{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.993607305936073,
  "eval_steps": 100,
  "global_step": 136,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 15.59375,
      "epoch": 0.0365296803652968,
      "grad_norm": 10.368077278137207,
      "kl": 0.15047144889831543,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 0.0295,
      "reward": 0.13515625,
      "reward_std": 0.19950873199850322,
      "rewards/accuracy_reward": 0.13203125,
      "rewards/format_reward": 0.003125,
      "step": 5
    },
    {
      "completion_length": 74.09296875,
      "epoch": 0.0730593607305936,
      "grad_norm": 6.168639183044434,
      "kl": 1.47294921875,
      "learning_rate": 1.4285714285714287e-05,
      "loss": 0.1038,
      "reward": 0.51015625,
      "reward_std": 0.4266123466193676,
      "rewards/accuracy_reward": 0.23203125,
      "rewards/format_reward": 0.278125,
      "step": 10
    },
    {
      "completion_length": 109.55703125,
      "epoch": 0.1095890410958904,
      "grad_norm": 10.966035842895508,
      "kl": 0.523876953125,
      "learning_rate": 1.9996684675143132e-05,
      "loss": 0.027,
      "reward": 1.046875,
      "reward_std": 0.46739846989512446,
      "rewards/accuracy_reward": 0.215625,
      "rewards/format_reward": 0.83125,
      "step": 15
    },
    {
      "completion_length": 125.8359375,
      "epoch": 0.1461187214611872,
      "grad_norm": 5.292937278747559,
      "kl": 0.46650390625,
      "learning_rate": 1.9880878960910772e-05,
      "loss": 0.0322,
      "reward": 1.22734375,
      "reward_std": 0.3152993652969599,
      "rewards/accuracy_reward": 0.2421875,
      "rewards/format_reward": 0.98515625,
      "step": 20
    },
    {
      "completion_length": 271.60078125,
      "epoch": 0.182648401826484,
      "grad_norm": 2.0227677822113037,
      "kl": 0.32373046875,
      "learning_rate": 1.960149873671602e-05,
      "loss": 0.005,
      "reward": 1.23359375,
      "reward_std": 0.41616563573479654,
      "rewards/accuracy_reward": 0.30078125,
      "rewards/format_reward": 0.9328125,
      "step": 25
    },
    {
      "completion_length": 134.89609375,
      "epoch": 0.2191780821917808,
      "grad_norm": 2.7977349758148193,
      "kl": 0.432275390625,
      "learning_rate": 1.916316904487005e-05,
      "loss": 0.0631,
      "reward": 1.353125,
      "reward_std": 0.26668183971196413,
      "rewards/accuracy_reward": 0.3578125,
      "rewards/format_reward": 0.9953125,
      "step": 30
    },
    {
      "completion_length": 93.9765625,
      "epoch": 0.2557077625570776,
      "grad_norm": 3.0016226768493652,
      "kl": 0.50341796875,
      "learning_rate": 1.8573146280763327e-05,
      "loss": 0.0361,
      "reward": 1.3046875,
      "reward_std": 0.26828942373394965,
      "rewards/accuracy_reward": 0.31640625,
      "rewards/format_reward": 0.98828125,
      "step": 35
    },
    {
      "completion_length": 117.1453125,
      "epoch": 0.2922374429223744,
      "grad_norm": 2.1085522174835205,
      "kl": 0.496875,
      "learning_rate": 1.7841198065767107e-05,
      "loss": -0.0058,
      "reward": 1.340625,
      "reward_std": 0.29750518389046193,
      "rewards/accuracy_reward": 0.353125,
      "rewards/format_reward": 0.9875,
      "step": 40
    },
    {
      "completion_length": 77.484375,
      "epoch": 0.3287671232876712,
      "grad_norm": 1.8194400072097778,
      "kl": 0.4939453125,
      "learning_rate": 1.6979441547663434e-05,
      "loss": -0.02,
      "reward": 1.3625,
      "reward_std": 0.2868320092558861,
      "rewards/accuracy_reward": 0.3734375,
      "rewards/format_reward": 0.9890625,
      "step": 45
    },
    {
      "completion_length": 57.68671875,
      "epoch": 0.365296803652968,
      "grad_norm": 1.7973569631576538,
      "kl": 0.652001953125,
      "learning_rate": 1.6002142805483686e-05,
      "loss": 0.0459,
      "reward": 1.3265625,
      "reward_std": 0.2394094867631793,
