{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9964664310954063,
  "eval_steps": 100,
  "global_step": 282,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 917.2587478637695,
      "epoch": 0.0176678445229682,
      "grad_norm": 0.05332557484507561,
      "kl": 0.00013451576232910157,
      "learning_rate": 3.448275862068966e-06,
      "loss": 0.0289,
      "reward": 0.27843193283770235,
      "reward_std": 0.14517987524159254,
      "rewards/accuracy_reward": 0.21607143768342213,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.062360493757296354,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 847.1094116210937,
      "epoch": 0.0353356890459364,
      "grad_norm": 0.053962819278240204,
      "kl": 0.001752650737762451,
      "learning_rate": 6.896551724137932e-06,
      "loss": 0.0249,
      "reward": 0.4628627423895523,
      "reward_std": 0.15404398429673166,
      "rewards/accuracy_reward": 0.3459821574622765,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.11688058525323868,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 732.8753692626954,
      "epoch": 0.053003533568904596,
      "grad_norm": 0.04866955429315567,
      "kl": 0.010287094116210937,
      "learning_rate": 1.0344827586206898e-05,
      "loss": 0.0172,
      "reward": 0.8053571786731482,
      "reward_std": 0.18580246912315487,
      "rewards/accuracy_reward": 0.5641741327941417,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.2411830423399806,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 706.6745796203613,
      "epoch": 0.0706713780918728,
      "grad_norm": 0.051155973225831985,
      "kl": 0.02039794921875,
      "learning_rate": 1.3793103448275863e-05,
      "loss": 0.0149,
      "reward": 0.86157928109169,
      "reward_std": 0.1875305467285216,
      "rewards/accuracy_reward": 0.6122768152505159,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24930245764553546,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 710.6496971130371,
      "epoch": 0.08833922261484099,
      "grad_norm": 0.05199262499809265,
      "kl": 0.02178192138671875,
      "learning_rate": 1.7241379310344828e-05,
      "loss": 0.0149,
      "reward": 0.8626953538507223,
      "reward_std": 0.17808558829128743,
      "rewards/accuracy_reward": 0.6143973529338836,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.2482979955151677,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 714.621907043457,
      "epoch": 0.10600706713780919,
      "grad_norm": 0.05970175191760063,
      "kl": 0.02830352783203125,
      "learning_rate": 1.999923511388017e-05,
      "loss": 0.015,
      "reward": 0.860267897695303,
      "reward_std": 0.1923450878355652,
      "rewards/accuracy_reward": 0.6117187786847353,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24854911137372254,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 706.1554992675781,
      "epoch": 0.12367491166077739,
      "grad_norm": 0.064034104347229,
      "kl": 0.03692779541015625,
      "learning_rate": 1.9972476383747748e-05,
      "loss": 0.0133,
      "reward": 0.8735770478844642,
      "reward_std": 0.19234329210594298,
      "rewards/accuracy_reward": 0.625111635401845,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24846540708094836,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 695.877042388916,
      "epoch": 0.1413427561837456,
      "grad_norm": 0.09493730962276459,
      "kl": 0.0706024169921875,
      "learning_rate": 1.9907590277344582e-05,
      "loss": 0.0091,
      "reward": 0.854743342846632,
      "reward_std": 0.21510332841426133,
      "rewards/accuracy_reward": 0.6077009219676256,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24704242013394834,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 704.042106628418,
      "epoch": 0.15901060070671377,
      "grad_norm": 0.06780606508255005,
      "kl": 0.129840087890625,
      "learning_rate": 1.9804824871166254e-05,
      "loss": 0.0127,
      "reward": 0.8354911088943482,
      "reward_std": 0.2541661562398076,
      "rewards/accuracy_reward": 0.5992187757045031,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.23627233169972897,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 723.2655464172364,
      "epoch": 0.17667844522968199,
      "grad_norm": 0.13506263494491577,
      "kl": 0.2400848388671875,
      "learning_rate": 1.9664573064143604e-05,
      "loss": 0.0784,
      "reward": 0.817243343964219,
      "reward_std": 0.28774973265826703,
