{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9230769230769231,
  "eval_steps": 500,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.19230769230769232,
      "grad_norm": 8.625,
      "learning_rate": 1.1600077849629454e-05,
      "logits/chosen": -3.838188886642456,
      "logits/rejected": -3.871121644973755,
      "logps/chosen": -29.310901641845703,
      "logps/rejected": -31.50802230834961,
      "loss": 0.6997,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.0766543298959732,
      "rewards/margins": -0.004551754333078861,
      "rewards/rejected": -0.07210256904363632,
      "step": 5
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 7.65625,
      "learning_rate": 2.610017516166627e-05,
      "logits/chosen": -3.865746259689331,
      "logits/rejected": -3.8615641593933105,
      "logps/chosen": -28.9598445892334,
      "logps/rejected": -31.76461410522461,
      "loss": 0.6813,
      "rewards/accuracies": 0.5531250238418579,
      "rewards/chosen": -0.06299854069948196,
      "rewards/margins": 0.03494865447282791,
      "rewards/rejected": -0.09794719517230988,
      "step": 10
    },
    {
      "epoch": 0.5769230769230769,
      "grad_norm": 8.125,
      "learning_rate": 4.060027247370309e-05,
      "logits/chosen": -3.7930521965026855,
      "logits/rejected": -3.7284348011016846,
      "logps/chosen": -29.46489906311035,
      "logps/rejected": -32.92306900024414,
      "loss": 0.6365,
      "rewards/accuracies": 0.6656249761581421,
      "rewards/chosen": -0.05547923967242241,
      "rewards/margins": 0.16162915527820587,
      "rewards/rejected": -0.21710839867591858,
      "step": 15
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 8.75,
      "learning_rate": 5.51003697857399e-05,
      "logits/chosen": -3.5052037239074707,
      "logits/rejected": -3.504037380218506,
      "logps/chosen": -30.033306121826172,
      "logps/rejected": -34.85963439941406,
      "loss": 0.5924,
      "rewards/accuracies": 0.6468750238418579,
      "rewards/chosen": -0.08518950641155243,
      "rewards/margins": 0.3314022421836853,
      "rewards/rejected": -0.41659173369407654,
      "step": 20
    },
    {
      "epoch": 0.9615384615384616,
      "grad_norm": 8.9375,
      "learning_rate": 6.960046709777673e-05,
      "logits/chosen": -3.224011182785034,
      "logits/rejected": -3.1782708168029785,
      "logps/chosen": -29.593297958374023,
      "logps/rejected": -36.04486846923828,
      "loss": 0.5711,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.07951204478740692,
      "rewards/margins": 0.46893948316574097,
      "rewards/rejected": -0.5484515428543091,
      "step": 25
    },
    {
      "epoch": 0.9615384615384616,
      "eval_logits/chosen": -3.0905661582946777,
      "eval_logits/rejected": -3.0327272415161133,
      "eval_logps/chosen": -28.399259567260742,
      "eval_logps/rejected": -35.722633361816406,
      "eval_loss": 0.5570524334907532,
      "eval_rewards/accuracies": 0.7200000286102295,
      "eval_rewards/chosen": -0.03617068752646446,
      "eval_rewards/margins": 0.4665118455886841,
      "eval_rewards/rejected": -0.5026825070381165,
      "eval_runtime": 2.4762,
      "eval_samples_per_second": 80.77,
      "eval_steps_per_second": 80.77,
      "step": 25
    },
    {
      "epoch": 1.1538461538461537,
      "grad_norm": 8.25,
      "learning_rate": 8.410056440981355e-05,
      "logits/chosen": -2.9339802265167236,
      "logits/rejected": -2.896488904953003,
      "logps/chosen": -27.779123306274414,
      "logps/rejected": -36.54551315307617,
      "loss": 0.4732,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 0.12445543706417084,
      "rewards/margins": 0.7573782801628113,
      "rewards/rejected": -0.632922887802124,
      "step": 30
    },
    {
      "epoch": 1.3461538461538463,
      "grad_norm": 8.75,
      "learning_rate": 9.860066172185035e-05,
      "logits/chosen": -2.742870330810547,
      "logits/rejected": -2.764082670211792,
      "logps/chosen": -25.611190795898438,
      "logps/rejected": -39.37268829345703,
      "loss": 0.419,
      "rewards/accuracies": 0.8343750238418579,
      "rewards/chosen": 0.31286072731018066,
      "rewards/margins": 1.1506929397583008,
      "rewards/rejected": -0.8378321528434753,
      "step": 35
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 8.6875,
      "learning_rate": 9.156664142733769e-05,
      "logits/chosen": -2.7378015518188477,
      "logits/rejected": -2.7470693588256836,
      "logps/chosen": -28.239633560180664,
      "logps/rejected": -43.21832275390625,
      "loss": 0.4162,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 0.05981593206524849,
      "rewards/margins": 1.2637542486190796,
      "rewards/rejected": -1.203938364982605,
      "step": 40
    },
    {
      "epoch": 1.7307692307692308,
      "grad_norm": 8.4375,
      "learning_rate": 5.992593773868142e-05,
      "logits/chosen": -2.7000434398651123,
      "logits/rejected": -2.720876932144165,
      "logps/chosen": -28.078105926513672,
      "logps/rejected": -43.50415802001953,
      "loss": 0.3869,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 0.0571039617061615,
      "rewards/margins": 1.3791061639785767,
      "rewards/rejected": -1.3220020532608032,
      "step": 45
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 8.4375,
      "learning_rate": 3.1076318828504456e-05,
      "logits/chosen": -2.6507811546325684,
      "logits/rejected": -2.614593505859375,
      "logps/chosen": -27.470378875732422,
      "logps/rejected": -42.90967559814453,
      "loss": 0.3851,
      "rewards/accuracies": 0.846875011920929,
      "rewards/chosen": 0.16307835280895233,
      "rewards/margins": 1.390608787536621,
      "rewards/rejected": -1.2275304794311523,
      "step": 50
    },
    {
      "epoch": 1.9230769230769231,
      "eval_logits/chosen": -2.7256622314453125,
      "eval_logits/rejected": -2.653876304626465,
      "eval_logps/chosen": -28.530363082885742,
      "eval_logps/rejected": -39.620460510253906,
      "eval_loss": 0.5317701101303101,
      "eval_rewards/accuracies": 0.7250000238418579,
      "eval_rewards/chosen": -0.049281273037195206,
      "eval_rewards/margins": 0.843184232711792,
      "eval_rewards/rejected": -0.8924655318260193,
      "eval_runtime": 2.3888,
      "eval_samples_per_second": 83.724,
      "eval_steps_per_second": 83.724,
      "step": 50
    }
  ],
  "logging_steps": 5,
  "max_steps": 52,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}