{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 50,
  "global_step": 216,
  "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
    },
    {
      "epoch": 1.4583333333333333,
      "grad_norm": 20.618955156266612,
      "learning_rate": 5.608505776324157e-07,
      "logits/chosen": -2.6715927124023438,
      "logits/rejected": -2.678304433822632,
      "logps/chosen": -160.5848388671875,
      "logps/rejected": -191.11619567871094,
      "loss": 0.1956,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.33186909556388855,
      "rewards/margins": 2.948606491088867,
      "rewards/rejected": -3.280475616455078,
      "step": 105
    },
    {
      "epoch": 1.5277777777777777,
      "grad_norm": 27.47403889294586,
      "learning_rate": 5.228676949903973e-07,
      "logits/chosen": -2.677685022354126,
      "logits/rejected": -2.6649279594421387,
      "logps/chosen": -167.76614379882812,
      "logps/rejected": -201.45071411132812,
      "loss": 0.2051,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.7044020295143127,
      "rewards/margins": 3.09299373626709,
      "rewards/rejected": -3.7973952293395996,
      "step": 110
    },
    {
      "epoch": 1.5972222222222223,
      "grad_norm": 25.78263768190195,
      "learning_rate": 4.847519147099294e-07,
      "logits/chosen": -2.673830986022949,
      "logits/rejected": -2.667365789413452,
      "logps/chosen": -165.1271209716797,
      "logps/rejected": -192.34664916992188,
      "loss": 0.2023,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": -0.7367149591445923,
      "rewards/margins": 2.9380996227264404,
      "rewards/rejected": -3.674814224243164,
      "step": 115
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 20.347389718573403,
      "learning_rate": 4.46724750030062e-07,
      "logits/chosen": -2.6994948387145996,
      "logits/rejected": -2.657327175140381,
      "logps/chosen": -167.97816467285156,
      "logps/rejected": -196.3680877685547,
      "loss": 0.1879,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": -0.4436143934726715,
      "rewards/margins": 3.0548558235168457,
      "rewards/rejected": -3.4984703063964844,
      "step": 120
    },
    {
      "epoch": 1.7361111111111112,
      "grad_norm": 23.654938219435277,
      "learning_rate": 4.0900719919241935e-07,
      "logits/chosen": -2.6885862350463867,
      "logits/rejected": -2.6681549549102783,
      "logps/chosen": -174.38291931152344,
      "logps/rejected": -210.3325653076172,
      "loss": 0.1744,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": -0.4140690863132477,
      "rewards/margins": 3.468705654144287,
      "rewards/rejected": -3.8827743530273438,
      "step": 125
    },
    {
      "epoch": 1.8055555555555556,
      "grad_norm": 24.500980504365117,
      "learning_rate": 3.7181846109031e-07,
      "logits/chosen": -2.6993517875671387,
      "logits/rejected": -2.6847097873687744,
      "logps/chosen": -162.849609375,
      "logps/rejected": -192.41830444335938,
      "loss": 0.1788,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.6508005857467651,
      "rewards/margins": 3.144455671310425,
      "rewards/rejected": -3.7952563762664795,
      "step": 130
    },
    {
      "epoch": 1.875,
      "grad_norm": 26.706394086119467,
      "learning_rate": 3.353746613749093e-07,
      "logits/chosen": -2.6868338584899902,
      "logits/rejected": -2.6891016960144043,
      "logps/chosen": -171.8800506591797,
      "logps/rejected": -201.1620330810547,
      "loss": 0.1877,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.5446206331253052,
      "rewards/margins": 3.2972817420959473,
      "rewards/rejected": -3.841902256011963,
      "step": 135
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 21.039975863861176,
      "learning_rate": 2.9988759642186093e-07,
      "logits/chosen": -2.677610397338867,
      "logits/rejected": -2.6499438285827637,
      "logps/chosen": -178.9335174560547,
      "logps/rejected": -216.8883056640625,
      "loss": 0.1809,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.4429135322570801,
      "rewards/margins": 3.727440595626831,
      "rewards/rejected": -4.17035436630249,
      "step": 140
    },
    {
      "epoch": 2.013888888888889,
      "grad_norm": 15.18765450974654,
      "learning_rate": 2.655635024578483e-07,
      "logits/chosen": -2.678591012954712,
      "logits/rejected": -2.6765263080596924,
      "logps/chosen": -175.04360961914062,
