{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 14.285714285714286,
  "eval_steps": 100,
  "global_step": 400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 54.0,
      "learning_rate": 1.44e-06,
      "log_odds_chosen": 0.12235047668218613,
      "log_odds_ratio": -0.653965413570404,
      "logits/chosen": 1.3041611909866333,
      "logits/rejected": 1.3057336807250977,
      "logps/chosen": -0.30839473009109497,
      "logps/rejected": -0.3456656336784363,
      "loss": 0.9624,
      "nll_loss": 0.3083947002887726,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -0.30839473009109497,
      "rewards/margins": 0.03727094456553459,
      "rewards/rejected": -0.3456656336784363,
      "step": 10
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 48.5,
      "learning_rate": 3.0399999999999997e-06,
      "log_odds_chosen": 0.08590120822191238,
      "log_odds_ratio": -0.680418848991394,
      "logits/chosen": 1.3281370401382446,
      "logits/rejected": 1.3296444416046143,
      "logps/chosen": -0.2094467133283615,
      "logps/rejected": -0.22896623611450195,
      "loss": 0.8899,
      "nll_loss": 0.2094467431306839,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.2094467133283615,
      "rewards/margins": 0.019519509747624397,
      "rewards/rejected": -0.22896623611450195,
      "step": 20
    },
    {
      "epoch": 1.0714285714285714,
      "grad_norm": 37.5,
      "learning_rate": 4.64e-06,
      "log_odds_chosen": 0.39145174622535706,
      "log_odds_ratio": -0.5761328935623169,
      "logits/chosen": 1.2879161834716797,
      "logits/rejected": 1.2894957065582275,
      "logps/chosen": -0.10628323256969452,
      "logps/rejected": -0.16520357131958008,
      "loss": 0.6824,
      "nll_loss": 0.10628323256969452,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.10628323256969452,
      "rewards/margins": 0.058920323848724365,
      "rewards/rejected": -0.16520357131958008,
      "step": 30
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 38.0,
      "learning_rate": 6.2399999999999995e-06,
      "log_odds_chosen": 0.9595848917961121,
      "log_odds_ratio": -0.38301774859428406,
      "logits/chosen": 1.2750675678253174,
      "logits/rejected": 1.2761198282241821,
      "logps/chosen": -0.08441177010536194,
      "logps/rejected": -0.1940467208623886,
      "loss": 0.4674,
      "nll_loss": 0.08441177010536194,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.08441177010536194,
      "rewards/margins": 0.10963495075702667,
      "rewards/rejected": -0.1940467208623886,
      "step": 40
    },
    {
      "epoch": 1.7857142857142856,
      "grad_norm": 47.0,
      "learning_rate": 7.84e-06,
      "log_odds_chosen": 1.2792123556137085,
      "log_odds_ratio": -0.32951468229293823,
      "logits/chosen": 1.1444017887115479,
      "logits/rejected": 1.146170973777771,
      "logps/chosen": -0.07942221313714981,
      "logps/rejected": -0.22475595772266388,
      "loss": 0.4089,
      "nll_loss": 0.07942221313714981,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.07942221313714981,
      "rewards/margins": 0.14533378183841705,
      "rewards/rejected": -0.22475595772266388,
      "step": 50
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 17.125,
      "learning_rate": 7.364596943186588e-06,
      "log_odds_chosen": 2.03440260887146,
      "log_odds_ratio": -0.17314136028289795,
      "logits/chosen": 0.9538688659667969,
      "logits/rejected": 0.9553494453430176,
      "logps/chosen": -0.05729224532842636,
      "logps/rejected": -0.3126146197319031,
      "loss": 0.2304,
      "nll_loss": 0.05729224532842636,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -0.05729224532842636,
      "rewards/margins": 0.2553223669528961,
      "rewards/rejected": -0.3126146197319031,
      "step": 60
    },
    {
      "epoch": 2.5,
      "grad_norm": 28.25,
      "learning_rate": 6.810052246069988e-06,
      "log_odds_chosen": 2.9335992336273193,
      "log_odds_ratio": -0.09323172271251678,
      "logits/chosen": 0.752956211566925,
      "logits/rejected": 0.7527356147766113,
