{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9988571428571429,
  "eval_steps": 50,
  "global_step": 437,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.022857142857142857,
      "grad_norm": 6.122519788916474,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.653984546661377,
      "logits/rejected": -2.6347525119781494,
      "logps/chosen": -287.884765625,
      "logps/rejected": -266.53350830078125,
      "loss": 0.6931,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 0.0003025080659426749,
      "rewards/margins": 0.0007510354043915868,
      "rewards/rejected": -0.0004485272220335901,
      "step": 10
    },
    {
      "epoch": 0.045714285714285714,
      "grad_norm": 5.658247958378644,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.655301332473755,
      "logits/rejected": -2.6238667964935303,
      "logps/chosen": -271.40631103515625,
      "logps/rejected": -260.2054748535156,
      "loss": 0.6924,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.0030373495537787676,
      "rewards/margins": 0.001921061659231782,
      "rewards/rejected": 0.0011162875453010201,
      "step": 20
    },
    {
      "epoch": 0.06857142857142857,
      "grad_norm": 5.381026859065415,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.6506433486938477,
      "logits/rejected": -2.55997896194458,
      "logps/chosen": -271.79742431640625,
      "logps/rejected": -254.66366577148438,
      "loss": 0.689,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.020913880318403244,
      "rewards/margins": 0.012432877905666828,
      "rewards/rejected": 0.008481001481413841,
      "step": 30
    },
    {
      "epoch": 0.09142857142857143,
      "grad_norm": 4.9623708021127495,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.615001916885376,
      "logits/rejected": -2.600529909133911,
      "logps/chosen": -261.85809326171875,
      "logps/rejected": -255.6659698486328,
      "loss": 0.6791,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.0391639769077301,
      "rewards/margins": 0.04127795249223709,
      "rewards/rejected": -0.0021139776799827814,
      "step": 40
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 6.716802270370761,
      "learning_rate": 4.997124959943201e-07,
      "logits/chosen": -2.596109390258789,
      "logits/rejected": -2.527108907699585,
      "logps/chosen": -288.6273193359375,
      "logps/rejected": -264.6972351074219,
      "loss": 0.6615,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.008229808881878853,
      "rewards/margins": 0.11277468502521515,
      "rewards/rejected": -0.10454487800598145,
      "step": 50
    },
    {
      "epoch": 0.11428571428571428,
      "eval_logits/chosen": -2.622114658355713,
      "eval_logits/rejected": -2.575448989868164,
      "eval_logps/chosen": -264.4378662109375,
      "eval_logps/rejected": -265.5943603515625,
      "eval_loss": 0.6475189328193665,
      "eval_rewards/accuracies": 0.6594827771186829,
      "eval_rewards/chosen": -0.04476550966501236,
      "eval_rewards/margins": 0.1284063458442688,
      "eval_rewards/rejected": -0.17317183315753937,
      "eval_runtime": 93.389,
      "eval_samples_per_second": 19.606,
      "eval_steps_per_second": 0.311,
      "step": 50
    },
    {
      "epoch": 0.13714285714285715,
      "grad_norm": 10.665432103556162,
      "learning_rate": 4.979579212164186e-07,
      "logits/chosen": -2.6227073669433594,
      "logits/rejected": -2.568916082382202,
      "logps/chosen": -306.7157287597656,
      "logps/rejected": -274.0307312011719,
      "loss": 0.6358,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.07229714095592499,
      "rewards/margins": 0.2034287005662918,
      "rewards/rejected": -0.2757258713245392,
      "step": 60
    },
    {
      "epoch": 0.16,
      "grad_norm": 14.460690059842785,
      "learning_rate": 4.946196886175515e-07,
