{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.041736227045075125,
  "eval_steps": 500,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "cache_hit_rate": 0.6749276161193848,
      "cache_hit_rate_size16": 0.6550344030062357,
      "cache_hit_rate_size32": 0.8710664212703705,
      "cache_hit_rate_size8": 0.4304385483264923,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.2875,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 926.0,
      "completions/mean_length": 732.10625,
      "completions/mean_terminated_length": 595.8285858154297,
      "completions/min_length": 208.4,
      "completions/min_terminated_length": 208.4,
      "entropy": 0.6143155984580517,
      "epoch": 0.002086811352253756,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.1999109834432602,
      "kl": 0.002709078643238172,
      "learning_rate": 9e-05,
      "loss": -0.020685632526874543,
      "num_tokens": 148025.0,
      "reward": 0.6688930571079255,
      "reward_std": 0.10492694256827235,
      "rewards/cache_reward/mean": 0.6688930571079255,
      "rewards/cache_reward/std": 0.10492694256827235,
      "step": 10,
      "step_time": 109.66132322684862
    },
    {
      "cache_hit_rate": 0.6733510851860046,
      "cache_hit_rate_size16": 0.7422660708427429,
      "cache_hit_rate_size32": 0.8451784491539002,
      "cache_hit_rate_size8": 0.4291795551776886,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.38125,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 988.3,
      "completions/mean_length": 870.26875,
      "completions/mean_terminated_length": 780.20380859375,
      "completions/min_length": 517.0,
      "completions/min_terminated_length": 517.0,
      "entropy": 0.5574822902679444,
      "epoch": 0.004173622704507512,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.27785834670066833,
      "kl": 0.008998260926455259,
      "learning_rate": 0.0001,
      "loss": -0.04589595198631287,
      "num_tokens": 324300.0,
      "reward": 0.6649930059909821,
      "reward_std": 0.09102208893746137,
      "rewards/cache_reward/mean": 0.6649930059909821,
      "rewards/cache_reward/std": 0.09102208893746137,
      "step": 20,
      "step_time": 109.77420157198794
    },
    {
      "cache_hit_rate": 0.7234689474105835,
      "cache_hit_rate_size16": 0.7119116663932801,
      "cache_hit_rate_size32": 0.8416047170758247,
      "cache_hit_rate_size8": 0.3245641191800435,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.2625,
      "completions/max_length": 989.8,
      "completions/max_terminated_length": 914.9,
      "completions/mean_length": 701.85625,
      "completions/mean_terminated_length": 626.0504425048828,
      "completions/min_length": 341.9,
      "completions/min_terminated_length": 341.9,
      "entropy": 0.6420590102672576,
      "epoch": 0.006260434056761269,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.2189197838306427,
      "kl": 0.015754866227507592,
      "learning_rate": 0.0001,
      "loss": -0.057462894916534425,
      "num_tokens": 461973.0,
      "reward": 0.7092925369739532,
      "reward_std": 0.09992160452529789,
      "rewards/cache_reward/mean": 0.7092925369739532,
      "rewards/cache_reward/std": 0.09992160452529789,
      "step": 30,
      "step_time": 105.50118747297674
    },
    {
      "cache_hit_rate": 0.6422885358333588,
      "cache_hit_rate_size16": 0.7278285367148263,
      "cache_hit_rate_size32": 0.8580973446369171,
      "cache_hit_rate_size8": 0.37648922204971313,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.4125,
      "completions/max_length": 1017.0,
      "completions/max_terminated_length": 930.3,
      "completions/mean_length": 831.81875,
      "completions/mean_terminated_length": 727.4821502685547,
      "completions/min_length": 484.1,
      "completions/min_terminated_length": 484.1,
      "entropy": 0.5546563148498536,
      "epoch": 0.008347245409015025,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.2441343367099762,
      "kl": 0.021256055543199183,
      "learning_rate": 0.0001,
      "loss": -0.06346478462219238,
      "num_tokens": 636696.0,
      "reward": 0.6391598552465438,
      "reward_std": 0.13217019997537136,
      "rewards/cache_reward/mean": 0.6391598552465438,
      "rewards/cache_reward/std": 0.13217019997537136,
