{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9906542056074765,
  "eval_steps": 50,
  "global_step": 240,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06230529595015576,
      "grad_norm": 51.38829302969845,
      "learning_rate": 5e-07,
      "logits/chosen": -2.7407402992248535,
      "logits/rejected": -2.726320266723633,
      "logps/chosen": -261.4151916503906,
      "logps/rejected": -221.7285614013672,
      "loss": 0.692,
      "rewards/accuracies": 0.28125,
      "rewards/chosen": 0.0067472741939127445,
      "rewards/margins": -0.002469523111358285,
      "rewards/rejected": 0.009216798469424248,
      "step": 5
    },
    {
      "epoch": 0.12461059190031153,
      "grad_norm": 54.32144262101448,
      "learning_rate": 1e-06,
      "logits/chosen": -2.679516315460205,
      "logits/rejected": -2.6851108074188232,
      "logps/chosen": -244.4600067138672,
      "logps/rejected": -194.29598999023438,
      "loss": 0.6422,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 0.4551638066768646,
      "rewards/margins": 0.20188376307487488,
      "rewards/rejected": 0.2532801032066345,
      "step": 10
    },
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 44.80793063993453,
      "learning_rate": 9.988343845952696e-07,
      "logits/chosen": -2.4403023719787598,
      "logits/rejected": -2.412696361541748,
      "logps/chosen": -237.53103637695312,
      "logps/rejected": -197.9112091064453,
      "loss": 0.6663,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": 1.520341396331787,
      "rewards/margins": 0.5765382051467896,
      "rewards/rejected": 0.9438031315803528,
      "step": 15
    },
    {
      "epoch": 0.24922118380062305,
      "grad_norm": 49.11699733468642,
      "learning_rate": 9.953429730181652e-07,
      "logits/chosen": -2.2657809257507324,
      "logits/rejected": -2.2547507286071777,
      "logps/chosen": -225.1350860595703,
      "logps/rejected": -232.2615203857422,
      "loss": 0.669,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 1.3284717798233032,
      "rewards/margins": 0.6173279285430908,
      "rewards/rejected": 0.7111440896987915,
      "step": 20
    },
    {
      "epoch": 0.3115264797507788,
      "grad_norm": 43.743825005172326,
      "learning_rate": 9.895420438411615e-07,
      "logits/chosen": -2.223740577697754,
      "logits/rejected": -2.239903688430786,
      "logps/chosen": -245.8108673095703,
      "logps/rejected": -246.0238037109375,
      "loss": 0.6389,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 1.2945916652679443,
      "rewards/margins": 0.8869680166244507,
      "rewards/rejected": 0.407623827457428,
      "step": 25
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 38.88842184904586,
      "learning_rate": 9.814586436738997e-07,
      "logits/chosen": -2.302748680114746,
      "logits/rejected": -2.31158185005188,
      "logps/chosen": -246.70578002929688,
      "logps/rejected": -217.460205078125,
      "loss": 0.6192,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": 1.3854382038116455,
      "rewards/margins": 1.0373018980026245,
      "rewards/rejected": 0.34813636541366577,
      "step": 30
    },
    {
      "epoch": 0.43613707165109034,
      "grad_norm": 51.72443769242076,
      "learning_rate": 9.711304610594102e-07,
      "logits/chosen": -2.3440041542053223,
      "logits/rejected": -2.31417179107666,
      "logps/chosen": -224.3390350341797,
      "logps/rejected": -194.77621459960938,
      "loss": 0.5888,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.972150981426239,
      "rewards/margins": 0.8106788396835327,
      "rewards/rejected": 0.1614721715450287,
      "step": 35
    },
    {
      "epoch": 0.4984423676012461,
      "grad_norm": 41.52935591539919,
      "learning_rate": 9.586056507527264e-07,
      "logits/chosen": -2.383333683013916,
      "logits/rejected": -2.34763765335083,
