{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9988571428571429,
  "eval_steps": 50,
  "global_step": 437,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.022857142857142857,
      "grad_norm": 4.698039299265843,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.654461622238159,
      "logits/rejected": -2.6349782943725586,
      "logps/chosen": -287.92938232421875,
      "logps/rejected": -266.479248046875,
      "loss": 0.6932,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": -0.00014368369011208415,
      "rewards/margins": -0.00023763992066960782,
      "rewards/rejected": 9.395615052198991e-05,
      "step": 10
    },
    {
      "epoch": 0.045714285714285714,
      "grad_norm": 4.697985911511002,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.655208110809326,
      "logits/rejected": -2.623675584793091,
      "logps/chosen": -271.43304443359375,
      "logps/rejected": -260.2268981933594,
      "loss": 0.6927,
      "rewards/accuracies": 0.581250011920929,
      "rewards/chosen": 0.002769916784018278,
      "rewards/margins": 0.0018681169021874666,
      "rewards/rejected": 0.0009017998236231506,
      "step": 20
    },
    {
      "epoch": 0.06857142857142857,
      "grad_norm": 4.400775078217756,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.6501145362854004,
      "logits/rejected": -2.5595479011535645,
      "logps/chosen": -271.99346923828125,
      "logps/rejected": -254.75973510742188,
      "loss": 0.6903,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.018953289836645126,
      "rewards/margins": 0.01143287681043148,
      "rewards/rejected": 0.007520412094891071,
      "step": 30
    },
    {
      "epoch": 0.09142857142857143,
      "grad_norm": 4.224410491085399,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.610691547393799,
      "logits/rejected": -2.596449375152588,
      "logps/chosen": -262.10076904296875,
      "logps/rejected": -255.63131713867188,
      "loss": 0.6838,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 0.03673754259943962,
      "rewards/margins": 0.03850514069199562,
      "rewards/rejected": -0.0017675980925559998,
      "step": 40
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 5.26619927871512,
      "learning_rate": 4.997124959943201e-07,
      "logits/chosen": -2.59323787689209,
      "logits/rejected": -2.5254805088043213,
      "logps/chosen": -288.6198425292969,
      "logps/rejected": -263.8785095214844,
      "loss": 0.6718,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.008304540999233723,
      "rewards/margins": 0.10466225445270538,
      "rewards/rejected": -0.09635771811008453,
      "step": 50
    },
    {
      "epoch": 0.11428571428571428,
      "eval_logits/chosen": -2.613680839538574,
      "eval_logits/rejected": -2.570056200027466,
      "eval_logps/chosen": -261.988525390625,
      "eval_logps/rejected": -261.6897277832031,
      "eval_loss": 0.6623334884643555,
      "eval_rewards/accuracies": 0.6594827771186829,
      "eval_rewards/chosen": -0.020272402092814445,
      "eval_rewards/margins": 0.11385296285152435,
      "eval_rewards/rejected": -0.13412536680698395,
      "eval_runtime": 94.9862,
      "eval_samples_per_second": 19.276,
      "eval_steps_per_second": 0.305,
      "step": 50
    },
    {
      "epoch": 0.13714285714285715,
      "grad_norm": 6.932576674963505,
      "learning_rate": 4.979579212164186e-07,
      "logits/chosen": -2.6151022911071777,
      "logits/rejected": -2.563739776611328,
      "logps/chosen": -303.95672607421875,
      "logps/rejected": -269.7010498046875,
      "loss": 0.6526,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.04470657557249069,
      "rewards/margins": 0.18772244453430176,
      "rewards/rejected": -0.23242902755737305,
      "step": 60
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.775877567674739,
