{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 292,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 7.59609375,
      "epoch": 0.03432003432003432,
      "grad_norm": 406.9186326479053,
      "kl": 7.42265625,
      "learning_rate": 1.5e-07,
      "logits/chosen": -107538022.4,
      "logits/rejected": -105647308.8,
      "logps/chosen": -9728.8,
      "logps/rejected": -10080.8,
      "loss": 0.504168701171875,
      "num_tokens": 1000098.0,
      "rewards/chosen": -0.14722747802734376,
      "rewards/margins": -0.14902572631835936,
      "rewards/rejected": 0.001798248291015625,
      "step": 10
    },
    {
      "entropy": 7.5984375,
      "epoch": 0.06864006864006864,
      "grad_norm": 409.3229066351965,
      "kl": 4.01533203125,
      "learning_rate": 3.166666666666666e-07,
      "logits/chosen": -110265958.4,
      "logits/rejected": -109350092.8,
      "logps/chosen": -10545.2,
      "logps/rejected": -10896.0,
      "loss": 0.498638916015625,
      "num_tokens": 2059302.0,
      "rewards/chosen": -1.3995697021484375,
      "rewards/margins": 0.15903167724609374,
      "rewards/rejected": -1.5586013793945312,
      "step": 20
    },
    {
      "entropy": 7.57578125,
      "epoch": 0.10296010296010295,
      "grad_norm": 404.8670770902758,
      "kl": 0.562109375,
      "learning_rate": 4.833333333333333e-07,
      "logits/chosen": -107849318.4,
      "logits/rejected": -107046502.4,
      "logps/chosen": -10034.4,
      "logps/rejected": -10251.4,
      "loss": 0.498687744140625,
      "num_tokens": 3087752.0,
      "rewards/chosen": -2.3518798828125,
      "rewards/margins": 0.034576416015625,
      "rewards/rejected": -2.386456298828125,
      "step": 30
    },
    {
      "entropy": 7.551171875,
      "epoch": 0.13728013728013727,
      "grad_norm": 187.83701636824927,
      "kl": 1.7640625,
      "learning_rate": 4.985456442051682e-07,
      "logits/chosen": -105958604.8,
      "logits/rejected": -102563840.0,
      "logps/chosen": -9802.6,
      "logps/rejected": -10256.8,
      "loss": 0.51365966796875,
      "num_tokens": 4100561.0,
      "rewards/chosen": -2.747271728515625,
      "rewards/margins": -0.363616943359375,
      "rewards/rejected": -2.38365478515625,
      "step": 40
    },
    {
      "entropy": 7.560546875,
      "epoch": 0.1716001716001716,
      "grad_norm": 390.9188206397092,
      "kl": 0.0,
      "learning_rate": 4.935399618791793e-07,
      "logits/chosen": -104621670.4,
      "logits/rejected": -103025868.8,
      "logps/chosen": -9854.5,
      "logps/rejected": -10193.1,
      "loss": 0.49798583984375,
      "num_tokens": 5111368.0,
      "rewards/chosen": -5.695751953125,
      "rewards/margins": -0.0411865234375,
      "rewards/rejected": -5.6545654296875,
      "step": 50
    },
    {
      "entropy": 7.540625,
      "epoch": 0.2059202059202059,
      "grad_norm": 184.75414677190986,
      "kl": 0.0,
      "learning_rate": 4.850368660819092e-07,
      "logits/chosen": -104896921.6,
      "logits/rejected": -104657715.2,
      "logps/chosen": -10527.6,
      "logps/rejected": -10888.4,
      "loss": 0.5055633544921875,
      "num_tokens": 6159442.0,
      "rewards/chosen": -6.3327789306640625,
      "rewards/margins": 0.1228485107421875,
      "rewards/rejected": -6.45562744140625,
      "step": 60
    },
    {
      "entropy": 7.571484375,
      "epoch": 0.24024024024024024,
      "grad_norm": 169.7800809090956,
      "kl": 0.0,
      "learning_rate": 4.731584675403184e-07,
      "logits/chosen": -103940096.0,
      "logits/rejected": -100905779.2,
      "logps/chosen": -9513.0,
      "logps/rejected": -9829.6,
      "loss": 0.4996856689453125,
      "num_tokens": 7148746.0,
      "rewards/chosen": -9.118505859375,
      "rewards/margins": 0.374853515625,
      "rewards/rejected": -9.493359375,
      "step": 70
    },
    {
      "entropy": 7.480859375,
      "epoch": 0.27456027456027454,
      "grad_norm": 408.3393740518918,
      "kl": 0.0,
      "learning_rate": 4.5807534881817183e-07,
      "logits/chosen": -83179929.6,