      "logps/chosen": -246.3532257080078,
      "logps/rejected": -218.7850341796875,
      "loss": 0.6391,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 1.0201754570007324,
      "rewards/margins": 1.0443416833877563,
      "rewards/rejected": -0.02416619285941124,
      "step": 40
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 37.760052208675,
      "learning_rate": 9.439426092011875e-07,
      "logits/chosen": -2.4347915649414062,
      "logits/rejected": -2.3998470306396484,
      "logps/chosen": -232.26266479492188,
      "logps/rejected": -218.72866821289062,
      "loss": 0.5939,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 1.1044284105300903,
      "rewards/margins": 0.9438824653625488,
      "rewards/rejected": 0.16054585576057434,
      "step": 45
    },
    {
      "epoch": 0.6230529595015576,
      "grad_norm": 37.571273896281745,
      "learning_rate": 9.272097022732443e-07,
      "logits/chosen": -2.45988130569458,
      "logits/rejected": -2.4507269859313965,
      "logps/chosen": -249.6595916748047,
      "logps/rejected": -218.61581420898438,
      "loss": 0.5547,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 1.192905306816101,
      "rewards/margins": 1.339156985282898,
      "rewards/rejected": -0.14625166356563568,
      "step": 50
    },
    {
      "epoch": 0.6230529595015576,
      "eval_logits/chosen": -2.489163637161255,
      "eval_logits/rejected": -2.487579107284546,
      "eval_logps/chosen": -237.33294677734375,
      "eval_logps/rejected": -211.63194274902344,
      "eval_loss": 0.5780755281448364,
      "eval_rewards/accuracies": 0.7222222089767456,
      "eval_rewards/chosen": 0.9537798762321472,
      "eval_rewards/margins": 1.0286411046981812,
      "eval_rewards/rejected": -0.07486122101545334,
      "eval_runtime": 152.1375,
      "eval_samples_per_second": 14.986,
      "eval_steps_per_second": 0.237,
      "step": 50
    },
    {
      "epoch": 0.6853582554517134,
      "grad_norm": 54.80320093683614,
      "learning_rate": 9.084849465052209e-07,
      "logits/chosen": -2.485079765319824,
      "logits/rejected": -2.4753665924072266,
      "logps/chosen": -235.5215301513672,
      "logps/rejected": -220.1900177001953,
      "loss": 0.5909,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": 0.7905256152153015,
      "rewards/margins": 1.0766956806182861,
      "rewards/rejected": -0.286170095205307,
      "step": 55
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 39.23815457631208,
      "learning_rate": 8.878556453522099e-07,
      "logits/chosen": -2.4538919925689697,
      "logits/rejected": -2.454207181930542,
      "logps/chosen": -251.35598754882812,
      "logps/rejected": -231.8341522216797,
      "loss": 0.5733,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.9427865743637085,
      "rewards/margins": 1.1441007852554321,
      "rewards/rejected": -0.2013140171766281,
      "step": 60
    },
    {
      "epoch": 0.8099688473520249,
      "grad_norm": 47.13645309746499,
      "learning_rate": 8.654179821390621e-07,
      "logits/chosen": -2.4390835762023926,
      "logits/rejected": -2.449341297149658,
      "logps/chosen": -262.5836486816406,
      "logps/rejected": -209.0801239013672,
      "loss": 0.5441,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 1.3532841205596924,
      "rewards/margins": 1.4332042932510376,
      "rewards/rejected": -0.07991998642683029,
      "step": 65
    },
    {
      "epoch": 0.8722741433021807,
      "grad_norm": 40.00645275405077,
      "learning_rate": 8.41276571609327e-07,
      "logits/chosen": -2.4375357627868652,
      "logits/rejected": -2.4228920936584473,
      "logps/chosen": -246.0580291748047,
      "logps/rejected": -219.20260620117188,
      "loss": 0.5816,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 1.0882443189620972,
      "rewards/margins": 1.3165982961654663,
