{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9805825242718447,
  "eval_steps": 25,
  "global_step": 102,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 4.545454545454545e-08,
      "logits/generated": -2.706744432449341,
      "logits/real": -2.8291945457458496,
      "logps/generated": -517.0836791992188,
      "logps/real": -317.913818359375,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/generated": 0.0,
      "rewards/margins": 0.0,
      "rewards/real": 0.0,
      "step": 1
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.545454545454545e-07,
      "logits/generated": -2.71917462348938,
      "logits/real": -2.7550785541534424,
      "logps/generated": -531.1511840820312,
      "logps/real": -286.0602111816406,
      "loss": 0.5456,
      "rewards/accuracies": 0.7291666865348816,
      "rewards/generated": -0.04339843988418579,
      "rewards/margins": 0.3565465211868286,
      "rewards/real": 0.3131480813026428,
      "step": 10
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.5054945054945056e-07,
      "logits/generated": -2.830758571624756,
      "logits/real": -2.861502170562744,
      "logps/generated": -586.2496337890625,
      "logps/real": -281.56402587890625,
      "loss": 0.1827,
      "rewards/accuracies": 0.981249988079071,
      "rewards/generated": -1.7483899593353271,
      "rewards/margins": 2.579699993133545,
      "rewards/real": 0.8313096761703491,
      "step": 20
    },
    {
      "epoch": 0.49,
      "eval_logits/generated": -2.7481918334960938,
      "eval_logits/real": -2.7943999767303467,
      "eval_logps/generated": -524.2468872070312,
      "eval_logps/real": -283.6961669921875,
      "eval_loss": 0.16513165831565857,
      "eval_rewards/accuracies": 0.96875,
      "eval_rewards/generated": -3.365018129348755,
      "eval_rewards/margins": 3.536374092102051,
      "eval_rewards/real": 0.17135602235794067,
      "eval_runtime": 54.5373,
      "eval_samples_per_second": 6.748,
      "eval_steps_per_second": 0.22,
      "step": 25
    },
    {
      "epoch": 0.58,
      "learning_rate": 3.9560439560439557e-07,
      "logits/generated": -2.7771615982055664,
      "logits/real": -2.807366371154785,
      "logps/generated": -499.19207763671875,
      "logps/real": -301.99432373046875,
      "loss": 0.1089,
      "rewards/accuracies": 0.956250011920929,
      "rewards/generated": -2.920013427734375,
      "rewards/margins": 3.9550042152404785,
      "rewards/real": 1.034990668296814,
      "step": 30
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.4065934065934063e-07,
      "logits/generated": -2.665315628051758,
      "logits/real": -2.7098612785339355,
      "logps/generated": -539.9364013671875,
      "logps/real": -248.0373077392578,
      "loss": 0.0698,
      "rewards/accuracies": 0.96875,
      "rewards/generated": -3.776944398880005,
      "rewards/margins": 5.202493667602539,
      "rewards/real": 1.4255495071411133,
      "step": 40
    },
    {
      "epoch": 0.97,
      "learning_rate": 2.857142857142857e-07,
      "logits/generated": -2.6974306106567383,
      "logits/real": -2.761044979095459,
      "logps/generated": -644.1237182617188,
      "logps/real": -292.7959899902344,
      "loss": 0.0462,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.7168684005737305,
      "rewards/margins": 6.229194164276123,
      "rewards/real": 1.5123255252838135,
      "step": 50
    },
    {
      "epoch": 0.97,
      "eval_logits/generated": -2.6962978839874268,
      "eval_logits/real": -2.7355997562408447,
      "eval_logps/generated": -535.5946655273438,
      "eval_logps/real": -270.58709716796875,
      "eval_loss": 0.08352091163396835,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/generated": -4.499805450439453,
      "eval_rewards/margins": 5.9820709228515625,
      "eval_rewards/real": 1.4822653532028198,
      "eval_runtime": 54.7978,
