{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.888888888888889,
  "eval_steps": 50,
  "global_step": 39,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 63.770220457157116,
      "learning_rate": 5e-07,
      "logits/chosen": -2.7140142917633057,
      "logits/rejected": -2.7324626445770264,
      "logps/chosen": -298.1173400878906,
      "logps/rejected": -224.2225799560547,
      "loss": 0.69,
      "rewards/accuracies": 0.375,
      "rewards/chosen": 0.016858218237757683,
      "rewards/margins": 0.008287839591503143,
      "rewards/rejected": 0.00857037864625454,
      "step": 5
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 53.75428697597037,
      "learning_rate": 1e-06,
      "logits/chosen": -2.687696933746338,
      "logits/rejected": -2.678734302520752,
      "logps/chosen": -295.27423095703125,
      "logps/rejected": -219.59884643554688,
      "loss": 0.6109,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.633124828338623,
      "rewards/margins": 0.3239261209964752,
      "rewards/rejected": 0.3091987073421478,
      "step": 10
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 28.262241530375384,
      "learning_rate": 9.284285880837946e-07,
      "logits/chosen": -2.508101224899292,
      "logits/rejected": -2.5198655128479004,
      "logps/chosen": -254.76724243164062,
      "logps/rejected": -196.24545288085938,
      "loss": 0.4795,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 1.860290765762329,
      "rewards/margins": 1.0879584550857544,
      "rewards/rejected": 0.7723323106765747,
      "step": 15
    },
    {
      "epoch": 1.4814814814814814,
      "grad_norm": 27.47403396305747,
      "learning_rate": 7.342042203498951e-07,
      "logits/chosen": -2.3905763626098633,
      "logits/rejected": -2.3961052894592285,
      "logps/chosen": -260.18084716796875,
      "logps/rejected": -234.7816619873047,
      "loss": 0.3137,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 2.7386152744293213,
      "rewards/margins": 2.408665418624878,
      "rewards/rejected": 0.32995009422302246,
      "step": 20
    },
    {
      "epoch": 1.8518518518518519,
      "grad_norm": 26.89872618716188,
      "learning_rate": 4.7293054570729126e-07,
      "logits/chosen": -2.3210275173187256,
      "logits/rejected": -2.3373942375183105,
      "logps/chosen": -246.31887817382812,
      "logps/rejected": -212.325439453125,
      "loss": 0.3025,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 3.0468170642852783,
      "rewards/margins": 3.434434413909912,
      "rewards/rejected": -0.3876175880432129,
      "step": 25
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 13.707539353724,
      "learning_rate": 2.1940646731880885e-07,
      "logits/chosen": -2.3717896938323975,
      "logits/rejected": -2.3635499477386475,
      "logps/chosen": -247.0383758544922,
      "logps/rejected": -233.0450439453125,
      "loss": 0.2036,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 3.3301596641540527,
      "rewards/margins": 4.297868251800537,
      "rewards/rejected": -0.967708945274353,
      "step": 30
    },
    {
      "epoch": 2.5925925925925926,
      "grad_norm": 13.424342173549588,
      "learning_rate": 4.621229016452155e-08,
      "logits/chosen": -2.376643657684326,
      "logits/rejected": -2.4049289226531982,
      "logps/chosen": -252.43588256835938,
      "logps/rejected": -251.13314819335938,
      "loss": 0.167,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 3.558809757232666,
      "rewards/margins": 5.30623722076416,
      "rewards/rejected": -1.747427225112915,
      "step": 35
    },
    {
      "epoch": 2.888888888888889,
      "step": 39,
      "total_flos": 459455468666880.0,
      "train_loss": 0.3703473011652629,
      "train_runtime": 1882.4229,
      "train_samples_per_second": 5.449,
      "train_steps_per_second": 0.021
    }
  ],
  "logging_steps": 5,
  "max_steps": 39,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 459455468666880.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}