      "learning_rate": 4.946196886175515e-07,
      "logits/chosen": -2.5894806385040283,
      "logits/rejected": -2.5019643306732178,
      "logps/chosen": -320.10382080078125,
      "logps/rejected": -296.3632507324219,
      "loss": 0.6419,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.1674882471561432,
      "rewards/margins": 0.30879640579223633,
      "rewards/rejected": -0.47628456354141235,
      "step": 70
    },
    {
      "epoch": 0.18285714285714286,
      "grad_norm": 8.903665160814306,
      "learning_rate": 4.897191188239667e-07,
      "logits/chosen": -2.577723979949951,
      "logits/rejected": -2.5037548542022705,
      "logps/chosen": -308.7467346191406,
      "logps/rejected": -314.86712646484375,
      "loss": 0.6373,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.41550832986831665,
      "rewards/margins": 0.2658213675022125,
      "rewards/rejected": -0.6813297271728516,
      "step": 80
    },
    {
      "epoch": 0.2057142857142857,
      "grad_norm": 13.187818767437374,
      "learning_rate": 4.832875107981763e-07,
      "logits/chosen": -1.7901182174682617,
      "logits/rejected": -1.6417770385742188,
      "logps/chosen": -292.3073425292969,
      "logps/rejected": -347.75286865234375,
      "loss": 0.6171,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.507135272026062,
      "rewards/margins": 0.4524330496788025,
      "rewards/rejected": -0.9595683813095093,
      "step": 90
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 18.455967644618873,
      "learning_rate": 4.753659419387223e-07,
      "logits/chosen": -1.0523039102554321,
      "logits/rejected": -0.7287885546684265,
      "logps/chosen": -313.5967712402344,
      "logps/rejected": -351.8410949707031,
      "loss": 0.5837,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.5234693884849548,
      "rewards/margins": 0.6886458396911621,
      "rewards/rejected": -1.2121151685714722,
      "step": 100
    },
    {
      "epoch": 0.22857142857142856,
      "eval_logits/chosen": -0.5771592259407043,
      "eval_logits/rejected": -0.20762209594249725,
      "eval_logps/chosen": -339.2017517089844,
      "eval_logps/rejected": -399.70703125,
      "eval_loss": 0.5854533314704895,
      "eval_rewards/accuracies": 0.6594827771186829,
      "eval_rewards/chosen": -0.7924044132232666,
      "eval_rewards/margins": 0.7218940258026123,
      "eval_rewards/rejected": -1.5142985582351685,
      "eval_runtime": 93.9256,
      "eval_samples_per_second": 19.494,
      "eval_steps_per_second": 0.309,
      "step": 100
    },
    {
      "epoch": 0.25142857142857145,
      "grad_norm": 20.061982050119,
      "learning_rate": 4.660050057270191e-07,
      "logits/chosen": -0.07258351147174835,
      "logits/rejected": 0.5934920310974121,
      "logps/chosen": -378.9754333496094,
      "logps/rejected": -401.3129577636719,
      "loss": 0.5614,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.7271420955657959,
      "rewards/margins": 0.9416759610176086,
      "rewards/rejected": -1.6688178777694702,
      "step": 110
    },
    {
      "epoch": 0.2742857142857143,
      "grad_norm": 19.936668067723307,
      "learning_rate": 4.5526448859687144e-07,
      "logits/chosen": 0.4706380367279053,
      "logits/rejected": 0.9462782740592957,
      "logps/chosen": -361.02105712890625,
      "logps/rejected": -419.98455810546875,
      "loss": 0.5824,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.8818947076797485,
      "rewards/margins": 0.8208815455436707,
      "rewards/rejected": -1.702776312828064,
      "step": 120
    },
    {
      "epoch": 0.29714285714285715,
      "grad_norm": 19.241668087031446,
      "learning_rate": 4.432129880904388e-07,
      "logits/chosen": 0.959698498249054,
      "logits/rejected": 1.6635558605194092,
      "logps/chosen": -376.22796630859375,
      "logps/rejected": -403.8036193847656,
