{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 329,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 2.09392729051305,
      "learning_rate": 1.5151515151515152e-07,
      "logits/chosen": -2.6820077896118164,
      "logits/rejected": -2.6930205821990967,
      "logps/chosen": -281.2528381347656,
      "logps/rejected": -258.0622253417969,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/margins_max": 0.0,
      "rewards/margins_min": 0.0,
      "rewards/margins_std": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.0925129894368433,
      "learning_rate": 1.5151515151515152e-06,
      "logits/chosen": -2.7695705890655518,
      "logits/rejected": -2.755211353302002,
      "logps/chosen": -284.5762634277344,
      "logps/rejected": -249.84103393554688,
      "loss": 0.6932,
      "rewards/accuracies": 0.4166666567325592,
      "rewards/chosen": 0.00027878841501660645,
      "rewards/margins": 0.00013094051973894238,
      "rewards/margins_max": 0.0033118026331067085,
      "rewards/margins_min": -0.002655874704942107,
      "rewards/margins_std": 0.0026171421632170677,
      "rewards/rejected": 0.0001478478661738336,
      "step": 10
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.9815822382919812,
      "learning_rate": 3.0303030303030305e-06,
      "logits/chosen": -2.833542823791504,
      "logits/rejected": -2.7807106971740723,
      "logps/chosen": -291.4638977050781,
      "logps/rejected": -270.51416015625,
      "loss": 0.6925,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.0006784057477489114,
      "rewards/margins": 0.0018588269595056772,
      "rewards/margins_max": 0.006385893560945988,
      "rewards/margins_min": -0.0022843112237751484,
      "rewards/margins_std": 0.0037722717970609665,
      "rewards/rejected": -0.001180421095341444,
      "step": 20
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.67492468935064,
      "learning_rate": 4.5454545454545455e-06,
      "logits/chosen": -2.860595226287842,
      "logits/rejected": -2.8134350776672363,
      "logps/chosen": -259.4276123046875,
      "logps/rejected": -228.0315704345703,
      "loss": 0.6913,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.001823395723477006,
      "rewards/margins": 0.004928162787109613,
      "rewards/margins_max": 0.017859386280179024,
      "rewards/margins_min": -0.005589204374700785,
      "rewards/margins_std": 0.01042240671813488,
      "rewards/rejected": -0.006751558743417263,
      "step": 30
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5968086960496506,
      "learning_rate": 4.993103596812269e-06,
      "logits/chosen": -2.8253955841064453,
      "logits/rejected": -2.762451648712158,
      "logps/chosen": -317.4245910644531,
      "logps/rejected": -226.83029174804688,
      "loss": 0.6847,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.001176043995656073,
      "rewards/margins": 0.022631246596574783,
      "rewards/margins_max": 0.05663127824664116,
      "rewards/margins_min": -0.002739027375355363,
      "rewards/margins_std": 0.027225341647863388,
      "rewards/rejected": -0.021455200389027596,
      "step": 40
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.7771351450812785,
      "learning_rate": 4.95941685833271e-06,
      "logits/chosen": -2.7818782329559326,
      "logits/rejected": -2.796020984649658,
      "logps/chosen": -243.5937957763672,
      "logps/rejected": -282.7391662597656,
      "loss": 0.6835,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.006589516066014767,
      "rewards/margins": 0.02196229062974453,
      "rewards/margins_max": 0.07266667485237122,
      "rewards/margins_min": -0.01875842735171318,
      "rewards/margins_std": 0.04057556018233299,
      "rewards/rejected": -0.02855180762708187,
      "step": 50
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.962252962150773,
      "learning_rate": 4.898051734555676e-06,
      "logits/chosen": -2.818239688873291,
      "logits/rejected": -2.8286662101745605,
