "eval_rewards/rejected": 0.8025919795036316,
      "eval_runtime": 376.5489,
      "eval_samples_per_second": 1.145,
      "eval_steps_per_second": 0.574,
      "step": 50
    },
    {
      "epoch": 0.19646365422396855,
      "grad_norm": 1.431640625,
      "learning_rate": 1e-06,
      "logits/chosen": -1.9131250381469727,
      "logits/rejected": -2.0282812118530273,
      "logps/chosen": -584.5037231445312,
      "logps/rejected": -919.989990234375,
      "loss": 0.3185,
      "rewards/accuracies": 0.5550000071525574,
      "rewards/chosen": 4.304075241088867,
      "rewards/margins": 5.650895595550537,
      "rewards/rejected": -1.3468197584152222,
      "step": 100
    },
    {
      "epoch": 0.19646365422396855,
      "eval_logits/chosen": -1.8304036855697632,
      "eval_logits/rejected": -1.8951867818832397,
      "eval_logps/chosen": -808.5324096679688,
      "eval_logps/rejected": -775.2742919921875,
      "eval_loss": 0.2037331759929657,
      "eval_rewards/accuracies": 0.7083333134651184,
      "eval_rewards/chosen": 6.5165486335754395,
      "eval_rewards/margins": 6.269077777862549,
      "eval_rewards/rejected": 0.24747034907341003,
      "eval_runtime": 376.7007,
      "eval_samples_per_second": 1.144,
      "eval_steps_per_second": 0.573,
      "step": 100
    },
    {
      "epoch": 0.29469548133595286,
      "grad_norm": 0.00038909912109375,
      "learning_rate": 5e-07,
      "logits/chosen": -1.900234341621399,
      "logits/rejected": -1.9972655773162842,
      "logps/chosen": -539.0875244140625,
      "logps/rejected": -859.8375244140625,
      "loss": 0.334,
      "rewards/accuracies": 0.5274999737739563,
      "rewards/chosen": 4.29106330871582,
      "rewards/margins": 8.129581451416016,
      "rewards/rejected": -3.838517427444458,
      "step": 150
    },
    {
      "epoch": 0.29469548133595286,
      "eval_logits/chosen": -1.8221435546875,
      "eval_logits/rejected": -1.8842185735702515,
      "eval_logps/chosen": -818.3194580078125,
      "eval_logps/rejected": -790.0972290039062,
      "eval_loss": 0.20332251489162445,
      "eval_rewards/accuracies": 0.7083333134651184,
      "eval_rewards/chosen": 5.53895902633667,
      "eval_rewards/margins": 6.775155067443848,
      "eval_rewards/rejected": -1.236195683479309,
      "eval_runtime": 376.4567,
      "eval_samples_per_second": 1.145,
      "eval_steps_per_second": 0.574,
      "step": 150
    },
    {
      "epoch": 0.3929273084479371,
      "grad_norm": 0.80126953125,
      "learning_rate": 0.0,
      "logits/chosen": -1.889101505279541,
      "logits/rejected": NaN,
      "logps/chosen": -537.0787353515625,
      "logps/rejected": -937.9553833007812,
      "loss": 0.3,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 4.2822651863098145,
      "rewards/margins": 9.660238265991211,
      "rewards/rejected": -5.377972602844238,
      "step": 200
    },
    {
      "epoch": 0.3929273084479371,
      "eval_logits/chosen": -1.8241690397262573,
      "eval_logits/rejected": -1.8862168788909912,
      "eval_logps/chosen": -818.3449096679688,
      "eval_logps/rejected": -792.08447265625,
      "eval_loss": 0.20316672325134277,
      "eval_rewards/accuracies": 0.7083333134651184,
      "eval_rewards/chosen": 5.535775661468506,
      "eval_rewards/margins": 6.967732906341553,
      "eval_rewards/rejected": -1.4319567680358887,
      "eval_runtime": 376.1581,
      "eval_samples_per_second": 1.146,
      "eval_steps_per_second": 0.574,
      "step": 200
    },
    {
      "epoch": 0.3929273084479371,
      "step": 200,
      "total_flos": 0.0,
      "train_loss": 0.39305298328399657,
      "train_runtime": 44907.6338,
      "train_samples_per_second": 0.036,
      "train_steps_per_second": 0.004
    }
  ],
  "logging_steps": 50,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}