{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.32,
  "eval_steps": 500,
  "global_step": 40,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 192.625,
      "epoch": 0.008,
      "grad_norm": 10.768404960632324,
      "kl": 0.0009242900705430657,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.0,
      "reward": 0.2329830303788185,
      "reward_std": 0.14905179431661963,
      "rewards/<lambda>": 0.2329830303788185,
      "step": 1
    },
    {
      "completion_length": 226.25,
      "epoch": 0.016,
      "grad_norm": 0.5803981423377991,
      "kl": 0.00016602075265836902,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.0,
      "reward": 0.20506704226136208,
      "reward_std": 0.07600272889249027,
      "rewards/<lambda>": 0.20506704226136208,
      "step": 2
    },
    {
      "completion_length": 195.5,
      "epoch": 0.024,
      "grad_norm": 1.928178071975708,
      "kl": 0.0007406917939078994,
      "learning_rate": 1.5e-06,
      "loss": 0.0,
      "reward": 0.24738124385476112,
      "reward_std": 0.12009324785321951,
      "rewards/<lambda>": 0.24738124385476112,
      "step": 3
    },
    {
      "completion_length": 165.25,
      "epoch": 0.032,
      "grad_norm": 4.732048034667969,
      "kl": 0.0003298191677458817,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.0,
      "reward": 0.20593957975506783,
      "reward_std": 0.13085306528955698,
      "rewards/<lambda>": 0.20593957975506783,
      "step": 4
    },
    {
      "completion_length": 229.25,
      "epoch": 0.04,
      "grad_norm": 0.6906648278236389,
      "kl": 0.00029802451899740845,
      "learning_rate": 2.5e-06,
      "loss": 0.0,
      "reward": 0.2962687499821186,
      "reward_std": 0.11282773199491203,
      "rewards/<lambda>": 0.2962687499821186,
      "step": 5
    },
    {
      "completion_length": 235.75,
      "epoch": 0.048,
      "grad_norm": 1.7818254232406616,
      "kl": 0.0009266494062103448,
      "learning_rate": 3e-06,
      "loss": 0.0,
      "reward": 0.17379014939069748,
      "reward_std": 0.03908907831646502,
      "rewards/<lambda>": 0.17379014939069748,
      "step": 6
    },
    {
      "completion_length": 141.25,
      "epoch": 0.056,
      "grad_norm": 5.83347225189209,
      "kl": 0.0010952446609735489,
      "learning_rate": 3.5e-06,
      "loss": 0.0,
      "reward": 0.18913621827960014,
      "reward_std": 0.13922751136124134,
      "rewards/<lambda>": 0.18913621827960014,
      "step": 7
    },
    {
      "completion_length": 193.125,
      "epoch": 0.064,
      "grad_norm": 4.481875896453857,
      "kl": 0.002368737303186208,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0001,
      "reward": 0.15933124721050262,
      "reward_std": 0.033755510827177204,
      "rewards/<lambda>": 0.15933124721050262,
      "step": 8
    },
    {
      "completion_length": 256.0,
      "epoch": 0.072,
      "grad_norm": 0.335154265165329,
      "kl": 0.00025119713973253965,
      "learning_rate": 4.5e-06,
      "loss": 0.0,
      "reward": 0.36442025750875473,
      "reward_std": 0.21805795654654503,
      "rewards/<lambda>": 0.36442025750875473,
      "step": 9
    },
    {
      "completion_length": 256.0,
      "epoch": 0.08,
      "grad_norm": 0.5085432529449463,
      "kl": 0.00022752030054107308,
      "learning_rate": 5e-06,
      "loss": 0.0,
      "reward": 0.26649028062820435,
      "reward_std": 0.05319603928364813,
      "rewards/<lambda>": 0.26649028062820435,
      "step": 10
    },
    {
      "completion_length": 194.875,