      "logits/chosen": -2.6028428077697754,
      "logits/rejected": -2.5148682594299316,
      "logps/chosen": -322.69781494140625,
      "logps/rejected": -301.1668395996094,
      "loss": 0.6208,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.1934281885623932,
      "rewards/margins": 0.33089178800582886,
      "rewards/rejected": -0.5243200063705444,
      "step": 70
    },
    {
      "epoch": 0.18285714285714286,
      "grad_norm": 13.62881948148791,
      "learning_rate": 4.897191188239667e-07,
      "logits/chosen": -2.143632411956787,
      "logits/rejected": -2.028029203414917,
      "logps/chosen": -308.3243713378906,
      "logps/rejected": -315.1154479980469,
      "loss": 0.6067,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.4112843871116638,
      "rewards/margins": 0.27252885699272156,
      "rewards/rejected": -0.683813214302063,
      "step": 80
    },
    {
      "epoch": 0.2057142857142857,
      "grad_norm": 14.003540136427612,
      "learning_rate": 4.832875107981763e-07,
      "logits/chosen": -0.7628787755966187,
      "logits/rejected": -0.48756521940231323,
      "logps/chosen": -307.3790588378906,
      "logps/rejected": -373.88568115234375,
      "loss": 0.5805,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.657852053642273,
      "rewards/margins": 0.5630446672439575,
      "rewards/rejected": -1.22089684009552,
      "step": 90
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 27.196144010198164,
      "learning_rate": 4.753659419387223e-07,
      "logits/chosen": -0.3219766616821289,
      "logits/rejected": 0.034670639783144,
      "logps/chosen": -338.6977233886719,
      "logps/rejected": -383.44903564453125,
      "loss": 0.5507,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.7744790315628052,
      "rewards/margins": 0.7537156343460083,
      "rewards/rejected": -1.528194546699524,
      "step": 100
    },
    {
      "epoch": 0.22857142857142856,
      "eval_logits/chosen": -0.1843080073595047,
      "eval_logits/rejected": 0.2567289173603058,
      "eval_logps/chosen": -335.9164123535156,
      "eval_logps/rejected": -404.4942626953125,
      "eval_loss": 0.5504318475723267,
      "eval_rewards/accuracies": 0.6896551847457886,
      "eval_rewards/chosen": -0.7595511674880981,
      "eval_rewards/margins": 0.8026195764541626,
      "eval_rewards/rejected": -1.5621706247329712,
      "eval_runtime": 94.019,
      "eval_samples_per_second": 19.475,
      "eval_steps_per_second": 0.308,
      "step": 100
    },
    {
      "epoch": 0.25142857142857145,
      "grad_norm": 20.09944606036678,
      "learning_rate": 4.660050057270191e-07,
      "logits/chosen": 0.22648504376411438,
      "logits/rejected": 0.9535753130912781,
      "logps/chosen": -380.6333923339844,
      "logps/rejected": -404.15185546875,
      "loss": 0.5258,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.7437217831611633,
      "rewards/margins": 0.9534850120544434,
      "rewards/rejected": -1.6972068548202515,
      "step": 110
    },
    {
      "epoch": 0.2742857142857143,
      "grad_norm": 18.922641880361788,
      "learning_rate": 4.5526448859687144e-07,
      "logits/chosen": 0.726629376411438,
      "logits/rejected": 1.2900234460830688,
      "logps/chosen": -386.16644287109375,
      "logps/rejected": -446.7608947753906,
      "loss": 0.5563,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -1.1333491802215576,
      "rewards/margins": 0.8371905088424683,
      "rewards/rejected": -1.9705396890640259,
      "step": 120
    },
    {
      "epoch": 0.29714285714285715,
      "grad_norm": 18.05413983916241,
      "learning_rate": 4.432129880904388e-07,
      "logits/chosen": 1.176317572593689,
      "logits/rejected": 1.9044733047485352,
      "logps/chosen": -381.20587158203125,
      "logps/rejected": -410.0304260253906,