      "loss": 0.5653,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": -1.0274622440338135,
      "rewards/margins": 0.6216489672660828,
      "rewards/rejected": -1.6491111516952515,
      "step": 130
    },
    {
      "epoch": 0.32,
      "grad_norm": 14.721763869214124,
      "learning_rate": 4.299274747394055e-07,
      "logits/chosen": 1.4800783395767212,
      "logits/rejected": 2.052093029022217,
      "logps/chosen": -443.85260009765625,
      "logps/rejected": -478.54522705078125,
      "loss": 0.5663,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.2601962089538574,
      "rewards/margins": 0.9224704504013062,
      "rewards/rejected": -2.182666540145874,
      "step": 140
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 18.13176648345513,
      "learning_rate": 4.1549280046953653e-07,
      "logits/chosen": 1.882054328918457,
      "logits/rejected": 2.4975225925445557,
      "logps/chosen": -417.7086486816406,
      "logps/rejected": -473.6971740722656,
      "loss": 0.5481,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -1.2165459394454956,
      "rewards/margins": 0.8646339178085327,
      "rewards/rejected": -2.0811798572540283,
      "step": 150
    },
    {
      "epoch": 0.34285714285714286,
      "eval_logits/chosen": 1.6485474109649658,
      "eval_logits/rejected": 2.4336249828338623,
      "eval_logps/chosen": -394.3633728027344,
      "eval_logps/rejected": -477.4252624511719,
      "eval_loss": 0.5498274564743042,
      "eval_rewards/accuracies": 0.693965494632721,
      "eval_rewards/chosen": -1.3440207242965698,
      "eval_rewards/margins": 0.9474601745605469,
      "eval_rewards/rejected": -2.291480779647827,
      "eval_runtime": 94.1545,
      "eval_samples_per_second": 19.447,
      "eval_steps_per_second": 0.308,
      "step": 150
    },
    {
      "epoch": 0.3657142857142857,
      "grad_norm": 25.73162601058156,
      "learning_rate": 4.000011566683401e-07,
      "logits/chosen": 0.9372028112411499,
      "logits/rejected": 2.038717746734619,
      "logps/chosen": -432.73846435546875,
      "logps/rejected": -488.79315185546875,
      "loss": 0.5381,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.2358500957489014,
      "rewards/margins": 0.9547734260559082,
      "rewards/rejected": -2.1906232833862305,
      "step": 160
    },
    {
      "epoch": 0.38857142857142857,
      "grad_norm": 22.65560343905013,
      "learning_rate": 3.8355148537705047e-07,
      "logits/chosen": 1.915405511856079,
      "logits/rejected": 2.607452869415283,
      "logps/chosen": -435.5879821777344,
      "logps/rejected": -534.2739868164062,
      "loss": 0.5443,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.4189122915267944,
      "rewards/margins": 1.1850039958953857,
      "rewards/rejected": -2.6039161682128906,
      "step": 170
    },
    {
      "epoch": 0.4114285714285714,
      "grad_norm": 23.399452190018653,
      "learning_rate": 3.662488473675315e-07,
      "logits/chosen": 2.181429624557495,
      "logits/rejected": 3.391533374786377,
      "logps/chosen": -434.38287353515625,
      "logps/rejected": -499.68768310546875,
      "loss": 0.5411,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -1.3194260597229004,
      "rewards/margins": 1.2128746509552002,
      "rewards/rejected": -2.5323009490966797,
      "step": 180
    },
    {
      "epoch": 0.4342857142857143,
      "grad_norm": 20.359816434441168,
      "learning_rate": 3.48203751140067e-07,
      "logits/chosen": 1.7659626007080078,
      "logits/rejected": 3.134269952774048,
      "logps/chosen": -422.34130859375,
      "logps/rejected": -507.91058349609375,
      "loss": 0.5419,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -1.6150566339492798,
      "rewards/margins": 1.1147284507751465,
      "rewards/rejected": -2.729785203933716,
      "step": 190
