{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 478,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 27.376848082151543,
      "learning_rate": 1.0416666666666666e-08,
      "logits/chosen": -1.8783892393112183,
      "logits/rejected": -1.8756425380706787,
      "logps/chosen": -298.4870300292969,
      "logps/rejected": -398.0157165527344,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.02,
      "grad_norm": 25.183613892194092,
      "learning_rate": 1.0416666666666667e-07,
      "logits/chosen": -1.7508937120437622,
      "logits/rejected": -1.7071995735168457,
      "logps/chosen": -280.5123596191406,
      "logps/rejected": -271.9100341796875,
      "loss": 0.6931,
      "rewards/accuracies": 0.4791666567325592,
      "rewards/chosen": -0.00033082481240853667,
      "rewards/margins": 0.0001291988301090896,
      "rewards/rejected": -0.00046002367162145674,
      "step": 10
    },
    {
      "epoch": 0.04,
      "grad_norm": 23.230240384482265,
      "learning_rate": 2.0833333333333333e-07,
      "logits/chosen": -1.832267165184021,
      "logits/rejected": -1.7261158227920532,
      "logps/chosen": -298.87811279296875,
      "logps/rejected": -320.8003845214844,
      "loss": 0.6919,
      "rewards/accuracies": 0.581250011920929,
      "rewards/chosen": 0.009286092594265938,
      "rewards/margins": 0.0007404378848150373,
      "rewards/rejected": 0.008545654825866222,
      "step": 20
    },
    {
      "epoch": 0.06,
      "grad_norm": 22.82438334979188,
      "learning_rate": 3.1249999999999997e-07,
      "logits/chosen": -1.8619463443756104,
      "logits/rejected": -1.8105783462524414,
      "logps/chosen": -315.0911865234375,
      "logps/rejected": -281.7936096191406,
      "loss": 0.6843,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": 0.06353694945573807,
      "rewards/margins": 0.020365100353956223,
      "rewards/rejected": 0.043171852827072144,
      "step": 30
    },
    {
      "epoch": 0.08,
      "grad_norm": 20.30578423378571,
      "learning_rate": 4.1666666666666667e-07,
      "logits/chosen": -1.7271950244903564,
      "logits/rejected": -1.6917632818222046,
      "logps/chosen": -269.0830078125,
      "logps/rejected": -258.13482666015625,
      "loss": 0.6707,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": 0.15124677121639252,
      "rewards/margins": 0.058850765228271484,
      "rewards/rejected": 0.09239600598812103,
      "step": 40
    },
    {
      "epoch": 0.1,
      "grad_norm": 18.991767469871117,
      "learning_rate": 4.999733114418725e-07,
      "logits/chosen": -1.7609294652938843,
      "logits/rejected": -1.749925971031189,
      "logps/chosen": -274.7580261230469,
      "logps/rejected": -298.2468566894531,
      "loss": 0.6568,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": 0.22399310767650604,
      "rewards/margins": 0.05384649708867073,
      "rewards/rejected": 0.1701466143131256,
      "step": 50
    },
    {
      "epoch": 0.13,
      "grad_norm": 22.512008467267577,
      "learning_rate": 4.990398100856366e-07,
      "logits/chosen": -1.8478819131851196,
      "logits/rejected": -1.8104110956192017,
      "logps/chosen": -268.5568542480469,
      "logps/rejected": -318.2205505371094,
      "loss": 0.6419,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.22708478569984436,
      "rewards/margins": 0.11861655861139297,
      "rewards/rejected": 0.108468197286129,
      "step": 60
    },
    {
      "epoch": 0.15,
      "grad_norm": 24.33517976511266,
      "learning_rate": 4.967775735898179e-07,
      "logits/chosen": -1.6735728979110718,
      "logits/rejected": -1.6890596151351929,
      "logps/chosen": -274.67156982421875,
      "logps/rejected": -289.94256591796875,
      "loss": 0.6121,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.11328981816768646,