      "step": 40,
      "step_time": 109.29803132868838
    },
    {
      "cache_hit_rate": 0.6705451369285583,
      "cache_hit_rate_size16": 0.6974249720573426,
      "cache_hit_rate_size32": 0.8658615776470729,
      "cache_hit_rate_size8": 0.4653993273774783,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.34375,
      "completions/max_length": 1021.8,
      "completions/max_terminated_length": 879.7,
      "completions/mean_length": 794.4875,
      "completions/mean_terminated_length": 615.1027160644531,
      "completions/min_length": 435.1,
      "completions/min_terminated_length": 332.7,
      "entropy": 0.6008389577269554,
      "epoch": 0.010434056761268781,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.23809342086315155,
      "kl": 0.03263230957090855,
      "learning_rate": 0.0001,
      "loss": -0.04350569248199463,
      "num_tokens": 798422.0,
      "reward": 0.688891452550888,
      "reward_std": 0.1476906482130289,
      "rewards/cache_reward/mean": 0.688891452550888,
      "rewards/cache_reward/std": 0.1476906482130289,
      "step": 50,
      "step_time": 109.32008387609385
    },
    {
      "cache_hit_rate": 0.5735559463500977,
      "cache_hit_rate_size16": 0.6513433158397675,
      "cache_hit_rate_size32": 0.8832110166549683,
      "cache_hit_rate_size8": 0.4417416916953193,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.39375,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 950.8,
      "completions/mean_length": 820.71875,
      "completions/mean_terminated_length": 705.5888244628907,
      "completions/min_length": 393.1,
      "completions/min_terminated_length": 393.1,
      "entropy": 0.5146959632635116,
      "epoch": 0.012520868113522538,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.2036655694246292,
      "kl": 0.03261517193168402,
      "learning_rate": 0.0001,
      "loss": -0.06676955223083496,
      "num_tokens": 977289.0,
      "reward": 0.5551395505666733,
      "reward_std": 0.130782650411129,
      "rewards/cache_reward/mean": 0.5551395505666733,
      "rewards/cache_reward/std": 0.130782650411129,
      "step": 60,
      "step_time": 110.19092689147219
    },
    {
      "cache_hit_rate": 0.7626929998397827,
      "cache_hit_rate_size16": 0.6949937641620636,
      "cache_hit_rate_size32": 0.8773939145935906,
      "cache_hit_rate_size8": 0.5052446524302164,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.375,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 960.0,
      "completions/mean_length": 831.425,
      "completions/mean_terminated_length": 736.7794067382813,
      "completions/min_length": 410.3,
      "completions/min_terminated_length": 410.3,
      "entropy": 0.550393845140934,
      "epoch": 0.014607679465776294,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.2976628541946411,
      "kl": 0.04141218531876802,
      "learning_rate": 0.0001,
      "loss": -0.028155168890953063,
      "num_tokens": 1149221.0,
      "reward": 0.765302586555481,
      "reward_std": 0.10612610904499889,
      "rewards/cache_reward/mean": 0.765302586555481,
      "rewards/cache_reward/std": 0.10612610904499889,
      "step": 70,
      "step_time": 109.80838733934797
    },
    {
      "cache_hit_rate": 0.6532172083854675,
      "cache_hit_rate_size16": 0.7496759465762547,
      "cache_hit_rate_size32": 0.8642965952555338,
      "cache_hit_rate_size8": 0.4807915687561035,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.475,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 958.0,
      "completions/mean_length": 828.0625,
      "completions/mean_terminated_length": 696.7795776367187,
      "completions/min_length": 460.7,
      "completions/min_terminated_length": 460.7,
      "entropy": 0.5280586190521717,
      "epoch": 0.01669449081803005,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.21478967368602753,
      "kl": 0.05440526399761438,
      "learning_rate": 0.0001,
      "loss": -0.047751826047897336,
      "num_tokens": 1317623.0,
      "reward": 0.6507435321807862,
      "reward_std": 0.1329113408923149,
      "rewards/cache_reward/mean": 0.6507435321807862,
      "rewards/cache_reward/std": 0.1329113408923149,
      "step": 80,
      "step_time": 109.61795915367547
    },
    {
