{
  "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": 3.2870738344316774,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.700951099395752,
      "logits/rejected": -2.6252384185791016,
      "logps/chosen": -301.28485107421875,
      "logps/rejected": -281.7684020996094,
      "loss": 0.0556,
      "rewards/accuracies": 0.41874998807907104,
      "rewards/chosen": 1.3337824384507257e-05,
      "rewards/margins": 0.00020079524256289005,
      "rewards/rejected": -0.0001874573645181954,
      "step": 10
    },
    {
      "epoch": 0.045714285714285714,
      "grad_norm": 2.5214503350239097,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.641275405883789,
      "logits/rejected": -2.606044292449951,
      "logps/chosen": -278.9186706542969,
      "logps/rejected": -254.6912384033203,
      "loss": 0.0552,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.002525148680433631,
      "rewards/margins": 0.0017468780279159546,
      "rewards/rejected": 0.0007782711763866246,
      "step": 20
    },
    {
      "epoch": 0.06857142857142857,
      "grad_norm": 2.7331110185448306,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.6379804611206055,
      "logits/rejected": -2.617119550704956,
      "logps/chosen": -263.29449462890625,
      "logps/rejected": -263.40509033203125,
      "loss": 0.0535,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": 0.01388559676706791,
      "rewards/margins": 0.008006090298295021,
      "rewards/rejected": 0.00587950786575675,
      "step": 30
    },
    {
      "epoch": 0.09142857142857143,
      "grad_norm": 2.830903095609413,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.6505517959594727,
      "logits/rejected": -2.5873801708221436,
      "logps/chosen": -290.2440185546875,
      "logps/rejected": -268.03863525390625,
      "loss": 0.0502,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.037532415241003036,
      "rewards/margins": 0.04160536080598831,
      "rewards/rejected": -0.004072942305356264,
      "step": 40
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 6.7770519006865815,
      "learning_rate": 4.997124959943201e-07,
      "logits/chosen": -2.6873106956481934,
      "logits/rejected": -2.6081948280334473,
      "logps/chosen": -290.84735107421875,
      "logps/rejected": -249.636474609375,
      "loss": 0.048,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.05462931841611862,
      "rewards/margins": 0.08235044032335281,
      "rewards/rejected": -0.027721121907234192,
      "step": 50
    },
    {
      "epoch": 0.11428571428571428,
      "eval_logits/chosen": -2.552030086517334,
      "eval_logits/rejected": -2.4545059204101562,
      "eval_logps/chosen": -269.4535217285156,
      "eval_logps/rejected": -224.7078094482422,
      "eval_loss": 0.04283829778432846,
      "eval_rewards/accuracies": 0.7025862336158752,
      "eval_rewards/chosen": 0.06150422990322113,
      "eval_rewards/margins": 0.11782074719667435,
      "eval_rewards/rejected": -0.05631651729345322,
      "eval_runtime": 91.0985,
      "eval_samples_per_second": 20.099,
      "eval_steps_per_second": 0.318,
      "step": 50
    },
    {
      "epoch": 0.13714285714285715,
      "grad_norm": 3.063883877478527,
      "learning_rate": 4.979579212164186e-07,
      "logits/chosen": -2.5849452018737793,
      "logits/rejected": -2.476766347885132,
      "logps/chosen": -280.53155517578125,
      "logps/rejected": -259.91339111328125,
      "loss": 0.0438,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.00025545060634613037,
      "rewards/margins": 0.11335714906454086,
      "rewards/rejected": -0.1136125922203064,
      "step": 60
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.212195825091894,
      "learning_rate": 4.946196886175515e-07,
      "logits/chosen": -2.527914524078369,
      "logits/rejected": -2.468449354171753,