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 19.72704748785004,
      "learning_rate": 3.2953144712759537e-07,
      "logits/chosen": 2.624401569366455,
      "logits/rejected": 3.5372815132141113,
      "logps/chosen": -405.11773681640625,
      "logps/rejected": -503.3150939941406,
      "loss": 0.5289,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -1.5478131771087646,
      "rewards/margins": 1.097571611404419,
      "rewards/rejected": -2.6453845500946045,
      "step": 200
    },
    {
      "epoch": 0.45714285714285713,
      "eval_logits/chosen": 2.6372268199920654,
      "eval_logits/rejected": 3.5667386054992676,
      "eval_logps/chosen": -424.2229919433594,
      "eval_logps/rejected": -523.725341796875,
      "eval_loss": 0.5408812165260315,
      "eval_rewards/accuracies": 0.6982758641242981,
      "eval_rewards/chosen": -1.642616629600525,
      "eval_rewards/margins": 1.111864686012268,
      "eval_rewards/rejected": -2.754481554031372,
      "eval_runtime": 94.7878,
      "eval_samples_per_second": 19.317,
      "eval_steps_per_second": 0.306,
      "step": 200
    },
    {
      "epoch": 0.48,
      "grad_norm": 22.170086283776122,
      "learning_rate": 3.103511916141658e-07,
      "logits/chosen": 2.3618617057800293,
      "logits/rejected": 3.551234722137451,
      "logps/chosen": -444.750244140625,
      "logps/rejected": -544.8475341796875,
      "loss": 0.5315,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.5890655517578125,
      "rewards/margins": 1.23880934715271,
      "rewards/rejected": -2.8278753757476807,
      "step": 210
    },
    {
      "epoch": 0.5028571428571429,
      "grad_norm": 20.88239892734878,
      "learning_rate": 2.9078548506882117e-07,
      "logits/chosen": 2.4734537601470947,
      "logits/rejected": 3.3690295219421387,
      "logps/chosen": -479.52294921875,
      "logps/rejected": -594.2535400390625,
      "loss": 0.542,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.751405119895935,
      "rewards/margins": 1.2897474765777588,
      "rewards/rejected": -3.0411527156829834,
      "step": 220
    },
    {
      "epoch": 0.5257142857142857,
      "grad_norm": 17.747062332553714,
      "learning_rate": 2.709592897595191e-07,
      "logits/chosen": 1.912064790725708,
      "logits/rejected": 2.7421936988830566,
      "logps/chosen": -438.20208740234375,
      "logps/rejected": -520.5653076171875,
      "loss": 0.5653,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -1.534075140953064,
      "rewards/margins": 0.8896983861923218,
      "rewards/rejected": -2.4237735271453857,
      "step": 230
    },
    {
      "epoch": 0.5485714285714286,
      "grad_norm": 18.030085534322904,
      "learning_rate": 2.509992316440332e-07,
      "logits/chosen": 2.9936442375183105,
      "logits/rejected": 4.284924507141113,
      "logps/chosen": -503.80523681640625,
      "logps/rejected": -576.6605834960938,
      "loss": 0.5468,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -1.9611694812774658,
      "rewards/margins": 1.1891307830810547,
      "rewards/rejected": -3.1503000259399414,
      "step": 240
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 24.141080092797733,
      "learning_rate": 2.3103279163519918e-07,
      "logits/chosen": 2.715038299560547,
      "logits/rejected": 3.2168846130371094,
      "logps/chosen": -476.09295654296875,
      "logps/rejected": -553.4679565429688,
      "loss": 0.54,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -1.9882991313934326,
      "rewards/margins": 0.7364269495010376,
      "rewards/rejected": -2.7247262001037598,
      "step": 250
    },
    {
      "epoch": 0.5714285714285714,
      "eval_logits/chosen": 1.416563630104065,
      "eval_logits/rejected": 2.6436657905578613,
      "eval_logps/chosen": -413.8667297363281,
      "eval_logps/rejected": -518.8563232421875,
      "eval_loss": 0.5280377268791199,
