{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.798330122029544,
  "eval_steps": 500,
  "global_step": 350,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05,
      "grad_norm": 6.844920635223389,
      "learning_rate": 5e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.0625,
      "logps/chosen": -258.0,
      "logps/rejected": -322.0,
      "loss": 1.0697,
      "max_memory_allocated (GB)": 94.25,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.30078125,
      "rewards/chosen": 0.67578125,
      "rewards/margins": -0.4140625,
      "rewards/rejected": 1.0859375,
      "step": 10,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.300484657287598,
      "learning_rate": 0.0001,
      "logits/chosen": -3.046875,
      "logits/rejected": -3.03125,
      "logps/chosen": -286.0,
      "logps/rejected": -336.0,
      "loss": 0.5736,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.69140625,
      "rewards/chosen": -0.150390625,
      "rewards/margins": 0.8046875,
      "rewards/rejected": -0.95703125,
      "step": 20,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.9487425088882446,
      "learning_rate": 0.00015,
      "logits/chosen": -2.859375,
      "logits/rejected": -2.84375,
      "logps/chosen": -294.0,
      "logps/rejected": -394.0,
      "loss": 0.2394,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": -1.984375,
      "rewards/margins": 4.8125,
      "rewards/rejected": -6.8125,
      "step": 30,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0650818347930908,
      "learning_rate": 0.0002,
      "logits/chosen": -2.875,
      "logits/rejected": -2.828125,
      "logps/chosen": -284.0,
      "logps/rejected": -388.0,
      "loss": 0.1181,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.95703125,
      "rewards/chosen": -0.62109375,
      "rewards/margins": 5.0625,
      "rewards/rejected": -5.6875,
      "step": 40,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5032941699028015,
      "learning_rate": 0.00025,
      "logits/chosen": -2.890625,
      "logits/rejected": -2.875,
      "logps/chosen": -278.0,
      "logps/rejected": -412.0,
      "loss": 0.0583,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.984375,
      "rewards/chosen": -0.8828125,
      "rewards/margins": 6.96875,
      "rewards/rejected": -7.84375,
      "step": 50,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.8771920800209045,
      "learning_rate": 0.0003,
      "logits/chosen": -2.828125,
      "logits/rejected": -2.828125,
      "logps/chosen": -312.0,
      "logps/rejected": -474.0,
      "loss": 0.0469,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.984375,
      "rewards/chosen": -2.609375,
      "rewards/margins": 11.5,
      "rewards/rejected": -14.125,
      "step": 60,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.42379140853881836,
      "learning_rate": 0.00035,
      "logits/chosen": -2.84375,
      "logits/rejected": -2.84375,
      "logps/chosen": -280.0,
      "logps/rejected": -436.0,
      "loss": 0.0285,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.9921875,
      "rewards/chosen": -0.9609375,
      "rewards/margins": 9.9375,
      "rewards/rejected": -10.9375,
      "step": 70,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5195761919021606,
      "learning_rate": 0.0004,
      "logits/chosen": -2.78125,
      "logits/rejected": -2.765625,
      "logps/chosen": -292.0,
      "logps/rejected": -450.0,
      "loss": 0.0266,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.98828125,
      "rewards/chosen": -1.4453125,
      "rewards/margins": 10.75,
      "rewards/rejected": -12.1875,
      "step": 80,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.770358085632324,
      "learning_rate": 0.00045000000000000004,
      "logits/chosen": -2.921875,
      "logits/rejected": -2.9375,
      "logps/chosen": -276.0,
      "logps/rejected": -478.0,
      "loss": 0.0486,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.984375,
      "rewards/chosen": -1.5703125,
      "rewards/margins": 13.375,
      "rewards/rejected": -15.0,
      "step": 90,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1556715965270996,
      "learning_rate": 0.0005,
      "logits/chosen": -2.75,
      "logits/rejected": -2.71875,
      "logps/chosen": -284.0,
      "logps/rejected": -482.0,
      "loss": 0.053,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.9765625,
      "rewards/chosen": -1.3125,
      "rewards/margins": 13.0,
      "rewards/rejected": -14.375,
      "step": 100,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.40476056933403015,
      "learning_rate": 0.0004998477067547739,
      "logits/chosen": -2.875,
      "logits/rejected": -2.875,
      "logps/chosen": -296.0,
      "logps/rejected": -444.0,
      "loss": 0.0384,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.98828125,
      "rewards/chosen": -1.2109375,
      "rewards/margins": 10.5,
      "rewards/rejected": -11.75,
      "step": 110,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8060222864151001,
      "learning_rate": 0.0004993910125649561,
      "logits/chosen": -3.234375,
      "logits/rejected": -3.234375,
      "logps/chosen": -282.0,
      "logps/rejected": -456.0,
      "loss": 0.0183,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.9921875,
