{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.983177570093458,
  "eval_steps": 50,
  "global_step": 399,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.037383177570093455,
      "grad_norm": 53.10315006693553,
      "learning_rate": 5e-07,
      "logits/chosen": -2.7264351844787598,
      "logits/rejected": -2.7314915657043457,
      "logps/chosen": -233.46450805664062,
      "logps/rejected": -215.2651824951172,
      "loss": 0.6911,
      "rewards/accuracies": 0.29374998807907104,
      "rewards/chosen": 0.011523213237524033,
      "rewards/margins": 0.00106804131064564,
      "rewards/rejected": 0.010455173440277576,
      "step": 5
    },
    {
      "epoch": 0.07476635514018691,
      "grad_norm": 47.36575434115236,
      "learning_rate": 1e-06,
      "logits/chosen": -2.7007861137390137,
      "logits/rejected": -2.6771092414855957,
      "logps/chosen": -243.54736328125,
      "logps/rejected": -216.7264404296875,
      "loss": 0.6571,
      "rewards/accuracies": 0.606249988079071,
      "rewards/chosen": 0.35093382000923157,
      "rewards/margins": 0.08091190457344055,
      "rewards/rejected": 0.270021915435791,
      "step": 10
    },
    {
      "epoch": 0.11214953271028037,
      "grad_norm": 47.16978215311432,
      "learning_rate": 9.995924118521016e-07,
      "logits/chosen": -2.430677652359009,
      "logits/rejected": -2.3965601921081543,
      "logps/chosen": -245.1031951904297,
      "logps/rejected": -206.2293701171875,
      "loss": 0.6246,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 1.259174108505249,
      "rewards/margins": 0.5380627512931824,
      "rewards/rejected": 0.7211112380027771,
      "step": 15
    },
    {
      "epoch": 0.14953271028037382,
      "grad_norm": 43.635723418114395,
      "learning_rate": 9.983703119207998e-07,
      "logits/chosen": -2.1696972846984863,
      "logits/rejected": -2.1355605125427246,
      "logps/chosen": -241.0774383544922,
      "logps/rejected": -203.36190795898438,
      "loss": 0.6793,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 1.169782280921936,
      "rewards/margins": 0.6064848899841309,
      "rewards/rejected": 0.5632972717285156,
      "step": 20
    },
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 48.55163726023584,
      "learning_rate": 9.963356926598848e-07,
      "logits/chosen": -2.0636093616485596,
      "logits/rejected": -2.068882942199707,
      "logps/chosen": -245.859130859375,
      "logps/rejected": -225.98214721679688,
      "loss": 0.6303,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 1.4211041927337646,
      "rewards/margins": 0.8779617547988892,
      "rewards/rejected": 0.5431426167488098,
      "step": 25
    },
    {
      "epoch": 0.22429906542056074,
      "grad_norm": 42.28573315120618,
      "learning_rate": 9.934918712161414e-07,
      "logits/chosen": -2.1142563819885254,
      "logits/rejected": -2.0855462551116943,
      "logps/chosen": -239.36471557617188,
      "logps/rejected": -208.6727294921875,
      "loss": 0.5921,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.9205316305160522,
      "rewards/margins": 0.793667733669281,
      "rewards/rejected": 0.12686386704444885,
      "step": 30
    },
    {
      "epoch": 0.2616822429906542,
      "grad_norm": 43.204924277825036,
      "learning_rate": 9.898434840212305e-07,
      "logits/chosen": -2.1356260776519775,
      "logits/rejected": -2.0982658863067627,
      "logps/chosen": -249.84848022460938,
      "logps/rejected": -232.4295196533203,
      "loss": 0.5949,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.6800674200057983,
      "rewards/margins": 0.8049441576004028,
      "rewards/rejected": -0.12487666308879852,
      "step": 35
    },
    {
      "epoch": 0.29906542056074764,
      "grad_norm": 38.34647398013497,
      "learning_rate": 9.853964792326704e-07,
      "logits/chosen": -2.1388490200042725,
      "logits/rejected": -2.106875419616699,
      "logps/chosen": -231.1354217529297,
      "logps/rejected": -210.77197265625,
      "loss": 0.6446,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.6789754033088684,
      "rewards/margins": 1.0355161428451538,
      "rewards/rejected": -0.35654082894325256,
      "step": 40
    },
    {
      "epoch": 0.3364485981308411,
      "grad_norm": 44.46300001215897,
      "learning_rate": 9.80158107036243e-07,
      "logits/chosen": -2.182988405227661,
      "logits/rejected": -2.139224052429199,
      "logps/chosen": -253.672119140625,
      "logps/rejected": -198.7143096923828,
      "loss": 0.5967,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.711867094039917,
      "rewards/margins": 0.8050382733345032,
      "rewards/rejected": -0.093171127140522,
      "step": 45
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 36.368588934194,
      "learning_rate": 9.741369078256344e-07,
      "logits/chosen": -2.1803622245788574,
      "logits/rejected": -2.1714465618133545,
      "logps/chosen": -229.3830108642578,
      "logps/rejected": -214.208251953125,
      "loss": 0.577,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.7632301449775696,
      "rewards/margins": 0.9728155136108398,
      "rewards/rejected": -0.20958539843559265,
      "step": 50
    },
    {
      "epoch": 0.37383177570093457,
      "eval_logits/chosen": -2.168900966644287,
      "eval_logits/rejected": -2.154597759246826,
      "eval_logps/chosen": -240.57708740234375,
      "eval_logps/rejected": -220.16871643066406,
      "eval_loss": 0.578223705291748,
      "eval_rewards/accuracies": 0.7250000238418579,
      "eval_rewards/chosen": 0.7959616780281067,
      "eval_rewards/margins": 0.9759488701820374,