      "cache_hit_rate": 0.628938740491867,
      "cache_hit_rate_size16": 0.6589138408501943,
      "cache_hit_rate_size32": 0.8045709133148193,
      "cache_hit_rate_size8": 0.4631062958921705,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.39375,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 922.1,
      "completions/mean_length": 763.0125,
      "completions/mean_terminated_length": 620.8271697998047,
      "completions/min_length": 354.5,
      "completions/min_terminated_length": 354.5,
      "entropy": 0.6081107705831528,
      "epoch": 0.018781302170283808,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.2587888240814209,
      "kl": 0.06952012721449137,
      "learning_rate": 0.0001,
      "loss": -0.006095578521490097,
      "num_tokens": 1485377.0,
      "reward": 0.6154629409313201,
      "reward_std": 0.10154437553137541,
      "rewards/cache_reward/mean": 0.6154629409313201,
      "rewards/cache_reward/std": 0.10154437553137541,
      "step": 90,
      "step_time": 110.01140693970956
    },
    {
      "cache_hit_rate": 0.6613744765520095,
      "cache_hit_rate_size16": 0.7402904033660889,
      "cache_hit_rate_size32": 0.8814258575439453,
      "cache_hit_rate_size8": 0.42935203909873965,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.425,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 943.8,
      "completions/mean_length": 829.99375,
      "completions/mean_terminated_length": 697.3199035644532,
      "completions/min_length": 391.9,
      "completions/min_terminated_length": 391.9,
      "entropy": 0.5821917943656445,
      "epoch": 0.020868113522537562,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.23848511278629303,
      "kl": 0.07018006853759289,
      "learning_rate": 0.0001,
      "loss": -0.031068474054336548,
      "num_tokens": 1647416.0,
      "reward": 0.6516854614019394,
      "reward_std": 0.08088633958250284,
      "rewards/cache_reward/mean": 0.6516854614019394,
      "rewards/cache_reward/std": 0.08088633958250284,
      "step": 100,
      "step_time": 109.22561302329413
    },
    {
      "cache_hit_rate": 0.7659814059734344,
      "cache_hit_rate_size16": 0.7210325002670288,
      "cache_hit_rate_size32": 0.8733292288250394,
      "cache_hit_rate_size8": 0.5074326276779175,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.39375,
      "completions/max_length": 1000.1,
      "completions/max_terminated_length": 903.8,
      "completions/mean_length": 738.875,
      "completions/mean_terminated_length": 597.613003540039,
      "completions/min_length": 244.4,
      "completions/min_terminated_length": 244.4,
      "entropy": 0.6063698619604111,
      "epoch": 0.022954924874791317,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.22822512686252594,
      "kl": 0.0519987877458334,
      "learning_rate": 0.0001,
      "loss": -0.004398582130670547,
      "num_tokens": 1809740.0,
      "reward": 0.7583289802074432,
      "reward_std": 0.11587669281288981,
      "rewards/cache_reward/mean": 0.7583289802074432,
      "rewards/cache_reward/std": 0.11587669281288981,
      "step": 110,
      "step_time": 107.61979042319581
    },
    {
      "cache_hit_rate": 0.667035973072052,
      "cache_hit_rate_size16": 0.6942717552185058,
      "cache_hit_rate_size32": 0.8527272343635559,
      "cache_hit_rate_size8": 0.40293918251991273,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.175,
      "completions/max_length": 1003.6,
      "completions/max_terminated_length": 930.3,
      "completions/mean_length": 665.3625,
      "completions/mean_terminated_length": 598.4001220703125,
      "completions/min_length": 247.0,
      "completions/min_terminated_length": 247.0,
      "entropy": 0.7441994078457356,
      "epoch": 0.025041736227045076,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.21977545320987701,
      "kl": 0.06097110863775015,
      "learning_rate": 0.0001,
      "loss": -0.05008636116981506,
      "num_tokens": 1947502.0,
      "reward": 0.6475550562143326,
      "reward_std": 0.11349433939903975,
      "rewards/cache_reward/mean": 0.6475550562143326,
      "rewards/cache_reward/std": 0.11349433939903975,
      "step": 120,
      "step_time": 107.30077534515877
    },
    {
      "cache_hit_rate": 0.686054265499115,
      "cache_hit_rate_size16": 0.6798208653926849,
