{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.3888888888888888,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06944444444444445,
      "grad_norm": 34.922038802469906,
      "learning_rate": 5e-07,
      "logits/chosen": -2.745856761932373,
      "logits/rejected": -2.7519428730010986,
      "logps/chosen": -158.59893798828125,
      "logps/rejected": -160.2094268798828,
      "loss": 0.6939,
      "rewards/accuracies": 0.33125001192092896,
      "rewards/chosen": 0.000581090513151139,
      "rewards/margins": -0.00047404784709215164,
      "rewards/rejected": 0.0010551378363743424,
      "step": 5
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 36.413747880379944,
      "learning_rate": 1e-06,
      "logits/chosen": -2.736849546432495,
      "logits/rejected": -2.7453856468200684,
      "logps/chosen": -174.9836883544922,
      "logps/rejected": -171.3789825439453,
      "loss": 0.6883,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.022891324013471603,
      "rewards/margins": 0.0072159310802817345,
      "rewards/rejected": 0.015675392001867294,
      "step": 10
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 37.39831623643963,
      "learning_rate": 9.985471028179154e-07,
      "logits/chosen": -2.6100494861602783,
      "logits/rejected": -2.613814115524292,
      "logps/chosen": -173.74456787109375,
      "logps/rejected": -171.9447784423828,
      "loss": 0.6733,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": 0.21788981556892395,
      "rewards/margins": 0.09981251507997513,
      "rewards/rejected": 0.11807730048894882,
      "step": 15
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 32.21436012894337,
      "learning_rate": 9.94196854912548e-07,
      "logits/chosen": -2.52358341217041,
      "logits/rejected": -2.5229320526123047,
      "logps/chosen": -155.1029815673828,
      "logps/rejected": -159.9917755126953,
      "loss": 0.653,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.3672500252723694,
      "rewards/margins": 0.20777833461761475,
      "rewards/rejected": 0.15947169065475464,
      "step": 20
    },
    {
      "epoch": 0.3472222222222222,
      "grad_norm": 32.0223445235898,
      "learning_rate": 9.869745381355905e-07,
      "logits/chosen": -2.4307353496551514,
      "logits/rejected": -2.429652452468872,
      "logps/chosen": -159.20545959472656,
      "logps/rejected": -158.57797241210938,
      "loss": 0.6327,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": 0.3310449421405792,
      "rewards/margins": 0.22794027626514435,
      "rewards/rejected": 0.10310468822717667,
      "step": 25
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 36.79676539267814,
      "learning_rate": 9.769221256218162e-07,
      "logits/chosen": -2.4173130989074707,
      "logits/rejected": -2.3855977058410645,
      "logps/chosen": -165.49618530273438,
      "logps/rejected": -164.41111755371094,
      "loss": 0.6264,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.17511680722236633,
      "rewards/margins": 0.2607855200767517,
      "rewards/rejected": -0.43590235710144043,
      "step": 30
    },
    {
      "epoch": 0.4861111111111111,
      "grad_norm": 32.497757396344305,
      "learning_rate": 9.64098037858483e-07,
      "logits/chosen": -2.482489824295044,
      "logits/rejected": -2.4674153327941895,
      "logps/chosen": -162.7029571533203,
      "logps/rejected": -164.48619079589844,
      "loss": 0.5914,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.3191075921058655,
      "rewards/margins": 0.362353652715683,
      "rewards/rejected": -0.6814612150192261,
      "step": 35
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 34.40764043054464,
      "learning_rate": 9.485768031694871e-07,
      "logits/chosen": -2.518897771835327,
      "logits/rejected": -2.5143680572509766,
      "logps/chosen": -167.1140899658203,
      "logps/rejected": -168.67987060546875,
      "loss": 0.58,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.037984561175107956,
      "rewards/margins": 0.48854589462280273,
      "rewards/rejected": -0.5265304446220398,
      "step": 40
    },
    {
      "epoch": 0.625,
      "grad_norm": 33.97876407643923,
      "learning_rate": 9.304486245873971e-07,
      "logits/chosen": -2.542407989501953,
      "logits/rejected": -2.510068416595459,
      "logps/chosen": -163.48536682128906,
      "logps/rejected": -164.54147338867188,
      "loss": 0.5815,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.062438905239105225,
      "rewards/margins": 0.5621680021286011,
      "rewards/rejected": -0.6246069073677063,
      "step": 45
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 32.726531127211416,
      "learning_rate": 9.098188556305262e-07,
      "logits/chosen": -2.513942241668701,
      "logits/rejected": -2.51139235496521,
      "logps/chosen": -165.90362548828125,
      "logps/rejected": -170.9777069091797,
      "loss": 0.5342,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.2061791867017746,
      "rewards/margins": 0.6521804928779602,
      "rewards/rejected": -0.8583596348762512,
      "step": 50
    },
    {
      "epoch": 0.6944444444444444,
      "eval_logits/chosen": -2.4995272159576416,
      "eval_logits/rejected": -2.4678475856781006,
      "eval_logps/chosen": -159.87913513183594,
      "eval_logps/rejected": -170.17739868164062,
      "eval_loss": 0.5209046602249146,
      "eval_rewards/accuracies": 0.71484375,
      "eval_rewards/chosen": -0.20708096027374268,
      "eval_rewards/margins": 0.7420069575309753,
      "eval_rewards/rejected": -0.9490878582000732,
      "eval_runtime": 129.0562,
      "eval_samples_per_second": 15.83,
      "eval_steps_per_second": 0.248,
      "step": 50
    },
    {