      "eval_rewards/accuracies": 0.7025862336158752,
      "eval_rewards/chosen": -1.5390541553497314,
      "eval_rewards/margins": 1.1667375564575195,
      "eval_rewards/rejected": -2.70579195022583,
      "eval_runtime": 93.2943,
      "eval_samples_per_second": 19.626,
      "eval_steps_per_second": 0.311,
      "step": 250
    },
    {
      "epoch": 0.5942857142857143,
      "grad_norm": 19.312331708156304,
      "learning_rate": 2.1118749140573358e-07,
      "logits/chosen": 1.422081708908081,
      "logits/rejected": 2.8245503902435303,
      "logps/chosen": -426.8099670410156,
      "logps/rejected": -459.8812561035156,
      "loss": 0.5455,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.5098531246185303,
      "rewards/margins": 0.8642712831497192,
      "rewards/rejected": -2.374124526977539,
      "step": 260
    },
    {
      "epoch": 0.6171428571428571,
      "grad_norm": 17.556239848475172,
      "learning_rate": 1.9159007893272703e-07,
      "logits/chosen": 1.7777824401855469,
      "logits/rejected": 2.890540599822998,
      "logps/chosen": -466.5252990722656,
      "logps/rejected": -499.5956115722656,
      "loss": 0.5345,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": -1.783983588218689,
      "rewards/margins": 0.7515356540679932,
      "rewards/rejected": -2.5355191230773926,
      "step": 270
    },
    {
      "epoch": 0.64,
      "grad_norm": 16.478054364358847,
      "learning_rate": 1.7236571898357766e-07,
      "logits/chosen": 2.3310110569000244,
      "logits/rejected": 3.422163486480713,
      "logps/chosen": -419.81512451171875,
      "logps/rejected": -533.2318115234375,
      "loss": 0.5402,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -1.7008750438690186,
      "rewards/margins": 1.2960573434829712,
      "rewards/rejected": -2.9969325065612793,
      "step": 280
    },
    {
      "epoch": 0.6628571428571428,
      "grad_norm": 21.127488932057656,
      "learning_rate": 1.5363719371356882e-07,
      "logits/chosen": 2.3282363414764404,
      "logits/rejected": 3.5149283409118652,
      "logps/chosen": -463.74200439453125,
      "logps/rejected": -535.0039672851562,
      "loss": 0.5179,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.7781498432159424,
      "rewards/margins": 1.0403541326522827,
      "rewards/rejected": -2.8185038566589355,
      "step": 290
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 21.3761678454725,
      "learning_rate": 1.3552411848071565e-07,
      "logits/chosen": 2.3439698219299316,
      "logits/rejected": 3.820255994796753,
      "logps/chosen": -497.107666015625,
      "logps/rejected": -575.4114990234375,
      "loss": 0.5147,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.8375751972198486,
      "rewards/margins": 1.2724038362503052,
      "rewards/rejected": -3.1099791526794434,
      "step": 300
    },
    {
      "epoch": 0.6857142857142857,
      "eval_logits/chosen": 2.696895122528076,
      "eval_logits/rejected": 3.777069330215454,
      "eval_logps/chosen": -444.8299560546875,
      "eval_logps/rejected": -558.1808471679688,
      "eval_loss": 0.5203906297683716,
      "eval_rewards/accuracies": 0.7112069129943848,
      "eval_rewards/chosen": -1.8486860990524292,
      "eval_rewards/margins": 1.2503504753112793,
      "eval_rewards/rejected": -3.099036455154419,
      "eval_runtime": 93.3957,
      "eval_samples_per_second": 19.605,
      "eval_steps_per_second": 0.311,
      "step": 300
    },
    {
      "epoch": 0.7085714285714285,
      "grad_norm": 21.989201062769567,
      "learning_rate": 1.1814217788631473e-07,
      "logits/chosen": 2.4984259605407715,
      "logits/rejected": 3.986057996749878,
      "logps/chosen": -490.80255126953125,
      "logps/rejected": -586.6756591796875,
      "loss": 0.5285,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -1.775442361831665,