      "logps/chosen": -0.04857013747096062,
      "logps/rejected": -0.47925204038619995,
      "loss": 0.1418,
      "nll_loss": 0.04857013374567032,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.04857013747096062,
      "rewards/margins": 0.43068188428878784,
      "rewards/rejected": -0.47925204038619995,
      "step": 70
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 19.5,
      "learning_rate": 6.36445827340584e-06,
      "log_odds_chosen": 2.9088222980499268,
      "log_odds_ratio": -0.10187868773937225,
      "logits/chosen": 0.6701191067695618,
      "logits/rejected": 0.6702790856361389,
      "logps/chosen": -0.048370588570833206,
      "logps/rejected": -0.4567483961582184,
      "loss": 0.1502,
      "nll_loss": 0.04837058484554291,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -0.048370588570833206,
      "rewards/margins": 0.4083777368068695,
      "rewards/rejected": -0.4567483961582184,
      "step": 80
    },
    {
      "epoch": 3.2142857142857144,
      "grad_norm": 20.625,
      "learning_rate": 5.996253511966892e-06,
      "log_odds_chosen": 3.2659313678741455,
      "log_odds_ratio": -0.06950727105140686,
      "logits/chosen": 0.6168217062950134,
      "logits/rejected": 0.6155665516853333,
      "logps/chosen": -0.03940538316965103,
      "logps/rejected": -0.46053066849708557,
      "loss": 0.1089,
      "nll_loss": 0.03940537944436073,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.03940538316965103,
      "rewards/margins": 0.42112523317337036,
      "rewards/rejected": -0.46053066849708557,
      "step": 90
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 11.5,
      "learning_rate": 5.685352436149611e-06,
      "log_odds_chosen": 3.775956630706787,
      "log_odds_ratio": -0.03668229654431343,
      "logits/chosen": 0.5860403776168823,
      "logits/rejected": 0.5861865282058716,
      "logps/chosen": -0.025243883952498436,
      "logps/rejected": -0.5562105178833008,
      "loss": 0.0619,
      "nll_loss": 0.025243883952498436,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.025243883952498436,
      "rewards/margins": 0.5309666395187378,
      "rewards/rejected": -0.5562105178833008,
      "step": 100
    },
    {
      "epoch": 3.571428571428571,
      "eval_log_odds_chosen": 2.1404664516448975,
      "eval_log_odds_ratio": -0.16489247977733612,
      "eval_logits/chosen": 0.6243001818656921,
      "eval_logits/rejected": 0.6249269247055054,
      "eval_logps/chosen": -0.07904133200645447,
      "eval_logps/rejected": -0.42098668217658997,
      "eval_loss": 0.2439338117837906,
      "eval_nll_loss": 0.07904132455587387,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": -0.07904133200645447,
      "eval_rewards/margins": 0.3419453203678131,
      "eval_rewards/rejected": -0.42098668217658997,
      "eval_runtime": 0.775,
      "eval_samples_per_second": 15.484,
      "eval_steps_per_second": 15.484,
      "step": 100
    },
    {
      "epoch": 3.928571428571429,
      "grad_norm": 31.125,
      "learning_rate": 5.418283691828771e-06,
      "log_odds_chosen": 3.764214038848877,
      "log_odds_ratio": -0.04073845222592354,
      "logits/chosen": 0.5791864991188049,
      "logits/rejected": 0.5785669088363647,
      "logps/chosen": -0.029079342260956764,
      "logps/rejected": -0.5950478315353394,
      "loss": 0.0698,
      "nll_loss": 0.029079344123601913,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.029079342260956764,
      "rewards/margins": 0.56596839427948,
      "rewards/rejected": -0.5950478315353394,
      "step": 110
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 9.0,
      "learning_rate": 5.1856297884173145e-06,
      "log_odds_chosen": 3.9393794536590576,
      "log_odds_ratio": -0.029240131378173828,
      "logits/chosen": 0.4942861497402191,
      "logits/rejected": 0.4927688539028168,
      "logps/chosen": -0.021664725616574287,
      "logps/rejected": -0.597990870475769,
      "loss": 0.0509,
      "nll_loss": 0.021664723753929138,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.021664725616574287,