      "logps/chosen": -320.9726257324219,
      "logps/rejected": -286.5067138671875,
      "loss": 0.6723,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.00948464497923851,
      "rewards/margins": 0.042791008949279785,
      "rewards/margins_max": 0.1431531012058258,
      "rewards/margins_min": -0.06081182882189751,
      "rewards/margins_std": 0.08912219107151031,
      "rewards/rejected": -0.033306367695331573,
      "step": 60
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.8587103545341894,
      "learning_rate": 4.809698831278217e-06,
      "logits/chosen": -2.7206215858459473,
      "logits/rejected": -2.7101528644561768,
      "logps/chosen": -266.5228576660156,
      "logps/rejected": -251.3375244140625,
      "loss": 0.6729,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.00045425668940879405,
      "rewards/margins": 0.049175992608070374,
      "rewards/margins_max": 0.13846850395202637,
      "rewards/margins_min": -0.052676666527986526,
      "rewards/margins_std": 0.08763651549816132,
      "rewards/rejected": -0.04963024705648422,
      "step": 70
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.2136192637287415,
      "learning_rate": 4.695352475952706e-06,
      "logits/chosen": -2.807755947113037,
      "logits/rejected": -2.7845520973205566,
      "logps/chosen": -283.79571533203125,
      "logps/rejected": -282.36761474609375,
      "loss": 0.6639,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.015222731046378613,
      "rewards/margins": 0.05928920581936836,
      "rewards/margins_max": 0.19641625881195068,
      "rewards/margins_min": -0.08654795587062836,
      "rewards/margins_std": 0.12829935550689697,
      "rewards/rejected": -0.07451193034648895,
      "step": 80
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.428092801188637,
      "learning_rate": 4.556299527482029e-06,
      "logits/chosen": -2.7647454738616943,
      "logits/rejected": -2.7148241996765137,
      "logps/chosen": -301.64178466796875,
      "logps/rejected": -302.57427978515625,
      "loss": 0.6646,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.060897838324308395,
      "rewards/margins": 0.049183908849954605,
      "rewards/margins_max": 0.24967026710510254,
      "rewards/margins_min": -0.15425434708595276,
      "rewards/margins_std": 0.17774364352226257,
      "rewards/rejected": -0.110081747174263,
      "step": 90
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.9682951507719486,
      "learning_rate": 4.394104893853007e-06,
      "logits/chosen": -2.860042095184326,
      "logits/rejected": -2.823477268218994,
      "logps/chosen": -277.8373718261719,
      "logps/rejected": -274.130615234375,
      "loss": 0.6517,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.04375104233622551,
      "rewards/margins": 0.1265198141336441,
      "rewards/margins_max": 0.31651413440704346,
      "rewards/margins_min": -0.057662349194288254,
      "rewards/margins_std": 0.16801223158836365,
      "rewards/rejected": -0.17027084529399872,
      "step": 100
    },
    {
      "epoch": 0.3,
      "eval_logits/chosen": -2.7791192531585693,
      "eval_logits/rejected": -2.744123697280884,
      "eval_logps/chosen": -285.6097412109375,
      "eval_logps/rejected": -269.1304016113281,
      "eval_loss": 0.6559571623802185,
      "eval_rewards/accuracies": 0.6899999976158142,
      "eval_rewards/chosen": -0.0101632634177804,
      "eval_rewards/margins": 0.0953514501452446,
      "eval_rewards/margins_max": 0.410457581281662,
      "eval_rewards/margins_min": -0.20267993211746216,
      "eval_rewards/margins_std": 0.20197777450084686,
      "eval_rewards/rejected": -0.10551470518112183,
      "eval_runtime": 429.1012,
      "eval_samples_per_second": 4.661,
      "eval_steps_per_second": 0.291,
      "step": 100
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.460788298428497,
      "learning_rate": 4.210593920593201e-06,
      "logits/chosen": -2.7310755252838135,
