{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.9893390191897655,
  "eval_steps": 100,
  "global_step": 232,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 445.9810485839844,
      "epoch": 0.017057569296375266,
      "grad_norm": 0.44629430770874023,
      "kl": 0.0,
      "learning_rate": 1.25e-07,
      "loss": -0.0061,
      "reward": 0.2678571566939354,
      "reward_std": 0.3369061201810837,
      "rewards/accuracy_reward": 0.2154017947614193,
      "rewards/format_reward": 0.05245535937137902,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 450.8755760192871,
      "epoch": 0.08528784648187633,
      "grad_norm": 0.5371240377426147,
      "kl": 5.666911602020264e-05,
      "learning_rate": 6.25e-07,
      "loss": -0.0072,
      "reward": 0.26227679662406445,
      "reward_std": 0.3667756309732795,
      "rewards/accuracy_reward": 0.1936384029686451,
      "rewards/format_reward": 0.0686383958091028,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 440.6763626098633,
      "epoch": 0.17057569296375266,
      "grad_norm": 0.5557937026023865,
      "kl": 0.003149533271789551,
      "learning_rate": 1.25e-06,
      "loss": 0.0071,
      "reward": 0.30558037012815475,
      "reward_std": 0.4056690149009228,
      "rewards/accuracy_reward": 0.19308036658912897,
      "rewards/format_reward": 0.11250000540167093,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 456.58997573852537,
      "epoch": 0.255863539445629,
      "grad_norm": 4.744609832763672,
      "kl": 0.0615509033203125,
      "learning_rate": 1.875e-06,
      "loss": 0.0307,
      "reward": 0.674107177555561,
      "reward_std": 0.5331576444208622,
      "rewards/accuracy_reward": 0.27075893972069026,
      "rewards/format_reward": 0.4033482324331999,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 481.85694046020507,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.46378934383392334,
      "kl": 0.06910400390625,
      "learning_rate": 2.5e-06,
      "loss": 0.0287,
      "reward": 1.0837054058909417,
      "reward_std": 0.5833071276545525,
      "rewards/accuracy_reward": 0.4551339499652386,
      "rewards/format_reward": 0.6285714603960514,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 518.9658714294434,
      "epoch": 0.42643923240938164,
      "grad_norm": 3.33089017868042,
      "kl": 0.056927490234375,
      "learning_rate": 2.999828909426247e-06,
      "loss": 0.0487,
      "reward": 1.3687500655651093,
      "reward_std": 0.50418995693326,
      "rewards/accuracy_reward": 0.581250024586916,
      "rewards/format_reward": 0.7875000387430191,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 517.6306053161621,
      "epoch": 0.511727078891258,
      "grad_norm": 7.978788375854492,
      "kl": 0.0954803466796875,
      "learning_rate": 2.9938448364256362e-06,
      "loss": 0.0525,
      "reward": 1.4741072177886962,
      "reward_std": 0.43093371838331224,
      "rewards/accuracy_reward": 0.5995536006987094,
      "rewards/format_reward": 0.8745536118745804,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 505.315869140625,
      "epoch": 0.5970149253731343,
      "grad_norm": 175585.21875,
      "kl": 64.5933837890625,
      "learning_rate": 2.979345224048116e-06,
      "loss": 5.1443,
      "reward": 1.4098214894533156,
      "reward_std": 0.4775668144226074,
      "rewards/accuracy_reward": 0.5758928783237934,
      "rewards/format_reward": 0.833928607404232,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 498.5922119140625,
      "epoch": 0.6823027718550106,
      "grad_norm": 2.6241941452026367,
      "kl": 0.74183349609375,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.1161,
      "reward": 1.2595982730388642,
      "reward_std": 0.5859433703124524,
      "rewards/accuracy_reward": 0.5636160977184772,
      "rewards/format_reward": 0.695982177555561,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 519.4406471252441,
      "epoch": 0.767590618336887,
      "grad_norm": 52.20960998535156,
      "kl": 0.151666259765625,
      "learning_rate": 2.925178067512904e-06,
      "loss": 0.1406,
      "reward": 1.215401840209961,
      "reward_std": 0.6168530285358429,
      "rewards/accuracy_reward": 0.5671875238418579,
      "rewards/format_reward": 0.648214316368103,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 566.0794883728028,