      "logits/rejected": -82441011.2,
      "logps/chosen": -9965.6,
      "logps/rejected": -10375.6,
      "loss": 0.501361083984375,
      "num_tokens": 8139737.0,
      "rewards/chosen": -17.73671875,
      "rewards/margins": 0.77255859375,
      "rewards/rejected": -18.50927734375,
      "step": 80
    },
    {
      "entropy": 7.480859375,
      "epoch": 0.3088803088803089,
      "grad_norm": 195.0042260741351,
      "kl": 0.0,
      "learning_rate": 4.400041146246136e-07,
      "logits/chosen": -84013875.2,
      "logits/rejected": -81135206.4,
      "logps/chosen": -10246.0,
      "logps/rejected": -10755.6,
      "loss": 0.503289794921875,
      "num_tokens": 9171482.0,
      "rewards/chosen": -19.4990234375,
      "rewards/margins": 0.1666015625,
      "rewards/rejected": -19.665625,
      "step": 90
    },
    {
      "entropy": 7.59296875,
      "epoch": 0.3432003432003432,
      "grad_norm": 418.08759150457024,
      "kl": 14.2984375,
      "learning_rate": 4.1920428121079e-07,
      "logits/chosen": -95789056.0,
      "logits/rejected": -94894489.6,
      "logps/chosen": -9991.2,
      "logps/rejected": -10348.8,
      "loss": 0.5091827392578125,
      "num_tokens": 10204959.0,
      "rewards/chosen": -8.932275390625,
      "rewards/margins": -1.5104385375976563,
      "rewards/rejected": -7.421836853027344,
      "step": 100
    },
    {
      "entropy": 7.6796875,
      "epoch": 0.37752037752037754,
      "grad_norm": 304.7207533670506,
      "kl": 74.37421875,
      "learning_rate": 3.959745495250818e-07,
      "logits/chosen": -98575974.4,
      "logits/rejected": -96652492.8,
      "logps/chosen": -9406.8,
      "logps/rejected": -9838.0,
      "loss": 0.511468505859375,
      "num_tokens": 11188498.0,
      "rewards/chosen": 0.582354736328125,
      "rewards/margins": -1.033416748046875,
      "rewards/rejected": 1.615771484375,
      "step": 110
    },
    {
      "entropy": 7.675,
      "epoch": 0.4118404118404118,
      "grad_norm": 112.28682038073552,
      "kl": 81.974609375,
      "learning_rate": 3.7064851564718353e-07,
      "logits/chosen": -98258124.8,
      "logits/rejected": -97157120.0,
      "logps/chosen": -9821.6,
      "logps/rejected": -9996.0,
      "loss": 0.4924774169921875,
      "num_tokens": 12220511.0,
      "rewards/chosen": 1.986285400390625,
      "rewards/margins": 0.4836639404296875,
      "rewards/rejected": 1.5026214599609375,
      "step": 120
    },
    {
      "entropy": 7.737890625,
      "epoch": 0.44616044616044614,
      "grad_norm": 88.46830115313492,
      "kl": 169.4015625,
      "learning_rate": 3.4358988010236103e-07,
      "logits/chosen": -109287833.6,
      "logits/rejected": -109248512.0,
      "logps/chosen": -9719.6,
      "logps/rejected": -9905.6,
      "loss": 0.4920684814453125,
      "num_tokens": 13250325.0,
      "rewards/chosen": 7.7063720703125,
      "rewards/margins": -0.6712646484375,
      "rewards/rejected": 8.37763671875,
      "step": 130
    },
    {
      "entropy": 7.6921875,
      "epoch": 0.4804804804804805,
      "grad_norm": 130.00094058289008,
      "kl": 205.4296875,
      "learning_rate": 3.1518722485366754e-07,
      "logits/chosen": -112984064.0,
      "logits/rejected": -110991769.6,
      "logps/chosen": -10208.4,
      "logps/rejected": -10518.8,
      "loss": 0.4965087890625,
      "num_tokens": 14326309.0,
      "rewards/chosen": 8.865472412109375,
      "rewards/margins": -1.158734130859375,
      "rewards/rejected": 10.02420654296875,
      "step": 140
    },
    {
      "entropy": 7.8125,
      "epoch": 0.5148005148005148,
      "grad_norm": 113.36547850330345,
      "kl": 338.4109375,
      "learning_rate": 2.8584843297836277e-07,
      "logits/chosen": -114737152.0,
      "logits/rejected": -115002572.8,
      "logps/chosen": -9504.2,
      "logps/rejected": -9669.2,
      "loss": 0.501788330078125,
      "num_tokens": 15346427.0,
      "rewards/chosen": 16.750439453125,
      "rewards/margins": 0.296142578125,
      "rewards/rejected": 16.454296875,
