{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 313,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003194888178913738,
      "grad_norm": 63.76829168210548,
      "learning_rate": 5.3191489361702125e-09,
      "logits/chosen": -3.703125,
      "logits/rejected": -3.859375,
      "logps/chosen": -139.0,
      "logps/rejected": -80.0,
      "loss": 0.6914,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.03194888178913738,
      "grad_norm": 92.94577098387109,
      "learning_rate": 5.3191489361702123e-08,
      "logits/chosen": -3.625,
      "logits/rejected": -3.53125,
      "logps/chosen": -99.5,
      "logps/rejected": -96.0,
      "loss": 0.6939,
      "rewards/accuracies": 0.1527777761220932,
      "rewards/chosen": -0.0010528564453125,
      "rewards/margins": -0.001739501953125,
      "rewards/rejected": 0.0006866455078125,
      "step": 10
    },
    {
      "epoch": 0.06389776357827476,
      "grad_norm": 70.79325392606067,
      "learning_rate": 1.0638297872340425e-07,
      "logits/chosen": -3.375,
      "logits/rejected": -3.5625,
      "logps/chosen": -96.0,
      "logps/rejected": -105.0,
      "loss": 0.6872,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": -0.0242919921875,
      "rewards/margins": 0.017333984375,
      "rewards/rejected": -0.04150390625,
      "step": 20
    },
    {
      "epoch": 0.09584664536741214,
      "grad_norm": 63.190582339280105,
      "learning_rate": 1.5957446808510638e-07,
      "logits/chosen": -3.5625,
      "logits/rejected": -3.453125,
      "logps/chosen": -99.5,
      "logps/rejected": -108.0,
      "loss": 0.671,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.08935546875,
      "rewards/margins": 0.0361328125,
      "rewards/rejected": -0.125,
      "step": 30
    },
    {
      "epoch": 0.12779552715654952,
      "grad_norm": 70.46556001874696,
      "learning_rate": 2.127659574468085e-07,
      "logits/chosen": -3.578125,
      "logits/rejected": -3.59375,
      "logps/chosen": -110.0,
      "logps/rejected": -95.5,
      "loss": 0.6578,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.1669921875,
      "rewards/margins": 0.0164794921875,
      "rewards/rejected": -0.1826171875,
      "step": 40
    },
    {
      "epoch": 0.1597444089456869,
      "grad_norm": 53.92661468897121,
      "learning_rate": 2.659574468085106e-07,
      "logits/chosen": -3.6875,
      "logits/rejected": -3.640625,
      "logps/chosen": -95.0,
      "logps/rejected": -106.5,
      "loss": 0.6153,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.251953125,
      "rewards/margins": 0.1376953125,
      "rewards/rejected": -0.388671875,
      "step": 50
    },
    {
      "epoch": 0.19169329073482427,
      "grad_norm": 61.31819398041802,
      "learning_rate": 3.1914893617021275e-07,
      "logits/chosen": -3.453125,
      "logits/rejected": -3.328125,
      "logps/chosen": -101.0,
      "logps/rejected": -107.0,
      "loss": 0.7335,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.283203125,
      "rewards/margins": 0.478515625,
      "rewards/rejected": -0.76171875,
      "step": 60
    },
    {
      "epoch": 0.22364217252396165,
      "grad_norm": 55.44332583758203,
      "learning_rate": 3.7234042553191484e-07,
      "logits/chosen": -3.421875,
      "logits/rejected": -3.546875,
      "logps/chosen": -103.0,
      "logps/rejected": -112.0,
      "loss": 0.6863,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.1220703125,
      "rewards/margins": 0.6328125,
      "rewards/rejected": -0.75390625,
      "step": 70
    },
    {
      "epoch": 0.25559105431309903,
      "grad_norm": 51.149001898192076,
      "learning_rate": 4.25531914893617e-07,
      "logits/chosen": -3.546875,
      "logits/rejected": -3.53125,
      "logps/chosen": -110.5,
      "logps/rejected": -110.5,
      "loss": 0.5805,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.0986328125,
      "rewards/margins": 0.62890625,
      "rewards/rejected": -0.7265625,
      "step": 80
    },
    {
      "epoch": 0.28753993610223644,
      "grad_norm": 55.06999142181094,
      "learning_rate": 4.787234042553192e-07,
      "logits/chosen": -3.671875,
      "logits/rejected": -3.59375,
      "logps/chosen": -107.0,
      "logps/rejected": -116.0,
      "loss": 0.5765,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.1494140625,
      "rewards/margins": 0.7734375,
      "rewards/rejected": -0.921875,
      "step": 90
    },
    {
      "epoch": 0.3194888178913738,
      "grad_norm": 33.54407800330637,
      "learning_rate": 4.964497041420119e-07,
      "logits/chosen": -3.546875,
      "logits/rejected": -3.53125,
      "logps/chosen": -105.0,
      "logps/rejected": -112.5,
      "loss": 0.5228,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": -0.1162109375,
      "rewards/margins": 0.61328125,
      "rewards/rejected": -0.7265625,
      "step": 100
    },
    {