      "logps/chosen": -273.3289489746094,
      "logps/rejected": -270.74847412109375,
      "loss": 0.0414,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 0.024492865428328514,
      "rewards/margins": 0.1357099711894989,
      "rewards/rejected": -0.11121710389852524,
      "step": 70
    },
    {
      "epoch": 0.18285714285714286,
      "grad_norm": 2.9871578323865613,
      "learning_rate": 4.897191188239667e-07,
      "logits/chosen": -2.435612678527832,
      "logits/rejected": -2.3714098930358887,
      "logps/chosen": -269.2018127441406,
      "logps/rejected": -276.71612548828125,
      "loss": 0.0424,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.02172762155532837,
      "rewards/margins": 0.19234482944011688,
      "rewards/rejected": -0.21407246589660645,
      "step": 80
    },
    {
      "epoch": 0.2057142857142857,
      "grad_norm": 7.2824339333866,
      "learning_rate": 4.832875107981763e-07,
      "logits/chosen": -2.3872673511505127,
      "logits/rejected": -2.300445318222046,
      "logps/chosen": -275.88641357421875,
      "logps/rejected": -275.3648681640625,
      "loss": 0.0423,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": 0.003988614305853844,
      "rewards/margins": 0.20215804874897003,
      "rewards/rejected": -0.19816945493221283,
      "step": 90
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 3.336291542294826,
      "learning_rate": 4.753659419387223e-07,
      "logits/chosen": -2.3031492233276367,
      "logits/rejected": -2.170569658279419,
      "logps/chosen": -289.19390869140625,
      "logps/rejected": -258.89019775390625,
      "loss": 0.0399,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.028075385838747025,
      "rewards/margins": 0.2263123095035553,
      "rewards/rejected": -0.25438767671585083,
      "step": 100
    },
    {
      "epoch": 0.22857142857142856,
      "eval_logits/chosen": -2.1601011753082275,
      "eval_logits/rejected": -1.989274263381958,
      "eval_logps/chosen": -283.0706481933594,
      "eval_logps/rejected": -250.25137329101562,
      "eval_loss": 0.03850938379764557,
      "eval_rewards/accuracies": 0.75,
      "eval_rewards/chosen": -0.07466702908277512,
      "eval_rewards/margins": 0.23708537220954895,
      "eval_rewards/rejected": -0.3117523789405823,
      "eval_runtime": 91.3269,
      "eval_samples_per_second": 20.049,
      "eval_steps_per_second": 0.318,
      "step": 100
    },
    {
      "epoch": 0.25142857142857145,
      "grad_norm": 3.177801599242846,
      "learning_rate": 4.660050057270191e-07,
      "logits/chosen": -2.190119504928589,
      "logits/rejected": -2.104891061782837,
      "logps/chosen": -322.0883483886719,
      "logps/rejected": -316.6219482421875,
      "loss": 0.0407,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.05784715339541435,
      "rewards/margins": 0.12927614152431488,
      "rewards/rejected": -0.18712328374385834,
      "step": 110
    },
    {
      "epoch": 0.2742857142857143,
      "grad_norm": 3.162850908555599,
      "learning_rate": 4.5526448859687144e-07,
      "logits/chosen": -2.1028637886047363,
      "logits/rejected": -1.9295556545257568,
      "logps/chosen": -316.31634521484375,
      "logps/rejected": -253.72195434570312,
      "loss": 0.0376,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.05411834642291069,
      "rewards/margins": 0.16595926880836487,
      "rewards/rejected": -0.22007760405540466,
      "step": 120
    },
    {
      "epoch": 0.29714285714285715,
      "grad_norm": 3.4487456432507195,
      "learning_rate": 4.432129880904388e-07,
      "logits/chosen": -2.1982297897338867,
      "logits/rejected": -2.1048200130462646,
      "logps/chosen": -314.8988037109375,
      "logps/rejected": -276.97674560546875,
      "loss": 0.0373,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": -0.09742867946624756,
