{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8316008316008316,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04158004158004158,
      "grad_norm": 60.93168963253538,
      "learning_rate": 5e-07,
      "logits/chosen": -2.7216269969940186,
      "logits/rejected": -2.719656467437744,
      "logps/chosen": -279.2793884277344,
      "logps/rejected": -217.73977661132812,
      "loss": 0.6903,
      "rewards/accuracies": 0.3187499940395355,
      "rewards/chosen": 0.02068181149661541,
      "rewards/margins": 0.008613892830908298,
      "rewards/rejected": 0.012067919597029686,
      "step": 5
    },
    {
      "epoch": 0.08316008316008316,
      "grad_norm": 56.89106976222382,
      "learning_rate": 1e-06,
      "logits/chosen": -2.6707825660705566,
      "logits/rejected": -2.6669201850891113,
      "logps/chosen": -256.79901123046875,
      "logps/rejected": -231.99880981445312,
      "loss": 0.655,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": 0.5268703103065491,
      "rewards/margins": 0.16865657269954681,
      "rewards/rejected": 0.35821372270584106,
      "step": 10
    },
    {
      "epoch": 0.12474012474012475,
      "grad_norm": 63.52271997282497,
      "learning_rate": 9.994965332706572e-07,
      "logits/chosen": -2.513793468475342,
      "logits/rejected": -2.5043387413024902,
      "logps/chosen": -264.4665222167969,
      "logps/rejected": -220.4278564453125,
      "loss": 0.6536,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 1.3569260835647583,
      "rewards/margins": 0.6852619051933289,
      "rewards/rejected": 0.6716641187667847,
      "step": 15
    },
    {
      "epoch": 0.16632016632016633,
      "grad_norm": 43.71783786864763,
      "learning_rate": 9.979871469976195e-07,
      "logits/chosen": -2.389570713043213,
      "logits/rejected": -2.362312078475952,
      "logps/chosen": -237.1497802734375,
      "logps/rejected": -194.1293182373047,
      "loss": 0.5961,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 1.4138034582138062,
      "rewards/margins": 0.9070841073989868,
      "rewards/rejected": 0.5067193508148193,
      "step": 20
    },
    {
      "epoch": 0.2079002079002079,
      "grad_norm": 46.44405380817153,
      "learning_rate": 9.954748808839674e-07,
      "logits/chosen": -2.2999889850616455,
      "logits/rejected": -2.317258358001709,
      "logps/chosen": -260.94281005859375,
      "logps/rejected": -250.1718292236328,
      "loss": 0.6178,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 1.4349501132965088,
      "rewards/margins": 1.189349889755249,
      "rewards/rejected": 0.2456001341342926,
      "step": 25
    },
    {
      "epoch": 0.2494802494802495,
      "grad_norm": 42.03082336502749,
      "learning_rate": 9.919647942993147e-07,
      "logits/chosen": -2.3956313133239746,
      "logits/rejected": -2.3552544116973877,
      "logps/chosen": -241.47103881835938,
      "logps/rejected": -215.29940795898438,
      "loss": 0.6159,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.977073073387146,
      "rewards/margins": 1.020427942276001,
      "rewards/rejected": -0.04335489124059677,
      "step": 30
    },
    {
      "epoch": 0.2910602910602911,
      "grad_norm": 46.49384193628988,
      "learning_rate": 9.874639560909118e-07,
      "logits/chosen": -2.493499994277954,
      "logits/rejected": -2.4809508323669434,
      "logps/chosen": -251.2248992919922,
      "logps/rejected": -201.01251220703125,
      "loss": 0.6348,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 0.6344128847122192,
      "rewards/margins": 0.7719027400016785,
      "rewards/rejected": -0.13748976588249207,
      "step": 35
    },
    {
      "epoch": 0.33264033264033266,
      "grad_norm": 36.79345771283175,
      "learning_rate": 9.819814303479267e-07,
      "logits/chosen": -2.4675540924072266,
      "logits/rejected": -2.46543288230896,