      "rewards/accuracy_reward": 0.5887277062982321,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.22851563505828382,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 732.3425590515137,
      "epoch": 0.19434628975265017,
      "grad_norm": 0.1647893786430359,
      "kl": 0.31966552734375,
      "learning_rate": 1.948737107548771e-05,
      "loss": 0.1074,
      "reward": 0.7718192338943481,
      "reward_std": 0.3275716399773955,
      "rewards/accuracy_reward": 0.5561384186148643,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.21568081323057414,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 713.8921081542969,
      "epoch": 0.21201413427561838,
      "grad_norm": 0.14291171729564667,
      "kl": 0.58829345703125,
      "learning_rate": 1.9273896394584103e-05,
      "loss": 0.1926,
      "reward": 0.6627232447266579,
      "reward_std": 0.39091945756226776,
      "rewards/accuracy_reward": 0.47834823541343213,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.18437500866129994,
      "step": 60
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 610.0054931640625,
      "epoch": 0.22968197879858657,
      "grad_norm": 0.10107492655515671,
      "kl": 0.25198974609375,
      "learning_rate": 1.9024965190774262e-05,
      "loss": 0.1113,
      "reward": 0.8659877650439739,
      "reward_std": 0.27730538016185163,
      "rewards/accuracy_reward": 0.6375000283122063,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.22848773319274188,
      "step": 65
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 578.1935543060303,
      "epoch": 0.24734982332155478,
      "grad_norm": 0.07951921969652176,
      "kl": 0.2242431640625,
      "learning_rate": 1.8741529192927528e-05,
      "loss": 0.0939,
      "reward": 0.8772321827709675,
      "reward_std": 0.26367146205157044,
      "rewards/accuracy_reward": 0.6401786014437676,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.23705358244478703,
      "step": 70
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 606.7115242004395,
      "epoch": 0.26501766784452296,
      "grad_norm": 0.09012111276388168,
      "kl": 0.3244873046875,
      "learning_rate": 1.8424672050733577e-05,
      "loss": 0.1685,
      "reward": 0.8198382072150707,
      "reward_std": 0.3401181472465396,
      "rewards/accuracy_reward": 0.5989955652505159,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.22084264531731607,
      "step": 75
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 571.5734634399414,
      "epoch": 0.2826855123674912,
      "grad_norm": 0.06397970765829086,
      "kl": 0.164495849609375,
      "learning_rate": 1.8075605191627242e-05,
      "loss": 0.0724,
      "reward": 0.9197545051574707,
      "reward_std": 0.22428997224196792,
      "rewards/accuracy_reward": 0.6797991372644901,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.23995536621659994,
      "step": 80
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 606.1346244812012,
      "epoch": 0.3003533568904594,
      "grad_norm": 0.07042359560728073,
      "kl": 0.234246826171875,
      "learning_rate": 1.7695663189185703e-05,
      "loss": 0.0965,
      "reward": 0.8753627657890319,
      "reward_std": 0.25722263269126416,
      "rewards/accuracy_reward": 0.6383928876370192,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.23696987684816123,
      "step": 85
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 593.6537105560303,
      "epoch": 0.31802120141342755,
      "grad_norm": 0.06578419357538223,
      "kl": 0.224627685546875,
      "learning_rate": 1.7286298660705877e-05,
      "loss": 0.0928,
      "reward": 0.8682199046015739,
      "reward_std": 0.2509442804381251,
      "rewards/accuracy_reward": 0.6315848492085934,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.236635054461658,
      "step": 90
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 563.4647586822509,
      "epoch": 0.33568904593639576,
      "grad_norm": 0.06993943452835083,
      "kl": 0.238482666015625,
      "learning_rate": 1.6849076713469914e-05,
      "loss": 0.1089,
      "reward": 0.8941406659781933,
      "reward_std": 0.2569141635671258,
      "rewards/accuracy_reward": 0.6537946693599224,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24034599158912898,
      "step": 95
    },
    {
      "epoch": 0.35335689045936397,
      "grad_norm": 0.06800527125597,
      "learning_rate": 1.6385668960932143e-05,
      "loss": 0.0909,
      "step": 100
    },
    {
      "epoch": 0.35335689045936397,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 590.76025390625,