      "rewards/margins": 0.16763603687286377,
      "rewards/rejected": -0.26506468653678894,
      "step": 130
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.9041569099317925,
      "learning_rate": 4.299274747394055e-07,
      "logits/chosen": -2.3467020988464355,
      "logits/rejected": -2.3200185298919678,
      "logps/chosen": -315.0827941894531,
      "logps/rejected": -290.0022277832031,
      "loss": 0.0375,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.0036486200988292694,
      "rewards/margins": 0.18455657362937927,
      "rewards/rejected": -0.1809079647064209,
      "step": 140
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 3.1721506985515817,
      "learning_rate": 4.1549280046953653e-07,
      "logits/chosen": -2.452073097229004,
      "logits/rejected": -2.371002674102783,
      "logps/chosen": -296.94329833984375,
      "logps/rejected": -307.01837158203125,
      "loss": 0.0367,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.09957350790500641,
      "rewards/margins": 0.2076205015182495,
      "rewards/rejected": -0.3071940243244171,
      "step": 150
    },
    {
      "epoch": 0.34285714285714286,
      "eval_logits/chosen": -2.4217660427093506,
      "eval_logits/rejected": -2.3057165145874023,
      "eval_logps/chosen": -294.94464111328125,
      "eval_logps/rejected": -263.3892822265625,
      "eval_loss": 0.037056975066661835,
      "eval_rewards/accuracies": 0.767241358757019,
      "eval_rewards/chosen": -0.19340714812278748,
      "eval_rewards/margins": 0.24972420930862427,
      "eval_rewards/rejected": -0.44313138723373413,
      "eval_runtime": 91.4385,
      "eval_samples_per_second": 20.024,
      "eval_steps_per_second": 0.317,
      "step": 150
    },
    {
      "epoch": 0.3657142857142857,
      "grad_norm": 3.2106332008237772,
      "learning_rate": 4.000011566683401e-07,
      "logits/chosen": -2.493586540222168,
      "logits/rejected": -2.4255142211914062,
      "logps/chosen": -316.421630859375,
      "logps/rejected": -294.204345703125,
      "loss": 0.0372,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.17081338167190552,
      "rewards/margins": 0.21480269730091095,
      "rewards/rejected": -0.38561612367630005,
      "step": 160
    },
    {
      "epoch": 0.38857142857142857,
      "grad_norm": 3.305163626494215,
      "learning_rate": 3.8355148537705047e-07,
      "logits/chosen": -2.473874568939209,
      "logits/rejected": -2.4223687648773193,
      "logps/chosen": -309.713623046875,
      "logps/rejected": -291.6881103515625,
      "loss": 0.0373,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.056086935102939606,
      "rewards/margins": 0.2034837305545807,
      "rewards/rejected": -0.2595706582069397,
      "step": 170
    },
    {
      "epoch": 0.4114285714285714,
      "grad_norm": 3.211306956770945,
      "learning_rate": 3.662488473675315e-07,
      "logits/chosen": -2.486910104751587,
      "logits/rejected": -2.3967292308807373,
      "logps/chosen": -334.092041015625,
      "logps/rejected": -292.355712890625,
      "loss": 0.0373,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.03912293165922165,
      "rewards/margins": 0.2849285304546356,
      "rewards/rejected": -0.3240514397621155,
      "step": 180
    },
    {
      "epoch": 0.4342857142857143,
      "grad_norm": 3.4199998150652564,
      "learning_rate": 3.48203751140067e-07,
      "logits/chosen": -2.507291078567505,
      "logits/rejected": -2.4220941066741943,
      "logps/chosen": -285.0407409667969,
      "logps/rejected": -261.77392578125,
      "loss": 0.0365,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.08783163130283356,
      "rewards/margins": 0.1849258691072464,
      "rewards/rejected": -0.27275753021240234,
      "step": 190
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 2.966792213983748,
      "learning_rate": 3.2953144712759537e-07,