      "rewards/rejected": -0.22835393249988556,
      "step": 70
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 34.63673701954024,
      "learning_rate": 8.155439721630264e-07,
      "logits/chosen": -2.4634451866149902,
      "logits/rejected": -2.4578163623809814,
      "logps/chosen": -227.31924438476562,
      "logps/rejected": -217.36508178710938,
      "loss": 0.5334,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.8839371800422668,
      "rewards/margins": 1.3334705829620361,
      "rewards/rejected": -0.4495334029197693,
      "step": 75
    },
    {
      "epoch": 0.9968847352024922,
      "grad_norm": 36.51497796461376,
      "learning_rate": 7.883401610574336e-07,
      "logits/chosen": -2.4859111309051514,
      "logits/rejected": -2.465956211090088,
      "logps/chosen": -242.3914794921875,
      "logps/rejected": -210.991943359375,
      "loss": 0.5281,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 0.5210741758346558,
      "rewards/margins": 1.2326009273529053,
      "rewards/rejected": -0.7115266919136047,
      "step": 80
    },
    {
      "epoch": 1.0591900311526479,
      "grad_norm": 20.825789256058748,
      "learning_rate": 7.597919750177168e-07,
      "logits/chosen": -2.437802791595459,
      "logits/rejected": -2.436861515045166,
      "logps/chosen": -260.0554504394531,
      "logps/rejected": -234.86740112304688,
      "loss": 0.2507,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 0.9926292300224304,
      "rewards/margins": 2.598191261291504,
      "rewards/rejected": -1.6055622100830078,
      "step": 85
    },
    {
      "epoch": 1.1214953271028036,
      "grad_norm": 17.02799613983615,
      "learning_rate": 7.30032518865576e-07,
      "logits/chosen": -2.417109489440918,
      "logits/rejected": -2.404921770095825,
      "logps/chosen": -233.2266387939453,
      "logps/rejected": -226.57186889648438,
      "loss": 0.2109,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 1.7625309228897095,
      "rewards/margins": 3.178365707397461,
      "rewards/rejected": -1.4158347845077515,
      "step": 90
    },
    {
      "epoch": 1.1838006230529594,
      "grad_norm": 20.45306812774076,
      "learning_rate": 6.992005449231207e-07,
      "logits/chosen": -2.3889553546905518,
      "logits/rejected": -2.38620924949646,
      "logps/chosen": -251.99356079101562,
      "logps/rejected": -243.1147918701172,
      "loss": 0.2198,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.9874979257583618,
      "rewards/margins": 3.548366069793701,
      "rewards/rejected": -1.5608683824539185,
      "step": 95
    },
    {
      "epoch": 1.2461059190031152,
      "grad_norm": 16.12410058725181,
      "learning_rate": 6.67439806085493e-07,
      "logits/chosen": -2.338783025741577,
      "logits/rejected": -2.325284004211426,
      "logps/chosen": -240.7113800048828,
      "logps/rejected": -222.7356414794922,
      "loss": 0.2103,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 2.1200156211853027,
      "rewards/margins": 3.5902743339538574,
      "rewards/rejected": -1.4702587127685547,
      "step": 100
    },
    {
      "epoch": 1.2461059190031152,
      "eval_logits/chosen": -2.3138608932495117,
      "eval_logits/rejected": -2.2990024089813232,
      "eval_logps/chosen": -233.84922790527344,
      "eval_logps/rejected": -214.24310302734375,
      "eval_loss": 0.6053553819656372,
      "eval_rewards/accuracies": 0.7777777910232544,
      "eval_rewards/chosen": 1.3021522760391235,
      "eval_rewards/margins": 1.638129472732544,
      "eval_rewards/rejected": -0.33597710728645325,
      "eval_runtime": 151.0351,
      "eval_samples_per_second": 15.096,
      "eval_steps_per_second": 0.238,
      "step": 100
    },
    {
      "epoch": 1.308411214953271,
      "grad_norm": 24.653682630335922,
      "learning_rate": 6.348983855785121e-07,