      "rewards/chosen": -0.76171875,
      "rewards/margins": 12.4375,
      "rewards/rejected": -13.1875,
      "step": 120,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.5584326982498169,
      "learning_rate": 0.0004986304738420684,
      "logits/chosen": -3.140625,
      "logits/rejected": -3.125,
      "logps/chosen": -296.0,
      "logps/rejected": -496.0,
      "loss": 0.0294,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.984375,
      "rewards/chosen": -1.875,
      "rewards/margins": 14.8125,
      "rewards/rejected": -16.75,
      "step": 130,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.2188811302185059,
      "learning_rate": 0.0004975670171853926,
      "logits/chosen": -3.21875,
      "logits/rejected": -3.21875,
      "logps/chosen": -300.0,
      "logps/rejected": -478.0,
      "loss": 0.0312,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.98828125,
      "rewards/chosen": -1.3515625,
      "rewards/margins": 13.0,
      "rewards/rejected": -14.375,
      "step": 140,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.9386507272720337,
      "learning_rate": 0.000496201938253052,
      "logits/chosen": -3.1875,
      "logits/rejected": -3.140625,
      "logps/chosen": -288.0,
      "logps/rejected": -488.0,
      "loss": 0.0504,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.984375,
      "rewards/chosen": -1.6171875,
      "rewards/margins": 13.25,
      "rewards/rejected": -14.8125,
      "step": 150,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.11973602324724197,
      "learning_rate": 0.0004945369001834514,
      "logits/chosen": -2.9375,
      "logits/rejected": -2.84375,
      "logps/chosen": -298.0,
      "logps/rejected": -486.0,
      "loss": 0.0395,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.984375,
      "rewards/chosen": -1.9296875,
      "rewards/margins": 14.5625,
      "rewards/rejected": -16.5,
      "step": 160,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.0885164737701416,
      "learning_rate": 0.0004925739315689991,
      "logits/chosen": -2.828125,
      "logits/rejected": -2.78125,
      "logps/chosen": -292.0,
      "logps/rejected": -490.0,
      "loss": 0.0511,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.984375,
      "rewards/chosen": -1.890625,
      "rewards/margins": 14.25,
      "rewards/rejected": -16.125,
      "step": 170,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.92,
      "grad_norm": 5.371945381164551,
      "learning_rate": 0.0004903154239845797,
      "logits/chosen": -2.71875,
      "logits/rejected": -2.671875,
      "logps/chosen": -304.0,
      "logps/rejected": -474.0,
      "loss": 0.028,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.98828125,
      "rewards/chosen": -2.03125,
      "rewards/margins": 12.75,
      "rewards/rejected": -14.75,
      "step": 180,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.7487589120864868,
      "learning_rate": 0.0004877641290737884,
      "logits/chosen": -2.90625,
      "logits/rejected": -2.875,
      "logps/chosen": -288.0,
      "logps/rejected": -484.0,
      "loss": 0.053,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.98046875,
      "rewards/chosen": -1.71875,
      "rewards/margins": 13.3125,
      "rewards/rejected": -15.0,
      "step": 190,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.1586950570344925,
      "learning_rate": 0.0004849231551964771,
      "logits/chosen": -3.0,
      "logits/rejected": -2.96875,
      "logps/chosen": -296.0,
      "logps/rejected": -516.0,
      "loss": 0.0173,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.9921875,
      "rewards/chosen": -2.375,
      "rewards/margins": 16.75,
      "rewards/rejected": -19.125,
      "step": 200,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.385784387588501,
      "learning_rate": 0.00048179596364169685,
      "logits/chosen": -3.015625,
      "logits/rejected": -3.0,
      "logps/chosen": -312.0,
      "logps/rejected": -528.0,
      "loss": 0.0372,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.98828125,
      "rewards/chosen": -2.28125,
      "rewards/margins": 17.25,
      "rewards/rejected": -19.625,
      "step": 210,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.10692762583494186,
      "learning_rate": 0.0004783863644106502,
      "logits/chosen": -3.125,
      "logits/rejected": -3.09375,
      "logps/chosen": -320.0,
      "logps/rejected": -572.0,
      "loss": 0.012,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.9921875,
      "rewards/chosen": -3.65625,
      "rewards/margins": 20.25,
      "rewards/rejected": -23.875,
      "step": 220,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.407893419265747,
      "learning_rate": 0.00047469851157479177,
      "logits/chosen": -3.171875,
      "logits/rejected": -3.109375,
      "logps/chosen": -308.0,
      "logps/rejected": -568.0,
      "loss": 0.008,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.99609375,
      "rewards/chosen": -3.234375,
      "rewards/margins": 20.625,
      "rewards/rejected": -23.875,
      "step": 230,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.15119591355323792,
      "learning_rate": 0.00047073689821473173,
      "logits/chosen": -3.203125,
      "logits/rejected": -3.1875,
      "logps/chosen": -290.0,