      "eval_kl": 0.19091796875,
      "eval_loss": 0.074989914894104,
      "eval_reward": 0.9082031697034836,
      "eval_reward_std": 0.2306336797773838,
      "eval_rewards/accuracy_reward": 0.667410746216774,
      "eval_rewards/format_reward": 0.0,
      "eval_rewards/tag_count_reward": 0.2407924234867096,
      "eval_runtime": 131.5321,
      "eval_samples_per_second": 0.753,
      "eval_steps_per_second": 0.008,
      "step": 100
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 586.0422695159912,
      "epoch": 0.3710247349823322,
      "grad_norm": 0.07303127646446228,
      "kl": 0.2162933349609375,
      "learning_rate": 1.5897847131705194e-05,
      "loss": 0.1027,
      "reward": 0.8979353092610836,
      "reward_std": 0.25196963488124313,
      "rewards/accuracy_reward": 0.6575335115194321,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.2404017962515354,
      "step": 105
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 577.3520366668702,
      "epoch": 0.38869257950530034,
      "grad_norm": 0.07516764849424362,
      "kl": 0.196807861328125,
      "learning_rate": 1.5387476295779737e-05,
      "loss": 0.0907,
      "reward": 0.9079520504921674,
      "reward_std": 0.252688442543149,
      "rewards/accuracy_reward": 0.6669643156230449,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24098773431032897,
      "step": 110
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 578.9473468780518,
      "epoch": 0.40636042402826855,
      "grad_norm": 0.0895974189043045,
      "kl": 0.27685546875,
      "learning_rate": 1.4856507733875837e-05,
      "loss": 0.1298,
      "reward": 0.8826172292232514,
      "reward_std": 0.28199427481740713,
      "rewards/accuracy_reward": 0.6393973510712385,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24321987628936767,
      "step": 115
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 564.4585056304932,
      "epoch": 0.42402826855123676,
      "grad_norm": 0.07208839058876038,
      "kl": 0.219805908203125,
      "learning_rate": 1.4306971477188223e-05,
      "loss": 0.0953,
      "reward": 0.900697585940361,
      "reward_std": 0.25064537590369584,
      "rewards/accuracy_reward": 0.6555803909897804,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.2451171986758709,
      "step": 120
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 619.9358505249023,
      "epoch": 0.4416961130742049,
      "grad_norm": 0.08019658923149109,
      "kl": 0.235772705078125,
      "learning_rate": 1.3740968546047935e-05,
      "loss": 0.105,
      "reward": 0.8835379853844643,
      "reward_std": 0.2642295270226896,
      "rewards/accuracy_reward": 0.6447544928640128,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.23878349252045156,
      "step": 125
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 602.9432163238525,
      "epoch": 0.45936395759717313,
      "grad_norm": 0.07495882362127304,
      "kl": 0.228643798828125,
      "learning_rate": 1.3160662917174045e-05,
      "loss": 0.1076,
      "reward": 0.9091239273548126,
      "reward_std": 0.2478904782794416,
      "rewards/accuracy_reward": 0.6672991372644901,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24182478711009026,
      "step": 130
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 571.6796016693115,
      "epoch": 0.47703180212014135,
      "grad_norm": 0.08212456852197647,
      "kl": 0.197271728515625,
      "learning_rate": 1.2568273250226681e-05,
      "loss": 0.11,
      "reward": 0.9201730333268643,
      "reward_std": 0.24961013691499828,
      "rewards/accuracy_reward": 0.6765625298023223,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24361050203442575,
      "step": 135
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 570.3125244140625,
      "epoch": 0.49469964664310956,
      "grad_norm": 0.07115107774734497,
      "kl": 0.2134033203125,
      "learning_rate": 1.1966064405292887e-05,
      "loss": 0.1111,
      "reward": 0.9121094167232513,
      "reward_std": 0.25566268572583795,
      "rewards/accuracy_reward": 0.6696428872644902,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24246652834117413,
      "step": 140
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 599.4550518035888,
      "epoch": 0.5123674911660777,
      "grad_norm": 0.07841041684150696,
      "kl": 0.18359375,
      "learning_rate": 1.1356338783736256e-05,
      "loss": 0.0858,
      "reward": 0.9083984784781933,
      "reward_std": 0.22585448753088713,