      "logps/chosen": -261.19268798828125,
      "logps/rejected": -202.8013153076172,
      "loss": 0.5763,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.6426112651824951,
      "rewards/margins": 0.9732248187065125,
      "rewards/rejected": -0.33061355352401733,
      "step": 40
    },
    {
      "epoch": 0.37422037422037424,
      "grad_norm": 41.675486172068894,
      "learning_rate": 9.755282581475767e-07,
      "logits/chosen": -2.425602436065674,
      "logits/rejected": -2.4132189750671387,
      "logps/chosen": -243.4677734375,
      "logps/rejected": -185.2170867919922,
      "loss": 0.6023,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.7012235522270203,
      "rewards/margins": 0.7697752714157104,
      "rewards/rejected": -0.06855180859565735,
      "step": 45
    },
    {
      "epoch": 0.4158004158004158,
      "grad_norm": 41.93176697972712,
      "learning_rate": 9.681174353198686e-07,
      "logits/chosen": -2.4606122970581055,
      "logits/rejected": -2.451716184616089,
      "logps/chosen": -232.999267578125,
      "logps/rejected": -211.801025390625,
      "loss": 0.6087,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 0.8648797869682312,
      "rewards/margins": 1.025213599205017,
      "rewards/rejected": -0.16033388674259186,
      "step": 50
    },
    {
      "epoch": 0.4158004158004158,
      "eval_logits/chosen": -2.5455965995788574,
      "eval_logits/rejected": -2.528465509414673,
      "eval_logps/chosen": -241.17100524902344,
      "eval_logps/rejected": -207.0612335205078,
      "eval_loss": 0.5778272747993469,
      "eval_rewards/accuracies": 0.7870370149612427,
      "eval_rewards/chosen": 0.7418816685676575,
      "eval_rewards/margins": 1.2493042945861816,
      "eval_rewards/rejected": -0.5074228644371033,
      "eval_runtime": 229.2387,
      "eval_samples_per_second": 14.919,
      "eval_steps_per_second": 0.236,
      "step": 50
    },
    {
      "epoch": 0.4573804573804574,
      "grad_norm": 37.969958241527884,
      "learning_rate": 9.597638862757253e-07,
      "logits/chosen": -2.5910000801086426,
      "logits/rejected": -2.5817267894744873,
      "logps/chosen": -235.71316528320312,
      "logps/rejected": -194.326171875,
      "loss": 0.5877,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.866040825843811,
      "rewards/margins": 1.1515779495239258,
      "rewards/rejected": -0.28553709387779236,
      "step": 55
    },
    {
      "epoch": 0.498960498960499,
      "grad_norm": 55.72037568304586,
      "learning_rate": 9.504844339512094e-07,
      "logits/chosen": -2.643787145614624,
      "logits/rejected": -2.643979787826538,
      "logps/chosen": -255.2833709716797,
      "logps/rejected": -219.94290161132812,
      "loss": 0.604,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 1.1731880903244019,
      "rewards/margins": 1.4057620763778687,
      "rewards/rejected": -0.23257413506507874,
      "step": 60
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 42.3606140855208,
      "learning_rate": 9.402977659283689e-07,
      "logits/chosen": -2.645413875579834,
      "logits/rejected": -2.6578667163848877,
      "logps/chosen": -233.7159423828125,
      "logps/rejected": -209.6158447265625,
      "loss": 0.556,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 1.1114470958709717,
      "rewards/margins": 1.3722739219665527,
      "rewards/rejected": -0.2608269453048706,
      "step": 65
    },
    {
      "epoch": 0.5821205821205822,
      "grad_norm": 36.14760649472159,
      "learning_rate": 9.29224396800933e-07,
      "logits/chosen": -2.5539443492889404,
      "logits/rejected": -2.582876682281494,
      "logps/chosen": -252.9326171875,
      "logps/rejected": -204.24783325195312,
      "loss": 0.5409,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": 0.9803892970085144,
      "rewards/margins": 1.488160252571106,
      "rewards/rejected": -0.507770836353302,