      "rewards/margins": 0.5763261318206787,
      "rewards/rejected": -0.597990870475769,
      "step": 120
    },
    {
      "epoch": 4.642857142857143,
      "grad_norm": 21.75,
      "learning_rate": 4.980582450917523e-06,
      "log_odds_chosen": 4.069719314575195,
      "log_odds_ratio": -0.03151845932006836,
      "logits/chosen": 0.47455301880836487,
      "logits/rejected": 0.4722704291343689,
      "logps/chosen": -0.02563317120075226,
      "logps/rejected": -0.6308635473251343,
      "loss": 0.0572,
      "nll_loss": 0.02563317120075226,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.02563317120075226,
      "rewards/margins": 0.605230450630188,
      "rewards/rejected": -0.6308635473251343,
      "step": 130
    },
    {
      "epoch": 5.0,
      "grad_norm": 20.375,
      "learning_rate": 4.798081151232537e-06,
      "log_odds_chosen": 3.8989360332489014,
      "log_odds_ratio": -0.03524245321750641,
      "logits/chosen": 0.4603108763694763,
      "logits/rejected": 0.4602002501487732,
      "logps/chosen": -0.029795462265610695,
      "logps/rejected": -0.6790590286254883,
      "loss": 0.065,
      "nll_loss": 0.029795462265610695,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.029795462265610695,
      "rewards/margins": 0.6492635607719421,
      "rewards/rejected": -0.6790590286254883,
      "step": 140
    },
    {
      "epoch": 5.357142857142857,
      "grad_norm": 14.125,
      "learning_rate": 4.634275570907938e-06,
      "log_odds_chosen": 4.399620056152344,
      "log_odds_ratio": -0.020217038691043854,
      "logits/chosen": 0.46009063720703125,
      "logits/rejected": 0.4598237872123718,
      "logps/chosen": -0.01865455135703087,
      "logps/rejected": -0.6864362955093384,
      "loss": 0.0389,
      "nll_loss": 0.01865455135703087,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.01865455135703087,
      "rewards/margins": 0.6677817702293396,
      "rewards/rejected": -0.6864362955093384,
      "step": 150
    },
    {
      "epoch": 5.714285714285714,
      "grad_norm": 11.125,
      "learning_rate": 4.4861772327363545e-06,
      "log_odds_chosen": 4.303354740142822,
      "log_odds_ratio": -0.026399653404951096,
      "logits/chosen": 0.4165441393852234,
      "logits/rejected": 0.4147323966026306,
      "logps/chosen": -0.024661768227815628,
      "logps/rejected": -0.7016408443450928,
      "loss": 0.0511,
      "nll_loss": 0.024661768227815628,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.024661768227815628,
      "rewards/margins": 0.6769790649414062,
      "rewards/rejected": -0.7016408443450928,
      "step": 160
    },
    {
      "epoch": 6.071428571428571,
      "grad_norm": 10.0,
      "learning_rate": 4.3514263457633695e-06,
      "log_odds_chosen": 4.128013610839844,
      "log_odds_ratio": -0.029339054599404335,
      "logits/chosen": 0.42463621497154236,
      "logits/rejected": 0.4222441613674164,
      "logps/chosen": -0.024488601833581924,
      "logps/rejected": -0.6450936794281006,
      "loss": 0.0538,
      "nll_loss": 0.024488601833581924,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.024488601833581924,
      "rewards/margins": 0.6206051111221313,
      "rewards/rejected": -0.6450936794281006,
      "step": 170
    },
    {
      "epoch": 6.428571428571429,
      "grad_norm": 18.375,
      "learning_rate": 4.228131380653519e-06,
      "log_odds_chosen": 4.4497575759887695,
      "log_odds_ratio": -0.019353585317730904,
      "logits/chosen": 0.45868927240371704,
      "logits/rejected": 0.4576541483402252,
      "logps/chosen": -0.019276883453130722,
      "logps/rejected": -0.7127623558044434,
      "loss": 0.0386,
      "nll_loss": 0.019276883453130722,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.019276883453130722,
      "rewards/margins": 0.6934854388237,
      "rewards/rejected": -0.7127623558044434,
      "step": 180
    },
    {
      "epoch": 6.785714285714286,
      "grad_norm": 8.6875,
      "learning_rate": 4.1147559989891175e-06,
      "log_odds_chosen": 4.529023170471191,