      "epoch": 0.3514376996805112,
      "grad_norm": 71.5169022650209,
      "learning_rate": 4.905325443786982e-07,
      "logits/chosen": -3.5625,
      "logits/rejected": -3.515625,
      "logps/chosen": -101.5,
      "logps/rejected": -123.0,
      "loss": 0.5408,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.1337890625,
      "rewards/margins": 0.8515625,
      "rewards/rejected": -0.984375,
      "step": 110
    },
    {
      "epoch": 0.38338658146964855,
      "grad_norm": 65.33074960619584,
      "learning_rate": 4.846153846153846e-07,
      "logits/chosen": -3.546875,
      "logits/rejected": -3.5625,
      "logps/chosen": -100.5,
      "logps/rejected": -139.0,
      "loss": 0.5381,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.021484375,
      "rewards/margins": 1.671875,
      "rewards/rejected": -1.6484375,
      "step": 120
    },
    {
      "epoch": 0.41533546325878595,
      "grad_norm": 49.04091358646178,
      "learning_rate": 4.78698224852071e-07,
      "logits/chosen": -3.515625,
      "logits/rejected": -3.5,
      "logps/chosen": -95.0,
      "logps/rejected": -116.5,
      "loss": 0.5387,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.10791015625,
      "rewards/margins": 0.99609375,
      "rewards/rejected": -1.1015625,
      "step": 130
    },
    {
      "epoch": 0.4472843450479233,
      "grad_norm": 49.48725125304273,
      "learning_rate": 4.727810650887574e-07,
      "logits/chosen": -3.671875,
      "logits/rejected": -3.515625,
      "logps/chosen": -101.0,
      "logps/rejected": -121.0,
      "loss": 0.4801,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.0260009765625,
      "rewards/margins": 1.609375,
      "rewards/rejected": -1.6328125,
      "step": 140
    },
    {
      "epoch": 0.4792332268370607,
      "grad_norm": 88.63757443724184,
      "learning_rate": 4.668639053254438e-07,
      "logits/chosen": -3.671875,
      "logits/rejected": -3.53125,
      "logps/chosen": -104.5,
      "logps/rejected": -107.5,
      "loss": 0.5525,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.0185546875,
      "rewards/margins": 0.88671875,
      "rewards/rejected": -0.90625,
      "step": 150
    },
    {
      "epoch": 0.5111821086261981,
      "grad_norm": 65.41898185778713,
      "learning_rate": 4.6094674556213014e-07,
      "logits/chosen": -3.53125,
      "logits/rejected": -3.375,
      "logps/chosen": -102.5,
      "logps/rejected": -110.0,
      "loss": 0.4824,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.20703125,
      "rewards/margins": 1.0390625,
      "rewards/rejected": -1.2421875,
      "step": 160
    },
    {
      "epoch": 0.5431309904153354,
      "grad_norm": 64.3627134555641,
      "learning_rate": 4.5502958579881655e-07,
      "logits/chosen": -3.671875,
      "logits/rejected": -3.5625,
      "logps/chosen": -95.5,
      "logps/rejected": -114.0,
      "loss": 0.4837,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.04736328125,
      "rewards/margins": 1.3984375,
      "rewards/rejected": -1.3515625,
      "step": 170
    },
    {
      "epoch": 0.5750798722044729,
      "grad_norm": 50.45380597678468,
      "learning_rate": 4.491124260355029e-07,
      "logits/chosen": -3.53125,
      "logits/rejected": -3.453125,
      "logps/chosen": -100.5,
      "logps/rejected": -114.5,
      "loss": 0.4732,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -0.031005859375,
      "rewards/margins": 1.1328125,
      "rewards/rejected": -1.1640625,
      "step": 180
    },
    {
      "epoch": 0.6070287539936102,
      "grad_norm": 47.054283179898476,
      "learning_rate": 4.4319526627218936e-07,
      "logits/chosen": -3.59375,
      "logits/rejected": -3.53125,
      "logps/chosen": -103.5,
      "logps/rejected": -112.0,
      "loss": 0.4777,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.08154296875,
      "rewards/margins": 1.203125,
      "rewards/rejected": -1.125,
      "step": 190
    },
    {
      "epoch": 0.6389776357827476,
      "grad_norm": 60.50447069634289,
      "learning_rate": 4.372781065088757e-07,
      "logits/chosen": -3.625,
      "logits/rejected": -3.46875,
      "logps/chosen": -100.0,
      "logps/rejected": -112.0,
      "loss": 0.4606,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.162109375,
      "rewards/margins": 1.375,
      "rewards/rejected": -1.2109375,
      "step": 200
    },
    {
      "epoch": 0.670926517571885,
      "grad_norm": 42.756358795311726,
      "learning_rate": 4.313609467455621e-07,
      "logits/chosen": -3.59375,
      "logits/rejected": -3.46875,
      "logps/chosen": -92.5,
      "logps/rejected": -119.0,
      "loss": 0.3956,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.058837890625,
      "rewards/margins": 1.7578125,
      "rewards/rejected": -1.8203125,
      "step": 210
    },
    {
      "epoch": 0.7028753993610224,
      "grad_norm": 37.12579027519649,