      "rewards/accuracy_reward": 0.33671875,
      "rewards/format_reward": 0.98984375,
      "step": 50
    },
    {
      "completion_length": 64.959375,
      "epoch": 0.4018264840182648,
      "grad_norm": 2.802725076675415,
      "kl": 0.602783203125,
      "learning_rate": 1.4925480679538646e-05,
      "loss": 0.0103,
      "reward": 1.3625,
      "reward_std": 0.1871818160638213,
      "rewards/accuracy_reward": 0.37734375,
      "rewards/format_reward": 0.98515625,
      "step": 55
    },
    {
      "completion_length": 94.5359375,
      "epoch": 0.4383561643835616,
      "grad_norm": 2.600384473800659,
      "kl": 0.4912109375,
      "learning_rate": 1.3767278936351853e-05,
      "loss": 0.0076,
      "reward": 1.33203125,
      "reward_std": 0.1461961718276143,
      "rewards/accuracy_reward": 0.33828125,
      "rewards/format_reward": 0.99375,
      "step": 60
    },
    {
      "completion_length": 60.94140625,
      "epoch": 0.4748858447488584,
      "grad_norm": 2.3066651821136475,
      "kl": 0.913916015625,
      "learning_rate": 1.2546711202412287e-05,
      "loss": 0.0148,
      "reward": 1.25546875,
      "reward_std": 0.26173087432980535,
      "rewards/accuracy_reward": 0.3359375,
      "rewards/format_reward": 0.91953125,
      "step": 65
    },
    {
      "completion_length": 20.51796875,
      "epoch": 0.5114155251141552,
      "grad_norm": 13.061442375183105,
      "kl": 6.19736328125,
      "learning_rate": 1.1283983551465512e-05,
      "loss": 0.2759,
      "reward": 1.31796875,
      "reward_std": 0.2789971936494112,
      "rewards/accuracy_reward": 0.39375,
      "rewards/format_reward": 0.92421875,
      "step": 70
    },
    {
      "completion_length": 43.41796875,
      "epoch": 0.547945205479452,
      "grad_norm": 441.85443115234375,
      "kl": 2.45595703125,
      "learning_rate": 1e-05,
      "loss": 0.1115,
      "reward": 1.0578125,
      "reward_std": 0.48872011750936506,
      "rewards/accuracy_reward": 0.278125,
      "rewards/format_reward": 0.7796875,
      "step": 75
    },
    {
      "completion_length": 46.90703125,
      "epoch": 0.5844748858447488,
      "grad_norm": 1.9799107313156128,
      "kl": 1.23125,
      "learning_rate": 8.71601644853449e-06,
      "loss": 0.0292,
      "reward": 0.984375,
      "reward_std": 0.4891943013295531,
      "rewards/accuracy_reward": 0.25703125,
      "rewards/format_reward": 0.72734375,
      "step": 80
    },
    {
      "completion_length": 13.75546875,
      "epoch": 0.6210045662100456,
      "grad_norm": 4.793896198272705,
      "kl": 1.34541015625,
      "learning_rate": 7.453288797587714e-06,
      "loss": 0.0482,
      "reward": 1.31484375,
      "reward_std": 0.14597927127033472,
      "rewards/accuracy_reward": 0.33671875,
      "rewards/format_reward": 0.978125,
      "step": 85
    },
    {
      "completion_length": 13.13359375,
      "epoch": 0.6575342465753424,
      "grad_norm": 12.646719932556152,
      "kl": 1.4255859375,
      "learning_rate": 6.232721063648148e-06,
      "loss": 0.0557,
      "reward": 1.34296875,
      "reward_std": 0.10623529590666295,
      "rewards/accuracy_reward": 0.3515625,
      "rewards/format_reward": 0.99140625,
      "step": 90
    },
    {
      "completion_length": 15.2046875,
      "epoch": 0.6940639269406392,
      "grad_norm": 2.1448068618774414,
      "kl": 1.299609375,
      "learning_rate": 5.074519320461358e-06,
      "loss": 0.0404,
      "reward": 1.3171875,
      "reward_std": 0.14243474025279285,
      "rewards/accuracy_reward": 0.33125,
      "rewards/format_reward": 0.9859375,
      "step": 95
    },
    {
      "completion_length": 22.12265625,
      "epoch": 0.730593607305936,
      "grad_norm": 13.79468822479248,
      "kl": 1.4185546875,
      "learning_rate": 3.997857194516319e-06,
      "loss": 0.0562,
      "reward": 1.3484375,
      "reward_std": 0.19392489716410638,
      "rewards/accuracy_reward": 0.3953125,