      "epoch": 0.8528784648187633,
      "grad_norm": 9.431961059570312,
      "kl": 0.23187255859375,
      "learning_rate": 2.88581929876693e-06,
      "loss": 0.2278,
      "reward": 1.245089340209961,
      "reward_std": 0.6514639720320702,
      "rewards/accuracy_reward": 0.5736607365310192,
      "rewards/format_reward": 0.6714285984635353,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 614.7377502441407,
      "epoch": 0.9381663113006397,
      "grad_norm": 28.97673988342285,
      "kl": 3.216015625,
      "learning_rate": 2.8385607813186967e-06,
      "loss": 0.434,
      "reward": 1.1569196969270705,
      "reward_std": 0.7305632084608078,
      "rewards/accuracy_reward": 0.5412946678698063,
      "rewards/format_reward": 0.615625025331974,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 559.5640083312989,
      "epoch": 1.0341151385927505,
      "grad_norm": 21.23921775817871,
      "kl": 6.149072265625,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.6512,
      "reward": 1.2779018491506577,
      "reward_std": 0.6680191181600094,
      "rewards/accuracy_reward": 0.5743303805589676,
      "rewards/format_reward": 0.7035714641213417,
      "step": 60
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 550.1716812133789,
      "epoch": 1.1194029850746268,
      "grad_norm": 22.82264518737793,
      "kl": 2.95859375,
      "learning_rate": 2.7214655696635407e-06,
      "loss": 0.4399,
      "reward": 1.30558041036129,
      "reward_std": 0.6594597332179546,
      "rewards/accuracy_reward": 0.5687500216066838,
      "rewards/format_reward": 0.7368303894996643,
      "step": 65
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 569.2410942077637,
      "epoch": 1.2046908315565032,
      "grad_norm": 22.326448440551758,
      "kl": 2.64033203125,
      "learning_rate": 2.652296367060421e-06,
      "loss": 0.4324,
      "reward": 1.283482199907303,
      "reward_std": 0.6751217097043991,
      "rewards/accuracy_reward": 0.5662946663796902,
      "rewards/format_reward": 0.7171875342726708,
      "step": 70
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 589.6875289916992,
      "epoch": 1.2899786780383795,
      "grad_norm": 11.764822959899902,
      "kl": 2.82216796875,
      "learning_rate": 2.5765585939817676e-06,
      "loss": 0.454,
      "reward": 1.2718750685453415,
      "reward_std": 0.6881907656788826,
      "rewards/accuracy_reward": 0.5497768148779869,
      "rewards/format_reward": 0.7220982432365417,
      "step": 75
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 588.3623039245606,
      "epoch": 1.375266524520256,
      "grad_norm": 16.239723205566406,
      "kl": 1.092626953125,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.3543,
      "reward": 1.2966518431901932,
      "reward_std": 0.7087490603327751,
      "rewards/accuracy_reward": 0.5763393104076385,
      "rewards/format_reward": 0.7203125327825546,
      "step": 80
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 554.1701164245605,
      "epoch": 1.4605543710021323,
      "grad_norm": 225.1170196533203,
      "kl": 3.343701171875,
      "learning_rate": 2.4071392666461563e-06,
      "loss": 0.4732,
      "reward": 1.348883980512619,
      "reward_std": 0.6181702017784119,
      "rewards/accuracy_reward": 0.5640625275671483,
      "rewards/format_reward": 0.7848214700818061,
      "step": 85
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 557.5998046875,
      "epoch": 1.5458422174840085,
      "grad_norm": 35.329673767089844,
      "kl": 3.09814453125,
      "learning_rate": 2.314423473302218e-06,
      "loss": 0.4767,
      "reward": 1.3281250685453414,
      "reward_std": 0.6315318033099174,
      "rewards/accuracy_reward": 0.5370535977184773,
      "rewards/format_reward": 0.7910714611411095,
      "step": 90
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 565.7868560791015,
      "epoch": 1.6311300639658848,
      "grad_norm": 64.45903778076172,
      "kl": 2.35947265625,
      "learning_rate": 2.2170651260682927e-06,
      "loss": 0.424,
      "reward": 1.3314732700586318,
      "reward_std": 0.6345132827758789,
      "rewards/accuracy_reward": 0.5468750260770321,
      "rewards/format_reward": 0.7845982551574707,
      "step": 95
    },