      "epoch": 0.088,
      "grad_norm": 5.396770000457764,
      "kl": 0.0003068408841500059,
      "learning_rate": 4.944444444444445e-06,
      "loss": 0.0,
      "reward": 0.1397937536239624,
      "reward_std": 0.049082053592428565,
      "rewards/<lambda>": 0.1397937536239624,
      "step": 11
    },
    {
      "completion_length": 228.25,
      "epoch": 0.096,
      "grad_norm": 0.9890621900558472,
      "kl": 0.0005189948133192956,
      "learning_rate": 4.888888888888889e-06,
      "loss": 0.0,
      "reward": 0.3081398792564869,
      "reward_std": 0.21328446036204696,
      "rewards/<lambda>": 0.3081398792564869,
      "step": 12
    },
    {
      "completion_length": 78.125,
      "epoch": 0.104,
      "grad_norm": 6.991876602172852,
      "kl": 0.0028572251394507475,
      "learning_rate": 4.833333333333333e-06,
      "loss": 0.0001,
      "reward": 0.16984067671000957,
      "reward_std": 0.09841607791895512,
      "rewards/<lambda>": 0.16984067671000957,
      "step": 13
    },
    {
      "completion_length": 224.5,
      "epoch": 0.112,
      "grad_norm": 2.380413770675659,
      "kl": 0.00024483678862452507,
      "learning_rate": 4.777777777777778e-06,
      "loss": 0.0,
      "reward": 0.3446955271065235,
      "reward_std": 0.07521216722670943,
      "rewards/<lambda>": 0.3446955271065235,
      "step": 14
    },
    {
      "completion_length": 198.25,
      "epoch": 0.12,
      "grad_norm": 0.9101364612579346,
      "kl": 0.000599912746110931,
      "learning_rate": 4.722222222222222e-06,
      "loss": 0.0,
      "reward": 0.3019299991428852,
      "reward_std": 0.11653121118433774,
      "rewards/<lambda>": 0.3019299991428852,
      "step": 15
    },
    {
      "completion_length": 256.0,
      "epoch": 0.128,
      "grad_norm": 0.3337076008319855,
      "kl": 0.00042263357318006456,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.0,
      "reward": 0.22962501272559166,
      "reward_std": 0.12568823376204818,
      "rewards/<lambda>": 0.22962501272559166,
      "step": 16
    },
    {
      "completion_length": 194.5,
      "epoch": 0.136,
      "grad_norm": 2.9452309608459473,
      "kl": 0.0003321969779790379,
      "learning_rate": 4.611111111111112e-06,
      "loss": 0.0,
      "reward": 0.2720954604446888,
      "reward_std": 0.11709045059978962,
      "rewards/<lambda>": 0.2720954604446888,
      "step": 17
    },
    {
      "completion_length": 138.0,
      "epoch": 0.144,
      "grad_norm": 6.265693664550781,
      "kl": 0.008480061980662867,
      "learning_rate": 4.555555555555556e-06,
      "loss": 0.0003,
      "reward": 0.21834762021899223,
      "reward_std": 0.15078714862465858,
      "rewards/<lambda>": 0.21834762021899223,
      "step": 18
    },
    {
      "completion_length": 207.25,
      "epoch": 0.152,
      "grad_norm": 1.309558391571045,
      "kl": 0.0008433158218394965,
      "learning_rate": 4.5e-06,
      "loss": 0.0,
      "reward": 0.2984776981174946,
      "reward_std": 0.10264036408625543,
      "rewards/<lambda>": 0.2984776981174946,
      "step": 19
    },
    {
      "completion_length": 198.0,
      "epoch": 0.16,
      "grad_norm": 3.6814424991607666,
      "kl": 0.0018292521999683231,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.0001,
      "reward": 0.21348833292722702,
      "reward_std": 0.032345420273486525,
      "rewards/<lambda>": 0.21348833292722702,
      "step": 20
    },
    {
      "completion_length": 253.0,
      "epoch": 0.168,