      "logps/rejected": -588.0,
      "loss": 0.0236,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.99609375,
      "rewards/chosen": -3.328125,
      "rewards/margins": 22.125,
      "rewards/rejected": -25.375,
      "step": 240,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.0850694179534912,
      "learning_rate": 0.00046650635094610973,
      "logits/chosen": -3.25,
      "logits/rejected": -3.265625,
      "logps/chosen": -286.0,
      "logps/rejected": -484.0,
      "loss": 0.0246,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.98828125,
      "rewards/chosen": -1.921875,
      "rewards/margins": 13.75,
      "rewards/rejected": -15.6875,
      "step": 250,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.0764814615249634,
      "learning_rate": 0.00046201202403910646,
      "logits/chosen": -3.234375,
      "logits/rejected": -3.21875,
      "logps/chosen": -298.0,
      "logps/rejected": -498.0,
      "loss": 0.0307,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.9921875,
      "rewards/chosen": -2.421875,
      "rewards/margins": 15.5,
      "rewards/rejected": -17.875,
      "step": 260,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.04254325106739998,
      "learning_rate": 0.00045725939313876043,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.0625,
      "logps/chosen": -304.0,
      "logps/rejected": -548.0,
      "loss": 0.0161,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.9921875,
      "rewards/chosen": -2.703125,
      "rewards/margins": 18.75,
      "rewards/rejected": -21.5,
      "step": 270,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.2640644311904907,
      "learning_rate": 0.0004522542485937369,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.078125,
      "logps/chosen": -294.0,
      "logps/rejected": -552.0,
      "loss": 0.0116,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.99609375,
      "rewards/chosen": -2.203125,
      "rewards/margins": 19.5,
      "rewards/rejected": -21.75,
      "step": 280,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.6378334164619446,
      "learning_rate": 0.00044700268840168044,
      "logits/chosen": -3.046875,
      "logits/rejected": -3.046875,
      "logps/chosen": -296.0,
      "logps/rejected": -600.0,
      "loss": 0.0159,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.53125,
      "rewards/margins": 24.0,
      "rewards/rejected": -27.625,
      "step": 290,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.9890326261520386,
      "learning_rate": 0.0004415111107797445,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.0625,
      "logps/chosen": -320.0,
      "logps/rejected": -588.0,
      "loss": 0.0083,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.640625,
      "rewards/margins": 22.375,
      "rewards/rejected": -26.0,
      "step": 300,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.59,
      "grad_norm": 2.6053104400634766,
      "learning_rate": 0.00043578620636934855,
      "logits/chosen": -3.109375,
      "logits/rejected": -3.09375,
      "logps/chosen": -338.0,
      "logps/rejected": -628.0,
      "loss": 0.0038,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -5.0,
      "rewards/margins": 24.75,
      "rewards/rejected": -29.75,
      "step": 310,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.0414825677871704,
      "learning_rate": 0.0004298349500846628,
      "logits/chosen": -3.25,
      "logits/rejected": -3.21875,
      "logps/chosen": -302.0,
      "logps/rejected": -540.0,
      "loss": 0.0174,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.9921875,
      "rewards/chosen": -3.015625,
      "rewards/margins": 18.25,
      "rewards/rejected": -21.25,
      "step": 320,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.3290166854858398,
      "learning_rate": 0.00042366459261474935,
      "logits/chosen": -3.171875,
      "logits/rejected": -3.140625,
      "logps/chosen": -334.0,
      "logps/rejected": -568.0,
      "loss": 0.0209,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.98828125,
      "rewards/chosen": -4.1875,
      "rewards/margins": 19.875,
      "rewards/rejected": -24.125,
      "step": 330,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.75,
      "grad_norm": 3.274353504180908,
      "learning_rate": 0.0004172826515897146,
      "logits/chosen": -3.203125,
      "logits/rejected": -3.171875,
      "logps/chosen": -316.0,
      "logps/rejected": -604.0,
      "loss": 0.0095,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -2.96875,
      "rewards/margins": 23.625,
      "rewards/rejected": -26.625,
      "step": 340,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.1814924478530884,
      "learning_rate": 0.0004106969024216348,
      "logits/chosen": -3.125,
      "logits/rejected": -3.09375,
      "logps/chosen": -294.0,
      "logps/rejected": -612.0,
      "loss": 0.0129,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.99609375,
      "rewards/chosen": -3.0625,
      "rewards/margins": 25.625,
      "rewards/rejected": -28.625,
      "step": 350,
      "total_memory_available (GB)": 94.62
    }
  ],
  "logging_steps": 10,
  "max_steps": 1000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 10,
  "total_flos": 2.939722980655104e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}