{
  "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": 9.09090909090909e-09,
      "logits/generated": -2.911376476287842,
      "logits/real": -2.89278507232666,
      "logps/generated": -668.016845703125,
      "logps/real": -324.0830383300781,
      "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": 9.09090909090909e-08,
      "logits/generated": -2.848540782928467,
      "logits/real": -2.8350822925567627,
      "logps/generated": -617.5262451171875,
      "logps/real": -255.12559509277344,
      "loss": 0.8027,
      "rewards/accuracies": 0.6527777910232544,
      "rewards/generated": -6.766346454620361,
      "rewards/margins": 4.9929962158203125,
      "rewards/real": -1.7733497619628906,
      "step": 10
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.010989010989011e-08,
      "logits/generated": -2.8388895988464355,
      "logits/real": -2.8330814838409424,
      "logps/generated": -707.9961547851562,
      "logps/real": -283.4620056152344,
      "loss": 0.2928,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/generated": -20.150367736816406,
      "rewards/margins": 17.525970458984375,
      "rewards/real": -2.6243953704833984,
      "step": 20
    },
    {
      "epoch": 0.49,
      "eval_logits/generated": -2.8098323345184326,
      "eval_logits/real": -2.8089659214019775,
      "eval_logps/generated": -700.5637817382812,
      "eval_logps/real": -278.0876159667969,
      "eval_loss": 0.3950892686843872,
      "eval_rewards/accuracies": 0.90625,
      "eval_rewards/generated": -20.326812744140625,
      "eval_rewards/margins": 17.705642700195312,
      "eval_rewards/real": -2.6211698055267334,
      "eval_runtime": 75.909,
      "eval_samples_per_second": 4.848,
      "eval_steps_per_second": 0.158,
      "step": 25
    },
    {
      "epoch": 0.58,
      "learning_rate": 7.912087912087911e-08,
      "logits/generated": -2.877678394317627,
      "logits/real": -2.8707680702209473,
      "logps/generated": -755.6437377929688,
      "logps/real": -290.04058837890625,
      "loss": 0.2615,
      "rewards/accuracies": 0.96875,
      "rewards/generated": -25.02899742126465,
      "rewards/margins": 23.474206924438477,
      "rewards/real": -1.5547949075698853,
      "step": 30
    },
    {
      "epoch": 0.78,
      "learning_rate": 6.813186813186813e-08,
      "logits/generated": -2.839968681335449,
      "logits/real": -2.8121862411499023,
      "logps/generated": -673.220703125,
      "logps/real": -257.77001953125,
      "loss": 0.1816,
      "rewards/accuracies": 0.956250011920929,
      "rewards/generated": -24.732364654541016,
      "rewards/margins": 26.96199607849121,
      "rewards/real": 2.2296276092529297,
      "step": 40
    },
    {
      "epoch": 0.97,
      "learning_rate": 5.714285714285714e-08,
      "logits/generated": -2.8406929969787598,
      "logits/real": -2.8268864154815674,
      "logps/generated": -727.8557739257812,
      "logps/real": -274.7761535644531,
      "loss": 0.1487,
      "rewards/accuracies": 0.981249988079071,
      "rewards/generated": -29.384685516357422,
      "rewards/margins": 30.1076717376709,
      "rewards/real": 0.7229812145233154,
      "step": 50
    },
    {
      "epoch": 0.97,
      "eval_logits/generated": -2.821841239929199,
      "eval_logits/real": -2.8065903186798096,
      "eval_logps/generated": -702.32763671875,
      "eval_logps/real": -278.1448974609375,
      "eval_loss": 0.131888747215271,
      "eval_rewards/accuracies": 0.9375,
      "eval_rewards/generated": -29.145910263061523,
      "eval_rewards/margins": 26.238248825073242,
      "eval_rewards/real": -2.9076578617095947,
      "eval_runtime": 74.4579,
      "eval_samples_per_second": 4.942,
      "eval_steps_per_second": 0.161,
      "step": 50
    },
    {
