{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 100,
  "global_step": 201,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 601.5053852081298,
      "epoch": 0.14925373134328357,
      "grad_norm": 6.400096893310547,
      "kl": 0.000240480899810791,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 0.0,
      "reward": 0.671540210954845,
      "reward_std": 0.3266597702167928,
      "rewards/accuracy_reward": 0.671540210954845,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "completion_length": 627.042663192749,
      "epoch": 0.29850746268656714,
      "grad_norm": 0.10202192515134811,
      "kl": 0.0013910531997680664,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 0.0001,
      "reward": 0.7212053906172514,
      "reward_std": 0.26295756818726657,
      "rewards/accuracy_reward": 0.7212053906172514,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "completion_length": 598.0592895507813,
      "epoch": 0.44776119402985076,
      "grad_norm": 0.07559198141098022,
      "kl": 0.003163862228393555,
      "learning_rate": 2.981532510892707e-06,
      "loss": 0.0001,
      "reward": 0.7546875355765224,
      "reward_std": 0.18810537890531123,
      "rewards/accuracy_reward": 0.7546875355765224,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "completion_length": 594.8722389221191,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.09930674731731415,
      "kl": 0.0029309749603271484,
      "learning_rate": 2.9182778633989753e-06,
      "loss": 0.0001,
      "reward": 0.7710937844589353,
      "reward_std": 0.17225130782462655,
      "rewards/accuracy_reward": 0.7710937844589353,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "completion_length": 580.8643112182617,
      "epoch": 0.746268656716418,
      "grad_norm": 0.07715580612421036,
      "kl": 0.0032186985015869142,
      "learning_rate": 2.811929560709094e-06,
      "loss": 0.0001,
      "reward": 0.7782366409897804,
      "reward_std": 0.168875193875283,
      "rewards/accuracy_reward": 0.7782366409897804,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "completion_length": 567.9614114761353,
      "epoch": 0.8955223880597015,
      "grad_norm": 0.1152808889746666,
      "kl": 0.0035730838775634766,
      "learning_rate": 2.6657189421854562e-06,
      "loss": 0.0001,
      "reward": 0.7762277115136385,
      "reward_std": 0.16049637314863502,
      "rewards/accuracy_reward": 0.7762277115136385,
      "rewards/format_reward": 0.0,
      "step": 60
    },
    {
      "completion_length": 550.2290409088134,
      "epoch": 1.044776119402985,
      "grad_norm": 0.09865733981132507,
      "kl": 0.0038765430450439452,
      "learning_rate": 2.484088543485761e-06,
      "loss": 0.0002,
      "reward": 0.7911458652466535,
      "reward_std": 0.150591628998518,
      "rewards/accuracy_reward": 0.7911458652466535,
      "rewards/format_reward": 0.0,
      "step": 70
    },
    {
      "completion_length": 537.502813911438,
      "epoch": 1.1940298507462686,
      "grad_norm": 0.09184063225984573,
      "kl": 0.004058647155761719,
      "learning_rate": 2.2725571123650813e-06,
      "loss": 0.0002,
      "reward": 0.8042411081492901,
      "reward_std": 0.1556034866720438,
      "rewards/accuracy_reward": 0.8042411081492901,
      "rewards/format_reward": 0.0,
      "step": 80
    },
    {
      "completion_length": 540.724467086792,
      "epoch": 1.3432835820895521,
      "grad_norm": 0.13207751512527466,
      "kl": 0.0038210868835449217,
      "learning_rate": 2.03755192431795e-06,
      "loss": 0.0002,
      "reward": 0.7966518238186836,
      "reward_std": 0.16528417109511792,
      "rewards/accuracy_reward": 0.7966518238186836,
      "rewards/format_reward": 0.0,
      "step": 90
    },
    {
      "completion_length": 541.9515859603882,
      "epoch": 1.4925373134328357,
      "grad_norm": 0.07470321655273438,
      "kl": 0.003845405578613281,
      "learning_rate": 1.7862134930648174e-06,
      "loss": 0.0002,
      "reward": 0.796093787625432,
      "reward_std": 0.14775474979542197,
      "rewards/accuracy_reward": 0.796093787625432,
      "rewards/format_reward": 0.0,
      "step": 100
    },
    {
      "epoch": 1.4925373134328357,
      "eval_completion_length": 540.39465481418,
      "eval_kl": 0.004954966298350088,
      "eval_loss": 0.00019825901836156845,
      "eval_reward": 0.6999750570400611,
      "eval_reward_std": 0.19887771770045473,
      "eval_rewards/accuracy_reward": 0.6999750570400611,
      "eval_rewards/format_reward": 0.0,
      "eval_runtime": 7024.4867,
      "eval_samples_per_second": 0.712,
      "eval_steps_per_second": 0.102,
      "step": 100
    },
    {
      "completion_length": 561.2457847595215,
      "epoch": 1.6417910447761193,
      "grad_norm": 0.08826518058776855,
      "kl": 0.003845024108886719,
      "learning_rate": 1.5261786096559255e-06,
      "loss": 0.0002,
      "reward": 0.7954241424798966,
      "reward_std": 0.17211202825419605,