      "grad_norm": 0.3527783751487732,
      "kl": 0.000525618303072406,
      "learning_rate": 4.388888888888889e-06,
      "loss": 0.0,
      "reward": 0.32108124345541,
      "reward_std": 0.13784162979573011,
      "rewards/<lambda>": 0.32108124345541,
      "step": 21
    },
    {
      "completion_length": 256.0,
      "epoch": 0.176,
      "grad_norm": 0.34396272897720337,
      "kl": 0.00019161385716870427,
      "learning_rate": 4.333333333333334e-06,
      "loss": 0.0,
      "reward": 0.2716740742325783,
      "reward_std": 0.0921019627712667,
      "rewards/<lambda>": 0.2716740742325783,
      "step": 22
    },
    {
      "completion_length": 256.0,
      "epoch": 0.184,
      "grad_norm": 0.38449183106422424,
      "kl": 0.0004789537051692605,
      "learning_rate": 4.277777777777778e-06,
      "loss": 0.0,
      "reward": 0.3738388866186142,
      "reward_std": 0.2010854547843337,
      "rewards/<lambda>": 0.3738388866186142,
      "step": 23
    },
    {
      "completion_length": 256.0,
      "epoch": 0.192,
      "grad_norm": 0.35797202587127686,
      "kl": 0.0003909100778400898,
      "learning_rate": 4.222222222222223e-06,
      "loss": 0.0,
      "reward": 0.3206794075667858,
      "reward_std": 0.20672890054993331,
      "rewards/<lambda>": 0.3206794075667858,
      "step": 24
    },
    {
      "completion_length": 256.0,
      "epoch": 0.2,
      "grad_norm": 0.22144320607185364,
      "kl": 0.0001992348989006132,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.0,
      "reward": 0.1902062501758337,
      "reward_std": 0.07741935132071376,
      "rewards/<lambda>": 0.1902062501758337,
      "step": 25
    },
    {
      "completion_length": 194.0,
      "epoch": 0.208,
      "grad_norm": 7.687119007110596,
      "kl": 0.0075769436080008745,
      "learning_rate": 4.111111111111111e-06,
      "loss": 0.0003,
      "reward": 0.27252499759197235,
      "reward_std": 0.11423310358077288,
      "rewards/<lambda>": 0.27252499759197235,
      "step": 26
    },
    {
      "completion_length": 226.75,
      "epoch": 0.216,
      "grad_norm": 4.88610315322876,
      "kl": 0.005719653330743313,
      "learning_rate": 4.055555555555556e-06,
      "loss": 0.0002,
      "reward": 0.28929195925593376,
      "reward_std": 0.2355259140022099,
      "rewards/<lambda>": 0.28929195925593376,
      "step": 27
    },
    {
      "completion_length": 163.875,
      "epoch": 0.224,
      "grad_norm": 0.8618195056915283,
      "kl": 0.001144724345067516,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0,
      "reward": 0.2521233316510916,
      "reward_std": 0.18312063068151474,
      "rewards/<lambda>": 0.2521233316510916,
      "step": 28
    },
    {
      "completion_length": 193.375,
      "epoch": 0.232,
      "grad_norm": 5.040801048278809,
      "kl": 0.003916982212103903,
      "learning_rate": 3.944444444444445e-06,
      "loss": 0.0002,
      "reward": 0.2782749943435192,
      "reward_std": 0.09372699866071343,
      "rewards/<lambda>": 0.2782749943435192,
      "step": 29
    },
    {
      "completion_length": 203.5,
      "epoch": 0.24,
      "grad_norm": 1.5515941381454468,
      "kl": 0.0015425277961185202,
      "learning_rate": 3.88888888888889e-06,
      "loss": 0.0001,
      "reward": 0.327726773917675,
      "reward_std": 0.13597915810532868,
      "rewards/<lambda>": 0.327726773917675,
      "step": 30
    },
    {
      "completion_length": 224.375,
      "epoch": 0.248,
      "grad_norm": 6.099854946136475,
      "kl": 0.0065212192421313375,