      "rewards/margins": 0.22474534809589386,
      "rewards/rejected": -0.11145554482936859,
      "step": 70
    },
    {
      "epoch": 0.17,
      "grad_norm": 32.81066991350388,
      "learning_rate": 4.931986719649298e-07,
      "logits/chosen": -1.9151760339736938,
      "logits/rejected": -1.802167534828186,
      "logps/chosen": -356.9640808105469,
      "logps/rejected": -325.5661315917969,
      "loss": 0.5878,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.08058764040470123,
      "rewards/margins": 0.30372124910354614,
      "rewards/rejected": -0.38430893421173096,
      "step": 80
    },
    {
      "epoch": 0.19,
      "grad_norm": 31.216011823878958,
      "learning_rate": 4.883222001996351e-07,
      "logits/chosen": -1.7752355337142944,
      "logits/rejected": -1.7595351934432983,
      "logps/chosen": -325.2732849121094,
      "logps/rejected": -370.2122497558594,
      "loss": 0.5628,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.44604864716529846,
      "rewards/margins": 0.5209277272224426,
      "rewards/rejected": -0.9669763445854187,
      "step": 90
    },
    {
      "epoch": 0.21,
      "grad_norm": 30.757968765158964,
      "learning_rate": 4.821741763807186e-07,
      "logits/chosen": -1.8407386541366577,
      "logits/rejected": -1.818633794784546,
      "logps/chosen": -356.88604736328125,
      "logps/rejected": -402.81597900390625,
      "loss": 0.5759,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.7781068682670593,
      "rewards/margins": 0.501887321472168,
      "rewards/rejected": -1.279994249343872,
      "step": 100
    },
    {
      "epoch": 0.21,
      "eval_logits/chosen": -1.9150919914245605,
      "eval_logits/rejected": -1.8988227844238281,
      "eval_logps/chosen": -335.8161315917969,
      "eval_logps/rejected": -395.6910095214844,
      "eval_loss": 0.5723333954811096,
      "eval_rewards/accuracies": 0.71484375,
      "eval_rewards/chosen": -0.5418255925178528,
      "eval_rewards/margins": 0.5374687910079956,
      "eval_rewards/rejected": -1.0792944431304932,
      "eval_runtime": 97.7859,
      "eval_samples_per_second": 20.453,
      "eval_steps_per_second": 0.327,
      "step": 100
    },
    {
      "epoch": 0.23,
      "grad_norm": 31.68583269080582,
      "learning_rate": 4.747874028753375e-07,
      "logits/chosen": -1.8968604803085327,
      "logits/rejected": -1.7719805240631104,
      "logps/chosen": -367.7148742675781,
      "logps/rejected": -371.1282653808594,
      "loss": 0.5798,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.36616355180740356,
      "rewards/margins": 0.49689429998397827,
      "rewards/rejected": -0.8630577921867371,
      "step": 110
    },
    {
      "epoch": 0.25,
      "grad_norm": 30.73893119347065,
      "learning_rate": 4.662012913161997e-07,
      "logits/chosen": -1.7683579921722412,
      "logits/rejected": -1.7642805576324463,
      "logps/chosen": -332.87030029296875,
      "logps/rejected": -358.53521728515625,
      "loss": 0.547,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.2855382263660431,
      "rewards/margins": 0.4636038839817047,
      "rewards/rejected": -0.7491421103477478,
      "step": 120
    },
    {
      "epoch": 0.27,
      "grad_norm": 36.07936832382646,
      "learning_rate": 4.5646165232345103e-07,
      "logits/chosen": -1.8693767786026,
      "logits/rejected": -1.794420599937439,
      "logps/chosen": -345.6004638671875,
      "logps/rejected": -402.54608154296875,
      "loss": 0.5353,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.4933016300201416,
      "rewards/margins": 0.5930127501487732,
      "rewards/rejected": -1.0863142013549805,
      "step": 130
    },
    {
      "epoch": 0.29,
      "grad_norm": 39.43015574676247,