      "loss": 0.5387,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": -1.07724130153656,
      "rewards/margins": 0.6341379284858704,
      "rewards/rejected": -1.7113792896270752,
      "step": 130
    },
    {
      "epoch": 0.32,
      "grad_norm": 16.462235206322415,
      "learning_rate": 4.299274747394055e-07,
      "logits/chosen": 1.4854234457015991,
      "logits/rejected": 2.0871422290802,
      "logps/chosen": -424.3130798339844,
      "logps/rejected": -455.3707580566406,
      "loss": 0.5369,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -1.0648014545440674,
      "rewards/margins": 0.886120617389679,
      "rewards/rejected": -1.9509223699569702,
      "step": 140
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 28.494521300429405,
      "learning_rate": 4.1549280046953653e-07,
      "logits/chosen": 1.2995119094848633,
      "logits/rejected": 1.8770694732666016,
      "logps/chosen": -386.13653564453125,
      "logps/rejected": -437.58349609375,
      "loss": 0.521,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.9008251428604126,
      "rewards/margins": 0.8192182779312134,
      "rewards/rejected": -1.7200431823730469,
      "step": 150
    },
    {
      "epoch": 0.34285714285714286,
      "eval_logits/chosen": 1.1096802949905396,
      "eval_logits/rejected": 1.8187731504440308,
      "eval_logps/chosen": -354.0176086425781,
      "eval_logps/rejected": -437.4890441894531,
      "eval_loss": 0.5233712196350098,
      "eval_rewards/accuracies": 0.7198275923728943,
      "eval_rewards/chosen": -0.9405630826950073,
      "eval_rewards/margins": 0.9515554308891296,
      "eval_rewards/rejected": -1.8921183347702026,
      "eval_runtime": 95.0191,
      "eval_samples_per_second": 19.27,
      "eval_steps_per_second": 0.305,
      "step": 150
    },
    {
      "epoch": 0.3657142857142857,
      "grad_norm": 19.256664361058306,
      "learning_rate": 4.000011566683401e-07,
      "logits/chosen": 1.1306709051132202,
      "logits/rejected": 1.9139797687530518,
      "logps/chosen": -414.0078125,
      "logps/rejected": -468.516357421875,
      "loss": 0.511,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -1.0485427379608154,
      "rewards/margins": 0.9393121600151062,
      "rewards/rejected": -1.9878549575805664,
      "step": 160
    },
    {
      "epoch": 0.38857142857142857,
      "grad_norm": 24.708541517576087,
      "learning_rate": 3.8355148537705047e-07,
      "logits/chosen": 1.1292288303375244,
      "logits/rejected": 1.7640559673309326,
      "logps/chosen": -385.39031982421875,
      "logps/rejected": -465.3546447753906,
      "loss": 0.5204,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.91693514585495,
      "rewards/margins": 0.99778813123703,
      "rewards/rejected": -1.9147233963012695,
      "step": 170
    },
    {
      "epoch": 0.4114285714285714,
      "grad_norm": 19.42104847899406,
      "learning_rate": 3.662488473675315e-07,
      "logits/chosen": 1.3040357828140259,
      "logits/rejected": 2.3787357807159424,
      "logps/chosen": -400.56915283203125,
      "logps/rejected": -451.06890869140625,
      "loss": 0.519,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.9812885522842407,
      "rewards/margins": 1.0648243427276611,
      "rewards/rejected": -2.0461127758026123,
      "step": 180
    },
    {
      "epoch": 0.4342857142857143,
      "grad_norm": 23.726334837975923,
      "learning_rate": 3.48203751140067e-07,
      "logits/chosen": 1.348807692527771,
      "logits/rejected": 2.427119731903076,
      "logps/chosen": -415.06341552734375,
      "logps/rejected": -491.79290771484375,
      "loss": 0.517,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -1.542277455329895,
      "rewards/margins": 1.0263307094573975,
      "rewards/rejected": -2.568608283996582,
      "step": 190
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 19.830821329672027,
