{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.998109640831758,
  "eval_steps": 51,
  "global_step": 198,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "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": 0.15538194444444445,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1014.3333333333334,
      "completions/mean_length": 317.7908020019531,
      "completions/mean_terminated_length": 188.92760213216147,
      "completions/min_length": 22.666666666666668,
      "completions/min_terminated_length": 22.666666666666668,
      "epoch": 0.045368620037807186,
      "grad_norm": 0.21497704088687897,
      "learning_rate": 4e-07,
      "loss": 0.0113,
      "num_tokens": 948815.0,
      "reward": 0.4324238797028859,
      "reward_std": 0.3621105949083964,
      "rewards/get_embedding_sim/mean": 0.34388219316800434,
      "rewards/get_embedding_sim/std": 0.06503158435225487,
      "rewards/reward_num_unique_chars/mean": 0.02973337098956108,
      "rewards/reward_num_unique_chars/std": 0.16630694766839346,
      "step": 3
    },
    {
      "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": 0.11805555555555554,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1010.0,
      "completions/mean_length": 297.98351033528644,
      "completions/mean_terminated_length": 199.93633015950522,
      "completions/min_length": 11.666666666666666,
      "completions/min_terminated_length": 11.666666666666666,
      "epoch": 0.09073724007561437,
      "grad_norm": 0.12115804105997086,
      "learning_rate": 1e-06,
      "loss": 0.0221,
      "num_tokens": 1879180.0,
      "reward": 0.7094011704126993,
      "reward_std": 0.5900625089804331,
      "rewards/get_embedding_sim/mean": 0.3682552973429362,
      "rewards/get_embedding_sim/std": 0.07593787461519241,
      "rewards/reward_num_unique_chars/mean": 0.11492967108885448,
      "rewards/reward_num_unique_chars/std": 0.3019222468137741,
      "step": 6
    },
    {
      "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": 0.07899305555555554,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1009.6666666666666,
      "completions/mean_length": 240.75695292154947,
      "completions/mean_terminated_length": 174.28186543782553,
      "completions/min_length": 17.0,
      "completions/min_terminated_length": 17.0,
      "epoch": 0.13610586011342155,
      "grad_norm": 0.14687585830688477,
      "learning_rate": 1e-06,
      "loss": 0.0089,
      "num_tokens": 2738532.0,
      "reward": 0.4804896612962087,
      "reward_std": 0.3203038076559703,
      "rewards/get_embedding_sim/mean": 0.32944798469543457,
      "rewards/get_embedding_sim/std": 0.07377548639973004,
      "rewards/reward_num_unique_chars/mean": 0.05086651133994261,
      "rewards/reward_num_unique_chars/std": 0.2023822863896688,
      "step": 9
    },
    {
      "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": 0.09114583333333333,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 938.3333333333334,
      "completions/mean_length": 246.99132792154947,
      "completions/mean_terminated_length": 169.27056884765625,
      "completions/min_length": 13.333333333333334,
      "completions/min_terminated_length": 13.333333333333334,
      "epoch": 0.18147448015122875,
      "grad_norm": 0.14165282249450684,
      "learning_rate": 1e-06,
      "loss": -0.0128,
      "num_tokens": 3612890.0,
      "reward": 0.5771736900011698,
      "reward_std": 0.5188641945521036,
      "rewards/get_embedding_sim/mean": 0.3323819935321808,
      "rewards/get_embedding_sim/std": 0.0759697953859965,
      "rewards/reward_num_unique_chars/mean": 0.0817720244328181,
      "rewards/reward_num_unique_chars/std": 0.2566508601109187,
      "step": 12
    },
    {
      "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": 0.13107638888888892,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1014.3333333333334,
      "completions/mean_length": 309.9053955078125,
      "completions/mean_terminated_length": 201.6662139892578,
      "completions/min_length": 10.666666666666666,
      "completions/min_terminated_length": 10.666666666666666,
      "epoch": 0.22684310018903592,
      "grad_norm": 0.19028598070144653,
      "learning_rate": 1e-06,
      "loss": 0.064,
      "num_tokens": 4532125.0,
      "reward": 0.6042946378389994,
      "reward_std": 0.5824080804983774,
      "rewards/get_embedding_sim/mean": 0.351690411567688,
      "rewards/get_embedding_sim/std": 0.07434952010711034,
      "rewards/reward_num_unique_chars/mean": 0.08517500261465709,
      "rewards/reward_num_unique_chars/std": 0.2722396006186803,
      "step": 15
    },
    {
      "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": 0.11545138888888891,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 953.6666666666666,
      "completions/mean_length": 282.27431233723956,
      "completions/mean_terminated_length": 185.10088094075522,
      "completions/min_length": 23.333333333333332,
      "completions/min_terminated_length": 23.333333333333332,
      "epoch": 0.2722117202268431,
      "grad_norm": 0.16469748318195343,
      "learning_rate": 1e-06,
      "loss": 0.0329,
      "num_tokens": 5440073.0,
      "reward": 0.6671707034111023,
      "reward_std": 0.6297763784726461,
      "rewards/get_embedding_sim/mean": 0.3546706239382426,
      "rewards/get_embedding_sim/std": 0.08629031976064046,
      "rewards/reward_num_unique_chars/mean": 0.10469372570514679,
      "rewards/reward_num_unique_chars/std": 0.30532515048980713,
      "step": 18
    },
    {
      "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": 0.09461805555555554,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 971.0,
      "completions/mean_length": 239.50695292154947,
      "completions/mean_terminated_length": 157.83088175455728,
      "completions/min_length": 13.666666666666666,
      "completions/min_terminated_length": 13.666666666666666,
      "epoch": 0.31758034026465026,
      "grad_norm": 0.3001786768436432,
      "learning_rate": 1e-06,
      "loss": 0.0196,
      "num_tokens": 6298945.0,
      "reward": 0.6787747740745544,
      "reward_std": 0.595538596312205,
      "rewards/get_embedding_sim/mean": 0.34804557760556537,
      "rewards/get_embedding_sim/std": 0.06324811031421025,
      "rewards/reward_num_unique_chars/mean": 0.11097987244526546,
      "rewards/reward_num_unique_chars/std": 0.3113570312658946,
      "step": 21
    },
    {
      "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": 0.09722222222222221,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 937.3333333333334,
      "completions/mean_length": 232.53819783528647,
      "completions/mean_terminated_length": 147.56163533528647,
      "completions/min_length": 14.666666666666666,
      "completions/min_terminated_length": 14.666666666666666,
      "epoch": 0.3629489603024575,
      "grad_norm": 0.14932705461978912,
      "learning_rate": 1e-06,
      "loss": 0.0234,
      "num_tokens": 7151613.0,
      "reward": 0.7232053279876709,
      "reward_std": 0.5765545765558878,
      "rewards/get_embedding_sim/mean": 0.3794552783171336,
      "rewards/get_embedding_sim/std": 0.07909785211086273,
      "rewards/reward_num_unique_chars/mean": 0.11477824797232945,
      "rewards/reward_num_unique_chars/std": 0.31598811348279315,
      "step": 24
    },
