{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9952153110047847,
  "eval_steps": 500,
  "global_step": 52,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 885.7360992431641,
      "epoch": 0.019138755980861243,
      "grad_norm": 0.05717110271611869,
      "kl": 0.0,
      "learning_rate": 1.6666666666666665e-07,
      "loss": 0.0224,
      "reward": 0.6623263955116272,
      "reward_std": 0.20746709778904915,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4123263880610466,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 855.6076354980469,
      "epoch": 0.03827751196172249,
      "grad_norm": 0.06704605474432411,
      "kl": 0.0,
      "learning_rate": 3.333333333333333e-07,
      "loss": 0.0199,
      "reward": 0.6901041716337204,
      "reward_std": 0.22956417128443718,
      "rewards/accuracy_reward": 0.2465277761220932,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4435763880610466,
      "step": 2
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 917.3541717529297,
      "epoch": 0.05741626794258373,
      "grad_norm": 0.05091560707629358,
      "kl": 2.9027462005615234e-05,
      "learning_rate": 5e-07,
      "loss": 0.0137,
      "reward": 0.5538194477558136,
      "reward_std": 0.19887378066778183,
      "rewards/accuracy_reward": 0.1770833358168602,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.3767361119389534,
      "step": 3
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 869.5937652587891,
      "epoch": 0.07655502392344497,
      "grad_norm": 0.06917728317111242,
      "kl": 3.2007694244384766e-05,
      "learning_rate": 6.666666666666666e-07,
      "loss": 0.0258,
      "reward": 0.634548619389534,
      "reward_std": 0.2516612634062767,
      "rewards/accuracy_reward": 0.2256944440305233,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4088541716337204,
      "step": 4
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 860.3229217529297,
      "epoch": 0.09569377990430622,
      "grad_norm": 0.06655951788671848,
      "kl": 2.7626752853393555e-05,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.0175,
      "reward": 0.636284738779068,
      "reward_std": 0.21483279019594193,
      "rewards/accuracy_reward": 0.2222222238779068,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4140625074505806,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 872.6909790039062,
      "epoch": 0.11483253588516747,
      "grad_norm": 0.06205063415295795,
      "kl": 2.2977590560913086e-05,
      "learning_rate": 1e-06,
      "loss": 0.0139,
      "reward": 0.6840277761220932,
      "reward_std": 0.2086946927011013,
      "rewards/accuracy_reward": 0.27083334140479565,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4131944477558136,
      "step": 6
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 881.4861297607422,
      "epoch": 0.1339712918660287,
      "grad_norm": 0.06390474824070583,
      "kl": 2.709031105041504e-05,
      "learning_rate": 9.989509461357426e-07,
      "loss": 0.0214,
      "reward": 0.618923619389534,
      "reward_std": 0.20255662873387337,
      "rewards/accuracy_reward": 0.2256944477558136,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.3932291641831398,
      "step": 7
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 856.1423645019531,
      "epoch": 0.15311004784688995,
      "grad_norm": 0.06611386863145982,
      "kl": 2.002716064453125e-05,
      "learning_rate": 9.958086757163488e-07,
      "loss": 0.0205,
      "reward": 0.659722238779068,
      "reward_std": 0.24306795373558998,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.430555559694767,
      "step": 8
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 875.7395935058594,
      "epoch": 0.1722488038277512,
      "grad_norm": 0.06272616959518927,
      "kl": 1.7598271369934082e-05,
      "learning_rate": 9.905878394570453e-07,
      "loss": 0.0155,
      "reward": 0.6875,
      "reward_std": 0.22097085043787956,
      "rewards/accuracy_reward": 0.2708333432674408,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4166666716337204,
      "step": 9
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 885.6041717529297,
      "epoch": 0.19138755980861244,
      "grad_norm": 0.0566584544534715,
      "kl": 1.3478100299835205e-05,
      "learning_rate": 9.833127793065097e-07,
      "loss": 0.0147,
      "reward": 0.6840277910232544,
      "reward_std": 0.21606039628386497,
      "rewards/accuracy_reward": 0.2881944477558136,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.3958333358168602,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 849.3645782470703,
      "epoch": 0.21052631578947367,
      "grad_norm": 0.07330325365220204,
      "kl": 2.1457672119140625e-05,
      "learning_rate": 9.740174149534692e-07,
      "loss": 0.0204,
      "reward": 0.7526041716337204,
      "reward_std": 0.2590269520878792,