      "logits/rejected": -2.716763973236084,
      "logps/chosen": -314.26788330078125,
      "logps/rejected": -245.84921264648438,
      "loss": 0.6531,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.02511635422706604,
      "rewards/margins": 0.07915838807821274,
      "rewards/margins_max": 0.2661080062389374,
      "rewards/margins_min": -0.1229986920952797,
      "rewards/margins_std": 0.17364470660686493,
      "rewards/rejected": -0.10427474975585938,
      "step": 110
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.4585744987342597,
      "learning_rate": 4.007831848252212e-06,
      "logits/chosen": -2.713639736175537,
      "logits/rejected": -2.717987537384033,
      "logps/chosen": -329.2998046875,
      "logps/rejected": -291.00775146484375,
      "loss": 0.6411,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.05735304206609726,
      "rewards/margins": 0.10872934013605118,
      "rewards/margins_max": 0.3730432391166687,
      "rewards/margins_min": -0.18109610676765442,
      "rewards/margins_std": 0.2501201629638672,
      "rewards/rejected": -0.16608235239982605,
      "step": 120
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.296742130510628,
      "learning_rate": 3.7881005700938635e-06,
      "logits/chosen": -2.784316301345825,
      "logits/rejected": -2.794018507003784,
      "logps/chosen": -270.2655334472656,
      "logps/rejected": -249.17416381835938,
      "loss": 0.6356,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.0387195460498333,
      "rewards/margins": 0.11184574663639069,
      "rewards/margins_max": 0.4147607386112213,
      "rewards/margins_min": -0.18675808608531952,
      "rewards/margins_std": 0.26902341842651367,
      "rewards/rejected": -0.15056531131267548,
      "step": 130
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.5520149273689916,
      "learning_rate": 3.553872951569236e-06,
      "logits/chosen": -2.7240898609161377,
      "logits/rejected": -2.7166969776153564,
      "logps/chosen": -307.4956970214844,
      "logps/rejected": -302.90875244140625,
      "loss": 0.6294,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.13102450966835022,
      "rewards/margins": 0.19427050650119781,
      "rewards/margins_max": 0.5367092490196228,
      "rewards/margins_min": -0.12927336990833282,
      "rewards/margins_std": 0.2990504503250122,
      "rewards/rejected": -0.32529503107070923,
      "step": 140
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.409073201436679,
      "learning_rate": 3.307785000580313e-06,
      "logits/chosen": -2.7703697681427,
      "logits/rejected": -2.758200168609619,
      "logps/chosen": -292.15093994140625,
      "logps/rejected": -284.210693359375,
      "loss": 0.6342,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.21509556472301483,
      "rewards/margins": 0.17580775916576385,
      "rewards/margins_max": 0.5677517056465149,
      "rewards/margins_min": -0.20906031131744385,
      "rewards/margins_std": 0.3507745862007141,
      "rewards/rejected": -0.3909032940864563,
      "step": 150
    },
    {
      "epoch": 0.49,
      "grad_norm": 3.016300738429993,
      "learning_rate": 3.052606201731325e-06,
      "logits/chosen": -2.735063076019287,
      "logits/rejected": -2.718566417694092,
      "logps/chosen": -261.0966796875,
      "logps/rejected": -258.08099365234375,
      "loss": 0.6538,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.05592544749379158,
      "rewards/margins": 0.11891476809978485,
      "rewards/margins_max": 0.4770866930484772,
      "rewards/margins_min": -0.23589050769805908,
      "rewards/margins_std": 0.3210330605506897,
      "rewards/rejected": -0.17484021186828613,
      "step": 160
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.650519079076739,
      "learning_rate": 2.7912083484274266e-06,
      "logits/chosen": -2.747620105743408,
      "logits/rejected": -2.6757848262786865,
      "logps/chosen": -310.66448974609375,