    {
      "epoch": 1.716417910447761,
      "grad_norm": 31.969350814819336,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 2.0191,
      "step": 100
    },
    {
      "epoch": 1.716417910447761,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 573.7114408937887,
      "eval_kl": 6.245007987220447,
      "eval_loss": 0.679384171962738,
      "eval_reward": 1.3160087325321599,
      "eval_reward_std": 0.6340273570138425,
      "eval_rewards/accuracy_reward": 0.5342024423824713,
      "eval_rewards/format_reward": 0.7818062878645267,
      "eval_runtime": 4416.6844,
      "eval_samples_per_second": 1.132,
      "eval_steps_per_second": 0.01,
      "step": 100
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 576.5287105560303,
      "epoch": 1.8017057569296375,
      "grad_norm": 38.427772521972656,
      "kl": 14.87158203125,
      "learning_rate": 2.010664003729149e-06,
      "loss": 0.5592,
      "reward": 1.361272382736206,
      "reward_std": 0.6285464949905872,
      "rewards/accuracy_reward": 0.5775669913738966,
      "rewards/format_reward": 0.7837053917348384,
      "step": 105
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 571.036637878418,
      "epoch": 1.886993603411514,
      "grad_norm": 13.744680404663086,
      "kl": 4.437109375,
      "learning_rate": 1.9027978012115653e-06,
      "loss": 0.5515,
      "reward": 1.3685268431901931,
      "reward_std": 0.6233541399240494,
      "rewards/accuracy_reward": 0.5756696701049805,
      "rewards/format_reward": 0.7928571805357933,
      "step": 110
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 561.5263648986817,
      "epoch": 1.9722814498933903,
      "grad_norm": 11.377424240112305,
      "kl": 5.195703125,
      "learning_rate": 1.7926354830241926e-06,
      "loss": 0.5897,
      "reward": 1.3955357789993286,
      "reward_std": 0.5930619619786739,
      "rewards/accuracy_reward": 0.5919643118977547,
      "rewards/format_reward": 0.8035714611411094,
      "step": 115
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 543.7195793151856,
      "epoch": 2.068230277185501,
      "grad_norm": 21.994367599487305,
      "kl": 3.9482421875,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.4929,
      "reward": 1.4200893491506577,
      "reward_std": 0.5751285634934902,
      "rewards/accuracy_reward": 0.594196455925703,
      "rewards/format_reward": 0.8258928969502449,
      "step": 120
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 528.3288215637207,
      "epoch": 2.1535181236673773,
      "grad_norm": 23.414892196655273,
      "kl": 3.4708984375,
      "learning_rate": 1.5679438936238768e-06,
      "loss": 0.4547,
      "reward": 1.4633929193019868,
      "reward_std": 0.5353646248579025,
      "rewards/accuracy_reward": 0.6140625290572643,
      "rewards/format_reward": 0.8493303969502449,
      "step": 125
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 532.4109573364258,
      "epoch": 2.2388059701492535,
      "grad_norm": 5.716637134552002,
      "kl": 3.59921875,
      "learning_rate": 1.454695458298667e-06,
      "loss": 0.4406,
      "reward": 1.4633929222822188,
      "reward_std": 0.5372500255703926,
      "rewards/accuracy_reward": 0.6140625238418579,
      "rewards/format_reward": 0.8493303999304771,
      "step": 130
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 529.8064964294433,
      "epoch": 2.3240938166311302,
      "grad_norm": 11.957624435424805,
      "kl": 3.2958984375,
      "learning_rate": 1.341705277779715e-06,
      "loss": 0.4071,
      "reward": 1.4368304163217545,
      "reward_std": 0.5057652793824673,
      "rewards/accuracy_reward": 0.5850446701049805,
      "rewards/format_reward": 0.8517857536673545,
      "step": 135
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 513.0975715637207,
      "epoch": 2.4093816631130065,
      "grad_norm": 409.32232666015625,
      "kl": 3.7796875,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.452,
      "reward": 1.4669643461704254,
      "reward_std": 0.5007161863148213,
      "rewards/accuracy_reward": 0.6037946708500386,
      "rewards/format_reward": 0.8631696820259094,
      "step": 140
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 513.7160926818848,
      "epoch": 2.4946695095948828,
      "grad_norm": 4.6824445724487305,