      "rewards/accuracy_reward": 0.6655134223401546,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24288505520671605,
      "step": 145
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 589.7945564270019,
      "epoch": 0.5300353356890459,
      "grad_norm": 0.07215780019760132,
      "kl": 0.148455810546875,
      "learning_rate": 1.0741427525516463e-05,
      "loss": 0.0668,
      "reward": 0.9342076353728771,
      "reward_std": 0.22251177271828054,
      "rewards/accuracy_reward": 0.6893973518162966,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.2448102779686451,
      "step": 150
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 577.0310535430908,
      "epoch": 0.5477031802120141,
      "grad_norm": 0.0739278495311737,
      "kl": 0.190936279296875,
      "learning_rate": 1.012368159663363e-05,
      "loss": 0.098,
      "reward": 0.9136719211935997,
      "reward_std": 0.2538995613344014,
      "rewards/accuracy_reward": 0.6722098514437675,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.2414620641618967,
      "step": 155
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 565.5535991668701,
      "epoch": 0.5653710247349824,
      "grad_norm": 0.08016849309206009,
      "kl": 0.1893310546875,
      "learning_rate": 9.505462800772612e-06,
      "loss": 0.0997,
      "reward": 0.9155971385538578,
      "reward_std": 0.2556414810940623,
      "rewards/accuracy_reward": 0.6753348506987095,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24026228599250316,
      "step": 160
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 584.8990207672119,
      "epoch": 0.5830388692579506,
      "grad_norm": 0.07024497538805008,
      "kl": 0.178204345703125,
      "learning_rate": 8.889134749511956e-06,
      "loss": 0.0891,
      "reward": 0.9097935684025288,
      "reward_std": 0.2374617088586092,
      "rewards/accuracy_reward": 0.668080385774374,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24171318039298056,
      "step": 165
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 617.6932182312012,
      "epoch": 0.6007067137809188,
      "grad_norm": 0.06575129926204681,
      "kl": 0.177227783203125,
      "learning_rate": 8.277053825620836e-06,
      "loss": 0.0803,
      "reward": 0.8907087415456771,
      "reward_std": 0.24254968659952283,
      "rewards/accuracy_reward": 0.6465402048081159,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24416853692382573,
      "step": 170
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 592.3628578186035,
      "epoch": 0.6183745583038869,
      "grad_norm": 0.07182252407073975,
      "kl": 0.195758056640625,
      "learning_rate": 7.671560173993588e-06,
      "loss": 0.0894,
      "reward": 0.9130301713943482,
      "reward_std": 0.23459300631657243,
      "rewards/accuracy_reward": 0.665736635401845,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24729353692382575,
      "step": 175
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 559.4955589294434,
      "epoch": 0.6360424028268551,
      "grad_norm": 0.09263365715742111,
      "kl": 0.205853271484375,
      "learning_rate": 7.07496875466589e-06,
      "loss": 0.0981,
      "reward": 0.9218192398548126,
      "reward_std": 0.23992226729169489,
      "rewards/accuracy_reward": 0.6737723559141159,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24804688580334186,
      "step": 180
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 577.3553829193115,
      "epoch": 0.6537102473498233,
      "grad_norm": 0.07366114854812622,
      "kl": 0.200531005859375,
      "learning_rate": 6.489560492119225e-06,
      "loss": 0.096,
      "reward": 0.9261161163449287,
      "reward_std": 0.2363728086464107,
      "rewards/accuracy_reward": 0.6770089607685804,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.2491071542724967,
      "step": 185
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 593.0000240325928,
      "epoch": 0.6713780918727915,
      "grad_norm": 0.06559525430202484,
      "kl": 0.21243896484375,
      "learning_rate": 5.9175735547120975e-06,
      "loss": 0.0995,
      "reward": 0.9242466956377029,
      "reward_std": 0.24610999850556253,
      "rewards/accuracy_reward": 0.6748884223401547,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24935827050358056,
      "step": 190
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 587.670785522461,
      "epoch": 0.6890459363957597,
      "grad_norm": 0.09491094201803207,
      "kl": 0.19505615234375,
      "learning_rate": 5.361194797579108e-06,
      "loss": 0.0807,
      "reward": 0.9143694624304771,
