{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9967637540453075,
  "eval_steps": 100,
  "global_step": 77,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.025889967637540454,
      "grad_norm": 9.941023016098997,
      "learning_rate": 5e-08,
      "logits/chosen": 34.564327239990234,
      "logits/rejected": 21.430946350097656,
      "logps/chosen": -193.26638793945312,
      "logps/rejected": -169.8800506591797,
      "loss": 0.7032,
      "rewards/accuracies": 0.46875,
      "rewards/chosen": 0.005037403665482998,
      "rewards/margins": -0.0025963345542550087,
      "rewards/rejected": 0.007633737754076719,
      "step": 2
    },
    {
      "epoch": 0.05177993527508091,
      "grad_norm": 9.84426167557295,
      "learning_rate": 1e-07,
      "logits/chosen": 34.75300598144531,
      "logits/rejected": 25.311954498291016,
      "logps/chosen": -189.80618286132812,
      "logps/rejected": -129.24330139160156,
      "loss": 0.6961,
      "rewards/accuracies": 0.484375,
      "rewards/chosen": -0.003958990331739187,
      "rewards/margins": 0.0072118965908885,
      "rewards/rejected": -0.011170885525643826,
      "step": 4
    },
    {
      "epoch": 0.07766990291262135,
      "grad_norm": 9.460230767868643,
      "learning_rate": 1.5e-07,
      "logits/chosen": 37.28773498535156,
      "logits/rejected": 27.513395309448242,
      "logps/chosen": -206.36117553710938,
      "logps/rejected": -171.872802734375,
      "loss": 0.6974,
      "rewards/accuracies": 0.53125,
      "rewards/chosen": -0.034006692469120026,
      "rewards/margins": -0.018697379156947136,
      "rewards/rejected": -0.015309315174818039,
      "step": 6
    },
    {
      "epoch": 0.10355987055016182,
      "grad_norm": 10.186609614004015,
      "learning_rate": 2e-07,
      "logits/chosen": 36.74854278564453,
      "logits/rejected": 31.93742561340332,
      "logps/chosen": -184.870361328125,
      "logps/rejected": -124.89741516113281,
      "loss": 0.6975,
      "rewards/accuracies": 0.46875,
      "rewards/chosen": -0.013519443571567535,
      "rewards/margins": -0.024582354351878166,
      "rewards/rejected": 0.011062909848988056,
      "step": 8
    },
    {
      "epoch": 0.12944983818770225,
      "grad_norm": 9.457766196917632,
      "learning_rate": 2.5e-07,
      "logits/chosen": 36.366363525390625,
      "logits/rejected": 29.217514038085938,
      "logps/chosen": -189.72547912597656,
      "logps/rejected": -133.34400939941406,
      "loss": 0.6916,
      "rewards/accuracies": 0.484375,
      "rewards/chosen": 0.003550291759893298,
      "rewards/margins": -0.00579411955550313,
      "rewards/rejected": 0.009344411082565784,
      "step": 10
    },
    {
      "epoch": 0.1553398058252427,
      "grad_norm": 9.820715630245564,
      "learning_rate": 3e-07,
      "logits/chosen": 35.559329986572266,
      "logits/rejected": 24.949851989746094,
      "logps/chosen": -191.26722717285156,
      "logps/rejected": -136.72520446777344,
      "loss": 0.6927,
      "rewards/accuracies": 0.546875,
      "rewards/chosen": 0.020757507532835007,
      "rewards/margins": 0.039941225200891495,
      "rewards/rejected": -0.019183719530701637,
      "step": 12
    },
    {
      "epoch": 0.18122977346278318,
      "grad_norm": 10.000904242116592,
      "learning_rate": 3.5e-07,
      "logits/chosen": 36.79935073852539,
      "logits/rejected": 26.168617248535156,
      "logps/chosen": -189.4495849609375,
      "logps/rejected": -138.61245727539062,
      "loss": 0.7015,
      "rewards/accuracies": 0.453125,
      "rewards/chosen": 0.0028548368718475103,
      "rewards/margins": 0.00897835474461317,
      "rewards/rejected": -0.006123516708612442,
      "step": 14
    },
    {
      "epoch": 0.20711974110032363,
      "grad_norm": 9.798236588997984,
      "learning_rate": 4e-07,
      "logits/chosen": 38.9239387512207,
      "logits/rejected": 30.127975463867188,
      "logps/chosen": -193.46170043945312,
