{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9999344906649198,
  "eval_steps": 500,
  "global_step": 3816,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.050049132001310184,
      "grad_norm": 4.843938827514648,
      "learning_rate": 5e-07,
      "logits/chosen": -0.5328307747840881,
      "logits/rejected": -0.46889373660087585,
      "logps/chosen": -74.61598205566406,
      "logps/rejected": -13.190138816833496,
      "loss": 0.6891,
      "rewards/accuracies": 0.5595549941062927,
      "rewards/chosen": 0.007412092760205269,
      "rewards/margins": 0.008624909445643425,
      "rewards/rejected": -0.001212816801853478,
      "step": 191
    },
    {
      "epoch": 0.10009826400262037,
      "grad_norm": 3.657888174057007,
      "learning_rate": 1e-06,
      "logits/chosen": -0.5486236810684204,
      "logits/rejected": -0.48813965916633606,
      "logps/chosen": -70.09341430664062,
      "logps/rejected": -13.666769981384277,
      "loss": 0.5809,
      "rewards/accuracies": 0.8239529132843018,
      "rewards/chosen": 0.24100865423679352,
      "rewards/margins": 0.2950609624385834,
      "rewards/rejected": -0.054052311927080154,
      "step": 382
    },
    {
      "epoch": 0.15014739600393057,
      "grad_norm": 1.726294755935669,
      "learning_rate": 9.443797320908561e-07,
      "logits/chosen": -0.5495401620864868,
      "logits/rejected": -0.4997619390487671,
      "logps/chosen": -61.40409469604492,
      "logps/rejected": -19.322750091552734,
      "loss": 0.2712,
      "rewards/accuracies": 0.9463350772857666,
      "rewards/chosen": 1.1437506675720215,
      "rewards/margins": 1.791763424873352,
      "rewards/rejected": -0.648012638092041,
      "step": 573
    },
    {
      "epoch": 0.20019652800524074,
      "grad_norm": 9.303996086120605,
      "learning_rate": 8.887594641817122e-07,
      "logits/chosen": -0.5349320769309998,
      "logits/rejected": -0.4746370315551758,
      "logps/chosen": -61.47517776489258,
      "logps/rejected": -35.966888427734375,
      "loss": 0.0598,
      "rewards/accuracies": 0.9869109988212585,
      "rewards/chosen": 1.5467382669448853,
      "rewards/margins": 3.8473398685455322,
      "rewards/rejected": -2.3006019592285156,
      "step": 764
    },
    {
      "epoch": 0.2502456600065509,
      "grad_norm": 0.23748359084129333,
      "learning_rate": 8.331391962725683e-07,
      "logits/chosen": -0.4757286310195923,
      "logits/rejected": -0.3977571725845337,
      "logps/chosen": -57.084896087646484,
      "logps/rejected": -45.92123794555664,
      "loss": 0.0267,
      "rewards/accuracies": 0.9901832342147827,
      "rewards/chosen": 1.6708416938781738,
      "rewards/margins": 5.011633396148682,
      "rewards/rejected": -3.340791702270508,
      "step": 955
    },
    {
      "epoch": 0.30029479200786113,
      "grad_norm": 0.43927711248397827,
      "learning_rate": 7.775189283634246e-07,
      "logits/chosen": -0.4342685043811798,
      "logits/rejected": -0.33603060245513916,
      "logps/chosen": -57.238773345947266,
      "logps/rejected": -50.37546920776367,
      "loss": 0.0331,
      "rewards/accuracies": 0.9869109988212585,
      "rewards/chosen": 1.7030671834945679,
      "rewards/margins": 5.485808372497559,
      "rewards/rejected": -3.7827417850494385,
      "step": 1146
    },
    {
      "epoch": 0.3503439240091713,
      "grad_norm": 0.08208785951137543,
      "learning_rate": 7.218986604542807e-07,
      "logits/chosen": -0.4299260973930359,
      "logits/rejected": -0.32499420642852783,
      "logps/chosen": -56.95598220825195,