      "logits/chosen": -2.2989373207092285,
      "logits/rejected": -2.290553569793701,
      "logps/chosen": -242.45651245117188,
      "logps/rejected": -214.2149658203125,
      "loss": 0.2427,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.9658845663070679,
      "rewards/margins": 3.365657091140747,
      "rewards/rejected": -1.3997727632522583,
      "step": 105
    },
    {
      "epoch": 1.3707165109034267,
      "grad_norm": 23.99763993646845,
      "learning_rate": 6.01728006526317e-07,
      "logits/chosen": -2.2708828449249268,
      "logits/rejected": -2.22404408454895,
      "logps/chosen": -240.62039184570312,
      "logps/rejected": -212.97265625,
      "loss": 0.2437,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 2.018810749053955,
      "rewards/margins": 3.461576461791992,
      "rewards/rejected": -1.442765474319458,
      "step": 110
    },
    {
      "epoch": 1.4330218068535825,
      "grad_norm": 28.797618169798415,
      "learning_rate": 5.680833245481234e-07,
      "logits/chosen": -2.270700693130493,
      "logits/rejected": -2.254040002822876,
      "logps/chosen": -244.4936981201172,
      "logps/rejected": -235.03982543945312,
      "loss": 0.3121,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 2.1543164253234863,
      "rewards/margins": 3.647374391555786,
      "rewards/rejected": -1.4930576086044312,
      "step": 115
    },
    {
      "epoch": 1.4953271028037383,
      "grad_norm": 23.473075262058984,
      "learning_rate": 5.341212066823355e-07,
      "logits/chosen": -2.3658766746520996,
      "logits/rejected": -2.3159148693084717,
      "logps/chosen": -232.2969970703125,
      "logps/rejected": -243.86593627929688,
      "loss": 0.2509,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.186516284942627,
      "rewards/margins": 3.8642489910125732,
      "rewards/rejected": -1.6777331829071045,
      "step": 120
    },
    {
      "epoch": 1.557632398753894,
      "grad_norm": 20.1682623902596,
      "learning_rate": 5e-07,
      "logits/chosen": -2.3667149543762207,
      "logits/rejected": -2.360989809036255,
      "logps/chosen": -233.6787109375,
      "logps/rejected": -218.07400512695312,
      "loss": 0.2581,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 2.068718671798706,
      "rewards/margins": 3.0430657863616943,
      "rewards/rejected": -0.9743471145629883,
      "step": 125
    },
    {
      "epoch": 1.6199376947040498,
      "grad_norm": 25.280683256501696,
      "learning_rate": 4.6587879331766457e-07,
      "logits/chosen": -2.388986825942993,
      "logits/rejected": -2.3711230754852295,
      "logps/chosen": -235.60311889648438,
      "logps/rejected": -224.765625,
      "loss": 0.2676,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.094482183456421,
      "rewards/margins": 3.472238063812256,
      "rewards/rejected": -1.377756118774414,
      "step": 130
    },
    {
      "epoch": 1.6822429906542056,
      "grad_norm": 27.72459722923311,
      "learning_rate": 4.3191667545187675e-07,
      "logits/chosen": -2.3894925117492676,
      "logits/rejected": -2.373619556427002,
      "logps/chosen": -227.37002563476562,
      "logps/rejected": -210.1580047607422,
      "loss": 0.3106,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 1.8648887872695923,
      "rewards/margins": 3.196420192718506,
      "rewards/rejected": -1.3315311670303345,
      "step": 135
    },
    {
      "epoch": 1.7445482866043613,
      "grad_norm": 31.87683565676033,
      "learning_rate": 3.9827199347368317e-07,
      "logits/chosen": -2.4143717288970947,
      "logits/rejected": -2.3826475143432617,
      "logps/chosen": -239.6693572998047,
      "logps/rejected": -235.18310546875,
      "loss": 0.3051,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.101874589920044,
      "rewards/margins": 3.821915864944458,
      "rewards/rejected": -1.7200415134429932,