      "learning_rate": 4.456204510851956e-07,
      "logits/chosen": -1.7766962051391602,
      "logits/rejected": -1.7008291482925415,
      "logps/chosen": -394.8687744140625,
      "logps/rejected": -457.8002014160156,
      "loss": 0.5255,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.5899434089660645,
      "rewards/margins": 0.7221861481666565,
      "rewards/rejected": -1.3121296167373657,
      "step": 140
    },
    {
      "epoch": 0.31,
      "grad_norm": 48.649979456976475,
      "learning_rate": 4.337355301007335e-07,
      "logits/chosen": -1.7205531597137451,
      "logits/rejected": -1.6475614309310913,
      "logps/chosen": -382.6226806640625,
      "logps/rejected": -433.13409423828125,
      "loss": 0.5514,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.8077861666679382,
      "rewards/margins": 0.5227850675582886,
      "rewards/rejected": -1.3305714130401611,
      "step": 150
    },
    {
      "epoch": 0.33,
      "grad_norm": 44.8404423652529,
      "learning_rate": 4.2087030056579986e-07,
      "logits/chosen": -1.698359727859497,
      "logits/rejected": -1.583348035812378,
      "logps/chosen": -338.5320129394531,
      "logps/rejected": -390.748291015625,
      "loss": 0.5328,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.4389611780643463,
      "rewards/margins": 0.7732844352722168,
      "rewards/rejected": -1.2122455835342407,
      "step": 160
    },
    {
      "epoch": 0.36,
      "grad_norm": 44.50813022956455,
      "learning_rate": 4.070934040463998e-07,
      "logits/chosen": -1.607373833656311,
      "logits/rejected": -1.4848835468292236,
      "logps/chosen": -323.30230712890625,
      "logps/rejected": -363.1103515625,
      "loss": 0.5341,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.5660237669944763,
      "rewards/margins": 0.5307204723358154,
      "rewards/rejected": -1.0967442989349365,
      "step": 170
    },
    {
      "epoch": 0.38,
      "grad_norm": 36.42011775174975,
      "learning_rate": 3.9247834624635404e-07,
      "logits/chosen": -1.4740314483642578,
      "logits/rejected": -1.3969428539276123,
      "logps/chosen": -355.79022216796875,
      "logps/rejected": -386.48846435546875,
      "loss": 0.5082,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.8436854481697083,
      "rewards/margins": 0.5756739377975464,
      "rewards/rejected": -1.4193594455718994,
      "step": 180
    },
    {
      "epoch": 0.4,
      "grad_norm": 37.5415107419379,
      "learning_rate": 3.7710310482256523e-07,
      "logits/chosen": -1.5718328952789307,
      "logits/rejected": -1.524204134941101,
      "logps/chosen": -348.835693359375,
      "logps/rejected": -401.84857177734375,
      "loss": 0.5283,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.5492793321609497,
      "rewards/margins": 0.6370285153388977,
      "rewards/rejected": -1.186307668685913,
      "step": 190
    },
    {
      "epoch": 0.42,
      "grad_norm": 86.8216257987796,
      "learning_rate": 3.610497133404795e-07,
      "logits/chosen": -1.508467674255371,
      "logits/rejected": -1.409771203994751,
      "logps/chosen": -339.861328125,
      "logps/rejected": -404.37200927734375,
      "loss": 0.5391,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.6251331567764282,
      "rewards/margins": 0.7106647491455078,
      "rewards/rejected": -1.335797905921936,
      "step": 200
    },
    {
      "epoch": 0.42,
      "eval_logits/chosen": -1.6569650173187256,
      "eval_logits/rejected": -1.577364206314087,
      "eval_logps/chosen": -321.1681823730469,
      "eval_logps/rejected": -405.005615234375,
      "eval_loss": 0.5137878060340881,
      "eval_rewards/accuracies": 0.74609375,
      "eval_rewards/chosen": -0.3953460156917572,