    {
      "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": 0.03819444444444442,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 935.0,
      "completions/mean_length": 177.9071248372396,
      "completions/mean_terminated_length": 144.04273478190103,
      "completions/min_length": 13.0,
      "completions/min_terminated_length": 13.0,
      "epoch": 0.40831758034026466,
      "grad_norm": 0.10965846478939056,
      "learning_rate": 1e-06,
      "loss": 0.0195,
      "num_tokens": 7940146.0,
      "reward": 0.9970683455467224,
      "reward_std": 0.7138840556144714,
      "rewards/get_embedding_sim/mean": 0.36425574620564777,
      "rewards/get_embedding_sim/std": 0.0799053559700648,
      "rewards/reward_num_unique_chars/mean": 0.21126614014307657,
      "rewards/reward_num_unique_chars/std": 0.406161626180013,
      "step": 27
    },
    {
      "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": 0.03125,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 919.3333333333334,
      "completions/mean_length": 189.5008748372396,
      "completions/mean_terminated_length": 162.68724060058594,
      "completions/min_length": 15.333333333333334,
      "completions/min_terminated_length": 15.333333333333334,
      "epoch": 0.45368620037807184,
      "grad_norm": 0.14693303406238556,
      "learning_rate": 1e-06,
      "loss": 0.0088,
      "num_tokens": 8735891.0,
      "reward": 0.6218339800834656,
      "reward_std": 0.5519726971785227,
      "rewards/get_embedding_sim/mean": 0.3483964701493581,
      "rewards/get_embedding_sim/std": 0.08077059934536616,
      "rewards/reward_num_unique_chars/mean": 0.09167342260479927,
      "rewards/reward_num_unique_chars/std": 0.26420795420805615,
      "step": 30
    },
    {
      "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": 0.036458333333333336,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 894.3333333333334,
      "completions/mean_length": 170.3498331705729,
      "completions/mean_terminated_length": 138.13396199544272,
      "completions/min_length": 13.0,
      "completions/min_terminated_length": 13.0,
      "epoch": 0.499054820415879,
      "grad_norm": 0.08968470990657806,
      "learning_rate": 1e-06,
      "loss": 0.0073,
      "num_tokens": 9511590.0,
      "reward": 0.6786622206370035,
      "reward_std": 0.6337552865346273,
      "rewards/get_embedding_sim/mean": 0.36355799436569214,
      "rewards/get_embedding_sim/std": 0.07963384687900543,
      "rewards/reward_num_unique_chars/mean": 0.10508231942852338,
      "rewards/reward_num_unique_chars/std": 0.30025769273440045,
      "step": 33
    },
    {
      "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": 0.029513888888888912,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 912.6666666666666,
      "completions/mean_length": 173.1371612548828,
      "completions/mean_terminated_length": 147.0882339477539,
      "completions/min_length": 8.333333333333334,
      "completions/min_terminated_length": 8.333333333333334,
      "epoch": 0.5444234404536862,
      "grad_norm": 0.13164524734020233,
      "learning_rate": 1e-06,
      "loss": 0.0119,
      "num_tokens": 10291652.0,
      "reward": 0.5599844257036845,
      "reward_std": 0.570480744043986,
      "rewards/get_embedding_sim/mean": 0.36988019943237305,
      "rewards/get_embedding_sim/std": 0.08963250865538915,
      "rewards/reward_num_unique_chars/mean": 0.06345440695683162,
      "rewards/reward_num_unique_chars/std": 0.24292809267838797,
      "step": 36
    },
    {
      "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": 0.035590277777777755,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 881.6666666666666,
      "completions/mean_length": 168.06597900390625,
      "completions/mean_terminated_length": 136.42411295572916,
      "completions/min_length": 12.333333333333334,
      "completions/min_terminated_length": 12.333333333333334,
      "epoch": 0.5897920604914934,
      "grad_norm": 0.11388170719146729,
      "learning_rate": 1e-06,
      "loss": 0.0148,
      "num_tokens": 11075088.0,
      "reward": 1.0007078846295674,
      "reward_std": 0.7608418663342794,
      "rewards/get_embedding_sim/mean": 0.37049947182337445,
      "rewards/get_embedding_sim/std": 0.09431915730237961,
      "rewards/reward_num_unique_chars/mean": 0.21018956104914346,
      "rewards/reward_num_unique_chars/std": 0.39977312088012695,
      "step": 39
    },
    {
      "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": 0.017361111111111088,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 931.0,
      "completions/mean_length": 183.0703125,
      "completions/mean_terminated_length": 168.1489054361979,
      "completions/min_length": 14.0,
      "completions/min_terminated_length": 14.0,
      "epoch": 0.6351606805293005,
      "grad_norm": 0.11133608222007751,
      "learning_rate": 1e-06,
      "loss": 0.0043,
      "num_tokens": 11875809.0,
      "reward": 0.8215018113454183,
      "reward_std": 0.6714911858240763,
      "rewards/get_embedding_sim/mean": 0.3735850652058919,
      "rewards/get_embedding_sim/std": 0.100465623040994,
      "rewards/reward_num_unique_chars/mean": 0.1498913218577703,
      "rewards/reward_num_unique_chars/std": 0.34748714168866474,
      "step": 42
    },
    {
      "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": 0.024305555555555542,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 944.0,
      "completions/mean_length": 180.48697916666666,
      "completions/mean_terminated_length": 159.57064310709634,
      "completions/min_length": 20.0,
      "completions/min_terminated_length": 20.0,
      "epoch": 0.6805293005671077,
      "grad_norm": 0.1150749996304512,
      "learning_rate": 1e-06,
      "loss": 0.0028,
      "num_tokens": 12659586.0,
      "reward": 0.7888058622678121,
      "reward_std": 0.591193675994873,
      "rewards/get_embedding_sim/mean": 0.39557669560114544,
      "rewards/get_embedding_sim/std": 0.10181870808204015,
      "rewards/reward_num_unique_chars/mean": 0.13228891293207803,
      "rewards/reward_num_unique_chars/std": 0.3239031334718068,
      "step": 45
    },
    {
      "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": 0.014756944444444456,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 930.6666666666666,
      "completions/mean_length": 151.30382792154947,
      "completions/mean_terminated_length": 138.23454793294272,
      "completions/min_length": 10.666666666666666,
      "completions/min_terminated_length": 10.666666666666666,
      "epoch": 0.725897920604915,
      "grad_norm": 0.1305595338344574,
      "learning_rate": 1e-06,
      "loss": 0.016,
      "num_tokens": 13415168.0,
      "reward": 0.8962518771489462,
      "reward_std": 0.6784158746401469,
      "rewards/get_embedding_sim/mean": 0.41448097427686054,
      "rewards/get_embedding_sim/std": 0.10920101900895436,
      "rewards/reward_num_unique_chars/mean": 0.16110383967558542,
      "rewards/reward_num_unique_chars/std": 0.3513130197922389,
      "step": 48
    },
    {
      "epoch": 0.7712665406427222,
      "grad_norm": 0.3789539039134979,
      "learning_rate": 1e-06,
      "loss": 0.0057,
      "step": 51
    },
    {
      "epoch": 0.7712665406427222,
      "eval_clip_ratio/high_max": 0.0,
      "eval_clip_ratio/high_mean": 0.0,