      "rewards/accuracy_reward": 0.3020833320915699,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4505208432674408,
      "step": 11
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 876.4757080078125,
      "epoch": 0.22966507177033493,
      "grad_norm": 0.05630416735365695,
      "kl": 1.9222497940063477e-05,
      "learning_rate": 9.627450856774539e-07,
      "loss": 0.0098,
      "reward": 0.661458358168602,
      "reward_std": 0.18905285000801086,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4322916716337204,
      "step": 12
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 863.5590362548828,
      "epoch": 0.24880382775119617,
      "grad_norm": 0.060825286332779054,
      "kl": 1.8551945686340332e-05,
      "learning_rate": 9.495483482810687e-07,
      "loss": 0.0179,
      "reward": 0.6163194477558136,
      "reward_std": 0.22097087278962135,
      "rewards/accuracy_reward": 0.2048611119389534,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4114583358168602,
      "step": 13
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 859.7778015136719,
      "epoch": 0.2679425837320574,
      "grad_norm": 0.06028261705382758,
      "kl": 4.965066909790039e-05,
      "learning_rate": 9.344887320459198e-07,
      "loss": 0.0167,
      "reward": 0.6354166641831398,
      "reward_std": 0.2086947001516819,
      "rewards/accuracy_reward": 0.21527778171002865,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4201388955116272,
      "step": 14
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 889.1319427490234,
      "epoch": 0.28708133971291866,
      "grad_norm": 0.06077195852297118,
      "kl": 7.545948028564453e-05,
      "learning_rate": 9.176364518546988e-07,
      "loss": 0.0159,
      "reward": 0.6475694477558136,
      "reward_std": 0.19641854241490364,
      "rewards/accuracy_reward": 0.2465277723968029,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4010416716337204,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 875.8854217529297,
      "epoch": 0.3062200956937799,
      "grad_norm": 0.06912072388582151,
      "kl": 9.357929229736328e-05,
      "learning_rate": 8.990700808169889e-07,
      "loss": 0.0254,
      "reward": 0.709201380610466,
      "reward_std": 0.21483278647065163,
      "rewards/accuracy_reward": 0.2708333320915699,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.438368059694767,
      "step": 16
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 865.875,
      "epoch": 0.3253588516746411,
      "grad_norm": 0.06267659137438868,
      "kl": 0.00013184547424316406,
      "learning_rate": 8.788761839251558e-07,
      "loss": 0.0206,
      "reward": 0.7309027910232544,
      "reward_std": 0.262709803879261,
      "rewards/accuracy_reward": 0.3090277835726738,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4218750074505806,
      "step": 17
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 864.7777862548828,
      "epoch": 0.3444976076555024,
      "grad_norm": 0.06921820728455384,
      "kl": 0.0001842975616455078,
      "learning_rate": 8.571489144483944e-07,
      "loss": 0.023,
      "reward": 0.6979166716337204,
      "reward_std": 0.23079178109765053,
      "rewards/accuracy_reward": 0.2777777798473835,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4201388880610466,
      "step": 18
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 864.5208435058594,
      "epoch": 0.36363636363636365,
      "grad_norm": 0.06828931603666485,
      "kl": 0.0002467632293701172,
      "learning_rate": 8.339895749467237e-07,
      "loss": 0.021,
      "reward": 0.7291666865348816,
      "reward_std": 0.2357022613286972,
      "rewards/accuracy_reward": 0.2986111082136631,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.430555559694767,
      "step": 19
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 857.4965362548828,
      "epoch": 0.3827751196172249,
      "grad_norm": 0.057114973470776605,
      "kl": 0.00024008750915527344,
      "learning_rate": 8.095061449516902e-07,
      "loss": 0.0101,
      "reward": 0.7317708283662796,
      "reward_std": 0.23692986369132996,
      "rewards/accuracy_reward": 0.3090277835726738,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4227430671453476,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 855.9201354980469,
      "epoch": 0.4019138755980861,
      "grad_norm": 0.05706380401976011,
      "kl": 0.0002942085266113281,
      "learning_rate": 7.838127775159451e-07,
      "loss": 0.0221,
      "reward": 0.7118055671453476,
      "reward_std": 0.25534411519765854,
      "rewards/accuracy_reward": 0.2847222313284874,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4270833358168602,
      "step": 21
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 860.5173645019531,
      "epoch": 0.42105263157894735,
