{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 50,
  "global_step": 201,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 57.12193824365218,
      "learning_rate": 5e-07,
      "logits/chosen": -2.716521978378296,
      "logits/rejected": -2.7164063453674316,
      "logps/chosen": -263.66815185546875,
      "logps/rejected": -226.99600219726562,
      "loss": 0.6899,
      "rewards/accuracies": 0.3499999940395355,
      "rewards/chosen": 0.014930379576981068,
      "rewards/margins": 0.005484213586896658,
      "rewards/rejected": 0.009446167387068272,
      "step": 5
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 50.98436154009925,
      "learning_rate": 1e-06,
      "logits/chosen": -2.659832239151001,
      "logits/rejected": -2.6706981658935547,
      "logps/chosen": -240.44741821289062,
      "logps/rejected": -202.6971893310547,
      "loss": 0.6353,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": 0.4928322732448578,
      "rewards/margins": 0.11446012556552887,
      "rewards/rejected": 0.3783721327781677,
      "step": 10
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 49.5897720682669,
      "learning_rate": 9.983100718730718e-07,
      "logits/chosen": -2.469956874847412,
      "logits/rejected": -2.457557201385498,
      "logps/chosen": -224.4867706298828,
      "logps/rejected": -186.89651489257812,
      "loss": 0.6504,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 1.3739125728607178,
      "rewards/margins": 0.8372632265090942,
      "rewards/rejected": 0.5366495847702026,
      "step": 15
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 38.57601617352061,
      "learning_rate": 9.932517109205849e-07,
      "logits/chosen": -2.329310894012451,
      "logits/rejected": -2.304492235183716,
      "logps/chosen": -216.98519897460938,
      "logps/rejected": -203.6713104248047,
      "loss": 0.6103,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 1.4170403480529785,
      "rewards/margins": 1.1797821521759033,
      "rewards/rejected": 0.2372581958770752,
      "step": 20
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 41.87199823484069,
      "learning_rate": 9.848591102083375e-07,
      "logits/chosen": -2.1940865516662598,
      "logits/rejected": -2.1780757904052734,
      "logps/chosen": -250.27261352539062,
      "logps/rejected": -203.70118713378906,
      "loss": 0.5969,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 1.575352430343628,
      "rewards/margins": 1.1331461668014526,
      "rewards/rejected": 0.44220608472824097,
      "step": 25
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 34.36864819717641,
      "learning_rate": 9.731890013043367e-07,
      "logits/chosen": -2.177009105682373,
      "logits/rejected": -2.1458096504211426,
      "logps/chosen": -257.85015869140625,
      "logps/rejected": -217.74462890625,
      "loss": 0.5814,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 1.5730830430984497,
      "rewards/margins": 1.5190694332122803,
      "rewards/rejected": 0.054013729095458984,
      "step": 30
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 40.73655696053411,
      "learning_rate": 9.583202707897073e-07,
      "logits/chosen": -2.2473671436309814,
      "logits/rejected": -2.1966567039489746,
      "logps/chosen": -247.3220672607422,
      "logps/rejected": -202.4593963623047,
      "loss": 0.5595,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.881485104560852,
      "rewards/margins": 1.353430986404419,
      "rewards/rejected": -0.47194600105285645,
      "step": 35
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 41.80314492549501,
      "learning_rate": 9.403534270080829e-07,
      "logits/chosen": -2.27215838432312,
      "logits/rejected": -2.2764623165130615,
      "logps/chosen": -231.4678497314453,
      "logps/rejected": -203.9759521484375,
      "loss": 0.578,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 0.9200285077095032,
      "rewards/margins": 1.0039114952087402,
      "rewards/rejected": -0.0838831439614296,
      "step": 40
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 38.467845430667744,
      "learning_rate": 9.19409920658098e-07,
      "logits/chosen": -2.338601589202881,
      "logits/rejected": -2.3325324058532715,
      "logps/chosen": -247.09738159179688,
      "logps/rejected": -206.18026733398438,
      "loss": 0.5433,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": 1.363398551940918,
      "rewards/margins": 1.4603192806243896,
      "rewards/rejected": -0.09692087769508362,
      "step": 45
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 37.05742903664642,