      "eval_clip_ratio/low_mean": 0.0,
      "eval_clip_ratio/low_min": 0.0,
      "eval_clip_ratio/region_mean": 0.0,
      "eval_completions/clipped_ratio": 0.07700892857142858,
      "eval_completions/max_length": 839.75,
      "eval_completions/max_terminated_length": 648.2678571428571,
      "eval_completions/mean_length": 198.52493231637138,
      "eval_completions/mean_terminated_length": 132.5261221272605,
      "eval_completions/min_length": 23.035714285714285,
      "eval_completions/min_terminated_length": 23.035714285714285,
      "eval_loss": 0.03166242688894272,
      "eval_num_tokens": 14154815.0,
      "eval_reward": 0.7619188348097461,
      "eval_reward_std": 0.6427728646833982,
      "eval_rewards/get_embedding_sim/mean": 0.44160632097295355,
      "eval_rewards/get_embedding_sim/std": 0.09290120118696775,
      "eval_rewards/reward_num_unique_chars/mean": 0.10697663322623287,
      "eval_rewards/reward_num_unique_chars/std": 0.20512951059000833,
      "eval_runtime": 6200.6791,
      "eval_samples_per_second": 0.009,
      "eval_steps_per_second": 0.0,
      "step": 51
    },
    {
      "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": 0.024305555555555542,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 913.0,
      "completions/mean_length": 172.71180979410806,
      "completions/mean_terminated_length": 151.74237823486328,
      "completions/min_length": 11.5,
      "completions/min_terminated_length": 11.5,
      "epoch": 0.8166351606805293,
      "grad_norm": 0.34116339683532715,
      "learning_rate": 1e-06,
      "loss": 0.0168,
      "num_tokens": 14970664.0,
      "reward": 0.804489662249883,
      "reward_std": 0.6523752709229788,
      "rewards/get_embedding_sim/mean": 0.438604195912679,
      "rewards/get_embedding_sim/std": 0.11339207986990611,
      "rewards/reward_num_unique_chars/mean": 0.12225343783696492,
      "rewards/reward_num_unique_chars/std": 0.3136557762821515,
      "step": 54
    },
    {
      "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": 0.01475694444444442,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 796.3333333333334,
      "completions/mean_length": 170.29340616861978,
      "completions/mean_terminated_length": 157.51276143391928,
      "completions/min_length": 11.0,
      "completions/min_terminated_length": 11.0,
      "epoch": 0.8620037807183365,
      "grad_norm": 0.19843681156635284,
      "learning_rate": 1e-06,
      "loss": 0.0085,
      "num_tokens": 15756666.0,
      "reward": 0.8036239345868429,
      "reward_std": 0.7330554525057474,
      "rewards/get_embedding_sim/mean": 0.45206138491630554,
      "rewards/get_embedding_sim/std": 0.11286439994970958,
      "rewards/reward_num_unique_chars/mean": 0.11766015738248825,
      "rewards/reward_num_unique_chars/std": 0.31892595688501996,
      "step": 57
    },
    {
      "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": 0.024305555555555542,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 969.3333333333334,
      "completions/mean_length": 174.41407267252603,
      "completions/mean_terminated_length": 153.36661275227866,
      "completions/min_length": 9.666666666666666,
      "completions/min_terminated_length": 9.666666666666666,
      "epoch": 0.9073724007561437,
      "grad_norm": 0.12094840407371521,
      "learning_rate": 1e-06,
      "loss": 0.0103,
      "num_tokens": 16540071.0,
      "reward": 0.9701792995134989,
      "reward_std": 0.7476302186648051,
      "rewards/get_embedding_sim/mean": 0.4727833966414134,
      "rewards/get_embedding_sim/std": 0.1110062524676323,
      "rewards/reward_num_unique_chars/mean": 0.17127168675263724,
      "rewards/reward_num_unique_chars/std": 0.3443687508503596,
      "step": 60
    },
    {
      "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": 0.0234375,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 890.0,
      "completions/mean_length": 189.61371866861978,
      "completions/mean_terminated_length": 169.4032440185547,
      "completions/min_length": 8.333333333333334,
      "completions/min_terminated_length": 8.333333333333334,
      "epoch": 0.9527410207939508,
      "grad_norm": 0.13162142038345337,
      "learning_rate": 1e-06,
      "loss": 0.014,
      "num_tokens": 17348330.0,
      "reward": 0.8747870127360026,
      "reward_std": 0.7496191461881002,
      "rewards/get_embedding_sim/mean": 0.49978697299957275,
      "rewards/get_embedding_sim/std": 0.1058602233727773,
      "rewards/reward_num_unique_chars/mean": 0.125,
      "rewards/reward_num_unique_chars/std": 0.32470036546389264,
      "step": 63
    },
    {
      "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": 0.03197437739463601,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 849.3333333333334,
      "completions/mean_length": 193.41219075520834,
      "completions/mean_terminated_length": 165.51568094889322,
      "completions/min_length": 8.333333333333334,
      "completions/min_terminated_length": 8.333333333333334,
      "epoch": 0.998109640831758,
      "grad_norm": 0.12491139769554138,
      "learning_rate": 1e-06,
      "loss": 0.0221,
      "num_tokens": 18146229.0,
      "reward": 1.0248117248217266,
      "reward_std": 0.8998820185661316,
      "rewards/get_embedding_sim/mean": 0.5143949886163076,
      "rewards/get_embedding_sim/std": 0.10505472123622894,
      "rewards/reward_num_unique_chars/mean": 0.17013889054457346,
      "rewards/reward_num_unique_chars/std": 0.3742268184820811,
      "step": 66
    },
    {
      "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": 0.0546875,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 946.0,
      "completions/mean_length": 234.43750508626303,
      "completions/mean_terminated_length": 188.7455037434896,
      "completions/min_length": 7.0,
      "completions/min_terminated_length": 7.0,
      "epoch": 1.0453686200378072,
      "grad_norm": 0.2972349524497986,
      "learning_rate": 1e-06,
      "loss": 0.0338,
      "num_tokens": 19006125.0,
      "reward": 0.7983500361442566,
      "reward_std": 0.811311940352122,
      "rewards/get_embedding_sim/mean": 0.5249124765396118,
      "rewards/get_embedding_sim/std": 0.08944002538919449,
      "rewards/reward_num_unique_chars/mean": 0.09124782433112462,
      "rewards/reward_num_unique_chars/std": 0.28635749220848083,
      "step": 69
    },
    {
      "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": 0.06510416666666667,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 976.6666666666666,
      "completions/mean_length": 244.2170155843099,
      "completions/mean_terminated_length": 189.76802571614584,
      "completions/min_length": 5.333333333333333,
      "completions/min_terminated_length": 5.333333333333333,
      "epoch": 1.0907372400756143,
      "grad_norm": 1.1331943273544312,
      "learning_rate": 1e-06,
      "loss": 0.039,
      "num_tokens": 19862839.0,
      "reward": 0.8927220304807028,
      "reward_std": 0.7463525732358297,
      "rewards/get_embedding_sim/mean": 0.5437636772791544,
      "rewards/get_embedding_sim/std": 0.09080894043048222,
      "rewards/reward_num_unique_chars/mean": 0.1166065161426862,
      "rewards/reward_num_unique_chars/std": 0.3212928573290507,
      "step": 72
    },
    {