      "log_odds_ratio": -0.01822759583592415,
      "logits/chosen": 0.4120427668094635,
      "logits/rejected": 0.40883350372314453,
      "logps/chosen": -0.016605500131845474,
      "logps/rejected": -0.7014130353927612,
      "loss": 0.0348,
      "nll_loss": 0.016605500131845474,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.016605500131845474,
      "rewards/margins": 0.6848075985908508,
      "rewards/rejected": -0.7014130353927612,
      "step": 190
    },
    {
      "epoch": 7.142857142857143,
      "grad_norm": 9.0625,
      "learning_rate": 4.010037656936684e-06,
      "log_odds_chosen": 4.466317176818848,
      "log_odds_ratio": -0.020451849326491356,
      "logits/chosen": 0.4540196359157562,
      "logits/rejected": 0.45342007279396057,
      "logps/chosen": -0.017516599968075752,
      "logps/rejected": -0.703050971031189,
      "loss": 0.038,
      "nll_loss": 0.017516599968075752,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.017516599968075752,
      "rewards/margins": 0.6855343580245972,
      "rewards/rejected": -0.703050971031189,
      "step": 200
    },
    {
      "epoch": 7.142857142857143,
      "eval_log_odds_chosen": 3.0318822860717773,
      "eval_log_odds_ratio": -0.0740598663687706,
      "eval_logits/chosen": 0.445597380399704,
      "eval_logits/rejected": 0.4458562433719635,
      "eval_logps/chosen": -0.05568574741482735,
      "eval_logps/rejected": -0.5331005454063416,
      "eval_loss": 0.12974560260772705,
      "eval_nll_loss": 0.05568574368953705,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": -0.05568574741482735,
      "eval_rewards/margins": 0.47741472721099854,
      "eval_rewards/rejected": -0.5331005454063416,
      "eval_runtime": 0.7718,
      "eval_samples_per_second": 15.548,
      "eval_steps_per_second": 15.548,
      "step": 200
    },
    {
      "epoch": 7.5,
      "grad_norm": 14.375,
      "learning_rate": 3.9129279043561475e-06,
      "log_odds_chosen": 4.380311965942383,
      "log_odds_ratio": -0.018879514187574387,
      "logits/chosen": 0.40403443574905396,
      "logits/rejected": 0.4024459719657898,
      "logps/chosen": -0.018743885681033134,
      "logps/rejected": -0.7524177432060242,
      "loss": 0.0376,
      "nll_loss": 0.018743883818387985,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.018743885681033134,
      "rewards/margins": 0.7336738705635071,
      "rewards/rejected": -0.7524177432060242,
      "step": 210
    },
    {
      "epoch": 7.857142857142857,
      "grad_norm": 14.75,
      "learning_rate": 3.822547860539972e-06,
      "log_odds_chosen": 4.6587018966674805,
      "log_odds_ratio": -0.016453776508569717,
      "logits/chosen": 0.39058271050453186,
      "logits/rejected": 0.38876551389694214,
      "logps/chosen": -0.017794523388147354,
      "logps/rejected": -0.6825803518295288,
      "loss": 0.0342,
      "nll_loss": 0.017794519662857056,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.017794523388147354,
      "rewards/margins": 0.6647858023643494,
      "rewards/rejected": -0.6825803518295288,
      "step": 220
    },
    {
      "epoch": 8.214285714285714,
      "grad_norm": 8.0625,
      "learning_rate": 3.738154508127982e-06,
      "log_odds_chosen": 4.862104415893555,
      "log_odds_ratio": -0.015109481289982796,
      "logits/chosen": 0.3413431644439697,
      "logits/rejected": 0.33844298124313354,
      "logps/chosen": -0.015027455985546112,
      "logps/rejected": -0.8296634554862976,
      "loss": 0.0301,
      "nll_loss": 0.015027457848191261,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.015027455985546112,
      "rewards/margins": 0.8146359324455261,
      "rewards/rejected": -0.8296634554862976,
      "step": 230
    },
    {
      "epoch": 8.571428571428571,
      "grad_norm": 14.6875,
      "learning_rate": 3.6591148299707855e-06,
      "log_odds_chosen": 4.68946647644043,
      "log_odds_ratio": -0.014953074976801872,
      "logits/chosen": 0.4132888913154602,
      "logits/rejected": 0.4114198088645935,
      "logps/chosen": -0.01745295524597168,
      "logps/rejected": -0.7407559752464294,