      "logps/rejected": -136.24960327148438,
      "loss": 0.699,
      "rewards/accuracies": 0.40625,
      "rewards/chosen": -0.002874697558581829,
      "rewards/margins": -0.022318758070468903,
      "rewards/rejected": 0.0194440595805645,
      "step": 16
    },
    {
      "epoch": 0.23300970873786409,
      "grad_norm": 9.959562684371269,
      "learning_rate": 4.5e-07,
      "logits/chosen": 34.25529479980469,
      "logits/rejected": 25.71901512145996,
      "logps/chosen": -204.4850616455078,
      "logps/rejected": -169.4137725830078,
      "loss": 0.6928,
      "rewards/accuracies": 0.453125,
      "rewards/chosen": 0.02741333469748497,
      "rewards/margins": 0.01749654859304428,
      "rewards/rejected": 0.009916786104440689,
      "step": 18
    },
    {
      "epoch": 0.2588996763754045,
      "grad_norm": 9.793123390682076,
      "learning_rate": 5e-07,
      "logits/chosen": 36.560420989990234,
      "logits/rejected": 24.802711486816406,
      "logps/chosen": -207.83480834960938,
      "logps/rejected": -159.98675537109375,
      "loss": 0.692,
      "rewards/accuracies": 0.546875,
      "rewards/chosen": 0.004708289634436369,
      "rewards/margins": 0.004980792291462421,
      "rewards/rejected": -0.00027250126004219055,
      "step": 20
    },
    {
      "epoch": 0.284789644012945,
      "grad_norm": 10.276248281197612,
      "learning_rate": 4.997252228714278e-07,
      "logits/chosen": 36.67436981201172,
      "logits/rejected": 25.816286087036133,
      "logps/chosen": -203.99571228027344,
      "logps/rejected": -157.78639221191406,
      "loss": 0.689,
      "rewards/accuracies": 0.578125,
      "rewards/chosen": 0.04165205731987953,
      "rewards/margins": 0.02663404308259487,
      "rewards/rejected": 0.015018017962574959,
      "step": 22
    },
    {
      "epoch": 0.3106796116504854,
      "grad_norm": 9.36691332286361,
      "learning_rate": 4.989014955054745e-07,
      "logits/chosen": 32.67152404785156,
      "logits/rejected": 26.35110092163086,
      "logps/chosen": -222.406982421875,
      "logps/rejected": -174.97457885742188,
      "loss": 0.6728,
      "rewards/accuracies": 0.515625,
      "rewards/chosen": 0.034607432782649994,
      "rewards/margins": 0.0227619968354702,
      "rewards/rejected": 0.011845439672470093,
      "step": 24
    },
    {
      "epoch": 0.3365695792880259,
      "grad_norm": 10.075140623553729,
      "learning_rate": 4.975306286336627e-07,
      "logits/chosen": 34.61298751831055,
      "logits/rejected": 22.848501205444336,
      "logps/chosen": -190.12135314941406,
      "logps/rejected": -169.7635955810547,
      "loss": 0.6757,
      "rewards/accuracies": 0.609375,
      "rewards/chosen": 0.05975666269659996,
      "rewards/margins": 0.05689358711242676,
      "rewards/rejected": 0.002863079309463501,
      "step": 26
    },
    {
      "epoch": 0.36245954692556637,
      "grad_norm": 9.766651095371804,
      "learning_rate": 4.956156357188939e-07,
      "logits/chosen": 37.388824462890625,
      "logits/rejected": 24.61412239074707,
      "logps/chosen": -183.52816772460938,
      "logps/rejected": -142.87033081054688,
      "loss": 0.6649,
      "rewards/accuracies": 0.703125,
      "rewards/chosen": 0.09949050843715668,
      "rewards/margins": 0.07790307700634003,
      "rewards/rejected": 0.0215874332934618,
      "step": 28
    },
    {
      "epoch": 0.3883495145631068,
      "grad_norm": 9.534993155007124,
      "learning_rate": 4.931607263312032e-07,
      "logits/chosen": 35.072879791259766,
      "logits/rejected": 23.343503952026367,
      "logps/chosen": -201.98580932617188,
      "logps/rejected": -145.196044921875,
      "loss": 0.6501,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 0.16385386884212494,
      "rewards/margins": 0.10670199245214462,
      "rewards/rejected": 0.05715188756585121,
      "step": 30
    },
    {
      "epoch": 0.41423948220064727,
      "grad_norm": 9.757243010304897,