      "learning_rate": 4.2544378698224847e-07,
      "logits/chosen": -3.453125,
      "logits/rejected": -3.484375,
      "logps/chosen": -101.5,
      "logps/rejected": -112.0,
      "loss": 0.4621,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.06884765625,
      "rewards/margins": 1.3046875,
      "rewards/rejected": -1.234375,
      "step": 220
    },
    {
      "epoch": 0.7348242811501597,
      "grad_norm": 34.37003470756731,
      "learning_rate": 4.195266272189349e-07,
      "logits/chosen": -3.40625,
      "logits/rejected": -3.421875,
      "logps/chosen": -102.0,
      "logps/rejected": -119.0,
      "loss": 0.4478,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.06640625,
      "rewards/margins": 1.59375,
      "rewards/rejected": -1.6640625,
      "step": 230
    },
    {
      "epoch": 0.7667731629392971,
      "grad_norm": 39.975701583088714,
      "learning_rate": 4.1360946745562133e-07,
      "logits/chosen": -3.484375,
      "logits/rejected": -3.4375,
      "logps/chosen": -95.5,
      "logps/rejected": -123.0,
      "loss": 0.4043,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.0322265625,
      "rewards/margins": 1.5546875,
      "rewards/rejected": -1.5234375,
      "step": 240
    },
    {
      "epoch": 0.7987220447284346,
      "grad_norm": 40.31896174589532,
      "learning_rate": 4.076923076923077e-07,
      "logits/chosen": -3.53125,
      "logits/rejected": -3.484375,
      "logps/chosen": -107.0,
      "logps/rejected": -122.0,
      "loss": 0.4742,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 0.54296875,
      "rewards/margins": 2.46875,
      "rewards/rejected": -1.921875,
      "step": 250
    },
    {
      "epoch": 0.8306709265175719,
      "grad_norm": 32.53695466017645,
      "learning_rate": 4.017751479289941e-07,
      "logits/chosen": -3.578125,
      "logits/rejected": -3.5,
      "logps/chosen": -81.0,
      "logps/rejected": -135.0,
      "loss": 0.3528,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 0.044921875,
      "rewards/margins": 2.765625,
      "rewards/rejected": -2.71875,
      "step": 260
    },
    {
      "epoch": 0.8626198083067093,
      "grad_norm": 35.58565159605434,
      "learning_rate": 3.9585798816568044e-07,
      "logits/chosen": -3.625,
      "logits/rejected": -3.46875,
      "logps/chosen": -106.5,
      "logps/rejected": -107.5,
      "loss": 0.3774,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 0.318359375,
      "rewards/margins": 1.8984375,
      "rewards/rejected": -1.578125,
      "step": 270
    },
    {
      "epoch": 0.8945686900958466,
      "grad_norm": 50.853618748566205,
      "learning_rate": 3.8994082840236685e-07,
      "logits/chosen": -3.515625,
      "logits/rejected": -3.484375,
      "logps/chosen": -99.5,
      "logps/rejected": -119.0,
      "loss": 0.4586,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.26953125,
      "rewards/margins": 1.671875,
      "rewards/rejected": -1.40625,
      "step": 280
    },
    {
      "epoch": 0.9265175718849841,
      "grad_norm": 49.95783582965489,
      "learning_rate": 3.840236686390532e-07,
      "logits/chosen": -3.640625,
      "logits/rejected": -3.453125,
      "logps/chosen": -100.5,
      "logps/rejected": -113.0,
      "loss": 0.4365,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.37109375,
      "rewards/margins": 1.8671875,
      "rewards/rejected": -1.5,
      "step": 290
    },
    {
      "epoch": 0.9584664536741214,
      "grad_norm": 26.77848071496543,
      "learning_rate": 3.7810650887573966e-07,
      "logits/chosen": -3.5625,
      "logits/rejected": -3.59375,
      "logps/chosen": -99.5,
      "logps/rejected": -135.0,
      "loss": 0.4772,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": 0.25390625,
      "rewards/margins": 2.046875,
      "rewards/rejected": -1.7890625,
      "step": 300
    },
    {
      "epoch": 0.9904153354632588,
      "grad_norm": 44.271261713875845,
      "learning_rate": 3.72189349112426e-07,
      "logits/chosen": -3.5625,
      "logits/rejected": -3.484375,
      "logps/chosen": -108.5,
      "logps/rejected": -122.5,
      "loss": 0.3872,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.0625,
      "rewards/margins": 1.625,
      "rewards/rejected": -1.5703125,
      "step": 310
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -3.5625,
      "eval_logits/rejected": -3.5,
      "eval_logps/chosen": -98.5,
      "eval_logps/rejected": -125.0,
      "eval_loss": 0.5349218845367432,
      "eval_rewards/accuracies": 0.7857142686843872,
      "eval_rewards/chosen": -0.12060546875,
      "eval_rewards/margins": 1.203125,
      "eval_rewards/rejected": -1.3203125,
      "eval_runtime": 12.3829,
      "eval_samples_per_second": 16.151,
      "eval_steps_per_second": 0.565,
      "step": 313
    }
  ],
  "logging_steps": 10,
  "max_steps": 939,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}