      "loss": 0.0324,
      "nll_loss": 0.01745295524597168,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.01745295524597168,
      "rewards/margins": 0.723302960395813,
      "rewards/rejected": -0.7407559752464294,
      "step": 240
    },
    {
      "epoch": 8.928571428571429,
      "grad_norm": 10.3125,
      "learning_rate": 3.584885719585766e-06,
      "log_odds_chosen": 4.460143566131592,
      "log_odds_ratio": -0.01899607665836811,
      "logits/chosen": 0.3702036440372467,
      "logits/rejected": 0.3682222366333008,
      "logps/chosen": -0.01728779450058937,
      "logps/rejected": -0.7273628115653992,
      "loss": 0.0363,
      "nll_loss": 0.01728779450058937,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.01728779450058937,
      "rewards/margins": 0.7100751399993896,
      "rewards/rejected": -0.7273628115653992,
      "step": 250
    },
    {
      "epoch": 9.285714285714286,
      "grad_norm": 7.84375,
      "learning_rate": 3.514998201309974e-06,
      "log_odds_chosen": 4.600634574890137,
      "log_odds_ratio": -0.01513788290321827,
      "logits/chosen": 0.38737642765045166,
      "logits/rejected": 0.3857085108757019,
      "logps/chosen": -0.015935596078634262,
      "logps/rejected": -0.7662953734397888,
      "loss": 0.0311,
      "nll_loss": 0.015935596078634262,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.015935596078634262,
      "rewards/margins": 0.7503597736358643,
      "rewards/rejected": -0.7662953734397888,
      "step": 260
    },
    {
      "epoch": 9.642857142857142,
      "grad_norm": 12.5,
      "learning_rate": 3.4490449084738148e-06,
      "log_odds_chosen": 4.695713996887207,
      "log_odds_ratio": -0.01749172806739807,
      "logits/chosen": 0.4047303795814514,
      "logits/rejected": 0.4035166800022125,
      "logps/chosen": -0.018705394119024277,
      "logps/rejected": -0.7980560064315796,
      "loss": 0.0362,
      "nll_loss": 0.018705394119024277,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.018705394119024277,
      "rewards/margins": 0.7793506383895874,
      "rewards/rejected": -0.7980560064315796,
      "step": 270
    },
    {
      "epoch": 10.0,
      "grad_norm": 3.3125,
      "learning_rate": 3.3866700533384128e-06,
      "log_odds_chosen": 4.833355903625488,
      "log_odds_ratio": -0.014932689256966114,
      "logits/chosen": 0.3466309607028961,
      "logits/rejected": 0.34371376037597656,
      "logps/chosen": -0.017113910987973213,
      "logps/rejected": -0.7418341040611267,
      "loss": 0.032,
      "nll_loss": 0.017113909125328064,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.017113910987973213,
      "rewards/margins": 0.7247201800346375,
      "rewards/rejected": -0.7418341040611267,
      "step": 280
    },
    {
      "epoch": 10.357142857142858,
      "grad_norm": 13.125,
      "learning_rate": 3.3275613232308115e-06,
      "log_odds_chosen": 4.978538513183594,
      "log_odds_ratio": -0.012340125627815723,
      "logits/chosen": 0.36500948667526245,
      "logits/rejected": 0.3617595136165619,
      "logps/chosen": -0.014922281727194786,
      "logps/rejected": -0.8379764556884766,
      "loss": 0.0273,
      "nll_loss": 0.014922281727194786,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.014922281727194786,
      "rewards/margins": 0.8230541348457336,
      "rewards/rejected": -0.8379764556884766,
      "step": 290
    },
    {
      "epoch": 10.714285714285714,
      "grad_norm": 8.625,
      "learning_rate": 3.2714432804381223e-06,
      "log_odds_chosen": 4.741899490356445,
      "log_odds_ratio": -0.01588159240782261,
      "logits/chosen": 0.3850441575050354,
      "logits/rejected": 0.38340434432029724,
      "logps/chosen": -0.016395803540945053,
      "logps/rejected": -0.7625649571418762,
      "loss": 0.0323,
      "nll_loss": 0.016395803540945053,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.016395803540945053,
      "rewards/margins": 0.7461690902709961,
      "rewards/rejected": -0.7625649571418762,
      "step": 300
    },
    {
      "epoch": 10.714285714285714,