      "eval_rewards/margins": 0.777094841003418,
      "eval_rewards/rejected": -1.172440767288208,
      "eval_runtime": 97.4997,
      "eval_samples_per_second": 20.513,
      "eval_steps_per_second": 0.328,
      "step": 200
    },
    {
      "epoch": 0.44,
      "grad_norm": 38.50117908255001,
      "learning_rate": 3.4440382358952115e-07,
      "logits/chosen": -1.443800687789917,
      "logits/rejected": -1.3134807348251343,
      "logps/chosen": -354.0191345214844,
      "logps/rejected": -393.02264404296875,
      "loss": 0.5328,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.40636110305786133,
      "rewards/margins": 0.759956419467926,
      "rewards/rejected": -1.1663174629211426,
      "step": 210
    },
    {
      "epoch": 0.46,
      "grad_norm": 50.76318175549852,
      "learning_rate": 3.272542485937368e-07,
      "logits/chosen": -1.2096138000488281,
      "logits/rejected": -1.064009189605713,
      "logps/chosen": -339.54339599609375,
      "logps/rejected": -401.81622314453125,
      "loss": 0.5149,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.570379376411438,
      "rewards/margins": 0.8607443571090698,
      "rewards/rejected": -1.4311238527297974,
      "step": 220
    },
    {
      "epoch": 0.48,
      "grad_norm": 56.74639801937906,
      "learning_rate": 3.096924887558854e-07,
      "logits/chosen": -1.1264355182647705,
      "logits/rejected": -1.0245306491851807,
      "logps/chosen": -354.9466857910156,
      "logps/rejected": -463.1206970214844,
      "loss": 0.5149,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.8869966268539429,
      "rewards/margins": 1.0312917232513428,
      "rewards/rejected": -1.9182884693145752,
      "step": 230
    },
    {
      "epoch": 0.5,
      "grad_norm": 42.18851613816063,
      "learning_rate": 2.9181224366319943e-07,
      "logits/chosen": -1.2307146787643433,
      "logits/rejected": -1.139776349067688,
      "logps/chosen": -370.49853515625,
      "logps/rejected": -442.11114501953125,
      "loss": 0.487,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.8160694241523743,
      "rewards/margins": 0.8950576782226562,
      "rewards/rejected": -1.7111270427703857,
      "step": 240
    },
    {
      "epoch": 0.52,
      "grad_norm": 42.578136485563924,
      "learning_rate": 2.7370891215954565e-07,
      "logits/chosen": -1.104014277458191,
      "logits/rejected": -0.9260801076889038,
      "logps/chosen": -386.5050964355469,
      "logps/rejected": -449.59722900390625,
      "loss": 0.5122,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.6448307633399963,
      "rewards/margins": 0.9931491017341614,
      "rewards/rejected": -1.6379798650741577,
      "step": 250
    },
    {
      "epoch": 0.54,
      "grad_norm": 45.2823901088297,
      "learning_rate": 2.55479083351317e-07,
      "logits/chosen": -0.9789173007011414,
      "logits/rejected": -0.7768954634666443,
      "logps/chosen": -395.30548095703125,
      "logps/rejected": -449.7933044433594,
      "loss": 0.505,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.7873316407203674,
      "rewards/margins": 0.9275019764900208,
      "rewards/rejected": -1.7148334980010986,
      "step": 260
    },
    {
      "epoch": 0.56,
      "grad_norm": 44.49352654305102,
      "learning_rate": 2.3722002126275822e-07,
      "logits/chosen": -0.9071099162101746,
      "logits/rejected": -0.8342393636703491,
      "logps/chosen": -374.5643310546875,
      "logps/rejected": -433.95098876953125,
      "loss": 0.5205,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.729364275932312,
      "rewards/margins": 0.7837658524513245,
      "rewards/rejected": -1.5131301879882812,
      "step": 270
    },
    {
      "epoch": 0.59,
      "grad_norm": 42.661081149871585,
      "learning_rate": 2.19029145890313e-07,