      "step": 150
    },
    {
      "entropy": 7.848828125,
      "epoch": 0.5491205491205491,
      "grad_norm": 121.67889562055869,
      "kl": 537.35,
      "learning_rate": 2.5599483116609544e-07,
      "logits/chosen": -128991232.0,
      "logits/rejected": -126107648.0,
      "logps/chosen": -9370.0,
      "logps/rejected": -9774.6,
      "loss": 0.497845458984375,
      "num_tokens": 16366642.0,
      "rewards/chosen": 26.253515625,
      "rewards/margins": -1.16259765625,
      "rewards/rejected": 27.41611328125,
      "step": 160
    },
    {
      "entropy": 7.9390625,
      "epoch": 0.5834405834405835,
      "grad_norm": 143.6403128818565,
      "kl": 801.4125,
      "learning_rate": 2.2605513915689874e-07,
      "logits/chosen": -146440192.0,
      "logits/rejected": -141308723.2,
      "logps/chosen": -9034.6,
      "logps/rejected": -9302.8,
      "loss": 0.5146453857421875,
      "num_tokens": 17380939.0,
      "rewards/chosen": 39.715625,
      "rewards/margins": -0.73203125,
      "rewards/rejected": 40.44765625,
      "step": 170
    },
    {
      "entropy": 7.980078125,
      "epoch": 0.6177606177606177,
      "grad_norm": 42.14785908826522,
      "kl": 1004.125,
      "learning_rate": 1.9645931300952795e-07,
      "logits/chosen": -160088064.0,
      "logits/rejected": -157659955.2,
      "logps/chosen": -8898.6,
      "logps/rejected": -9383.6,
      "loss": 0.516741943359375,
      "num_tokens": 18421034.0,
      "rewards/chosen": 48.4921875,
      "rewards/margins": -3.44765625,
      "rewards/rejected": 51.93984375,
      "step": 180
    },
    {
      "entropy": 8.00078125,
      "epoch": 0.6520806520806521,
      "grad_norm": 104.10846117229305,
      "kl": 1096.575,
      "learning_rate": 1.6763237061535008e-07,
      "logits/chosen": -159468748.8,
      "logits/rejected": -158800281.6,
      "logps/chosen": -8481.4,
      "logps/rejected": -8729.6,
      "loss": 0.508721923828125,
      "num_tokens": 19430090.0,
      "rewards/chosen": 52.7296875,
      "rewards/margins": -4.18125,
      "rewards/rejected": 56.9109375,
      "step": 190
    },
    {
      "entropy": 7.983203125,
      "epoch": 0.6864006864006864,
      "grad_norm": 89.80201439869198,
      "kl": 1178.15,
      "learning_rate": 1.3998828812795932e-07,
      "logits/chosen": -164646092.8,
      "logits/rejected": -162535833.6,
      "logps/chosen": -8834.4,
      "logps/rejected": -9109.2,
      "loss": 0.50364990234375,
      "num_tokens": 20459910.0,
      "rewards/chosen": 57.29375,
      "rewards/margins": -3.2859375,
      "rewards/rejected": 60.5796875,
      "step": 200
    },
    {
      "entropy": 8.003515625,
      "epoch": 0.7207207207207207,
      "grad_norm": 126.30100335052259,
      "kl": 1175.75,
      "learning_rate": 1.1392405496029076e-07,
      "logits/chosen": -165543936.0,
      "logits/rejected": -164678860.8,
      "logps/chosen": -9119.3,
      "logps/rejected": -9405.0,
      "loss": 0.5057098388671875,
      "num_tokens": 21493036.0,
      "rewards/chosen": 56.555859375,
      "rewards/margins": -4.447265625,
      "rewards/rejected": 61.003125,
      "step": 210
    },
    {
      "entropy": 8.0046875,
      "epoch": 0.7550407550407551,
      "grad_norm": 87.58008225171027,
      "kl": 1176.325,
      "learning_rate": 8.981397272385738e-08,
      "logits/chosen": -164039884.8,
      "logits/rejected": -162191769.6,
      "logps/chosen": -8580.2,
      "logps/rejected": -9027.8,
      "loss": 0.5058837890625,
      "num_tokens": 22497542.0,
      "rewards/chosen": 57.2875,
      "rewards/margins": -3.028125,
      "rewards/rejected": 60.315625,
      "step": 220
    },
    {
      "entropy": 8.004296875,
      "epoch": 0.7893607893607893,
      "grad_norm": 175.38563131735916,
      "kl": 1219.65,
      "learning_rate": 6.800427998154968e-08,
      "logits/chosen": -171258675.2,
      "logits/rejected": -168919040.0,
      "logps/chosen": -8726.4,
      "logps/rejected": -9050.0,
      "loss": 0.4997528076171875,
      "num_tokens": 23534012.0,