      "learning_rate": 8.956313238215823e-07,
      "logits/chosen": -2.4232733249664307,
      "logits/rejected": -2.413396120071411,
      "logps/chosen": -252.5549774169922,
      "logps/rejected": -190.77883911132812,
      "loss": 0.6027,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 1.3205119371414185,
      "rewards/margins": 1.3863776922225952,
      "rewards/rejected": -0.06586578488349915,
      "step": 50
    },
    {
      "epoch": 0.746268656716418,
      "eval_logits/chosen": -2.457545757293701,
      "eval_logits/rejected": -2.435551404953003,
      "eval_logps/chosen": -259.58868408203125,
      "eval_logps/rejected": -201.051025390625,
      "eval_loss": 0.5490748882293701,
      "eval_rewards/accuracies": 0.8208333253860474,
      "eval_rewards/chosen": 1.3802965879440308,
      "eval_rewards/margins": 1.6785677671432495,
      "eval_rewards/rejected": -0.29827114939689636,
      "eval_runtime": 125.9419,
      "eval_samples_per_second": 15.086,
      "eval_steps_per_second": 0.238,
      "step": 50
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 33.60802656137982,
      "learning_rate": 8.691783729769873e-07,
      "logits/chosen": -2.4649384021759033,
      "logits/rejected": -2.4393675327301025,
      "logps/chosen": -247.6337432861328,
      "logps/rejected": -225.37429809570312,
      "loss": 0.513,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 1.384879469871521,
      "rewards/margins": 1.5727264881134033,
      "rewards/rejected": -0.1878468245267868,
      "step": 55
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 33.47722052756408,
      "learning_rate": 8.402298824670029e-07,
      "logits/chosen": -2.4343879222869873,
      "logits/rejected": -2.429543972015381,
      "logps/chosen": -243.7395782470703,
      "logps/rejected": -216.7984161376953,
      "loss": 0.5325,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": 1.061714768409729,
      "rewards/margins": 1.736096739768982,
      "rewards/rejected": -0.6743819713592529,
      "step": 60
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 30.472130365224405,
      "learning_rate": 8.089815357650089e-07,
      "logits/chosen": -2.3370931148529053,
      "logits/rejected": -2.2993381023406982,
      "logps/chosen": -250.01687622070312,
      "logps/rejected": -205.3022003173828,
      "loss": 0.5415,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": 1.2970752716064453,
      "rewards/margins": 1.5023232698440552,
      "rewards/rejected": -0.2052478790283203,
      "step": 65
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 19.840952878286068,
      "learning_rate": 7.756445627110522e-07,
      "logits/chosen": -2.2309017181396484,
      "logits/rejected": -2.2050414085388184,
      "logps/chosen": -237.7880096435547,
      "logps/rejected": -231.38265991210938,
      "loss": 0.3696,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": 1.7277215719223022,
      "rewards/margins": 2.358109474182129,
      "rewards/rejected": -0.6303879618644714,
      "step": 70
    },
    {
      "epoch": 1.1194029850746268,
      "grad_norm": 21.240934428524795,
      "learning_rate": 7.404443116588547e-07,
      "logits/chosen": -2.2093091011047363,
      "logits/rejected": -2.1994729042053223,
      "logps/chosen": -260.18426513671875,
      "logps/rejected": -221.51742553710938,
      "loss": 0.2473,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.7431342601776123,
      "rewards/margins": 2.91957950592041,
      "rewards/rejected": -1.1764450073242188,
      "step": 75
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 19.68722107584437,
      "learning_rate": 7.036187261857288e-07,
      "logits/chosen": -2.1707935333251953,
      "logits/rejected": -2.151038408279419,
      "logps/chosen": -219.67929077148438,
      "logps/rejected": -219.01358032226562,
      "loss": 0.2201,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.7701470851898193,
      "rewards/margins": 3.296602964401245,
      "rewards/rejected": -1.5264555215835571,
      "step": 80
    },
    {
      "epoch": 1.2686567164179103,
      "grad_norm": 19.753765139786655,
      "learning_rate": 6.654167366624008e-07,
      "logits/chosen": -2.137498140335083,
      "logits/rejected": -2.1242105960845947,
      "logps/chosen": -235.6524658203125,
      "logps/rejected": -222.9441680908203,
      "loss": 0.2428,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 1.8704173564910889,
      "rewards/margins": 3.3603546619415283,
      "rewards/rejected": -1.4899370670318604,
      "step": 85
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 18.35032830894322,