      "logits/chosen": -0.9833901524543762,
      "logits/rejected": -0.7881035804748535,
      "logps/chosen": -336.73388671875,
      "logps/rejected": -407.16326904296875,
      "loss": 0.5186,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.5388863682746887,
      "rewards/margins": 0.8700063824653625,
      "rewards/rejected": -1.4088926315307617,
      "step": 280
    },
    {
      "epoch": 0.61,
      "grad_norm": 42.8181212203347,
      "learning_rate": 2.0100351342479216e-07,
      "logits/chosen": -1.1338340044021606,
      "logits/rejected": -1.0056653022766113,
      "logps/chosen": -327.16217041015625,
      "logps/rejected": -406.1835021972656,
      "loss": 0.5062,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.5975332260131836,
      "rewards/margins": 0.854636549949646,
      "rewards/rejected": -1.4521698951721191,
      "step": 290
    },
    {
      "epoch": 0.63,
      "grad_norm": 43.557515582129994,
      "learning_rate": 1.8323929841460178e-07,
      "logits/chosen": -1.1290686130523682,
      "logits/rejected": -0.9314834475517273,
      "logps/chosen": -405.9638671875,
      "logps/rejected": -460.133544921875,
      "loss": 0.4788,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.8333964347839355,
      "rewards/margins": 0.9296092987060547,
      "rewards/rejected": -1.7630058526992798,
      "step": 300
    },
    {
      "epoch": 0.63,
      "eval_logits/chosen": -0.986372709274292,
      "eval_logits/rejected": -0.8825947046279907,
      "eval_logps/chosen": -367.72772216796875,
      "eval_logps/rejected": -475.55322265625,
      "eval_loss": 0.49388447403907776,
      "eval_rewards/accuracies": 0.7578125,
      "eval_rewards/chosen": -0.8609416484832764,
      "eval_rewards/margins": 1.016974687576294,
      "eval_rewards/rejected": -1.8779162168502808,
      "eval_runtime": 97.7186,
      "eval_samples_per_second": 20.467,
      "eval_steps_per_second": 0.327,
      "step": 300
    },
    {
      "epoch": 0.65,
      "grad_norm": 50.71884572569721,
      "learning_rate": 1.6583128063291573e-07,
      "logits/chosen": -0.8109232783317566,
      "logits/rejected": -0.6804934144020081,
      "logps/chosen": -421.7939453125,
      "logps/rejected": -471.0542907714844,
      "loss": 0.4943,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.045866847038269,
      "rewards/margins": 0.8514853715896606,
      "rewards/rejected": -1.8973522186279297,
      "step": 310
    },
    {
      "epoch": 0.67,
      "grad_norm": 50.530325658635704,
      "learning_rate": 1.488723393865766e-07,
      "logits/chosen": -0.697489321231842,
      "logits/rejected": -0.5368015170097351,
      "logps/chosen": -416.4817810058594,
      "logps/rejected": -457.8753356933594,
      "loss": 0.4844,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.9900579452514648,
      "rewards/margins": 0.9124795198440552,
      "rewards/rejected": -1.9025375843048096,
      "step": 320
    },
    {
      "epoch": 0.69,
      "grad_norm": 51.29954417938159,
      "learning_rate": 1.3245295796480788e-07,
      "logits/chosen": -0.9539875984191895,
      "logits/rejected": -0.7579910159111023,
      "logps/chosen": -368.8915710449219,
      "logps/rejected": -450.0838317871094,
      "loss": 0.4783,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.8678079843521118,
      "rewards/margins": 0.7714605331420898,
      "rewards/rejected": -1.6392685174942017,
      "step": 330
    },
    {
      "epoch": 0.71,
      "grad_norm": 47.146951845139505,
      "learning_rate": 1.1666074087171627e-07,
      "logits/chosen": -0.9075170755386353,
      "logits/rejected": -0.7373852133750916,
      "logps/chosen": -388.6153259277344,
      "logps/rejected": -477.04217529296875,