      "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": 0.09895833333333337,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 960.6666666666666,
      "completions/mean_length": 278.218755086263,
      "completions/mean_terminated_length": 196.66878763834634,
      "completions/min_length": 6.333333333333333,
      "completions/min_terminated_length": 6.333333333333333,
      "epoch": 1.1361058601134215,
      "grad_norm": 1.4549461603164673,
      "learning_rate": 1e-06,
      "loss": 0.0488,
      "num_tokens": 20767987.0,
      "reward": 0.840681274731954,
      "reward_std": 0.767987052599589,
      "rewards/get_embedding_sim/mean": 0.551618734995524,
      "rewards/get_embedding_sim/std": 0.08721146732568741,
      "rewards/reward_num_unique_chars/mean": 0.09680009384950002,
      "rewards/reward_num_unique_chars/std": 0.29205257693926495,
      "step": 75
    },
    {
      "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": 0.1354166666666667,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 975.6666666666666,
      "completions/mean_length": 325.6267395019531,
      "completions/mean_terminated_length": 216.13731384277344,
      "completions/min_length": 9.0,
      "completions/min_terminated_length": 9.0,
      "epoch": 1.1814744801512287,
      "grad_norm": 1.139538049697876,
      "learning_rate": 1e-06,
      "loss": 0.0302,
      "num_tokens": 21725925.0,
      "reward": 0.7460880279541016,
      "reward_std": 0.4740897516409556,
      "rewards/get_embedding_sim/mean": 0.5429630080858866,
      "rewards/get_embedding_sim/std": 0.07056275258461635,
      "rewards/reward_num_unique_chars/mean": 0.06794072687625885,
      "rewards/reward_num_unique_chars/std": 0.24171357850233713,
      "step": 78
    },
    {
      "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": 0.36111111111111116,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1004.6666666666666,
      "completions/mean_length": 532.0043640136719,
      "completions/mean_terminated_length": 255.43873596191406,
      "completions/min_length": 4.0,
      "completions/min_terminated_length": 4.0,
      "epoch": 1.2268431001890359,
      "grad_norm": 33.07490158081055,
      "learning_rate": 1e-06,
      "loss": 0.0512,
      "num_tokens": 22928618.0,
      "reward": 0.5259575247764587,
      "reward_std": 0.13220050682624182,
      "rewards/get_embedding_sim/mean": 0.5077283481756846,
      "rewards/get_embedding_sim/std": 0.06187118341525396,
      "rewards/reward_num_unique_chars/mean": 0.006076388992369175,
      "rewards/reward_num_unique_chars/std": 0.06327294061581294,
      "step": 81
    },
    {
      "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": 0.3463541666666667,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 996.0,
      "completions/mean_length": 495.05469767252606,
      "completions/mean_terminated_length": 215.88401794433594,
      "completions/min_length": 5.333333333333333,
      "completions/min_terminated_length": 5.333333333333333,
      "epoch": 1.272211720226843,
      "grad_norm": 24.829124450683594,
      "learning_rate": 1e-06,
      "loss": 0.0565,
      "num_tokens": 24088745.0,
      "reward": 0.523603618144989,
      "reward_std": 0.051181082924207054,
      "rewards/get_embedding_sim/mean": 0.5236035784085592,
      "rewards/get_embedding_sim/std": 0.06259416912992795,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 84
    },
    {
      "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": 0.2595486111111111,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 993.3333333333334,
      "completions/mean_length": 425.3038330078125,
      "completions/mean_terminated_length": 215.5502471923828,
      "completions/min_length": 4.0,
      "completions/min_terminated_length": 4.0,
      "epoch": 1.3175803402646502,
      "grad_norm": 3.4278340339660645,
      "learning_rate": 1e-06,
      "loss": 0.038,
      "num_tokens": 25168519.0,
      "reward": 0.6717002590497335,
      "reward_std": 0.36602258682250977,
      "rewards/get_embedding_sim/mean": 0.5493044058481852,
      "rewards/get_embedding_sim/std": 0.06710439547896385,
      "rewards/reward_num_unique_chars/mean": 0.040798611318071686,
      "rewards/reward_num_unique_chars/std": 0.19389309485753378,
      "step": 87
    },
    {
      "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": 0.24305555555555555,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 996.6666666666666,
      "completions/mean_length": 408.6744791666667,
      "completions/mean_terminated_length": 212.26323445638022,
      "completions/min_length": 7.0,
      "completions/min_terminated_length": 7.0,
      "epoch": 1.3629489603024574,
      "grad_norm": 3.8435463905334473,
      "learning_rate": 1e-06,
      "loss": 0.0128,
      "num_tokens": 26226592.0,
      "reward": 0.6131470998128256,
      "reward_std": 0.31071917215983075,
      "rewards/get_embedding_sim/mean": 0.5376262267430624,
      "rewards/get_embedding_sim/std": 0.05832842489083608,
      "rewards/reward_num_unique_chars/mean": 0.025184943340718746,
      "rewards/reward_num_unique_chars/std": 0.15304673463106155,
      "step": 90
    },
    {
      "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": 0.22222222222222224,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1008.0,
      "completions/mean_length": 373.02952067057294,
      "completions/mean_terminated_length": 186.92364501953125,
      "completions/min_length": 4.333333333333333,
      "completions/min_terminated_length": 4.333333333333333,
      "epoch": 1.4083175803402646,
      "grad_norm": 12.08054256439209,
      "learning_rate": 1e-06,
      "loss": 0.0535,
      "num_tokens": 27238274.0,
      "reward": 0.8090014656384786,
      "reward_std": 0.46231311559677124,
      "rewards/get_embedding_sim/mean": 0.5459806124369303,
      "rewards/get_embedding_sim/std": 0.07225656509399414,
      "rewards/reward_num_unique_chars/mean": 0.08768494582424562,
      "rewards/reward_num_unique_chars/std": 0.25595557938019436,
      "step": 93
    },
    {
      "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": 0.23958333333333334,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1010.0,
      "completions/mean_length": 384.4331766764323,
      "completions/mean_terminated_length": 183.23039754231772,
      "completions/min_length": 5.666666666666667,
      "completions/min_terminated_length": 5.666666666666667,
      "epoch": 1.4536862003780717,
      "grad_norm": 16.734275817871094,
      "learning_rate": 1e-06,
      "loss": 0.0641,
      "num_tokens": 28261845.0,
      "reward": 0.7301652232805887,
      "reward_std": 0.4747513731320699,
      "rewards/get_embedding_sim/mean": 0.5478734771410624,
      "rewards/get_embedding_sim/std": 0.06370775277415912,
      "rewards/reward_num_unique_chars/mean": 0.060809217393398285,
      "rewards/reward_num_unique_chars/std": 0.23756111164887747,
      "step": 96
    },
    {
      "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": 0.296875,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 997.6666666666666,
      "completions/mean_length": 471.3073018391927,
      "completions/mean_terminated_length": 240.4876708984375,
      "completions/min_length": 5.0,
      "completions/min_terminated_length": 5.0,
      "epoch": 1.499054820415879,