      "rewards/accuracy_reward": 0.7954241424798966,
      "rewards/format_reward": 0.0,
      "step": 110
    },
    {
      "completion_length": 552.2967874526978,
      "epoch": 1.7910447761194028,
      "grad_norm": 0.22947834432125092,
      "kl": 0.003882312774658203,
      "learning_rate": 1.2653483024396534e-06,
      "loss": 0.0002,
      "reward": 0.7954241406172514,
      "reward_std": 0.1543394286185503,
      "rewards/accuracy_reward": 0.7954241406172514,
      "rewards/format_reward": 0.0,
      "step": 120
    },
    {
      "completion_length": 534.2573900222778,
      "epoch": 1.9402985074626866,
      "grad_norm": 0.14124220609664917,
      "kl": 0.004029464721679687,
      "learning_rate": 1.0116477683142654e-06,
      "loss": 0.0002,
      "reward": 0.8051339661702513,
      "reward_std": 0.15696909232065082,
      "rewards/accuracy_reward": 0.8051339661702513,
      "rewards/format_reward": 0.0,
      "step": 130
    },
    {
      "completion_length": 544.0703746795655,
      "epoch": 2.08955223880597,
      "grad_norm": 0.10850396007299423,
      "kl": 0.003820991516113281,
      "learning_rate": 7.727855696304945e-07,
      "loss": 0.0002,
      "reward": 0.8154018180444836,
      "reward_std": 0.15202449564822018,
      "rewards/accuracy_reward": 0.8154018180444836,
      "rewards/format_reward": 0.0,
      "step": 140
    },
    {
      "completion_length": 549.2021450042724,
      "epoch": 2.2388059701492535,
      "grad_norm": 0.11342689394950867,
      "kl": 0.0039272308349609375,
      "learning_rate": 5.560194134252441e-07,
      "loss": 0.0002,
      "reward": 0.795312536507845,
      "reward_std": 0.158915360737592,
      "rewards/accuracy_reward": 0.795312536507845,
      "rewards/format_reward": 0.0,
      "step": 150
    },
    {
      "completion_length": 532.3224586486816,
      "epoch": 2.388059701492537,
      "grad_norm": 0.14040400087833405,
      "kl": 0.004098033905029297,
      "learning_rate": 3.67935629665842e-07,
      "loss": 0.0002,
      "reward": 0.816071467846632,
      "reward_std": 0.1492480550892651,
      "rewards/accuracy_reward": 0.816071467846632,
      "rewards/format_reward": 0.0,
      "step": 160
    },
    {
      "completion_length": 539.504601097107,
      "epoch": 2.5373134328358207,
      "grad_norm": 0.09679315984249115,
      "kl": 0.0059491157531738285,
      "learning_rate": 2.1424904894683168e-07,
      "loss": 0.0002,
      "reward": 0.8157366402447224,
      "reward_std": 0.1421221026685089,
      "rewards/accuracy_reward": 0.8157366402447224,
      "rewards/format_reward": 0.0,
      "step": 170
    },
    {
      "completion_length": 547.4501371383667,
      "epoch": 2.6865671641791042,
      "grad_norm": 0.10368100553750992,
      "kl": 0.003998279571533203,
      "learning_rate": 9.962936025419756e-08,
      "loss": 0.0002,
      "reward": 0.8022321790456772,
      "reward_std": 0.1627280049957335,
      "rewards/accuracy_reward": 0.8022321790456772,
      "rewards/format_reward": 0.0,
      "step": 180
    },
    {
      "completion_length": 533.6452249526977,
      "epoch": 2.835820895522388,
      "grad_norm": 0.09484420716762543,
      "kl": 0.0039805412292480465,
      "learning_rate": 2.7559224828504036e-08,
      "loss": 0.0002,
      "reward": 0.8222098549827933,
      "reward_std": 0.14822139525786043,
      "rewards/accuracy_reward": 0.8222098549827933,
      "rewards/format_reward": 0.0,
      "step": 190
    },
    {
      "completion_length": 546.433731842041,
      "epoch": 2.9850746268656714,
      "grad_norm": 0.11888264119625092,
      "kl": 0.003950977325439453,
      "learning_rate": 2.2845726541309565e-10,
      "loss": 0.0002,
      "reward": 0.8064732495695353,
      "reward_std": 0.15783933801576494,
      "rewards/accuracy_reward": 0.8064732495695353,
      "rewards/format_reward": 0.0,
      "step": 200
    },
    {
      "epoch": 2.9850746268656714,
      "eval_completion_length": 538.2288205660307,
      "eval_kl": 0.004983488496366914,
      "eval_loss": 0.00019947961845900863,
      "eval_reward": 0.7005245058478176,
      "eval_reward_std": 0.19627383342989674,
      "eval_rewards/accuracy_reward": 0.7005245058478176,
      "eval_rewards/format_reward": 0.0,
      "eval_runtime": 7022.2087,
      "eval_samples_per_second": 0.712,
      "eval_steps_per_second": 0.102,
      "step": 200
    },
    {
      "completion_length": 547.8125114440918,
      "epoch": 3.0,
      "kl": 0.005582332611083984,
      "reward": 0.8750000018626451,
      "reward_std": 0.0765409953892231,
      "rewards/accuracy_reward": 0.8750000018626451,
      "rewards/format_reward": 0.0,
      "step": 201,
      "total_flos": 0.0,
      "train_loss": 0.0001433196965044953,
      "train_runtime": 49513.7568,
      "train_samples_per_second": 0.454,
      "train_steps_per_second": 0.004
    }
  ],
  "logging_steps": 10,
  "max_steps": 201,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}