      "learning_rate": 6.260965775552713e-07,
      "logits/chosen": -2.0957112312316895,
      "logits/rejected": -2.025146961212158,
      "logps/chosen": -224.27377319335938,
      "logps/rejected": -204.6040496826172,
      "loss": 0.2536,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.086984395980835,
      "rewards/margins": 3.093162775039673,
      "rewards/rejected": -1.0061782598495483,
      "step": 90
    },
    {
      "epoch": 1.417910447761194,
      "grad_norm": 22.277576113780217,
      "learning_rate": 5.859240418356614e-07,
      "logits/chosen": -2.0116684436798096,
      "logits/rejected": -1.9373754262924194,
      "logps/chosen": -225.5041046142578,
      "logps/rejected": -224.0210723876953,
      "loss": 0.2394,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 1.8472282886505127,
      "rewards/margins": 3.6459250450134277,
      "rewards/rejected": -1.7986968755722046,
      "step": 95
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 21.864051938399246,
      "learning_rate": 5.451706842957421e-07,
      "logits/chosen": -1.985772728919983,
      "logits/rejected": -1.9631462097167969,
      "logps/chosen": -228.11532592773438,
      "logps/rejected": -222.6101837158203,
      "loss": 0.2795,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 1.6982160806655884,
      "rewards/margins": 3.457064151763916,
      "rewards/rejected": -1.758847951889038,
      "step": 100
    },
    {
      "epoch": 1.4925373134328357,
      "eval_logits/chosen": -1.975551724433899,
      "eval_logits/rejected": -1.910164713859558,
      "eval_logps/chosen": -261.8016357421875,
      "eval_logps/rejected": -213.1614227294922,
      "eval_loss": 0.5111880302429199,
      "eval_rewards/accuracies": 0.8416666388511658,
      "eval_rewards/chosen": 1.159003496170044,
      "eval_rewards/margins": 2.6683123111724854,
      "eval_rewards/rejected": -1.509308934211731,
      "eval_runtime": 126.7081,
      "eval_samples_per_second": 14.995,
      "eval_steps_per_second": 0.237,
      "step": 100
    },
    {
      "epoch": 1.5671641791044775,
      "grad_norm": 24.045736817846212,
      "learning_rate": 5.041119859162068e-07,
      "logits/chosen": -1.96713387966156,
      "logits/rejected": -1.9250026941299438,
      "logps/chosen": -253.3628387451172,
      "logps/rejected": -219.9677276611328,
      "loss": 0.2536,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 1.8924213647842407,
      "rewards/margins": 3.974386692047119,
      "rewards/rejected": -2.081965446472168,
      "step": 105
    },
    {
      "epoch": 1.6417910447761193,
      "grad_norm": 21.61477181672177,
      "learning_rate": 4.630254916940423e-07,
      "logits/chosen": -1.9149868488311768,
      "logits/rejected": -1.8783435821533203,
      "logps/chosen": -229.2848663330078,
      "logps/rejected": -201.11834716796875,
      "loss": 0.2682,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.406848669052124,
      "rewards/margins": 3.400202989578247,
      "rewards/rejected": -1.9933545589447021,
      "step": 110
    },
    {
      "epoch": 1.716417910447761,
      "grad_norm": 21.07628352255977,
      "learning_rate": 4.2218893451814e-07,
      "logits/chosen": -1.9863373041152954,
      "logits/rejected": -1.9536094665527344,
      "logps/chosen": -241.6915740966797,
      "logps/rejected": -234.1923828125,
      "loss": 0.2425,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.814729928970337,
      "rewards/margins": 3.653728485107422,
      "rewards/rejected": -1.8389990329742432,
      "step": 115
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 20.726924515407703,
      "learning_rate": 3.8187835777481375e-07,
      "logits/chosen": -2.0750255584716797,
      "logits/rejected": -2.0614113807678223,
      "logps/chosen": -259.0155029296875,
      "logps/rejected": -212.9644012451172,
      "loss": 0.2864,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.9931504726409912,
      "rewards/margins": 3.505798816680908,
      "rewards/rejected": -1.512648582458496,
      "step": 120
    },
    {
      "epoch": 1.8656716417910446,
      "grad_norm": 20.134285007421816,
      "learning_rate": 3.423662493738687e-07,
      "logits/chosen": -2.130033493041992,
      "logits/rejected": -2.0896992683410645,
      "logps/chosen": -227.1628875732422,
      "logps/rejected": -232.8570556640625,
      "loss": 0.2575,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 1.4646759033203125,
      "rewards/margins": 3.994136095046997,
      "rewards/rejected": -2.5294601917266846,
      "step": 125
    },
    {