      "logps/rejected": -301.0819091796875,
      "loss": 0.6027,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.06945960223674774,
      "rewards/margins": 0.2121526002883911,
      "rewards/margins_max": 0.5673162937164307,
      "rewards/margins_min": -0.14554527401924133,
      "rewards/margins_std": 0.3131377100944519,
      "rewards/rejected": -0.28161221742630005,
      "step": 170
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.9314631663166417,
      "learning_rate": 2.526533223585641e-06,
      "logits/chosen": -2.7748751640319824,
      "logits/rejected": -2.718724489212036,
      "logps/chosen": -268.07562255859375,
      "logps/rejected": -257.9107971191406,
      "loss": 0.6109,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.11941893398761749,
      "rewards/margins": 0.17082905769348145,
      "rewards/margins_max": 0.5415267944335938,
      "rewards/margins_min": -0.1771378219127655,
      "rewards/margins_std": 0.31447431445121765,
      "rewards/rejected": -0.29024797677993774,
      "step": 180
    },
    {
      "epoch": 0.58,
      "grad_norm": 3.113744990751754,
      "learning_rate": 2.2615594926807554e-06,
      "logits/chosen": -2.7187552452087402,
      "logits/rejected": -2.709470748901367,
      "logps/chosen": -306.37664794921875,
      "logps/rejected": -302.543701171875,
      "loss": 0.598,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.05795687437057495,
      "rewards/margins": 0.2614460289478302,
      "rewards/margins_max": 0.7187456488609314,
      "rewards/margins_min": -0.1500331312417984,
      "rewards/margins_std": 0.38883161544799805,
      "rewards/rejected": -0.31940287351608276,
      "step": 190
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.46704837439659,
      "learning_rate": 1.9992691817133025e-06,
      "logits/chosen": -2.725795269012451,
      "logits/rejected": -2.700993776321411,
      "logps/chosen": -298.82855224609375,
      "logps/rejected": -336.44140625,
      "loss": 0.5795,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.17227962613105774,
      "rewards/margins": 0.30934515595436096,
      "rewards/margins_max": 0.8330885171890259,
      "rewards/margins_min": -0.24815456569194794,
      "rewards/margins_std": 0.4841969907283783,
      "rewards/rejected": -0.4816247820854187,
      "step": 200
    },
    {
      "epoch": 0.61,
      "eval_logits/chosen": -2.770137310028076,
      "eval_logits/rejected": -2.738154649734497,
      "eval_logps/chosen": -305.9859313964844,
      "eval_logps/rejected": -303.88079833984375,
      "eval_loss": 0.6177265644073486,
      "eval_rewards/accuracies": 0.6819999814033508,
      "eval_rewards/chosen": -0.21392498910427094,
      "eval_rewards/margins": 0.23909391462802887,
      "eval_rewards/margins_max": 0.9738174080848694,
      "eval_rewards/margins_min": -0.4424648880958557,
      "eval_rewards/margins_std": 0.47772130370140076,
      "eval_rewards/rejected": -0.4530189037322998,
      "eval_runtime": 428.0295,
      "eval_samples_per_second": 4.673,
      "eval_steps_per_second": 0.292,
      "step": 200
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.3590225389683472,
      "learning_rate": 1.742614117358029e-06,
      "logits/chosen": -2.7592720985412598,
      "logits/rejected": -2.724670886993408,
      "logps/chosen": -328.0078430175781,
      "logps/rejected": -345.1807556152344,
      "loss": 0.5664,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.22690674662590027,
      "rewards/margins": 0.34196537733078003,
      "rewards/margins_max": 0.8763746023178101,
      "rewards/margins_min": -0.14572417736053467,
      "rewards/margins_std": 0.465751975774765,
      "rewards/rejected": -0.5688720941543579,
      "step": 210
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.608754753313119,
      "learning_rate": 1.4944827069769125e-06,
      "logits/chosen": -2.813154458999634,
      "logits/rejected": -2.796706199645996,
      "logps/chosen": -338.57867431640625,