      "eval_rewards/rejected": -0.17998719215393066,
      "eval_runtime": 252.8699,
      "eval_samples_per_second": 15.024,
      "eval_steps_per_second": 0.237,
      "step": 50
    },
    {
      "epoch": 0.411214953271028,
      "grad_norm": 36.85921784944549,
      "learning_rate": 9.673426982785825e-07,
      "logits/chosen": -2.1165783405303955,
      "logits/rejected": -2.133802890777588,
      "logps/chosen": -227.85147094726562,
      "logps/rejected": -229.4573516845703,
      "loss": 0.5854,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.668209969997406,
      "rewards/margins": 0.9655311703681946,
      "rewards/rejected": -0.29732123017311096,
      "step": 55
    },
    {
      "epoch": 0.4485981308411215,
      "grad_norm": 37.099348027626476,
      "learning_rate": 9.597865553522297e-07,
      "logits/chosen": -2.1299071311950684,
      "logits/rejected": -2.1265180110931396,
      "logps/chosen": -246.6769256591797,
      "logps/rejected": -218.6841583251953,
      "loss": 0.5468,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.3689742088317871,
      "rewards/margins": 1.112032175064087,
      "rewards/rejected": -0.7430580258369446,
      "step": 60
    },
    {
      "epoch": 0.48598130841121495,
      "grad_norm": 35.927665101781095,
      "learning_rate": 9.514807982237785e-07,
      "logits/chosen": -2.298119068145752,
      "logits/rejected": -2.2940573692321777,
      "logps/chosen": -265.78155517578125,
      "logps/rejected": -213.521240234375,
      "loss": 0.5918,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.9371916055679321,
      "rewards/margins": 1.3884761333465576,
      "rewards/rejected": -0.45128464698791504,
      "step": 65
    },
    {
      "epoch": 0.5233644859813084,
      "grad_norm": 39.11873955408514,
      "learning_rate": 9.424389682058886e-07,
      "logits/chosen": -2.3393406867980957,
      "logits/rejected": -2.309872627258301,
      "logps/chosen": -218.0382080078125,
      "logps/rejected": -194.01332092285156,
      "loss": 0.5295,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.9512729644775391,
      "rewards/margins": 1.144668698310852,
      "rewards/rejected": -0.19339559972286224,
      "step": 70
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 37.634486315960864,
      "learning_rate": 9.326758066695624e-07,
      "logits/chosen": -2.34443736076355,
      "logits/rejected": -2.325118064880371,
      "logps/chosen": -259.53143310546875,
      "logps/rejected": -198.3692626953125,
      "loss": 0.5487,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.9949586987495422,
      "rewards/margins": 1.7472797632217407,
      "rewards/rejected": -0.7523208856582642,
      "step": 75
    },
    {
      "epoch": 0.5981308411214953,
      "grad_norm": 35.50580413675745,
      "learning_rate": 9.222072310105126e-07,
      "logits/chosen": -2.3364174365997314,
      "logits/rejected": -2.312894105911255,
      "logps/chosen": -236.32666015625,
      "logps/rejected": -261.32843017578125,
      "loss": 0.6152,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 0.37070125341415405,
      "rewards/margins": 1.0360088348388672,
      "rewards/rejected": -0.6653076410293579,
      "step": 80
    },
    {
      "epoch": 0.6355140186915887,
      "grad_norm": 33.7374435363093,
      "learning_rate": 9.110503086981955e-07,
      "logits/chosen": -2.2803781032562256,
      "logits/rejected": -2.267977237701416,
      "logps/chosen": -255.5703887939453,
      "logps/rejected": -206.46066284179688,
      "loss": 0.553,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.18914642930030823,
      "rewards/margins": 1.2585302591323853,
      "rewards/rejected": -1.0693838596343994,
      "step": 85
    },
    {
      "epoch": 0.6728971962616822,
      "grad_norm": 39.897529811340966,
      "learning_rate": 8.992232294498169e-07,
      "logits/chosen": -2.1736109256744385,
      "logits/rejected": -2.1623623371124268,
      "logps/chosen": -255.2686309814453,
      "logps/rejected": -225.6684112548828,
      "loss": 0.5359,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.11371274292469025,
      "rewards/margins": 1.4066712856292725,
      "rewards/rejected": -1.2929584980010986,
      "step": 90
    },
    {
      "epoch": 0.7102803738317757,
      "grad_norm": 36.88770484599986,
      "learning_rate": 8.867452755746805e-07,
      "logits/chosen": -2.1795907020568848,
      "logits/rejected": -2.1622931957244873,
      "logps/chosen": -267.36358642578125,
      "logps/rejected": -237.27359008789062,
      "loss": 0.4889,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.24791303277015686,
      "rewards/margins": 1.365724802017212,
      "rewards/rejected": -1.6136376857757568,
      "step": 95
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 29.575103655541206,
      "learning_rate": 8.736367905372246e-07,
      "logits/chosen": -2.1824848651885986,
      "logits/rejected": -2.164578914642334,
      "logps/chosen": -262.33575439453125,
      "logps/rejected": -242.2086639404297,
      "loss": 0.5388,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.11618832498788834,
      "rewards/margins": 1.8072330951690674,
      "rewards/rejected": -1.9234212636947632,
      "step": 100
    },
    {
      "epoch": 0.7476635514018691,
      "eval_logits/chosen": -2.199050188064575,
      "eval_logits/rejected": -2.174029588699341,
      "eval_logps/chosen": -252.93431091308594,
      "eval_logps/rejected": -238.5013885498047,
      "eval_loss": 0.5390673875808716,
      "eval_rewards/accuracies": 0.7479166388511658,
      "eval_rewards/chosen": -0.4397614300251007,