      "logits/chosen": -2.508540630340576,
      "logits/rejected": -2.376711130142212,
      "logps/chosen": -271.0993347167969,
      "logps/rejected": -254.6019744873047,
      "loss": 0.0375,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.02756611444056034,
      "rewards/margins": 0.2296477109193802,
      "rewards/rejected": -0.257213830947876,
      "step": 200
    },
    {
      "epoch": 0.45714285714285713,
      "eval_logits/chosen": -2.290728807449341,
      "eval_logits/rejected": -2.1435999870300293,
      "eval_logps/chosen": -281.01300048828125,
      "eval_logps/rejected": -252.2786407470703,
      "eval_loss": 0.03527820482850075,
      "eval_rewards/accuracies": 0.767241358757019,
      "eval_rewards/chosen": -0.054090581834316254,
      "eval_rewards/margins": 0.2779344916343689,
      "eval_rewards/rejected": -0.33202511072158813,
      "eval_runtime": 91.8763,
      "eval_samples_per_second": 19.929,
      "eval_steps_per_second": 0.316,
      "step": 200
    },
    {
      "epoch": 0.48,
      "grad_norm": 3.030911456207702,
      "learning_rate": 3.103511916141658e-07,
      "logits/chosen": -2.2013955116271973,
      "logits/rejected": -2.0953996181488037,
      "logps/chosen": -276.37432861328125,
      "logps/rejected": -280.0777893066406,
      "loss": 0.0362,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.12920817732810974,
      "rewards/margins": 0.2187701165676117,
      "rewards/rejected": -0.3479783535003662,
      "step": 210
    },
    {
      "epoch": 0.5028571428571429,
      "grad_norm": 3.521691949347102,
      "learning_rate": 2.9078548506882117e-07,
      "logits/chosen": -2.1590688228607178,
      "logits/rejected": -2.0675222873687744,
      "logps/chosen": -301.6694030761719,
      "logps/rejected": -281.8822326660156,
      "loss": 0.0387,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.16603410243988037,
      "rewards/margins": 0.1909538209438324,
      "rewards/rejected": -0.35698795318603516,
      "step": 220
    },
    {
      "epoch": 0.5257142857142857,
      "grad_norm": 3.271109178578225,
      "learning_rate": 2.709592897595191e-07,
      "logits/chosen": -2.195312023162842,
      "logits/rejected": -2.095759391784668,
      "logps/chosen": -293.96038818359375,
      "logps/rejected": -271.3498229980469,
      "loss": 0.0366,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.1290125697851181,
      "rewards/margins": 0.19161902368068695,
      "rewards/rejected": -0.32063156366348267,
      "step": 230
    },
    {
      "epoch": 0.5485714285714286,
      "grad_norm": 3.788405528579881,
      "learning_rate": 2.509992316440332e-07,
      "logits/chosen": -2.3450160026550293,
      "logits/rejected": -2.233682870864868,
      "logps/chosen": -311.3532409667969,
      "logps/rejected": -312.9554138183594,
      "loss": 0.0359,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.12643404304981232,
      "rewards/margins": 0.2889019250869751,
      "rewards/rejected": -0.4153359532356262,
      "step": 240
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 2.975794895949038,
      "learning_rate": 2.3103279163519918e-07,
      "logits/chosen": -2.361924648284912,
      "logits/rejected": -2.305227041244507,
      "logps/chosen": -292.22430419921875,
      "logps/rejected": -304.44610595703125,
      "loss": 0.0371,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.11598072946071625,
      "rewards/margins": 0.22647562623023987,
      "rewards/rejected": -0.3424563705921173,
      "step": 250
    },
    {
      "epoch": 0.5714285714285714,
      "eval_logits/chosen": -2.378539800643921,
      "eval_logits/rejected": -2.2614917755126953,
      "eval_logps/chosen": -283.72186279296875,
      "eval_logps/rejected": -254.03253173828125,
      "eval_loss": 0.0344076044857502,
      "eval_rewards/accuracies": 0.7629310488700867,