      "cache_hit_rate_size32": 0.871903528769811,
      "cache_hit_rate_size8": 0.443097785115242,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.2875,
      "completions/max_length": 986.6,
      "completions/max_terminated_length": 909.5,
      "completions/mean_length": 661.2125,
      "completions/mean_terminated_length": 539.6267791748047,
      "completions/min_length": 202.2,
      "completions/min_terminated_length": 202.2,
      "entropy": 0.685629041492939,
      "epoch": 0.02712854757929883,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.2505522072315216,
      "kl": 0.06346707474440336,
      "learning_rate": 0.0001,
      "loss": -0.08610177040100098,
      "num_tokens": 2106072.0,
      "reward": 0.6472916662693023,
      "reward_std": 0.10670399386435747,
      "rewards/cache_reward/mean": 0.6472916662693023,
      "rewards/cache_reward/std": 0.10670399386435747,
      "step": 130,
      "step_time": 106.4104538091924
    },
    {
      "cache_hit_rate": 0.7193193137645721,
      "cache_hit_rate_size16": 0.6384188731511434,
      "cache_hit_rate_size32": 0.9035206958651543,
      "cache_hit_rate_size8": 0.45546102098056246,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.49375,
      "completions/max_length": 1011.1,
      "completions/max_terminated_length": 949.7,
      "completions/mean_length": 824.2125,
      "completions/mean_terminated_length": 655.7207336425781,
      "completions/min_length": 250.9,
      "completions/min_terminated_length": 250.9,
      "entropy": 0.5345787741243839,
      "epoch": 0.02921535893155259,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.24551904201507568,
      "kl": 0.05889502316713333,
      "learning_rate": 0.0001,
      "loss": -0.10549832582473755,
      "num_tokens": 2277234.0,
      "reward": 0.6919173955917358,
      "reward_std": 0.18916533272713423,
      "rewards/cache_reward/mean": 0.6919173955917358,
      "rewards/cache_reward/std": 0.18916533272713423,
      "step": 140,
      "step_time": 108.70422725854442
    },
    {
      "cache_hit_rate": 0.6837528288364411,
      "cache_hit_rate_size16": 0.743753457069397,
      "cache_hit_rate_size32": 0.8899944126605988,
      "cache_hit_rate_size8": 0.43498979695141315,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.29375,
      "completions/max_length": 997.1,
      "completions/max_terminated_length": 943.5,
      "completions/mean_length": 775.30625,
      "completions/mean_terminated_length": 693.9155212402344,
      "completions/min_length": 411.5,
      "completions/min_terminated_length": 411.5,
      "entropy": 0.5689274027943612,
      "epoch": 0.031302170283806344,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.22718407213687897,
      "kl": 0.06529534254223109,
      "learning_rate": 0.0001,
      "loss": -0.04616514444351196,
      "num_tokens": 2444147.0,
      "reward": 0.6700104534626007,
      "reward_std": 0.17748181223869325,
      "rewards/cache_reward/mean": 0.6700104534626007,
      "rewards/cache_reward/std": 0.17748181223869325,
      "step": 150,
      "step_time": 106.98494991599583
    },
    {
      "cache_hit_rate": 0.6439329028129578,
      "cache_hit_rate_size16": 0.7306425826890128,
      "cache_hit_rate_size32": 0.865531712770462,
      "cache_hit_rate_size8": 0.4455534170071284,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.24375,
      "completions/max_length": 1010.9,
      "completions/max_terminated_length": 928.8,
      "completions/mean_length": 701.30625,
      "completions/mean_terminated_length": 617.9401885986329,
      "completions/min_length": 373.5,
      "completions/min_terminated_length": 373.5,
      "entropy": 0.6062380790710449,
      "epoch": 0.0333889816360601,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.311809241771698,
      "kl": 0.06560975145548582,
      "learning_rate": 0.0001,
      "loss": -0.11374449729919434,
      "num_tokens": 2588116.0,
      "reward": 0.629342782497406,
      "reward_std": 0.09073512144386768,
      "rewards/cache_reward/mean": 0.629342782497406,
      "rewards/cache_reward/std": 0.09073512144386768,
      "step": 160,
      "step_time": 107.92681602598168
    },
    {
      "cache_hit_rate": 0.737691605091095,
      "cache_hit_rate_size16": 0.7616987451910973,
