{
  "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": 36.21096742368932,
      "learning_rate": 5e-07,
      "logits/chosen": -2.735914945602417,
      "logits/rejected": -2.7412195205688477,
      "logps/chosen": -166.00094604492188,
      "logps/rejected": -162.81643676757812,
      "loss": 0.6928,
      "rewards/accuracies": 0.3375000059604645,
      "rewards/chosen": 0.00429560337215662,
      "rewards/margins": 0.0009204222005791962,
      "rewards/rejected": 0.0033751812297850847,
      "step": 5
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 35.97443184449595,
      "learning_rate": 1e-06,
      "logits/chosen": -2.742196798324585,
      "logits/rejected": -2.7352712154388428,
      "logps/chosen": -163.42056274414062,
      "logps/rejected": -168.62094116210938,
      "loss": 0.6867,
      "rewards/accuracies": 0.48124998807907104,
      "rewards/chosen": 0.10343559086322784,
      "rewards/margins": 0.005917676724493504,
      "rewards/rejected": 0.09751791507005692,
      "step": 10
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 35.479696664348296,
      "learning_rate": 9.985471028179154e-07,
      "logits/chosen": -2.715827465057373,
      "logits/rejected": -2.7099735736846924,
      "logps/chosen": -164.28744506835938,
      "logps/rejected": -166.86209106445312,
      "loss": 0.6686,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.31278976798057556,
      "rewards/margins": 0.10602164268493652,
      "rewards/rejected": 0.20676811039447784,
      "step": 15
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 34.26437345645622,
      "learning_rate": 9.94196854912548e-07,
      "logits/chosen": -2.6752734184265137,
      "logits/rejected": -2.670536518096924,
      "logps/chosen": -162.92506408691406,
      "logps/rejected": -162.58132934570312,
      "loss": 0.653,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": 0.23271910846233368,
      "rewards/margins": 0.16488614678382874,
      "rewards/rejected": 0.06783294677734375,
      "step": 20
    },
    {
      "epoch": 0.3472222222222222,
      "grad_norm": 38.12636989971247,
      "learning_rate": 9.869745381355905e-07,
      "logits/chosen": -2.612743854522705,
      "logits/rejected": -2.601036310195923,
      "logps/chosen": -169.65054321289062,
      "logps/rejected": -170.94589233398438,
      "loss": 0.6341,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.23453514277935028,
      "rewards/margins": 0.1853707879781723,
      "rewards/rejected": 0.04916436970233917,
      "step": 25
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 37.15638790112506,
      "learning_rate": 9.769221256218162e-07,
      "logits/chosen": -2.6376729011535645,
      "logits/rejected": -2.6211869716644287,
      "logps/chosen": -169.70230102539062,
      "logps/rejected": -169.1073760986328,
      "loss": 0.6276,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.197641983628273,
      "rewards/margins": 0.23505587875843048,
      "rewards/rejected": -0.03741389513015747,
      "step": 30
    },
    {
      "epoch": 0.4861111111111111,
      "grad_norm": 31.96578650923538,
      "learning_rate": 9.64098037858483e-07,
      "logits/chosen": -2.6476080417633057,
      "logits/rejected": -2.638826847076416,
      "logps/chosen": -164.2353515625,
      "logps/rejected": -171.78424072265625,
      "loss": 0.6001,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": 0.011483956128358841,
      "rewards/margins": 0.3633750379085541,
      "rewards/rejected": -0.35189107060432434,
      "step": 35
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 35.60629409012632,
      "learning_rate": 9.485768031694871e-07,
      "logits/chosen": -2.6523194313049316,
      "logits/rejected": -2.621492385864258,
      "logps/chosen": -168.99270629882812,
      "logps/rejected": -177.50718688964844,
      "loss": 0.5939,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": -0.2566075921058655,
      "rewards/margins": 0.5398613214492798,
      "rewards/rejected": -0.79646897315979,
      "step": 40
    },
    {
      "epoch": 0.625,
      "grad_norm": 32.80701192573668,
      "learning_rate": 9.304486245873971e-07,
      "logits/chosen": -2.657984495162964,
      "logits/rejected": -2.6483747959136963,
      "logps/chosen": -163.6527557373047,
      "logps/rejected": -167.71705627441406,
      "loss": 0.5942,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.24404068291187286,
      "rewards/margins": 0.5225220918655396,
      "rewards/rejected": -0.766562819480896,
      "step": 45
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 32.94692651420362,
      "learning_rate": 9.098188556305262e-07,
      "logits/chosen": -2.732595682144165,
      "logits/rejected": -2.7179951667785645,
      "logps/chosen": -159.82009887695312,
      "logps/rejected": -163.01516723632812,
      "loss": 0.575,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": -0.31872302293777466,
      "rewards/margins": 0.4740700125694275,
      "rewards/rejected": -0.7927930951118469,
      "step": 50
    },
    {
      "epoch": 0.6944444444444444,
      "eval_logits/chosen": -2.7981717586517334,
      "eval_logits/rejected": -2.7966415882110596,
      "eval_logps/chosen": -171.31138610839844,
      "eval_logps/rejected": -180.07443237304688,
      "eval_loss": 0.5679606199264526,
      "eval_rewards/accuracies": 0.69140625,
      "eval_rewards/chosen": -0.5232083201408386,
      "eval_rewards/margins": 0.6266617178916931,
      "eval_rewards/rejected": -1.1498699188232422,
      "eval_runtime": 127.2891,
      "eval_samples_per_second": 16.05,
      "eval_steps_per_second": 0.251,
      "step": 50
    },
    {