      "learning_rate": 3.2953144712759537e-07,
      "logits/chosen": 1.331316351890564,
      "logits/rejected": 2.264349937438965,
      "logps/chosen": -363.25860595703125,
      "logps/rejected": -451.0553283691406,
      "loss": 0.5162,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -1.1292212009429932,
      "rewards/margins": 0.9935659170150757,
      "rewards/rejected": -2.1227869987487793,
      "step": 200
    },
    {
      "epoch": 0.45714285714285713,
      "eval_logits/chosen": 1.2298647165298462,
      "eval_logits/rejected": 2.2567214965820312,
      "eval_logps/chosen": -383.05267333984375,
      "eval_logps/rejected": -473.64532470703125,
      "eval_loss": 0.5107715129852295,
      "eval_rewards/accuracies": 0.732758641242981,
      "eval_rewards/chosen": -1.230913519859314,
      "eval_rewards/margins": 1.0227679014205933,
      "eval_rewards/rejected": -2.2536816596984863,
      "eval_runtime": 94.5713,
      "eval_samples_per_second": 19.361,
      "eval_steps_per_second": 0.307,
      "step": 200
    },
    {
      "epoch": 0.48,
      "grad_norm": 22.298035621509914,
      "learning_rate": 3.103511916141658e-07,
      "logits/chosen": 1.5307371616363525,
      "logits/rejected": 2.729835033416748,
      "logps/chosen": -424.962890625,
      "logps/rejected": -514.0426025390625,
      "loss": 0.5071,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.3911924362182617,
      "rewards/margins": 1.1286324262619019,
      "rewards/rejected": -2.5198252201080322,
      "step": 210
    },
    {
      "epoch": 0.5028571428571429,
      "grad_norm": 14.694863409372582,
      "learning_rate": 2.9078548506882117e-07,
      "logits/chosen": 1.7180284261703491,
      "logits/rejected": 2.68153715133667,
      "logps/chosen": -447.5650939941406,
      "logps/rejected": -545.7744140625,
      "loss": 0.5181,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -1.4318263530731201,
      "rewards/margins": 1.1245338916778564,
      "rewards/rejected": -2.5563602447509766,
      "step": 220
    },
    {
      "epoch": 0.5257142857142857,
      "grad_norm": 21.866604136555903,
      "learning_rate": 2.709592897595191e-07,
      "logits/chosen": 1.7002506256103516,
      "logits/rejected": 2.5147674083709717,
      "logps/chosen": -424.66607666015625,
      "logps/rejected": -503.14837646484375,
      "loss": 0.539,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -1.3987152576446533,
      "rewards/margins": 0.8508893251419067,
      "rewards/rejected": -2.2496042251586914,
      "step": 230
    },
    {
      "epoch": 0.5485714285714286,
      "grad_norm": 23.461922171727267,
      "learning_rate": 2.509992316440332e-07,
      "logits/chosen": 2.7161831855773926,
      "logits/rejected": 3.887934446334839,
      "logps/chosen": -481.4347229003906,
      "logps/rejected": -544.0018310546875,
      "loss": 0.5234,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -1.737464189529419,
      "rewards/margins": 1.0862486362457275,
      "rewards/rejected": -2.8237128257751465,
      "step": 240
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 23.055743791457775,
      "learning_rate": 2.3103279163519918e-07,
      "logits/chosen": 2.5470688343048096,
      "logits/rejected": 3.076658248901367,
      "logps/chosen": -456.06646728515625,
      "logps/rejected": -522.9896850585938,
      "loss": 0.5186,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": -1.7880340814590454,
      "rewards/margins": 0.6319095492362976,
      "rewards/rejected": -2.419943332672119,
      "step": 250
    },
    {
      "epoch": 0.5714285714285714,
      "eval_logits/chosen": 1.5761785507202148,
      "eval_logits/rejected": 2.8322033882141113,
      "eval_logps/chosen": -396.9323425292969,
      "eval_logps/rejected": -496.5982971191406,
      "eval_loss": 0.5027161836624146,