      "eval_rewards/margins": 1.5734889507293701,
      "eval_rewards/rejected": -2.0132501125335693,
      "eval_runtime": 252.4297,
      "eval_samples_per_second": 15.05,
      "eval_steps_per_second": 0.238,
      "step": 100
    },
    {
      "epoch": 0.7850467289719626,
      "grad_norm": 33.93716964639979,
      "learning_rate": 8.599191457900016e-07,
      "logits/chosen": -2.2423758506774902,
      "logits/rejected": -2.215951919555664,
      "logps/chosen": -248.74136352539062,
      "logps/rejected": -238.6445770263672,
      "loss": 0.5277,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.2006298005580902,
      "rewards/margins": 1.6982009410858154,
      "rewards/rejected": -1.898830771446228,
      "step": 105
    },
    {
      "epoch": 0.822429906542056,
      "grad_norm": 34.93462881131966,
      "learning_rate": 8.456147059306757e-07,
      "logits/chosen": -2.3176093101501465,
      "logits/rejected": -2.312403440475464,
      "logps/chosen": -260.4902648925781,
      "logps/rejected": -202.03927612304688,
      "loss": 0.5658,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.13997402787208557,
      "rewards/margins": 1.6242597103118896,
      "rewards/rejected": -1.7642338275909424,
      "step": 110
    },
    {
      "epoch": 0.8598130841121495,
      "grad_norm": 33.1410955281443,
      "learning_rate": 8.307467922398432e-07,
      "logits/chosen": -2.4065065383911133,
      "logits/rejected": -2.4101977348327637,
      "logps/chosen": -249.76541137695312,
      "logps/rejected": -247.932861328125,
      "loss": 0.5152,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.18386869132518768,
      "rewards/margins": 1.8899492025375366,
      "rewards/rejected": -2.073817729949951,
      "step": 115
    },
    {
      "epoch": 0.897196261682243,
      "grad_norm": 31.607304609592937,
      "learning_rate": 8.15339644659121e-07,
      "logits/chosen": -2.4259490966796875,
      "logits/rejected": -2.41045880317688,
      "logps/chosen": -258.97723388671875,
      "logps/rejected": -221.7887420654297,
      "loss": 0.5595,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.14788323640823364,
      "rewards/margins": 1.9169807434082031,
      "rewards/rejected": -2.064863681793213,
      "step": 120
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 32.67833778299361,
      "learning_rate": 7.994183822714968e-07,
      "logits/chosen": -2.4193129539489746,
      "logits/rejected": -2.4144270420074463,
      "logps/chosen": -262.0440368652344,
      "logps/rejected": -223.1062469482422,
      "loss": 0.4636,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.24569320678710938,
      "rewards/margins": 2.0063912868499756,
      "rewards/rejected": -2.252084970474243,
      "step": 125
    },
    {
      "epoch": 0.9719626168224299,
      "grad_norm": 28.586935737564,
      "learning_rate": 7.830089623483656e-07,
      "logits/chosen": -2.431533098220825,
      "logits/rejected": -2.414226531982422,
      "logps/chosen": -250.75613403320312,
      "logps/rejected": -233.8451385498047,
      "loss": 0.4432,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.2264728844165802,
      "rewards/margins": 2.210639476776123,
      "rewards/rejected": -2.4371120929718018,
      "step": 130
    },
    {
      "epoch": 1.0093457943925233,
      "grad_norm": 21.434099176546553,
      "learning_rate": 7.661381380300253e-07,
      "logits/chosen": -2.442919969558716,
      "logits/rejected": -2.422956705093384,
      "logps/chosen": -274.38555908203125,
      "logps/rejected": -250.0774688720703,
      "loss": 0.4499,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 0.05850023776292801,
      "rewards/margins": 2.6394431591033936,
      "rewards/rejected": -2.5809426307678223,
      "step": 135
    },
    {
      "epoch": 1.0467289719626167,
      "grad_norm": 18.591470106307337,
      "learning_rate": 7.488334147086263e-07,
      "logits/chosen": -2.431138515472412,
      "logits/rejected": -2.4030303955078125,
      "logps/chosen": -250.7612762451172,
      "logps/rejected": -224.02401733398438,
      "loss": 0.2627,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 0.4705049395561218,
      "rewards/margins": 2.965265989303589,
      "rewards/rejected": -2.494760751724243,
      "step": 140
    },
    {
      "epoch": 1.0841121495327102,
      "grad_norm": 26.220775795974554,
      "learning_rate": 7.311230051846819e-07,
      "logits/chosen": -2.4043233394622803,
      "logits/rejected": -2.3778297901153564,
      "logps/chosen": -225.55380249023438,
      "logps/rejected": -237.52963256835938,
      "loss": 0.2753,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.8029943704605103,
      "rewards/margins": 3.111466646194458,
      "rewards/rejected": -2.3084726333618164,
      "step": 145
    },
    {
      "epoch": 1.1214953271028036,
      "grad_norm": 20.647616721664097,
      "learning_rate": 7.130357836702577e-07,
      "logits/chosen": -2.3966336250305176,
      "logits/rejected": -2.3554282188415527,
      "logps/chosen": -260.1661376953125,
      "logps/rejected": -240.3435821533203,
      "loss": 0.2653,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.162467360496521,
      "rewards/margins": 3.1894516944885254,
      "rewards/rejected": -2.026984214782715,
      "step": 150
    },
    {
      "epoch": 1.1214953271028036,
      "eval_logits/chosen": -2.3485267162323,
      "eval_logits/rejected": -2.32659649848938,
      "eval_logps/chosen": -245.67454528808594,
      "eval_logps/rejected": -235.21473693847656,
      "eval_loss": 0.5246723890304565,
      "eval_rewards/accuracies": 0.7645833492279053,