      "logps/rejected": -206.4744415283203,
      "loss": 0.1621,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": -0.7178301215171814,
      "rewards/margins": 3.4394123554229736,
      "rewards/rejected": -4.1572418212890625,
      "step": 145
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 11.856035922703338,
      "learning_rate": 2.3260185700046292e-07,
      "logits/chosen": -2.6947238445281982,
      "logits/rejected": -2.6655373573303223,
      "logps/chosen": -173.5478057861328,
      "logps/rejected": -222.5522918701172,
      "loss": 0.0912,
      "rewards/accuracies": 0.9937499761581421,
      "rewards/chosen": -0.3790014088153839,
      "rewards/margins": 4.613499164581299,
      "rewards/rejected": -4.9925007820129395,
      "step": 150
    },
    {
      "epoch": 2.0833333333333335,
      "eval_logits/chosen": -2.6673920154571533,
      "eval_logits/rejected": -2.6701459884643555,
      "eval_logps/chosen": -187.42156982421875,
      "eval_logps/rejected": -204.27391052246094,
      "eval_loss": 0.5558860898017883,
      "eval_rewards/accuracies": 0.71875,
      "eval_rewards/chosen": -2.1342270374298096,
      "eval_rewards/margins": 1.435591697692871,
      "eval_rewards/rejected": -3.5698184967041016,
      "eval_runtime": 126.9326,
      "eval_samples_per_second": 16.095,
      "eval_steps_per_second": 0.252,
      "step": 150
    },
    {
      "epoch": 2.1527777777777777,
      "grad_norm": 19.648989383169503,
      "learning_rate": 2.0119421957691218e-07,
      "logits/chosen": -2.655550956726074,
      "logits/rejected": -2.667914390563965,
      "logps/chosen": -181.70692443847656,
      "logps/rejected": -226.82400512695312,
      "loss": 0.0896,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.7839339375495911,
      "rewards/margins": 4.58644962310791,
      "rewards/rejected": -5.370383262634277,
      "step": 155
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 15.624018161097496,
      "learning_rate": 1.7152311845883094e-07,
      "logits/chosen": -2.6638529300689697,
      "logits/rejected": -2.619544267654419,
      "logps/chosen": -173.3532257080078,
      "logps/rejected": -220.2028350830078,
      "loss": 0.0812,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -1.1546170711517334,
      "rewards/margins": 4.321033000946045,
      "rewards/rejected": -5.475650310516357,
      "step": 160
    },
    {
      "epoch": 2.2916666666666665,
      "grad_norm": 17.11020248803683,
      "learning_rate": 1.4376098988303404e-07,
      "logits/chosen": -2.63069224357605,
      "logits/rejected": -2.6121063232421875,
      "logps/chosen": -173.0404510498047,
      "logps/rejected": -211.6892852783203,
      "loss": 0.0818,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -1.21498703956604,
      "rewards/margins": 4.448957443237305,
      "rewards/rejected": -5.663943767547607,
      "step": 165
    },
    {
      "epoch": 2.361111111111111,
      "grad_norm": 17.112572527447032,
      "learning_rate": 1.1806917592302761e-07,
      "logits/chosen": -2.6293816566467285,
      "logits/rejected": -2.597991943359375,
      "logps/chosen": -176.73304748535156,
      "logps/rejected": -224.79714965820312,
      "loss": 0.078,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": -0.8656209707260132,
      "rewards/margins": 5.170679092407227,
      "rewards/rejected": -6.036300182342529,
      "step": 170
    },
    {
      "epoch": 2.4305555555555554,
      "grad_norm": 16.22760001506656,
      "learning_rate": 9.459698683523204e-08,
      "logits/chosen": -2.6276373863220215,
      "logits/rejected": -2.627812147140503,
      "logps/chosen": -182.4290008544922,
      "logps/rejected": -230.18063354492188,
      "loss": 0.0849,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": -1.0544099807739258,
      "rewards/margins": 5.050833702087402,
      "rewards/rejected": -6.1052446365356445,
      "step": 175
    },
    {
      "epoch": 2.5,
      "grad_norm": 18.33379110443006,
      "learning_rate": 7.348083332917926e-08,
      "logits/chosen": -2.625272035598755,
      "logits/rejected": -2.597888946533203,
      "logps/chosen": -174.5447235107422,
      "logps/rejected": -223.33822631835938,
      "loss": 0.0853,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -1.2181237936019897,
      "rewards/margins": 4.818295478820801,
      "rewards/rejected": -6.0364203453063965,
      "step": 180
    },
    {
      "epoch": 2.5694444444444446,
      "grad_norm": 19.469578336412205,
