{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.971563981042654,
  "eval_steps": 100,
  "global_step": 104,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "grad_norm": 133.0057671141645,
      "learning_rate": 4.545454545454545e-08,
      "logits/chosen": 123.11854553222656,
      "logits/rejected": 97.00198364257812,
      "logps/chosen": -425.18585205078125,
      "logps/rejected": -424.1869201660156,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.19,
      "grad_norm": 182.25849293095231,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": 117.33037567138672,
      "logits/rejected": 136.23989868164062,
      "logps/chosen": -442.517333984375,
      "logps/rejected": -524.100830078125,
      "loss": 0.7193,
      "rewards/accuracies": 0.4791666567325592,
      "rewards/chosen": 0.04355551674962044,
      "rewards/margins": 0.04285159707069397,
      "rewards/rejected": 0.0007039175252430141,
      "step": 10
    },
    {
      "epoch": 0.38,
      "grad_norm": 98.47580983421109,
      "learning_rate": 4.885348141000122e-07,
      "logits/chosen": 124.34559631347656,
      "logits/rejected": 131.87884521484375,
      "logps/chosen": -425.8033142089844,
      "logps/rejected": -496.58966064453125,
      "loss": 0.5981,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.02306445501744747,
      "rewards/margins": 0.4770180284976959,
      "rewards/rejected": -0.4539535939693451,
      "step": 20
    },
    {
      "epoch": 0.57,
      "grad_norm": 114.087847259367,
      "learning_rate": 4.5025027361734613e-07,
      "logits/chosen": 121.997314453125,
      "logits/rejected": 125.30668640136719,
      "logps/chosen": -473.92950439453125,
      "logps/rejected": -546.4301147460938,
      "loss": 0.5828,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.9404676556587219,
      "rewards/margins": 1.0203006267547607,
      "rewards/rejected": -1.9607683420181274,
      "step": 30
    },
    {
      "epoch": 0.76,
      "grad_norm": 102.23106994021863,
      "learning_rate": 3.893311157806091e-07,
      "logits/chosen": 123.2125473022461,
      "logits/rejected": 118.11393737792969,
      "logps/chosen": -498.98480224609375,
      "logps/rejected": -529.0015869140625,
      "loss": 0.5432,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -1.29017174243927,
      "rewards/margins": 0.921674907207489,
      "rewards/rejected": -2.2118465900421143,
      "step": 40
    },
    {
      "epoch": 0.95,
      "grad_norm": 116.89709640403532,
      "learning_rate": 3.126631330646801e-07,
      "logits/chosen": 131.3624267578125,
      "logits/rejected": 131.5041046142578,
      "logps/chosen": -517.221923828125,
      "logps/rejected": -547.076904296875,
      "loss": 0.5007,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -1.7199589014053345,
      "rewards/margins": 1.0714657306671143,
      "rewards/rejected": -2.791424512863159,
      "step": 50
    },
    {
      "epoch": 1.14,
      "grad_norm": 53.67728401735822,
      "learning_rate": 2.2891223348923882e-07,
      "logits/chosen": 130.25340270996094,
      "logits/rejected": 133.3688507080078,
      "logps/chosen": -482.68798828125,
      "logps/rejected": -531.1278076171875,
      "loss": 0.2897,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": -1.2961251735687256,
      "rewards/margins": 2.0452933311462402,
      "rewards/rejected": -3.341418504714966,
      "step": 60
    },
    {
      "epoch": 1.33,
      "grad_norm": 46.49732419269788,
      "learning_rate": 1.4754491880085317e-07,
      "logits/chosen": 125.44151306152344,
      "logits/rejected": 126.0068588256836,
      "logps/chosen": -451.8477478027344,
      "logps/rejected": -573.8129272460938,
      "loss": 0.1884,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": -1.1466975212097168,
      "rewards/margins": 2.4370884895324707,
      "rewards/rejected": -3.5837860107421875,
      "step": 70
    },
    {
      "epoch": 1.52,
      "grad_norm": 48.86628168194261,
      "learning_rate": 7.775827023107834e-08,
      "logits/chosen": 129.2606201171875,
      "logits/rejected": 130.1272735595703,
      "logps/chosen": -510.95819091796875,
      "logps/rejected": -606.6770629882812,
      "loss": 0.1877,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": -1.0929005146026611,
      "rewards/margins": 2.9041507244110107,
      "rewards/rejected": -3.9970507621765137,
      "step": 80
    },
    {
      "epoch": 1.71,
      "grad_norm": 51.734548215896965,
      "learning_rate": 2.7440387297912122e-08,
      "logits/chosen": 116.0444564819336,
      "logits/rejected": 118.35429382324219,
      "logps/chosen": -488.314697265625,
      "logps/rejected": -571.8872680664062,
      "loss": 0.1774,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": -1.3073575496673584,
      "rewards/margins": 3.0684092044830322,
      "rewards/rejected": -4.375766277313232,
      "step": 90
    },
    {
      "epoch": 1.9,
      "grad_norm": 58.09774424636419,
      "learning_rate": 2.27878296044029e-09,
      "logits/chosen": 129.8248748779297,
      "logits/rejected": 120.5040512084961,
      "logps/chosen": -505.01513671875,
      "logps/rejected": -610.0936279296875,
      "loss": 0.193,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -1.331251859664917,
      "rewards/margins": 3.034428596496582,
      "rewards/rejected": -4.36568021774292,
      "step": 100
    },
    {
      "epoch": 1.9,
      "eval_logits/chosen": 99.29187774658203,
      "eval_logits/rejected": 93.2354736328125,
      "eval_logps/chosen": -493.443359375,
      "eval_logps/rejected": -520.8292236328125,
      "eval_loss": 0.4772518575191498,
      "eval_rewards/accuracies": 0.7291666865348816,
      "eval_rewards/chosen": -2.0019426345825195,
      "eval_rewards/margins": 1.4620394706726074,
      "eval_rewards/rejected": -3.463982343673706,
      "eval_runtime": 49.3044,
      "eval_samples_per_second": 15.212,
      "eval_steps_per_second": 0.487,
      "step": 100
    },
    {
      "epoch": 1.97,
      "step": 104,
      "total_flos": 0.0,
      "train_loss": 0.3880494168171516,
      "train_runtime": 1239.5887,
      "train_samples_per_second": 10.891,
      "train_steps_per_second": 0.084
    }
  ],
  "logging_steps": 10,
  "max_steps": 104,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}