      "logps/rejected": -311.73773193359375,
      "loss": 0.6278,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.25605496764183044,
      "rewards/margins": 0.19726695120334625,
      "rewards/margins_max": 0.692109227180481,
      "rewards/margins_min": -0.2970424294471741,
      "rewards/margins_std": 0.4508973956108093,
      "rewards/rejected": -0.4533218741416931,
      "step": 220
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.028562977399784,
      "learning_rate": 1.257667432355893e-06,
      "logits/chosen": -2.784775972366333,
      "logits/rejected": -2.8116180896759033,
      "logps/chosen": -320.9100646972656,
      "logps/rejected": -302.4966125488281,
      "loss": 0.6612,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.3226456046104431,
      "rewards/margins": 0.07472896575927734,
      "rewards/margins_max": 0.6551424264907837,
      "rewards/margins_min": -0.5510781407356262,
      "rewards/margins_std": 0.5290972590446472,
      "rewards/rejected": -0.39737454056739807,
      "step": 230
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.214920370095605,
      "learning_rate": 1.0348334229922677e-06,
      "logits/chosen": -2.8330750465393066,
      "logits/rejected": -2.793381452560425,
      "logps/chosen": -310.33575439453125,
      "logps/rejected": -318.92852783203125,
      "loss": 0.623,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.19318023324012756,
      "rewards/margins": 0.24430537223815918,
      "rewards/margins_max": 0.8275845646858215,
      "rewards/margins_min": -0.34578949213027954,
      "rewards/margins_std": 0.5300652384757996,
      "rewards/rejected": -0.43748560547828674,
      "step": 240
    },
    {
      "epoch": 0.76,
      "grad_norm": 3.4120571086616582,
      "learning_rate": 8.284884626103165e-07,
      "logits/chosen": -2.783628463745117,
      "logits/rejected": -2.7553200721740723,
      "logps/chosen": -310.9916687011719,
      "logps/rejected": -346.27276611328125,
      "loss": 0.6019,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.09901777654886246,
      "rewards/margins": 0.3181927502155304,
      "rewards/margins_max": 0.9294508099555969,
      "rewards/margins_min": -0.20651724934577942,
      "rewards/margins_std": 0.5166513919830322,
      "rewards/rejected": -0.41721048951148987,
      "step": 250
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.910453405014883,
      "learning_rate": 6.409547664531734e-07,
      "logits/chosen": -2.799520254135132,
      "logits/rejected": -2.77909255027771,
      "logps/chosen": -341.7703552246094,
      "logps/rejected": -347.35455322265625,
      "loss": 0.6077,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.07788446545600891,
      "rewards/margins": 0.2706037759780884,
      "rewards/margins_max": 0.7577456831932068,
      "rewards/margins_min": -0.09968160092830658,
      "rewards/margins_std": 0.3826812505722046,
      "rewards/rejected": -0.3484882414340973,
      "step": 260
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.466170918075581,
      "learning_rate": 4.743428469705336e-07,
      "logits/chosen": -2.751262903213501,
      "logits/rejected": -2.7571630477905273,
      "logps/chosen": -323.39410400390625,
      "logps/rejected": -356.8424987792969,
      "loss": 0.6027,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.19586281478405,
      "rewards/margins": 0.29274672269821167,
      "rewards/margins_max": 0.7976349592208862,
      "rewards/margins_min": -0.1788657009601593,
      "rewards/margins_std": 0.4405595362186432,
      "rewards/rejected": -0.4886094927787781,
      "step": 270
    },
    {
      "epoch": 0.85,
      "grad_norm": 3.1691510008843458,
      "learning_rate": 3.3052776201888266e-07,
      "logits/chosen": -2.800191879272461,
      "logits/rejected": -2.795988082885742,
      "logps/chosen": -346.5174255371094,
      "logps/rejected": -325.9556579589844,
      "loss": 0.5704,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.2095174342393875,