      "step": 70
    },
    {
      "epoch": 0.6237006237006237,
      "grad_norm": 40.66552692069641,
      "learning_rate": 9.172866268606513e-07,
      "logits/chosen": -2.556718349456787,
      "logits/rejected": -2.5814919471740723,
      "logps/chosen": -252.392333984375,
      "logps/rejected": -230.672607421875,
      "loss": 0.5972,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.6901723146438599,
      "rewards/margins": 1.7339065074920654,
      "rewards/rejected": -1.0437343120574951,
      "step": 75
    },
    {
      "epoch": 0.6652806652806653,
      "grad_norm": 37.41696546929204,
      "learning_rate": 9.045084971874737e-07,
      "logits/chosen": -2.5521421432495117,
      "logits/rejected": -2.558589220046997,
      "logps/chosen": -238.3732452392578,
      "logps/rejected": -218.8362274169922,
      "loss": 0.5516,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": 0.7391923666000366,
      "rewards/margins": 1.9195365905761719,
      "rewards/rejected": -1.1803443431854248,
      "step": 80
    },
    {
      "epoch": 0.7068607068607069,
      "grad_norm": 32.554352102245765,
      "learning_rate": 8.909157412340149e-07,
      "logits/chosen": -2.5185506343841553,
      "logits/rejected": -2.5202035903930664,
      "logps/chosen": -244.92233276367188,
      "logps/rejected": -222.07748413085938,
      "loss": 0.5411,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.8945465087890625,
      "rewards/margins": 1.484235167503357,
      "rewards/rejected": -0.589688777923584,
      "step": 85
    },
    {
      "epoch": 0.7484407484407485,
      "grad_norm": 41.80775702009017,
      "learning_rate": 8.765357330018055e-07,
      "logits/chosen": -2.527893543243408,
      "logits/rejected": -2.4960551261901855,
      "logps/chosen": -244.02468872070312,
      "logps/rejected": -226.1230926513672,
      "loss": 0.5433,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.863143265247345,
      "rewards/margins": 1.4539084434509277,
      "rewards/rejected": -0.5907653570175171,
      "step": 90
    },
    {
      "epoch": 0.7900207900207901,
      "grad_norm": 35.599529291827764,
      "learning_rate": 8.613974319136957e-07,
      "logits/chosen": -2.546417713165283,
      "logits/rejected": -2.5654358863830566,
      "logps/chosen": -240.69375610351562,
      "logps/rejected": -204.02964782714844,
      "loss": 0.4936,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": 1.335752248764038,
      "rewards/margins": 1.7393795251846313,
      "rewards/rejected": -0.40362730622291565,
      "step": 95
    },
    {
      "epoch": 0.8316008316008316,
      "grad_norm": 42.869388172320754,
      "learning_rate": 8.455313244934324e-07,
      "logits/chosen": -2.597480297088623,
      "logits/rejected": -2.5727450847625732,
      "logps/chosen": -242.670166015625,
      "logps/rejected": -230.5889129638672,
      "loss": 0.5039,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.7388527393341064,
      "rewards/margins": 1.7855724096298218,
      "rewards/rejected": -1.0467197895050049,
      "step": 100
    },
    {
      "epoch": 0.8316008316008316,
      "eval_logits/chosen": -2.6104719638824463,
      "eval_logits/rejected": -2.5932347774505615,
      "eval_logps/chosen": -245.64724731445312,
      "eval_logps/rejected": -217.64056396484375,
      "eval_loss": 0.5226317048072815,
      "eval_rewards/accuracies": 0.8055555820465088,
      "eval_rewards/chosen": 0.2942599654197693,
      "eval_rewards/margins": 1.8596160411834717,
      "eval_rewards/rejected": -1.5653562545776367,
      "eval_runtime": 229.0549,
      "eval_samples_per_second": 14.931,
      "eval_steps_per_second": 0.236,
      "step": 100
    }
  ],
  "logging_steps": 5,
  "max_steps": 360,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1178822762299392.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}