      "rewards/format_reward": 0.953125,
      "step": 100
    },
    {
      "epoch": 0.730593607305936,
      "eval_completion_length": 61.36258561643836,
      "eval_kl": 2.0426084474885844,
      "eval_loss": 0.08902593702077866,
      "eval_reward": 1.2051084474885845,
      "eval_reward_std": 0.4570505227672455,
      "eval_rewards/accuracy_reward": 0.39023972602739726,
      "eval_rewards/format_reward": 0.8148687214611872,
      "eval_runtime": 25621.0164,
      "eval_samples_per_second": 0.171,
      "eval_steps_per_second": 0.043,
      "step": 100
    },
    {
      "completion_length": 50.4015625,
      "epoch": 0.7671232876712328,
      "grad_norm": 3.6549317836761475,
      "kl": 1.721875,
      "learning_rate": 3.0205584523365626e-06,
      "loss": 0.035,
      "reward": 1.27421875,
      "reward_std": 0.3447163349017501,
      "rewards/accuracy_reward": 0.38671875,
      "rewards/format_reward": 0.8875,
      "step": 105
    },
    {
      "completion_length": 43.98359375,
      "epoch": 0.8036529680365296,
      "grad_norm": 2.842573404312134,
      "kl": 1.8482421875,
      "learning_rate": 2.158801934232897e-06,
      "loss": 0.0649,
      "reward": 1.240625,
      "reward_std": 0.2724386781454086,
      "rewards/accuracy_reward": 0.321875,
      "rewards/format_reward": 0.91875,
      "step": 110
    },
    {
      "completion_length": 50.07265625,
      "epoch": 0.8401826484018264,
      "grad_norm": 5.093225479125977,
      "kl": 2.44453125,
      "learning_rate": 1.426853719236676e-06,
      "loss": 0.1351,
      "reward": 1.1875,
      "reward_std": 0.3050973150879145,
      "rewards/accuracy_reward": 0.29296875,
      "rewards/format_reward": 0.89453125,
      "step": 115
    },
    {
      "completion_length": 66.58828125,
      "epoch": 0.8767123287671232,
      "grad_norm": 3.9608869552612305,
      "kl": 2.5376953125,
      "learning_rate": 8.368309551299536e-07,
      "loss": 0.1227,
      "reward": 1.29765625,
      "reward_std": 0.35280441734939816,
      "rewards/accuracy_reward": 0.40390625,
      "rewards/format_reward": 0.89375,
      "step": 120
    },
    {
      "completion_length": 86.17265625,
      "epoch": 0.91324200913242,
      "grad_norm": 4.511857509613037,
      "kl": 2.7826171875,
      "learning_rate": 3.985012632839824e-07,
      "loss": 0.1011,
      "reward": 1.24921875,
      "reward_std": 0.31032323855906724,
      "rewards/accuracy_reward": 0.33828125,
      "rewards/format_reward": 0.9109375,
      "step": 125
    },
    {
      "completion_length": 81.040625,
      "epoch": 0.9497716894977168,
      "grad_norm": 3.5829803943634033,
      "kl": 3.201953125,
      "learning_rate": 1.1912103908922945e-07,
      "loss": 0.1484,
      "reward": 1.275,
      "reward_std": 0.31014324594289067,
      "rewards/accuracy_reward": 0.35859375,
      "rewards/format_reward": 0.91640625,
      "step": 130
    },
    {
      "completion_length": 89.2578125,
      "epoch": 0.9863013698630136,
      "grad_norm": 6.557438373565674,
      "kl": 3.0515625,
      "learning_rate": 3.315324856869584e-09,
      "loss": 0.1109,
      "reward": 1.20390625,
      "reward_std": 0.32457808069884775,
      "rewards/accuracy_reward": 0.3,
      "rewards/format_reward": 0.90390625,
      "step": 135
    },
    {
      "completion_length": 95.453125,
      "epoch": 0.993607305936073,
      "kl": 3.2041015625,
      "reward": 1.2265625,
      "reward_std": 0.43310215324163437,
      "rewards/accuracy_reward": 0.3359375,
      "rewards/format_reward": 0.890625,
      "step": 136,
      "total_flos": 0.0,
      "train_loss": 0.0634942120351993,
      "train_runtime": 123112.1165,
      "train_samples_per_second": 0.036,
      "train_steps_per_second": 0.001
    }
  ],
  "logging_steps": 5,
  "max_steps": 136,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 15,
  "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": 1,
  "trial_name": null,
  "trial_params": null
}