      "epoch": 1.9402985074626866,
      "grad_norm": 18.272474363996036,
      "learning_rate": 3.039196998086687e-07,
      "logits/chosen": -2.158094882965088,
      "logits/rejected": -2.131124973297119,
      "logps/chosen": -219.0237274169922,
      "logps/rejected": -218.6309051513672,
      "loss": 0.2712,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.447300910949707,
      "rewards/margins": 3.195319414138794,
      "rewards/rejected": -1.748018503189087,
      "step": 130
    },
    {
      "epoch": 2.014925373134328,
      "grad_norm": 15.864915551280724,
      "learning_rate": 2.667985967011878e-07,
      "logits/chosen": -2.1461024284362793,
      "logits/rejected": -2.168231248855591,
      "logps/chosen": -228.489501953125,
      "logps/rejected": -218.523193359375,
      "loss": 0.2636,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.7504692077636719,
      "rewards/margins": 3.347080945968628,
      "rewards/rejected": -1.596611738204956,
      "step": 135
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 20.70108938067205,
      "learning_rate": 2.3125386803640183e-07,
      "logits/chosen": -2.1393442153930664,
      "logits/rejected": -2.13488507270813,
      "logps/chosen": -227.1748046875,
      "logps/rejected": -206.73635864257812,
      "loss": 0.1687,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 2.1466095447540283,
      "rewards/margins": 3.6835296154022217,
      "rewards/rejected": -1.5369203090667725,
      "step": 140
    },
    {
      "epoch": 2.1641791044776117,
      "grad_norm": 15.3203590693748,
      "learning_rate": 1.9752578596124952e-07,
      "logits/chosen": -2.1212143898010254,
      "logits/rejected": -2.108987808227539,
      "logps/chosen": -229.7302703857422,
      "logps/rejected": -227.28109741210938,
      "loss": 0.1473,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 2.296015739440918,
      "rewards/margins": 4.074695110321045,
      "rewards/rejected": -1.7786792516708374,
      "step": 145
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 23.44280872584024,
      "learning_rate": 1.6584234261399532e-07,
      "logits/chosen": -2.1407978534698486,
      "logits/rejected": -2.152918815612793,
      "logps/chosen": -229.5863800048828,
      "logps/rejected": -212.45986938476562,
      "loss": 0.1557,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 2.0357882976531982,
      "rewards/margins": 4.068355560302734,
      "rewards/rejected": -2.032567024230957,
      "step": 150
    },
    {
      "epoch": 2.2388059701492535,
      "eval_logits/chosen": -2.1696274280548096,
      "eval_logits/rejected": -2.1170217990875244,
      "eval_logps/chosen": -259.6371765136719,
      "eval_logps/rejected": -211.39309692382812,
      "eval_loss": 0.5032579898834229,
      "eval_rewards/accuracies": 0.8583333492279053,
      "eval_rewards/chosen": 1.3754442930221558,
      "eval_rewards/margins": 2.7079200744628906,
      "eval_rewards/rejected": -1.3324757814407349,
      "eval_runtime": 125.6191,
      "eval_samples_per_second": 15.125,
      "eval_steps_per_second": 0.239,
      "step": 150
    },
    {
      "epoch": 2.3134328358208958,
      "grad_norm": 15.277790126219001,
      "learning_rate": 1.3641770896292082e-07,
      "logits/chosen": -2.171910524368286,
      "logits/rejected": -2.1707565784454346,
      "logps/chosen": -250.38583374023438,
      "logps/rejected": -238.89022827148438,
      "loss": 0.1381,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 2.417271137237549,
      "rewards/margins": 4.274667739868164,
      "rewards/rejected": -1.8573967218399048,
      "step": 155
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 14.868033858341619,
      "learning_rate": 1.0945078707215221e-07,
      "logits/chosen": -2.1999287605285645,
      "logits/rejected": -2.1457407474517822,
      "logps/chosen": -241.4682159423828,
      "logps/rejected": -240.2734832763672,
      "loss": 0.1412,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.0974485874176025,
      "rewards/margins": 4.184462547302246,
      "rewards/rejected": -2.0870139598846436,
      "step": 160
    },
    {
      "epoch": 2.4626865671641793,
      "grad_norm": 23.298138909290813,
      "learning_rate": 8.512386558088919e-08,
      "logits/chosen": -2.1854496002197266,
      "logits/rejected": -2.173795223236084,
      "logps/chosen": -261.7022399902344,
      "logps/rejected": -242.4682159423828,
      "loss": 0.1478,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.057788133621216,
      "rewards/margins": 4.235553741455078,
      "rewards/rejected": -2.1777658462524414,