      "eval_log_odds_chosen": 3.306226968765259,
      "eval_log_odds_ratio": -0.06318668276071548,
      "eval_logits/chosen": 0.3989677429199219,
      "eval_logits/rejected": 0.3996124565601349,
      "eval_logps/chosen": -0.05453824624419212,
      "eval_logps/rejected": -0.580333411693573,
      "eval_loss": 0.1177249327301979,
      "eval_nll_loss": 0.05453824624419212,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": -0.05453824624419212,
      "eval_rewards/margins": 0.5257951617240906,
      "eval_rewards/rejected": -0.580333411693573,
      "eval_runtime": 0.7728,
      "eval_samples_per_second": 15.527,
      "eval_steps_per_second": 15.527,
      "step": 300
    },
    {
      "epoch": 11.071428571428571,
      "grad_norm": 9.6875,
      "learning_rate": 3.2180719468365186e-06,
      "log_odds_chosen": 4.641325950622559,
      "log_odds_ratio": -0.014844359830021858,
      "logits/chosen": 0.3421401083469391,
      "logits/rejected": 0.34065181016921997,
      "logps/chosen": -0.015925418585538864,
      "logps/rejected": -0.7471505403518677,
      "loss": 0.0308,
      "nll_loss": 0.015925418585538864,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.015925418585538864,
      "rewards/margins": 0.7312251329421997,
      "rewards/rejected": -0.7471505403518677,
      "step": 310
    },
    {
      "epoch": 11.428571428571429,
      "grad_norm": 13.9375,
      "learning_rate": 3.167230329859583e-06,
      "log_odds_chosen": 4.906919956207275,
      "log_odds_ratio": -0.011877222917973995,
      "logits/chosen": 0.3373512625694275,
      "logits/rejected": 0.33693403005599976,
      "logps/chosen": -0.012469630688428879,
      "logps/rejected": -0.7439812421798706,
      "loss": 0.0243,
      "nll_loss": 0.012469631619751453,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.012469630688428879,
      "rewards/margins": 0.7315117120742798,
      "rewards/rejected": -0.7439812421798706,
      "step": 320
    },
    {
      "epoch": 11.785714285714286,
      "grad_norm": 22.375,
      "learning_rate": 3.118724702341802e-06,
      "log_odds_chosen": 4.819491863250732,
      "log_odds_ratio": -0.014139850623905659,
      "logits/chosen": 0.36182940006256104,
      "logits/rejected": 0.3593204915523529,
      "logps/chosen": -0.016692567616701126,
      "logps/rejected": -0.835942268371582,
      "loss": 0.0308,
      "nll_loss": 0.016692567616701126,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.016692567616701126,
      "rewards/margins": 0.8192497491836548,
      "rewards/rejected": -0.835942268371582,
      "step": 330
    },
    {
      "epoch": 12.142857142857142,
      "grad_norm": 17.625,
      "learning_rate": 3.072381490570274e-06,
      "log_odds_chosen": 4.758606910705566,
      "log_odds_ratio": -0.015324095264077187,
      "logits/chosen": 0.3854551613330841,
      "logits/rejected": 0.3818650245666504,
      "logps/chosen": -0.01669556275010109,
      "logps/rejected": -0.8267823457717896,
      "loss": 0.032,
      "nll_loss": 0.01669556275010109,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.01669556275010109,
      "rewards/margins": 0.8100866079330444,
      "rewards/rejected": -0.8267823457717896,
      "step": 340
    },
    {
      "epoch": 12.5,
      "grad_norm": 12.125,
      "learning_rate": 3.028044656417858e-06,
      "log_odds_chosen": 4.943177223205566,
      "log_odds_ratio": -0.011840974912047386,
      "logits/chosen": 0.3681556284427643,
      "logits/rejected": 0.364575058221817,
      "logps/chosen": -0.014707336202263832,
      "logps/rejected": -0.8508695363998413,
      "loss": 0.0265,
      "nll_loss": 0.014707337133586407,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.014707336202263832,
      "rewards/margins": 0.8361622095108032,
      "rewards/rejected": -0.8508695363998413,
      "step": 350
    },
    {
      "epoch": 12.857142857142858,
      "grad_norm": 14.375,
      "learning_rate": 2.9855734834483035e-06,
      "log_odds_chosen": 4.955853462219238,
      "log_odds_ratio": -0.013498658314347267,