      "epoch": 0.7638888888888888,
      "grad_norm": 34.97882221943595,
      "learning_rate": 8.868073880316123e-07,
      "logits/chosen": -2.835651397705078,
      "logits/rejected": -2.836982250213623,
      "logps/chosen": -173.93702697753906,
      "logps/rejected": -180.1125030517578,
      "loss": 0.5562,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.5288220643997192,
      "rewards/margins": 0.7004331350326538,
      "rewards/rejected": -1.229255199432373,
      "step": 55
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 31.166211388759624,
      "learning_rate": 8.615479549763755e-07,
      "logits/chosen": -2.8652548789978027,
      "logits/rejected": -2.8450732231140137,
      "logps/chosen": -169.28530883789062,
      "logps/rejected": -171.57772827148438,
      "loss": 0.5334,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.237405925989151,
      "rewards/margins": 0.7358155846595764,
      "rewards/rejected": -0.973221480846405,
      "step": 60
    },
    {
      "epoch": 0.9027777777777778,
      "grad_norm": 31.846767846888632,
      "learning_rate": 8.341873539012443e-07,
      "logits/chosen": -2.8254175186157227,
      "logits/rejected": -2.8189828395843506,
      "logps/chosen": -178.87318420410156,
      "logps/rejected": -185.26193237304688,
      "loss": 0.5483,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.3370054364204407,
      "rewards/margins": 0.6962798237800598,
      "rewards/rejected": -1.0332852602005005,
      "step": 65
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 30.952350544641195,
      "learning_rate": 8.048845933670271e-07,
      "logits/chosen": -2.7731075286865234,
      "logits/rejected": -2.7470154762268066,
      "logps/chosen": -181.693359375,
      "logps/rejected": -192.0342254638672,
      "loss": 0.5378,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.617510974407196,
      "rewards/margins": 0.8088364601135254,
      "rewards/rejected": -1.4263474941253662,
      "step": 70
    },
    {
      "epoch": 1.0416666666666667,
      "grad_norm": 20.244570418413698,
      "learning_rate": 7.738099689665539e-07,
      "logits/chosen": -2.679137706756592,
      "logits/rejected": -2.676011800765991,
      "logps/chosen": -172.0131072998047,
      "logps/rejected": -184.72222900390625,
      "loss": 0.3552,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.7075360417366028,
      "rewards/margins": 1.6730060577392578,
      "rewards/rejected": -2.380542278289795,
      "step": 75
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 18.51901755323729,
      "learning_rate": 7.41144073636728e-07,
      "logits/chosen": -2.663628339767456,
      "logits/rejected": -2.6579511165618896,
      "logps/chosen": -181.416748046875,
      "logps/rejected": -201.2371063232422,
      "loss": 0.2457,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.018878469243645668,
      "rewards/margins": 2.686278820037842,
      "rewards/rejected": -2.667400360107422,
      "step": 80
    },
    {
      "epoch": 1.1805555555555556,
      "grad_norm": 20.24347077505837,
      "learning_rate": 7.070767481266492e-07,
      "logits/chosen": -2.68660569190979,
      "logits/rejected": -2.6812427043914795,
      "logps/chosen": -160.11874389648438,
      "logps/rejected": -179.59771728515625,
      "loss": 0.227,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.016312014311552048,
      "rewards/margins": 2.060859203338623,
      "rewards/rejected": -2.0771713256835938,
      "step": 85
    },
    {
      "epoch": 1.25,
      "grad_norm": 20.15132514330672,
      "learning_rate": 6.718059777212565e-07,
      "logits/chosen": -2.69787859916687,
      "logits/rejected": -2.7063913345336914,
      "logps/chosen": -165.69448852539062,
      "logps/rejected": -190.65296936035156,
      "loss": 0.2041,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.18171457946300507,
      "rewards/margins": 2.763362407684326,
      "rewards/rejected": -2.9450771808624268,
      "step": 90
    },
    {
      "epoch": 1.3194444444444444,
      "grad_norm": 21.629755831470078,
      "learning_rate": 6.355367416322778e-07,
      "logits/chosen": -2.7282795906066895,
      "logits/rejected": -2.7191052436828613,
      "logps/chosen": -176.59262084960938,
      "logps/rejected": -204.98123168945312,
      "loss": 0.2028,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.1713067591190338,
      "rewards/margins": 3.3725147247314453,
      "rewards/rejected": -3.543820858001709,
      "step": 95
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 26.9593328849758,
      "learning_rate": 5.984798217433531e-07,
      "logits/chosen": -2.690068006515503,
      "logits/rejected": -2.69694185256958,
      "logps/chosen": -170.9009246826172,
      "logps/rejected": -199.11679077148438,
      "loss": 0.2161,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": -0.24792905151844025,
      "rewards/margins": 3.1425349712371826,
      "rewards/rejected": -3.3904640674591064,
      "step": 100
    },
    {
      "epoch": 1.3888888888888888,
      "eval_logits/chosen": -2.6713719367980957,
      "eval_logits/rejected": -2.6708080768585205,
      "eval_logps/chosen": -177.34860229492188,
      "eval_logps/rejected": -191.46810913085938,
      "eval_loss": 0.541614830493927,
      "eval_rewards/accuracies": 0.74609375,
      "eval_rewards/chosen": -1.1269280910491943,
      "eval_rewards/margins": 1.162311315536499,
      "eval_rewards/rejected": -2.2892394065856934,
      "eval_runtime": 126.9171,
      "eval_samples_per_second": 16.097,
      "eval_steps_per_second": 0.252,
      "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
}