      "eval_rewards/accuracies": 0.7025862336158752,
      "eval_rewards/chosen": -1.3697103261947632,
      "eval_rewards/margins": 1.1135003566741943,
      "eval_rewards/rejected": -2.483210563659668,
      "eval_runtime": 93.6382,
      "eval_samples_per_second": 19.554,
      "eval_steps_per_second": 0.31,
      "step": 250
    },
    {
      "epoch": 0.5942857142857143,
      "grad_norm": 23.966486101527014,
      "learning_rate": 2.1118749140573358e-07,
      "logits/chosen": 2.044908046722412,
      "logits/rejected": 3.2726006507873535,
      "logps/chosen": -411.2015075683594,
      "logps/rejected": -446.53857421875,
      "loss": 0.514,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -1.3537694215774536,
      "rewards/margins": 0.8869277834892273,
      "rewards/rejected": -2.240697145462036,
      "step": 260
    },
    {
      "epoch": 0.6171428571428571,
      "grad_norm": 19.203425356174254,
      "learning_rate": 1.9159007893272703e-07,
      "logits/chosen": 2.4072651863098145,
      "logits/rejected": 3.3252644538879395,
      "logps/chosen": -457.94830322265625,
      "logps/rejected": -485.1842346191406,
      "loss": 0.5102,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -1.6982141733169556,
      "rewards/margins": 0.6931914687156677,
      "rewards/rejected": -2.3914055824279785,
      "step": 270
    },
    {
      "epoch": 0.64,
      "grad_norm": 17.06179349785014,
      "learning_rate": 1.7236571898357766e-07,
      "logits/chosen": 1.9787667989730835,
      "logits/rejected": 3.012749195098877,
      "logps/chosen": -375.3614807128906,
      "logps/rejected": -471.38397216796875,
      "loss": 0.514,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -1.2563382387161255,
      "rewards/margins": 1.122116208076477,
      "rewards/rejected": -2.3784542083740234,
      "step": 280
    },
    {
      "epoch": 0.6628571428571428,
      "grad_norm": 22.052326742230484,
      "learning_rate": 1.5363719371356882e-07,
      "logits/chosen": 2.3124561309814453,
      "logits/rejected": 3.362464189529419,
      "logps/chosen": -446.782958984375,
      "logps/rejected": -505.8094177246094,
      "loss": 0.496,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -1.6085598468780518,
      "rewards/margins": 0.9179994463920593,
      "rewards/rejected": -2.5265591144561768,
      "step": 290
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 21.6246754175424,
      "learning_rate": 1.3552411848071565e-07,
      "logits/chosen": 2.0221407413482666,
      "logits/rejected": 3.554128646850586,
      "logps/chosen": -473.49298095703125,
      "logps/rejected": -550.6085815429688,
      "loss": 0.4883,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -1.6014286279678345,
      "rewards/margins": 1.2605220079421997,
      "rewards/rejected": -2.861950397491455,
      "step": 300
    },
    {
      "epoch": 0.6857142857142857,
      "eval_logits/chosen": 2.1394236087799072,
      "eval_logits/rejected": 3.3275508880615234,
      "eval_logps/chosen": -413.999755859375,
      "eval_logps/rejected": -518.487548828125,
      "eval_loss": 0.4935988783836365,
      "eval_rewards/accuracies": 0.7370689511299133,
      "eval_rewards/chosen": -1.5403847694396973,
      "eval_rewards/margins": 1.1617193222045898,
      "eval_rewards/rejected": -2.702104091644287,
      "eval_runtime": 93.5025,
      "eval_samples_per_second": 19.582,
      "eval_steps_per_second": 0.31,
      "step": 300
    },
    {
      "epoch": 0.7085714285714285,
      "grad_norm": 20.645058605492473,
      "learning_rate": 1.1814217788631473e-07,
      "logits/chosen": 1.9714610576629639,
      "logits/rejected": 3.4288382530212402,
      "logps/chosen": -459.2438049316406,
      "logps/rejected": -542.9857177734375,
      "loss": 0.5034,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -1.459855079650879,