      "eval_rewards/chosen": -0.08117903023958206,
      "eval_rewards/margins": 0.2683848738670349,
      "eval_rewards/rejected": -0.3495638966560364,
      "eval_runtime": 90.3763,
      "eval_samples_per_second": 20.26,
      "eval_steps_per_second": 0.321,
      "step": 250
    },
    {
      "epoch": 0.5942857142857143,
      "grad_norm": 3.4496793515012127,
      "learning_rate": 2.1118749140573358e-07,
      "logits/chosen": -2.378929615020752,
      "logits/rejected": -2.367784023284912,
      "logps/chosen": -292.9862365722656,
      "logps/rejected": -303.8517150878906,
      "loss": 0.0378,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.1319950819015503,
      "rewards/margins": 0.17309217154979706,
      "rewards/rejected": -0.30508726835250854,
      "step": 260
    },
    {
      "epoch": 0.6171428571428571,
      "grad_norm": 3.309695226000979,
      "learning_rate": 1.9159007893272703e-07,
      "logits/chosen": -2.4140243530273438,
      "logits/rejected": -2.2664575576782227,
      "logps/chosen": -284.2572326660156,
      "logps/rejected": -267.2565612792969,
      "loss": 0.0344,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.11731357872486115,
      "rewards/margins": 0.2501775622367859,
      "rewards/rejected": -0.3674911856651306,
      "step": 270
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.6035910497695895,
      "learning_rate": 1.7236571898357766e-07,
      "logits/chosen": -2.3951687812805176,
      "logits/rejected": -2.2973971366882324,
      "logps/chosen": -286.1737365722656,
      "logps/rejected": -301.73150634765625,
      "loss": 0.0359,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.13135617971420288,
      "rewards/margins": 0.25489118695259094,
      "rewards/rejected": -0.38624733686447144,
      "step": 280
    },
    {
      "epoch": 0.6628571428571428,
      "grad_norm": 4.1650025414210665,
      "learning_rate": 1.5363719371356882e-07,
      "logits/chosen": -2.3873696327209473,
      "logits/rejected": -2.2862212657928467,
      "logps/chosen": -316.511962890625,
      "logps/rejected": -304.41741943359375,
      "loss": 0.0349,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.15606102347373962,
      "rewards/margins": 0.2326282560825348,
      "rewards/rejected": -0.3886892795562744,
      "step": 290
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 2.987221364071687,
      "learning_rate": 1.3552411848071565e-07,
      "logits/chosen": -2.387683391571045,
      "logits/rejected": -2.2113590240478516,
      "logps/chosen": -317.22491455078125,
      "logps/rejected": -293.2046813964844,
      "loss": 0.0345,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.1531573385000229,
      "rewards/margins": 0.25246819853782654,
      "rewards/rejected": -0.40562552213668823,
      "step": 300
    },
    {
      "epoch": 0.6857142857142857,
      "eval_logits/chosen": -2.3475093841552734,
      "eval_logits/rejected": -2.2130320072174072,
      "eval_logps/chosen": -282.4234313964844,
      "eval_logps/rejected": -254.02647399902344,
      "eval_loss": 0.03413059934973717,
      "eval_rewards/accuracies": 0.7456896305084229,
      "eval_rewards/chosen": -0.06819500029087067,
      "eval_rewards/margins": 0.28130805492401123,
      "eval_rewards/rejected": -0.3495030105113983,
      "eval_runtime": 90.5949,
      "eval_samples_per_second": 20.211,
      "eval_steps_per_second": 0.32,
      "step": 300
    },
    {
      "epoch": 0.7085714285714285,
      "grad_norm": 2.83762479667315,
      "learning_rate": 1.1814217788631473e-07,
      "logits/chosen": -2.3416621685028076,
      "logits/rejected": -2.2815122604370117,
      "logps/chosen": -267.0755310058594,
      "logps/rejected": -267.9870300292969,
      "loss": 0.0367,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.10771362483501434,
      "rewards/margins": 0.20020349323749542,