      "step": 140
    },
    {
      "epoch": 1.8068535825545171,
      "grad_norm": 22.52672353938926,
      "learning_rate": 3.651016144214878e-07,
      "logits/chosen": -2.409569263458252,
      "logits/rejected": -2.3817169666290283,
      "logps/chosen": -234.31216430664062,
      "logps/rejected": -235.0973358154297,
      "loss": 0.2524,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.0497496128082275,
      "rewards/margins": 3.438047409057617,
      "rewards/rejected": -1.3882980346679688,
      "step": 145
    },
    {
      "epoch": 1.8691588785046729,
      "grad_norm": 18.620721977843708,
      "learning_rate": 3.325601939145069e-07,
      "logits/chosen": -2.389521360397339,
      "logits/rejected": -2.376431465148926,
      "logps/chosen": -227.772216796875,
      "logps/rejected": -229.7588348388672,
      "loss": 0.2095,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.8472219705581665,
      "rewards/margins": 3.8442161083221436,
      "rewards/rejected": -1.9969940185546875,
      "step": 150
    },
    {
      "epoch": 1.8691588785046729,
      "eval_logits/chosen": -2.387160539627075,
      "eval_logits/rejected": -2.3737096786499023,
      "eval_logps/chosen": -232.7999725341797,
      "eval_logps/rejected": -216.12271118164062,
      "eval_loss": 0.5997537970542908,
      "eval_rewards/accuracies": 0.7743055820465088,
      "eval_rewards/chosen": 1.4070783853530884,
      "eval_rewards/margins": 1.9310154914855957,
      "eval_rewards/rejected": -0.5239372253417969,
      "eval_runtime": 151.4198,
      "eval_samples_per_second": 15.057,
      "eval_steps_per_second": 0.238,
      "step": 150
    },
    {
      "epoch": 1.9314641744548287,
      "grad_norm": 21.107555847149946,
      "learning_rate": 3.007994550768793e-07,
      "logits/chosen": -2.34552264213562,
      "logits/rejected": -2.3452506065368652,
      "logps/chosen": -232.08834838867188,
      "logps/rejected": -241.10519409179688,
      "loss": 0.2781,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.9657140970230103,
      "rewards/margins": 3.589655637741089,
      "rewards/rejected": -1.6239420175552368,
      "step": 155
    },
    {
      "epoch": 1.9937694704049844,
      "grad_norm": 27.047203132356024,
      "learning_rate": 2.699674811344239e-07,
      "logits/chosen": -2.3365378379821777,
      "logits/rejected": -2.3133881092071533,
      "logps/chosen": -262.26690673828125,
      "logps/rejected": -240.2264862060547,
      "loss": 0.285,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": 2.5078001022338867,
      "rewards/margins": 3.6130905151367188,
      "rewards/rejected": -1.105290174484253,
      "step": 160
    },
    {
      "epoch": 2.05607476635514,
      "grad_norm": 15.009687081782982,
      "learning_rate": 2.4020802498228334e-07,
      "logits/chosen": -2.3232502937316895,
      "logits/rejected": -2.25954008102417,
      "logps/chosen": -228.8788604736328,
      "logps/rejected": -219.2600860595703,
      "loss": 0.1447,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 2.339390277862549,
      "rewards/margins": 3.8433260917663574,
      "rewards/rejected": -1.5039361715316772,
      "step": 165
    },
    {
      "epoch": 2.1183800623052957,
      "grad_norm": 13.575529236402222,
      "learning_rate": 2.1165983894256646e-07,
      "logits/chosen": -2.3263421058654785,
      "logits/rejected": -2.2757723331451416,
      "logps/chosen": -231.72366333007812,
      "logps/rejected": -209.7088623046875,
      "loss": 0.1306,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": 2.5836586952209473,
      "rewards/margins": 4.131028175354004,
      "rewards/rejected": -1.5473694801330566,
      "step": 170
    },
    {
      "epoch": 2.1806853582554515,
      "grad_norm": 21.912200560703297,
      "learning_rate": 1.8445602783697373e-07,