      "loss": 0.4968,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.7279259562492371,
      "rewards/margins": 1.0357873439788818,
      "rewards/rejected": -1.7637134790420532,
      "step": 340
    },
    {
      "epoch": 0.73,
      "grad_norm": 40.51519007652658,
      "learning_rate": 1.0157994641835734e-07,
      "logits/chosen": -0.9085396528244019,
      "logits/rejected": -0.7681713104248047,
      "logps/chosen": -352.57073974609375,
      "logps/rejected": -430.44903564453125,
      "loss": 0.4725,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.7491118311882019,
      "rewards/margins": 0.9513088464736938,
      "rewards/rejected": -1.7004207372665405,
      "step": 350
    },
    {
      "epoch": 0.75,
      "grad_norm": 44.22500120445586,
      "learning_rate": 8.729103716819111e-08,
      "logits/chosen": -1.0195329189300537,
      "logits/rejected": -0.7282712459564209,
      "logps/chosen": -405.8896179199219,
      "logps/rejected": -458.53826904296875,
      "loss": 0.5133,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.7919691801071167,
      "rewards/margins": 0.9254425168037415,
      "rewards/rejected": -1.7174116373062134,
      "step": 360
    },
    {
      "epoch": 0.77,
      "grad_norm": 42.97964211979383,
      "learning_rate": 7.387025063449081e-08,
      "logits/chosen": -0.7996746301651001,
      "logits/rejected": -0.5788342356681824,
      "logps/chosen": -381.4356689453125,
      "logps/rejected": -423.28570556640625,
      "loss": 0.5002,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.8598394393920898,
      "rewards/margins": 0.7896124124526978,
      "rewards/rejected": -1.6494518518447876,
      "step": 370
    },
    {
      "epoch": 0.79,
      "grad_norm": 40.60703994205003,
      "learning_rate": 6.138919252022435e-08,
      "logits/chosen": -0.7264416217803955,
      "logits/rejected": -0.689493715763092,
      "logps/chosen": -355.46893310546875,
      "logps/rejected": -487.8539123535156,
      "loss": 0.485,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.9955714344978333,
      "rewards/margins": 1.0758744478225708,
      "rewards/rejected": -2.071445941925049,
      "step": 380
    },
    {
      "epoch": 0.82,
      "grad_norm": 46.93787360588045,
      "learning_rate": 4.991445467064689e-08,
      "logits/chosen": -0.8655691146850586,
      "logits/rejected": -0.7559862732887268,
      "logps/chosen": -414.95660400390625,
      "logps/rejected": -487.86029052734375,
      "loss": 0.4776,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.9033805727958679,
      "rewards/margins": 0.8835199475288391,
      "rewards/rejected": -1.786900281906128,
      "step": 390
    },
    {
      "epoch": 0.84,
      "grad_norm": 44.554590531311376,
      "learning_rate": 3.9507259776993954e-08,
      "logits/chosen": -0.6984928846359253,
      "logits/rejected": -0.5499908924102783,
      "logps/chosen": -390.16436767578125,
      "logps/rejected": -478.6742248535156,
      "loss": 0.4937,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.9803212285041809,
      "rewards/margins": 0.9358729124069214,
      "rewards/rejected": -1.916194200515747,
      "step": 400
    },
    {
      "epoch": 0.84,
      "eval_logits/chosen": -0.9126077890396118,
      "eval_logits/rejected": -0.760056734085083,
      "eval_logps/chosen": -361.4586486816406,
      "eval_logps/rejected": -470.95220947265625,
      "eval_loss": 0.4905379116535187,
      "eval_rewards/accuracies": 0.75390625,
      "eval_rewards/chosen": -0.7982508540153503,
      "eval_rewards/margins": 1.0336554050445557,
      "eval_rewards/rejected": -1.8319063186645508,
      "eval_runtime": 97.7293,
      "eval_samples_per_second": 20.465,
      "eval_steps_per_second": 0.327,
      "step": 400
    },
    {
      "epoch": 0.86,
