{
  "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": -2.5097975730895996,
      "logits/rejected": -2.419074058532715,
      "logps/chosen": -193.5599365234375,
      "logps/pi_response": -171.87442016601562,
      "logps/ref_response": -171.87442016601562,
      "logps/rejected": -224.4684295654297,
      "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": -2.5170841217041016,
      "logits/rejected": -2.499955415725708,
      "logps/chosen": -268.5122375488281,
      "logps/pi_response": -200.62173461914062,
      "logps/ref_response": -200.7644805908203,
      "logps/rejected": -275.15576171875,
      "loss": 0.692,
      "rewards/accuracies": 0.4513888955116272,
      "rewards/chosen": 0.0004006973758805543,
      "rewards/margins": 0.0025936374440789223,
      "rewards/rejected": -0.0021929400973021984,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.990353313429303e-07,
      "logits/chosen": -2.5279006958007812,
      "logits/rejected": -2.48834228515625,
      "logps/chosen": -280.61029052734375,
      "logps/pi_response": -218.712158203125,
      "logps/ref_response": -220.61703491210938,
      "logps/rejected": -302.9790344238281,
      "loss": 0.6799,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.0002802034723572433,
      "rewards/margins": 0.06487339735031128,
      "rewards/rejected": -0.06459320336580276,
      "step": 20
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.882681251368548e-07,
      "logits/chosen": -2.3786978721618652,
      "logits/rejected": -2.309803009033203,
      "logps/chosen": -268.9042053222656,
      "logps/pi_response": -195.0692901611328,
      "logps/ref_response": -189.48886108398438,
      "logps/rejected": -271.2804260253906,
      "loss": 0.6702,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": -0.04177286848425865,
      "rewards/margins": 0.0969323068857193,
      "rewards/rejected": -0.13870516419410706,
      "step": 30
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.6604720940421207e-07,
      "logits/chosen": -2.210906505584717,
      "logits/rejected": -2.181436061859131,
      "logps/chosen": -283.42681884765625,
      "logps/pi_response": -215.8210906982422,
      "logps/ref_response": -189.73727416992188,
      "logps/rejected": -321.3996276855469,
      "loss": 0.6576,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.26275861263275146,
      "rewards/margins": 0.13055923581123352,
      "rewards/rejected": -0.3933178782463074,
      "step": 40
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.3344075855595097e-07,
      "logits/chosen": -2.1676063537597656,
      "logits/rejected": -2.122574806213379,
      "logps/chosen": -321.0862121582031,
      "logps/pi_response": -257.51531982421875,
      "logps/ref_response": -199.2998504638672,
      "logps/rejected": -330.80474853515625,
      "loss": 0.6442,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -0.47685688734054565,
      "rewards/margins": 0.16487360000610352,
      "rewards/rejected": -0.6417304873466492,
      "step": 50
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.920161866827889e-07,
      "logits/chosen": -2.14129900932312,
      "logits/rejected": -2.106346607208252,
      "logps/chosen": -291.825439453125,
      "logps/pi_response": -247.85244750976562,
      "logps/ref_response": -182.53189086914062,
      "logps/rejected": -323.88409423828125,
      "loss": 0.6403,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.41668352484703064,
      "rewards/margins": 0.3306669592857361,
      "rewards/rejected": -0.7473504543304443,
      "step": 60
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4376480090239047e-07,
      "logits/chosen": -2.1696386337280273,
      "logits/rejected": -2.1270456314086914,
      "logps/chosen": -331.155029296875,
      "logps/pi_response": -269.52734375,
      "logps/ref_response": -191.21128845214844,
      "logps/rejected": -335.54974365234375,
      "loss": 0.6304,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": -0.6724362373352051,
      "rewards/margins": 0.23178334534168243,
      "rewards/rejected": -0.9042196273803711,
      "step": 70
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.910060778827554e-07,
      "logits/chosen": -2.203618049621582,
      "logits/rejected": -2.166483163833618,
      "logps/chosen": -287.53997802734375,
      "logps/pi_response": -254.5912322998047,