      "logits/chosen": -2.274022340774536,
      "logits/rejected": -2.27970290184021,
      "logps/chosen": -220.47201538085938,
      "logps/rejected": -237.96298217773438,
      "loss": 0.1521,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 2.145082950592041,
      "rewards/margins": 4.457543849945068,
      "rewards/rejected": -2.3124611377716064,
      "step": 175
    },
    {
      "epoch": 2.2429906542056073,
      "grad_norm": 13.769383169619562,
      "learning_rate": 1.5872342839067304e-07,
      "logits/chosen": -2.27915620803833,
      "logits/rejected": -2.2511637210845947,
      "logps/chosen": -234.8832550048828,
      "logps/rejected": -230.9884796142578,
      "loss": 0.1365,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 2.4758851528167725,
      "rewards/margins": 4.525899887084961,
      "rewards/rejected": -2.050014019012451,
      "step": 180
    },
    {
      "epoch": 2.305295950155763,
      "grad_norm": 15.714968300704413,
      "learning_rate": 1.3458201786093794e-07,
      "logits/chosen": -2.2769675254821777,
      "logits/rejected": -2.2372546195983887,
      "logps/chosen": -254.29257202148438,
      "logps/rejected": -231.6654815673828,
      "loss": 0.1485,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 2.103557586669922,
      "rewards/margins": 4.048580169677734,
      "rewards/rejected": -1.9450223445892334,
      "step": 185
    },
    {
      "epoch": 2.367601246105919,
      "grad_norm": 17.400239039982335,
      "learning_rate": 1.1214435464779003e-07,
      "logits/chosen": -2.257563591003418,
      "logits/rejected": -2.221086025238037,
      "logps/chosen": -246.0885009765625,
      "logps/rejected": -243.9921417236328,
      "loss": 0.1361,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 2.082960605621338,
      "rewards/margins": 4.483183860778809,
      "rewards/rejected": -2.4002232551574707,
      "step": 190
    },
    {
      "epoch": 2.4299065420560746,
      "grad_norm": 18.536501030355655,
      "learning_rate": 9.1515053494779e-08,
      "logits/chosen": -2.2309863567352295,
      "logits/rejected": -2.231677532196045,
      "logps/chosen": -252.19198608398438,
      "logps/rejected": -236.55419921875,
      "loss": 0.153,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 2.158655881881714,
      "rewards/margins": 4.387732982635498,
      "rewards/rejected": -2.229076862335205,
      "step": 195
    },
    {
      "epoch": 2.4922118380062304,
      "grad_norm": 16.28538830494891,
      "learning_rate": 7.279029772675571e-08,
      "logits/chosen": -2.2156031131744385,
      "logits/rejected": -2.206735849380493,
      "logps/chosen": -233.7646942138672,
      "logps/rejected": -236.1497039794922,
      "loss": 0.1498,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.1195907592773438,
      "rewards/margins": 3.97522234916687,
      "rewards/rejected": -1.8556314706802368,
      "step": 200
    },
    {
      "epoch": 2.4922118380062304,
      "eval_logits/chosen": -2.2468643188476562,
      "eval_logits/rejected": -2.225817918777466,
      "eval_logps/chosen": -235.95494079589844,
      "eval_logps/rejected": -220.96902465820312,
      "eval_loss": 0.5972098112106323,
      "eval_rewards/accuracies": 0.7847222089767456,
      "eval_rewards/chosen": 1.0915825366973877,
      "eval_rewards/margins": 2.100149393081665,
      "eval_rewards/rejected": -1.008566975593567,
      "eval_runtime": 151.4569,
      "eval_samples_per_second": 15.054,
      "eval_steps_per_second": 0.238,
      "step": 200
    },
    {
      "epoch": 2.554517133956386,
      "grad_norm": 21.925763245524834,
      "learning_rate": 5.605739079881239e-08,
      "logits/chosen": -2.2283740043640137,
      "logits/rejected": -2.1990256309509277,
      "logps/chosen": -223.1180877685547,
      "logps/rejected": -212.6169891357422,
      "loss": 0.16,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 1.6127411127090454,
