{
  "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": -2.7517285346984863,
      "logits/real": -2.7709789276123047,
      "logps/generated": -844.9539794921875,
      "logps/real": -335.89251708984375,
      "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": -2.6863832473754883,
      "logits/real": -2.7313873767852783,
      "logps/generated": -753.656005859375,
      "logps/real": -260.1191711425781,
      "loss": 0.5813,
      "rewards/accuracies": 0.7430555820465088,
      "rewards/generated": -0.3012603223323822,
      "rewards/margins": 0.32114675641059875,
      "rewards/real": 0.019886476919054985,
      "step": 10
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.010989010989011e-08,
      "logits/generated": -2.4782943725585938,
      "logits/real": -2.5394017696380615,
      "logps/generated": -1096.2259521484375,
      "logps/real": -633.4000854492188,
      "loss": 5.8769,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/generated": -33.22383499145508,
      "rewards/margins": -1.7550216913223267,
      "rewards/real": -34.978858947753906,
      "step": 20
    },
    {
      "epoch": 0.49,
      "eval_logits/generated": -2.7939765453338623,
      "eval_logits/real": -2.838216781616211,
      "eval_logps/generated": -719.6648559570312,
      "eval_logps/real": -274.2817077636719,
      "eval_loss": 0.18901990354061127,
      "eval_rewards/accuracies": 0.9375,
      "eval_rewards/generated": -2.9832763671875,
      "eval_rewards/margins": 2.815262794494629,
      "eval_rewards/real": -0.16801361739635468,
      "eval_runtime": 56.889,
      "eval_samples_per_second": 6.469,
      "eval_steps_per_second": 0.211,
      "step": 25
    },
    {
      "epoch": 0.58,
      "learning_rate": 7.912087912087911e-08,
      "logits/generated": -2.823836088180542,
      "logits/real": -2.8936166763305664,
      "logps/generated": -856.4396362304688,
      "logps/real": -293.3896179199219,
      "loss": 0.1538,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -3.060887336730957,
      "rewards/margins": 2.894258975982666,
      "rewards/real": -0.1666283905506134,
      "step": 30
    },
    {
      "epoch": 0.78,
      "learning_rate": 6.813186813186813e-08,
      "logits/generated": -2.8231825828552246,
      "logits/real": -2.8379809856414795,
      "logps/generated": -769.75048828125,
      "logps/real": -256.47027587890625,
      "loss": 0.1241,
      "rewards/accuracies": 0.9937499761581421,
      "rewards/generated": -3.725442409515381,
      "rewards/margins": 3.5178775787353516,
      "rewards/real": -0.20756463706493378,
      "step": 40
    },
    {
      "epoch": 0.97,
      "learning_rate": 5.714285714285714e-08,
      "logits/generated": -2.8316774368286133,
      "logits/real": -2.8536953926086426,
      "logps/generated": -803.0462646484375,
      "logps/real": -274.77337646484375,
      "loss": 0.1202,
      "rewards/accuracies": 0.949999988079071,
      "rewards/generated": -3.983602523803711,
      "rewards/margins": 3.5648624897003174,
      "rewards/real": -0.4187401831150055,
      "step": 50
    },
    {
      "epoch": 0.97,
      "eval_logits/generated": -2.83947491645813,
      "eval_logits/real": -2.8439435958862305,
      "eval_logps/generated": -732.087890625,
      "eval_logps/real": -276.7652282714844,
      "eval_loss": 0.14404349029064178,
      "eval_rewards/accuracies": 0.9479166865348816,
      "eval_rewards/generated": -4.225581645965576,
      "eval_rewards/margins": 3.809215784072876,
      "eval_rewards/real": -0.41636598110198975,
      "eval_runtime": 55.8686,
      "eval_samples_per_second": 6.587,
      "eval_steps_per_second": 0.215,
      "step": 50
    },
    {
      "epoch": 1.17,
      "learning_rate": 4.615384615384615e-08,
