{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.992,
  "eval_steps": 200,
  "global_step": 62,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.016,
      "grad_norm": 198.7469829190425,
      "learning_rate": 7.142857142857142e-08,
      "logits/generated": -2.8518388271331787,
      "logits/real": -2.9862747192382812,
      "logps/generated": -241.12493896484375,
      "logps/real": -248.8436279296875,
      "loss": 0.9204,
      "rewards/accuracies": 0.0,
      "rewards/generated": 0.0,
      "rewards/margins": 0.0,
      "rewards/real": 0.0,
      "step": 1
    },
    {
      "epoch": 0.16,
      "grad_norm": 191.54263414525988,
      "learning_rate": 4.727272727272727e-07,
      "logits/generated": -3.05552339553833,
      "logits/real": -3.0378544330596924,
      "logps/generated": -243.76849365234375,
      "logps/real": -230.84719848632812,
      "loss": 0.8961,
      "rewards/accuracies": 0.5833333134651184,
      "rewards/generated": 0.18684834241867065,
      "rewards/margins": 0.20998667180538177,
      "rewards/real": 0.396835058927536,
      "step": 10
    },
    {
      "epoch": 0.32,
      "grad_norm": 139.26955693169643,
      "learning_rate": 3.818181818181818e-07,
      "logits/generated": -2.9985575675964355,
      "logits/real": -2.984205484390259,
      "logps/generated": -250.49853515625,
      "logps/real": -217.1739044189453,
      "loss": 0.8668,
      "rewards/accuracies": 0.675000011920929,
      "rewards/generated": -1.4985097646713257,
      "rewards/margins": 0.8482303619384766,
      "rewards/real": -0.6502794623374939,
      "step": 20
    },
    {
      "epoch": 0.48,
      "grad_norm": 117.51802509366077,
      "learning_rate": 2.909090909090909e-07,
      "logits/generated": -2.9175362586975098,
      "logits/real": -2.964449644088745,
      "logps/generated": -262.0832824707031,
      "logps/real": -228.2645721435547,
      "loss": 0.7592,
      "rewards/accuracies": 0.8125,
      "rewards/generated": -1.5061546564102173,
      "rewards/margins": 1.8053114414215088,
      "rewards/real": 0.29915687441825867,
      "step": 30
    },
    {
      "epoch": 0.64,
      "grad_norm": 176.64698900823225,
      "learning_rate": 2e-07,
      "logits/generated": -2.9176225662231445,
      "logits/real": -2.9303154945373535,
      "logps/generated": -260.50018310546875,
      "logps/real": -230.49288940429688,
      "loss": 0.6242,
      "rewards/accuracies": 0.824999988079071,
      "rewards/generated": -2.2268033027648926,
      "rewards/margins": 2.359907627105713,
      "rewards/real": 0.13310480117797852,
      "step": 40
    },
    {
      "epoch": 0.8,
      "grad_norm": 183.16237322543932,
      "learning_rate": 1.0909090909090908e-07,
      "logits/generated": -2.88456654548645,
      "logits/real": -2.944023609161377,
      "logps/generated": -260.68035888671875,
      "logps/real": -228.1358184814453,
      "loss": 0.604,
      "rewards/accuracies": 0.800000011920929,
      "rewards/generated": -1.7078344821929932,
      "rewards/margins": 1.941307783126831,
      "rewards/real": 0.23347334563732147,
      "step": 50
    },
    {
      "epoch": 0.96,
      "grad_norm": 95.0210404383474,
      "learning_rate": 1.818181818181818e-08,
      "logits/generated": -2.917457342147827,
      "logits/real": -2.9479122161865234,
      "logps/generated": -249.3968048095703,
      "logps/real": -225.28262329101562,
      "loss": 0.6543,
      "rewards/accuracies": 0.824999988079071,
      "rewards/generated": -1.3457591533660889,
      "rewards/margins": 1.6494863033294678,
      "rewards/real": 0.30372729897499084,
      "step": 60
    },
    {
      "epoch": 0.992,
      "step": 62,
      "total_flos": 0.0,
      "train_loss": 0.7342928696063257,
      "train_runtime": 746.6525,
      "train_samples_per_second": 2.677,
      "train_steps_per_second": 0.083
    }
  ],
  "logging_steps": 10,
  "max_steps": 62,
  "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
}