      "eval_rewards/chosen": 0.2862185537815094,
      "eval_rewards/margins": 1.9708030223846436,
      "eval_rewards/rejected": -1.6845842599868774,
      "eval_runtime": 253.0283,
      "eval_samples_per_second": 15.014,
      "eval_steps_per_second": 0.237,
      "step": 150
    },
    {
      "epoch": 1.158878504672897,
      "grad_norm": 23.50919746826828,
      "learning_rate": 6.946012387138247e-07,
      "logits/chosen": -2.3319590091705322,
      "logits/rejected": -2.305656909942627,
      "logps/chosen": -240.41323852539062,
      "logps/rejected": -236.7122802734375,
      "loss": 0.2499,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 0.9150522947311401,
      "rewards/margins": 3.2040953636169434,
      "rewards/rejected": -2.2890431880950928,
      "step": 155
    },
    {
      "epoch": 1.1962616822429906,
      "grad_norm": 16.10335481559186,
      "learning_rate": 6.758494251235274e-07,
      "logits/chosen": -2.288930654525757,
      "logits/rejected": -2.2800326347351074,
      "logps/chosen": -252.2129364013672,
      "logps/rejected": -242.2833251953125,
      "loss": 0.2442,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 1.06269371509552,
      "rewards/margins": 3.869558334350586,
      "rewards/rejected": -2.8068645000457764,
      "step": 160
    },
    {
      "epoch": 1.233644859813084,
      "grad_norm": 18.48336670058505,
      "learning_rate": 6.568109149672496e-07,
      "logits/chosen": -2.2865896224975586,
      "logits/rejected": -2.2526893615722656,
      "logps/chosen": -253.0467529296875,
      "logps/rejected": -230.2500762939453,
      "loss": 0.2703,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.3167855739593506,
      "rewards/margins": 3.570039749145508,
      "rewards/rejected": -2.2532541751861572,
      "step": 165
    },
    {
      "epoch": 1.2710280373831775,
      "grad_norm": 25.61953738965752,
      "learning_rate": 6.375167477293648e-07,
      "logits/chosen": -2.304429292678833,
      "logits/rejected": -2.293348789215088,
      "logps/chosen": -239.5231170654297,
      "logps/rejected": -229.12246704101562,
      "loss": 0.2725,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 1.3146635293960571,
      "rewards/margins": 3.4736862182617188,
      "rewards/rejected": -2.159022808074951,
      "step": 170
    },
    {
      "epoch": 1.308411214953271,
      "grad_norm": 25.42815054266923,
      "learning_rate": 6.179983797054321e-07,
      "logits/chosen": -2.3991851806640625,
      "logits/rejected": -2.3383266925811768,
      "logps/chosen": -236.6082000732422,
      "logps/rejected": -269.2547607421875,
      "loss": 0.2646,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.4701949954032898,
      "rewards/margins": 4.1486945152282715,
      "rewards/rejected": -3.6785004138946533,
      "step": 175
    },
    {
      "epoch": 1.3457943925233644,
      "grad_norm": 25.323140928512764,
      "learning_rate": 5.982876327173427e-07,
      "logits/chosen": -2.4732565879821777,
      "logits/rejected": -2.43239164352417,
      "logps/chosen": -274.3072814941406,
      "logps/rejected": -239.09585571289062,
      "loss": 0.3069,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 0.0698857456445694,
      "rewards/margins": 3.551178455352783,
      "rewards/rejected": -3.481292247772217,
      "step": 180
    },
    {
      "epoch": 1.3831775700934579,
      "grad_norm": 20.834596330130136,
      "learning_rate": 5.78416642232531e-07,
      "logits/chosen": -2.4427692890167236,
      "logits/rejected": -2.4098360538482666,
      "logps/chosen": -258.0860900878906,
      "logps/rejected": -243.33206176757812,
      "loss": 0.2444,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.3224690556526184,
      "rewards/margins": 3.5159637928009033,
      "rewards/rejected": -3.1934947967529297,
      "step": 185
    },
    {
      "epoch": 1.4205607476635513,
      "grad_norm": 18.851032795980213,
      "learning_rate": 5.584178049718314e-07,
      "logits/chosen": -2.4080841541290283,
      "logits/rejected": -2.3888661861419678,
      "logps/chosen": -242.3834686279297,
      "logps/rejected": -247.5576629638672,
      "loss": 0.2604,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.38662514090538025,
      "rewards/margins": 3.3747589588165283,
      "rewards/rejected": -2.988133668899536,
      "step": 190
    },
    {
      "epoch": 1.4579439252336448,
      "grad_norm": 24.324007820476073,
      "learning_rate": 5.38323726091401e-07,
      "logits/chosen": -2.436607837677002,
      "logits/rejected": -2.391335964202881,
      "logps/chosen": -250.54647827148438,
      "logps/rejected": -239.52590942382812,
      "loss": 0.3042,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": -0.005544149782508612,
      "rewards/margins": 3.300248384475708,
      "rewards/rejected": -3.3057925701141357,
      "step": 195
    },
    {
      "epoch": 1.4953271028037383,
      "grad_norm": 21.752837131187086,
      "learning_rate": 5.181671660248178e-07,
      "logits/chosen": -2.4886248111724854,
      "logits/rejected": -2.4703927040100098,
      "logps/chosen": -239.51376342773438,
      "logps/rejected": -247.77114868164062,
      "loss": 0.2571,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": -0.03603418171405792,
      "rewards/margins": 3.4513862133026123,
      "rewards/rejected": -3.4874203205108643,
      "step": 200
    },
    {
      "epoch": 1.4953271028037383,
      "eval_logits/chosen": -2.5015809535980225,
      "eval_logits/rejected": -2.475208282470703,
      "eval_logps/chosen": -254.51597595214844,
      "eval_logps/rejected": -249.1765594482422,