      "cache_hit_rate_size32": 0.8484096378087997,
      "cache_hit_rate_size8": 0.47391005357106525,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.5375,
      "completions/max_length": 1015.0,
      "completions/max_terminated_length": 968.0,
      "completions/mean_length": 876.60625,
      "completions/mean_terminated_length": 778.4466735839844,
      "completions/min_length": 541.3,
      "completions/min_terminated_length": 541.3,
      "entropy": 0.5039393544197083,
      "epoch": 0.03547579298831385,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.22193852066993713,
      "kl": 0.07721416000276804,
      "learning_rate": 0.0001,
      "loss": 0.009571240842342376,
      "num_tokens": 2774853.0,
      "reward": 0.7360832095146179,
      "reward_std": 0.09169260896742344,
      "rewards/cache_reward/mean": 0.7360832095146179,
      "rewards/cache_reward/std": 0.09169260896742344,
      "step": 170,
      "step_time": 108.83163746409119
    },
    {
      "cache_hit_rate": 0.6702417641878128,
      "cache_hit_rate_size16": 0.7594628930091858,
      "cache_hit_rate_size32": 0.8569953143596649,
      "cache_hit_rate_size8": 0.4382890040675799,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.3625,
      "completions/max_length": 1016.3,
      "completions/max_terminated_length": 938.8,
      "completions/mean_length": 702.325,
      "completions/mean_terminated_length": 549.5508026123047,
      "completions/min_length": 199.8,
      "completions/min_terminated_length": 199.8,
      "entropy": 0.6205699190497398,
      "epoch": 0.037562604340567615,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.28704333305358887,
      "kl": 0.08045898843556643,
      "learning_rate": 0.0001,
      "loss": -0.04709939360618591,
      "num_tokens": 2935177.0,
      "reward": 0.6992827504873276,
      "reward_std": 0.131203200109303,
      "rewards/cache_reward/mean": 0.6992827504873276,
      "rewards/cache_reward/std": 0.131203200109303,
      "step": 180,
      "step_time": 109.50919925523922
    },
    {
      "cache_hit_rate": 0.6852899014949798,
      "cache_hit_rate_size16": 0.6899512580462864,
      "cache_hit_rate_size32": 0.8868090361356735,
      "cache_hit_rate_size8": 0.4937669684489568,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.40625,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 867.3,
      "completions/mean_length": 734.975,
      "completions/mean_terminated_length": 540.3156127929688,
      "completions/min_length": 192.5,
      "completions/min_terminated_length": 192.5,
      "entropy": 0.5207937769591808,
      "epoch": 0.03964941569282137,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.2835640013217926,
      "kl": 0.0956730455160141,
      "learning_rate": 0.0001,
      "loss": -0.044486325979232785,
      "num_tokens": 3104349.0,
      "reward": 0.6792330205440521,
      "reward_std": 0.14855806417763234,
      "rewards/cache_reward/mean": 0.6792330205440521,
      "rewards/cache_reward/std": 0.14855806417763234,
      "step": 190,
      "step_time": 110.20991073080805
    },
    {
      "cache_hit_rate": 0.6421143919229507,
      "cache_hit_rate_size16": 0.6826490312814713,
      "cache_hit_rate_size32": 0.8723512291908264,
      "cache_hit_rate_size8": 0.44266503552595776,
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.425,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 871.2,
      "completions/mean_length": 670.93125,
      "completions/mean_terminated_length": 457.2080528259277,
      "completions/min_length": 191.1,
      "completions/min_terminated_length": 191.1,
      "entropy": 0.6987772159278393,
      "epoch": 0.041736227045075125,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 0.28543853759765625,
      "kl": 0.09522749911993741,
      "learning_rate": 0.0001,
      "loss": -0.05127257704734802,
      "num_tokens": 3267042.0,
      "reward": 0.6385201156139374,
      "reward_std": 0.12131488006561994,
      "rewards/cache_reward/mean": 0.6385201156139374,
      "rewards/cache_reward/std": 0.12131488006561994,
      "step": 200,
      "step_time": 110.4871686702827
    }
  ],
  "logging_steps": 10,
  "max_steps": 200,
  "num_input_tokens_seen": 3267042,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}