      "rewards/margins": 1.4152400493621826,
      "rewards/rejected": -3.1906826496124268,
      "step": 310
    },
    {
      "epoch": 0.7314285714285714,
      "grad_norm": 22.181868921213347,
      "learning_rate": 1.0160238692045331e-07,
      "logits/chosen": 2.8167648315429688,
      "logits/rejected": 4.0772247314453125,
      "logps/chosen": -484.88201904296875,
      "logps/rejected": -577.0535888671875,
      "loss": 0.533,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -1.8776414394378662,
      "rewards/margins": 1.1431219577789307,
      "rewards/rejected": -3.0207631587982178,
      "step": 320
    },
    {
      "epoch": 0.7542857142857143,
      "grad_norm": 21.423699121141595,
      "learning_rate": 8.601038193139438e-08,
      "logits/chosen": 2.9268698692321777,
      "logits/rejected": 3.6747875213623047,
      "logps/chosen": -416.2300720214844,
      "logps/rejected": -506.71966552734375,
      "loss": 0.5385,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -1.7054163217544556,
      "rewards/margins": 0.9577444195747375,
      "rewards/rejected": -2.663160800933838,
      "step": 330
    },
    {
      "epoch": 0.7771428571428571,
      "grad_norm": 22.736161602909004,
      "learning_rate": 7.146574594727572e-08,
      "logits/chosen": 2.4303653240203857,
      "logits/rejected": 3.7860004901885986,
      "logps/chosen": -426.480224609375,
      "logps/rejected": -512.9637451171875,
      "loss": 0.5184,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -1.8047637939453125,
      "rewards/margins": 1.1890126466751099,
      "rewards/rejected": -2.9937758445739746,
      "step": 340
    },
    {
      "epoch": 0.8,
      "grad_norm": 23.477309940264664,
      "learning_rate": 5.8061372659157306e-08,
      "logits/chosen": 2.426809310913086,
      "logits/rejected": 4.1134934425354,
      "logps/chosen": -493.06414794921875,
      "logps/rejected": -577.4049072265625,
      "loss": 0.5033,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -1.8289105892181396,
      "rewards/margins": 1.2613272666931152,
      "rewards/rejected": -3.090238094329834,
      "step": 350
    },
    {
      "epoch": 0.8,
      "eval_logits/chosen": 2.7948429584503174,
      "eval_logits/rejected": 3.863651990890503,
      "eval_logps/chosen": -447.5259094238281,
      "eval_logps/rejected": -562.8871459960938,
      "eval_loss": 0.5168038606643677,
      "eval_rewards/accuracies": 0.7413793206214905,
      "eval_rewards/chosen": -1.8756457567214966,
      "eval_rewards/margins": 1.2704533338546753,
      "eval_rewards/rejected": -3.146099328994751,
      "eval_runtime": 92.678,
      "eval_samples_per_second": 19.757,
      "eval_steps_per_second": 0.313,
      "step": 350
    },
    {
      "epoch": 0.8228571428571428,
      "grad_norm": 22.089570552576745,
      "learning_rate": 4.5882873127531614e-08,
      "logits/chosen": 3.1440396308898926,
      "logits/rejected": 4.210689067840576,
      "logps/chosen": -474.7491760253906,
      "logps/rejected": -611.8050537109375,
      "loss": 0.5176,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -1.8223841190338135,
      "rewards/margins": 1.6220409870147705,
      "rewards/rejected": -3.444424867630005,
      "step": 360
    },
    {
      "epoch": 0.8457142857142858,
      "grad_norm": 26.5380528583338,
      "learning_rate": 3.500802900154412e-08,
      "logits/chosen": 3.0336647033691406,
      "logits/rejected": 3.8006350994110107,
      "logps/chosen": -432.41278076171875,
      "logps/rejected": -567.8787231445312,
      "loss": 0.5204,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -1.7640663385391235,
      "rewards/margins": 1.2935981750488281,
      "rewards/rejected": -3.057664394378662,
      "step": 370
    },
    {
      "epoch": 0.8685714285714285,
      "grad_norm": 24.05648133941441,