      "rewards/rejected": -0.30791711807250977,
      "step": 310
    },
    {
      "epoch": 0.7314285714285714,
      "grad_norm": 3.3425045367761492,
      "learning_rate": 1.0160238692045331e-07,
      "logits/chosen": -2.3024659156799316,
      "logits/rejected": -2.2111096382141113,
      "logps/chosen": -257.63531494140625,
      "logps/rejected": -272.8826904296875,
      "loss": 0.0364,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.13352924585342407,
      "rewards/margins": 0.18541577458381653,
      "rewards/rejected": -0.3189450204372406,
      "step": 320
    },
    {
      "epoch": 0.7542857142857143,
      "grad_norm": 2.769882287509293,
      "learning_rate": 8.601038193139438e-08,
      "logits/chosen": -2.4251229763031006,
      "logits/rejected": -2.29219388961792,
      "logps/chosen": -307.23333740234375,
      "logps/rejected": -280.0276794433594,
      "loss": 0.0348,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.08015141636133194,
      "rewards/margins": 0.24224629998207092,
      "rewards/rejected": -0.3223976790904999,
      "step": 330
    },
    {
      "epoch": 0.7771428571428571,
      "grad_norm": 3.0959592695193763,
      "learning_rate": 7.146574594727572e-08,
      "logits/chosen": -2.3518755435943604,
      "logits/rejected": -2.3008193969726562,
      "logps/chosen": -277.1438293457031,
      "logps/rejected": -279.57427978515625,
      "loss": 0.0353,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.11192858219146729,
      "rewards/margins": 0.2635765075683594,
      "rewards/rejected": -0.3755050301551819,
      "step": 340
    },
    {
      "epoch": 0.8,
      "grad_norm": 3.411772564497867,
      "learning_rate": 5.8061372659157306e-08,
      "logits/chosen": -2.4039082527160645,
      "logits/rejected": -2.2675061225891113,
      "logps/chosen": -300.5389099121094,
      "logps/rejected": -285.20794677734375,
      "loss": 0.0373,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.10831709951162338,
      "rewards/margins": 0.20078572630882263,
      "rewards/rejected": -0.3091028332710266,
      "step": 350
    },
    {
      "epoch": 0.8,
      "eval_logits/chosen": -2.3320775032043457,
      "eval_logits/rejected": -2.178847312927246,
      "eval_logps/chosen": -284.68194580078125,
      "eval_logps/rejected": -257.5619201660156,
      "eval_loss": 0.03409451246261597,
      "eval_rewards/accuracies": 0.7413793206214905,
      "eval_rewards/chosen": -0.09078007191419601,
      "eval_rewards/margins": 0.2940778136253357,
      "eval_rewards/rejected": -0.38485783338546753,
      "eval_runtime": 90.2939,
      "eval_samples_per_second": 20.278,
      "eval_steps_per_second": 0.321,
      "step": 350
    },
    {
      "epoch": 0.8228571428571428,
      "grad_norm": 2.905033636374422,
      "learning_rate": 4.5882873127531614e-08,
      "logits/chosen": -2.4123928546905518,
      "logits/rejected": -2.2830798625946045,
      "logps/chosen": -295.31842041015625,
      "logps/rejected": -286.4782409667969,
      "loss": 0.0342,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -0.09913380444049835,
      "rewards/margins": 0.25478315353393555,
      "rewards/rejected": -0.3539169430732727,
      "step": 360
    },
    {
      "epoch": 0.8457142857142858,
      "grad_norm": 3.0252642177231617,
      "learning_rate": 3.500802900154412e-08,
      "logits/chosen": -2.3916738033294678,
      "logits/rejected": -2.244600296020508,
      "logps/chosen": -278.41217041015625,
      "logps/rejected": -273.110595703125,
      "loss": 0.034,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.07350718975067139,
      "rewards/margins": 0.273443341255188,
      "rewards/rejected": -0.346950501203537,
      "step": 370
    },
    {
      "epoch": 0.8685714285714285,
      "grad_norm": 3.1477774029837864,
      "learning_rate": 2.550629574310309e-08,