      "epoch": 0.7638888888888888,
      "grad_norm": 30.615472770103583,
      "learning_rate": 8.868073880316123e-07,
      "logits/chosen": -2.500612497329712,
      "logits/rejected": -2.520470142364502,
      "logps/chosen": -177.12451171875,
      "logps/rejected": -183.56619262695312,
      "loss": 0.5577,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.32517191767692566,
      "rewards/margins": 0.8004968762397766,
      "rewards/rejected": -1.1256687641143799,
      "step": 55
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 32.6842144140191,
      "learning_rate": 8.615479549763755e-07,
      "logits/chosen": -2.5329596996307373,
      "logits/rejected": -2.5310966968536377,
      "logps/chosen": -183.05056762695312,
      "logps/rejected": -193.45396423339844,
      "loss": 0.5335,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.4021238386631012,
      "rewards/margins": 0.8954024314880371,
      "rewards/rejected": -1.297526240348816,
      "step": 60
    },
    {
      "epoch": 0.9027777777777778,
      "grad_norm": 31.818374577978716,
      "learning_rate": 8.341873539012443e-07,
      "logits/chosen": -2.518287420272827,
      "logits/rejected": -2.5078444480895996,
      "logps/chosen": -168.2073974609375,
      "logps/rejected": -179.64515686035156,
      "loss": 0.5272,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.6193049550056458,
      "rewards/margins": 0.8720922470092773,
      "rewards/rejected": -1.4913971424102783,
      "step": 65
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 30.74282933667216,
      "learning_rate": 8.048845933670271e-07,
      "logits/chosen": -2.474501371383667,
      "logits/rejected": -2.486248731613159,
      "logps/chosen": -170.53952026367188,
      "logps/rejected": -183.97750854492188,
      "loss": 0.516,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": -1.0543142557144165,
      "rewards/margins": 0.9860897064208984,
      "rewards/rejected": -2.0404040813446045,
      "step": 70
    },
    {
      "epoch": 1.0416666666666667,
      "grad_norm": 20.857882794392346,
      "learning_rate": 7.738099689665539e-07,
      "logits/chosen": -2.4805214405059814,
      "logits/rejected": -2.4716594219207764,
      "logps/chosen": -179.7224884033203,
      "logps/rejected": -194.9944610595703,
      "loss": 0.3668,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.5534920692443848,
      "rewards/margins": 1.5888177156448364,
      "rewards/rejected": -2.1423099040985107,
      "step": 75
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 19.198293152522673,
      "learning_rate": 7.41144073636728e-07,
      "logits/chosen": -2.4745230674743652,
      "logits/rejected": -2.449141025543213,
      "logps/chosen": -166.55409240722656,
      "logps/rejected": -190.5934295654297,
      "loss": 0.2405,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 0.21927456557750702,
      "rewards/margins": 2.280102014541626,
      "rewards/rejected": -2.0608274936676025,
      "step": 80
    },
    {
      "epoch": 1.1805555555555556,
      "grad_norm": 22.513577769114043,
      "learning_rate": 7.070767481266492e-07,
      "logits/chosen": -2.457094192504883,
      "logits/rejected": -2.434199571609497,
      "logps/chosen": -168.0836944580078,
      "logps/rejected": -188.27374267578125,
      "loss": 0.2396,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.004831351339817047,
      "rewards/margins": 2.4242899417877197,
      "rewards/rejected": -2.429121494293213,
      "step": 85
    },
    {
      "epoch": 1.25,
      "grad_norm": 18.850259072813216,
      "learning_rate": 6.718059777212565e-07,
      "logits/chosen": -2.4347877502441406,
      "logits/rejected": -2.432413101196289,
      "logps/chosen": -154.73831176757812,
      "logps/rejected": -174.75621032714844,
      "loss": 0.2034,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.1518968939781189,
      "rewards/margins": 2.676344394683838,
      "rewards/rejected": -2.8282413482666016,
      "step": 90
    },
    {
      "epoch": 1.3194444444444444,
      "grad_norm": 19.10543784815799,
      "learning_rate": 6.355367416322778e-07,
      "logits/chosen": -2.433638572692871,
      "logits/rejected": -2.4281392097473145,
      "logps/chosen": -173.74961853027344,
      "logps/rejected": -198.87208557128906,
      "loss": 0.2048,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": -0.30015555024147034,
      "rewards/margins": 2.652449369430542,
      "rewards/rejected": -2.9526050090789795,
      "step": 95
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 19.991814546360683,
      "learning_rate": 5.984798217433531e-07,
      "logits/chosen": -2.434572696685791,
      "logits/rejected": -2.417861223220825,
      "logps/chosen": -158.5247802734375,
      "logps/rejected": -190.9485626220703,
      "loss": 0.217,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": -0.37773507833480835,
      "rewards/margins": 2.8572323322296143,
      "rewards/rejected": -3.2349674701690674,
      "step": 100
    },
    {
      "epoch": 1.3888888888888888,
      "eval_logits/chosen": -2.4349896907806396,
      "eval_logits/rejected": -2.40444278717041,
      "eval_logps/chosen": -169.47793579101562,
      "eval_logps/rejected": -184.55125427246094,
      "eval_loss": 0.5166749954223633,
      "eval_rewards/accuracies": 0.70703125,
      "eval_rewards/chosen": -1.1669622659683228,
      "eval_rewards/margins": 1.2195117473602295,
      "eval_rewards/rejected": -2.386474132537842,
      "eval_runtime": 128.5113,
      "eval_samples_per_second": 15.897,
      "eval_steps_per_second": 0.249,
      "step": 100
    }
  ],
  "logging_steps": 5,
  "max_steps": 216,
  "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
}