      "logps/rejected": -55.50334548950195,
      "loss": 0.0327,
      "rewards/accuracies": 0.985602080821991,
      "rewards/chosen": 1.6657171249389648,
      "rewards/margins": 5.907327651977539,
      "rewards/rejected": -4.241610527038574,
      "step": 1337
    },
    {
      "epoch": 0.4003930560104815,
      "grad_norm": 0.18074464797973633,
      "learning_rate": 6.662783925451368e-07,
      "logits/chosen": -0.389539510011673,
      "logits/rejected": -0.26959964632987976,
      "logps/chosen": -56.53901672363281,
      "logps/rejected": -59.943992614746094,
      "loss": 0.0247,
      "rewards/accuracies": 0.9914921522140503,
      "rewards/chosen": 1.684263825416565,
      "rewards/margins": 6.446012496948242,
      "rewards/rejected": -4.761748790740967,
      "step": 1528
    },
    {
      "epoch": 0.4504421880117917,
      "grad_norm": 0.07140897214412689,
      "learning_rate": 6.10658124635993e-07,
      "logits/chosen": -0.38458481431007385,
      "logits/rejected": -0.25353050231933594,
      "logps/chosen": -56.987735748291016,
      "logps/rejected": -65.21515655517578,
      "loss": 0.0252,
      "rewards/accuracies": 0.9882199168205261,
      "rewards/chosen": 1.7450363636016846,
      "rewards/margins": 6.95145845413208,
      "rewards/rejected": -5.206421375274658,
      "step": 1719
    },
    {
      "epoch": 0.5004913200131018,
      "grad_norm": 0.13229215145111084,
      "learning_rate": 5.550378567268491e-07,
      "logits/chosen": -0.3671688139438629,
      "logits/rejected": -0.23125909268856049,
      "logps/chosen": -55.54978942871094,
      "logps/rejected": -69.9174575805664,
      "loss": 0.0311,
      "rewards/accuracies": 0.9908376932144165,
      "rewards/chosen": 1.6499972343444824,
      "rewards/margins": 7.394839286804199,
      "rewards/rejected": -5.744842529296875,
      "step": 1910
    },
    {
      "epoch": 0.550540452014412,
      "grad_norm": 0.03863115608692169,
      "learning_rate": 4.994175888177053e-07,
      "logits/chosen": -0.34299236536026,
      "logits/rejected": -0.21112458407878876,
      "logps/chosen": -56.28550338745117,
      "logps/rejected": -74.3562240600586,
      "loss": 0.0203,
      "rewards/accuracies": 0.9908376932144165,
      "rewards/chosen": 1.6116416454315186,
      "rewards/margins": 7.8128886222839355,
      "rewards/rejected": -6.201247215270996,
      "step": 2101
    },
    {
      "epoch": 0.6005895840157223,
      "grad_norm": 0.16626648604869843,
      "learning_rate": 4.4379732090856143e-07,
      "logits/chosen": -0.317820280790329,
      "logits/rejected": -0.16657008230686188,
      "logps/chosen": -58.859397888183594,
      "logps/rejected": -77.39883422851562,
      "loss": 0.0317,
      "rewards/accuracies": 0.9875654578208923,
      "rewards/chosen": 1.6255024671554565,
      "rewards/margins": 8.082979202270508,
      "rewards/rejected": -6.4574761390686035,
      "step": 2292
    },
    {
      "epoch": 0.6506387160170324,
      "grad_norm": 0.01698416657745838,
      "learning_rate": 3.8817705299941754e-07,
      "logits/chosen": -0.3418547213077545,
      "logits/rejected": -0.19180215895175934,
      "logps/chosen": -57.59861373901367,
      "logps/rejected": -80.31951141357422,
      "loss": 0.0122,
      "rewards/accuracies": 0.9934555292129517,
      "rewards/chosen": 1.641247034072876,
      "rewards/margins": 8.408109664916992,
      "rewards/rejected": -6.766862392425537,
      "step": 2483
    },
    {
      "epoch": 0.7006878480183426,
      "grad_norm": 0.03158177435398102,