      "rewards/chosen": 59.96875,
      "rewards/margins": -2.05234375,
      "rewards/rejected": 62.02109375,
      "step": 230
    },
    {
      "entropy": 8.005859375,
      "epoch": 0.8236808236808236,
      "grad_norm": 179.66444536530696,
      "kl": 1322.8,
      "learning_rate": 4.8808180006509356e-08,
      "logits/chosen": -182104883.2,
      "logits/rejected": -178900172.8,
      "logps/chosen": -9388.4,
      "logps/rejected": -9730.8,
      "loss": 0.503857421875,
      "num_tokens": 24628737.0,
      "rewards/chosen": 64.2625,
      "rewards/margins": -3.753125,
      "rewards/rejected": 68.015625,
      "step": 240
    },
    {
      "entropy": 8.018359375,
      "epoch": 0.858000858000858,
      "grad_norm": 41.96169751613372,
      "kl": 1284.6,
      "learning_rate": 3.250134295213053e-08,
      "logits/chosen": -171111219.2,
      "logits/rejected": -169240166.4,
      "logps/chosen": -9261.2,
      "logps/rejected": -9670.2,
      "loss": 0.50284423828125,
      "num_tokens": 25690351.0,
      "rewards/chosen": 62.85625,
      "rewards/margins": -2.76875,
      "rewards/rejected": 65.625,
      "step": 250
    },
    {
      "entropy": 7.95703125,
      "epoch": 0.8923208923208923,
      "grad_norm": 74.44782326863705,
      "kl": 1305.4,
      "learning_rate": 1.9317947025340232e-08,
      "logits/chosen": -177713971.2,
      "logits/rejected": -177976115.2,
      "logps/chosen": -9279.6,
      "logps/rejected": -9792.2,
      "loss": 0.5191864013671875,
      "num_tokens": 26763232.0,
      "rewards/chosen": 63.3265625,
      "rewards/margins": -3.925,
      "rewards/rejected": 67.2515625,
      "step": 260
    },
    {
      "entropy": 7.959375,
      "epoch": 0.9266409266409267,
      "grad_norm": 78.92089799828031,
      "kl": 1155.175,
      "learning_rate": 9.447315514833353e-09,
      "logits/chosen": -167346176.0,
      "logits/rejected": -163191193.6,
      "logps/chosen": -8660.6,
      "logps/rejected": -9027.4,
      "loss": 0.4955169677734375,
      "num_tokens": 27765238.0,
      "rewards/chosen": 56.3484375,
      "rewards/margins": -2.79375,
      "rewards/rejected": 59.1421875,
      "step": 270
    },
    {
      "entropy": 7.9859375,
      "epoch": 0.960960960960961,
      "grad_norm": 90.537992127442,
      "kl": 1152.825,
      "learning_rate": 3.0311979690147426e-09,
      "logits/chosen": -157574758.4,
      "logits/rejected": -156021555.2,
      "logps/chosen": -8495.6,
      "logps/rejected": -8591.4,
      "loss": 0.500689697265625,
      "num_tokens": 28731039.0,
      "rewards/chosen": 56.9328125,
      "rewards/margins": -1.3921875,
      "rewards/rejected": 58.325,
      "step": 280
    },
    {
      "entropy": 7.99765625,
      "epoch": 0.9952809952809952,
      "grad_norm": 128.53404483419018,
      "kl": 1228.95,
      "learning_rate": 1.6173456793908135e-10,
      "logits/chosen": -168358707.2,
      "logits/rejected": -166900531.2,
      "logps/chosen": -9022.6,
      "logps/rejected": -9187.5,
      "loss": 0.5062225341796875,
      "num_tokens": 29763064.0,
      "rewards/chosen": 60.1203125,
      "rewards/margins": -2.6171875,
      "rewards/rejected": 62.7375,
      "step": 290
    },
    {
      "epoch": 1.0,
      "eval_entropy": 8.009457236842104,
      "eval_kl": 1195.4473684210527,
      "eval_logits/chosen": -173201300.21052632,
      "eval_logits/rejected": -171587045.0526316,
      "eval_logps/chosen": -8729.052631578947,
      "eval_logps/rejected": -9016.21052631579,
      "eval_loss": 0.5066904425621033,
      "eval_num_tokens": 29885865.0,
      "eval_rewards/chosen": 58.98766447368421,
      "eval_rewards/margins": -1.5361842105263157,
      "eval_rewards/rejected": 60.52384868421053,
      "eval_runtime": 43.5216,
      "eval_samples_per_second": 27.756,
      "eval_steps_per_second": 1.746,
      "step": 292
    }
  ],
  "logging_steps": 10,
  "max_steps": 292,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7245703888896.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}