      "grad_norm": 39.65148650158285,
      "learning_rate": 3.022313472693447e-08,
      "logits/chosen": -0.938845157623291,
      "logits/rejected": -0.6247340440750122,
      "logps/chosen": -400.84222412109375,
      "logps/rejected": -476.00018310546875,
      "loss": 0.4863,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.8032888174057007,
      "rewards/margins": 1.0230791568756104,
      "rewards/rejected": -1.8263680934906006,
      "step": 410
    },
    {
      "epoch": 0.88,
      "grad_norm": 42.93092965398867,
      "learning_rate": 2.2111614344599684e-08,
      "logits/chosen": -0.8916628956794739,
      "logits/rejected": -0.7629222869873047,
      "logps/chosen": -404.97088623046875,
      "logps/rejected": -487.2880859375,
      "loss": 0.4797,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.836542010307312,
      "rewards/margins": 1.0462357997894287,
      "rewards/rejected": -1.8827779293060303,
      "step": 420
    },
    {
      "epoch": 0.9,
      "grad_norm": 50.64982947415171,
      "learning_rate": 1.521597710086439e-08,
      "logits/chosen": -0.7213121056556702,
      "logits/rejected": -0.5332973003387451,
      "logps/chosen": -401.74871826171875,
      "logps/rejected": -478.0919494628906,
      "loss": 0.4735,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.9987181425094604,
      "rewards/margins": 0.9951263666152954,
      "rewards/rejected": -1.9938443899154663,
      "step": 430
    },
    {
      "epoch": 0.92,
      "grad_norm": 45.64074270515632,
      "learning_rate": 9.57301420397924e-09,
      "logits/chosen": -0.8010743856430054,
      "logits/rejected": -0.6153632998466492,
      "logps/chosen": -389.0445251464844,
      "logps/rejected": -466.53387451171875,
      "loss": 0.4956,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.842692494392395,
      "rewards/margins": 0.8713605999946594,
      "rewards/rejected": -1.7140531539916992,
      "step": 440
    },
    {
      "epoch": 0.94,
      "grad_norm": 42.58278664732128,
      "learning_rate": 5.212833302556258e-09,
      "logits/chosen": -0.7098689675331116,
      "logits/rejected": -0.6174753308296204,
      "logps/chosen": -411.04345703125,
      "logps/rejected": -515.0621948242188,
      "loss": 0.4986,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.983228325843811,
      "rewards/margins": 0.8576449155807495,
      "rewards/rejected": -1.84087336063385,
      "step": 450
    },
    {
      "epoch": 0.96,
      "grad_norm": 60.928874193313085,
      "learning_rate": 2.158697848236607e-09,
      "logits/chosen": -0.7517565488815308,
      "logits/rejected": -0.5577678084373474,
      "logps/chosen": -384.75677490234375,
      "logps/rejected": -444.00030517578125,
      "loss": 0.4922,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.859534740447998,
      "rewards/margins": 0.9033729434013367,
      "rewards/rejected": -1.76290762424469,
      "step": 460
    },
    {
      "epoch": 0.98,
      "grad_norm": 45.081509845975475,
      "learning_rate": 4.269029751107489e-10,
      "logits/chosen": -0.7836319804191589,
      "logits/rejected": -0.5721665620803833,
      "logps/chosen": -376.28497314453125,
      "logps/rejected": -472.9750061035156,
      "loss": 0.4796,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.814776599407196,
      "rewards/margins": 0.9141017198562622,
      "rewards/rejected": -1.7288782596588135,
      "step": 470
    },
    {
      "epoch": 1.0,
      "step": 478,
      "total_flos": 0.0,
      "train_loss": 0.07911034258838479,
      "train_runtime": 1140.8874,
      "train_samples_per_second": 53.585,
      "train_steps_per_second": 0.419
    }
  ],
  "logging_steps": 10,
  "max_steps": 478,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}