      "learning_rate": 4.9017129689421e-07,
      "logits/chosen": 38.24150466918945,
      "logits/rejected": 25.80931854248047,
      "logps/chosen": -182.25128173828125,
      "logps/rejected": -137.34368896484375,
      "loss": 0.65,
      "rewards/accuracies": 0.734375,
      "rewards/chosen": 0.15896882116794586,
      "rewards/margins": 0.09183251112699509,
      "rewards/rejected": 0.06713630259037018,
      "step": 32
    },
    {
      "epoch": 0.4401294498381877,
      "grad_norm": 9.351248951840846,
      "learning_rate": 4.866539188226085e-07,
      "logits/chosen": 36.365047454833984,
      "logits/rejected": 26.499774932861328,
      "logps/chosen": -186.34649658203125,
      "logps/rejected": -148.74656677246094,
      "loss": 0.6368,
      "rewards/accuracies": 0.734375,
      "rewards/chosen": 0.16049998998641968,
      "rewards/margins": 0.12190230935811996,
      "rewards/rejected": 0.03859768062829971,
      "step": 34
    },
    {
      "epoch": 0.46601941747572817,
      "grad_norm": 9.747161361199119,
      "learning_rate": 4.826163240767716e-07,
      "logits/chosen": 32.18162155151367,
      "logits/rejected": 21.5031681060791,
      "logps/chosen": -211.55101013183594,
      "logps/rejected": -170.30624389648438,
      "loss": 0.6312,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": 0.20323500037193298,
      "rewards/margins": 0.14077112078666687,
      "rewards/rejected": 0.06246389076113701,
      "step": 36
    },
    {
      "epoch": 0.4919093851132686,
      "grad_norm": 8.77950594997527,
      "learning_rate": 4.780673881662242e-07,
      "logits/chosen": 36.65592956542969,
      "logits/rejected": 28.596372604370117,
      "logps/chosen": -178.170166015625,
      "logps/rejected": -115.30381774902344,
      "loss": 0.6193,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 0.2410201132297516,
      "rewards/margins": 0.1858924776315689,
      "rewards/rejected": 0.05512763559818268,
      "step": 38
    },
    {
      "epoch": 0.517799352750809,
      "grad_norm": 8.94781303872067,
      "learning_rate": 4.730171106393466e-07,
      "logits/chosen": 31.84834098815918,
      "logits/rejected": 25.41543197631836,
      "logps/chosen": -211.7396240234375,
      "logps/rejected": -159.92942810058594,
      "loss": 0.6189,
      "rewards/accuracies": 0.796875,
      "rewards/chosen": 0.2539159655570984,
      "rewards/margins": 0.1977871209383011,
      "rewards/rejected": 0.0561288520693779,
      "step": 40
    },
    {
      "epoch": 0.5436893203883495,
      "grad_norm": 9.230235665250099,
      "learning_rate": 4.6747659310219757e-07,
      "logits/chosen": 39.43976593017578,
      "logits/rejected": 28.925722122192383,
      "logps/chosen": -206.06393432617188,
      "logps/rejected": -152.7125244140625,
      "loss": 0.6154,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 0.2500843405723572,
      "rewards/margins": 0.19579222798347473,
      "rewards/rejected": 0.05429207533597946,
      "step": 42
    },
    {
      "epoch": 0.56957928802589,
      "grad_norm": 9.122707874165032,
      "learning_rate": 4.6145801481477433e-07,
      "logits/chosen": 31.554033279418945,
      "logits/rejected": 22.213409423828125,
      "logps/chosen": -214.81248474121094,
      "logps/rejected": -200.99317932128906,
      "loss": 0.5942,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 0.3934122622013092,
      "rewards/margins": 0.2697424292564392,
      "rewards/rejected": 0.12366984784603119,
      "step": 44
    },
    {
      "epoch": 0.5954692556634305,
      "grad_norm": 8.293775972177585,
      "learning_rate": 4.549746059183561e-07,
      "logits/chosen": 40.96059799194336,
      "logits/rejected": 31.219024658203125,
      "logps/chosen": -186.52183532714844,
      "logps/rejected": -130.83743286132812,
      "loss": 0.5531,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 0.4351516366004944,
      "rewards/margins": 0.29554393887519836,