      "grad_norm": 78.62782287597656,
      "learning_rate": 1e-06,
      "loss": 0.0802,
      "num_tokens": 29394615.0,
      "reward": 0.5468668937683105,
      "reward_std": 0.1741823156674703,
      "rewards/get_embedding_sim/mean": 0.5156168540318807,
      "rewards/get_embedding_sim/std": 0.06206041698654493,
      "rewards/reward_num_unique_chars/mean": 0.010416666666666666,
      "rewards/reward_num_unique_chars/std": 0.10111008336146672,
      "step": 99
    },
    {
      "epoch": 1.544423440453686,
      "grad_norm": 38.67718505859375,
      "learning_rate": 1e-06,
      "loss": 0.1045,
      "step": 102
    },
    {
      "epoch": 1.544423440453686,
      "eval_clip_ratio/high_max": 0.0,
      "eval_clip_ratio/high_mean": 0.0,
      "eval_clip_ratio/low_mean": 0.0,
      "eval_clip_ratio/low_min": 0.0,
      "eval_clip_ratio/region_mean": 0.0,
      "eval_completions/clipped_ratio": 0.6886160714285714,
      "eval_completions/max_length": 1024.0,
      "eval_completions/max_terminated_length": 611.75,
      "eval_completions/mean_length": 827.2779154096331,
      "eval_completions/mean_terminated_length": 360.4234839166914,
      "eval_completions/min_length": 518.75,
      "eval_completions/min_terminated_length": 189.60714285714286,
      "eval_loss": 0.04136611148715019,
      "eval_num_tokens": 30813833.0,
      "eval_reward": 0.5307553862886769,
      "eval_reward_std": 0.07386554186398696,
      "eval_rewards/get_embedding_sim/mean": 0.5128982285303729,
      "eval_rewards/get_embedding_sim/std": 0.022365678313820223,
      "eval_rewards/reward_num_unique_chars/mean": 0.005952380996729646,
      "eval_rewards/reward_num_unique_chars/std": 0.01812402212194034,
      "eval_runtime": 7571.0613,
      "eval_samples_per_second": 0.007,
      "eval_steps_per_second": 0.0,
      "step": 102
    },
    {
      "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": 0.6297743055555555,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 991.5,
      "completions/mean_length": 752.4883015950521,
      "completions/mean_terminated_length": 286.00857798258465,
      "completions/min_length": 7.166666666666667,
      "completions/min_terminated_length": 7.166666666666667,
      "epoch": 1.5897920604914932,
      "grad_norm": 48.53269958496094,
      "learning_rate": 1e-06,
      "loss": 0.1121,
      "num_tokens": 32283132.0,
      "reward": 0.5582722375790278,
      "reward_std": 0.1687685896952947,
      "rewards/get_embedding_sim/mean": 0.5009805411100388,
      "rewards/get_embedding_sim/std": 0.06131617228190104,
      "rewards/reward_num_unique_chars/mean": 0.019097222480922937,
      "rewards/reward_num_unique_chars/std": 0.1062369520465533,
      "step": 105
    },
    {
      "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": 0.4835069444444445,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1001.3333333333334,
      "completions/mean_length": 599.3819580078125,
      "completions/mean_terminated_length": 200.02798970540366,
      "completions/min_length": 11.0,
      "completions/min_terminated_length": 11.0,
      "epoch": 1.6351606805293004,
      "grad_norm": 79.72682189941406,
      "learning_rate": 1e-06,
      "loss": 0.0964,
      "num_tokens": 33563444.0,
      "reward": 0.6361258228619894,
      "reward_std": 0.2207108438014984,
      "rewards/get_embedding_sim/mean": 0.5345633029937744,
      "rewards/get_embedding_sim/std": 0.06109160929918289,
      "rewards/reward_num_unique_chars/mean": 0.033854166666666664,
      "rewards/reward_num_unique_chars/std": 0.17674431701501211,
      "step": 108
    },
    {
      "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": 0.4791666666666667,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 983.3333333333334,
      "completions/mean_length": 582.7769368489584,
      "completions/mean_terminated_length": 177.35736083984375,
      "completions/min_length": 6.666666666666667,
      "completions/min_terminated_length": 6.666666666666667,
      "epoch": 1.6805293005671076,
      "grad_norm": 66.84724426269531,
      "learning_rate": 1e-06,
      "loss": 0.1361,
      "num_tokens": 34817523.0,
      "reward": 0.6592132250467936,
      "reward_std": 0.30980920791625977,
      "rewards/get_embedding_sim/mean": 0.5211923718452454,
      "rewards/get_embedding_sim/std": 0.07001716395219167,
      "rewards/reward_num_unique_chars/mean": 0.04600694387530287,
      "rewards/reward_num_unique_chars/std": 0.1816587597131729,
      "step": 111
    },
    {
      "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": 0.5286458333333334,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 959.3333333333334,
      "completions/mean_length": 604.1206868489584,
      "completions/mean_terminated_length": 133.16328938802084,
      "completions/min_length": 10.0,
      "completions/min_terminated_length": 10.0,
      "epoch": 1.725897920604915,
      "grad_norm": 89.8062744140625,
      "learning_rate": 1e-06,
      "loss": 0.1364,
      "num_tokens": 36095950.0,
      "reward": 0.8402608036994934,
      "reward_std": 0.35089851419130963,
      "rewards/get_embedding_sim/mean": 0.5277607639630636,
      "rewards/get_embedding_sim/std": 0.07844971617062886,
      "rewards/reward_num_unique_chars/mean": 0.10852466026941936,
      "rewards/reward_num_unique_chars/std": 0.3103277087211609,
      "step": 114
    },
    {
      "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": 0.6354166666666666,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 921.3333333333334,
      "completions/mean_length": 707.2665201822916,
      "completions/mean_terminated_length": 148.84990437825522,
      "completions/min_length": 10.0,
      "completions/min_terminated_length": 10.0,
      "epoch": 1.7712665406427222,
      "grad_norm": 40.22359848022461,
      "learning_rate": 1e-06,
      "loss": 0.1332,
      "num_tokens": 37474913.0,
      "reward": 0.6136655608812968,
      "reward_std": 0.293540820479393,
      "rewards/get_embedding_sim/mean": 0.5094988644123077,
      "rewards/get_embedding_sim/std": 0.08142028003931046,
      "rewards/reward_num_unique_chars/mean": 0.034722223257025085,
      "rewards/reward_num_unique_chars/std": 0.17292212943236032,
      "step": 117
    },
    {
      "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": 0.732638888888889,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 996.0,
      "completions/mean_length": 784.4765828450521,
      "completions/mean_terminated_length": 130.17157745361328,
      "completions/min_length": 13.333333333333334,
      "completions/min_terminated_length": 13.333333333333334,
      "epoch": 1.8166351606805293,
      "grad_norm": 89.25932312011719,
      "learning_rate": 1e-06,
      "loss": 0.137,
      "num_tokens": 38955302.0,
      "reward": 0.5185406506061554,
      "reward_std": 0.09952469418446223,
      "rewards/get_embedding_sim/mean": 0.4951031009356181,
      "rewards/get_embedding_sim/std": 0.0700259196261565,
      "rewards/reward_num_unique_chars/mean": 0.0078125,
      "rewards/reward_num_unique_chars/std": 0.05049526194731394,
      "step": 120
    },
    {
      "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": 0.8177083333333334,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 984.6666666666666,
      "completions/mean_length": 867.1640828450521,