      "grad_norm": 0.05882812256894408,
      "kl": 0.0003452301025390625,
      "learning_rate": 7.570292669790184e-07,
      "loss": 0.0031,
      "reward": 0.6935763955116272,
      "reward_std": 0.20501185953617096,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.422743059694767,
      "step": 22
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 859.2639007568359,
      "epoch": 0.44019138755980863,
      "grad_norm": 0.06173500079510746,
      "kl": 0.000537872314453125,
      "learning_rate": 7.292804904308086e-07,
      "loss": 0.0154,
      "reward": 0.7161458283662796,
      "reward_std": 0.21483277902007103,
      "rewards/accuracy_reward": 0.2951388917863369,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4210069552063942,
      "step": 23
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 858.9097290039062,
      "epoch": 0.45933014354066987,
      "grad_norm": 0.06893952606573064,
      "kl": 0.0005884170532226562,
      "learning_rate": 7.006958254769437e-07,
      "loss": 0.0264,
      "reward": 0.701388880610466,
      "reward_std": 0.2602545768022537,
      "rewards/accuracy_reward": 0.2881944440305233,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4131944477558136,
      "step": 24
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 860.0416717529297,
      "epoch": 0.4784688995215311,
      "grad_norm": 0.0687100634503083,
      "kl": 0.0006256103515625,
      "learning_rate": 6.714085470206609e-07,
      "loss": 0.0255,
      "reward": 0.7230902910232544,
      "reward_std": 0.24184033274650574,
      "rewards/accuracy_reward": 0.2881944552063942,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4348958283662796,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 864.3715209960938,
      "epoch": 0.49760765550239233,
      "grad_norm": 0.061459175011899785,
      "kl": 0.0009174346923828125,
      "learning_rate": 6.415552058736853e-07,
      "loss": 0.0191,
      "reward": 0.7335069477558136,
      "reward_std": 0.2786688134074211,
      "rewards/accuracy_reward": 0.3229166641831398,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4105902835726738,
      "step": 26
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 896.4722290039062,
      "epoch": 0.5167464114832536,
      "grad_norm": 0.06334722551499229,
      "kl": 0.000942230224609375,
      "learning_rate": 6.11274992093311e-07,
      "loss": 0.0245,
      "reward": 0.6371527761220932,
      "reward_std": 0.24306796491146088,
      "rewards/accuracy_reward": 0.2430555559694767,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.3940972313284874,
      "step": 27
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 853.0451354980469,
      "epoch": 0.5358851674641149,
      "grad_norm": 0.05677441196904184,
      "kl": 0.0012645721435546875,
      "learning_rate": 5.80709086014102e-07,
      "loss": 0.0077,
      "reward": 0.740451380610466,
      "reward_std": 0.16818338260054588,
      "rewards/accuracy_reward": 0.3229166641831398,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4175347238779068,
      "step": 28
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 859.9132080078125,
      "epoch": 0.5550239234449761,
      "grad_norm": 0.08121565074078244,
      "kl": 0.001453399658203125,
      "learning_rate": 5.5e-07,
      "loss": 0.0325,
      "reward": 0.743923619389534,
      "reward_std": 0.3105868399143219,
      "rewards/accuracy_reward": 0.3090277835726738,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4348958358168602,
      "step": 29
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 843.4513854980469,
      "epoch": 0.5741626794258373,
      "grad_norm": 0.06778365437453367,
      "kl": 0.001445770263671875,
      "learning_rate": 5.192909139858981e-07,
      "loss": 0.028,
      "reward": 0.7717013955116272,
      "reward_std": 0.27375834435224533,
      "rewards/accuracy_reward": 0.3368055522441864,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4348958432674408,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 887.6354370117188,
      "epoch": 0.5933014354066986,
      "grad_norm": 0.0590498329750237,
      "kl": 0.0012683868408203125,
      "learning_rate": 4.887250079066891e-07,
      "loss": 0.0255,
      "reward": 0.7057291865348816,
      "reward_std": 0.21237755194306374,
      "rewards/accuracy_reward": 0.2916666716337204,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4140625074505806,
      "step": 31
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 832.625,
      "epoch": 0.6124401913875598,
      "grad_norm": 0.07056325724098401,
      "kl": 0.0017757415771484375,
      "learning_rate": 4.584447941263149e-07,
      "loss": 0.0178,
      "reward": 0.7578125149011612,
      "reward_std": 0.26393740251660347,
      "rewards/accuracy_reward": 0.3159722238779068,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4418402761220932,