      "learning_rate": 3.833333333333334e-06,
      "loss": 0.0003,
      "reward": 0.26245963387191296,
      "reward_std": 0.23274014610797167,
      "rewards/<lambda>": 0.26245963387191296,
      "step": 31
    },
    {
      "completion_length": 225.5,
      "epoch": 0.256,
      "grad_norm": 0.9755590558052063,
      "kl": 0.00134121990413405,
      "learning_rate": 3.777777777777778e-06,
      "loss": 0.0001,
      "reward": 0.2514333389699459,
      "reward_std": 0.07464690878987312,
      "rewards/<lambda>": 0.2514333389699459,
      "step": 32
    },
    {
      "completion_length": 228.875,
      "epoch": 0.264,
      "grad_norm": 0.5367397665977478,
      "kl": 0.0005917141388636082,
      "learning_rate": 3.7222222222222225e-06,
      "loss": 0.0,
      "reward": 0.3105039447546005,
      "reward_std": 0.029746492160484195,
      "rewards/<lambda>": 0.3105039447546005,
      "step": 33
    },
    {
      "completion_length": 194.5,
      "epoch": 0.272,
      "grad_norm": 3.941100597381592,
      "kl": 0.010093522083479911,
      "learning_rate": 3.6666666666666666e-06,
      "loss": 0.0004,
      "reward": 0.36005016416311264,
      "reward_std": 0.20752966683357954,
      "rewards/<lambda>": 0.36005016416311264,
      "step": 34
    },
    {
      "completion_length": 162.125,
      "epoch": 0.28,
      "grad_norm": 2.5163803100585938,
      "kl": 0.015705711644841358,
      "learning_rate": 3.6111111111111115e-06,
      "loss": 0.0006,
      "reward": 0.2547324914485216,
      "reward_std": 0.13099860399961472,
      "rewards/<lambda>": 0.2547324914485216,
      "step": 35
    },
    {
      "completion_length": 256.0,
      "epoch": 0.288,
      "grad_norm": 0.40727710723876953,
      "kl": 0.00078230255167,
      "learning_rate": 3.555555555555556e-06,
      "loss": 0.0,
      "reward": 0.32325625233352184,
      "reward_std": 0.1363072171807289,
      "rewards/<lambda>": 0.32325625233352184,
      "step": 36
    },
    {
      "completion_length": 196.75,
      "epoch": 0.296,
      "grad_norm": 3.323727607727051,
      "kl": 0.003316763584734872,
      "learning_rate": 3.5e-06,
      "loss": 0.0001,
      "reward": 0.21428124606609344,
      "reward_std": 0.10128421382978559,
      "rewards/<lambda>": 0.21428124606609344,
      "step": 37
    },
    {
      "completion_length": 224.75,
      "epoch": 0.304,
      "grad_norm": 4.83845853805542,
      "kl": 0.005240194499492645,
      "learning_rate": 3.444444444444445e-06,
      "loss": 0.0002,
      "reward": 0.21497500129044056,
      "reward_std": 0.15779088297858834,
      "rewards/<lambda>": 0.21497500129044056,
      "step": 38
    },
    {
      "completion_length": 230.125,
      "epoch": 0.312,
      "grad_norm": 0.7527925372123718,
      "kl": 0.0013605846324935555,
      "learning_rate": 3.3888888888888893e-06,
      "loss": 0.0001,
      "reward": 0.35972169786691666,
      "reward_std": 0.2259513009339571,
      "rewards/<lambda>": 0.35972169786691666,
      "step": 39
    },
    {
      "completion_length": 256.0,
      "epoch": 0.32,
      "grad_norm": 0.7770674824714661,
      "kl": 0.0018776137148961425,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.0001,
      "reward": 0.3781879246234894,
      "reward_std": 0.17891038954257965,
      "rewards/<lambda>": 0.3781879246234894,
      "step": 40
    }
  ],
  "logging_steps": 1,
  "max_steps": 100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 20,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}