      "completions/mean_terminated_length": 161.21670532226562,
      "completions/min_length": 16.333333333333332,
      "completions/min_terminated_length": 16.333333333333332,
      "epoch": 1.8620037807183365,
      "grad_norm": 82.80584716796875,
      "learning_rate": 1e-06,
      "loss": 0.0882,
      "num_tokens": 40533683.0,
      "reward": 0.5915247996648153,
      "reward_std": 0.15163771559794745,
      "rewards/get_embedding_sim/mean": 0.5212122996648153,
      "rewards/get_embedding_sim/std": 0.06101148078838984,
      "rewards/reward_num_unique_chars/mean": 0.023437500620881718,
      "rewards/reward_num_unique_chars/std": 0.12005388736724854,
      "step": 123
    },
    {
      "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": 0.8498263888888888,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 586.3333333333334,
      "completions/mean_length": 880.1041870117188,
      "completions/mean_terminated_length": 75.05465443929036,
      "completions/min_length": 24.666666666666668,
      "completions/min_terminated_length": 24.666666666666668,
      "epoch": 1.9073724007561437,
      "grad_norm": 91.38998413085938,
      "learning_rate": 1e-06,
      "loss": 0.0944,
      "num_tokens": 42126011.0,
      "reward": 0.5572214623292288,
      "reward_std": 0.09933711464206378,
      "rewards/get_embedding_sim/mean": 0.510346402724584,
      "rewards/get_embedding_sim/std": 0.05711419011155764,
      "rewards/reward_num_unique_chars/mean": 0.015625,
      "rewards/reward_num_unique_chars/std": 0.07054894665877025,
      "step": 126
    },
    {
      "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": 0.8116319444444443,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 118.66666666666667,
      "completions/mean_length": 837.5321451822916,
      "completions/mean_terminated_length": 37.19810231526693,
      "completions/min_length": 12.666666666666666,
      "completions/min_terminated_length": 12.666666666666666,
      "epoch": 1.9527410207939508,
      "grad_norm": 119.43608856201172,
      "learning_rate": 1e-06,
      "loss": 0.088,
      "num_tokens": 43665792.0,
      "reward": 0.6588172515233358,
      "reward_std": 0.08685253312190373,
      "rewards/get_embedding_sim/mean": 0.5312130351861318,
      "rewards/get_embedding_sim/std": 0.07142255455255508,
      "rewards/reward_num_unique_chars/mean": 0.04253472310180465,
      "rewards/reward_num_unique_chars/std": 0.1334148347377777,
      "step": 129
    },
    {
      "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": 0.9213122605363985,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 109.0,
      "completions/mean_length": 945.9847208658854,
      "completions/mean_terminated_length": 33.62779172261556,
      "completions/min_length": 23.333333333333332,
      "completions/min_terminated_length": 23.333333333333332,
      "epoch": 1.998109640831758,
      "grad_norm": 44.33031463623047,
      "learning_rate": 1e-06,
      "loss": 0.0602,
      "num_tokens": 45313394.0,
      "reward": 0.6258882085482279,
      "reward_std": 0.11284429579973221,
      "rewards/get_embedding_sim/mean": 0.5034923354784647,
      "rewards/get_embedding_sim/std": 0.06357006107767423,
      "rewards/reward_num_unique_chars/mean": 0.0407986119389534,
      "rewards/reward_num_unique_chars/std": 0.10939000050226848,
      "step": 132
    },
    {
      "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": 0.8880208333333334,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 71.0,
      "completions/mean_length": 852.623291015625,
      "completions/mean_terminated_length": 35.400139490763344,
      "completions/min_length": 27.333333333333332,
      "completions/min_terminated_length": 27.333333333333332,
      "epoch": 2.045368620037807,
      "grad_norm": 75.32003784179688,
      "learning_rate": 1e-06,
      "loss": 0.0494,
      "num_tokens": 46871568.0,
      "reward": 0.7583642403284708,
      "reward_std": 0.15150074660778046,
      "rewards/get_embedding_sim/mean": 0.5291975339253744,
      "rewards/get_embedding_sim/std": 0.06528133650620778,
      "rewards/reward_num_unique_chars/mean": 0.07638889054457347,
      "rewards/reward_num_unique_chars/std": 0.21262266238530478,
      "step": 135
    },
    {
      "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": 0.9305555555555555,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 581.0,
      "completions/mean_length": 960.4965413411459,
      "completions/mean_terminated_length": 102.24935150146484,
      "completions/min_length": 21.666666666666668,
      "completions/min_terminated_length": 21.666666666666668,
      "epoch": 2.0907372400756143,
      "grad_norm": 118.80703735351562,
      "learning_rate": 1e-06,
      "loss": 0.0293,
      "num_tokens": 48565148.0,
      "reward": 0.5344894230365753,
      "reward_std": 0.1357189305126667,
      "rewards/get_embedding_sim/mean": 0.48501022656758624,
      "rewards/get_embedding_sim/std": 0.056427846352259316,
      "rewards/reward_num_unique_chars/mean": 0.0164930559694767,
      "rewards/reward_num_unique_chars/std": 0.07238306601842244,
      "step": 138
    },
    {
      "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": 0.9331597222222223,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 697.6666666666666,
      "completions/mean_length": 968.9010620117188,
      "completions/mean_terminated_length": 194.02562967936197,
      "completions/min_length": 37.0,
      "completions/min_terminated_length": 37.0,
      "epoch": 2.1361058601134215,
      "grad_norm": 75.15093231201172,
      "learning_rate": 1e-06,
      "loss": 0.0129,
      "num_tokens": 50255834.0,
      "reward": 0.47950167457262677,
      "reward_std": 0.07561548550923665,
      "rewards/get_embedding_sim/mean": 0.4638766249020894,
      "rewards/get_embedding_sim/std": 0.055329278111457825,
      "rewards/reward_num_unique_chars/mean": 0.005208333333333333,
      "rewards/reward_num_unique_chars/std": 0.041393796602884926,
      "step": 141
    },
    {
      "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": 0.96875,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 621.3333333333334,
      "completions/mean_length": 1001.4574788411459,
      "completions/mean_terminated_length": 315.28333536783856,
      "completions/min_length": 129.0,
      "completions/min_terminated_length": 129.0,
      "epoch": 2.1814744801512287,
      "grad_norm": 29.92278480529785,
      "learning_rate": 1e-06,
      "loss": 0.0005,
      "num_tokens": 51989497.0,
      "reward": 0.4565551280975342,
      "reward_std": 0.02663045997420947,
      "rewards/get_embedding_sim/mean": 0.4565551280975342,
      "rewards/get_embedding_sim/std": 0.03693879644076029,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 144
    },
    {
      "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": 0.9496527777777778,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 792.0,
      "completions/mean_length": 981.279541015625,
      "completions/mean_terminated_length": 333.9038492838542,
      "completions/min_length": 204.66666666666666,
      "completions/min_terminated_length": 204.66666666666666,
      "epoch": 2.226843100189036,
      "grad_norm": 17.615537643432617,
      "learning_rate": 1e-06,
      "loss": 0.0063,
      "num_tokens": 53701931.0,