      "rewards/margins": 3.6744377613067627,
      "rewards/rejected": -2.0616965293884277,
      "step": 205
    },
    {
      "epoch": 2.616822429906542,
      "grad_norm": 13.249950588083076,
      "learning_rate": 4.139434924727358e-08,
      "logits/chosen": -2.2427330017089844,
      "logits/rejected": -2.2058417797088623,
      "logps/chosen": -240.4376983642578,
      "logps/rejected": -229.5666046142578,
      "loss": 0.1269,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 2.0897271633148193,
      "rewards/margins": 4.362399101257324,
      "rewards/rejected": -2.272672176361084,
      "step": 210
    },
    {
      "epoch": 2.6791277258566977,
      "grad_norm": 17.32476459302876,
      "learning_rate": 2.88695389405898e-08,
      "logits/chosen": -2.241701602935791,
      "logits/rejected": -2.215984344482422,
      "logps/chosen": -217.4446258544922,
      "logps/rejected": -217.5822296142578,
      "loss": 0.1386,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 1.7757539749145508,
      "rewards/margins": 4.135173797607422,
      "rewards/rejected": -2.3594202995300293,
      "step": 215
    },
    {
      "epoch": 2.7414330218068534,
      "grad_norm": 14.04290551232067,
      "learning_rate": 1.8541356326100433e-08,
      "logits/chosen": -2.241448402404785,
      "logits/rejected": -2.20662522315979,
      "logps/chosen": -247.5508575439453,
      "logps/rejected": -217.6520233154297,
      "loss": 0.1336,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 2.255281925201416,
      "rewards/margins": 4.285429000854492,
      "rewards/rejected": -2.030146837234497,
      "step": 220
    },
    {
      "epoch": 2.803738317757009,
      "grad_norm": 12.131700338579662,
      "learning_rate": 1.0457956158838544e-08,
      "logits/chosen": -2.2289175987243652,
      "logits/rejected": -2.201352596282959,
      "logps/chosen": -231.2293701171875,
      "logps/rejected": -224.2309112548828,
      "loss": 0.1421,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.050419330596924,
      "rewards/margins": 4.02677583694458,
      "rewards/rejected": -1.9763565063476562,
      "step": 225
    },
    {
      "epoch": 2.866043613707165,
      "grad_norm": 19.353963778331245,
      "learning_rate": 4.657026981834622e-09,
      "logits/chosen": -2.2360243797302246,
      "logits/rejected": -2.2264018058776855,
      "logps/chosen": -240.57394409179688,
      "logps/rejected": -242.8083953857422,
      "loss": 0.1692,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 2.1738040447235107,
      "rewards/margins": 4.646971225738525,
      "rewards/rejected": -2.4731671810150146,
      "step": 230
    },
    {
      "epoch": 2.9283489096573208,
      "grad_norm": 16.204646402305528,
      "learning_rate": 1.165615404730369e-09,
      "logits/chosen": -2.2358551025390625,
      "logits/rejected": -2.183411121368408,
      "logps/chosen": -230.99368286132812,
      "logps/rejected": -217.8062286376953,
      "loss": 0.1504,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.135359287261963,
      "rewards/margins": 4.111185550689697,
      "rewards/rejected": -1.9758259057998657,
      "step": 235
    },
    {
      "epoch": 2.9906542056074765,
      "grad_norm": 16.826133770583578,
      "learning_rate": 0.0,
      "logits/chosen": -2.233426570892334,
      "logits/rejected": -2.203826904296875,
      "logps/chosen": -232.7841796875,
      "logps/rejected": -218.5866241455078,
      "loss": 0.1512,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 1.8168675899505615,
      "rewards/margins": 4.000540733337402,
      "rewards/rejected": -2.18367338180542,
      "step": 240
    }
  ],
  "logging_steps": 5,
  "max_steps": 240,
  "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": 2829829665718272.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}