      "step": 32
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 856.9514007568359,
      "epoch": 0.631578947368421,
      "grad_norm": 0.06693116072370171,
      "kl": 0.001903533935546875,
      "learning_rate": 4.285914529793391e-07,
      "loss": 0.0283,
      "reward": 0.747395858168602,
      "reward_std": 0.23447464406490326,
      "rewards/accuracy_reward": 0.3229166716337204,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4244791716337204,
      "step": 33
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 837.0069427490234,
      "epoch": 0.6507177033492823,
      "grad_norm": 0.06430130826482054,
      "kl": 0.0016841888427734375,
      "learning_rate": 3.993041745230562e-07,
      "loss": 0.0254,
      "reward": 0.8125000149011612,
      "reward_std": 0.22833656147122383,
      "rewards/accuracy_reward": 0.3888888955116272,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.423611119389534,
      "step": 34
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 833.3854217529297,
      "epoch": 0.6698564593301436,
      "grad_norm": 0.06623663931865324,
      "kl": 0.0020465850830078125,
      "learning_rate": 3.707195095691913e-07,
      "loss": 0.0112,
      "reward": 0.8081597238779068,
      "reward_std": 0.254116490483284,
      "rewards/accuracy_reward": 0.3576388955116272,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4505208358168602,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 888.3264007568359,
      "epoch": 0.6889952153110048,
      "grad_norm": 0.05215266241295597,
      "kl": 0.001735687255859375,
      "learning_rate": 3.4297073302098155e-07,
      "loss": 0.012,
      "reward": 0.7022569477558136,
      "reward_std": 0.22956418246030807,
      "rewards/accuracy_reward": 0.2986111156642437,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4036458432674408,
      "step": 36
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 855.5243225097656,
      "epoch": 0.7081339712918661,
      "grad_norm": 0.06846168763574559,
      "kl": 0.0021762847900390625,
      "learning_rate": 3.16187222484055e-07,
      "loss": 0.0195,
      "reward": 0.7786458283662796,
      "reward_std": 0.29831067472696304,
      "rewards/accuracy_reward": 0.3402777761220932,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4383680522441864,
      "step": 37
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 867.7257080078125,
      "epoch": 0.7272727272727273,
      "grad_norm": 0.06807879530009997,
      "kl": 0.0018215179443359375,
      "learning_rate": 2.904938550483098e-07,
      "loss": 0.0285,
      "reward": 0.6675347238779068,
      "reward_std": 0.21728801727294922,
      "rewards/accuracy_reward": 0.2638888917863369,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4036458358168602,
      "step": 38
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 877.5381927490234,
      "epoch": 0.7464114832535885,
      "grad_norm": 0.06659774200582641,
      "kl": 0.001922607421875,
      "learning_rate": 2.6601042505327635e-07,
      "loss": 0.0224,
      "reward": 0.732638880610466,
      "reward_std": 0.2946278154850006,
      "rewards/accuracy_reward": 0.3055555559694767,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4270833358168602,
      "step": 39
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 848.8159790039062,
      "epoch": 0.7655502392344498,
      "grad_norm": 0.06715335856665118,
      "kl": 0.0018100738525390625,
      "learning_rate": 2.4285108555160575e-07,
      "loss": 0.0216,
      "reward": 0.8125000149011612,
      "reward_std": 0.2357022687792778,
      "rewards/accuracy_reward": 0.3854166716337204,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4270833358168602,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 862.9201507568359,
      "epoch": 0.784688995215311,
      "grad_norm": 0.06259931940680961,
      "kl": 0.00200653076171875,
      "learning_rate": 2.2112381607484416e-07,
      "loss": 0.0198,
      "reward": 0.714409738779068,
      "reward_std": 0.24429557099938393,
      "rewards/accuracy_reward": 0.2916666716337204,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.422743059694767,
      "step": 41
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 863.4201354980469,
      "epoch": 0.8038277511961722,
      "grad_norm": 0.07414881526494982,
      "kl": 0.002193450927734375,
      "learning_rate": 2.0092991918301106e-07,
      "loss": 0.022,
      "reward": 0.7604166865348816,
      "reward_std": 0.22833655402064323,
      "rewards/accuracy_reward": 0.329861119389534,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4305555522441864,
      "step": 42
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 843.4826507568359,
      "epoch": 0.8229665071770335,
      "grad_norm": 0.06426826229363446,
      "kl": 0.0018749237060546875,