      "kl": 3.0923828125,
      "learning_rate": 1.1190709022599545e-06,
      "loss": 0.3687,
      "reward": 1.501339355111122,
      "reward_std": 0.501213763654232,
      "rewards/accuracy_reward": 0.6316964566707611,
      "rewards/format_reward": 0.8696429014205933,
      "step": 145
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 503.14399871826174,
      "epoch": 2.579957356076759,
      "grad_norm": 5.0907368659973145,
      "kl": 2.658984375,
      "learning_rate": 1.0106958161686963e-06,
      "loss": 0.3415,
      "reward": 1.4787947177886962,
      "reward_std": 0.47073035016655923,
      "rewards/accuracy_reward": 0.6015625268220901,
      "rewards/format_reward": 0.8772321864962578,
      "step": 150
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 493.3506935119629,
      "epoch": 2.6652452025586353,
      "grad_norm": 5.612578868865967,
      "kl": 3.1716796875,
      "learning_rate": 9.051099682520474e-07,
      "loss": 0.3894,
      "reward": 1.5064732819795608,
      "reward_std": 0.45394266620278356,
      "rewards/accuracy_reward": 0.6234375290572644,
      "rewards/format_reward": 0.8830357596278191,
      "step": 155
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 491.6685501098633,
      "epoch": 2.750533049040512,
      "grad_norm": 9.828149795532227,
      "kl": 2.4236328125,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.3284,
      "reward": 1.5341518610715865,
      "reward_std": 0.44977339953184126,
      "rewards/accuracy_reward": 0.6435268133878708,
      "rewards/format_reward": 0.8906250447034836,
      "step": 160
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 485.81029205322267,
      "epoch": 2.835820895522388,
      "grad_norm": 4.9597859382629395,
      "kl": 2.07568359375,
      "learning_rate": 7.046941898307347e-07,
      "loss": 0.3018,
      "reward": 1.5156250655651093,
      "reward_std": 0.4370568677783012,
      "rewards/accuracy_reward": 0.6238839596509933,
      "rewards/format_reward": 0.8917411118745804,
      "step": 165
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 497.5314971923828,
      "epoch": 2.9211087420042645,
      "grad_norm": 84.52203369140625,
      "kl": 3.3498046875,
      "learning_rate": 6.11006712953975e-07,
      "loss": 0.4164,
      "reward": 1.5022322148084641,
      "reward_std": 0.4752186842262745,
      "rewards/accuracy_reward": 0.6212053842842579,
      "rewards/format_reward": 0.8810268267989159,
      "step": 170
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 495.55774307250977,
      "epoch": 3.0170575692963753,
      "grad_norm": 16.004627227783203,
      "kl": 3.1123046875,
      "learning_rate": 5.223868690448817e-07,
      "loss": 0.4053,
      "reward": 1.4607143461704255,
      "reward_std": 0.49991785958409307,
      "rewards/accuracy_reward": 0.5937500283122062,
      "rewards/format_reward": 0.8669643327593803,
      "step": 175
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 504.62011337280273,
      "epoch": 3.1023454157782515,
      "grad_norm": 4.859574794769287,
      "kl": 3.499609375,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.4582,
      "reward": 1.4575893461704255,
      "reward_std": 0.5223984435200691,
      "rewards/accuracy_reward": 0.6051339566707611,
      "rewards/format_reward": 0.8524553969502449,
      "step": 180
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 505.25716705322264,
      "epoch": 3.1876332622601278,
      "grad_norm": 6.827749729156494,
      "kl": 3.01005859375,
      "learning_rate": 3.6233899329188115e-07,
      "loss": 0.3983,
      "reward": 1.4602679193019867,
      "reward_std": 0.49594502747058866,
      "rewards/accuracy_reward": 0.5908482372760773,
      "rewards/format_reward": 0.8694196835160255,
      "step": 185
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 499.8402046203613,
      "epoch": 3.272921108742004,
      "grad_norm": 5.723031520843506,
      "kl": 2.642578125,
      "learning_rate": 2.9182330117358096e-07,
      "loss": 0.381,
      "reward": 1.4939732760190965,
      "reward_std": 0.4914128452539444,
      "rewards/accuracy_reward": 0.615401814877987,
      "rewards/format_reward": 0.8785714715719223,
      "step": 190
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 495.40426177978514,