      "learning_rate": 5.484343380457124e-08,
      "logits/chosen": -2.604079008102417,
      "logits/rejected": -2.6138792037963867,
      "logps/chosen": -168.84449768066406,
      "logps/rejected": -218.6962432861328,
      "loss": 0.0846,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": -1.0644819736480713,
      "rewards/margins": 4.62592077255249,
      "rewards/rejected": -5.690402030944824,
      "step": 185
    },
    {
      "epoch": 2.638888888888889,
      "grad_norm": 21.805307710932198,
      "learning_rate": 3.879310116241041e-08,
      "logits/chosen": -2.6105732917785645,
      "logits/rejected": -2.596642017364502,
      "logps/chosen": -174.59005737304688,
      "logps/rejected": -219.40878295898438,
      "loss": 0.0821,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -1.2059953212738037,
      "rewards/margins": 4.756261348724365,
      "rewards/rejected": -5.962257385253906,
      "step": 190
    },
    {
      "epoch": 2.7083333333333335,
      "grad_norm": 18.608644466504245,
      "learning_rate": 2.5423113334966218e-08,
      "logits/chosen": -2.6064913272857666,
      "logits/rejected": -2.616105794906616,
      "logps/chosen": -185.19998168945312,
      "logps/rejected": -225.87014770507812,
      "loss": 0.0772,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": -0.8651708364486694,
      "rewards/margins": 4.7674455642700195,
      "rewards/rejected": -5.6326165199279785,
      "step": 195
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 17.74237497662627,
      "learning_rate": 1.4811171192794624e-08,
      "logits/chosen": -2.6110963821411133,
      "logits/rejected": -2.6080284118652344,
      "logps/chosen": -171.61239624023438,
      "logps/rejected": -220.25830078125,
      "loss": 0.0828,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -1.1769065856933594,
      "rewards/margins": 4.8281941413879395,
      "rewards/rejected": -6.005099773406982,
      "step": 200
    },
    {
      "epoch": 2.7777777777777777,
      "eval_logits/chosen": -2.60565185546875,
      "eval_logits/rejected": -2.607104539871216,
      "eval_logps/chosen": -193.79588317871094,
      "eval_logps/rejected": -210.87051391601562,
      "eval_loss": 0.590232253074646,
      "eval_rewards/accuracies": 0.72265625,
      "eval_rewards/chosen": -2.7716591358184814,
      "eval_rewards/margins": 1.457817554473877,
      "eval_rewards/rejected": -4.2294769287109375,
      "eval_runtime": 127.0051,
      "eval_samples_per_second": 16.086,
      "eval_steps_per_second": 0.252,
      "step": 200
    },
    {
      "epoch": 2.8472222222222223,
      "grad_norm": 20.206548432698874,
      "learning_rate": 7.018946979234997e-09,
      "logits/chosen": -2.6121597290039062,
      "logits/rejected": -2.6113522052764893,
      "logps/chosen": -171.62460327148438,
      "logps/rejected": -223.3975830078125,
      "loss": 0.088,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": -1.0721832513809204,
      "rewards/margins": 4.880965232849121,
      "rewards/rejected": -5.95314884185791,
      "step": 205
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 16.26368007303619,
      "learning_rate": 2.0917258966953734e-09,
      "logits/chosen": -2.628420114517212,
      "logits/rejected": -2.607481002807617,
      "logps/chosen": -178.46401977539062,
      "logps/rejected": -225.849365234375,
      "loss": 0.0771,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -1.2336267232894897,
      "rewards/margins": 4.748377323150635,
      "rewards/rejected": -5.982003211975098,
      "step": 210
    },
    {
      "epoch": 2.986111111111111,
      "grad_norm": 13.799681183452929,
      "learning_rate": 5.814292768108187e-11,
      "logits/chosen": -2.610588550567627,
      "logits/rejected": -2.587486505508423,
      "logps/chosen": -188.3278045654297,
      "logps/rejected": -241.2031707763672,
      "loss": 0.0775,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": -0.9807193875312805,
      "rewards/margins": 5.365724563598633,
      "rewards/rejected": -6.346443176269531,
      "step": 215
    },
    {
      "epoch": 3.0,
      "step": 216,
      "total_flos": 2546799910846464.0,
      "train_loss": 0.29657687264046184,
      "train_runtime": 7656.725,
      "train_samples_per_second": 7.203,
      "train_steps_per_second": 0.028
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2546799910846464.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}