      "learning_rate": 1.8236354814530112e-07,
      "loss": 0.0145,
      "reward": 0.796875,
      "reward_std": 0.2111499235033989,
      "rewards/accuracy_reward": 0.357638880610466,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4392361119389534,
      "step": 43
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 878.6527709960938,
      "epoch": 0.8421052631578947,
      "grad_norm": 0.058633524755595696,
      "kl": 0.0018100738525390625,
      "learning_rate": 1.6551126795408015e-07,
      "loss": 0.0143,
      "reward": 0.6909722238779068,
      "reward_std": 0.18659762665629387,
      "rewards/accuracy_reward": 0.2951388880610466,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.3958333358168602,
      "step": 44
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 840.09375,
      "epoch": 0.861244019138756,
      "grad_norm": 0.07730488608092184,
      "kl": 0.0020885467529296875,
      "learning_rate": 1.5045165171893116e-07,
      "loss": 0.0289,
      "reward": 0.8420138955116272,
      "reward_std": 0.2676202580332756,
      "rewards/accuracy_reward": 0.3888888955116272,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.453125,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 847.3611297607422,
      "epoch": 0.8803827751196173,
      "grad_norm": 0.06479288468708434,
      "kl": 0.002285003662109375,
      "learning_rate": 1.3725491432254623e-07,
      "loss": 0.0195,
      "reward": 0.7777777910232544,
      "reward_std": 0.27989643067121506,
      "rewards/accuracy_reward": 0.3541666716337204,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4236111119389534,
      "step": 46
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 851.6944427490234,
      "epoch": 0.8995215311004785,
      "grad_norm": 0.05948308775990511,
      "kl": 0.0021038055419921875,
      "learning_rate": 1.259825850465308e-07,
      "loss": 0.0056,
      "reward": 0.7213541716337204,
      "reward_std": 0.19764616712927818,
      "rewards/accuracy_reward": 0.3159722276031971,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.405381940305233,
      "step": 47
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 831.1041717529297,
      "epoch": 0.9186602870813397,
      "grad_norm": 0.06976945326641983,
      "kl": 0.0020198822021484375,
      "learning_rate": 1.166872206934904e-07,
      "loss": 0.0315,
      "reward": 0.7795138955116272,
      "reward_std": 0.25043364614248276,
      "rewards/accuracy_reward": 0.3402777835726738,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4392361119389534,
      "step": 48
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 871.7847137451172,
      "epoch": 0.937799043062201,
      "grad_norm": 0.06380948814810407,
      "kl": 0.002155303955078125,
      "learning_rate": 1.0941216054295468e-07,
      "loss": 0.0197,
      "reward": 0.7526041716337204,
      "reward_std": 0.23447464033961296,
      "rewards/accuracy_reward": 0.3298611119389534,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4227430671453476,
      "step": 49
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 847.6632080078125,
      "epoch": 0.9569377990430622,
      "grad_norm": 0.07507185787901022,
      "kl": 0.0022182464599609375,
      "learning_rate": 1.0419132428365116e-07,
      "loss": 0.0201,
      "reward": 0.7786458283662796,
      "reward_std": 0.2590269520878792,
      "rewards/accuracy_reward": 0.329861119389534,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4487847238779068,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 848.7222290039062,
      "epoch": 0.9760765550239234,
      "grad_norm": 0.05934554686543501,
      "kl": 0.0019321441650390625,
      "learning_rate": 1.0104905386425732e-07,
      "loss": 0.0147,
      "reward": 0.7803819626569748,
      "reward_std": 0.22219848446547985,
      "rewards/accuracy_reward": 0.3402777835726738,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4401041641831398,
      "step": 51
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 920.5416870117188,
      "epoch": 0.9952153110047847,
      "grad_norm": 0.05965196947908554,
      "kl": 0.002071380615234375,
      "learning_rate": 1e-07,
      "loss": 0.0123,
      "reward": 0.7968750149011612,
      "reward_std": 0.2651650495827198,
      "rewards/accuracy_reward": 0.3715277761220932,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.4253472313284874,
      "step": 52
    },
    {
      "epoch": 0.9952153110047847,
      "step": 52,
      "total_flos": 0.0,
      "train_loss": 0.01933252687404792,
      "train_runtime": 2498.5955,
      "train_samples_per_second": 3.002,
      "train_steps_per_second": 0.021
    }
  ],
  "logging_steps": 1,
  "max_steps": 52,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 10,
  "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": 24,
  "trial_name": null,
  "trial_params": null
}