      "reward_std": 0.2263300753198564,
      "rewards/accuracy_reward": 0.667410746961832,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24695871509611605,
      "step": 195
    },
    {
      "epoch": 0.7067137809187279,
      "grad_norm": 0.07209863513708115,
      "learning_rate": 4.8225514017138205e-06,
      "loss": 0.1019,
      "step": 200
    },
    {
      "epoch": 0.7067137809187279,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 597.6492767333984,
      "eval_kl": 0.2119140625,
      "eval_loss": 0.09467914700508118,
      "eval_reward": 0.9132254868745804,
      "eval_reward_std": 0.22664131596684456,
      "eval_rewards/accuracy_reward": 0.6685268133878708,
      "eval_rewards/format_reward": 0.0,
      "eval_rewards/tag_count_reward": 0.2446986697614193,
      "eval_runtime": 131.7381,
      "eval_samples_per_second": 0.751,
      "eval_steps_per_second": 0.008,
      "step": 200
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 593.5689434051513,
      "epoch": 0.7243816254416962,
      "grad_norm": 0.06886138767004013,
      "kl": 0.2204437255859375,
      "learning_rate": 4.303702741201431e-06,
      "loss": 0.0962,
      "reward": 0.9143136594444513,
      "reward_std": 0.2504740643315017,
      "rewards/accuracy_reward": 0.6694754760712385,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24483818002045155,
      "step": 205
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 590.6283740997314,
      "epoch": 0.7420494699646644,
      "grad_norm": 0.07291878759860992,
      "kl": 0.195550537109375,
      "learning_rate": 3.8066325096949153e-06,
      "loss": 0.0884,
      "reward": 0.914955398440361,
      "reward_std": 0.23502108193933963,
      "rewards/accuracy_reward": 0.666517884656787,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.2484375100582838,
      "step": 210
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 573.4894218444824,
      "epoch": 0.7597173144876325,
      "grad_norm": 0.07844824343919754,
      "kl": 0.20074462890625,
      "learning_rate": 3.3332411362372063e-06,
      "loss": 0.0823,
      "reward": 0.9080915614962578,
      "reward_std": 0.22488524662330747,
      "rewards/accuracy_reward": 0.658370564877987,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24972099326550962,
      "step": 215
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 572.3277076721191,
      "epoch": 0.7773851590106007,
      "grad_norm": 0.06696169078350067,
      "kl": 0.189166259765625,
      "learning_rate": 2.8853385194256677e-06,
      "loss": 0.0805,
      "reward": 0.9161830775439739,
      "reward_std": 0.22096735332161188,
      "rewards/accuracy_reward": 0.6659598544239997,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.25022322554141285,
      "step": 220
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 591.3379741668701,
      "epoch": 0.7950530035335689,
      "grad_norm": 0.1020839512348175,
      "kl": 0.213531494140625,
      "learning_rate": 2.464637107698046e-06,
      "loss": 0.0864,
      "reward": 0.9009207993745804,
      "reward_std": 0.23703113198280334,
      "rewards/accuracy_reward": 0.6501116339117289,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.25080916434526446,
      "step": 225
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 579.4515911102295,
      "epoch": 0.8127208480565371,
      "grad_norm": 0.08135361969470978,
      "kl": 0.2225341796875,
      "learning_rate": 2.072745352195794e-06,
      "loss": 0.0975,
      "reward": 0.9147600896656514,
      "reward_std": 0.23770583430305123,
      "rewards/accuracy_reward": 0.6649553872644901,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24980470035225152,
      "step": 230
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 573.1327266693115,
      "epoch": 0.8303886925795053,
      "grad_norm": 0.07584529370069504,
      "kl": 0.2205078125,
      "learning_rate": 1.7111615572361628e-06,
      "loss": 0.1078,
      "reward": 0.9006138816475868,
      "reward_std": 0.24386549470946192,
      "rewards/accuracy_reward": 0.6527902077883482,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24782367032021285,
      "step": 235
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 574.2280414581298,
      "epoch": 0.8480565371024735,
      "grad_norm": 0.07971746474504471,
      "kl": 0.22808837890625,
      "learning_rate": 1.381268151904298e-06,
      "loss": 0.1048,
      "reward": 0.889174147695303,
      "reward_std": 0.24307329086586832,