      "logits/generated": -2.8372151851654053,
      "logits/real": -2.857466697692871,
      "logps/generated": -920.998046875,
      "logps/real": -288.32086181640625,
      "loss": 0.0744,
      "rewards/accuracies": 1.0,
      "rewards/generated": -4.382575035095215,
      "rewards/margins": 4.1766462326049805,
      "rewards/real": -0.20592932403087616,
      "step": 60
    },
    {
      "epoch": 1.36,
      "learning_rate": 3.516483516483517e-08,
      "logits/generated": -2.832484483718872,
      "logits/real": -2.8351359367370605,
      "logps/generated": -778.0480346679688,
      "logps/real": -260.9442443847656,
      "loss": 0.0754,
      "rewards/accuracies": 0.987500011920929,
      "rewards/generated": -4.6376495361328125,
      "rewards/margins": 4.451912879943848,
      "rewards/real": -0.18573713302612305,
      "step": 70
    },
    {
      "epoch": 1.46,
      "eval_logits/generated": -2.841118574142456,
      "eval_logits/real": -2.838819742202759,
      "eval_logps/generated": -737.3972778320312,
      "eval_logps/real": -278.0699768066406,
      "eval_loss": 0.1297575831413269,
      "eval_rewards/accuracies": 0.9583333134651184,
      "eval_rewards/generated": -4.7565155029296875,
      "eval_rewards/margins": 4.2096757888793945,
      "eval_rewards/real": -0.546840488910675,
      "eval_runtime": 57.5597,
      "eval_samples_per_second": 6.393,
      "eval_steps_per_second": 0.208,
      "step": 75
    },
    {
      "epoch": 1.55,
      "learning_rate": 2.4175824175824175e-08,
      "logits/generated": -2.8195009231567383,
      "logits/real": -2.84141206741333,
      "logps/generated": -862.6012573242188,
      "logps/real": -276.63531494140625,
      "loss": 0.0666,
      "rewards/accuracies": 0.9937499761581421,
      "rewards/generated": -5.086130619049072,
      "rewards/margins": 4.704850196838379,
      "rewards/real": -0.3812801241874695,
      "step": 80
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.3186813186813187e-08,
      "logits/generated": -2.8254876136779785,
      "logits/real": -2.823962450027466,
      "logps/generated": -835.1746215820312,
      "logps/real": -272.4798278808594,
      "loss": 0.0572,
      "rewards/accuracies": 1.0,
      "rewards/generated": -5.038660526275635,
      "rewards/margins": 4.638853073120117,
      "rewards/real": -0.39980727434158325,
      "step": 90
    },
    {
      "epoch": 1.94,
      "learning_rate": 2.197802197802198e-09,
      "logits/generated": -2.8575081825256348,
      "logits/real": -2.859795093536377,
      "logps/generated": -838.92919921875,
      "logps/real": -267.28704833984375,
      "loss": 0.0621,
      "rewards/accuracies": 0.9937499761581421,
      "rewards/generated": -5.040513038635254,
      "rewards/margins": 4.70792818069458,
      "rewards/real": -0.33258455991744995,
      "step": 100
    },
    {
      "epoch": 1.94,
      "eval_logits/generated": -2.84299635887146,
      "eval_logits/real": -2.8375051021575928,
      "eval_logps/generated": -739.3701171875,
      "eval_logps/real": -278.2850646972656,
      "eval_loss": 0.12532015144824982,
      "eval_rewards/accuracies": 0.9479166865348816,
      "eval_rewards/generated": -4.953795433044434,
      "eval_rewards/margins": 4.385446071624756,
      "eval_rewards/real": -0.5683497786521912,
      "eval_runtime": 56.3549,
      "eval_samples_per_second": 6.53,
      "eval_steps_per_second": 0.213,
      "step": 100
    },
    {
      "epoch": 1.98,
      "step": 102,
      "total_flos": 0.0,
      "train_loss": 0.7072318076503044,
      "train_runtime": 2304.549,
      "train_samples_per_second": 2.86,
      "train_steps_per_second": 0.044
    }
  ],
  "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
}