      "rewards/rejected": 0.13960765302181244,
      "step": 46
    },
    {
      "epoch": 0.6213592233009708,
      "grad_norm": 8.320484480615326,
      "learning_rate": 4.480406183527823e-07,
      "logits/chosen": 34.98313522338867,
      "logits/rejected": 23.058082580566406,
      "logps/chosen": -194.0576934814453,
      "logps/rejected": -143.09188842773438,
      "loss": 0.5291,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.5364688634872437,
      "rewards/margins": 0.3940187096595764,
      "rewards/rejected": 0.14245013892650604,
      "step": 48
    },
    {
      "epoch": 0.6472491909385113,
      "grad_norm": 7.915902357485634,
      "learning_rate": 4.4067129452759546e-07,
      "logits/chosen": 39.90618133544922,
      "logits/rejected": 27.514339447021484,
      "logps/chosen": -188.83193969726562,
      "logps/rejected": -126.44863891601562,
      "loss": 0.5153,
      "rewards/accuracies": 0.921875,
      "rewards/chosen": 0.5586945414543152,
      "rewards/margins": 0.3602239787578583,
      "rewards/rejected": 0.19847053289413452,
      "step": 50
    },
    {
      "epoch": 0.6731391585760518,
      "grad_norm": 7.903680790440658,
      "learning_rate": 4.3288283381591725e-07,
      "logits/chosen": 33.0372428894043,
      "logits/rejected": 21.942291259765625,
      "logps/chosen": -183.81349182128906,
      "logps/rejected": -142.9326934814453,
      "loss": 0.5032,
      "rewards/accuracies": 0.859375,
      "rewards/chosen": 0.6302269697189331,
      "rewards/margins": 0.4071354568004608,
      "rewards/rejected": 0.22309155762195587,
      "step": 52
    },
    {
      "epoch": 0.6990291262135923,
      "grad_norm": 7.297509807618061,
      "learning_rate": 4.246923569447104e-07,
      "logits/chosen": 34.447837829589844,
      "logits/rejected": 20.60545539855957,
      "logps/chosen": -179.7378692626953,
      "logps/rejected": -145.105712890625,
      "loss": 0.4965,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.727641224861145,
      "rewards/margins": 0.47567129135131836,
      "rewards/rejected": 0.25196999311447144,
      "step": 54
    },
    {
      "epoch": 0.7249190938511327,
      "grad_norm": 7.283409160445081,
      "learning_rate": 4.161178683597054e-07,
      "logits/chosen": 35.52771759033203,
      "logits/rejected": 30.491806030273438,
      "logps/chosen": -211.5355224609375,
      "logps/rejected": -149.41921997070312,
      "loss": 0.4999,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 0.752875804901123,
      "rewards/margins": 0.49092933535575867,
      "rewards/rejected": 0.261946439743042,
      "step": 56
    },
    {
      "epoch": 0.7508090614886731,
      "grad_norm": 7.796223292724318,
      "learning_rate": 4.0717821664772124e-07,
      "logits/chosen": 40.88484573364258,
      "logits/rejected": 29.62355613708496,
      "logps/chosen": -196.11395263671875,
      "logps/rejected": -143.0408935546875,
      "loss": 0.491,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.7662146091461182,
      "rewards/margins": 0.5029052495956421,
      "rewards/rejected": 0.26330941915512085,
      "step": 58
    },
    {
      "epoch": 0.7766990291262136,
      "grad_norm": 7.057826513245534,
      "learning_rate": 3.978930531033806e-07,
      "logits/chosen": 40.54041290283203,
      "logits/rejected": 30.305131912231445,
      "logps/chosen": -174.91925048828125,
      "logps/rejected": -112.93276977539062,
      "loss": 0.4852,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.7946022748947144,
      "rewards/margins": 0.5570634603500366,
      "rewards/rejected": 0.23753879964351654,
      "step": 60
    },
    {
      "epoch": 0.8025889967637541,
      "grad_norm": 7.4086734288874645,
      "learning_rate": 3.882827885312998e-07,
      "logits/chosen": 39.428497314453125,
      "logits/rejected": 29.150917053222656,
      "logps/chosen": -191.66470336914062,
      "logps/rejected": -136.3485107421875,