      "reward": 0.5170595248540243,
      "reward_std": 0.08570278746386369,
      "rewards/get_embedding_sim/mean": 0.4467470149199168,
      "rewards/get_embedding_sim/std": 0.044603065277139343,
      "rewards/reward_num_unique_chars/mean": 0.0234375,
      "rewards/reward_num_unique_chars/std": 0.08533549308776855,
      "step": 147
    },
    {
      "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": 0.9418402777777777,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 986.0,
      "completions/mean_length": 987.9757080078125,
      "completions/mean_terminated_length": 411.26666259765625,
      "completions/min_length": 144.0,
      "completions/min_terminated_length": 144.0,
      "epoch": 2.272211720226843,
      "grad_norm": 8.74587631225586,
      "learning_rate": 1e-06,
      "loss": -0.0112,
      "num_tokens": 55429903.0,
      "reward": 0.47102906306584674,
      "reward_std": 0.014634329825639725,
      "rewards/get_embedding_sim/mean": 0.4710290531317393,
      "rewards/get_embedding_sim/std": 0.033305841808517776,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 150
    },
    {
      "epoch": 2.31758034026465,
      "grad_norm": 30.690135955810547,
      "learning_rate": 1e-06,
      "loss": -0.002,
      "step": 153
    },
    {
      "epoch": 2.31758034026465,
      "eval_clip_ratio/high_max": 0.0,
      "eval_clip_ratio/high_mean": 0.0,
      "eval_clip_ratio/low_mean": 0.0,
      "eval_clip_ratio/low_min": 0.0,
      "eval_clip_ratio/region_mean": 0.0,
      "eval_completions/clipped_ratio": 0.9482886904761905,
      "eval_completions/max_length": 1024.0,
      "eval_completions/max_terminated_length": 129.91071428571428,
      "eval_completions/mean_length": 989.9851215907505,
      "eval_completions/mean_terminated_length": 76.58274132864815,
      "eval_completions/min_length": 894.125,
      "eval_completions/min_terminated_length": 52.982142857142854,
      "eval_loss": -0.005262339487671852,
      "eval_num_tokens": 57173973.0,
      "eval_reward": 0.5007602897073541,
      "eval_reward_std": 0.004911966698541294,
      "eval_rewards/get_embedding_sim/mean": 0.5007602822567735,
      "eval_rewards/get_embedding_sim/std": 0.004911966153254558,
      "eval_rewards/reward_num_unique_chars/mean": 0.0,
      "eval_rewards/reward_num_unique_chars/std": 0.0,
      "eval_runtime": 7731.9419,
      "eval_samples_per_second": 0.007,
      "eval_steps_per_second": 0.0,
      "step": 153
    },
    {
      "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": 0.9375,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 525.6666666666666,
      "completions/mean_length": 982.8880310058594,
      "completions/mean_terminated_length": 230.5204086303711,
      "completions/min_length": 466.0,
      "completions/min_terminated_length": 124.66666666666667,
      "epoch": 2.3629489603024574,
      "grad_norm": 2.623792886734009,
      "learning_rate": 1e-06,
      "loss": 0.0178,
      "num_tokens": 58851517.0,
      "reward": 0.4892935703198115,
      "reward_std": 0.030512975354213268,
      "rewards/get_embedding_sim/mean": 0.47757480045159656,
      "rewards/get_embedding_sim/std": 0.02862175398816665,
      "rewards/reward_num_unique_chars/mean": 0.00390625,
      "rewards/reward_num_unique_chars/std": 0.02524763097365697,
      "step": 156
    },
    {
      "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": 0.9262152777777778,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1009.6666666666666,
      "completions/mean_length": 990.2352498372396,
      "completions/mean_terminated_length": 566.21484375,
      "completions/min_length": 283.0,
      "completions/min_terminated_length": 283.0,
      "epoch": 2.4083175803402646,
      "grad_norm": 9.202348709106445,
      "learning_rate": 1e-06,
      "loss": 0.0084,
      "num_tokens": 60579500.0,
      "reward": 0.47327958544095355,
      "reward_std": 0.004001262908180554,
      "rewards/get_embedding_sim/mean": 0.47327956557273865,
      "rewards/get_embedding_sim/std": 0.03640973443786303,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 159
    },
    {
      "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": 0.907986111111111,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 943.0,
      "completions/mean_length": 981.1085408528646,
      "completions/mean_terminated_length": 633.5793965657552,
      "completions/min_length": 343.0,
      "completions/min_terminated_length": 343.0,
      "epoch": 2.4536862003780717,
      "grad_norm": 69.65328979492188,
      "learning_rate": 1e-06,
      "loss": 0.0092,
      "num_tokens": 62292361.0,
      "reward": 0.4694495002428691,
      "reward_std": 0.004467625229153782,
      "rewards/get_embedding_sim/mean": 0.4694494605064392,
      "rewards/get_embedding_sim/std": 0.03462202536563078,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 162
    },
    {
      "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": 0.9513888888888888,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 675.0,
      "completions/mean_length": 1018.2404581705729,
      "completions/mean_terminated_length": 630.1222127278646,
      "completions/min_length": 914.0,
      "completions/min_terminated_length": 572.6666666666666,
      "epoch": 2.499054820415879,
      "grad_norm": 0.23563966155052185,
      "learning_rate": 1e-06,
      "loss": -0.0,
      "num_tokens": 64055198.0,
      "reward": 0.4772230386734009,
      "reward_std": 0.0006476088650136566,
      "rewards/get_embedding_sim/mean": 0.4772230088710785,
      "rewards/get_embedding_sim/std": 0.028875624140103657,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 165
    },
    {
      "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": 0.9305555555555555,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 952.0,
      "completions/mean_length": 1010.2786661783854,
      "completions/mean_terminated_length": 819.9857991536459,
      "completions/min_length": 694.6666666666666,
      "completions/min_terminated_length": 694.6666666666666,
      "epoch": 2.544423440453686,
      "grad_norm": 0.5504884719848633,
      "learning_rate": 1e-06,
      "loss": 0.0001,
      "num_tokens": 65808863.0,
      "reward": 0.4726286629835765,
      "reward_std": 3.934144478989765e-05,
      "rewards/get_embedding_sim/mean": 0.472628653049469,
      "rewards/get_embedding_sim/std": 0.027834332858522732,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 168
    },
    {
      "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": 0.9253472222222222,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 669.0,
      "completions/mean_length": 994.8368326822916,
      "completions/mean_terminated_length": 522.8986206054688,
      "completions/min_length": 689.3333333333334,
      "completions/min_terminated_length": 348.0,
      "epoch": 2.5897920604914932,
      "grad_norm": 0.6784894466400146,
      "learning_rate": 1e-06,
      "loss": 0.0001,
      "num_tokens": 67536867.0,
      "reward": 0.49061213930447894,
      "reward_std": 0.0015911554898290585,
      "rewards/get_embedding_sim/mean": 0.49061209956804913,
      "rewards/get_embedding_sim/std": 0.030627068132162094,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 171
    },
    {