      "rewards/margins": 0.35230007767677307,
      "rewards/margins_max": 0.9150822758674622,
      "rewards/margins_min": -0.2514013946056366,
      "rewards/margins_std": 0.5179048776626587,
      "rewards/rejected": -0.561817467212677,
      "step": 280
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.145673782819818,
      "learning_rate": 2.111280128780638e-07,
      "logits/chosen": -2.7347121238708496,
      "logits/rejected": -2.703662633895874,
      "logps/chosen": -293.16033935546875,
      "logps/rejected": -290.85107421875,
      "loss": 0.6088,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.1897042840719223,
      "rewards/margins": 0.26360541582107544,
      "rewards/margins_max": 0.8136995434761047,
      "rewards/margins_min": -0.21997782588005066,
      "rewards/margins_std": 0.4667910635471344,
      "rewards/rejected": -0.45330971479415894,
      "step": 290
    },
    {
      "epoch": 0.91,
      "grad_norm": 6.333030325350309,
      "learning_rate": 1.1748732956682023e-07,
      "logits/chosen": -2.8414254188537598,
      "logits/rejected": -2.791003704071045,
      "logps/chosen": -348.19757080078125,
      "logps/rejected": -340.9607849121094,
      "loss": 0.5791,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.24484387040138245,
      "rewards/margins": 0.30887269973754883,
      "rewards/margins_max": 0.9062288403511047,
      "rewards/margins_min": -0.2699830234050751,
      "rewards/margins_std": 0.5193021893501282,
      "rewards/rejected": -0.5537165999412537,
      "step": 300
    },
    {
      "epoch": 0.91,
      "eval_logits/chosen": -2.7722113132476807,
      "eval_logits/rejected": -2.740241289138794,
      "eval_logps/chosen": -304.6903381347656,
      "eval_logps/rejected": -305.08380126953125,
      "eval_loss": 0.6102678775787354,
      "eval_rewards/accuracies": 0.6859999895095825,
      "eval_rewards/chosen": -0.20096930861473083,
      "eval_rewards/margins": 0.26407960057258606,
      "eval_rewards/margins_max": 1.0352518558502197,
      "eval_rewards/margins_min": -0.4649355411529541,
      "eval_rewards/margins_std": 0.5062240362167358,
      "eval_rewards/rejected": -0.4650489091873169,
      "eval_runtime": 428.127,
      "eval_samples_per_second": 4.672,
      "eval_steps_per_second": 0.292,
      "step": 300
    },
    {
      "epoch": 0.94,
      "grad_norm": 6.7576880294714785,
      "learning_rate": 5.065954844616722e-08,
      "logits/chosen": -2.7763688564300537,
      "logits/rejected": -2.7473700046539307,
      "logps/chosen": -290.21368408203125,
      "logps/rejected": -329.4396667480469,
      "loss": 0.5829,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.13091912865638733,
      "rewards/margins": 0.3528442680835724,
      "rewards/margins_max": 0.8581522703170776,
      "rewards/margins_min": -0.14479844272136688,
      "rewards/margins_std": 0.44291505217552185,
      "rewards/rejected": -0.4837633967399597,
      "step": 310
    },
    {
      "epoch": 0.97,
      "grad_norm": 5.39765909106254,
      "learning_rate": 1.1396752298723501e-08,
      "logits/chosen": -2.820077896118164,
      "logits/rejected": -2.7777419090270996,
      "logps/chosen": -276.8497314453125,
      "logps/rejected": -314.1188049316406,
      "loss": 0.5654,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.2790452241897583,
      "rewards/margins": 0.28673863410949707,
      "rewards/margins_max": 0.7317686080932617,
      "rewards/margins_min": -0.12932872772216797,
      "rewards/margins_std": 0.3887925446033478,
      "rewards/rejected": -0.5657838582992554,
      "step": 320
    },
    {
      "epoch": 1.0,
      "step": 329,
      "total_flos": 0.0,
      "train_loss": 0.6308559574254741,
      "train_runtime": 3887.3613,
      "train_samples_per_second": 1.354,
      "train_steps_per_second": 0.085
    }
  ],
  "logging_steps": 10,
  "max_steps": 329,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}