{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9968652037617555,
  "eval_steps": 500,
  "global_step": 159,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 3.125e-08,
      "logits/chosen": -2.689429759979248,
      "logits/rejected": -2.605907440185547,
      "logps/chosen": -143.16458129882812,
      "logps/rejected": -148.07847595214844,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.06,
      "learning_rate": 3.1249999999999997e-07,
      "logits/chosen": -2.69895076751709,
      "logits/rejected": -2.6911985874176025,
      "logps/chosen": -234.76565551757812,
      "logps/rejected": -260.6521911621094,
      "loss": 0.6922,
      "rewards/accuracies": 0.4166666567325592,
      "rewards/chosen": -0.010081589221954346,
      "rewards/margins": -0.0022414023987948895,
      "rewards/rejected": -0.007840188220143318,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.990353313429303e-07,
      "logits/chosen": -2.6826164722442627,
      "logits/rejected": -2.676023006439209,
      "logps/chosen": -270.6432189941406,
      "logps/rejected": -290.40869140625,
      "loss": 0.6794,
      "rewards/accuracies": 0.5062500238418579,
      "rewards/chosen": -0.2684400975704193,
      "rewards/margins": 0.04983901605010033,
      "rewards/rejected": -0.31827908754348755,
      "step": 20
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.882681251368548e-07,
      "logits/chosen": -2.6072068214416504,
      "logits/rejected": -2.588265895843506,
      "logps/chosen": -228.98974609375,
      "logps/rejected": -268.8215637207031,
      "loss": 0.6895,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.09740053117275238,
      "rewards/margins": 0.05857750028371811,
      "rewards/rejected": -0.1559780389070511,
      "step": 30
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.6604720940421207e-07,
      "logits/chosen": -2.5079424381256104,
      "logits/rejected": -2.485220432281494,
      "logps/chosen": -233.1949462890625,
      "logps/rejected": -272.9608459472656,
      "loss": 0.6737,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.1403357833623886,
      "rewards/margins": 0.14330314099788666,
      "rewards/rejected": -0.28363892436027527,
      "step": 40
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.3344075855595097e-07,
      "logits/chosen": -2.430572986602783,
      "logits/rejected": -2.4246697425842285,
      "logps/chosen": -258.1377868652344,
      "logps/rejected": -303.17877197265625,
      "loss": 0.6657,
      "rewards/accuracies": 0.5687500238418579,
      "rewards/chosen": -0.4008946418762207,
      "rewards/margins": 0.11444135755300522,
      "rewards/rejected": -0.5153360366821289,
      "step": 50
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.920161866827889e-07,
      "logits/chosen": -2.3226020336151123,
      "logits/rejected": -2.313552141189575,
      "logps/chosen": -289.2204284667969,
      "logps/rejected": -296.24609375,
      "loss": 0.6682,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": -0.5691416263580322,
      "rewards/margins": 0.10885188728570938,
      "rewards/rejected": -0.6779934167861938,
      "step": 60
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4376480090239047e-07,
      "logits/chosen": -2.2763919830322266,
      "logits/rejected": -2.2820637226104736,
      "logps/chosen": -253.41159057617188,
      "logps/rejected": -296.48333740234375,
      "loss": 0.667,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.4127040505409241,
      "rewards/margins": 0.14596061408519745,
      "rewards/rejected": -0.5586646199226379,
      "step": 70
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.910060778827554e-07,
      "logits/chosen": -2.2447080612182617,
      "logits/rejected": -2.2723259925842285,
      "logps/chosen": -234.634521484375,
      "logps/rejected": -309.20513916015625,
      "loss": 0.653,
      "rewards/accuracies": 0.581250011920929,
      "rewards/chosen": -0.40885549783706665,
      "rewards/margins": 0.18530671298503876,
      "rewards/rejected": -0.5941622257232666,
      "step": 80
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.3627616503391812e-07,
      "logits/chosen": -2.330038070678711,
      "logits/rejected": -2.3056416511535645,
      "logps/chosen": -243.3378448486328,
      "logps/rejected": -270.39666748046875,
      "loss": 0.6571,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.4189552664756775,
      "rewards/margins": 0.1503242701292038,
      "rewards/rejected": -0.5692795515060425,
      "step": 90
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.8220596619089573e-07,
      "logits/chosen": -2.235563278198242,
      "logits/rejected": -2.2629618644714355,
      "logps/chosen": -269.2986145019531,
      "logps/rejected": -299.86138916015625,
      "loss": 0.6631,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.46052223443984985,
      "rewards/margins": 0.18169686198234558,
      "rewards/rejected": -0.6422190070152283,
      "step": 100
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.3139467229135998e-07,
      "logits/chosen": -2.3022847175598145,
      "logits/rejected": -2.3016762733459473,
      "logps/chosen": -234.77597045898438,
      "logps/rejected": -316.5335998535156,
      "loss": 0.6379,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": -0.4068765640258789,
      "rewards/margins": 0.2481352835893631,
      "rewards/rejected": -0.6550118923187256,
      "step": 110
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.628481651367875e-08,
      "logits/chosen": -2.2812705039978027,
      "logits/rejected": -2.2640042304992676,
      "logps/chosen": -265.8632507324219,
      "logps/rejected": -339.9316101074219,
      "loss": 0.6563,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": -0.49150800704956055,
      "rewards/margins": 0.2535227835178375,
      "rewards/rejected": -0.7450307607650757,
      "step": 120
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.904486005914027e-08,
      "logits/chosen": -2.3139283657073975,
      "logits/rejected": -2.327219247817993,
      "logps/chosen": -259.82452392578125,
      "logps/rejected": -339.85772705078125,
      "loss": 0.6411,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.4681639075279236,
      "rewards/margins": 0.2101978361606598,
      "rewards/rejected": -0.6783617734909058,
      "step": 130
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.1464952759020856e-08,
      "logits/chosen": -2.3203117847442627,
      "logits/rejected": -2.3070220947265625,
      "logps/chosen": -270.1856994628906,
      "logps/rejected": -336.08758544921875,
      "loss": 0.6446,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.48976272344589233,
      "rewards/margins": 0.19642502069473267,
      "rewards/rejected": -0.686187744140625,
      "step": 140
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.8708793644441086e-09,
      "logits/chosen": -2.2997617721557617,
      "logits/rejected": -2.2623441219329834,
      "logps/chosen": -261.05316162109375,
      "logps/rejected": -276.8531188964844,
      "loss": 0.6467,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.49157172441482544,
      "rewards/margins": 0.11737655103206635,
      "rewards/rejected": -0.6089483499526978,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 159,
      "total_flos": 0.0,
      "train_loss": 0.6627466093819097,
      "train_runtime": 2639.006,
      "train_samples_per_second": 7.722,
      "train_steps_per_second": 0.06
    }
  ],
  "logging_steps": 10,
  "max_steps": 159,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}