      "rewards/accuracy_reward": 0.641852705180645,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24732143860310316,
      "step": 240
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 580.1797130584716,
      "epoch": 0.8657243816254417,
      "grad_norm": 0.0819220319390297,
      "kl": 0.2184326171875,
      "learning_rate": 1.0843264046665558e-06,
      "loss": 0.1038,
      "reward": 0.9068638809025288,
      "reward_std": 0.24303168477490544,
      "rewards/accuracy_reward": 0.6598214615136385,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24704242143779992,
      "step": 245
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 586.1197792053223,
      "epoch": 0.8833922261484098,
      "grad_norm": 0.07211937755346298,
      "kl": 0.21861572265625,
      "learning_rate": 8.214716012124491e-07,
      "loss": 0.0972,
      "reward": 0.8969587460160255,
      "reward_std": 0.24261261774227022,
      "rewards/accuracy_reward": 0.6494419921189547,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24751675240695475,
      "step": 250
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 581.7469051361084,
      "epoch": 0.901060070671378,
      "grad_norm": 0.07039004564285278,
      "kl": 0.1945556640625,
      "learning_rate": 5.937087039615619e-07,
      "loss": 0.0798,
      "reward": 0.9000837452709675,
      "reward_std": 0.22726329311262816,
      "rewards/accuracy_reward": 0.6511161029338837,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24896764568984509,
      "step": 255
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 568.4496898651123,
      "epoch": 0.9187279151943463,
      "grad_norm": 0.07688862830400467,
      "kl": 0.1930908203125,
      "learning_rate": 4.019085098303077e-07,
      "loss": 0.0886,
      "reward": 0.925837092846632,
      "reward_std": 0.229900080896914,
      "rewards/accuracy_reward": 0.6780134219676256,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24782367069274186,
      "step": 260
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 585.3359642028809,
      "epoch": 0.9363957597173145,
      "grad_norm": 0.07333354651927948,
      "kl": 0.199957275390625,
      "learning_rate": 2.4680432094837394e-07,
      "loss": 0.0913,
      "reward": 0.9235212452709675,
      "reward_std": 0.2362320901826024,
      "rewards/accuracy_reward": 0.6742187783122062,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.2493024656549096,
      "step": 265
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 589.8091770172119,
      "epoch": 0.9540636042402827,
      "grad_norm": 0.07898814976215363,
      "kl": 0.199560546875,
      "learning_rate": 1.289891410535593e-07,
      "loss": 0.0874,
      "reward": 0.9116071857511997,
      "reward_std": 0.22943277563899755,
      "rewards/accuracy_reward": 0.6623884223401546,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.249218762293458,
      "step": 270
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 589.4594024658203,
      "epoch": 0.9717314487632509,
      "grad_norm": 0.07111751288175583,
      "kl": 0.20418701171875,
      "learning_rate": 4.8913408283934874e-08,
      "loss": 0.0867,
      "reward": 0.9023716941475868,
      "reward_std": 0.23179991263896227,
      "rewards/accuracy_reward": 0.6522321693599225,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.25013952106237414,
      "step": 275
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 583.0817214965821,
      "epoch": 0.9893992932862191,
      "grad_norm": 0.07049054652452469,
      "kl": 0.199383544921875,
      "learning_rate": 6.883273035447335e-09,
      "loss": 0.0902,
      "reward": 0.9261998198926449,
      "reward_std": 0.22382794730365277,
      "rewards/accuracy_reward": 0.6766741391271353,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.24952568095177413,
      "step": 280
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 574.5289487838745,
      "epoch": 0.9964664310954063,
      "kl": 0.199249267578125,
      "reward": 0.9313616473227739,
      "reward_std": 0.22618725849315524,
      "rewards/accuracy_reward": 0.6813616417348385,
      "rewards/format_reward": 0.0,
      "rewards/tag_count_reward": 0.2500000102445483,
      "step": 282,
      "total_flos": 0.0,
      "train_loss": 0.08527955120918176,
      "train_runtime": 119365.0462,
      "train_samples_per_second": 0.607,
      "train_steps_per_second": 0.002
    }
  ],
  "logging_steps": 5,
  "max_steps": 283,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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
}