      "eval_loss": 0.5108085870742798,
      "eval_rewards/accuracies": 0.7791666388511658,
      "eval_rewards/chosen": -0.5979260802268982,
      "eval_rewards/margins": 2.482844591140747,
      "eval_rewards/rejected": -3.080770254135132,
      "eval_runtime": 252.5989,
      "eval_samples_per_second": 15.04,
      "eval_steps_per_second": 0.238,
      "step": 200
    },
    {
      "epoch": 1.5327102803738317,
      "grad_norm": 22.788597725922575,
      "learning_rate": 4.979809870720242e-07,
      "logits/chosen": -2.5105209350585938,
      "logits/rejected": -2.48913836479187,
      "logps/chosen": -252.13818359375,
      "logps/rejected": -243.4073944091797,
      "loss": 0.2725,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": -0.011826371774077415,
      "rewards/margins": 3.6809163093566895,
      "rewards/rejected": -3.6927428245544434,
      "step": 205
    },
    {
      "epoch": 1.5700934579439252,
      "grad_norm": 22.065080050736203,
      "learning_rate": 4.777980998221901e-07,
      "logits/chosen": -2.4797415733337402,
      "logits/rejected": -2.4482269287109375,
      "logps/chosen": -224.90585327148438,
      "logps/rejected": -238.3304443359375,
      "loss": 0.283,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.0005477949744090438,
      "rewards/margins": 3.547239303588867,
      "rewards/rejected": -3.5477871894836426,
      "step": 210
    },
    {
      "epoch": 1.6074766355140186,
      "grad_norm": 24.51095857605576,
      "learning_rate": 4.5765140949784923e-07,
      "logits/chosen": -2.4635329246520996,
      "logits/rejected": -2.4240822792053223,
      "logps/chosen": -256.9695739746094,
      "logps/rejected": -243.4046630859375,
      "loss": 0.3058,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 0.43238821625709534,
      "rewards/margins": 3.786862850189209,
      "rewards/rejected": -3.3544750213623047,
      "step": 215
    },
    {
      "epoch": 1.644859813084112,
      "grad_norm": 20.419017482003486,
      "learning_rate": 4.3757376230778383e-07,
      "logits/chosen": -2.4464869499206543,
      "logits/rejected": -2.394662618637085,
      "logps/chosen": -249.76754760742188,
      "logps/rejected": -256.5289306640625,
      "loss": 0.2327,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.8897647857666016,
      "rewards/margins": 4.213340759277344,
      "rewards/rejected": -3.3235764503479004,
      "step": 220
    },
    {
      "epoch": 1.6822429906542056,
      "grad_norm": 53.385379065169666,
      "learning_rate": 4.1759789189612333e-07,
      "logits/chosen": -2.394674777984619,
      "logits/rejected": -2.3603250980377197,
      "logps/chosen": -253.1683349609375,
      "logps/rejected": -259.30975341796875,
      "loss": 0.2888,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.5363051891326904,
      "rewards/margins": 3.8216965198516846,
      "rewards/rejected": -3.285390853881836,
      "step": 225
    },
    {
      "epoch": 1.719626168224299,
      "grad_norm": 19.461317653850365,
      "learning_rate": 3.9775636597496285e-07,
      "logits/chosen": -2.3545804023742676,
      "logits/rejected": -2.356778621673584,
      "logps/chosen": -246.68734741210938,
      "logps/rejected": -241.13998413085938,
      "loss": 0.2588,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 0.10117466747760773,
      "rewards/margins": 3.4182426929473877,
      "rewards/rejected": -3.317068099975586,
      "step": 230
    },
    {
      "epoch": 1.7570093457943925,
      "grad_norm": 22.752036394982305,
      "learning_rate": 3.7808153322750893e-07,
      "logits/chosen": -2.379742383956909,
      "logits/rejected": -2.3354058265686035,
      "logps/chosen": -243.09249877929688,
      "logps/rejected": -249.0592498779297,
      "loss": 0.2849,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.22115659713745117,
      "rewards/margins": 3.8574492931365967,
      "rewards/rejected": -3.6362926959991455,
      "step": 235
    },
    {
      "epoch": 1.794392523364486,
      "grad_norm": 25.661272927251265,
      "learning_rate": 3.586054705683208e-07,
      "logits/chosen": -2.387585163116455,
      "logits/rejected": -2.333857774734497,
      "logps/chosen": -273.67315673828125,
      "logps/rejected": -271.0914306640625,
      "loss": 0.2909,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 0.6078623533248901,
      "rewards/margins": 3.6751747131347656,
      "rewards/rejected": -3.0673117637634277,
      "step": 240
    },
    {
      "epoch": 1.8317757009345794,
      "grad_norm": 25.180396150597772,
      "learning_rate": 3.393599308466285e-07,
      "logits/chosen": -2.401310920715332,
      "logits/rejected": -2.336974620819092,
      "logps/chosen": -256.4820251464844,
      "logps/rejected": -246.9813232421875,
      "loss": 0.282,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 0.10372234880924225,
      "rewards/margins": 3.3294150829315186,
      "rewards/rejected": -3.2256927490234375,
      "step": 245
    },
    {
      "epoch": 1.8691588785046729,
      "grad_norm": 18.724955408080817,
      "learning_rate": 3.203762910779944e-07,
      "logits/chosen": -2.404759645462036,
      "logits/rejected": -2.3678829669952393,
      "logps/chosen": -253.46896362304688,
      "logps/rejected": -234.8633575439453,
      "loss": 0.2803,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 0.2244938164949417,
      "rewards/margins": 3.192117214202881,
      "rewards/rejected": -2.967623233795166,
      "step": 250
    },
    {
      "epoch": 1.8691588785046729,
      "eval_logits/chosen": -2.4107203483581543,
      "eval_logits/rejected": -2.3852570056915283,