      "loss": 0.4617,
      "rewards/accuracies": 0.984375,
      "rewards/chosen": 0.7605564594268799,
      "rewards/margins": 0.5766815543174744,
      "rewards/rejected": 0.1838749349117279,
      "step": 62
    },
    {
      "epoch": 0.8284789644012945,
      "grad_norm": 6.8861010004738095,
      "learning_rate": 3.7836854837871044e-07,
      "logits/chosen": 35.57158279418945,
      "logits/rejected": 25.836395263671875,
      "logps/chosen": -196.1132049560547,
      "logps/rejected": -142.07156372070312,
      "loss": 0.4596,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 0.861929714679718,
      "rewards/margins": 0.6038858294487,
      "rewards/rejected": 0.2580438554286957,
      "step": 64
    },
    {
      "epoch": 0.8543689320388349,
      "grad_norm": 6.501368437260124,
      "learning_rate": 3.681721262971413e-07,
      "logits/chosen": 37.78006362915039,
      "logits/rejected": 27.594064712524414,
      "logps/chosen": -164.67950439453125,
      "logps/rejected": -110.48625183105469,
      "loss": 0.4547,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.8753312826156616,
      "rewards/margins": 0.6975246071815491,
      "rewards/rejected": 0.17780669033527374,
      "step": 66
    },
    {
      "epoch": 0.8802588996763754,
      "grad_norm": 6.543034078515488,
      "learning_rate": 3.577159362352426e-07,
      "logits/chosen": 37.94480514526367,
      "logits/rejected": 27.400300979614258,
      "logps/chosen": -161.64242553710938,
      "logps/rejected": -118.77005004882812,
      "loss": 0.4531,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.8563450574874878,
      "rewards/margins": 0.6019262075424194,
      "rewards/rejected": 0.25441890954971313,
      "step": 68
    },
    {
      "epoch": 0.9061488673139159,
      "grad_norm": 6.720711950688133,
      "learning_rate": 3.470229631680624e-07,
      "logits/chosen": 38.4688835144043,
      "logits/rejected": 28.796794891357422,
      "logps/chosen": -184.064697265625,
      "logps/rejected": -134.16714477539062,
      "loss": 0.4334,
      "rewards/accuracies": 0.921875,
      "rewards/chosen": 0.8126658201217651,
      "rewards/margins": 0.6111183166503906,
      "rewards/rejected": 0.2015475183725357,
      "step": 70
    },
    {
      "epoch": 0.9320388349514563,
      "grad_norm": 6.228494469432083,
      "learning_rate": 3.361167125710832e-07,
      "logits/chosen": 42.73847579956055,
      "logits/rejected": 31.657482147216797,
      "logps/chosen": -183.9188232421875,
      "logps/rejected": -129.78868103027344,
      "loss": 0.4365,
      "rewards/accuracies": 0.984375,
      "rewards/chosen": 0.8676687479019165,
      "rewards/margins": 0.6447817087173462,
      "rewards/rejected": 0.2228870689868927,
      "step": 72
    },
    {
      "epoch": 0.9579288025889967,
      "grad_norm": 6.71909319850701,
      "learning_rate": 3.2502115875008516e-07,
      "logits/chosen": 34.51922607421875,
      "logits/rejected": 20.536556243896484,
      "logps/chosen": -169.9495391845703,
      "logps/rejected": -114.22787475585938,
      "loss": 0.4202,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 0.9300148487091064,
      "rewards/margins": 0.7187902331352234,
      "rewards/rejected": 0.21122469007968903,
      "step": 74
    },
    {
      "epoch": 0.9838187702265372,
      "grad_norm": 6.3858630141282795,
      "learning_rate": 3.137606921404191e-07,
      "logits/chosen": 32.11302185058594,
      "logits/rejected": 20.860279083251953,
      "logps/chosen": -204.4998779296875,
      "logps/rejected": -183.44058227539062,
      "loss": 0.4167,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 0.9314306378364563,
      "rewards/margins": 0.751998782157898,
      "rewards/rejected": 0.17943182587623596,
      "step": 76
    }
  ],
  "logging_steps": 2,
  "max_steps": 154,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 50000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}