      "epoch": 3.3582089552238807,
      "grad_norm": 7.8606390953063965,
      "kl": 3.060546875,
      "learning_rate": 2.281947207567473e-07,
      "loss": 0.3953,
      "reward": 1.468303632736206,
      "reward_std": 0.4849529266357422,
      "rewards/accuracy_reward": 0.5866071686148644,
      "rewards/format_reward": 0.8816964745521545,
      "step": 195
    },
    {
      "epoch": 3.443496801705757,
      "grad_norm": 4.874986171722412,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.3222,
      "step": 200
    },
    {
      "epoch": 3.443496801705757,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 488.8427760700067,
      "eval_kl": 2.4132263378594248,
      "eval_loss": 0.32682499289512634,
      "eval_reward": 1.4757246209409671,
      "eval_reward_std": 0.4513232411858373,
      "eval_rewards/accuracy_reward": 0.5826677580039722,
      "eval_rewards/format_reward": 0.8930568643652212,
      "eval_runtime": 4221.4152,
      "eval_samples_per_second": 1.184,
      "eval_steps_per_second": 0.011,
      "step": 200
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 489.15783576965333,
      "epoch": 3.5287846481876333,
      "grad_norm": 5.266751289367676,
      "kl": 2.5323974609375,
      "learning_rate": 1.2300840593454622e-07,
      "loss": 0.3495,
      "reward": 1.5162947088479997,
      "reward_std": 0.4514313301071525,
      "rewards/accuracy_reward": 0.6287946715950966,
      "rewards/format_reward": 0.8875000402331352,
      "step": 205
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 491.2348434448242,
      "epoch": 3.6140724946695095,
      "grad_norm": 9.579411506652832,
      "kl": 2.3810546875,
      "learning_rate": 8.20502774480395e-08,
      "loss": 0.308,
      "reward": 1.5162946969270705,
      "reward_std": 0.4419292353093624,
      "rewards/accuracy_reward": 0.6263393133878707,
      "rewards/format_reward": 0.8899554029107094,
      "step": 210
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 489.42010803222655,
      "epoch": 3.699360341151386,
      "grad_norm": 4.795035362243652,
      "kl": 2.62646484375,
      "learning_rate": 4.917505449659615e-08,
      "loss": 0.3449,
      "reward": 1.5185268491506576,
      "reward_std": 0.45038328543305395,
      "rewards/accuracy_reward": 0.6337053865194321,
      "rewards/format_reward": 0.8848214730620384,
      "step": 215
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 495.2564956665039,
      "epoch": 3.7846481876332625,
      "grad_norm": 6.437369346618652,
      "kl": 2.59952392578125,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.3286,
      "reward": 1.495535773038864,
      "reward_std": 0.46331221498548986,
      "rewards/accuracy_reward": 0.6071428850293159,
      "rewards/format_reward": 0.8883929014205932,
      "step": 220
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 485.0676559448242,
      "epoch": 3.8699360341151388,
      "grad_norm": 4.596541404724121,
      "kl": 2.499609375,
      "learning_rate": 8.37579098581176e-09,
      "loss": 0.3586,
      "reward": 1.5156250715255737,
      "reward_std": 0.4558119185268879,
      "rewards/accuracy_reward": 0.6247768118977547,
      "rewards/format_reward": 0.8908482611179351,
      "step": 225
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 497.7335029602051,
      "epoch": 3.955223880597015,
      "grad_norm": 5.619830131530762,
      "kl": 2.5017578125,
      "learning_rate": 6.843232656998933e-10,
      "loss": 0.3316,
      "reward": 1.4997768551111221,
      "reward_std": 0.44890894591808317,
      "rewards/accuracy_reward": 0.6104911021888256,
      "rewards/format_reward": 0.8892857521772385,
      "step": 230
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 497.67967987060547,
      "epoch": 3.9893390191897655,
      "kl": 2.341064453125,
      "reward": 1.5100447237491608,
      "reward_std": 0.44926475919783115,
      "rewards/accuracy_reward": 0.6216518096625805,
      "rewards/format_reward": 0.8883928954601288,
      "step": 232,
      "total_flos": 0.0,
      "train_loss": 0.4856032096632307,
      "train_runtime": 39327.4213,
      "train_samples_per_second": 0.763,
      "train_steps_per_second": 0.006
    }
  ],
  "logging_steps": 5,
  "max_steps": 232,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}