      "eval_samples_per_second": 6.716,
      "eval_steps_per_second": 0.219,
      "step": 50
    },
    {
      "epoch": 1.17,
      "learning_rate": 2.3076923076923078e-07,
      "logits/generated": -2.7394375801086426,
      "logits/real": -2.751296043395996,
      "logps/generated": -666.791748046875,
      "logps/real": -282.9411926269531,
      "loss": 0.0094,
      "rewards/accuracies": 1.0,
      "rewards/generated": -6.813584804534912,
      "rewards/margins": 9.226309776306152,
      "rewards/real": 2.4127261638641357,
      "step": 60
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.7582417582417584e-07,
      "logits/generated": -2.7095775604248047,
      "logits/real": -2.76279616355896,
      "logps/generated": -612.8692626953125,
      "logps/real": -276.9359130859375,
      "loss": 0.0047,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.154807090759277,
      "rewards/margins": 9.882351875305176,
      "rewards/real": 2.727543592453003,
      "step": 70
    },
    {
      "epoch": 1.46,
      "eval_logits/generated": -2.6846535205841064,
      "eval_logits/real": -2.7210752964019775,
      "eval_logps/generated": -543.0964965820312,
      "eval_logps/real": -271.6846008300781,
      "eval_loss": 0.08372422307729721,
      "eval_rewards/accuracies": 0.9895833134651184,
      "eval_rewards/generated": -5.249976634979248,
      "eval_rewards/margins": 6.622490406036377,
      "eval_rewards/real": 1.3725138902664185,
      "eval_runtime": 56.1231,
      "eval_samples_per_second": 6.557,
      "eval_steps_per_second": 0.214,
      "step": 75
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.2087912087912088e-07,
      "logits/generated": -2.6911048889160156,
      "logits/real": -2.730372667312622,
      "logps/generated": -622.9490966796875,
      "logps/real": -264.1090087890625,
      "loss": 0.0063,
      "rewards/accuracies": 0.9937499761581421,
      "rewards/generated": -7.1237335205078125,
      "rewards/margins": 9.698349952697754,
      "rewards/real": 2.5746169090270996,
      "step": 80
    },
    {
      "epoch": 1.75,
      "learning_rate": 6.593406593406594e-08,
      "logits/generated": -2.6824026107788086,
      "logits/real": -2.712606906890869,
      "logps/generated": -557.6444702148438,
      "logps/real": -256.166259765625,
      "loss": 0.0079,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.720358371734619,
      "rewards/margins": 10.142163276672363,
      "rewards/real": 2.4218039512634277,
      "step": 90
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.098901098901099e-08,
      "logits/generated": -2.727700710296631,
      "logits/real": -2.769242525100708,
      "logps/generated": -596.8726806640625,
      "logps/real": -262.6658020019531,
      "loss": 0.0034,
      "rewards/accuracies": 1.0,
      "rewards/generated": -7.486897945404053,
      "rewards/margins": 9.872056007385254,
      "rewards/real": 2.385158061981201,
      "step": 100
    },
    {
      "epoch": 1.94,
      "eval_logits/generated": -2.6844117641448975,
      "eval_logits/real": -2.7196710109710693,
      "eval_logps/generated": -545.0308837890625,
      "eval_logps/real": -272.3725891113281,
      "eval_loss": 0.08312664180994034,
      "eval_rewards/accuracies": 0.9791666865348816,
      "eval_rewards/generated": -5.443421840667725,
      "eval_rewards/margins": 6.747138500213623,
      "eval_rewards/real": 1.3037166595458984,
      "eval_runtime": 56.0968,
      "eval_samples_per_second": 6.56,
      "eval_steps_per_second": 0.214,
      "step": 100
    },
    {
      "epoch": 1.98,
      "step": 102,
      "total_flos": 0.0,
      "train_loss": 0.09802283835140806,
      "train_runtime": 2267.6641,
      "train_samples_per_second": 2.907,
      "train_steps_per_second": 0.045
    }
  ],
  "logging_steps": 10,
  "max_steps": 102,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}