      "logps/ref_response": -190.11843872070312,
      "logps/rejected": -332.36590576171875,
      "loss": 0.6061,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.514179527759552,
      "rewards/margins": 0.23715217411518097,
      "rewards/rejected": -0.7513316869735718,
      "step": 80
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.3627616503391812e-07,
      "logits/chosen": -2.278022527694702,
      "logits/rejected": -2.2521653175354004,
      "logps/chosen": -271.2624206542969,
      "logps/pi_response": -235.70266723632812,
      "logps/ref_response": -180.7279815673828,
      "logps/rejected": -326.7520751953125,
      "loss": 0.6169,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.3808743357658386,
      "rewards/margins": 0.2905920445919037,
      "rewards/rejected": -0.6714664101600647,
      "step": 90
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.8220596619089573e-07,
      "logits/chosen": -2.2085721492767334,
      "logits/rejected": -2.194762706756592,
      "logps/chosen": -326.04888916015625,
      "logps/pi_response": -270.80810546875,
      "logps/ref_response": -195.6220703125,
      "logps/rejected": -389.5771179199219,
      "loss": 0.6167,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.691897988319397,
      "rewards/margins": 0.3410559296607971,
      "rewards/rejected": -1.0329539775848389,
      "step": 100
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.3139467229135998e-07,
      "logits/chosen": -2.2901813983917236,
      "logits/rejected": -2.2825167179107666,
      "logps/chosen": -300.0601806640625,
      "logps/pi_response": -242.42221069335938,
      "logps/ref_response": -179.70770263671875,
      "logps/rejected": -329.3973083496094,
      "loss": 0.6131,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": -0.5147222280502319,
      "rewards/margins": 0.23834428191184998,
      "rewards/rejected": -0.7530665993690491,
      "step": 110
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.628481651367875e-08,
      "logits/chosen": -2.2861225605010986,
      "logits/rejected": -2.2135860919952393,
      "logps/chosen": -316.3228454589844,
      "logps/pi_response": -282.2044982910156,
      "logps/ref_response": -202.98269653320312,
      "logps/rejected": -386.8232421875,
      "loss": 0.5922,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.6165128946304321,
      "rewards/margins": 0.42569026350975037,
      "rewards/rejected": -1.0422031879425049,
      "step": 120
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.904486005914027e-08,
      "logits/chosen": -2.304694890975952,
      "logits/rejected": -2.2709462642669678,
      "logps/chosen": -332.1279602050781,
      "logps/pi_response": -276.42706298828125,
      "logps/ref_response": -195.01205444335938,
      "logps/rejected": -367.41204833984375,
      "loss": 0.6018,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.6682323217391968,
      "rewards/margins": 0.37029170989990234,
      "rewards/rejected": -1.0385239124298096,
      "step": 130
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.1464952759020856e-08,
      "logits/chosen": -2.3100249767303467,
      "logits/rejected": -2.246779680252075,
      "logps/chosen": -328.9765930175781,
      "logps/pi_response": -279.7056579589844,
      "logps/ref_response": -195.75735473632812,
      "logps/rejected": -392.63818359375,
      "loss": 0.584,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": -0.7241796255111694,
      "rewards/margins": 0.34663844108581543,
      "rewards/rejected": -1.0708180665969849,
      "step": 140
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.8708793644441086e-09,
      "logits/chosen": -2.2483038902282715,
      "logits/rejected": -2.229506254196167,
      "logps/chosen": -307.9349060058594,
      "logps/pi_response": -257.32733154296875,
      "logps/ref_response": -179.74029541015625,
      "logps/rejected": -353.04681396484375,
      "loss": 0.5882,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.6363255381584167,
      "rewards/margins": 0.36607229709625244,
      "rewards/rejected": -1.002397894859314,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 159,
      "total_flos": 0.0,
      "train_loss": 0.6280672625175812,
      "train_runtime": 4264.6818,
      "train_samples_per_second": 4.778,
      "train_steps_per_second": 0.037
    }
  ],
  "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
}