      "logits/chosen": -2.4462573528289795,
      "logits/rejected": -2.309199094772339,
      "logps/chosen": -329.63348388671875,
      "logps/rejected": -289.4853515625,
      "loss": 0.0352,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.11577162891626358,
      "rewards/margins": 0.20508745312690735,
      "rewards/rejected": -0.32085904479026794,
      "step": 380
    },
    {
      "epoch": 0.8914285714285715,
      "grad_norm": 3.5873354511870312,
      "learning_rate": 1.7438359028687983e-08,
      "logits/chosen": -2.403637170791626,
      "logits/rejected": -2.321964740753174,
      "logps/chosen": -318.7025146484375,
      "logps/rejected": -324.0418395996094,
      "loss": 0.0358,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.07329531759023666,
      "rewards/margins": 0.21939659118652344,
      "rewards/rejected": -0.2926918864250183,
      "step": 390
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 5.030788992854401,
      "learning_rate": 1.0855747162029361e-08,
      "logits/chosen": -2.3374335765838623,
      "logits/rejected": -2.3087658882141113,
      "logps/chosen": -292.0116271972656,
      "logps/rejected": -298.6130065917969,
      "loss": 0.0367,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.13278374075889587,
      "rewards/margins": 0.18760989606380463,
      "rewards/rejected": -0.3203936815261841,
      "step": 400
    },
    {
      "epoch": 0.9142857142857143,
      "eval_logits/chosen": -2.3139522075653076,
      "eval_logits/rejected": -2.1562979221343994,
      "eval_logps/chosen": -285.25201416015625,
      "eval_logps/rejected": -258.45458984375,
      "eval_loss": 0.034024760127067566,
      "eval_rewards/accuracies": 0.7413793206214905,
      "eval_rewards/chosen": -0.09648066759109497,
      "eval_rewards/margins": 0.2973036468029022,
      "eval_rewards/rejected": -0.3937843143939972,
      "eval_runtime": 90.5725,
      "eval_samples_per_second": 20.216,
      "eval_steps_per_second": 0.32,
      "step": 400
    },
    {
      "epoch": 0.9371428571428572,
      "grad_norm": 3.2590392659418335,
      "learning_rate": 5.8005019731033615e-09,
      "logits/chosen": -2.3900809288024902,
      "logits/rejected": -2.28753924369812,
      "logps/chosen": -301.38275146484375,
      "logps/rejected": -294.66497802734375,
      "loss": 0.0352,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.13202829658985138,
      "rewards/margins": 0.21732142567634583,
      "rewards/rejected": -0.349349707365036,
      "step": 410
    },
    {
      "epoch": 0.96,
      "grad_norm": 3.0742600085581127,
      "learning_rate": 2.3049103053431886e-09,
      "logits/chosen": -2.326852798461914,
      "logits/rejected": -2.221895694732666,
      "logps/chosen": -288.41644287109375,
      "logps/rejected": -271.6980895996094,
      "loss": 0.0345,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.056683529168367386,
      "rewards/margins": 0.32042860984802246,
      "rewards/rejected": -0.37711212038993835,
      "step": 420
    },
    {
      "epoch": 0.9828571428571429,
      "grad_norm": 3.069144533981935,
      "learning_rate": 3.9129780600541397e-10,
      "logits/chosen": -2.312039613723755,
      "logits/rejected": -2.213510751724243,
      "logps/chosen": -293.47271728515625,
      "logps/rejected": -299.9048156738281,
      "loss": 0.0344,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.11240877956151962,
      "rewards/margins": 0.24994345009326935,
      "rewards/rejected": -0.3623522222042084,
      "step": 430
    },
    {
      "epoch": 0.9988571428571429,
      "step": 437,
      "total_flos": 0.0,
      "train_loss": 0.038760803656938146,
      "train_runtime": 11420.4923,
      "train_samples_per_second": 4.903,
      "train_steps_per_second": 0.038
    }
  ],
  "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
}