      "learning_rate": 3.3255678509027375e-07,
      "logits/chosen": -0.32275617122650146,
      "logits/rejected": -0.18107445538043976,
      "logps/chosen": -59.604644775390625,
      "logps/rejected": -81.69689178466797,
      "loss": 0.0214,
      "rewards/accuracies": 0.9908376932144165,
      "rewards/chosen": 1.6721757650375366,
      "rewards/margins": 8.503122329711914,
      "rewards/rejected": -6.83094596862793,
      "step": 2674
    },
    {
      "epoch": 0.7507369800196528,
      "grad_norm": 0.08755598962306976,
      "learning_rate": 2.7693651718112985e-07,
      "logits/chosen": -0.32179561257362366,
      "logits/rejected": -0.17599359154701233,
      "logps/chosen": -57.8769416809082,
      "logps/rejected": -83.73433685302734,
      "loss": 0.0192,
      "rewards/accuracies": 0.9908376932144165,
      "rewards/chosen": 1.6639045476913452,
      "rewards/margins": 8.702630043029785,
      "rewards/rejected": -7.038724899291992,
      "step": 2865
    },
    {
      "epoch": 0.800786112020963,
      "grad_norm": 0.015691732987761497,
      "learning_rate": 2.21316249271986e-07,
      "logits/chosen": -0.3434266149997711,
      "logits/rejected": -0.19399824738502502,
      "logps/chosen": -56.69347381591797,
      "logps/rejected": -85.22378540039062,
      "loss": 0.0277,
      "rewards/accuracies": 0.9914921522140503,
      "rewards/chosen": 1.6015362739562988,
      "rewards/margins": 8.894182205200195,
      "rewards/rejected": -7.2926459312438965,
      "step": 3056
    },
    {
      "epoch": 0.8508352440222732,
      "grad_norm": 0.010581793263554573,
      "learning_rate": 1.6569598136284217e-07,
      "logits/chosen": -0.33340033888816833,
      "logits/rejected": -0.18355616927146912,
      "logps/chosen": -58.657325744628906,
      "logps/rejected": -85.88097381591797,
      "loss": 0.0251,
      "rewards/accuracies": 0.9921466112136841,
      "rewards/chosen": 1.615267276763916,
      "rewards/margins": 8.870849609375,
      "rewards/rejected": -7.255581855773926,
      "step": 3247
    },
    {
      "epoch": 0.9008843760235834,
      "grad_norm": 0.02463706023991108,
      "learning_rate": 1.100757134536983e-07,
      "logits/chosen": -0.3216906785964966,
      "logits/rejected": -0.17967411875724792,
      "logps/chosen": -55.944915771484375,
      "logps/rejected": -87.04055786132812,
      "loss": 0.0236,
      "rewards/accuracies": 0.9882199168205261,
      "rewards/chosen": 1.5252350568771362,
      "rewards/margins": 8.986153602600098,
      "rewards/rejected": -7.460918426513672,
      "step": 3438
    },
    {
      "epoch": 0.9509335080248935,
      "grad_norm": 0.1806812733411789,
      "learning_rate": 5.445544554455445e-08,
      "logits/chosen": -0.3310312032699585,
      "logits/rejected": -0.17037709057331085,
      "logps/chosen": -56.92354202270508,
      "logps/rejected": -86.47986602783203,
      "loss": 0.0231,
      "rewards/accuracies": 0.9928010702133179,
      "rewards/chosen": 1.5716968774795532,
      "rewards/margins": 9.01201057434082,
      "rewards/rejected": -7.440313339233398,
      "step": 3629
    },
    {
      "epoch": 0.9999344906649198,
      "step": 3816,
      "total_flos": 2.3140851990119055e+18,
      "train_loss": 0.09998038786012421,
      "train_runtime": 57165.5253,
      "train_samples_per_second": 0.534,
      "train_steps_per_second": 0.067
    }
  ],
  "logging_steps": 191,
  "max_steps": 3816,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.3140851990119055e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}