      "step": 165
    },
    {
      "epoch": 2.5373134328358207,
      "grad_norm": 22.03814748434263,
      "learning_rate": 6.360138748461013e-08,
      "logits/chosen": -2.20246958732605,
      "logits/rejected": -2.161410093307495,
      "logps/chosen": -251.98989868164062,
      "logps/rejected": -254.8340301513672,
      "loss": 0.1586,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 1.9426151514053345,
      "rewards/margins": 4.301736831665039,
      "rewards/rejected": -2.359121799468994,
      "step": 170
    },
    {
      "epoch": 2.611940298507463,
      "grad_norm": 19.677590731179777,
      "learning_rate": 4.5028838547699346e-08,
      "logits/chosen": -2.1607601642608643,
      "logits/rejected": -2.127946376800537,
      "logps/chosen": -221.7189178466797,
      "logps/rejected": -245.9696502685547,
      "loss": 0.1676,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": 1.624866247177124,
      "rewards/margins": 4.7091498374938965,
      "rewards/rejected": -3.0842833518981934,
      "step": 175
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 22.8450636545283,
      "learning_rate": 2.9531763861505964e-08,
      "logits/chosen": -2.1652731895446777,
      "logits/rejected": -2.100562334060669,
      "logps/chosen": -219.1479949951172,
      "logps/rejected": -211.0758056640625,
      "loss": 0.1661,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.2030365467071533,
      "rewards/margins": 3.858271837234497,
      "rewards/rejected": -2.6552352905273438,
      "step": 180
    },
    {
      "epoch": 2.7611940298507465,
      "grad_norm": 13.900949161141034,
      "learning_rate": 1.7214919195619125e-08,
      "logits/chosen": -2.158569574356079,
      "logits/rejected": -2.139193058013916,
      "logps/chosen": -241.50634765625,
      "logps/rejected": -225.2047882080078,
      "loss": 0.1367,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 1.889639139175415,
      "rewards/margins": 4.425418853759766,
      "rewards/rejected": -2.5357797145843506,
      "step": 185
    },
    {
      "epoch": 2.835820895522388,
      "grad_norm": 22.822822022841606,
      "learning_rate": 8.161562878982398e-09,
      "logits/chosen": -2.134626865386963,
      "logits/rejected": -2.092273235321045,
      "logps/chosen": -233.07662963867188,
      "logps/rejected": -203.2852783203125,
      "loss": 0.1562,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 1.5627435445785522,
      "rewards/margins": 4.164488315582275,
      "rewards/rejected": -2.601745367050171,
      "step": 190
    },
    {
      "epoch": 2.91044776119403,
      "grad_norm": 13.470995740623104,
      "learning_rate": 2.432892997526026e-09,
      "logits/chosen": -2.160935878753662,
      "logits/rejected": -2.1190755367279053,
      "logps/chosen": -251.3184814453125,
      "logps/rejected": -268.527587890625,
      "loss": 0.1325,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 1.8667724132537842,
      "rewards/margins": 4.689689636230469,
      "rewards/rejected": -2.822917938232422,
      "step": 195
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 14.133379803838286,
      "learning_rate": 6.763371270035457e-11,
      "logits/chosen": -2.160177707672119,
      "logits/rejected": -2.123290777206421,
      "logps/chosen": -251.46255493164062,
      "logps/rejected": -241.31295776367188,
      "loss": 0.1338,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 1.7690976858139038,
      "rewards/margins": 4.538566589355469,
      "rewards/rejected": -2.7694690227508545,
      "step": 200
    },
    {
      "epoch": 2.9850746268656714,
      "eval_logits/chosen": -2.158780097961426,
      "eval_logits/rejected": -2.1046879291534424,
      "eval_logps/chosen": -263.8291320800781,
      "eval_logps/rejected": -217.83078002929688,
      "eval_loss": 0.4983350932598114,
      "eval_rewards/accuracies": 0.8500000238418579,
      "eval_rewards/chosen": 0.9562520384788513,
      "eval_rewards/margins": 2.932494878768921,
      "eval_rewards/rejected": -1.9762426614761353,
      "eval_runtime": 125.7438,
      "eval_samples_per_second": 15.11,
      "eval_steps_per_second": 0.239,
      "step": 200
    },
    {
      "epoch": 3.0,
      "step": 201,
      "total_flos": 2369906314051584.0,
      "train_loss": 0.33095499672996465,
      "train_runtime": 7348.1824,
      "train_samples_per_second": 6.979,
      "train_steps_per_second": 0.027
    }
  ],
  "logging_steps": 5,
  "max_steps": 201,
  "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": 2369906314051584.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}