      "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": 0.9331597222222222,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 318.0,
      "completions/mean_length": 984.9166666666666,
      "completions/mean_terminated_length": 146.42424519856772,
      "completions/min_length": 786.3333333333334,
      "completions/min_terminated_length": 103.66666666666667,
      "epoch": 2.6351606805293004,
      "grad_norm": 0.1935257911682129,
      "learning_rate": 1e-06,
      "loss": -0.0027,
      "num_tokens": 69251571.0,
      "reward": 0.4839080174763997,
      "reward_std": 0.0016439886846152756,
      "rewards/get_embedding_sim/mean": 0.4839079678058624,
      "rewards/get_embedding_sim/std": 0.029750562583406765,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 174
    },
    {
      "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": 0.88671875,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 976.0,
      "completions/mean_length": 1003.83203125,
      "completions/mean_terminated_length": 855.934326171875,
      "completions/min_length": 643.0,
      "completions/min_terminated_length": 643.0,
      "epoch": 2.6805293005671076,
      "grad_norm": 3.1704888343811035,
      "learning_rate": 1e-06,
      "loss": 0.0019,
      "num_tokens": 70958103.0,
      "reward": 0.4883721321821213,
      "reward_std": 0.0021593487645077403,
      "rewards/get_embedding_sim/mean": 0.4883721321821213,
      "rewards/get_embedding_sim/std": 0.025321779772639275,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 177
    },
    {
      "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": 0.8706597222222222,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 968.3333333333334,
      "completions/mean_length": 991.0763956705729,
      "completions/mean_terminated_length": 788.4104410807291,
      "completions/min_length": 603.0,
      "completions/min_terminated_length": 603.0,
      "epoch": 2.7258979206049148,
      "grad_norm": 3.6269359588623047,
      "learning_rate": 1e-06,
      "loss": 0.0003,
      "num_tokens": 72681103.0,
      "reward": 0.46831490596135456,
      "reward_std": 0.0003801115908572683,
      "rewards/get_embedding_sim/mean": 0.46831488609313965,
      "rewards/get_embedding_sim/std": 0.032175165290633835,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 180
    },
    {
      "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": 0.8628472222222222,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1018.0,
      "completions/mean_length": 997.5885416666666,
      "completions/mean_terminated_length": 854.9134724934896,
      "completions/min_length": 666.6666666666666,
      "completions/min_terminated_length": 666.6666666666666,
      "epoch": 2.7712665406427224,
      "grad_norm": 0.16258712112903595,
      "learning_rate": 1e-06,
      "loss": -0.0007,
      "num_tokens": 74405941.0,
      "reward": 0.47544575730959576,
      "reward_std": 0.0010224993523782662,
      "rewards/get_embedding_sim/mean": 0.4754457275072734,
      "rewards/get_embedding_sim/std": 0.032684508711099625,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 183
    },
    {
      "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": 0.9435763888888888,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 674.3333333333334,
      "completions/mean_length": 1004.1059163411459,
      "completions/mean_terminated_length": 516.867919921875,
      "completions/min_length": 631.3333333333334,
      "completions/min_terminated_length": 290.0,
      "epoch": 2.816635160680529,
      "grad_norm": 0.16970351338386536,
      "learning_rate": 1e-06,
      "loss": 0.0018,
      "num_tokens": 76144623.0,
      "reward": 0.4907430410385132,
      "reward_std": 0.0015885522558770997,
      "rewards/get_embedding_sim/mean": 0.4907430013020833,
      "rewards/get_embedding_sim/std": 0.03282180055975914,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 186
    },
    {
      "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": 0.9782986111111112,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 658.3333333333334,
      "completions/mean_length": 1019.5789998372396,
      "completions/mean_terminated_length": 608.4166666666666,
      "completions/min_length": 915.0,
      "completions/min_terminated_length": 573.6666666666666,
      "epoch": 2.8620037807183367,
      "grad_norm": 0.432363897562027,
      "learning_rate": 1e-06,
      "loss": 0.0002,
      "num_tokens": 77909002.0,
      "reward": 0.48864559332529706,
      "reward_std": 0.0002599236443832827,
      "rewards/get_embedding_sim/mean": 0.4886455436547597,
      "rewards/get_embedding_sim/std": 0.029337758819262188,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 189
    },
    {
      "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": 0.9661458333333334,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 319.6666666666667,
      "completions/mean_length": 1010.1883748372396,
      "completions/mean_terminated_length": 205.34187825520834,
      "completions/min_length": 822.0,
      "completions/min_terminated_length": 139.33333333333334,
      "epoch": 2.9073724007561434,
      "grad_norm": 0.5266261696815491,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "num_tokens": 79655219.0,
      "reward": 0.4859796365102132,
      "reward_std": 0.0003688052911456907,
      "rewards/get_embedding_sim/mean": 0.4859795967737834,
      "rewards/get_embedding_sim/std": 0.021221444631616276,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 192
    },
    {
      "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": 0.9262152777777777,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1017.0,
      "completions/mean_length": 1012.0989786783854,
      "completions/mean_terminated_length": 882.4782918294271,
      "completions/min_length": 770.6666666666666,
      "completions/min_terminated_length": 770.6666666666666,
      "epoch": 2.952741020793951,
      "grad_norm": 0.4420604109764099,
      "learning_rate": 1e-06,
      "loss": 0.0001,
      "num_tokens": 81410981.0,
      "reward": 0.48151273528734845,
      "reward_std": 0.00013237820166978054,
      "rewards/get_embedding_sim/mean": 0.48151270548502606,
      "rewards/get_embedding_sim/std": 0.0341289850572745,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 195
    },
    {
      "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": 0.9340277777777778,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1020.0,
      "completions/mean_length": 1000.9948120117188,
      "completions/mean_terminated_length": 821.5246988932291,
      "completions/min_length": 590.6666666666666,
      "completions/min_terminated_length": 590.6666666666666,
      "epoch": 2.998109640831758,
      "grad_norm": 0.9245023131370544,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "num_tokens": 83141135.0,
      "reward": 0.4751121997833252,
      "reward_std": 0.0013413634515018202,
      "rewards/get_embedding_sim/mean": 0.4751121699810028,
      "rewards/get_embedding_sim/std": 0.034396614879369736,
      "rewards/reward_num_unique_chars/mean": 0.0,
      "rewards/reward_num_unique_chars/std": 0.0,
      "step": 198
    }
  ],
  "logging_steps": 3,
  "max_steps": 198,
  "num_input_tokens_seen": 83141135,
  "num_train_epochs": 3,
  "save_steps": 25,
  "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
}