      "rewards/margins": 1.2939279079437256,
      "rewards/rejected": -2.7537827491760254,
      "step": 310
    },
    {
      "epoch": 0.7314285714285714,
      "grad_norm": 28.46682682513433,
      "learning_rate": 1.0160238692045331e-07,
      "logits/chosen": 2.388679265975952,
      "logits/rejected": 3.5533275604248047,
      "logps/chosen": -448.905029296875,
      "logps/rejected": -526.8614501953125,
      "loss": 0.5066,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -1.5178719758987427,
      "rewards/margins": 1.000969648361206,
      "rewards/rejected": -2.5188417434692383,
      "step": 320
    },
    {
      "epoch": 0.7542857142857143,
      "grad_norm": 21.745911663397,
      "learning_rate": 8.601038193139438e-08,
      "logits/chosen": 2.581099271774292,
      "logits/rejected": 3.3041462898254395,
      "logps/chosen": -381.9894714355469,
      "logps/rejected": -467.76580810546875,
      "loss": 0.5113,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -1.3630098104476929,
      "rewards/margins": 0.9106127023696899,
      "rewards/rejected": -2.273622512817383,
      "step": 330
    },
    {
      "epoch": 0.7771428571428571,
      "grad_norm": 22.95799998221957,
      "learning_rate": 7.146574594727572e-08,
      "logits/chosen": 2.3186068534851074,
      "logits/rejected": 3.4312567710876465,
      "logps/chosen": -391.4497985839844,
      "logps/rejected": -469.642333984375,
      "loss": 0.4892,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -1.4544594287872314,
      "rewards/margins": 1.1061028242111206,
      "rewards/rejected": -2.5605618953704834,
      "step": 340
    },
    {
      "epoch": 0.8,
      "grad_norm": 29.21776715425853,
      "learning_rate": 5.8061372659157306e-08,
      "logits/chosen": 2.201559066772461,
      "logits/rejected": 3.762953996658325,
      "logps/chosen": -459.7745056152344,
      "logps/rejected": -539.0149536132812,
      "loss": 0.4714,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -1.4960135221481323,
      "rewards/margins": 1.210325002670288,
      "rewards/rejected": -2.70633864402771,
      "step": 350
    },
    {
      "epoch": 0.8,
      "eval_logits/chosen": 2.394544839859009,
      "eval_logits/rejected": 3.541562557220459,
      "eval_logps/chosen": -410.9665832519531,
      "eval_logps/rejected": -516.7974853515625,
      "eval_loss": 0.4907006621360779,
      "eval_rewards/accuracies": 0.7456896305084229,
      "eval_rewards/chosen": -1.5100529193878174,
      "eval_rewards/margins": 1.1751505136489868,
      "eval_rewards/rejected": -2.6852033138275146,
      "eval_runtime": 93.0638,
      "eval_samples_per_second": 19.675,
      "eval_steps_per_second": 0.312,
      "step": 350
    },
    {
      "epoch": 0.8228571428571428,
      "grad_norm": 22.43994807327437,
      "learning_rate": 4.5882873127531614e-08,
      "logits/chosen": 2.859044075012207,
      "logits/rejected": 3.9125289916992188,
      "logps/chosen": -440.3011169433594,
      "logps/rejected": -565.3880615234375,
      "loss": 0.4928,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": -1.4779033660888672,
      "rewards/margins": 1.5023510456085205,
      "rewards/rejected": -2.9802544116973877,
      "step": 360
    },
    {
      "epoch": 0.8457142857142858,
      "grad_norm": 22.645094236252362,
      "learning_rate": 3.500802900154412e-08,
      "logits/chosen": 2.81022310256958,
      "logits/rejected": 3.5364959239959717,
      "logps/chosen": -403.3380126953125,
      "logps/rejected": -527.1024169921875,
      "loss": 0.4891,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -1.4733188152313232,
      "rewards/margins": 1.1765823364257812,
      "rewards/rejected": -2.6499013900756836,
      "step": 370
    },
    {
      "epoch": 0.8685714285714285,
      "grad_norm": 24.474487723268208,
      "learning_rate": 2.550629574310309e-08,