      "eval_logps/chosen": -251.44602966308594,
      "eval_logps/rejected": -245.23483276367188,
      "eval_loss": 0.48170700669288635,
      "eval_rewards/accuracies": 0.7854166626930237,
      "eval_rewards/chosen": -0.2909303605556488,
      "eval_rewards/margins": 2.3956663608551025,
      "eval_rewards/rejected": -2.686596632003784,
      "eval_runtime": 252.2818,
      "eval_samples_per_second": 15.059,
      "eval_steps_per_second": 0.238,
      "step": 250
    },
    {
      "epoch": 1.9065420560747663,
      "grad_norm": 29.924010084948176,
      "learning_rate": 3.0168550128871264e-07,
      "logits/chosen": -2.4268598556518555,
      "logits/rejected": -2.3924014568328857,
      "logps/chosen": -237.61965942382812,
      "logps/rejected": -245.6161346435547,
      "loss": 0.2684,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 0.2175990641117096,
      "rewards/margins": 3.640658140182495,
      "rewards/rejected": -3.4230587482452393,
      "step": 255
    },
    {
      "epoch": 1.9439252336448598,
      "grad_norm": 26.71214593633519,
      "learning_rate": 2.833180340563554e-07,
      "logits/chosen": -2.425642728805542,
      "logits/rejected": -2.399965763092041,
      "logps/chosen": -231.1629180908203,
      "logps/rejected": -250.3664093017578,
      "loss": 0.2788,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 0.05841640383005142,
      "rewards/margins": 3.254362106323242,
      "rewards/rejected": -3.1959457397460938,
      "step": 260
    },
    {
      "epoch": 1.9813084112149533,
      "grad_norm": 23.089942965590083,
      "learning_rate": 2.653038348287261e-07,
      "logits/chosen": -2.4249393939971924,
      "logits/rejected": -2.3924624919891357,
      "logps/chosen": -267.17547607421875,
      "logps/rejected": -267.37066650390625,
      "loss": 0.2624,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 0.4873844087123871,
      "rewards/margins": 3.554990291595459,
      "rewards/rejected": -3.0676064491271973,
      "step": 265
    },
    {
      "epoch": 2.0186915887850465,
      "grad_norm": 24.317059270551702,
      "learning_rate": 2.476722731022207e-07,
      "logits/chosen": -2.397399663925171,
      "logits/rejected": -2.3522555828094482,
      "logps/chosen": -270.63812255859375,
      "logps/rejected": -237.6252899169922,
      "loss": 0.2462,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.7779295444488525,
      "rewards/margins": 4.320176601409912,
      "rewards/rejected": -3.5422470569610596,
      "step": 270
    },
    {
      "epoch": 2.05607476635514,
      "grad_norm": 24.99157167857996,
      "learning_rate": 2.3045209453919407e-07,
      "logits/chosen": -2.3702712059020996,
      "logits/rejected": -2.362605571746826,
      "logps/chosen": -238.41159057617188,
      "logps/rejected": -242.57275390625,
      "loss": 0.1566,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 0.2242518663406372,
      "rewards/margins": 3.7089030742645264,
      "rewards/rejected": -3.4846510887145996,
      "step": 275
    },
    {
      "epoch": 2.0934579439252334,
      "grad_norm": 19.242571139212618,
      "learning_rate": 2.13671374102394e-07,
      "logits/chosen": -2.362250804901123,
      "logits/rejected": -2.343247175216675,
      "logps/chosen": -276.24969482421875,
      "logps/rejected": -272.6140441894531,
      "loss": 0.1735,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 0.516792893409729,
      "rewards/margins": 4.325055122375488,
      "rewards/rejected": -3.808262348175049,
      "step": 280
    },
    {
      "epoch": 2.130841121495327,
      "grad_norm": 15.533308886613643,
      "learning_rate": 1.9735747028287342e-07,
      "logits/chosen": -2.3508124351501465,
      "logits/rejected": -2.330045700073242,
      "logps/chosen": -239.631103515625,
      "logps/rejected": -263.2986145019531,
      "loss": 0.1593,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.07615267485380173,
      "rewards/margins": 4.113415718078613,
      "rewards/rejected": -4.037262916564941,
      "step": 285
    },
    {
      "epoch": 2.1682242990654204,
      "grad_norm": 15.699144927979777,
      "learning_rate": 1.815369804960034e-07,
      "logits/chosen": -2.35709547996521,
      "logits/rejected": -2.3328442573547363,
      "logps/chosen": -244.9105987548828,
      "logps/rejected": -255.9078826904297,
      "loss": 0.1619,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.3098369538784027,
      "rewards/margins": 4.1518449783325195,
      "rewards/rejected": -3.842008113861084,
      "step": 290
    },
    {
      "epoch": 2.205607476635514,
      "grad_norm": 20.58346646707518,
      "learning_rate": 1.6623569771830852e-07,
      "logits/chosen": -2.3732151985168457,
      "logits/rejected": -2.3346047401428223,
      "logps/chosen": -247.153564453125,
      "logps/rejected": -250.949951171875,
      "loss": 0.1841,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.4144899249076843,
      "rewards/margins": 3.991154909133911,
      "rewards/rejected": -3.5766654014587402,
      "step": 295
    },
    {
      "epoch": 2.2429906542056073,
      "grad_norm": 12.509918099598698,
      "learning_rate": 1.5147856843582002e-07,
      "logits/chosen": -2.349177360534668,
      "logits/rejected": -2.3184986114501953,
      "logps/chosen": -242.14108276367188,
      "logps/rejected": -242.0617218017578,
      "loss": 0.1739,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.6371382474899292,
      "rewards/margins": 4.000860214233398,
      "rewards/rejected": -3.3637218475341797,
      "step": 300
    },
    {
      "epoch": 2.2429906542056073,
      "eval_logits/chosen": -2.355950355529785,