      "learning_rate": 2.550629574310309e-08,
      "logits/chosen": 3.677473783493042,
      "logits/rejected": 4.557853698730469,
      "logps/chosen": -449.2403259277344,
      "logps/rejected": -553.65966796875,
      "loss": 0.5084,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -1.9337879419326782,
      "rewards/margins": 1.0252636671066284,
      "rewards/rejected": -2.9590516090393066,
      "step": 380
    },
    {
      "epoch": 0.8914285714285715,
      "grad_norm": 18.50726505627672,
      "learning_rate": 1.7438359028687983e-08,
      "logits/chosen": 3.2151970863342285,
      "logits/rejected": 4.2298479080200195,
      "logps/chosen": -456.96112060546875,
      "logps/rejected": -567.398681640625,
      "loss": 0.519,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -1.903903603553772,
      "rewards/margins": 1.2563444375991821,
      "rewards/rejected": -3.160248041152954,
      "step": 390
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 23.71785447914187,
      "learning_rate": 1.0855747162029361e-08,
      "logits/chosen": 2.804896831512451,
      "logits/rejected": 3.584916353225708,
      "logps/chosen": -462.6809997558594,
      "logps/rejected": -580.8098754882812,
      "loss": 0.52,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.885479211807251,
      "rewards/margins": 1.2495968341827393,
      "rewards/rejected": -3.1350760459899902,
      "step": 400
    },
    {
      "epoch": 0.9142857142857143,
      "eval_logits/chosen": 3.059108257293701,
      "eval_logits/rejected": 4.039414882659912,
      "eval_logps/chosen": -455.7142333984375,
      "eval_logps/rejected": -573.1553344726562,
      "eval_loss": 0.5141626000404358,
      "eval_rewards/accuracies": 0.7198275923728943,
      "eval_rewards/chosen": -1.9575295448303223,
      "eval_rewards/margins": 1.2912517786026,
      "eval_rewards/rejected": -3.248781442642212,
      "eval_runtime": 94.0826,
      "eval_samples_per_second": 19.462,
      "eval_steps_per_second": 0.308,
      "step": 400
    },
    {
      "epoch": 0.9371428571428572,
      "grad_norm": 24.860041582436438,
      "learning_rate": 5.8005019731033615e-09,
      "logits/chosen": 3.145111560821533,
      "logits/rejected": 4.010159492492676,
      "logps/chosen": -449.4214782714844,
      "logps/rejected": -553.9349365234375,
      "loss": 0.5035,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -1.891174077987671,
      "rewards/margins": 1.0715255737304688,
      "rewards/rejected": -2.9626996517181396,
      "step": 410
    },
    {
      "epoch": 0.96,
      "grad_norm": 20.06298675421982,
      "learning_rate": 2.3049103053431886e-09,
      "logits/chosen": 2.699260950088501,
      "logits/rejected": 3.91593861579895,
      "logps/chosen": -503.0846252441406,
      "logps/rejected": -636.1543579101562,
      "loss": 0.524,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.9241759777069092,
      "rewards/margins": 1.4948225021362305,
      "rewards/rejected": -3.4189987182617188,
      "step": 420
    },
    {
      "epoch": 0.9828571428571429,
      "grad_norm": 18.461900399401518,
      "learning_rate": 3.9129780600541397e-10,
      "logits/chosen": 3.151226043701172,
      "logits/rejected": 4.009431838989258,
      "logps/chosen": -474.90179443359375,
      "logps/rejected": -583.8973388671875,
      "loss": 0.4974,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -1.8930785655975342,
      "rewards/margins": 1.3667445182800293,
      "rewards/rejected": -3.2598233222961426,
      "step": 430
    },
    {
      "epoch": 0.9988571428571429,
      "step": 437,
      "total_flos": 0.0,
      "train_loss": 0.5629221896557841,
      "train_runtime": 11730.276,
      "train_samples_per_second": 4.774,
      "train_steps_per_second": 0.037
    }
  ],
  "logging_steps": 10,
  "max_steps": 437,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}