      "epoch": 1.17,
      "learning_rate": 4.615384615384615e-08,
      "logits/generated": -2.8634510040283203,
      "logits/real": -2.8573107719421387,
      "logps/generated": -768.2184448242188,
      "logps/real": -281.481201171875,
      "loss": 0.0249,
      "rewards/accuracies": 0.9937499761581421,
      "rewards/generated": -33.38450241088867,
      "rewards/margins": 33.606407165527344,
      "rewards/real": 0.22189739346504211,
      "step": 60
    },
    {
      "epoch": 1.36,
      "learning_rate": 3.516483516483517e-08,
      "logits/generated": -2.846198797225952,
      "logits/real": -2.818643093109131,
      "logps/generated": -667.6427001953125,
      "logps/real": -262.6141052246094,
      "loss": 0.006,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -29.605731964111328,
      "rewards/margins": 33.2759895324707,
      "rewards/real": 3.670259952545166,
      "step": 70
    },
    {
      "epoch": 1.46,
      "eval_logits/generated": -2.8174571990966797,
      "eval_logits/real": -2.8036582469940186,
      "eval_logps/generated": -702.328857421875,
      "eval_logps/real": -278.0840759277344,
      "eval_loss": 0.12692251801490784,
      "eval_rewards/accuracies": 0.9583333134651184,
      "eval_rewards/generated": -29.151870727539062,
      "eval_rewards/margins": 26.54817008972168,
      "eval_rewards/real": -2.6037025451660156,
      "eval_runtime": 74.2997,
      "eval_samples_per_second": 4.953,
      "eval_steps_per_second": 0.162,
      "step": 75
    },
    {
      "epoch": 1.55,
      "learning_rate": 2.4175824175824175e-08,
      "logits/generated": -2.8320305347442627,
      "logits/real": -2.810762882232666,
      "logps/generated": -708.1215209960938,
      "logps/real": -277.090576171875,
      "loss": 0.0292,
      "rewards/accuracies": 0.9937499761581421,
      "rewards/generated": -33.68733215332031,
      "rewards/margins": 36.958492279052734,
      "rewards/real": 3.271162748336792,
      "step": 80
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.3186813186813187e-08,
      "logits/generated": -2.8205840587615967,
      "logits/real": -2.8130674362182617,
      "logps/generated": -704.9281005859375,
      "logps/real": -273.8406066894531,
      "loss": 0.0322,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -31.48798179626465,
      "rewards/margins": 33.98870849609375,
      "rewards/real": 2.500725746154785,
      "step": 90
    },
    {
      "epoch": 1.94,
      "learning_rate": 2.197802197802198e-09,
      "logits/generated": -2.872133255004883,
      "logits/real": -2.856153964996338,
      "logps/generated": -711.3128662109375,
      "logps/real": -259.7333068847656,
      "loss": 0.0086,
      "rewards/accuracies": 1.0,
      "rewards/generated": -35.352970123291016,
      "rewards/margins": 39.97649383544922,
      "rewards/real": 4.623520851135254,
      "step": 100
    },
    {
      "epoch": 1.94,
      "eval_logits/generated": -2.817664384841919,
      "eval_logits/real": -2.805149793624878,
      "eval_logps/generated": -702.4378051757812,
      "eval_logps/real": -278.1470031738281,
      "eval_loss": 0.10993197560310364,
      "eval_rewards/accuracies": 0.9270833134651184,
      "eval_rewards/generated": -29.69696617126465,
      "eval_rewards/margins": 26.778860092163086,
      "eval_rewards/real": -2.918107748031616,
      "eval_runtime": 73.9005,
      "eval_samples_per_second": 4.98,
      "eval_steps_per_second": 0.162,
      "step": 100
    },
    {
      "epoch": 1.98,
      "step": 102,
      "total_flos": 0.0,
      "train_loss": 0.1742458962578573,
      "train_runtime": 2590.1164,
      "train_samples_per_second": 2.545,
      "train_steps_per_second": 0.039
    }
  ],
  "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
}