      "eval_logits/rejected": -2.3281238079071045,
      "eval_logps/chosen": -252.35202026367188,
      "eval_logps/rejected": -246.845947265625,
      "eval_loss": 0.49124717712402344,
      "eval_rewards/accuracies": 0.7916666865348816,
      "eval_rewards/chosen": -0.381531685590744,
      "eval_rewards/margins": 2.466177225112915,
      "eval_rewards/rejected": -2.8477089405059814,
      "eval_runtime": 252.5764,
      "eval_samples_per_second": 15.041,
      "eval_steps_per_second": 0.238,
      "step": 300
    },
    {
      "epoch": 2.2803738317757007,
      "grad_norm": 18.288756117670946,
      "learning_rate": 1.3728965197250781e-07,
      "logits/chosen": -2.35316801071167,
      "logits/rejected": -2.318678379058838,
      "logps/chosen": -256.4324645996094,
      "logps/rejected": -242.2520751953125,
      "loss": 0.1359,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.5630406141281128,
      "rewards/margins": 4.296410083770752,
      "rewards/rejected": -3.7333693504333496,
      "step": 305
    },
    {
      "epoch": 2.317757009345794,
      "grad_norm": 16.35995569615145,
      "learning_rate": 1.236920812651003e-07,
      "logits/chosen": -2.347511053085327,
      "logits/rejected": -2.3156919479370117,
      "logps/chosen": -234.4215545654297,
      "logps/rejected": -240.92129516601562,
      "loss": 0.1667,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 0.38125258684158325,
      "rewards/margins": 4.210916042327881,
      "rewards/rejected": -3.8296637535095215,
      "step": 310
    },
    {
      "epoch": 2.3551401869158877,
      "grad_norm": 15.341078879044524,
      "learning_rate": 1.1070802514823913e-07,
      "logits/chosen": -2.3151564598083496,
      "logits/rejected": -2.291027307510376,
      "logps/chosen": -245.5782928466797,
      "logps/rejected": -246.9814453125,
      "loss": 0.1652,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.24629132449626923,
      "rewards/margins": 4.182533264160156,
      "rewards/rejected": -3.9362423419952393,
      "step": 315
    },
    {
      "epoch": 2.392523364485981,
      "grad_norm": 23.7922647773062,
      "learning_rate": 9.835865221146389e-08,
      "logits/chosen": -2.301703691482544,
      "logits/rejected": -2.290492296218872,
      "logps/chosen": -223.250732421875,
      "logps/rejected": -263.16534423828125,
      "loss": 0.1591,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 0.41100868582725525,
      "rewards/margins": 4.717347145080566,
      "rewards/rejected": -4.306338310241699,
      "step": 320
    },
    {
      "epoch": 2.4299065420560746,
      "grad_norm": 16.439486015212957,
      "learning_rate": 8.666409628694693e-08,
      "logits/chosen": -2.319634199142456,
      "logits/rejected": -2.2700839042663574,
      "logps/chosen": -260.8030700683594,
      "logps/rejected": -255.59646606445312,
      "loss": 0.1595,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.4477875232696533,
      "rewards/margins": 4.375200271606445,
      "rewards/rejected": -3.927412748336792,
      "step": 325
    },
    {
      "epoch": 2.467289719626168,
      "grad_norm": 14.726547277918746,
      "learning_rate": 7.564342362424713e-08,
      "logits/chosen": -2.313271999359131,
      "logits/rejected": -2.2773654460906982,
      "logps/chosen": -262.1775817871094,
      "logps/rejected": -245.81216430664062,
      "loss": 0.1524,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 0.36651816964149475,
      "rewards/margins": 4.084885597229004,
      "rewards/rejected": -3.7183678150177,
      "step": 330
    },
    {
      "epoch": 2.5046728971962615,
      "grad_norm": 11.804281276567055,
      "learning_rate": 6.53146018056011e-08,
      "logits/chosen": -2.328040599822998,
      "logits/rejected": -2.275287389755249,
      "logps/chosen": -258.7254638671875,
      "logps/rejected": -244.4686737060547,
      "loss": 0.1497,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.21398362517356873,
      "rewards/margins": 4.004579067230225,
      "rewards/rejected": -3.790595531463623,
      "step": 335
    },
    {
      "epoch": 2.542056074766355,
      "grad_norm": 17.23503560362695,
      "learning_rate": 5.569447045242931e-08,
      "logits/chosen": -2.288806676864624,
      "logits/rejected": -2.2881455421447754,
      "logps/chosen": -238.40414428710938,
      "logps/rejected": -256.07183837890625,
      "loss": 0.1615,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.1942141056060791,
      "rewards/margins": 3.9404773712158203,
      "rewards/rejected": -3.7462635040283203,
      "step": 340
    },
    {
      "epoch": 2.5794392523364484,
      "grad_norm": 24.74669297879705,
      "learning_rate": 4.6798713770814625e-08,
      "logits/chosen": -2.3001134395599365,
      "logits/rejected": -2.2864573001861572,
      "logps/chosen": -238.6791534423828,
      "logps/rejected": -238.4395294189453,
      "loss": 0.1534,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 0.4987949728965759,
      "rewards/margins": 4.1639580726623535,
      "rewards/rejected": -3.665163040161133,
      "step": 345
    },
    {
      "epoch": 2.616822429906542,
      "grad_norm": 18.596924543370314,
      "learning_rate": 3.864183498071699e-08,
      "logits/chosen": -2.3193628787994385,
      "logits/rejected": -2.2843828201293945,
      "logps/chosen": -253.653076171875,
      "logps/rejected": -278.3705139160156,
      "loss": 0.1631,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.44818204641342163,
      "rewards/margins": 4.706423759460449,
      "rewards/rejected": -4.258241176605225,
      "step": 350
    },
    {