      "logits/chosen": 3.3984062671661377,
      "logits/rejected": 4.194012641906738,
      "logps/chosen": -419.423583984375,
      "logps/rejected": -518.23583984375,
      "loss": 0.4926,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -1.635619878768921,
      "rewards/margins": 0.9691934585571289,
      "rewards/rejected": -2.60481333732605,
      "step": 380
    },
    {
      "epoch": 0.8914285714285715,
      "grad_norm": 25.84913647388076,
      "learning_rate": 1.7438359028687983e-08,
      "logits/chosen": 2.8808953762054443,
      "logits/rejected": 3.839012861251831,
      "logps/chosen": -421.871337890625,
      "logps/rejected": -526.1373291015625,
      "loss": 0.497,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -1.5530052185058594,
      "rewards/margins": 1.1946289539337158,
      "rewards/rejected": -2.747633934020996,
      "step": 390
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 28.299780272584496,
      "learning_rate": 1.0855747162029361e-08,
      "logits/chosen": 2.4124693870544434,
      "logits/rejected": 3.1909306049346924,
      "logps/chosen": -425.010498046875,
      "logps/rejected": -537.2962646484375,
      "loss": 0.4977,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -1.5087743997573853,
      "rewards/margins": 1.1911652088165283,
      "rewards/rejected": -2.699939250946045,
      "step": 400
    },
    {
      "epoch": 0.9142857142857143,
      "eval_logits/chosen": 2.507979393005371,
      "eval_logits/rejected": 3.6379663944244385,
      "eval_logps/chosen": -418.5211486816406,
      "eval_logps/rejected": -526.59423828125,
      "eval_loss": 0.4890148639678955,
      "eval_rewards/accuracies": 0.7370689511299133,
      "eval_rewards/chosen": -1.5855982303619385,
      "eval_rewards/margins": 1.197571873664856,
      "eval_rewards/rejected": -2.783169984817505,
      "eval_runtime": 95.2901,
      "eval_samples_per_second": 19.215,
      "eval_steps_per_second": 0.304,
      "step": 400
    },
    {
      "epoch": 0.9371428571428572,
      "grad_norm": 19.845237856376826,
      "learning_rate": 5.8005019731033615e-09,
      "logits/chosen": 2.7473673820495605,
      "logits/rejected": 3.6581249237060547,
      "logps/chosen": -407.8973083496094,
      "logps/rejected": -511.965576171875,
      "loss": 0.4863,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": -1.4759327173233032,
      "rewards/margins": 1.0670742988586426,
      "rewards/rejected": -2.5430071353912354,
      "step": 410
    },
    {
      "epoch": 0.96,
      "grad_norm": 20.45469236378943,
      "learning_rate": 2.3049103053431886e-09,
      "logits/chosen": 2.3240396976470947,
      "logits/rejected": 3.4506924152374268,
      "logps/chosen": -462.15362548828125,
      "logps/rejected": -587.3756103515625,
      "loss": 0.4973,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.514866590499878,
      "rewards/margins": 1.4163448810577393,
      "rewards/rejected": -2.931211471557617,
      "step": 420
    },
    {
      "epoch": 0.9828571428571429,
      "grad_norm": 26.841510292915476,
      "learning_rate": 3.9129780600541397e-10,
      "logits/chosen": 2.6934516429901123,
      "logits/rejected": 3.5782058238983154,
      "logps/chosen": -432.9996643066406,
      "logps/rejected": -534.2493896484375,
      "loss": 0.4797,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.4740568399429321,
      "rewards/margins": 1.289287805557251,
      "rewards/rejected": -2.7633445262908936,
      "step": 430
    },
    {
      "epoch": 0.9988571428571429,
      "step": 437,
      "total_flos": 0.0,
      "train_loss": 0.5399287248912611,
      "train_runtime": 11756.1871,
      "train_samples_per_second": 4.763,
      "train_steps_per_second": 0.037
    }
  ],
  "logging_steps": 10,
  "max_steps": 437,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}