      "logits/chosen": 0.35510769486427307,
      "logits/rejected": 0.3516218066215515,
      "logps/chosen": -0.014407852664589882,
      "logps/rejected": -0.7824307680130005,
      "loss": 0.0279,
      "nll_loss": 0.014407852664589882,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.014407852664589882,
      "rewards/margins": 0.768022894859314,
      "rewards/rejected": -0.7824307680130005,
      "step": 360
    },
    {
      "epoch": 13.214285714285714,
      "grad_norm": 7.71875,
      "learning_rate": 2.944840695329241e-06,
      "log_odds_chosen": 4.901689529418945,
      "log_odds_ratio": -0.013260012492537498,
      "logits/chosen": 0.33003872632980347,
      "logits/rejected": 0.33072271943092346,
      "logps/chosen": -0.014115576632320881,
      "logps/rejected": -0.8084686994552612,
      "loss": 0.0274,
      "nll_loss": 0.014115576632320881,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.014115576632320881,
      "rewards/margins": 0.794353187084198,
      "rewards/rejected": -0.8084686994552612,
      "step": 370
    },
    {
      "epoch": 13.571428571428571,
      "grad_norm": 15.125,
      "learning_rate": 2.9057308491685917e-06,
      "log_odds_chosen": 4.83064603805542,
      "log_odds_ratio": -0.012847358360886574,
      "logits/chosen": 0.3825533092021942,
      "logits/rejected": 0.3795173466205597,
      "logps/chosen": -0.014799078926444054,
      "logps/rejected": -0.8248156309127808,
      "loss": 0.0276,
      "nll_loss": 0.014799078926444054,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.014799078926444054,
      "rewards/margins": 0.8100166320800781,
      "rewards/rejected": -0.8248156309127808,
      "step": 380
    },
    {
      "epoch": 13.928571428571429,
      "grad_norm": 13.625,
      "learning_rate": 2.868138957528299e-06,
      "log_odds_chosen": 4.913525581359863,
      "log_odds_ratio": -0.0131983058527112,
      "logits/chosen": 0.31087830662727356,
      "logits/rejected": 0.30931466817855835,
      "logps/chosen": -0.01525194477289915,
      "logps/rejected": -0.8281568288803101,
      "loss": 0.0285,
      "nll_loss": 0.0152519466355443,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.01525194477289915,
      "rewards/margins": 0.8129048347473145,
      "rewards/rejected": -0.8281568288803101,
      "step": 390
    },
    {
      "epoch": 14.285714285714286,
      "grad_norm": 5.25,
      "learning_rate": 2.8319693016191545e-06,
      "log_odds_chosen": 5.078604698181152,
      "log_odds_ratio": -0.011706714518368244,
      "logits/chosen": 0.3178618550300598,
      "logits/rejected": 0.31361934542655945,
      "logps/chosen": -0.01467773038893938,
      "logps/rejected": -0.8909094929695129,
      "loss": 0.0264,
      "nll_loss": 0.01467773038893938,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.01467773038893938,
      "rewards/margins": 0.8762317895889282,
      "rewards/rejected": -0.8909094929695129,
      "step": 400
    },
    {
      "epoch": 14.285714285714286,
      "eval_log_odds_chosen": 3.4341936111450195,
      "eval_log_odds_ratio": -0.0597212128341198,
      "eval_logits/chosen": 0.3752058744430542,
      "eval_logits/rejected": 0.376191109418869,
      "eval_logps/chosen": -0.052630532532930374,
      "eval_logps/rejected": -0.6063037514686584,
      "eval_loss": 0.11235173791646957,
      "eval_nll_loss": 0.052630528807640076,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": -0.052630532532930374,
      "eval_rewards/margins": 0.5536733269691467,
      "eval_rewards/rejected": -0.6063037514686584,
      "eval_runtime": 0.7736,
      "eval_samples_per_second": 15.512,
      "eval_steps_per_second": 15.512,
      "step": 400
    },
    {
      "epoch": 14.285714285714286,
      "step": 400,
      "total_flos": 0.0,
      "train_loss": 0.1303568182513118,
      "train_runtime": 673.2061,
      "train_samples_per_second": 4.753,
      "train_steps_per_second": 0.594
    }
  ],
  "logging_steps": 10,
  "max_steps": 400,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 400,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}