      "epoch": 2.616822429906542,
      "eval_logits/chosen": -2.3091881275177,
      "eval_logits/rejected": -2.2783515453338623,
      "eval_logps/chosen": -252.63780212402344,
      "eval_logps/rejected": -248.45175170898438,
      "eval_loss": 0.49646082520484924,
      "eval_rewards/accuracies": 0.7895833253860474,
      "eval_rewards/chosen": -0.4101119339466095,
      "eval_rewards/margins": 2.5981762409210205,
      "eval_rewards/rejected": -3.0082881450653076,
      "eval_runtime": 251.9742,
      "eval_samples_per_second": 15.077,
      "eval_steps_per_second": 0.238,
      "step": 350
    },
    {
      "epoch": 2.6542056074766354,
      "grad_norm": 24.678306348653066,
      "learning_rate": 3.1237132670611455e-08,
      "logits/chosen": -2.3217742443084717,
      "logits/rejected": -2.285179853439331,
      "logps/chosen": -245.7552032470703,
      "logps/rejected": -277.97393798828125,
      "loss": 0.1623,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 0.5618477463722229,
      "rewards/margins": 4.592735290527344,
      "rewards/rejected": -4.030888080596924,
      "step": 355
    },
    {
      "epoch": 2.691588785046729,
      "grad_norm": 18.396082804478937,
      "learning_rate": 2.4596679116099083e-08,
      "logits/chosen": -2.304999828338623,
      "logits/rejected": -2.295154094696045,
      "logps/chosen": -240.2224884033203,
      "logps/rejected": -248.5629425048828,
      "loss": 0.1735,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 0.357688844203949,
      "rewards/margins": 3.9913132190704346,
      "rewards/rejected": -3.63362455368042,
      "step": 360
    },
    {
      "epoch": 2.7289719626168223,
      "grad_norm": 18.24918146544934,
      "learning_rate": 1.8731300597841837e-08,
      "logits/chosen": -2.304487943649292,
      "logits/rejected": -2.260864496231079,
      "logps/chosen": -234.8140106201172,
      "logps/rejected": -260.767578125,
      "loss": 0.161,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.49128788709640503,
      "rewards/margins": 4.2879767417907715,
      "rewards/rejected": -3.7966887950897217,
      "step": 365
    },
    {
      "epoch": 2.7663551401869158,
      "grad_norm": 26.438822654562358,
      "learning_rate": 1.365055975090773e-08,
      "logits/chosen": -2.3118393421173096,
      "logits/rejected": -2.2817904949188232,
      "logps/chosen": -252.44638061523438,
      "logps/rejected": -268.03570556640625,
      "loss": 0.1686,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.5989893078804016,
      "rewards/margins": 4.2095770835876465,
      "rewards/rejected": -3.6105875968933105,
      "step": 370
    },
    {
      "epoch": 2.803738317757009,
      "grad_norm": 16.398293030996605,
      "learning_rate": 9.362739974303757e-09,
      "logits/chosen": -2.335395097732544,
      "logits/rejected": -2.2768070697784424,
      "logps/chosen": -243.970703125,
      "logps/rejected": -267.5694580078125,
      "loss": 0.1565,
      "rewards/accuracies": 0.9937499761581421,
      "rewards/chosen": 0.38310879468917847,
      "rewards/margins": 4.402381896972656,
      "rewards/rejected": -4.019272804260254,
      "step": 375
    },
    {
      "epoch": 2.8411214953271027,
      "grad_norm": 18.034442959036298,
      "learning_rate": 5.874831926114931e-09,
      "logits/chosen": -2.3112030029296875,
      "logits/rejected": -2.281367778778076,
      "logps/chosen": -240.7181854248047,
      "logps/rejected": -251.8282012939453,
      "loss": 0.1504,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.41617345809936523,
      "rewards/margins": 4.074304580688477,
      "rewards/rejected": -3.6581311225891113,
      "step": 380
    },
    {
      "epoch": 2.878504672897196,
      "grad_norm": 17.543650487814592,
      "learning_rate": 3.192522126266861e-09,
      "logits/chosen": -2.3081631660461426,
      "logits/rejected": -2.264885902404785,
      "logps/chosen": -244.57107543945312,
      "logps/rejected": -248.1013946533203,
      "loss": 0.1641,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 0.41191452741622925,
      "rewards/margins": 4.37111759185791,
      "rewards/rejected": -3.9592037200927734,
      "step": 385
    },
    {
      "epoch": 2.9158878504672896,
      "grad_norm": 15.080570821768335,
      "learning_rate": 1.3201836854931924e-09,
      "logits/chosen": -2.3092455863952637,
      "logits/rejected": -2.2830567359924316,
      "logps/chosen": -241.13278198242188,
      "logps/rejected": -266.1209716796875,
      "loss": 0.1715,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 0.32842570543289185,
      "rewards/margins": 4.37562370300293,
      "rewards/rejected": -4.047197341918945,
      "step": 390
    },
    {
      "epoch": 2.953271028037383,
      "grad_norm": 20.03660661880128,
      "learning_rate": 2.6086917562317957e-10,
      "logits/chosen": -2.3180956840515137,
      "logits/rejected": -2.277705669403076,
      "logps/chosen": -252.68295288085938,
      "logps/rejected": -274.5094909667969,
      "loss": 0.1565,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.44486260414123535,
      "rewards/margins": 4.4007062911987305,
      "rewards/rejected": -3.955843687057495,
      "step": 395
    },
    {
      "epoch": 2.983177570093458,
      "step": 399,
      "total_flos": 4704901791744000.0,
      "train_loss": 0.3356622438084213,
      "train_runtime": 15341.8197,
      "train_samples_per_second": 6.686,
      "train_steps_per_second": 0.026
    }
  ],
  "logging_steps": 5,
  "max_steps": 399,
  "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": 4704901791744000.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}